diff options
author | Slávek Banko <slavek.banko@axis.cz> | 2021-11-05 13:28:23 +0100 |
---|---|---|
committer | Slávek Banko <slavek.banko@axis.cz> | 2021-11-05 13:28:23 +0100 |
commit | 8c787c3591c1c885b91a54128835b400858c5cca (patch) | |
tree | eca1b776912a305c4d45b3964038278a2fae1ead /debian/htdig/htdig-3.2.0b6/htlib | |
parent | fe188b907cdf30dfdfe0eba9412e7f8749fec158 (diff) | |
download | extra-dependencies-8c787c35.tar.gz extra-dependencies-8c787c35.zip |
DEB htdig: Added to repository.
Signed-off-by: Slávek Banko <slavek.banko@axis.cz>
Diffstat (limited to 'debian/htdig/htdig-3.2.0b6/htlib')
98 files changed, 30190 insertions, 0 deletions
diff --git a/debian/htdig/htdig-3.2.0b6/htlib/.cvsignore b/debian/htdig/htdig-3.2.0b6/htlib/.cvsignore new file mode 100644 index 00000000..09dc8ef2 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/.cvsignore @@ -0,0 +1,7 @@ +Makefile +*.lo +*.la +.purify +.pure +.deps +.libs diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Configuration.cc b/debian/htdig/htdig-3.2.0b6/htlib/Configuration.cc new file mode 100644 index 00000000..56912736 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Configuration.cc @@ -0,0 +1,390 @@ +// +// Configuration.cc +// +// Configuration: This class provides an object lookup table. Each object +// in the Configuration is indexed with a string. The objects +// can be returned by mentioning their string index. Values may +// include files with `/path/to/file` or other configuration +// variables with ${variable} +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Configuration.cc,v 1.20 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include <stdio.h> +#include "Configuration.h" +#include "htString.h" +#include "ParsedString.h" + +#include <stdlib.h> +#include <ctype.h> +#include <locale.h> + + +//********************************************************************* +// Configuration::Configuration() +// +Configuration::Configuration() : separators("=:"), allow_multiple(0) +{ +} + + +//********************************************************************* +// void Configuration::NameValueSeparators(char *s) +// +void Configuration::NameValueSeparators(const String& s) +{ + separators = s; +} + + +//********************************************************************* +// Add an entry to the configuration table. +// +void Configuration::Add(const String& str_arg) +{ + const char* str = str_arg; + String name, value; + + while (str && *str) + { + while (isspace(*str)) + str++; + name = 0; + if (!isalpha(*str)) + break; + // Some isalnum() implementations don't allow all the letters that + // isalpha() does, e.g. accented ones. They're not POSIX.2 compliant + // but we won't punish them with an infinite loop... + if (!isalnum(*str)) + break; + while (isalnum(*str) || *str == '-' || *str == '_') + name << *str++; + + name.lowercase(); + + // + // We have the name. Let's see if we will get a value + // + while (isspace(*str)) + str++; + if (!*str) + { + // + // End of string. We need to store the name as a boolean TRUE + // + Add(name, "true"); + return; + } + + if (!strchr((char*)separators, *str)) + { + // + // We are now at a new name. The previous one needs to be set + // to boolean TRUE + // + Add(name, "true"); + continue; + } + + // + // We now need to deal with the value + // + str++; // Skip the separator + while (isspace(*str)) + str++; + if (!*str) + { + // + // End of string reached. The value must be blank + // + Add(name, ""); + break; + } + value = 0; + if (*str == '"') + { + // + // Ah! A quoted value. This should be easy to deal with... + // (Just kidding!) + // + str++; + while (*str && *str != '"') + { + value << *str++; + } + Add(name, value); + if (*str == '"') + str++; + continue; + } + else if (*str == '\'') + { + // A single quoted value. + str++; + while (*str && *str != '\'') + { + value << *str++; + } + Add(name, value); + if (*str == '\'') + str++; + continue; + } + else + { + // + // A non-quoted string. This string will terminate at the + // next blank + // + while (*str && !isspace(*str)) + { + value << *str++; + } + Add(name, value); + continue; + } + } +} + + +//********************************************************************* +// Add an entry to the configuration table, without allowing variable +// or file expansion of the value. +// +void Configuration::Add(const String& name, const String& value) +{ + String escaped; + const char *s = value.get(); + while (*s) + { + if (strchr("$`\\", *s)) + escaped << '\\'; + escaped << *s++; + } + ParsedString *ps = new ParsedString(escaped); + dcGlobalVars.Add(name, ps); +} + + +//********************************************************************* +// Add an entry to the configuration table, allowing parsing for variable +// or file expansion of the value. +// +void Configuration::AddParsed(const String& name, const String& value) +{ + ParsedString *ps = new ParsedString(value); + if (mystrcasecmp(name, "locale") == 0) + { + String str(setlocale(LC_ALL, ps->get(dcGlobalVars))); + ps->set(str); + + // + // Set time format to standard to avoid sending If-Modified-Since + // http headers in native format which http servers can't + // understand + // + setlocale(LC_TIME, "C"); + } + dcGlobalVars.Add(name, ps); +} + + +//********************************************************************* +// Remove an entry from both the hash table and from the list of keys. +// +int Configuration::Remove(const String& name) +{ + return dcGlobalVars.Remove(name); +} + + +//********************************************************************* +// char *Configuration::Find(const char *name) const +// Retrieve a variable from the configuration database. This variable +// will be parsed and a new String object will be returned. +// +const String Configuration::Find(const String& name) const +{ + ParsedString *ps = (ParsedString *) dcGlobalVars[name]; + if (ps) + { + return ps->get(dcGlobalVars); + } + else + { +#ifdef DEBUG + fprintf (stderr, "Could not find configuration option %s\n", (const char*)name); +#endif + return 0; + } +} + +//- +// Return 1 if the value of configuration attribute <b>name</b> has +// been set, 0 otherwise +int Configuration::Exists(const String& name) const +{ + return dcGlobalVars.Exists(name); +} + +//********************************************************************* +Object *Configuration::Get_Object(char *name) { +return dcGlobalVars[name]; +} + + +//********************************************************************* +// +int Configuration::Value(const String& name, int default_value) const +{ + return Find(name).as_integer(default_value); +} + + +//********************************************************************* +// +double Configuration::Double(const String& name, double default_value) const +{ + return Find(name).as_double(default_value); +} + + +//********************************************************************* +// int Configuration::Boolean(char *name, int default_value) +// +int Configuration::Boolean(const String& name, int default_value) const +{ + int value = default_value; + const String s = Find(name); + if (s[0]) + { + if (s.nocase_compare("true") == 0 || + s.nocase_compare("yes") == 0 || + s.nocase_compare("1") == 0) + value = 1; + else if (s.nocase_compare("false") == 0 || + s.nocase_compare("no") == 0 || + s.nocase_compare("0") == 0) + value = 0; + } + + return value; +} + + +//********************************************************************* +// +const String Configuration::operator[](const String& name) const +{ + return Find(name); +} + + +//********************************************************************* +// +int Configuration::Read(const String& filename) +{ + FILE* in = fopen((const char*)filename, "r"); + + if(!in) { + fprintf(stderr, "Configuration::Read: cannot open %s for reading : ", (const char*)filename); + perror(""); + return NOTOK; + } + +#define CONFIG_BUFFER_SIZE (50*1024) + // + // Make the line buffer large so that we can read long lists of start + // URLs. + // + char buffer[CONFIG_BUFFER_SIZE + 1]; + char *current; + String line; + String name; + char *value; + int len; + while (fgets(buffer, CONFIG_BUFFER_SIZE, in)) + { + line << buffer; + line.chop("\r\n"); + if (line.last() == '\\') + { + line.chop(1); + continue; // Append the next line to this one + } + + current = line.get(); + if (*current == '#' || *current == '\0') + { + line = 0; + continue; // Comments and blank lines are skipped + } + + name = strtok(current, ": =\t"); + value = strtok(0, "\r\n"); + if (!value) + value = ""; // Blank value + + // + // Skip any whitespace before the actual text + // + while (*value == ' ' || *value == '\t') + value++; + len = strlen(value) - 1; + // + // Skip any whitespace after the actual text + // + while (len >= 0 && (value[len] == ' ' || value[len] == '\t')) + { + value[len] = '\0'; + len--; + } + + if (mystrcasecmp((char*)name, "include") == 0) + { + ParsedString ps(value); + String str(ps.get(dcGlobalVars)); + if (str[0] != '/') // Given file name not fully qualified + { + str = filename; // so strip dir. name from current one + len = str.lastIndexOf('/') + 1; + if (len > 0) + str.chop(str.length() - len); + else + str = ""; // No slash in current filename + str << ps.get(dcGlobalVars); + } + Read(str); + line = 0; + continue; + } + + AddParsed(name, value); + line = 0; + } + fclose(in); + return OK; +} + + +//********************************************************************* +// void Configuration::Defaults(ConfigDefaults *array) +// +void Configuration::Defaults(const ConfigDefaults *array) +{ + for (int i = 0; array[i].name; i++) + { + AddParsed(array[i].name, array[i].value); + } +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Configuration.h b/debian/htdig/htdig-3.2.0b6/htlib/Configuration.h new file mode 100644 index 00000000..2628617d --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Configuration.h @@ -0,0 +1,229 @@ +// +// Configuration.h +// +// NAME +// +// reads the configuration file and manages it in memory. +// +// SYNOPSIS +// +// #include <Configuration.h> +// +// Configuration config; +// +// ConfigDefault config_defaults = { +// { "verbose", "true" }, +// { 0, 0 } +// }; +// +// config.Defaults(config_defaults); +// +// config.Read("~/.myconfig") ; +// +// config.Add("sync", "false"); +// +// if(config["sync"]) ... +// if(config.Value("rate") < 50) ... +// if(config.Boolean("sync")) ... +// +// DESCRIPTION +// +// The primary purpose of the <b>Configuration</b> class is to parse +// a configuration file and allow the application to modify the internal +// data structure. All values are strings and are converted by the +// appropriate accessors. For instance the <b>Boolean</b> method will +// return numerical true (not zero) if the string either contains +// a number that is different from zero or the string <i>true</i>. +// +// The <i>ConfigDefaults</i> type is a structure of two char pointers: +// the name of the configuration attribute and it's value. The end of +// the array is the first entry that contains a null pointer instead of +// the attribute name. Numerical +// values must be in strings. For instance: +// <pre> +// ConfigDefault* config_defaults = { +// { "wordlist_compress", "true" }, +// { "wordlist_page_size", "8192" }, +// { 0, 0 } +// }; +// </pre> +// Returns the configuration (object of type <i>Configuration</i>) +// built if a file was found or config_defaults +// provided, 0 otherwise. +// The additional +// fields of the <b>ConfigDefault</b> are purely informative. +// +// FILE FORMAT +// +// This configuration file is a plain ASCII text file. Each line in +// the file is either a comment or contains an attribute. +// Comment lines are blank lines or lines that start with a '#'. +// Attributes consist of a variable name and an associated +// value: +// +// <pre> +// <name>:<whitespace><value><newline> +// </pre> +// +// The <name> contains any alphanumeric character or +// underline (_) The <value> can include any character +// except newline. It also cannot start with spaces or tabs since +// those are considered part of the whitespace after the colon. It +// is important to keep in mind that any trailing spaces or tabs +// will be included. +// +// It is possible to split the <value> across several +// lines of the configuration file by ending each line with a +// backslash (\). The effect on the value is that a space is +// added where the line split occurs. +// +// A configuration file can include another file, by using the special +// <name>, <tt>include</tt>. The <value> is taken as +// the file name of another configuration file to be read in at +// this point. If the given file name is not fully qualified, it is +// taken relative to the directory in which the current configuration +// file is found. Variable expansion is permitted in the file name. +// Multiple include statements, and nested includes are also permitted. +// +// <pre> +// include: common.conf +// </pre> +// +// +// END +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Configuration.h,v 1.11 2004/05/28 13:15:20 lha Exp $ +// + +#ifndef _Configuration_h_ +#define _Configuration_h_ + +#include "Dictionary.h" +#include "htString.h" + +struct ConfigDefaults +{ + char *name; // Name of the attribute + char *value; // Default value + char *type; // Type of the value (string, integer, boolean) + char *programs; // Whitespace separated list of programs/modules using this attribute + char *block; // Configuration block this can be used in (can be blank) + char *version; // Version that introduced the attribute + char *category; // Attribute category (to split documentation) + char *example; // Example usage of the attribute (HTML) + char *description; // Long description of the attribute (HTML) +}; + + +class Configuration : public Object +{ +public: + //- + // Constructor + // + Configuration(); +#ifndef SWIG + Configuration(const Configuration& config) : + dcGlobalVars(config.dcGlobalVars), + separators(config.separators) + { + allow_multiple = config.allow_multiple; + } +#endif /* SWIG */ + //- + // Destructor + // + ~Configuration() {} + + // + // Adding and deleting items to and from the Configuration + // +#ifndef SWIG + //- + // Add configuration item <b>str</b> to the configuration. The value + // associated with it is undefined. + // + void Add(const String& str); +#endif /* SWIG */ + //- + // Add configuration item <b>name</b> to the configuration and associate + // it with <b>value</b>. + // + void Add(const String& name, const String& value); + void AddParsed(const String& name, const String& value); + //- + // Remove the <b>name</b> from the configuration. + // + int Remove(const String& name); + + //- + // Let the Configuration know how to parse name value pairs. + // Each character of string <b>s</b> is a valid separator between + // the <i>name</i> and the <i>value.</i> + // + void NameValueSeparators(const String& s); + + //- + // Read name/value configuration pairs from the file <b>filename</b>. + // + virtual int Read(const String& filename); + + //- + // Return the value of configuration attribute <b>name</b> as a + // <i>String</i>. + // + const String Find(const String& name) const; + + //- + // Return 1 if the value of configuration attribute <b>name</b> has + // been set, 0 otherwise + int Exists(const String& name) const; + +#ifndef SWIG + //- + // Alias to the <b>Find</b> method. + // + const String operator[](const String& name) const; +#endif /* SWIG */ + //- + // Return the value associated with the configuration attribute + // <b>name</b>, converted to integer using the atoi(3) function. + // If the attribute is not found in the configuration and + // a <b>default_value</b> is provided, return it. + // + int Value(const String& name, int default_value = 0) const; + //- + // Return the value associated with the configuration attribute + // <b>name</b>, converted to double using the atof(3) function. + // If the attribute is not found in the configuration and + // a <b>default_value</b> is provided, return it. + // + double Double(const String& name, double default_value = 0) const; + //- + // Return 1 if the value associated to <b>name</b> is + // either <b>1, yes</b> or <b>true</b>. + // Return 0 if the value associated to <b>name</b> is + // either <b>0, no</b> or <b>false</b>. + // + int Boolean(const String& name, int default_value = 0) const; + Object *Get_Object(char *name); + + //- + // Load configuration attributes from the <i>name</i> and <i>value</i> + // members of the <b>array</b> argument. + // + void Defaults(const ConfigDefaults *array); + +protected: + Dictionary dcGlobalVars; + String separators; + int allow_multiple; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/DB2_db.cc b/debian/htdig/htdig-3.2.0b6/htlib/DB2_db.cc new file mode 100644 index 00000000..46127056 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/DB2_db.cc @@ -0,0 +1,379 @@ +// +// DB2_db.cc +// +// DB2_db: Implements the Berkeley B-Tree database as a Database object +// (including duplicate values to allow duplicate word entries) +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: DB2_db.cc,v 1.26 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include <fcntl.h> +#include <errno.h> +#include <stdlib.h> + +#ifdef HAVE_STD +#include <fstream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <fstream.h> +#endif /* HAVE_STD */ + +#ifndef _MSC_VER /* _WIN32 */ +#include <unistd.h> +#endif + +#include "DB2_db.h" +#include "HtConfiguration.h" + +// Default cache size in kilobytes. +// Maybe this should be an config option, just for easy testing and +// determination for best system performance +// NOTE: page size is 1KB - do not change!! +#define CACHE_SIZE_IN_KB 64 + +//***************************************************************************** +// DB2_db::DB2_db() +// +DB2_db::DB2_db() +{ + isOpen = 0; + _compare = 0; + _prefix = 0; +} + + +//***************************************************************************** +// DB2_db::~DB2_db() +// +DB2_db::~DB2_db() +{ + Close(); +} + + +//***************************************************************************** +// +int +DB2_db::Open(const char *filename, int flags, int mode) +{ + // + // Initialize the database environment. + // + if((dbenv = db_init((char *)NULL)) == 0) return NOTOK; + + if(CDB_db_create(&dbp, dbenv, 0) != 0) return NOTOK; + + if(_compare) dbp->set_bt_compare(dbp, _compare); + if(_prefix) dbp->set_bt_prefix(dbp, _prefix); + + // + // Open the database. + // + if((errno = dbp->open(dbp, filename, NULL, db_type, flags, mode)) == 0) + { + // + // Acquire a cursor for the database. + // + if ((seqrc = dbp->cursor(dbp, NULL, &dbcp, 0)) != 0) + { + seqerr = seqrc; + Close(); + return NOTOK; + } + isOpen = 1; + return OK; + } + else + { + return NOTOK; + } +} + + +//***************************************************************************** +// int DB2_db::Close() +// +int +DB2_db::Close() +{ + if(isOpen) + { + // + // Close cursor, database and clean up environment + // + (void)(dbcp->c_close)(dbcp); + (void)(dbp->close)(dbp, 0); + (void)(dbenv->close(dbenv, 0)); + dbenv = 0; + } + isOpen = 0; + return OK; +} + + +//***************************************************************************** +// char *DB2_db::Get_Next(String &item, String &key) +// +char * +DB2_db::Get_Next(String &item, String &key) +{ + if (isOpen && !seqrc) + { + // + // Return values + // + key = skey; + lkey = skey; + item = data; + + // + // Search for the next record + // + DBT local_key; + DBT local_data; + + memset(&local_key, 0, sizeof(DBT)); + memset(&local_data, 0, sizeof(DBT)); + + local_key.data = skey.get(); + local_key.size = skey.length(); + + seqrc = dbcp->c_get(dbcp, &local_key, &local_data, DB_NEXT); + seqerr = seqrc; + + if(!seqrc) { + data = 0; + data.append((char*)local_data.data, (int)local_data.size); + skey = 0; + skey.append((char*)local_key.data, (int)local_key.size); + } + + return lkey.get(); + } + else + return 0; +} + +//***************************************************************************** +// void DB2_db::Start_Seq() +// +void +DB2_db::Start_Seq(const String& key) +{ + DBT local_key; + DBT local_data; + + memset(&local_key, 0, sizeof(DBT)); + memset(&local_data, 0, sizeof(DBT)); + + skey = key; + + local_key.data = skey.get(); + local_key.size = skey.length(); + + if (isOpen && dbp) + { + // + // Okay, get the first key. Use DB_SET_RANGE for finding partial + // keys also. If you set it to DB_SET, and the words book, books + // and bookstore do exists, it will find them if you specify + // book*. However if you specify boo* if will not find + // anything. Setting to DB_SET_RANGE will still find the `first' + // word after boo* (which is book). + // + seqrc = dbcp->c_get(dbcp, &local_key, &local_data, DB_SET_RANGE); + seqerr = seqrc; + + if(!seqrc) { + data = 0; + data.append((char*)local_data.data, (int)local_data.size); + skey = 0; + skey.append((char*)local_key.data, (int)local_key.size); + } + } +} + +//***************************************************************************** +// void DB2_db::Start_Get() +// +void +DB2_db::Start_Get() +{ + DBT local_key; + DBT local_data; + + memset(&local_key, 0, sizeof(DBT)); + memset(&local_data, 0, sizeof(DBT)); + + if (isOpen && dbp) + { + // + // Okay, get the first key. Use DB_SET_RANGE for finding partial + // keys also. If you set it to DB_SET, and the words book, books + // and bookstore do exists, it will find them if you specify + // book*. However if you specify boo* if will not find + // anything. Setting to DB_SET_RANGE will still find the `first' + // word after boo* (which is book). + // + seqrc = dbcp->c_get(dbcp, &local_key, &local_data, DB_FIRST); + seqerr = seqrc; + + if(!seqrc) { + data = 0; + data.append((char*)local_data.data, (int)local_data.size); + skey = 0; + skey.append((char*)local_key.data, (int)local_key.size); + } + } +} + +//***************************************************************************** +// int DB2_db::Put(const String &key, const String &data) +// +int +DB2_db::Put(const String &key, const String &data) +{ + DBT k, d; + + memset(&k, 0, sizeof(DBT)); + memset(&d, 0, sizeof(DBT)); + + if (!isOpen) + return NOTOK; + + k.data = (char*)key.get(); + k.size = key.length(); + + d.data = (char*)data.get(); + d.size = data.length(); + + // + // A 0 in the flags in put means replace, if you didn't specify DB_DUP + // somewhere else... + // + return (dbp->put)(dbp, NULL, &k, &d, 0) == 0 ? OK : NOTOK; +} + + +//***************************************************************************** +// int DB2_db::Get(const String &key, String &data) +// +int +DB2_db::Get(const String &key, String &data) +{ + DBT k, d; + + memset(&k, 0, sizeof(DBT)); + memset(&d, 0, sizeof(DBT)); + + // + // k arg of get should be const but is not. Harmless cast. + // + k.data = (char*)key.get(); + k.size = key.length(); + + int rc = dbp->get(dbp, NULL, &k, &d, 0); + if (rc) + return NOTOK; + + data = 0; + data.append((char *)d.data, d.size); + return OK; +} + + +//***************************************************************************** +// int DB2_db::Exists(const String &key) +// +int +DB2_db::Exists(const String &key) +{ + String data; + + if (!isOpen) + return 0; + + return Get(key, data); +} + + +//***************************************************************************** +// int DB2_db::Delete(const String &key) +// +int +DB2_db::Delete(const String &key) +{ + DBT k; + + memset(&k, 0, sizeof(DBT)); + + if (!isOpen) + return 0; + + k.data = (char*)key.get(); + k.size = key.length(); + + return (dbp->del)(dbp, NULL, &k, 0); +} + + +//***************************************************************************** +// DB2_db *DB2_db::getDatabaseInstance() +// +DB2_db * +DB2_db::getDatabaseInstance(DBTYPE) +{ + return new DB2_db(); +} + +//***************************************************************************** +// void Error(const char *error_prefix, char *message); +// +void Error(const char *error_prefix, char *message) +{ + // We don't do anything here, it's mostly a stub so we can set a breakpoint + // for debugging purposes + fprintf(stderr, "%s: %s\n", error_prefix, message); +} + +//****************************************************************************** + +/* + * db_init -- + * Initialize the environment. Only returns a pointer + */ +DB_ENV * +DB2_db::db_init(char *home) +{ + DB_ENV *dbenv; + char *progname = "DB2 problem..."; + + int error; + if((error = CDB_db_env_create(&dbenv, 0)) != 0) { + fprintf(stderr, "DB2_db: CDB_db_env_create %s\n", CDB_db_strerror(error)); + return 0; + } + + dbenv->set_errpfx(dbenv, progname); + dbenv->set_errcall(dbenv, &Error); + + if((error = dbenv->open(dbenv, (const char*)home, NULL, DB_CREATE | DB_PRIVATE | DB_INIT_LOCK | DB_INIT_MPOOL, 0666)) != 0) { + dbenv->err(dbenv, error, "open %s", (home ? home : "")); + return 0; + } + + return (dbenv); +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/DB2_db.h b/debian/htdig/htdig-3.2.0b6/htlib/DB2_db.h new file mode 100644 index 00000000..48a6e837 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/DB2_db.h @@ -0,0 +1,53 @@ +// +// DB2_db.h +// +// DB2_db: Implements the Berkeley B-Tree database as a Database object +// (including duplicate values to allow duplicate word entries) +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: DB2_db.h,v 1.11 2004/05/28 13:15:20 lha Exp $ +// + +#ifndef _DB2_db_h_ +#define _DB2_db_h_ + +#include "Database.h" +#include <db.h> +#include <fcntl.h> + +class DB2_db : public Database +{ + // + // Construction/Destruction + // +protected: + DB2_db(); +public: + ~DB2_db(); + + static DB2_db *getDatabaseInstance(DBTYPE type); + + virtual int OpenReadWrite(const char *filename, int mode) { return Open(filename, DB_CREATE, mode); } + virtual int OpenRead(const char *filename) { return Open(filename, DB_RDONLY, 0666); } + virtual int Close(); + virtual int Get(const String &, String &); + virtual int Put(const String &, const String &); + virtual int Exists(const String &); + virtual int Delete(const String &); + + virtual void Start_Get(); + virtual char *Get_Next(String &item, String &key); + virtual void Start_Seq(const String& key); + +private: + DB_ENV *db_init(char *); + + int Open(const char *filename, int flags, int mode); +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Database.cc b/debian/htdig/htdig-3.2.0b6/htlib/Database.cc new file mode 100644 index 00000000..968c6630 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Database.cc @@ -0,0 +1,53 @@ +// +// Database.cc +// +// Database: Class which defines the interface to a generic, +// simple database. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Database.cc,v 1.12 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "Database.h" +#include "DB2_db.h" + +//***************************************************************************** +// Database::Database() +// +Database::Database() +{ +} + + +//***************************************************************************** +// Database::~Database() +// +Database::~Database() +{ +} + + +//***************************************************************************** +// Database *Database::getDatabaseInstance() +// +Database * +Database::getDatabaseInstance(DBTYPE type = DB_BTREE) +{ + Database* db = DB2_db::getDatabaseInstance(type); + + db->db_type = type; + + return db; +} + + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Database.h b/debian/htdig/htdig-3.2.0b6/htlib/Database.h new file mode 100644 index 00000000..552de594 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Database.h @@ -0,0 +1,91 @@ +// +// Database.h +// +// Database: Class which defines the interface to a generic, +// simple database. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Database.h,v 1.15 2004/05/28 13:15:20 lha Exp $ +// + +#ifndef _Database_h_ +#define _Database_h_ + +#include "Object.h" +#include "htString.h" + +#include <db.h> + +// Database Types +// defined in db.h +// #define DB_BTREE 1 +// #define DB_HASH 2 +#ifndef GDBM_HASH +#define GDBM_HASH 2 +#endif + + +class Database : public Object +{ + // + // Make sure no one can actually create an object of this type or + // the derived types. The static getDatabaseInstance() method needs to be + // used. + // +protected: + Database(); +public: + ~Database(); + + // + // Since the contructor is protected, the only way to actually get + // a Database object is through the following class method. + // The idea here is that the particular type of database used by + // all the programs is to be defined in one place. + // + static Database *getDatabaseInstance(DBTYPE type); + + // + // Common interface + // + virtual int OpenReadWrite(const char *filename, int mode = 0666) = 0; + virtual int OpenRead(const char *filename) = 0; + void SetCompare(int (*func)(const DBT *a, const DBT *b)) { _compare = func; } + void SetPrefix(size_t (*func)(const DBT *a, const DBT *b)) { _prefix = func; } + virtual int Close() = 0; + virtual int Put(const String &key, const String &data) = 0; + virtual int Get(const String &key, String &data) = 0; + virtual int Exists(const String &key) = 0; + virtual int Delete(const String &key) = 0; + + virtual void Start_Get() = 0; + virtual char *Get_Next() { String item; String key; return Get_Next(item, key); } + virtual char *Get_Next(String &item) { String key; return Get_Next(item, key); } + virtual char *Get_Next(String &item, String &key) = 0; + virtual void Start_Seq(const String& str) = 0; + virtual char *Get_Next_Seq() { return Get_Next(); } + +protected: + int isOpen; + DB *dbp; // database + DBC *dbcp; // cursor + + String skey; // Next key to search for iterator + String data; // Next data to return for iterator + String lkey; // Contains the last key returned by iterator + + DB_ENV *dbenv; // database enviroment + int (*_compare)(const DBT *a, const DBT *b); // Key comparison + size_t (*_prefix)(const DBT *a, const DBT *b); // Key reduction + + int seqrc; + int seqerr; + DBTYPE db_type; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Dictionary.cc b/debian/htdig/htdig-3.2.0b6/htlib/Dictionary.cc new file mode 100644 index 00000000..8c0da019 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Dictionary.cc @@ -0,0 +1,416 @@ +// +// Dictionary.cc +// +// Dictionary: This class provides an object lookup table. +// Each object in the dictionary is indexed with a string. +// The objects can be returned by mentioning their +// string index. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1995-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Dictionary.cc,v 1.16 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "Dictionary.h" + +#include <stdlib.h> + +class DictionaryEntry +{ +public: + unsigned int hash; + char *key; + Object *value; + DictionaryEntry *next; + + ~DictionaryEntry(); + void release(); +}; + +DictionaryEntry::~DictionaryEntry() +{ + free(key); + delete value; +} + +void +DictionaryEntry::release() +{ + value = NULL; // Prevent the value from being deleted +} + + +//********************************************************************* +// +Dictionary::Dictionary() +{ + init(101, 10.0f); +} + +Dictionary::Dictionary(int initialCapacity, float loadFactor) +{ + init(initialCapacity, loadFactor); +} + +Dictionary::Dictionary(int initialCapacity) +{ + init(initialCapacity, 0.75f); +} + +Dictionary::Dictionary(const Dictionary& other) +{ + init(other.initialCapacity, other.loadFactor); + + DictionaryCursor cursor; + const char* key; + for(other.Start_Get(cursor); (key = other.Get_Next(cursor));) { + Add(key, other[key]); + } +} + +//********************************************************************* +// +Dictionary::~Dictionary() +{ + Destroy(); + + delete [] table; +} + + +//********************************************************************* +// +void +Dictionary::Destroy() +{ + DictionaryEntry *t, *n; + + for (int i = 0; i < tableLength; i++) + { + if (table[i] != NULL) + { + t = table[i]; + do { // clear out hash chain + n = t->next; + delete t; + t = n; + } while (n); + table[i] = NULL; + } + } + count = 0; +} + +//********************************************************************* +// +void +Dictionary::Release() +{ + DictionaryEntry *t, *n; + + for (int i = 0; i < tableLength; i++) + { + if (table[i] != NULL) + { + t = table[i]; + do { // clear out hash chain + n = t->next; + t->release(); + delete t; + t = n; + } while (n); + table[i] = NULL; + } + } + count = 0; +} + + +//********************************************************************* +// +void +Dictionary::init(int initialCapacity, float loadFactor) +{ + if (initialCapacity <= 0) + initialCapacity = 101; + if (loadFactor <= 0.0) + loadFactor = 0.75f; + Dictionary::loadFactor = loadFactor; + table = new DictionaryEntry*[initialCapacity]; + for (int i = 0; i < initialCapacity; i++) + { + table[i] = NULL; + } + threshold = (int)(initialCapacity * loadFactor); + tableLength = initialCapacity; + count = 0; +} + +//********************************************************************* +// +unsigned int +Dictionary::hashCode(const char *key) const +{ + char *test; + long conv_key = strtol(key, &test, 10); + if (key && *key && !*test) // Conversion succeeded + return conv_key; + + char *base = (char*)malloc(strlen(key) + 2); + char *tmp_key = base; + strcpy(tmp_key, key); + + unsigned int h = 0; + int length = strlen(tmp_key); + + if (length >= 16) + { + tmp_key += strlen(tmp_key) - 15; + length = strlen(tmp_key); + } + for (int i = length; i > 0; i--) + { + h = (h*37) + *tmp_key++; + } + + free(base); + return h; +} + +//********************************************************************* +// Add an entry to the hash table. This will replace the +// data associated with an already existing key. +// +void +Dictionary::Add(const String& name, Object *obj) +{ + unsigned int hash = hashCode(name); + int index = hash % tableLength; + DictionaryEntry *e; + + for (e = table[index]; e != NULL; e = e->next) + { + if (e->hash == hash && strcmp(e->key, name) == 0) + { + delete e->value; + e->value = obj; + return; + } + } + + if (count >= threshold) + { + rehash(); + Add(name, obj); + return; + } + + e = new DictionaryEntry(); + e->hash = hash; + e->key = strdup(name); + e->value = obj; + e->next = table[index]; + table[index] = e; + count++; +} + + +//********************************************************************* +// Remove an entry from the hash table. +// +int +Dictionary::Remove(const String& name) +{ + if (!count) + return 0; + + unsigned int hash = hashCode(name); + int index = hash % tableLength; + DictionaryEntry *e, *prev; + + for (e = table[index], prev = NULL; e != NULL; prev = e, e = e->next) + { + if (hash == e->hash && strcmp(e->key, name) == 0) + { + if (prev != NULL) + { + prev->next = e->next; + } + else + { + table[index] = e->next; + } + count--; + delete e; + return 1; + } + } + return 0; +} + + +//********************************************************************* +// +Object *Dictionary::Find(const String& name) const +{ + if (!count) + return NULL; + + unsigned int hash = hashCode(name); + int index = hash % tableLength; + DictionaryEntry *e; + + for (e = table[index]; e != NULL; e = e->next) + { + if (e->hash == hash && strcmp(e->key, name) == 0) + { + return e->value; + } + } + return NULL; +} + + +//********************************************************************* +// +Object *Dictionary::operator[](const String& name) const +{ + return Find(name); +} + + +//********************************************************************* +// +int Dictionary::Exists(const String& name) const +{ + if (!count) + return 0; + + unsigned int hash = hashCode(name); + int index = hash % tableLength; + DictionaryEntry *e; + + for (e = table[index]; e != NULL; e = e->next) + { + if (e->hash == hash && strcmp(e->key, name) == 0) + { + return 1; + } + } + return 0; +} + + +//********************************************************************* +// +void +Dictionary::rehash() +{ + DictionaryEntry **oldTable = table; + int oldCapacity = tableLength; + + int newCapacity; + DictionaryEntry *e; + int i, index; + + newCapacity = count > oldCapacity ? count * 2 + 1 : oldCapacity * 2 + 1; + + DictionaryEntry **newTable = new DictionaryEntry*[newCapacity]; + + for (i = 0; i < newCapacity; i++) + { + newTable[i] = NULL; + } + + threshold = (int) (newCapacity * loadFactor); + table = newTable; + tableLength = newCapacity; + + for (i = oldCapacity; i-- > 0;) + { + for (DictionaryEntry *old = oldTable[i]; old != NULL;) + { + e = old; + old = old->next; + index = e->hash % newCapacity; + e->next = newTable[index]; + newTable[index] = e; + } + } + delete [] oldTable; +} + + +//********************************************************************* +// +void +Dictionary::Start_Get(DictionaryCursor& cursor) const +{ + cursor.currentTableIndex = -1; + cursor.currentDictionaryEntry = NULL; +} + + +//********************************************************************* +// +char * +Dictionary::Get_Next(DictionaryCursor& cursor) const +{ + while (cursor.currentDictionaryEntry == NULL || + cursor.currentDictionaryEntry->next == NULL) + { + cursor.currentTableIndex++; + + if (cursor.currentTableIndex >= tableLength) + { + cursor.currentTableIndex--; + return NULL; + } + + cursor.currentDictionaryEntry = table[cursor.currentTableIndex]; + + if (cursor.currentDictionaryEntry != NULL) + { + return cursor.currentDictionaryEntry->key; + } + } + + cursor.currentDictionaryEntry = cursor.currentDictionaryEntry->next; + return cursor.currentDictionaryEntry->key; +} + +//********************************************************************* +// +Object * +Dictionary::Get_NextElement(DictionaryCursor& cursor) const +{ + while (cursor.currentDictionaryEntry == NULL || + cursor.currentDictionaryEntry->next == NULL) + { + cursor.currentTableIndex++; + + if (cursor.currentTableIndex >= tableLength) + { + cursor.currentTableIndex--; + return NULL; + } + + cursor.currentDictionaryEntry = table[cursor.currentTableIndex]; + + if (cursor.currentDictionaryEntry != NULL) + { + return cursor.currentDictionaryEntry->value; + } + } + + cursor.currentDictionaryEntry = cursor.currentDictionaryEntry->next; + return cursor.currentDictionaryEntry->value; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Dictionary.h b/debian/htdig/htdig-3.2.0b6/htlib/Dictionary.h new file mode 100644 index 00000000..8ff6cf1c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Dictionary.h @@ -0,0 +1,99 @@ +// +// Dictionary.h +// +// Dictionary: This class provides an object lookup table. +// Each object in the dictionary is indexed with a string. +// The objects can be returned by mentioning their +// string index. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Dictionary.h,v 1.10 2004/05/28 13:15:20 lha Exp $ +// + +#ifndef _Dictionary_h_ +#define _Dictionary_h_ + +#include "Object.h" +#include "htString.h" +#include "List.h" + +class Dictionary; +class DictionaryEntry; + +class DictionaryCursor { + public: + // + // Support for the Start_Get and Get_Next routines + // + int currentTableIndex; + DictionaryEntry *currentDictionaryEntry; +}; + +class Dictionary : public Object +{ +public: + // + // Construction/Destruction + // + Dictionary(); + Dictionary(const Dictionary& other); + Dictionary(int initialCapacity); + Dictionary(int initialCapacity, float loadFactor); + ~Dictionary(); + + // + // Adding and deleting items to and from the dictionary + // + void Add(const String& name, Object *obj); + int Remove(const String& name); + + // + // Searching can be done with the Find() member of the array indexing + // operator + // + Object *Find(const String& name) const; + Object *operator[](const String& name) const; + int Exists(const String& name) const; + + // + // We want to be able to go through all the entries in the + // dictionary in sequence. To do this, we have the same + // traversal interface as the List class + // + void Start_Get() { Start_Get(cursor); } + void Start_Get(DictionaryCursor& cursor) const; + // + // Get the next key + // + char *Get_Next() { return Get_Next(cursor); } + char *Get_Next(DictionaryCursor& cursor) const; + // + // Get the next entry + // + Object *Get_NextElement() { return Get_NextElement(cursor); } + Object *Get_NextElement(DictionaryCursor& cursor) const; + void Release(); + void Destroy(); + int Count() const { return count; } + +private: + DictionaryEntry **table; + int tableLength; + int initialCapacity; + int count; + int threshold; + float loadFactor; + + DictionaryCursor cursor; + + void rehash(); + void init(int, float); + unsigned int hashCode(const char *key) const; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtCodec.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtCodec.cc new file mode 100644 index 00000000..69b2682c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtCodec.cc @@ -0,0 +1,30 @@ +// +// HtCodec.cc +// +// HtCodec: Provide a generic means to take a String, code +// it, and return the encoded string. And vice versa. +// +// Keep constructor and destructor in a file of its own. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtCodec.cc,v 1.6 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtCodec.h" + +HtCodec::HtCodec() +{ } + +HtCodec::~HtCodec() +{ } + +// End of HtCodec.cc diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtCodec.h b/debian/htdig/htdig-3.2.0b6/htlib/HtCodec.h new file mode 100644 index 00000000..b3c56bf6 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtCodec.h @@ -0,0 +1,37 @@ +// +// HtCodec.h +// +// HtCodec: Provide a generic means to take a String, code +// it, and return the encoded string. And vice versa. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtCodec.h,v 1.6 2004/05/28 13:15:20 lha Exp $ +// +#ifndef __HtCodec_h +#define __HtCodec_h + +#include "htString.h" + +class HtCodec : public Object +{ +public: + HtCodec(); + virtual ~HtCodec(); + + // Code what's in this string. + virtual String encode(const String &) const = 0; + + // Decode what's in this string. + virtual String decode(const String &) const = 0; + +private: + HtCodec(const HtCodec &); // Not supposed to be implemented. + void operator= (const HtCodec &); // Not supposed to be implemented. +}; + +#endif /* __HtCodec_h */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtDateTime.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtDateTime.cc new file mode 100644 index 00000000..3ab40e47 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtDateTime.cc @@ -0,0 +1,1419 @@ +// +// HtDateTime.cc +// +// HtDateTime: Parse, split, compare and format dates and times. +// Uses locale. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtDateTime.cc,v 1.20 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtDateTime.h" + +#include <stdlib.h> +#include <string.h> +#include <ctype.h> + +#ifdef HAVE_STD +#include <iostream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <iostream.h> +#endif /* HAVE_STD */ + +#ifndef HAVE_STRPTIME +// mystrptime() declared in lib.h, defined in htlib/strptime.cc +#define strptime(s,f,t) mystrptime(s,f,t) +#else /* HAVE_STRPTIME */ +#ifndef HAVE_STRPTIME_DECL +extern "C" { +extern char *strptime(const char *__s, const char *__fmt, struct tm *__tp); +} +#endif /* HAVE_STRPTIME_DECL */ +#endif /* HAVE_STRPTIME */ + +/////// + // Static local variable : Visible only here !!! +/////// + +#define MAXSTRTIME 256 // Max length of my_strtime + +static struct tm Ht_tm; +static char my_strtime[MAXSTRTIME]; + + +/////// + // Recognized Date Formats +/////// + +// RFC1123: Sun, 06 Nov 1994 08:49:37 GMT +#define RFC1123_FORMAT "%a, %d %b %Y %H:%M:%S %Z" +#define LOOSE_RFC1123_FORMAT "%d %b %Y %H:%M:%S" + +// RFC850 : Sunday, 06-Nov-94 08:49:37 GMT +#define RFC850_FORMAT "%A, %d-%b-%y %H:%M:%S %Z" +#define LOOSE_RFC850_FORMAT "%d-%b-%y %H:%M:%S" + +// ANSI C's asctime() format : Sun Nov 6 08:49:37 1994 +#define ASCTIME_FORMAT "%a %b %e %H:%M:%S %Y" +#define LOOSE_ASCTIME_FORMAT "%b %e %H:%M:%S %Y" + +// ISO8601 : 1994-11-06 08:49:37 GMT +#define ISO8601_FORMAT "%Y-%m-%d %H:%M:%S %Z" + +// ISO8601 (short version): 1994-11-06 +#define ISO8601_SHORT_FORMAT "%Y-%m-%d" + +// Timestamp : 19941106084937 +#define TIMESTAMP_FORMAT "%Y%m%d%H%M%S" + + + +/////// + // Initialization +/////// + +const int HtDateTime::days[] = { 31, 28, 31, 30, 31, 30, + 31, 31, 30, 31, 30, 31}; + + +/////// // Input Formats // /////// + +/////// + // Generalized date/time parser for "LOOSE" formats + // - converts LOOSE RFC850 or RFC1123 date string into a time value + // - converts SHORT ISO8601 date string into a time value + // - autodetects which of these formats is used + // - assumes midnight if time portion omitted + // We've had problems using strptime() and timegm() on a few platforms + // while parsing these formats, so this is an attempt to sidestep them. + // + // Returns 0 if parsing failed, or returns number of characters parsed + // in date string otherwise, and sets Ht_t field to time_t value. +/////// +#define EPOCH 1970 + +int HtDateTime::Parse(const char *date) +{ + register const char *s; + register const char *t; + int day, month, year, hour, minute, second; + + // + // Three possible time designations: + // Tuesday, 01-Jul-97 16:48:02 GMT (RFC850) + // or + // Thu, 01 May 1997 00:40:42 GMT (RFC1123) + // or + // 1997-05-01 00:40:42 GMT (ISO8601) + // + // We strip off the weekday because we don't need it, and + // because some servers send invalid weekdays! + // (Some don't even send a weekday, but we'll be flexible...) + + s = date; + while (*s && *s != ',') + s++; + if (*s) + s++; + else + s = date; + while (isspace(*s)) + s++; + + // check for ISO8601 format + month = 0; + t = s; + while (isdigit(*t)) + t++; + if (t > s && *t == '-' && isdigit(t[1])) + day = -1; + else { + // not ISO8601, so try RFC850 or RFC1123 + // get day... + if (!isdigit(*s)) + return 0; + day = 0; + while (isdigit(*s)) + day = day * 10 + (*s++ - '0'); + if (day > 31) + return 0; + while (*s == '-' || isspace(*s)) + s++; + + // get month... + // (it's ugly, but it works) + switch (*s++) { + case 'J': case 'j': + switch (*s++) { + case 'A': case 'a': + month = 1; + s++; + break; + case 'U': case 'u': + switch (*s++) { + case 'N': case 'n': + month = 6; + break; + case 'L': case 'l': + month = 7; + break; + default: + return 0; + } + break; + default: + return 0; + } + break; + case 'F': case 'f': + month = 2; + s += 2; + break; + case 'M': case 'm': + switch (*s++) { + case 'A': case 'a': + switch (*s++) { + case 'R': case 'r': + month = 3; + break; + case 'Y': case 'y': + month = 5; + break; + default: + return 0; + } + break; + default: + return 0; + } + break; + case 'A': case 'a': + switch (*s++) { + case 'P': case 'p': + month = 4; + s++; + break; + case 'U': case 'u': + month = 8; + s++; + break; + default: + return 0; + } + break; + case 'S': case 's': + month = 9; + s += 2; + break; + case 'O': case 'o': + month = 10; + s += 2; + break; + case 'N': case 'n': + month = 11; + s += 2; + break; + case 'D': case 'd': + month = 12; + s += 2; + break; + default: + return 0; + } + while (*s == '-' || isspace(*s)) + s++; + } + + // get year... + if (!isdigit(*s)) + return 0; + year = 0; + while (isdigit(*s)) + year = year * 10 + (*s++ - '0'); + if (year < 69) + year += 2000; + else if (year < 1900) + year += 1900; + else if (year >= 19100) // seen some programs do it, why not check? + year -= (19100-2000); + while (*s == '-' || isspace(*s)) + s++; + + if (day < 0) { // still don't have day, so it's ISO8601 format + // get month... + if (!isdigit(*s)) + return 0; + month = 0; + while (isdigit(*s)) + month = month * 10 + (*s++ - '0'); + if (month < 1 || month > 12) + return 0; + while (*s == '-' || isspace(*s)) + s++; + + // get day... + if (!isdigit(*s)) + return 0; + day = 0; + while (isdigit(*s)) + day = day * 10 + (*s++ - '0'); + if (day < 1 || day > 31) + return 0; + while (*s == '-' || isspace(*s)) + s++; + } + + // optionally get hour... + hour = 0; + while (isdigit(*s)) + hour = hour * 10 + (*s++ - '0'); + if (hour > 23) + return 0; + while (*s == ':' || isspace(*s)) + s++; + + // optionally get minute... + minute = 0; + while (isdigit(*s)) + minute = minute * 10 + (*s++ - '0'); + if (minute > 59) + return 0; + while (*s == ':' || isspace(*s)) + s++; + + // optionally get second... + second = 0; + while (isdigit(*s)) + second = second * 10 + (*s++ - '0'); + if (second > 59) + return 0; + while (*s == ':' || isspace(*s)) + s++; + + // Assign the new value to time_t field + // + // Calculate date as seconds since 01 Jan 1970 00:00:00 GMT + // This is based somewhat on the date calculation code in NetBSD's + // cd9660_node.c code, for which I was unable to find a reference. + // It works, though! + // + Ht_t = (time_t) (((((367L*year - 7L*(year+(month+9)/12)/4 + - 3L*(((year)+((month)+9)/12-1)/100+1)/4 + + 275L*(month)/9 + day) - + (367L*EPOCH - 7L*(EPOCH+(1+9)/12)/4 + - 3L*((EPOCH+(1+9)/12-1)/100+1)/4 + + 275L*1/9 + 1)) + * 24 + hour) * 60 + minute) * 60 + second); + + // cerr << "Date string '" << date << "' converted to time_t " + // << (int)Ht_t << ", used " << (s-date) << " characters\n"; + + return s-date; +} + +/////// + // Personalized format such as C strftime function + // Overloaded version 1 + // It ignores, for now, Time Zone values +/////// + +char *HtDateTime::SetFTime(const char *buf, const char *format) +{ + + register char *p; + register int r; + + ToGMTime(); // This must be set cos strptime always stores in GM + + p = (char *) buf; + if (*format == '%') // skip any unexpected white space + while (isspace(*p)) + p++; + + // Special handling for LOOSE/SHORT formats... + if ((strcmp((char *) format, LOOSE_RFC850_FORMAT) == 0 || + strcmp((char *) format, LOOSE_RFC1123_FORMAT) == 0 || + strcmp((char *) format, ISO8601_SHORT_FORMAT) == 0) && + (r = Parse(p)) > 0) + return p+r; + + p = (char *) strptime (p, (char *) format, & Ht_tm); + +#ifdef TEST_HTDATETIME +// ViewStructTM(& Ht_tm); +#endif + + // Assign the new value to time_t value + SetDateTime(Ht_tm); + + return p; + +} + + +/////// + // C asctime() standard format +/////// + +void HtDateTime::SetAscTime(char *s) +{ + + // Unfortunately, I cannot think of an easy test to + // see if we have a weekday *FIX* + SetFTime(s, ASCTIME_FORMAT); + +} + +/////// + // RFC1123 standard Date format + // Sun, 06 Nov 1994 08:49:37 GMT +/////// + +void HtDateTime::SetRFC1123(char *s) +{ + + // abbreviated weekday name; + // day of the month; + // abbreviated month name; + // year as ccyy; + // hour ( 00 - 23); + // minute ( 00 - 59); + // seconds ( 00 - 59); + // time zone name; + + // First, if we have it, strip off the weekday + char *stripped; + stripped = strchr(s, ','); + if (stripped) + stripped++; + else + stripped = s; + + SetFTime(stripped, LOOSE_RFC1123_FORMAT); + +} + + +/////// + // RFC850 standard Date format + // Sunday, 06-Nov-1994 08:49:37 GMT +/////// + +void HtDateTime::SetRFC850(char *s) +{ + + // weekday name; + // day of the month; + // abbreviated month name; + // year within century; + // hour ( 00 - 23); + // minute ( 00 - 59); + // seconds ( 00 - 59); + // time zone name; + + // First, if we have it, strip off the weekday + char *stripped; + stripped = strchr(s, ','); + if (stripped) + stripped++; + else + stripped = s; + + SetFTime(stripped, LOOSE_RFC850_FORMAT); + +} + + +/////// + // ISO8601 standard Date format + // 1994-11-06 08:49:37 GMT +/////// + +void HtDateTime::SetISO8601(char *s) +{ + + // year as ccyy; + // month ( 01 - 12) + // day of the month + // hour ( 00 - 23) + // minute ( 00 - 59) + // seconds ( 00 - 59); + // time zone name; + + SetFTime(s, ISO8601_FORMAT); + +} + + +/////// + // Timestamp Date format (MySQL) without timezone + // 19941106084937 +/////// + +void HtDateTime::SetTimeStamp(char *s) +{ + + // year as ccyy; + // month ( 01 - 12) + // day of the month + // hour ( 00 - 23) + // minute ( 00 - 59) + // seconds ( 00 - 59); + + SetFTime(s, TIMESTAMP_FORMAT); + +} + + +/////// + // Default date and time format for the locale +/////// + +void HtDateTime::SetDateTimeDefault(char *s) +{ + + SetFTime(s, "%c"); + +} + + + + +/////// // Output Formats // /////// + + +/////// + // Personalized format such as C strftime function + // Overloaded version 1 +/////// + +size_t HtDateTime::GetFTime(char *s, size_t max, const char *format) const +{ + // Refresh static struct tm variable + + RefreshStructTM(); + + return strftime(s, max, format, & Ht_tm); + +} + +/////// + // Personalized format such as C strftime function + // Overloaded version 2 - The best to be used outside + // for temporary uses +/////// + +char *HtDateTime::GetFTime(const char *format) const +{ + + // Invoke GetFTime overloaded method + + if(GetFTime(my_strtime, MAXSTRTIME, format)) + return (char *)my_strtime; + else return 0; + +} + +/////// + // RFC1123 standard Date format + // Sun, 06 Nov 1994 08:49:37 GMT +/////// + +char *HtDateTime::GetRFC1123() const +{ + + // abbreviated weekday name; + // day of the month; + // abbreviated month name; + // year as ccyy; + // hour ( 00 - 23); + // minute ( 00 - 59); + // seconds ( 00 - 59); + // time zone name; + + GetFTime(my_strtime, MAXSTRTIME, RFC1123_FORMAT); + + return (char *)my_strtime; + +} + +/////// + // RFC850 standard Date format + // Sunday, 06-Nov-94 08:49:37 GMT +/////// + +char *HtDateTime::GetRFC850() const +{ + + // full weekday name + // day of the month + // abbreviated month name + // year within century ( 00 - 99 ) + // hour ( 00 - 23) + // minute ( 00 - 59) + // seconds ( 00 - 59); + // time zone name; + + GetFTime(my_strtime, MAXSTRTIME, RFC850_FORMAT); + + return (char *)my_strtime; + +} + +/////// + // C asctime() standard format +/////// + +char *HtDateTime::GetAscTime() const +{ + + GetFTime(my_strtime, MAXSTRTIME, ASCTIME_FORMAT); + return (char *)my_strtime; + +} + +/////// + // ISO8601 standard Date format + // 1994-11-06 08:49:37 GMT +/////// + +char *HtDateTime::GetISO8601() const +{ + + // year as ccyy; + // month ( 01 - 12) + // day of the month + // hour ( 00 - 23) + // minute ( 00 - 59) + // seconds ( 00 - 59); + // time zone name; + + GetFTime(my_strtime, MAXSTRTIME, ISO8601_FORMAT); + + return (char *)my_strtime; + +} + +/////// + // ISO8601 standard Date format + // 1994-11-06 08:49:37 GMT +/////// + +char *HtDateTime::GetShortISO8601() const +{ + + // year as ccyy; + // month ( 01 - 12) + // day of the month + + GetFTime(my_strtime, MAXSTRTIME, ISO8601_SHORT_FORMAT); + + return (char *)my_strtime; + +} + +/////// + // Timestamp Date format (MySQL) without timezone + // 19941106084937 +/////// + +char *HtDateTime::GetTimeStamp() const +{ + + // year as ccyy; + // month ( 01 - 12) + // day of the month + // hour ( 00 - 23) + // minute ( 00 - 59) + // seconds ( 00 - 59); + + GetFTime(my_strtime, MAXSTRTIME, TIMESTAMP_FORMAT); + + return (char *)my_strtime; + +} + +/////// + // Default date and time format for the locale +/////// + +char *HtDateTime::GetDateTimeDefault() const +{ + + GetFTime(my_strtime, MAXSTRTIME, "%c"); + + return (char *)my_strtime; + +} + +/////// + // Default date format for the locale +/////// + +char *HtDateTime::GetDateDefault() const +{ + + GetFTime(my_strtime, MAXSTRTIME, "%x"); + + return (char *)my_strtime; + +} + +/////// + // Default time format for the locale +/////// + +char *HtDateTime::GetTimeDefault() const +{ + + GetFTime(my_strtime, MAXSTRTIME, "%X"); + + return (char *)my_strtime; + +} + + + +/////// + // Set the static struct tm depending on localtime status +/////// + + +void HtDateTime::RefreshStructTM() const +{ + + if(local_time) + // Setting localtime + memcpy(& Ht_tm, localtime(&Ht_t), sizeof(struct tm)); + else + // Setting UTC or GM time + memcpy(& Ht_tm , gmtime(&Ht_t), sizeof(struct tm)); + +} + + +// Set the date time from a struct tm pointer + +void HtDateTime::SetDateTime(struct tm *ptm) +{ + + if(local_time) + Ht_t = mktime(ptm); // Invoke mktime + else + Ht_t = HtTimeGM(ptm); // Invoke timegm alike function + +} + + +// Set time to now + +void HtDateTime::SettoNow() +{ + Ht_t = time(0); +} + + +// Sets date by passing specific values +// The values are reffered to the GM date time +// Return false if failed + +bool HtDateTime::SetGMDateTime ( int year, int mon, int mday, + int hour, int min, int sec) +{ + struct tm tm_tmp; + + // Year + + if ( ! isAValidYear (year) ) return false; + + if( year < 100) + year=Year_From2To4digits (year); // For further checks it's converted + + // Assigning the year + + tm_tmp.tm_year=year-1900; + + + // Month + + if( ! isAValidMonth(mon) ) return false; + + tm_tmp.tm_mon=mon-1; // Assigning the month to the structure + + + // Day + + if ( ! isAValidDay ( mday, mon, year ) ) return false; + + tm_tmp.tm_mday=mday; // Assigning the day of the month + + + + if(hour >= 0 && hour < 24) tm_tmp.tm_hour = hour; + else return false; + + if(min >= 0 && min < 60) tm_tmp.tm_min = min; + else return false; + + if(sec >= 0 && sec < 60) tm_tmp.tm_sec = sec; + else return false; + + tm_tmp.tm_yday = 0; // day of the year (to be ignored) + tm_tmp.tm_isdst = 0; // default for GM (to be ignored) + + // Now we are going to insert the new values as time_t value + // This can only be done using GM Time and so ... + + if (isLocalTime()) + { + ToGMTime(); // Change to GM Time + SetDateTime(&tm_tmp); // commit it + ToLocalTime(); // And then return to Local Time + } + else SetDateTime(&tm_tmp); // only commit it + + return true; + +} + + +/////// + // Gets a struct tm from the value stored in the object + // It's a protected method. Not visible outside the class +/////// + +struct tm &HtDateTime::GetStructTM() const +{ + RefreshStructTM(); // refresh it + + return Ht_tm; +} + + +struct tm &HtDateTime::GetGMStructTM() const +{ + GetGMStructTM (Ht_tm); + return Ht_tm; +} + +void HtDateTime::GetGMStructTM(struct tm & t) const +{ + // Directly gets gmtime value + memcpy(& t , gmtime(& Ht_t), sizeof(struct tm)); +} + + +/////// + // Is a leap year? +/////// + +bool HtDateTime::LeapYear (int y) +{ + + if(y % 400 == 0 || ( y % 100 != 0 && y % 4 == 0)) + return true; // a leap year + else + return false; // and not +} + + +/////// + // Is a valid year number? +/////// + +bool HtDateTime::isAValidYear (int y) +{ + + if(y >= 1970 && y < 2069) return true; // simple check and most likely + + if(y >= 0 && y < 100) return true; // 2 digits year number + + return false; + +} + + +/////// + // Is a valid month number? +/////// + +bool HtDateTime::isAValidMonth (int m) +{ + + if( m >= 1 && m <= 12) return true; + else return false; + +} + + +/////// + // Is a valid day? +/////// + +bool HtDateTime::isAValidDay (int d, int m, int y) +{ + + if ( ! isAValidYear (y) ) return false; // Checks for the year + + if ( ! isAValidMonth (m) ) return false; // Checks for the month + + if(m == 2) + { + + // Expands the 2 digits year number + if ( y < 100 ) y=Year_From2To4digits(y); + + if ( LeapYear (y) ) // Checks for the leap year + { + if (d >= 1 && d <= 29) return true; + else return false; + } + } + + // Acts as default + + if (d >= 1 && d <= days [m -1]) return true; + else return false; + +} + + +/////// + // Comparison methods +/////// + + +int HtDateTime::DateTimeCompare (const HtDateTime & right) const +{ + int result; + + // Let's compare the date + + result=DateCompare(right); + + if(result) return result; + + // Same date. Let's compare the time + + result=TimeCompare(right); + + return result; // Nothing more to check + +} + + +int HtDateTime::GMDateTimeCompare (const HtDateTime & right) const +{ + // We must compare the whole time_t value + + if ( * this > right) return 1; // 1st greater than 2nd + if ( * this < right) return 1; // 1st lower than 2nd + + return 0; + +} + + +int HtDateTime::DateCompare (const HtDateTime & right) const +{ + + // We must transform them in 2 struct tm variables + + struct tm tm1, tm2; + + this->GetGMStructTM (tm1); + right.GetGMStructTM (tm2); + + // Let's compare them + return DateCompare (&tm1, &tm2); + +} + + +int HtDateTime::GMDateCompare (const HtDateTime & right) const +{ + + // We must transform them in 2 struct tm variables + // both referred to GM time + + struct tm tm1, tm2; + + this->GetGMStructTM (tm1); + right.GetGMStructTM (tm2); + + // Let's compare them + return DateCompare (&tm1, &tm2); + +} + + +int HtDateTime::TimeCompare (const HtDateTime & right) const +{ + + // We must transform them in 2 struct tm variables + + struct tm tm1, tm2; + + this->GetStructTM (tm1); + right.GetStructTM (tm2); + + return TimeCompare (&tm1, &tm2); + +} + + +int HtDateTime::GMTimeCompare (const HtDateTime & right) const +{ + + // We must transform them in 2 struct tm variables + + struct tm tm1, tm2; + + // We take the GM value of the time + this->GetGMStructTM (tm1); + right.GetGMStructTM (tm2); + + return TimeCompare (&tm1, &tm2); + +} + + + +/////// + // Static methods of comparison between 2 struct tm pointers +/////// + + +/////// + // Compares only the date (ignoring the time) +/////// + +int HtDateTime::DateCompare(const struct tm *tm1, const struct tm *tm2) +{ + + // Let's check the year + + if (tm1->tm_year < tm2->tm_year) return -1; + if (tm1->tm_year > tm2->tm_year) return 1; + + // Same year. Let's check the month + if (tm1->tm_mon < tm2->tm_mon) return -1; + if (tm1->tm_mon > tm2->tm_mon) return 1; + + // Same month. Let's check the day of the month + + if (tm1->tm_mday < tm2->tm_mday) return -1; + if (tm1->tm_mday > tm2->tm_mday) return 1; + + // They are equal for the date + return 0; +} + + +/////// + // Compares only the time (ignoring the date) +/////// + +int HtDateTime::TimeCompare(const struct tm *tm1, const struct tm *tm2) +{ + + // Let's check the hour + + if (tm1->tm_hour < tm2->tm_hour) return -1; + if (tm1->tm_hour > tm2->tm_hour) return 1; + + // Same hour . Let's check the minutes + + if (tm1->tm_min < tm2->tm_min) return -1; + if (tm1->tm_min > tm2->tm_min) return 1; + + // Ooops !!! Same minute. Let's check the seconds + + if (tm1->tm_sec < tm2->tm_sec) return -1; + if (tm1->tm_sec > tm2->tm_sec) return 1; + + // They are equal for the time + return 0; +} + + +/////// + // Compares both date and time +/////// + +int HtDateTime::DateTimeCompare(const struct tm *tm1, const struct tm *tm2) +{ + + int compare_date = DateCompare(tm1, tm2); + + if(compare_date) return compare_date; // Different days + + // We are in the same day. Let's check the time + + int compare_time = TimeCompare(tm1, tm2); + if(compare_time) return compare_time; // Different time + + // Equal + return 0; +} + +time_t HtDateTime::HtTimeGM (struct tm *tm) +{ + +#if HAVE_TIMEGM + return timegm (tm); +#else + return Httimegm (tm); // timegm replacement in timegm.c + // static time_t gmtime_offset; + // tm->tm_isdst = 0; + // return __mktime_internal (tm, gmtime, &gmtime_offset); +#endif + +} + + + +// Returns the difference in seconds between two HtDateTime Objects + +int HtDateTime::GetDiff(const HtDateTime &d1, const HtDateTime &d2) +{ + + return (int) ( d1.Ht_t - d2.Ht_t ); + +} + + + + + + +/////// + // Only for test and debug +/////// + +#ifdef TEST_HTDATETIME + + +/////// + // View of struct tm fields +/////// + +void HtDateTime::ViewStructTM() +{ + // Default viewing: refresh depending on time_t value + + RefreshStructTM(); // Refresh static variable + + ViewStructTM(&Ht_tm); +} + +void HtDateTime::ViewStructTM(struct tm *ptm) +{ + + cout << "Struct TM fields" << endl; + cout << "================" << endl; + cout << "tm_sec :\t" << ptm->tm_sec << endl; + cout << "tm_min :\t" << ptm->tm_min << endl; + cout << "tm_hour :\t" << ptm->tm_hour << endl; + cout << "tm_mday :\t" << ptm->tm_mday << endl; + cout << "tm_mon :\t" << ptm->tm_mon << endl; + cout << "tm_year :\t" << ptm->tm_year << endl; + cout << "tm_wday :\t" << ptm->tm_wday << endl; + cout << "tm_yday :\t" << ptm->tm_yday << endl; + cout << "tm_isdst :\t" << ptm->tm_isdst<< endl; + +} + + + + +int HtDateTime::Test(void) +{ + + int ok=1; + + const char *test_dates[] = + { + "1970.01.01 00:00:00", + "1970.01.01 00:00:01", + "1972.02.05 23:59:59", + "1972.02.28 00:59:59", + "1972.02.28 23:59:59", + "1972.02.29 00:00:00", + "1972.03.01 13:00:04", + "1973.03.01 12:00:00", + "1980.01.01 00:00:05", + "1984.12.31 23:00:00", + "1997.06.05 17:55:35", + "1999.12.31 23:00:00", + "2000.01.01 00:00:05", + "2000.02.28 23:00:05", + "2000.02.29 23:00:05", + "2000.03.01 00:00:05", + "2007.06.05 17:55:35", + "2038.01.19 03:14:07", + 0 + }; + + const char *test_dates_ISO8601[] = + { + "1970-01-01 00:00:00 GMT", + "1970-01-01 00:00:00 CET", + "1990-02-27 23:30:20 GMT", + "1999-02-28 06:53:40 GMT", + "1975-04-27 06:53:40 CET", + 0 + }; + + const char *test_dates_RFC1123[] = + { + "Sun, 06 Nov 1994 08:49:37 GMT", + "Sun, 25 Apr 1999 17:49:37 GMT", + "Sun, 25 Apr 1999 17:49:37 CET", + 0 + }; + + const char *test_dates_RFC850[] = + { + "Sunday, 06-Nov-94 08:49:37 GMT", + "Sunday, 25-Apr-99 17:49:37 GMT", + "Sunday, 25-Apr-99 17:49:37 CET", + 0 + }; + + + const char myformat[]="%Y.%m.%d %H:%M:%S"; + + // Tests a personal format + + cout << endl << "Beginning Test of a personal format such as " + << myformat << endl << endl; + + if (Test((char **)test_dates, (const char *)myformat)) + cout << "Test OK." << endl; + else + { + cout << "Test Failed." << endl; + ok=0; + } + + + // Tests ISO 8601 Format + + cout << endl << "Beginning Test of ISO 8601 format" << endl << endl; + + if(Test((char **)test_dates_ISO8601, (const char *)ISO8601_FORMAT)) + cout << "Test OK." << endl; + else + { + cout << "Test Failed." << endl; + ok=0; + } + + + // Tests RFC 1123 Format + + cout << endl << "Beginning Test of RFC 1123 format" << endl << endl; + + if (Test((char **)test_dates_RFC1123, (const char *)RFC1123_FORMAT)) + cout << "Test OK." << endl; + else + { + cout << "Test Failed." << endl; + ok=0; + } + + + // Tests RFC 850 Format + + cout << endl << "Beginning Test of RFC 850 format" << endl << endl; + + if (Test((char **)test_dates_RFC850, (const char *)RFC850_FORMAT)) + cout << "Test OK." << endl; + else + { + cout << "Test Failed." << endl; + ok=0; + } + + + return(ok ? 1 : 0); + +} + + + +int HtDateTime::Test(char **test_dates, const char *format) +{ + int i, ok = 1; + HtDateTime orig, conv; + + for (i = 0; (test_dates[i]); i++) + { + + cout << "\t " << i+1 << "\tDate string parsing of:" << endl; + cout << "\t\t" << test_dates[i] << endl; + cout << "\t\tusing format: " << format << endl << endl; + + orig.SetFTime(test_dates[i], format); + + orig.ComparisonTest(conv); + + conv=orig; + + if (orig != conv) + { + cout << "HtDateTime test failed!" << endl; + cout << "\t Original : " << orig.GetRFC1123() << endl; + cout << "\t Converted: " << orig.GetRFC1123() << endl; + ok = 0; + } + else + { + orig.ToLocalTime(); + cout << endl << "\t Localtime viewing" << endl; + orig.ViewFormats(); + orig.ToGMTime(); + cout << endl << "\t GMtime viewing" << endl; + orig.ViewFormats(); + //orig.ViewStructTM(); + } + + cout << endl; + + } + + return ok; +} + + +void HtDateTime::ComparisonTest (const HtDateTime &right) const +{ + int result; + + + cout << "Comparison between:" << endl; + + cout << " 1. " << this->GetRFC1123() << endl; + cout << " 2. " << right.GetRFC1123() << endl; + cout << endl; + + +/////// + // Complete comparison +/////// + + cout << "\tComplete comparison (date and time)" << endl; + result = this->DateTimeCompare (right); + + cout << "\t\t " << this->GetDateTimeDefault(); + + if (result > 0 ) + cout << " is greater than "; + else if (result < 0 ) + cout << " is lower than "; + else cout << " is equal to "; + + cout << " " << right.GetDateTimeDefault() << endl; + + + +/////// + // Date comparison +/////// + + cout << "\tDate comparison (ignoring time)" << endl; + result = this->DateCompare (right); + + cout << "\t\t " << this->GetDateDefault(); + + if (result > 0 ) + cout << " is greater than "; + else if (result < 0 ) + cout << " is lower than "; + else cout << " is equal to "; + + cout << " " << right.GetDateDefault() << endl; + + +/////// + // Date comparison (after GM time conversion) +/////// + + cout << "\tDate comparison (ignoring time) - GM time conversion" << endl; + result = this->GMDateCompare (right); + + cout << "\t\t " << this->GetDateDefault(); + + if (result > 0 ) + cout << " is greater than "; + else if (result < 0 ) + cout << " is lower than "; + else cout << " is equal to "; + + cout << " " << right.GetDateDefault() << endl; + + + +/////// + // Time comparison +/////// + + cout << "\tTime comparison (ignoring date)" << endl; + result = this->TimeCompare (right); + + cout << "\t\t " << this->GetTimeDefault(); + + if (result > 0 ) + cout << " is greater than "; + else if (result < 0 ) + cout << " is lower than "; + else cout << " is equal to "; + + cout << " " << right.GetTimeDefault() << endl; + + +/////// + // Time comparison (after GM time conversion) +/////// + + cout << "\tTime comparison (ignoring date) - GM time conversion" << endl; + result = this->GMTimeCompare (right); + + cout << "\t\t " << this->GetTimeDefault(); + + if (result > 0 ) + cout << " is greater than "; + else if (result < 0 ) + cout << " is lower than "; + else cout << " is equal to "; + + cout << " " << right.GetTimeDefault() << endl; + +} + + + +void HtDateTime::ViewFormats() +{ + + cout << "\t\t RFC 1123 Format : " << GetRFC1123() << endl; + cout << "\t\t RFC 850 Format : " << GetRFC850() << endl; + cout << "\t\t C Asctime Format: " << GetAscTime() << endl; + cout << "\t\t ISO 8601 Format : " << GetISO8601() << endl; + +} + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtDateTime.h b/debian/htdig/htdig-3.2.0b6/htlib/HtDateTime.h new file mode 100644 index 00000000..c212b49a --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtDateTime.h @@ -0,0 +1,533 @@ +// +// HtDateTime.h +// +// HtDateTime: Parse, split, compare and format dates and times. +// Uses locale. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtDateTime.h,v 1.19 2004/05/28 13:15:20 lha Exp $ + +/////// + // Class for Date and Time + // Gabriele Bartolini - Prato - Italia + // Started: 22.04.1999 +/////// + +// Version: 1.0 +// Release date: 07.05.1999 + +// +// General purpose of HtDateTime +// The general purpose of this class, is to provide an interface for +// date and time managing, and to unload the programmer to manage +// time_t, struct tm, time system functions and other related procedures +// locally ... Everything regarding time and date must be put here. +// D'you agree with me? Well, here is a summary of the class capabilities. + +// Attributes of the class: +// +// HtDateTime class has only 2 member attributes +// - time_t Ht_t +// - bool local_time +// +// Obviously Ht_t contains the most important piece of information. +// local_time assumes a true value if we wanna consider the date and +// time information as local. False means that our object value is +// referred to the Greenwich Meridian time. + +// Interface provided: +// +// Construction: +// - Default: set the date time value to now +// - By passing a time_t value or pointer: Set to it +// - By passing a struct tm value or pointer: Set to it +// The last one could be useful sometimes. But it had better not to +// be used. +// +// Parsing interface: +// Not yet implemented ... :-) +// +// Setting Interface: +// - from time_t: copy the time_t value into the object +// - from struct tm: set the object's time_t value by converting +// the value from the struct tm. If local_time is set to true, +// converts it with mktime, else uses HtTimeGM. +// - set to now +// - from a string, by passing the input format: the method uses +// strptime syntax (and invokes Htstrptime). For now, timezone +// is ignored, and so data are stored as a GM date time value. +// - from an int series, by specifying all the information (year, +// month, day, hour, minute and second). It's all stored as +// GM value. +// - from various standard formats, such as C asctime, RFC 1123, +// RFC 850 (these 3 types are suggested by the HTTP/1.1 standard), +// ISO 8601, date and time default representation for the locale. +// This list could get longer ... It all depends on us. +// - setting the date and time to be represented in a local value +// or universal (GM) one. +// +// Getting Interface +// - in a personalized output format, by passing a string with +// strftime values. +// - in various standard formats, like C asctime, RFC 1123, +// RFC 850, ISO 8601 (short too), date and time default +// representation for the locale. +// - getting the time_t value +// - queries the local time status +// - getting specific piece of information of both the date and the +// the time, like the year, the month, the day of the week, of +// the year or of the month, ... In short, every kind of thing +// a tm structure is able to store ... +// +// Operator overloading +// - Copy +// - Every kind of logical comparison between 2 objects +// +// Comparison interface +// This is divided in 2 sections. +// - Static section: +// comparison are made on a 2 struct tm values basis. +// It's possible to compare the whole date time value, or +// simply the date or the time value. +// - Member functions section: +// comparison are made between 2 HtDateTime objects. +// You can compare either the whole date time, or the date, or the +// time, both as they are or referring their values to the GM value. +// +// System functions interface +// They are all done with previous "configure" checks +// - for strptime +// - for timegm +// +// Static methods +// - check for a leap year +// - check for a valid year number (according with time_t capabilities) +// - check for a valid month number +// - check for a valid day +// - converts a 2 digits year number into a 4 digits one: from 1970 to 2069. +// - converts a 4 digits year number into a 2 digits one. +// - retrieve the difference in seconds between 2 HtDateTime objs +// +// Test Interface (only by defining TEST_HTDATETIME directive). +// + + +#ifndef _HTDATETIME_H +#define _HTDATETIME_H + +#ifdef HAVE_CONFIG_H +# include "htconfig.h" +#endif + +#if TIME_WITH_SYS_TIME +#include <sys/time.h> +#include <time.h> +#else +# if HAVE_SYS_TIME_H +# include <sys/time.h> +# else +# include <time.h> +# endif +#endif + +#include "htString.h" + +// If you wanna do some tests +#define TEST_HTDATETIME + + +class HtDateTime +{ +public: + +/////// + // Construction +/////// + + // Default: now and local + HtDateTime() {SettoNow(); ToLocalTime();} + + // From an integer (seconds from epoc) +// Causes ambiguity in systems with time_t an integer... +// HtDateTime(const int i) {SetDateTime((time_t)i); ToLocalTime();} + + // From a time_t value and pointer + HtDateTime(time_t &t) {SetDateTime(t); ToLocalTime();} + HtDateTime(time_t *t) {SetDateTime(t); ToLocalTime();} + + // From a struct tm value and pointer + HtDateTime(struct tm &t) {SetDateTime(t); ToLocalTime();} + HtDateTime(struct tm *t) {SetDateTime(t); ToLocalTime();} + + // Copy constructor + inline HtDateTime(const HtDateTime& rhs); + +/////// + // Interface methods +/////// + +/////// + // "Parsing" interface +/////// + + int Parse(const char *); // It looks for the similar format + // then sets the date by invoking + // right method + + +/////// + // "Setting" interface +/////// + + // Setting from a time_t value + void SetDateTime(const time_t &t) { Ht_t = t; } // by reference + void SetDateTime(const time_t *t) { Ht_t = *t; } // by pointer + + // Set object time_t value from a struct tm + void SetDateTime(struct tm *); // by pointer + inline void SetDateTime(struct tm &t) { SetDateTime(&t);} // by reference + + // Set GM Time from single values input + // Return true if it all went good, false else + bool SetGMDateTime( int year, int mon, int mday, + int hour=0, int min=0, int sec=0); + + // Set to Now + void SettoNow(); + + // Parsing various input string format + // It ignores time Zone value - always stores as GM + char *SetFTime(const char *, const char *); // as strptime + + void SetAscTime(char *); // Sun Nov 6 08:49:37 1994 + void SetRFC1123(char *); // Sun, 06 Nov 1994 08:49:37 GMT + void SetRFC850(char *); // Sunday, 06-Nov-94 08:49:37 GMT + void SetISO8601(char *); // 1994-11-06 08:49:37 GMT + void SetTimeStamp(char *); // 19941106084937 + + void SetDateTimeDefault(char *); // Default date and time representation + // for the locale + + /////// + // Methods for setting Local and GM time formats (Switches) + /////// + + void ToLocalTime() {local_time=true;} + void ToGMTime() {local_time=false;} + + +/////// + // "Getting" interface +/////// + + + /////// + // Output formats + /////// + + // Personalized output + char *GetFTime(const char *format) const; // as strftime + size_t GetFTime(char *, size_t, const char *) const; // as strftime + + char *GetAscTime() const; // Sun Nov 6 08:49:37 1994 + char *GetRFC1123() const; // Sun, 06 Nov 1994 08:49:37 GMT + char *GetRFC850() const; // Sunday, 06-Nov-94 08:49:37 GMT + char *GetISO8601() const; // 1994-11-06 08:49:37 GMT + char *GetTimeStamp() const; // 19941106084937 + + char *GetDateTimeDefault() const; // Default date and time representation + // for the locale + + // Partial (only date or only time) + char *GetShortISO8601() const; // 1994-11-06 + char *GetDateDefault() const; // Default date form for the locale + char *GetTimeDefault() const; // Default time form for the locale + + + /////// + // Gets the time_t value + /////// + + time_t GetTime_t() const {return Ht_t;} + + + /////// + // Gets specific date and time values (from a struct tm) + /////// + + // Gets the year + int GetYear() const { return ( GetStructTM().tm_year + 1900) ;} + + // Gets the month + int GetMonth() const { return (GetStructTM().tm_mon + 1);} + + // Gets the day of the week (since Sunday) + int GetWDay() const { return (GetStructTM().tm_wday + 1);} + + // Gets the day of the month + int GetMDay() const { return GetStructTM().tm_mday;} + + // Gets the day since january 1 + int GetYDay() const { return (GetStructTM().tm_yday + 1);} + + // Gets the hour + int GetHour() const { return GetStructTM().tm_hour;} + + // Gets the minute + int GetMinute() const { return GetStructTM().tm_min;} + + // Gets the second + int GetSecond() const { return GetStructTM().tm_sec;} + + // Daylight saving time is in effect at that time? + int GetIsDst() const { return GetStructTM().tm_isdst;} + + + /////// + // Methods for querying localtime status + /////// + + bool isLocalTime() const {return local_time;} + bool isGMTime() const {return !local_time;} + + + +/////// + // Methods for comparison +/////// + + // Returns 0 if equal, -1 if tm1 is lower than tm2, 1 if tm1 is greater than tm2 + + int DateTimeCompare (const HtDateTime &) const; // Compares both date and time + + int DateCompare (const HtDateTime &) const; // Compares the date + int TimeCompare (const HtDateTime &) const; // Compares the time + + // Refers the date and the time to a GM value, then compares + int GMDateTimeCompare (const HtDateTime &) const; // Compares both date and time + int GMDateCompare (const HtDateTime &) const; // Compares the date + int GMTimeCompare (const HtDateTime &) const; // Compares the time + + +/////// + // Operator overloading +/////// + + // For comparisons - between objects of the same class + + inline bool operator==(const HtDateTime &right) const; + inline bool operator<(const HtDateTime &right) const; + + bool operator!=(const HtDateTime &right) const + {return !( *this == right );} + + bool operator>=(const HtDateTime &right) const + {return !( *this < right);} + + bool operator<=(const HtDateTime &right) const + {return !( right < *this);} + + bool operator>(const HtDateTime &right) const + {return right < *this; } + + + // For comparisons - between HtDateTime objects and int + + bool operator==(const int right) const // with an int + {return ( Ht_t == (time_t) right );} + + bool operator<(const int right) const // with an int + {return ( Ht_t < (time_t) right );} + + bool operator!=(const int right) const // with an int + {return !( *this == right );} + + bool operator>=(const int right) const // with an int + {return !( *this < right);} + + bool operator<=(const int right) const // with an int + {return !( *this > right);} + + bool operator>(const int right) const // with an int + {return (Ht_t > (time_t) right); } + + + // For Copy + + inline HtDateTime &operator=(const HtDateTime &right); + inline HtDateTime &operator=(const int right); + + + + +/////// // STATIC METHODS // /////// + + // Here we can add static methods as we want more :-) + // Then invoke them with HtDateTime::MethodXXX () + + + inline static bool LeapYear(int); // Is a leap year? + + // These checks are made for time_t compatibility + inline static bool isAValidYear(int); // Is a valid year number + + inline static bool isAValidMonth(int); // Is a valid month number + inline static bool isAValidDay(int, int, int); // Is a valid day + + + // Converts a 2 digits year in a 4 one - with no checks + static int Year_From2To4digits (int y) + { + if ( y >= 70 ) return y+1900; + else return y+2000; + } + + // Converts a 4 digits year in a 2 one - with no checks + static int Year_From4To2digits (int y) + { + if ( y >= 2000 ) return y - 2000; + else return y - 1900; + } + + static int GetDiff(const HtDateTime &, const HtDateTime &); + + // Check equality from 2 struct tm pointers + // Returns 0 if equal, -1 if tm1 is lower than tm2, 1 if tm1 is greater than tm2 + + // Compares the whole time information (both date and time) + static int DateTimeCompare(const struct tm *tm1, + const struct tm *tm2); + + // Compares only date + static int DateCompare(const struct tm *tm1, + const struct tm *tm2); + + // Compares only time + static int TimeCompare(const struct tm *tm1, + const struct tm *tm2); + + + +/////// // HIDDEN ATTRIBUTES & METHODS // /////// + +protected: // to permit inheritance + + time_t Ht_t; + bool local_time; + + static const int days[]; + +/////// + // Sets and gets the struct tm depending on local_time status +/////// + + void RefreshStructTM() const; // Refresh its content + struct tm &GetStructTM() const; // gets it + void GetStructTM(struct tm & t) const { t=GetStructTM(); } // Gets and copy + + +/////// + // Gets the struct tm ignoring local_time status +/////// + + struct tm &GetGMStructTM() const; // gets it + void GetGMStructTM(struct tm &) const; // Gets and copy + + +/////// + // Interface for system functions +/////// + + // Interface for timegm + static time_t HtTimeGM (struct tm*); + + +#ifdef TEST_HTDATETIME + +/////// + // Only for debug: view of struct tm fields +/////// + +public: + + static void ViewStructTM(struct tm *); // view of struct tm fields + virtual void ViewStructTM(); // view of struct tm fields + void ViewFormats(); // View of various formats + + void ComparisonTest (const HtDateTime &) const; // comparison + + + // Test of the class + static int Test(void); + static int Test(char **test_dates, const char *format); + +#endif + +}; + + +/////// + // Copy constructor +/////// + +inline +HtDateTime::HtDateTime (const HtDateTime& rhs) +{ + // Copy the contents + Ht_t = rhs.Ht_t; + local_time = rhs.local_time; +} + +/////// + // Operator overloading +/////// + +inline +bool HtDateTime::operator==(const HtDateTime &right) const +{ + if(Ht_t==right.Ht_t) + return true; + else + return false; +} + +inline +bool HtDateTime::operator<(const HtDateTime &right) const +{ + if(Ht_t < right.Ht_t) return true; + else return false; +} + +/////// + // Copy +/////// + +inline +HtDateTime &HtDateTime::operator=(const HtDateTime &right) +{ + if (this != &right) + { + Ht_t=right.Ht_t; // Copy the time_t value + local_time=right.local_time; // Copy the local_time flag + } + + return *this; +} + +inline +HtDateTime &HtDateTime::operator=(const int right) +{ + Ht_t=(time_t)right; // Copy the int as a time_t value + ToLocalTime(); + return *this; +} + + + + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtHeap.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtHeap.cc new file mode 100644 index 00000000..92267767 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtHeap.cc @@ -0,0 +1,198 @@ +// +// HtHeap.cc +// +// HtHeap: A Heap class which holds objects of type Object. +// (A heap is a semi-ordered tree-like structure. +// it ensures that the first item is *always* the largest. +// NOTE: To use a heap, you must implement the Compare() function for +// your Object classes. The assumption used here is -1 means +// less-than, 0 means equal, and +1 means greater-than. Thus +// this is a "min heap" for that definition.) +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtHeap.cc,v 1.12 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtHeap.h" + +#ifdef HAVE_STD +#include <fstream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <fstream.h> +#endif /* HAVE_STD */ + +//********************************************************************* +// void HtHeap::HtHeap() +// Default constructor +// +HtHeap::HtHeap() +{ + data = new HtVector; +} + + +//********************************************************************* +// void HtHeap::HtHeap(HtVector vector) +// Constructor from vector +// (has the side effect of not allocating double memory) +// +HtHeap::HtHeap(HtVector vector) +{ + int size = vector.Count(); + data = static_cast<HtVector*>(vector.Copy()); + + // Now we have to "heapify" -- start at the first interior node + // And push each node down into its subtree + // (This is O(n)!) + for (int i = parentOf(size); i >= 0; i--) + pushDownRoot(i); +} + + +//********************************************************************* +// void HtHeap::~HtHeap() +// Destructor +// +HtHeap::~HtHeap() +{ + Destroy(); +} + + +//********************************************************************* +// void HtHeap::Destroy() +// Deletes all objects from the heap +// +void HtHeap::Destroy() +{ + data->Destroy(); + delete data; +} + + +//********************************************************************* +// void HtHeap::Add(Object *object) +// Add an object to the heap. +// +void HtHeap::Add(Object *object) +{ + data->Add(object); + percolateUp(data->Count() - 1); +} + + + +//********************************************************************* +// Object *HtHeap::Remove() +// Remove an object from the top of the heap +// This requires re-heapifying by placing the last element on the top +// and pushing it down. +// +Object *HtHeap::Remove() +{ + Object *min = Peek(); + + data->Assign(data->Last(), 0); + data->RemoveFrom(data->Count()-1); + + if (data->Count() > 1) + pushDownRoot(0); + + return min; +} + +//********************************************************************* +// HtHeap *HtHeap::Copy() const +// Return a deep copy of the heap. +// +Object *HtHeap::Copy() const +{ + HtHeap *heap = new HtHeap(*data); + + return heap; +} + + +//********************************************************************* +// HtHeap &HtHeap::operator=(HtHeap &heap) +// Return a deep copy of the heap. +// +HtHeap &HtHeap::operator=(HtHeap &heap) +{ + Destroy(); + data = heap.data; + return *this; +} + +//********************************************************************* +// voide HtHeap::percolateUp(int leaf) +// Pushes the node pointed to by leaf upwards +// it will travel as far as possible upwards to ensure the data is a heap +// +void HtHeap:: percolateUp(int leaf) +{ + int parent = parentOf(leaf); + Object *value = data->Nth(leaf); + while (leaf > 0 && + (value->compare(*(data->Nth(parent))) < 0)) + { + data->Assign(data->Nth(parent), leaf); + leaf = parent; + parent = parentOf(leaf); + } + data->Assign(value, leaf); +} + +//********************************************************************* +// void HtHeap::pushDownRoot(int root) +// Pushes the node pointed to by root into the heap +// it will go down as far as necessary to ensure the data is a heap +// +void HtHeap::pushDownRoot(int root) +{ + int size = data->Count() - 1; + Object *value = data->Nth(root); + while (root < size) + { + int childPos = leftChildOf(root); + if (childPos < size) + { + if ( rightChildOf(root) < size && + data->Nth(childPos + 1)->compare(*(data->Nth(childPos))) < 0 ) + { + childPos++; + } + if ( data->Nth(childPos)->compare(*value) < 0 ) // -1, so smaller + { + // We have to swap this node with the root and then loop + data->Assign(data->Nth(childPos), root); + data->Assign(value, childPos); + root = childPos; + } + else + { + // Found the right position, so we're done + data->Assign(value, root); + return; + } + } + else // childPos >= heapSize + { + // At a leaf, so we're done + data->Assign(value, root); + return; + } + } +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtHeap.h b/debian/htdig/htdig-3.2.0b6/htlib/HtHeap.h new file mode 100644 index 00000000..75e3c411 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtHeap.h @@ -0,0 +1,92 @@ +// +// HtHeap.h +// +// HtHeap: A Heap class which holds objects of type Object. +// (A heap is a semi-ordered tree-like structure. +// it ensures that the first item is *always* the largest. +// NOTE: To use a heap, you must implement the Compare() function for +// your Object classes. The assumption used here is -1 means +// less-than, 0 means equal, and +1 means greater-than. Thus +// this is a "min heap" for that definition.) +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtHeap.h,v 1.7 2004/05/28 13:15:20 lha Exp $ +// +// +#ifndef _HtHeap_h_ +#define _HtHeap_h_ +#include "Object.h" +#include "HtVector.h" + +class HtHeap : public Object +{ +public: + // + // Constructor/Destructor + // + HtHeap(); + HtHeap(HtVector vector); + ~HtHeap(); + + // + // Add() will add an Object to the heap in the appropriate location + // + void Add(Object *); + + // + // Destroy() will delete all the objects in the heap. This is + // equivalent to calling the destructor + // + void Destroy(); + + // + // Peek() will return a reference to the top object in the heap. + // + Object *Peek() {return data->Nth(0);} + + // + // Remove() will return a reference as Peek() but will also + // remove the reference from the heap and re-heapify + // + Object *Remove(); + + // + // Access to the number of elements + // + int Count() {return data->Count();} + int IsEmpty() {return data->IsEmpty();} + + // + // Deep copy member function + // + Object *Copy() const; + + // + // Assignment + // + HtHeap &operator= (HtHeap *heap) {return *this = *heap;} + HtHeap &operator= (HtHeap &heap); + +protected: + // The vector class should keep track of everything for us + HtVector *data; + + // Functions for establishing the relations between elements + int parentOf (int i) + { return (i - 1)/2; } + int leftChildOf (int i) + { return 2*i + 1; } + int rightChildOf (int i) + { return 2* (i+1); } + + // Protected procedures for performing heap-making operations + void percolateUp (int leaf); // pushes the node up as far as possible + void pushDownRoot (int root); // pushes the node down as necessary +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtMaxMin.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtMaxMin.cc new file mode 100644 index 00000000..59412b24 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtMaxMin.cc @@ -0,0 +1,66 @@ +// HtMaxMin +// +// macros and tools for computing max and min of values +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtMaxMin.cc,v 1.5 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include"HtMaxMin.h" + +unsigned int +HtMaxMin::max_v(unsigned int *vals,int n) +{ + unsigned int maxv=vals[0]; + for(int i=1;i<n;i++) + { + unsigned int v=vals[i]; + if(v>maxv){maxv=v;} + } + return(maxv); +} + +unsigned short +HtMaxMin::max_v(unsigned short *vals,int n) +{ + unsigned short maxv=vals[0]; + for(int i=1;i<n;i++) + { + unsigned short v=vals[i]; + if(v>maxv){maxv=v;} + } + return(maxv); +} + +unsigned int +HtMaxMin::min_v(unsigned int *vals,int n) +{ + unsigned int minv=vals[0]; + for(int i=1;i<n;i++) + { + unsigned int v=vals[i]; + if(v<minv){minv=v;} + } + return(minv); +} + +unsigned short +HtMaxMin::min_v(unsigned short *vals,int n) +{ + unsigned short minv=vals[0]; + for(int i=1;i<n;i++) + { + unsigned short v=vals[i]; + if(v<minv){minv=v;} + } + return(minv); +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtMaxMin.h b/debian/htdig/htdig-3.2.0b6/htlib/HtMaxMin.h new file mode 100644 index 00000000..26d0e44e --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtMaxMin.h @@ -0,0 +1,34 @@ +// HtMaxMin +// +// macros and tools for computing max and min of values +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtMaxMin.h,v 1.5 2004/05/28 13:15:20 lha Exp $ +// + +#ifndef _HtMaxMin_h_ +#define _HtMaxMin_h_ + +#define HtMAX(a,b) (((a)>(b)) ? (a) : (b)) +#define HtMIN(a,b) (((a)<(b)) ? (a) : (b)) + + +// Max/Min value of an array +class HtMaxMin +{ + public: + // compute max/min of an array of values + static unsigned int max_v(unsigned int *vals,int n); + static unsigned short max_v(unsigned short *vals,int n); + static unsigned int min_v(unsigned int *vals,int n); + static unsigned short min_v(unsigned short *vals,int n); +}; + + + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtPack.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtPack.cc new file mode 100644 index 00000000..8026622d --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtPack.cc @@ -0,0 +1,450 @@ +// +// HtPack.cc +// +// HtPack: Compress and uncompress data in e.g. simple structures. +// The structure must have the layout defined in the ABI; +// the layout the compiler generates. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtPack.cc,v 1.8 2004/05/28 13:15:20 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtPack.h" + +#include <ctype.h> +#include <stdlib.h> + +// For the moment, these formats are accepted: +// "i" native int, with most compressed value 0 +// "u" unsigned int, with most compressed value 0 +// "c" unsigned int, with most compressed value 1. +// +// If someone adds other formats (and uses them), please note +// that structure padding may give surprising effects on some +// (most) platforms, for example if you try to unpack a +// structure with the imagined signature "isi" (int, short, int). +// You will want to solve that portably. +// +// Compression is done to 2 bits description (overhead) each, +// plus variable-sized data. +// Theoretically, different formats can use different number of +// bits in the description with a few changes. +// The description is located in a byte before every four +// "fields". +String +htPack(const char format[], const char *data) +{ + const char *s = format; + + // We insert the encodings by number, rather than shifting and + // inserting at the "bottom". This should make it faster for + // decoding, which presumably is more important than the speed + // of encoding. + int code_no = 0; + + // Make a wild guess that we will compress some ordinary sized + // struct. This guess only has speed effects. + String compressed(60); + + // Accumulated codes. + unsigned int description = 0; + + // Store the encoding here. We cannot use a char *, as the + // string may be reallocated and moved. + int code_index = 0; + + // Make place for the first codes. + compressed << '\0'; + + // Format string loop. + while (*s) + { + int fchar = *s++; + int n; + + if (isdigit(*s)) + { + char* t; + n = strtol(s, &t, 10); + s = t; + } + else + n = 1; + + // Loop over N in e.g. "iN" (default 1). + while (n--) + { + // Format character handling. + switch (fchar) + { + case 'c': + { + // We compress an unsigned int with the most common + // value 1 as this: + // 00 - value is 1. + // 01 - value fits in unsigned char - appended. + // 10 - value fits in unsigned short - appended. + // 11 - just plain unsigned int - appended (you lose). + unsigned int value; + + // Initialize, but allow disalignment. + memcpy(&value, data, sizeof value); + data += sizeof(unsigned int); + + int mycode; + if (value == 1) + { + mycode = 0; + } + else + { + unsigned char charvalue = (unsigned char) value; + unsigned short shortvalue = (unsigned short) value; + if (value == charvalue) + { + mycode = 1; + compressed << charvalue; + } + else if (value == shortvalue) + { + mycode = 2; + compressed.append((char *) &shortvalue, sizeof shortvalue); + } + else + { + mycode = 3; + compressed.append((char *) &value, sizeof value); + } + } + + description |= mycode << (2*code_no++); + } + break; + + case 'i': + { + // We compress a (signed) int as follows: + // 00 - value is 0. + // 01 - value fits in char - appended. + // 10 - value fits in short - appended. + // 11 - just plain int - appended (you lose). + int value; + + // Initialize, but allow disalignment. + memcpy(&value, data, sizeof value); + data += sizeof(int); + + int mycode; + if (value == 0) + { + mycode = 0; + } + else + { + char charvalue = char(value); + short shortvalue = short(value); + if (value == charvalue) + { + mycode = 1; + compressed << charvalue; + } + else if (value == shortvalue) + { + mycode = 2; + compressed.append((char *) &shortvalue, sizeof shortvalue); + } + else + { + mycode = 3; + compressed.append((char *) &value, sizeof value); + } + } + + description |= mycode << (2*code_no++); + } + break; + + case 'u': + { + // We compress an unsigned int like an int: + // 00 - value is 0. + // 01 - value fits in unsigned char - appended. + // 10 - value fits in unsigned short - appended. + // 11 - just plain unsigned int - appended (you lose). + unsigned int value; + + // Initialize, but allow disalignment. + memcpy(&value, data, sizeof value); + data += sizeof(unsigned int); + + int mycode; + if (value == 0) + { + mycode = 0; + } + else + { + unsigned char charvalue = (unsigned char) value; + unsigned short shortvalue = (unsigned short) value; + if (value == charvalue) + { + mycode = 1; + compressed << charvalue; + } + else if (value == shortvalue) + { + mycode = 2; + compressed.append((char *) &shortvalue, sizeof shortvalue); + } + else + { + mycode = 3; + compressed.append((char *) &value, sizeof value); + } + } + + description |= mycode << (2*code_no++); + } + break; + + default: +#ifndef NOSTREAM +#ifdef DEBUG + if (1) + cerr << "Invalid char \'" << char(fchar) + << "\' in pack format \"" << format << "\"" + << endl; + return ""; +#endif +#endif + ; // Must always have a statement after a label. + } + + // Assuming 8-bit chars here. Flush encodings after 4 (2 bits + // each) or when the code-string is consumed. + if (code_no == 4 || (n == 0 && *s == 0)) + { + char *codepos = compressed.get() + code_index; + + *codepos = description; + description = 0; + code_no = 0; + + if (n || *s) + { + // If more data to be encoded, then we need a new place to + // store the encodings. + code_index = compressed.length(); + compressed << '\0'; + } + } + } + } + + return compressed; +} + + +// Reverse the effect of htPack. +String +htUnpack(const char format[], const char *data) +{ + const char *s = format; + + // The description needs to be renewed immediately. + unsigned int description = 1; + + // Make a wild guess about that we decompress to some ordinary + // sized struct and assume the cost of allocation some extra + // memory is much less than the cost of allocating more. + // This guess only has speed effects. + String decompressed(60); + + // Format string loop. + while (*s) + { + int fchar = *s++; + int n; + + if (isdigit(*s)) + { + char* t; + n = strtol(s, &t, 10); + s = t; + } + else + n = 1; + + // Loop over N in e.g. "iN" (default 1). + while (n--) + { + // Time to renew description? + if (description == 1) + description = 256 | *data++; + + // Format character handling. + switch (fchar) + { + case 'c': + { + // An unsigned int with the most common value 1 is + // compressed as follows: + // 00 - value is 1. + // 01 - value fits in unsigned char - appended. + // 10 - value fits in unsigned short - appended. + // 11 - just plain unsigned int - appended (you lose). + unsigned int value; + + switch (description & 3) + { + case 0: + value = 1; + break; + + case 1: + { + unsigned char charvalue; + memcpy(&charvalue, data, sizeof charvalue); + value = charvalue; + data++; + } + break; + + case 2: + { + unsigned short int shortvalue; + memcpy(&shortvalue, data, sizeof shortvalue); + value = shortvalue; + data += sizeof shortvalue; + } + break; + + case 3: + { + memcpy(&value, data, sizeof value); + data += sizeof value; + } + break; + } + decompressed.append((char *) &value, sizeof value); + } + break; + + case 'i': + { + // A (signed) int is compressed as follows: + // 00 - value is 0. + // 01 - value fits in char - appended. + // 10 - value fits in short - appended. + // 11 - just plain int - appended (you lose). + int value; + + switch (description & 3) + { + case 0: + value = 0; + break; + + case 1: + { + char charvalue; + memcpy(&charvalue, data, sizeof charvalue); + value = charvalue; + data++; + } + break; + + case 2: + { + short int shortvalue; + memcpy(&shortvalue, data, sizeof shortvalue); + value = shortvalue; + data += sizeof shortvalue; + } + break; + + case 3: + { + memcpy(&value, data, sizeof value); + data += sizeof value; + } + break; + } + decompressed.append((char *) &value, sizeof value); + } + break; + + case 'u': + { + // An unsigned int is compressed as follows: + // 00 - value is 0. + // 01 - value fits in unsigned char - appended. + // 10 - value fits in unsigned short - appended. + // 11 - just plain unsigned int - appended (you lose). + unsigned int value; + + switch (description & 3) + { + case 0: + value = 0; + break; + + case 1: + { + unsigned char charvalue; + memcpy(&charvalue, data, sizeof charvalue); + value = charvalue; + data++; + } + break; + + case 2: + { + unsigned short int shortvalue; + memcpy(&shortvalue, data, sizeof shortvalue); + value = shortvalue; + data += sizeof shortvalue; + } + break; + + case 3: + { + memcpy(&value, data, sizeof value); + data += sizeof value; + } + break; + } + decompressed.append((char *) &value, sizeof value); + } + break; + + default: +#ifndef NOSTREAM +#ifdef DEBUG + if (1) + cerr << "Invalid char \'" << char(fchar) + << "\' in unpack format \"" << format << "\"" + << endl; + return ""; +#endif +#endif + ; // Must always have a statement after a label. + } + + description >>= 2; + } + } + + return decompressed; +} + +// End of HtPack.cc diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtPack.h b/debian/htdig/htdig-3.2.0b6/htlib/HtPack.h new file mode 100644 index 00000000..3da6fa9a --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtPack.h @@ -0,0 +1,39 @@ +// +// HtPack.h +// +// HtPack: Compress and uncompress data in e.g. simple structures. +// The structure must have the layout defined in the ABI; +// the layout the compiler generates. +// +// Much like the pack()/unpack() function pair in perl, but +// compressing, not "packing into a binary structure". +// +// Note that the contents of the returned "String" is not +// necessarily aligned to allow using it as a struct. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtPack.h,v 1.7 2004/05/28 13:15:20 lha Exp $ +// + +#ifndef __HtPack_h +#define __HtPack_h + +#include "htString.h" + +// Pack. +// The parameter "format" is not const but should normally be. +extern String +htPack(const char format[], const char *theStruct); + +// Unpack. +// The parameter "theString" will be updated to point after the +// processed amount of data. +extern String +htUnpack(const char format[], const char *thePackedData); + +#endif // __HtPack_h diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRandom.h b/debian/htdig/htdig-3.2.0b6/htlib/HtRandom.h new file mode 100644 index 00000000..46c183d4 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRandom.h @@ -0,0 +1,49 @@ +// HtRandom.h +// +// class HtRandom: +// tools for random numbers +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRandom.h,v 1.5 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _HtRandom_h_ +#define _HtRandom_h_ + + +class HtRandom +{ + public: + // produce a random unsigned int between v0 and v1 + static inline unsigned int rnd(unsigned int v0,unsigned int v1) + { + return((rand()%(v1-v0)) + v0 ); + } + + // randomly mix up an array of int's + static int *randomize_v(int *vals,int n) + { + int i; + if(!vals) + { + vals=new int[n]; + for(i=0;i<n;i++){vals[i]=i;} + } + for(i=0;i<2*n;i++) + { + int i0=HtRandom::rnd(0,n); + int i1=HtRandom::rnd(0,n); + int t=vals[i0]; + vals[i0]=vals[i1]; + vals[i1]=t; + } + return(vals); + } + +}; +#endif // _HtRandom_h_ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegex.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtRegex.cc new file mode 100644 index 00000000..a0f74038 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegex.cc @@ -0,0 +1,105 @@ +// +// HtRegex.cc +// +// HtRegex: A simple C++ wrapper class for the system regex routines. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegex.cc,v 1.13 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtRegex.h" +#include <locale.h> + + +HtRegex::HtRegex() : compiled(0) { } + +HtRegex::HtRegex(const char *str, int case_sensitive) : compiled(0) +{ + set(str, case_sensitive); +} + +HtRegex::~HtRegex() +{ + if (compiled != 0) regfree(&re); + compiled = 0; +} + +const String &HtRegex::lastError() +{ + return lastErrorMessage; +} + +int +HtRegex::set(const char * str, int case_sensitive) +{ + if (compiled != 0) regfree(&re); + + int err; + compiled = 0; + if (str == NULL) return 0; + if (strlen(str) <= 0) return 0; + if (err = regcomp(&re, str, case_sensitive ? REG_EXTENDED : (REG_EXTENDED|REG_ICASE)), err == 0) + { + compiled = 1; + } + else + { + size_t len = regerror(err, &re, 0, 0); + char *buf = new char[len]; + regerror(err, &re, buf, len); + lastErrorMessage = buf; + delete buf; + } + return compiled; +} + +int +HtRegex::setEscaped(StringList &list, int case_sensitive) +{ + String *str; + String transformedLimits; + list.Start_Get(); + while ((str = (String *) list.Get_Next())) + { + if (str->indexOf('[') == 0 && str->lastIndexOf(']') == str->length()-1) + { + transformedLimits << str->sub(1,str->length()-2).get(); + } + else // Backquote any regex special characters + { + for (int pos = 0; pos < str->length(); pos++) + { + if (strchr("^.[$()|*+?{\\", str->Nth(pos))) + transformedLimits << '\\'; + transformedLimits << str->Nth(pos); + } + } + transformedLimits << "|"; + } + transformedLimits.chop(1); + + return set(transformedLimits, case_sensitive); +} + +int +HtRegex::match(const char * str, int nullpattern, int nullstr) +{ + int rval; + + if (compiled == 0) return(nullpattern); + if (str == NULL) return(nullstr); + if (strlen(str) <= 0) return(nullstr); + rval = regexec(&re, str, (size_t) 0, NULL, 0); + if (rval == 0) return(1); + else return(0); +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegex.h b/debian/htdig/htdig-3.2.0b6/htlib/HtRegex.h new file mode 100644 index 00000000..bb5c60d8 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegex.h @@ -0,0 +1,86 @@ +// +// HtRegex.h +// +// HtRegex: A simple C++ wrapper class for the system regex routines. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegex.h,v 1.13 2004/05/28 13:15:21 lha Exp $ +// +// + +#ifndef _HtRegex_h_ +#define _HtRegex_h_ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "Object.h" +#include "StringList.h" + +// This is an attempt to get around compatibility problems +// with the included regex + +#ifdef _MSC_VER /* _WIN32 */ +#include "regex_win32.h" +#else +# ifdef USE_RX +# include <rxposix.h> +# else // Use regex +# ifdef HAVE_BROKEN_REGEX +# include <regex.h> +# else // include regex code and header +# include "gregex.h" +# endif +# endif +#endif /* _WIN32 */ + +#include <sys/types.h> + +#ifdef HAVE_STD +#include <fstream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <fstream.h> +#endif /* HAVE_STD */ + +class HtRegex : public Object +{ +public: + // + // Construction/Destruction + // + HtRegex(); + HtRegex(const char *str, int case_sensitive = 0); + virtual ~HtRegex(); + + // + // Methods for setting the pattern + // + int set(const String& str, int case_sensitive = 0) { return set(str.get(), case_sensitive); } + int set(const char *str, int case_sensitive = 0); + int setEscaped(StringList &list, int case_sensitive = 0); + + virtual const String &lastError(); // returns the last error message + + // + // Methods for checking a match + // + int match(const String& str, int nullmatch, int nullstr) { return match(str.get(), nullmatch, nullstr); } + int match(const char *str, int nullmatch, int nullstr); + +protected: + int compiled; + regex_t re; + + String lastErrorMessage; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegexList.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexList.cc new file mode 100644 index 00000000..2e017627 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexList.cc @@ -0,0 +1,137 @@ +// +// HtRegexList.cc +// +// HtRegex: A list of HtRegex objects for handling large regex patterns +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegexList.cc,v 1.5 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtRegexList.h" +#include <locale.h> + +class listnode +{ +public: + listnode *next; + Object *object; +}; + +HtRegexList::HtRegexList() +{ + compiled = 0; +} + +HtRegexList::~HtRegexList() +{ + compiled = 0; +} + +const String &HtRegexList::lastError() +{ + return lastErrorMessage; +} + +int +HtRegexList::setEscaped(StringList &list, int case_sensitive) +{ + if (list.Count() == 0) + { + compiled = false; + return true; + } + + String *str; + String transformedLimits, currentPattern, prevPattern; + HtRegex *limit = new HtRegex; + + list.Start_Get(); + while ((str = (String *) list.Get_Next())) + { + if (str->indexOf('[') == 0 && str->lastIndexOf(']') == str->length()-1) + { + transformedLimits = str->sub(1,str->length()-2).get(); + } + else // Backquote any regex special characters + { + transformedLimits = 0; + for (int pos = 0; pos < str->length(); pos++) + { + if (strchr("^.[$()|*+?{\\", str->Nth(pos))) + transformedLimits << '\\'; + transformedLimits << str->Nth(pos); + } + } + if (!currentPattern.empty()) + currentPattern << "|"; + currentPattern << transformedLimits; + if (!limit->set(currentPattern.get(), case_sensitive)) + { + if (prevPattern.empty()) // we haven't set anything yet! + { + lastErrorMessage = limit->lastError(); + compiled = 0; + return false; + } + limit->set(prevPattern.get(), case_sensitive); // Go back a step + Add(limit); + limit = new HtRegex; + currentPattern = transformedLimits; + if (!limit->set(currentPattern.get(), case_sensitive)) + { + lastErrorMessage = limit->lastError(); + compiled = 0; + return false; + } + } + prevPattern = currentPattern; + } + Add(limit); // OK, we're done so just add the last compiled pattern + + compiled = 1; + return true; +} + +int +HtRegexList::match(const char * str, int nullpattern, int nullstr) +{ + HtRegex *regx; + + if (compiled == 0) return(nullpattern); + if (str == NULL) return(nullstr); + if (strlen(str) <= 0) return(nullstr); + + if (number == 0) return(1); // An empty pattern matches everything + + Start_Get(); + while ((regx = (HtRegex *) Get_Next())) + { + if (regx->match(str, nullpattern, nullstr)) + { + // Move this one to the front and update pointers + if (cursor.current_index != -1) + { + if (cursor.prev) + cursor.prev->next = cursor.current->next; + cursor.prev = 0; + cursor.current->next = head; + head = cursor.current; + cursor.current = head; + cursor.current_index = -1; + } + return(1); + } + } + + return(0); +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegexList.h b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexList.h new file mode 100644 index 00000000..bbc0ddf8 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexList.h @@ -0,0 +1,57 @@ +// +// HtRegexList.h +// +// HtRegexList: A list of HtRegex objects for handling large regex patterns +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegexList.h,v 1.4 2004/05/28 13:15:21 lha Exp $ +// +// + +#ifndef _HtRegexList_h_ +#define _HtRegexList_h_ + +#include "Object.h" +#include "List.h" +#include "StringList.h" +#include "HtRegex.h" + +class HtRegexList : public List +{ +public: + // + // Construction/Destruction + // + HtRegexList(); + virtual ~HtRegexList(); + + // + // Setting (construct from a list of patterns) + // + int setEscaped(StringList &list, int case_sensitive = 0); + + virtual const String &lastError(); // returns the last error message + + // + // Methods for checking a match + // + int match(const String& str, int nullmatch, int nullstr) + { return match(str.get(), nullmatch, nullstr); } + int match(const char *str, int nullmatch, int nullstr); + +protected: + int compiled; + + String lastErrorMessage; + +private: +}; + +#endif + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplace.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplace.cc new file mode 100644 index 00000000..eb409e80 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplace.cc @@ -0,0 +1,141 @@ +// +// HtRegexReplace.cc +// +// HtRegexReplace: A subclass of HtRegex that can perform replacements +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 2000-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegexReplace.cc,v 1.4 2004/05/28 13:15:21 lha Exp $ +// + +#include "HtRegexReplace.h" +#include <locale.h> + + +HtRegexReplace::HtRegexReplace() +{ +} + +HtRegexReplace::HtRegexReplace(const char *from, const char *to, int case_sensitive) + : HtRegex(from, case_sensitive) +{ + memset(®s, 0, sizeof(regs)); + repBuf = 0; + segSize = + segUsed = 0; + segMark = 0; + repLen = 0; + + setReplace(to); +} + +HtRegexReplace::~HtRegexReplace() +{ + empty(); +} + +int HtRegexReplace::replace(String &str, int nullpattern, int nullstr) +{ + const int regCount = sizeof(regs) / sizeof(regs[0]); + if (compiled == 0 || repBuf == 0) return nullpattern; + if (str.length() == 0) return nullstr; + + if (regexec(&re, str.get(), regCount, regs, 0) == 0) + { + // Firstly work out how long the result string will be. We think this will be more effecient + // than letting the buffer grow in stages as we build the result, but who knows? + //cout << "!!! Match !!!" << endl; + size_t resLen = repLen; + int i, reg, repPos; + const char *src = str.get(); + + for (i = 1; i < (int) segUsed; i += 2) + { + reg = segMark[i]; + if (reg < regCount && regs[reg].rm_so != -1) + resLen += regs[reg].rm_eo - regs[reg].rm_so; + } + //cout << "result will be " << resLen << " chars long" << endl; + String result(resLen); // Make the result string preallocating the buffer size + for (i = 0, repPos = 0;; ) + { + //cout << "appending segment " << i << endl; + result.append(repBuf + repPos, segMark[i] - repPos); // part of the replace string + repPos = segMark[i]; // move forward + if (++i == (int) segUsed) break; // was that the last segment? + reg = segMark[i++]; // get the register number + if (reg < regCount && regs[reg].rm_so != -1) + result.append((char *) src + regs[reg].rm_so, regs[reg].rm_eo - regs[reg].rm_so); + } + str = result; + //cout << "return " << result.get() << endl; + + return 1; + } + + return 0; +} + +// Private: place a mark in the mark buffer growing it if necessary. +void HtRegexReplace::putMark(int n) +{ + // assert(segUsed <= segSize); + if (segUsed == segSize) + { + size_t newSize = segSize * 2 + 5; // grow in chunks + int *newMark = new int[newSize]; // do we assume that new can't fail? + memcpy(newMark, segMark, segSize * sizeof(int)); + delete segMark; + segMark = newMark; + segSize = newSize; + } + segMark[segUsed++] = n; +} + +void HtRegexReplace::empty() +{ + // Destroy any existing replace pattern + delete repBuf; repBuf = 0; + segSize = segUsed = 0; + delete segMark; segMark = 0; + repLen = 0; +} + +void HtRegexReplace::setReplace(const char *to) +{ + empty(); + + repBuf = new char[strlen(to)]; // replace buffer can never contain more text than to string + int bufPos = 0; // our position within the output buffer + + while (*to) + { + if (*to == '\\') + { + if (*++to == '\0') break; + if (*to >= '0' && *to <= '9') + { + putMark(bufPos); + putMark(*to - '0'); + } + else + { + // We could handle some C style escapes here, but instead we just pass the character + // after the backslash through. This means that \\, \" and \' will do the right thing. + // It's unlikely that anyone will need any C style escapes in ht://Dig anyway. + repBuf[bufPos++] = *to; + } + to++; + } + else + { + repBuf[bufPos++] = *to++; + } + } + putMark(bufPos); + repLen = (size_t) bufPos; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplace.h b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplace.h new file mode 100644 index 00000000..6cf4d50f --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplace.h @@ -0,0 +1,58 @@ +// +// HtRegexReplace.h +// +// HtRegexReplace: A subclass of HtRegex that can perform replacements +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 2000-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegexReplace.h,v 1.4 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _HtRegexReplace_h_ +#define _HtRegexReplace_h_ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtRegex.h" + +class HtRegexReplace : public HtRegex +{ +public: + // + // Construction/Destruction + // + HtRegexReplace(); + HtRegexReplace(const char *from, const char *to, int case_sensitive = 0); + virtual ~HtRegexReplace(); + + // + // Methods for setting the replacement pattern + // + void setReplace(const String& str) { setReplace(str.get()); } + void setReplace(const char *str); + + // + // Methods for replacing + // + int replace(String &str, int nullpattern = 0, int nullstr = 0); + +protected: + char *repBuf; // Replace text. + size_t segSize, segUsed; + int *segMark; + size_t repLen; + + regmatch_t regs[10]; + + // Various private methods + void putMark(int n); + void empty(); +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplaceList.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplaceList.cc new file mode 100644 index 00000000..2ba7a4d2 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplaceList.cc @@ -0,0 +1,84 @@ +// +// HtRegexReplaceList.cc +// +// HtRegexReplaceList: Perform RegexReplace on a list of from/to pairs. +// Patterns are applied in order; pattern matching +// doesn't stop when a match occurs. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 2000-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegexReplaceList.cc,v 1.5 2004/05/28 13:15:21 lha Exp $ +// +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtRegexReplaceList.h" + +#ifdef HAVE_STD +#include <iostream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <iostream.h> +#endif /* HAVE_STD */ + +HtRegexReplaceList::HtRegexReplaceList(StringList &list, int case_sensitive ) +{ + if (list.Count() & 1) + { + lastErrorMessage = "HtRegexReplaceList needs an even number of strings"; + return; + } + + int i; + String err; + + for (i = 0; i < list.Count(); i += 2) + { + String from = list[i]; + String to = list[i+1]; + HtRegexReplace *replacer = new HtRegexReplace(from.get(), to.get(), case_sensitive); + replacers.Add(replacer); // Stash it even if there's an error so it will get destroyed later + const String &err = replacer->lastError(); + if (err.length() != 0) + { + lastErrorMessage = err; + return; + } + } +} + +HtRegexReplaceList::~HtRegexReplaceList() +{ + // replacers gets chucked away +} + +int HtRegexReplaceList::replace(String &str, int nullpattern , int nullstr ) +{ + int repCount = replacers.Count(); + int doneCount = 0; + + for (int rep = 0; rep < repCount; rep++) + { + HtRegexReplace *replacer = (HtRegexReplace *) replacers[rep]; + if (replacer->replace(str, nullpattern, nullstr) > 0) + doneCount++; + } + + return doneCount; +} + +const String &HtRegexReplaceList::lastError() +{ + return lastErrorMessage; +} + +// End of HtRegexReplaceList.cc diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplaceList.h b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplaceList.h new file mode 100644 index 00000000..b51cf2d1 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtRegexReplaceList.h @@ -0,0 +1,39 @@ +// +// HtRegexReplaceList.h +// +// HtRegexReplaceList: Perform RegexReplace on a list of from/to pairs. +// Patterns are applied in order; pattern matching +// doesn't stop when a match occurs. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 2000-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtRegexReplaceList.h,v 1.4 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef __HtRegexReplaceList_h +#define __HtRegexReplaceList_h + +#include "HtRegexReplace.h" +#include "List.h" +#include "StringList.h" + +class HtRegexReplaceList : public Object +{ +public: + // Construct a HtRegexReplaceList. |list| should contain an even + // number of strings that constitute from/to pairs. + HtRegexReplaceList(StringList &list, int case_sensitive = 0); + virtual ~HtRegexReplaceList(); + int replace(String &str, int nullpattern = 0, int nullstr = 0); + virtual const String &lastError(); + +private: + List replacers; + String lastErrorMessage; +}; + +#endif /* __HtRegexReplaceList_h */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtTime.h b/debian/htdig/htdig-3.2.0b6/htlib/HtTime.h new file mode 100644 index 00000000..0f6e5ae7 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtTime.h @@ -0,0 +1,128 @@ +// HtTime.h +// +// class HtTime: +// tools for timing +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtTime.h,v 1.8 2004/05/28 13:15:21 lha Exp $ +// +#ifndef _HtTime_h_ +#define _HtTime_h_ + +#if TIME_WITH_SYS_TIME +#include <sys/time.h> +#include <time.h> +#else +# if HAVE_SYS_TIME_H +# include <sys/time.h> +# else +# include <time.h> +# endif +#endif + +#ifdef _MSC_VER /* _WIN32 */ +#include <sys/timeb.h> +#endif + +class HtTime +{ + public: + // time in seconds (double format) + static inline double DTime() + { +#ifdef _MSC_VER /* _WIN32 */ + struct timeb tb; + ftime(&tb); + return((double)((tb.millitm/1000)+tb.time+tb.timezone)); +#else + struct timeval tv; + gettimeofday(&tv,NULL); + return(tv.tv_usec/1000000.0+tv.tv_sec); +#endif + + } + // time in seconds relative to T0 (double format) + static inline double DTime(double T0) + { +#ifdef _MSC_VER /* _WIN32 */ + struct timeb tb; + ftime(&tb); + return((double)(((tb.millitm/1000)+tb.time+tb.timezone))-T0); +#else + struct timeval tv; + gettimeofday(&tv,NULL); + return((tv.tv_usec/1000000.0+tv.tv_sec)-T0); +#endif + } + + // Do something every x seconds + class Periodic + { + double t0; + double last; + double period; + public: + double total(){return(HtTime::DTime(t0));} + void change_period(double nperiod){period=nperiod;} + int operator()(double *prperiod=NULL) + { + double t=HtTime::DTime(t0); + if(prperiod){*prperiod=t-last;} + if((t-last)>period) + { + last=t; + return(1); + } + return(0); + } + Periodic(double nperiod=.1) + { + period=nperiod; + t0=HtTime::DTime(); + last=0; + } + }; + + + +#ifdef NOTDEF + // print progression message every x seconds + class Progression + { + double t0; + double last; + double period; + char *label; + public: + double total(){return(HtTime::DTime()-t0);} + int operator()(double x) + { + double t=HtTime::DTime()-t0; + if((t-last)>period) + { + last=t; + printf("%s (%f): %f\n",label,t,x); + return(1); + } + return(0); + } + Progression(double nperiod=.1,char *nlabel=(char *)"progression") + { + label=nlabel; + period=nperiod; + t0=HtTime::DTime(); + last=0; + } + }; +#endif +}; +#endif // _HtTime_h_ + + + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVector.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtVector.cc new file mode 100644 index 00000000..b6addea3 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVector.cc @@ -0,0 +1,310 @@ +// +// HtVector.cc +// +// HtVector: A Vector class which holds objects of type Object. +// (A vector is an array that can expand as necessary) +// This class is very similar in interface to the List class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtVector.cc,v 1.11 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtVector.h" + +//********************************************************************* +// void HtVector::HtVector() +// Default constructor +// +HtVector::HtVector() +{ + data = new Object *[4]; // After all, why would anyone want an empty vector? + element_count = 0; + allocated = 4; + current_index = -1; +} + + +//********************************************************************* +// void HtVector::HtVector(int capacity) +// Constructor with known capacity +// (has the side effect of not allocating double memory) +// +HtVector::HtVector(int capacity) +{ + data = new Object *[capacity]; + element_count = 0; + allocated = capacity; + current_index = -1; +} + + +//********************************************************************* +// void HtVector::~HtVector() +// Destructor +// +HtVector::~HtVector() +{ + Destroy(); +} + + +//********************************************************************* +// void HtVector::Release() +// Remove all objects from the vector, but do not delete them +void HtVector::Release() +{ + for (current_index = 0; current_index < element_count; current_index++) + { + data[current_index] = NULL; + } + if (data) + delete [] data; + data = NULL; + allocated = 0; + element_count = 0; + current_index = -1; +} + +//********************************************************************* +// void HtVector::Destroy() +// Deletes all objects from the vector +// +void HtVector::Destroy() +{ + for (current_index = 0; current_index < element_count; current_index++) + if (data[current_index]) + { + delete data[current_index]; + data[current_index] = NULL; + } + if (data) + delete [] data; + data = NULL; + allocated = 0; + element_count = 0; + current_index = -1; +} + + +//********************************************************************* +// void HtVector::Add(Object *object) +// Add an object to the list. +// +void HtVector::Add(Object *object) +{ + Allocate(element_count+1); + data[element_count] = object; + element_count += 1; +} + + +//********************************************************************* +// void HtVector::Insert(Object *object, int position) +// Add an object into the list. +// +void HtVector::Insert(Object *object, int position) +{ + if (position < 0) return; + if (position >= element_count) + { + Add(object); + return; + } + + Allocate(element_count + 1); + for (int i = element_count; i > position; i--) + data[i] = data[i-1]; + data[position] = object; + element_count += 1; +} + + +//********************************************************************* +// void HtVector::Assign(Object *object, int position) +// Assign an object to the position +// +void HtVector:: Assign(Object *object, int position) +{ + // Simply perform an insert, followed by a remove! + Insert(object, position); + RemoveFrom(position + 1); + return; +} + + +//********************************************************************* +// int HtVector::Remove(Object *object) +// Remove an object from the list. +// +int HtVector::Remove(Object *object) +{ + return RemoveFrom(Index(object)); +} + + +//********************************************************************* +// int HtVector::RemoveFrom(int position) +// Remove an object from the list. +// +int HtVector::RemoveFrom(int position) +{ + if (position < 0 || position >= element_count) + return NOTOK; + + for (int i = position; i < element_count - 1; i++) + data[i] = data[i+1]; + element_count -= 1; + return OK; +} + + +//********************************************************************* +// Object *HtVector::Get_Next() +// Return the next object in the list. +// +Object *HtVector::Get_Next() +{ + current_index++; + if (current_index >= element_count) + return 0; + return data[current_index]; +} + + +//********************************************************************* +// Object *HtVector::Get_First() +// Return the first object in the list. +// +Object *HtVector::Get_First() +{ + if (!IsEmpty()) + { + current_index = 0; + return data[0]; + } + else + return 0; +} + + +//********************************************************************* +// int HtVector::Index(Object *obj) +// Return the index of an object in the list. +// +int HtVector::Index(Object *obj) +{ + int index = 0; + + while (index < element_count && data[index] != obj) + { + index++; + } + if (index >= element_count) + return -1; + else + return index; +} + + +//********************************************************************* +// Object *HtVector::Next(Object *prev) +// Return the next object in the list. Using this, the list will +// appear as a circular list. +// +Object *HtVector::Next(Object *prev) +{ + current_index = Index(prev); + if (current_index == -1) + return 0; + + current_index++; // We should probably do this with remainders + if (current_index >= element_count) + current_index = 0; + return data[current_index]; +} + + +//********************************************************************* +// Object *HtVector::Previous(Object *next) +// Return the previous object in the vector. Using this, the vector will +// appear as a circular list. +// +Object *HtVector::Previous(Object *next) +{ + current_index = Index(next); + if (current_index == -1) + return 0; + + current_index--; // We should probably do this with remainders + if (current_index < 0) + current_index = element_count - 1; + return data[current_index]; +} + + +//********************************************************************* +// Object *HtVector::Copy() const +// Return a deep copy of the vector. +// +Object *HtVector::Copy() const +{ + HtVector *vector = new HtVector(allocated); + + for(int i = 0; i < Count(); i++) + vector->Add(data[i]->Copy()); + + return vector; +} + + +//********************************************************************* +// HtVector &HtVector::operator=(HtVector &vector) +// Return a deep copy of the list. +// +HtVector &HtVector::operator=(HtVector &vector) +{ + Destroy(); + + for(int i = 0; i < vector.Count(); i++) + Add(vector.data[i]->Copy()); + + return *this; +} + + +//********************************************************************* +// int Allocate(int capacity) +// Ensure there is at least capacity space in the vector +// +void HtVector::Allocate(int capacity) +{ + if (capacity > allocated) // Darn, we actually have to do work :-) + { + Object **old_data = data; + + // Ensure we have more than the capacity and we aren't + // always rebuilding the vector (which leads to quadratic behavior) + while (allocated < capacity) + allocated *= 2; + + data = new Object *[allocated]; + + for (int i = 0; i < element_count; i++) + { + data[i] = old_data[i]; + old_data[i] = NULL; + } + + if (old_data) + delete [] old_data; + } +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVector.h b/debian/htdig/htdig-3.2.0b6/htlib/HtVector.h new file mode 100644 index 00000000..af1271b6 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVector.h @@ -0,0 +1,137 @@ +// +// HtVector.h +// +// HtVector: A Vector class which holds objects of type Object. +// (A vector is an array that can expand as necessary) +// This class is very similar in interface to the List class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtVector.h,v 1.10 2004/05/28 13:15:21 lha Exp $ +// +// +#ifndef _HtVector_h_ +#define _HtVector_h_ +#include "Object.h" + +class HtVector : public Object +{ +public: + // + // Constructor/Destructor + // + HtVector(); + HtVector(int capacity); + ~HtVector(); + + // + // Add() will append an Object to the end of the vector + // + void Add(Object *); + + // + // Insert() will insert an object at the given position. If the + // position is larger than the number of objects in the vector, the + // object is appended; no new objects are created between the end + // of the vector and the given position. + // + void Insert(Object *, int position); + + // + // Assign() will assign the object to the given position, replacing + // the object currently there. It is functionally equivalent to calling + // RemoveFrom() followed by Insert() + void Assign(Object *, int position); + + // + // Find the given object in the vector and remove it from the vector. + // The object will NOT be deleted. If the object is not found, + // NOTOK will be returned, else OK. + // + int Remove(Object *); + + // + // Remove the object at the given position + // (in some sense, the inverse of Insert) + // + int RemoveFrom(int position); + + // + // Release() will remove all the objects from the vector. + // This will NOT delete them + void Release(); + + // + // Destroy() will delete all the objects in the vector. This is + // equivalent to calling the destructor + // + void Destroy(); + + // + // Vector traversel (a bit redundant since you can use []) + // + void Start_Get() {current_index = -1;} + Object *Get_Next(); + Object *Get_First(); + Object *Next(Object *current); + Object *Previous(Object *current); + Object *Last() {return element_count<=0?(Object *)NULL:data[element_count-1];} + + // + // Direct access to vector items. To assign new objects, use + // Insert() or Add() or Assign() + // + Object *operator[] (int n) {return (n<0||n>=element_count)?(Object *)NULL:data[n];} + Object *Nth(int n) {return (n<0||n>=element_count)?(Object *)NULL:data[n];} + + // + // Access to the number of elements + // + int Count() const {return element_count;} + int IsEmpty() {return element_count==0;} + + // + // Get the index number of an object. If the object is not found, + // returns -1 + // + int Index(Object *); + + // + // Deep copy member function + // + Object *Copy() const; + + // + // Vector Assignment + // + HtVector &operator= (HtVector *vector) {return *this = *vector;} + HtVector &operator= (HtVector &vector); + +protected: + // + // The actual internal data array + Object **data; + + // + // For traversal it is nice to know where we are... + // + int current_index; + + // + // It's nice to keep track of how many things we contain... + // as well as how many slots we've declared + // + int element_count; + int allocated; + + // + // Protected function to ensure capacity + // + void Allocate(int ensureCapacity); +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGeneric.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGeneric.cc new file mode 100644 index 00000000..b3583e0c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGeneric.cc @@ -0,0 +1,91 @@ +// +// HtVectorGeneric.cc +// +// HtVectorGeneric: A Vector class which holds objects of type GType. +// (A vector is an array that can expand as necessary) +// This class is very similar in interface to the List class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtVectorGeneric.cc,v 1.5 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtVector_int.h" +#define GType int +#define HtVectorGType HtVector_int +#include "HtVectorGenericCode.h" + + +#define GType char +#define HtVectorGType HtVector_char +#include "HtVectorGeneric.h" +#define GType char +#define HtVectorGType HtVector_char +#include "HtVectorGenericCode.h" + +#include"HtVector_String.h" +#define GType String +#define HtVectorGType HtVector_String +#include "HtVectorGenericCode.h" + + +#define GType double +#define HtVectorGType HtVector_double +#include "HtVectorGeneric.h" +#define GType double +#define HtVectorGType HtVector_double +#include "HtVectorGenericCode.h" + +#include<stdio.h> + +// this is just to check if it compiles ok + +class ZOZO +{ + int a,b,c; +public: + void show(){printf("ZOZO SHOW:%d %d %d\n",a,b,c);} + ZOZO() + { + a=1; + b=2; + c=3; + } +}; + + +#define HTVECTORGENERIC_NOTCOMPARABLE +#define GType ZOZO +#define HtVectorGType HtVector_ZOZO +#include "HtVectorGeneric.h" +#define HTVECTORGENERIC_NOTCOMPARABLE +#define GType ZOZO +#define HtVectorGType HtVector_ZOZO +#include "HtVectorGenericCode.h" + +void +test_HtVectorGeneric() +{ + HtVector_int intv; + HtVector_char charv; + HtVector_ZOZO vz; + ZOZO zz; + vz.push_back(zz); + vz.push_back(zz); + vz.push_back(zz); + + int i; + for(i=0;i<vz.size();i++) + { + vz[i].show(); + } +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGeneric.h b/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGeneric.h new file mode 100644 index 00000000..8299aa2c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGeneric.h @@ -0,0 +1,247 @@ +// +// HtVectorGeneric.h +// +// HtVector: A Vector class which holds objects of type GType. +// (A vector is an array that can expand as necessary) +// This class is very similar in interface to the List class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtVectorGeneric.h,v 1.5 2004/05/28 13:15:21 lha Exp $ +// +// +// #ifndef _HtVectorGeneric_h_ +//#define _HtVectorGeneric_h_ + + +// HOWTO use: +// let's say you have a class called Zozo (that has to have an +// empty constructor Zozo::Zozo() ) +// You want to use vectors of Zozo's (called HtVector_Zozo ) +// +// In every file you use it you must include its declaration: +// +// #define GType Zozo +// #define HtVectorGType HtVector_Zozo +// #include "HtVectorGeneric.h" +// +// Then you can use it: +// HtVector_Zozo vect; +// Zozo a,b,c; +// vect.push_back(a); +// vect.push_back(b); +// vect[0]=c; +// +// Somewhere in some .cc file you have to +// +// #define GType Zozo +// #define HtVectorGType HtVector_Zozo +// #include "HtVectorGenericCode.h" +// +// +// Notes: +// *If you include the declaration more +// than once in the same scope you will get +// errors (HtVector_Zozo : already declared) +// +// *Bounds checking is on by default, this should be +// changed (it's performance critical), if your'e +// out of bounds an error is printed on cerr +// but the prg doesnt stop (maybe it should?) +// +// *vectors containing pointers DO NOT FREE +// them, this is deliberate +// + +#include "Object.h" + +class HtVectorGType : public Object +{ +public: + // + // Constructor/Destructor + // + HtVectorGType(); + HtVectorGType(int capacity); + ~HtVectorGType(); + +protected: + // + // this error checking should be made optional! + // + inline void CheckBounds(const int n) const {if(n<0 || n>=element_count){ fprintf(stderr, "HtVectorGType::CheckBounds: out of bounds.\n");}} + +public: + + + // + // Insert() will insert an object at the given position. If the + // position is larger than the number of objects in the vector, the + // object is appended; no new objects are created between the end + // of the vector and the given position. + // + void Insert(const GType &, int position); + + // *** this is obsolete in HtVectorGeneric ** use: vector[position]=value; +// void Assign(GType , int position); + + + // + // Remove the object at the given position + // (in some sense, the inverse of Insert) + // + void RemoveFrom(int position); + + // Release is obsolete since no deletions of pointers contained in + // the vector is done + // void Release(); + + // + // Destroy() will delete all the objects in the vector. + // Warning: no deletions of pointers contained in + // the vector are done (that's up to you) + // + void Destroy(); + + // + // Vector traversal (a bit redundant since you can use []) + // + void Start_Get() {current_index = -1;} + GType & Get_Next(); + GType & Get_First(); + GType & Last() {return Nth(element_count-1);} + + // + // Direct access to vector items. To assign new objects, use + // Insert() or Add() or Assign() + // + inline GType & Nth(int n) + { +#ifdef HtVectorGeneric_CheckBounds + CheckBounds(n); +#endif + return data[n]; + } + inline const GType & Nth(int n) const + { +#ifdef HtVectorGeneric_CheckBounds + CheckBounds(n); +#endif + return data[n]; + } + inline GType & operator[] (int n) {return Nth(n);} + inline const GType & operator[] (int n) const {return Nth(n);} + + // + // Access to the number of elements + // + inline int Count() const {return element_count;} + inline int IsEmpty() {return element_count==0;} + + + // + // Accesses wich involve finding an element (only possible if + // we can compare two elements) + // +#ifndef HTVECTORGENERIC_NOTCOMPARABLE + // + // Get the index number of an object. If the object is not found, + // returns -1 + // + int Index(const GType &); + GType & Next(const GType ¤t); + GType & Previous(const GType ¤t); + + // + // Find the given object in the vector and remove it from the vector. + // The object will NOT be deleted. If the object is not found, + // NOTOK will be returned, else OK. + // + void Remove(const GType &); +#endif + + // + // Deep copy member function + // + Object *Copy() const; + + // + // Vector Assignment + // + HtVectorGType &operator= (HtVectorGType *vector) {return *this = *vector;} + HtVectorGType &operator= (const HtVectorGType &vector); + + +protected: + // + // The actual internal data array + GType *data; + + // + // For traversal it is nice to know where we are... + // + int current_index; + + // + // It's nice to keep track of how many things we contain... + // as well as how many slots we've declared + // + int element_count; + int allocated; + protected: + // + // Protected function to ensure capacity + // + inline void Allocate(int capacity) + { + if (capacity > allocated){ActuallyAllocate(capacity);} + } + void ActuallyAllocate(int); + + public: + // + // Add() will append an object to the end of the vector + // + inline void Add(const GType &object) + { + Allocate(element_count+1); + data[element_count] = object; + element_count += 1; + } + + + // + // STL like accesors + // + public: + inline int size() const {return Count();} + inline void push_back(const GType &v) {Add( v);} + + inline GType * begin() {return(data);} + inline const GType * begin() const {return(data);} + inline GType * end() {return(data+element_count);} + inline const GType * end() const {return(data+element_count);} + + inline GType & back() {return Nth(element_count-1);} + inline const GType & back() const {return Nth(element_count-1);} + inline void pop_back() {RemoveFrom(size()-1);} + inline void clear() {;} + + void reserve (int n) {Allocate(n);} + +// TODO: erase clear resize insert(...) and many others + +}; + +// #endif + + +#ifdef HTVECTORGENERIC_NOTCOMPARABLE +#undef HTVECTORGENERIC_NOTCOMPARABLE +#endif +#undef HtVectorGType +#undef GType diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGenericCode.h b/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGenericCode.h new file mode 100644 index 00000000..ac809f3b --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVectorGenericCode.h @@ -0,0 +1,262 @@ +// +// HtVectorGenericCode.h +// +// HtVectorGeneric: A Vector class which holds objects of type GType. +// (A vector is an array that can expand as necessary) +// This class is very similar in interface to the List class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtVectorGenericCode.h,v 1.5 2004/05/28 13:15:21 lha Exp $ +// + + +//********************************************************************* +// void HtVectorGType::HtVectorGType() +// Default constructor +// +HtVectorGType::HtVectorGType() +{ + data = new GType[4]; // After all, why would anyone want an empty vector? + element_count = 0; + allocated = 4; + current_index = -1; +} + + +//********************************************************************* +// void HtVectorGType::HtVectorGType(int capacity) +// Constructor with known capacity +// (has the side effect of not allocating double memory) +// +HtVectorGType::HtVectorGType(int capacity) +{ + data = new GType[capacity]; + element_count = 0; + allocated = capacity; + current_index = -1; +} + + +//********************************************************************* +// void HtVectorGType::~HtVectorGType() +// Destructor +// +HtVectorGType::~HtVectorGType() +{ + Destroy(); +} + + + +//********************************************************************* +// void HtVectorGType::Destroy() +// Deletes all objects from the vector +// +void HtVectorGType::Destroy() +{ + if (data) + delete [] data; + data = NULL; + allocated = 0; + element_count = 0; + current_index = -1; +} + + + +//********************************************************************* +// void HtVectorGType::Insert(GType object, int position) +// Add an object into the list. +// +void HtVectorGType::Insert(const GType &object, int position) +{ + if (position < 0) {CheckBounds(position);} + if (position >= element_count) + { + Add(object); + return; + } + + Allocate(element_count + 1); + for (int i = element_count; i > position; i--) + data[i] = data[i-1]; + data[position] = object; + element_count += 1; +} + + +//********************************************************************* +// int HtVectorGType::RemoveFrom(int position) +// Remove an object from the list. +// +void HtVectorGType::RemoveFrom(int position) +{ + CheckBounds(position); + + for (int i = position; i < element_count - 1; i++) + { + data[i] = data[i+1]; + } + element_count -= 1; +} + + +//********************************************************************* +// GType HtVectorGType::Get_Next() +// Return the next object in the list. +// +GType &HtVectorGType::Get_Next() +{ + current_index++; + CheckBounds(current_index); + return data[current_index]; +} + + +//********************************************************************* +// GType HtVectorGType::Get_First() +// Return the first object in the list. +// +GType &HtVectorGType::Get_First() +{ + CheckBounds(0); + return data[0]; +} + +#ifndef HTVECTORGENERIC_NOTCOMPARABLE + +//********************************************************************* +// int HtVectorGType::Index(GType obj) +// Return the index of an object in the list. +// +int HtVectorGType::Index(const GType &obj) +{ + int index0 = 0; + + while (index0 < element_count && data[index0] != obj) + { + index0++; + } + if (index0 >= element_count) + return -1; + else + return index0; +} + + +//********************************************************************* +// GType HtVectorGType::Next(GType prev) +// Return the next object in the list. Using this, the list will +// appear as a circular list. +// +GType &HtVectorGType::Next(const GType & prev) +{ + current_index = Index(prev); + CheckBounds(current_index); + + current_index++; // We should probably do this with remainders + return Nth(current_index); +} + +//********************************************************************* +// GType HtVectorGType::Previous(GType next) +// Return the previous object in the vector. Using this, the vector will +// appear as a circular list. +// +GType &HtVectorGType::Previous(const GType & next) +{ + current_index = Index(next); + CheckBounds(current_index); + + current_index--; // We should probably do this with remainders + return Nth(current_index); +} + +//********************************************************************* +// int HtVectorGType::Remove(GType object) +// Remove an object from the list. +// +void HtVectorGType::Remove(const GType &object) +{ + int pos = Index(object); + CheckBounds(pos); + RemoveFrom(pos); +} +#endif + +//********************************************************************* +// HtVectorGType *HtVectorGType::Copy() const +// Return a deep copy of the vector. +// +Object *HtVectorGType::Copy() const +{ + HtVectorGType *vector = new HtVectorGType(allocated); + + for(int i = 0; i < Count(); i++) +{ +#ifdef HTVECTORGENERIC_OBJECTPTRTYPE + vector->Add(data[i]->Copy()); +#else + vector->Add(data[i]); +#endif +} + return vector; +} + + +//********************************************************************* +// HtVectorGType &HtVectorGType::operator=(HtVectorGType &vector) +// Return a deep copy of the list. +// +HtVectorGType &HtVectorGType::operator=(const HtVectorGType &vector) +{ + Destroy(); + + for(int i = 0; i < vector.Count(); i++) + { + Add(vector.data[i]); + } + return *this; +} + + +//********************************************************************* +// int Allocate(int capacity) +// Ensure there is at least capacity space in the vector +// +void HtVectorGType::ActuallyAllocate(int capacity) +{ + if (capacity > allocated) // Darn, we actually have to do work :-) + { + GType *old_data = data; + + // Ensure we have more than the capacity and we aren't + // always rebuilding the vector (which leads to quadratic behavior) + if(!allocated){allocated=1;} + while (allocated < capacity) + allocated *= 2; + + data = new GType[allocated]; + + for (int i = 0; i < element_count; i++) + { + data[i] = old_data[i]; + } + + if (old_data) + delete [] old_data; + } +} + + +#ifdef HTVECTORGENERIC_NOTCOMPARABLE +#undef HTVECTORGENERIC_NOTCOMPARABLE +#endif + +#undef HtVectorGType +#undef GType diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVector_String.h b/debian/htdig/htdig-3.2.0b6/htlib/HtVector_String.h new file mode 100644 index 00000000..5e72e117 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVector_String.h @@ -0,0 +1,23 @@ +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// + + +#ifndef _HtVector_String_h_ +#define _HtVector_String_h_ + + +#include"htString.h" + +#define GType String +#define HtVectorGType HtVector_String +#include "HtVectorGeneric.h" + +#endif + + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtVector_int.h b/debian/htdig/htdig-3.2.0b6/htlib/HtVector_int.h new file mode 100644 index 00000000..f2bf8a9a --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtVector_int.h @@ -0,0 +1,21 @@ +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// + + +#ifndef _HtVector_int_h_ +#define _HtVector_int_h_ + +#define GType int +#define HtVectorGType HtVector_int +#include "HtVectorGeneric.h" + + +#endif + + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtWordCodec.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtWordCodec.cc new file mode 100644 index 00000000..5a4474d4 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtWordCodec.cc @@ -0,0 +1,437 @@ +// +// HtWordCodec.cc +// +// HtWordCodec: Given two lists of pair of "words" 'from' and 'to'; +// simple one-to-one translations, use those lists to translate. +// Only restriction are that no null (0) characters must be +// used in "words", and that there is a character "joiner" that +// does not appear in any word. One-to-one consistency may be +// checked at construction. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtWordCodec.cc,v 1.9 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtWordCodec.h" + +// Do not use 0, so we can use "normal" string routines. +// Values 1..4 are used to describe how many bytes are used to +// keep the number. Do not use other than control-characters, +// as the first character for internal encodings, so the user +// can use "international" characters (128 .. 255) for cute +// encodings to use across different configuration files and +// databases. +#define JOIN_CHAR 5 +#define QUOTE_CHAR 6 +#define FIRST_INTERNAL_SINGLECHAR 7 +#define LAST_INTERNAL_SINGLECHAR 31 + + +HtWordCodec::HtWordCodec() +{ + myFrom = 0; + myTo = 0; + myFromMatch = 0; + myToMatch = 0; +} + + +HtWordCodec::~HtWordCodec() +{ + if (myFrom) + delete myFrom; + + if (myTo) + delete myTo; + + if (myFromMatch) + delete myFromMatch; + + if (myToMatch) + delete myToMatch; +} + + +// Straightforward filling of the encoding-lists. +HtWordCodec::HtWordCodec(StringList *from, StringList *to, char joiner) +{ + myFromMatch = new StringMatch; + myToMatch = new StringMatch; + + myTo = to; + myFrom = from; + + String to_pattern(myTo->Join(joiner)); + + // After being initialized with Join, the strings are not + // null-terminated, but that is done through "operator char*". + myToMatch->Pattern(to_pattern, joiner); + + String from_pattern(myFrom->Join(joiner)); + myFromMatch->Pattern(from_pattern, joiner); + +} + + +// This constructor is the most complicated function in this class. +// It handles consistency checking for the supplied code-lists. + +// Cleanups for anything except myTo, myFrom, myToMatch is +// necessary. The member myFromMatch is used as a sanity check +// for member functions to see that the constructor was +// successful in case the programmer forgets to check errmsg. +HtWordCodec::HtWordCodec(StringList &requested_encodings, + StringList &frequent_substrings, + String &errmsg) +{ + if ((requested_encodings.Count() % 2) != 0) + { + errmsg = + "Expected pairs, got odd number of strings"; + + return; + } + + myFrom = new StringList; + myTo = new StringList; + + // Go through requested_encodings and fill myTo and myFrom. + // Check that the "to" strings look remotely sane regarding + // reserved characters. + + // Iteration temporaries. + String *from; + String *to; + + int n_of_pairs = requested_encodings.Count() / 2; + + requested_encodings.Start_Get(); + while ((from = (String *) requested_encodings.Get_Next()) != NULL) + { + // Sanity check: Reserve empty strings as we cannot do + // anything sane with them. + + int templen = from->length(); + if (templen == 0) + { + errmsg = "Empty strings are not allowed"; + return; + } + + myFrom->Add(new String(*from)); + + // This must be non-null since we checked "oddness" above. + to = (String *) requested_encodings.Get_Next(); + + templen = to->length(); + if (templen == 0) + { + errmsg = "Empty strings are not allowed"; + return; + } + + // We just have to check that there's no JOIN_CHAR in the + // string. Since no "to" is allowed to be part of any other + // "to", there will be no ambiguity, even if one would + // contain a QUOTE_CHAR (which is documented as invalid anyway). + if (strchr(from->get(), JOIN_CHAR) != NULL) + { + errmsg = + form("(\"%s\" =>) \"%s\" contains a reserved character (number %d)", + from->get(), to->get(), int(JOIN_CHAR)); + return; + } + + // Loop over the other "to"-strings and check that this + // string is not a substring of any other "to", or vice versa. + // Return in error if it is so. + int i; + int count = myTo->Count(); + for (i = 0; i < count; i++) + { + String *ith = (String *) myTo->Nth(i); + + // Just check if the shorter string is part of the + // longer string. + if (to->length() < ith->length() + ? ith->indexOf(to->get()) != -1 + : to->indexOf(ith->get()) != -1) + { + errmsg = + form("\"%s\" => \"%s\" collides with (\"%s\" => \"%s\")", + from, to, (*myFrom)[i], ith->get()); + + return; + } + } + + // All ok, just add this one. + myTo->Add(new String(*to)); + } + + // Check that none of the "to"-strings is a substring of any + // of the "from" strings, since that's hard to support and + // most probably is a user mistake anyway. + + StringMatch req_tos; + String req_to_pattern(myTo->Join(JOIN_CHAR)); + int which, length; + + // The StringMatch functions want the strings + // zero-terminated, which is done through "operator char*". + req_tos.Pattern(req_to_pattern, JOIN_CHAR); + + // Check the requested encodings. + if (n_of_pairs != 0) + { + int i; + for (i = 0; i < n_of_pairs; i++) + { + from = (String *) myFrom->Nth(i); + if (req_tos.FindFirst(from->get(), which, length) != -1) + { + if (i != which) + { + errmsg = + form("(\"%s\" => \"%s\") overlaps (\"%s\" => \"%s\")", + (*myFrom)[which], (*myTo)[which], + from->get(), (*myTo)[i]); + } + else + { + errmsg = + form("Overlap in (\"%s\" => \"%s\")", + from->get(), (*myTo)[i]); + } + + return; + } + } + } + + if (frequent_substrings.Count() != 0) + { + // Make a temporary search-pattern of the requested + // from-strings. + + StringMatch req_froms; + String req_from_pattern(myFrom->Join(JOIN_CHAR)); + + req_froms.Pattern(req_from_pattern, JOIN_CHAR); + + // Continue filling "to" and "from" from frequent_substrings and + // internal encodings. If a frequent_substring is found in the + // requested from-strings, it is ignored, but the internal + // encoding is still ticked up, so that changes in + // requested_encodings (e.g. url_part_aliases) do not change + // an existing database (e.g. containing common_url_parts). + + int internal_encoding_no = 0; + + String *common_part; + frequent_substrings.Start_Get(); + String to; + + for (; + (common_part = (String *) frequent_substrings.Get_Next()) != NULL; + internal_encoding_no++) + { + int templen = common_part->length(); + if (templen == 0) + { + errmsg = "Empty strings are not allowed"; + return; + } + + // Is a "From" string in it, or is a "To" string in it? + // Note that checking if there are *any* requested + // encodings (n_of_pairs) is not just an "optimization"; + // it is necessary since StringMatch will return 0 (not + // -1) if the pattern is empty (FIXME: changing that + // breaks something else in another part of ht://Dig). + + if (n_of_pairs + && (req_froms.FindFirst(common_part->get()) != -1 + || req_tos.FindFirst(common_part->get()) != -1)) + continue; + + to = 0; // Clear previous run. + + // Dream up an encoding without zeroes. + // Use FIRST_INTERNAL_SINGLECHAR .. LAST_INTERNAL_SINGLECHAR + // for the first encodings, as much as possible. + + long int number_to_store = + internal_encoding_no + FIRST_INTERNAL_SINGLECHAR; + + if (number_to_store <= LAST_INTERNAL_SINGLECHAR) + { + to << char(number_to_store); + } + else + { + // Use <number-of-bytes-in-length> + // <number-as-nonzero-bytes> to code the rest. + // Note that we assume eight-bit chars here, which + // should be ok for all systems you run htdig on. + // At least it helps clarity here. + + number_to_store -= LAST_INTERNAL_SINGLECHAR; + + // Make sure highest bit in every byte is "1" by + // inserting one there. + char to_store[sizeof(number_to_store)+1]; + int j = 1; + + while (number_to_store > 0x7f) + { + number_to_store = ((number_to_store & ~0x7f) << 1) + | 0x80 | (number_to_store & 0x7f); + + to_store[j++] = char(number_to_store); + number_to_store >>= 8; + } + + // Finally, store the highest byte. It too shall have + // the highest bit set. This is the easiest way to + // adjust it not to be QUOTE_CHAR. + to_store[0] = j; + to_store[j] = char(number_to_store | 0x80); + + to.append(to_store, j+1); + } + + // Add to replacement pairs. + myFrom->Add(new String(*common_part)); + myTo->Add(new String(to)); + } + } + + // Now, add the quoted "to":s to the "to"-list, with the unquoted + // "to":s to the "from"-list. This way we do not have to + // check for quoting separately. Like this: + // From To + // foo : ! + // bar : > + // baz : $ + // ! : \! + // > : \> + // $ : \$ + // + // Since we checked that none of the "To":s are in a "From" we + // can do this. + + myTo->Start_Get(); + int to_count = myTo->Count(); + String *current; + String temp; + + int i; + for (i = 0; i < to_count; i++) + { + // It works to append *and* iterate through a + // StringList, despite not having an iterator class. + current = (String *) myTo->Nth(i); + + myFrom->Add(new String(*current)); + + temp = 0; // Reset any previous round. + temp.append(char(QUOTE_CHAR)); + temp.append(*current); + + myTo->Add(new String(temp)); + } + + myFromMatch = new StringMatch; + myToMatch = new StringMatch; + + String to_pattern(myTo->Join(JOIN_CHAR)); + String from_pattern(myFrom->Join(JOIN_CHAR)); + + // StringMatch class has unchecked limits, better check them. + // The length of each string in the pattern an the upper limit + // of the needs. + if (to_pattern.length() - (myTo->Count() - 1) > 0xffff + || from_pattern.length() - (myFrom->Count() - 1) > 0xffff) + { + errmsg = "Limit reached; use fewer encodings"; + return; + } + + myToMatch->Pattern(to_pattern, JOIN_CHAR); + myFromMatch->Pattern(from_pattern, JOIN_CHAR); + + errmsg = 0; +} + + +// We only need one "coding" function, since quoting and unquoting is +// handled through the to- and from-lists. +String +HtWordCodec::code(const String &orig_string, StringMatch &match, + StringList &replacements) const +{ + String retval; + String tempinput; + int offset, which, length; + const char *orig; + + // Get a null-terminated string, usable for FindFirst to look at. + orig = orig_string.get(); + + // Sanity check. If bad use, just return empty strings. + if (myFromMatch == NULL) + { + return retval; + } + + // Need to check if "replacements" is empty; that is, if no + // transformations should be done. FindFirst() does not return + // -1 in this case, it returns 0. + if (replacements.Count() == 0) + return orig_string; + + // Find the encodings and replace them. + while ((offset = match.FindFirst(orig, which, length)) != -1) + { + // Append the previous part that was not part of a code. + retval.append(orig, offset); + + // Replace with the original string. + retval.append(replacements[which]); + + orig += offset + length; + } + + // Add the final non-matched part. + retval.append(orig); + + return retval; +} + + +// The assymetry is caused by swapping both the matching and +// replacement lists. +String +HtWordCodec::decode(const String &orig) const +{ + return code(orig, *myToMatch, *myFrom); +} + + +String +HtWordCodec::encode(const String &orig) const +{ + return code(orig, *myFromMatch, *myTo); +} + +// End of HtWordCodec.cc diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtWordCodec.h b/debian/htdig/htdig-3.2.0b6/htlib/HtWordCodec.h new file mode 100644 index 00000000..96b752ff --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtWordCodec.h @@ -0,0 +1,70 @@ +// +// HtWordCodec.h +// +// HtWordCodec: Given two lists of pair of "words" 'from' and 'to'; +// simple one-to-one translations, use those lists to translate. +// Only restriction are that no null (0) characters must be +// used in "words", and that there is a character "joiner" that +// does not appear in any word. One-to-one consistency may be +// checked at construction. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtWordCodec.h,v 1.6 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef __HtWordCodec_h +#define __HtWordCodec_h + +#include "HtCodec.h" +#include "StringList.h" +#include "StringMatch.h" + +class HtWordCodec : public HtCodec +{ +public: + HtWordCodec(); + virtual ~HtWordCodec(); + + // Set the lists of asymmetric pairs of "words" in "from" and + // "to", using: + // * one list of requested encodings with two consecutive + // items "to" and "from" per translation + // * one list of just words which HtWordCodec will generate + // space-saving encodings for. + // Either may be empty. + // Items in frequent_substrings will be silently ignored if + // they collide with anything in requested_encoding_pairs. + // CodingError is empty on success, or has a failure message. + HtWordCodec(StringList &requested_encodings, + StringList &frequest_substrings, String &errmsg); + + // *Or*, set the lists directly, without checking coding + // consistency. HtWordCodec will delete these lists when + // destroyed. Not really recommended, but this class would be + // incomplete without it. + HtWordCodec (StringList *from, StringList *to, char joiner = char(1)); + + // Same as those in the parent class. Each string to + // encode/decode may contain zero or more of words from the + // lists. Those words will be replaced. + virtual String encode(const String &uncoded) const; + virtual String decode(const String &coded) const; + +private: + HtWordCodec(const HtWordCodec &); // Not supposed to be implemented. + void operator= (const HtWordCodec &); // Not supposed to be implemented. + StringList *myFrom; + StringList *myTo; + StringMatch *myFromMatch; + StringMatch *myToMatch; + + // Do coding/decoding symmetrically using the provided lookup and lists. + String code(const String &, StringMatch& match, StringList& replacements) const; +}; + +#endif /* __HtWordCodec_h */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtWordType.cc b/debian/htdig/htdig-3.2.0b6/htlib/HtWordType.cc new file mode 100644 index 00000000..cacb9342 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtWordType.cc @@ -0,0 +1,51 @@ +// +// HtWordType.h +// +// functions for determining valid words/characters +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtWordType.cc,v 1.11 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "HtWordType.h" +#include "WordType.h" + +int HtIsWordChar(char c) { return WordType::Instance()->IsChar(c); } +int HtIsStrictWordChar(char c) { return WordType::Instance()->IsStrictChar(c); } +int HtWordNormalize(String &w) { return WordType::Instance()->Normalize(w); } +int HtStripPunctuation(String &w) { return WordType::Instance()->StripPunctuation(w); } + + +// much like strtok(), and destructive of the source string like strtok(), +// but does word separation by our rules. +char * +HtWordToken(char *str) +{ + unsigned char *text = (unsigned char *)str; + char *ret = 0; + static unsigned char *prev = 0; + + if (!text) + text = prev; + while (text && *text && !HtIsStrictWordChar(*text)) + text++; + if (text && *text) + { + ret = (char *)text; + while (*text && HtIsWordChar(*text)) + text++; + if (*text) + *text++ = '\0'; + } + prev = text; + return ret; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/HtWordType.h b/debian/htdig/htdig-3.2.0b6/htlib/HtWordType.h new file mode 100644 index 00000000..ef4f9bf5 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/HtWordType.h @@ -0,0 +1,27 @@ +// +// HtWordType.h +// +// functions for determining valid words/characters +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: HtWordType.h,v 1.12 2004/05/28 13:15:21 lha Exp $ +// +#ifndef _HtWordType_h +#define _HtWordType_h + +#include "htString.h" + +extern int HtIsWordChar(char c); +extern int HtIsStrictWordChar(char c); +extern int HtWordNormalize(String &w); +extern int HtStripPunctuation(String &w); + +// Like strtok(), but using our rules for word separation. +extern char *HtWordToken(char *s); + +#endif /* _HtWordType_h */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/IntObject.cc b/debian/htdig/htdig-3.2.0b6/htlib/IntObject.cc new file mode 100644 index 00000000..0b500edf --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/IntObject.cc @@ -0,0 +1,37 @@ +// +// IntObject.cc +// +// IntObject: int variable encapsulated in Object derived class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: IntObject.cc,v 1.5 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "IntObject.h" + + +//******************************************************************************* +// IntObject::IntObject() +// +IntObject::IntObject() +{ +} + + +//******************************************************************************* +// IntObject::~IntObject() +// +IntObject::~IntObject() +{ +} + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/IntObject.h b/debian/htdig/htdig-3.2.0b6/htlib/IntObject.h new file mode 100644 index 00000000..461a1747 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/IntObject.h @@ -0,0 +1,39 @@ +// +// IntObject.h +// +// IntObject: int variable encapsulated in Object derived class +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: IntObject.h,v 1.7 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _IntObject_h_ +#define _IntObject_h_ + +#include "Object.h" + +class IntObject : public Object +{ +public: + // + // Construction/Destruction + // + IntObject(); + IntObject(int v) { value = v; } + ~IntObject(); + + int Value() {return value;} + void Value(int v) {value = v;} + +private: + int value; +}; + +#endif + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/List.cc b/debian/htdig/htdig-3.2.0b6/htlib/List.cc new file mode 100644 index 00000000..dd020aab --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/List.cc @@ -0,0 +1,511 @@ +// +// List.cc +// +// List: A List class which holds objects of type Object. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: List.cc,v 1.9 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "List.h" + +class listnode +{ +public: + + listnode *next; + Object *object; +}; + + +//********************************************************************* +// List::List() +// Constructor +// +List::List() +{ + head = tail = 0; + number = 0; +} + + +//********************************************************************* +// List::~List() +// Destructor +// +List::~List() +{ + Destroy(); +} + + +//********************************************************************* +// void List::Release() +// Release all the objects from our list. +// +void List::Release() +{ + listnode *node; + while (head) + { + node = head; + head = head->next; + delete node; + } + head = tail = 0; + number = 0; + cursor.Clear(); +} + + +//********************************************************************* +// void List::Destroy() +// Delete all the objects from our list. +// +void List::Destroy() +{ + listnode *node; + while (head) + { + node = head; + head = head->next; + delete node->object; + delete node; + } + head = tail = 0; + number = 0; + cursor.Clear(); +} + + +//********************************************************************* +// void List::Add(Object *object) +// Add an object to the list. +// +void List::Add(Object *object) +{ + listnode *node = new listnode; + node->next = 0; + node->object = object; + if (tail) + { + tail->next = node; + tail = node; + } + else + { + head = tail = node; + } + + number++; +} + + +//********************************************************************* +// void List::Insert(Object *object, int position) +// Add an object to the list. +// +void List::Insert(Object *object, int position) +{ + listnode *node = new listnode; + node->next = 0; + node->object = object; + + listnode *ln = head; + listnode *prev = 0; + + for (int i = 0; i < position && ln; i++, ln = ln->next) + prev = ln; + if (!ln) + { + if (tail) + tail->next = node; + tail = node; + + // + // The list is empty. This is a simple case, then. + // + if (!head) + head = node; + } + else + { + if (ln == head) + { + node->next = head; + head = node; + } + else + { + node->next = ln; + prev->next = node; + } + } + + cursor.current_index = -1; + number++; +} + + +//********************************************************************* +// void List::Assign(Object *object, int position) +// Assign a new value to an index. +// +void List::Assign(Object *object, int position) +{ + // + // First make sure that there is something there! + // + while (number < position + 1) + { + Add(0); + } + + // + // Now find the listnode to put the new object in + // + listnode *temp = head; + + for (int i = 0; temp && i < position; i++) + { + temp = temp->next; + } + + cursor.current_index = -1; + delete temp->object; + temp->object = object; +} + + +//********************************************************************* +// int List::Remove(Object *object) +// Remove an object from the list. +// +int List::Remove(Object *object) +{ + listnode *node = head; + listnode *prev = 0; + while (node) + { + if (node->object == object) + { + // + // Found it! + // + // + // If we are in the middle of a Get_Next() sequence, we need to + // fix up any problems with the current node. + // + if (cursor.current == node) + { + cursor.current = node->next; + } + + if (head == tail) + { + head = tail = 0; + } + else if (head == node) + { + head = head->next; + } + else if (tail == node) + { + tail = prev; + tail->next = 0; + } + else + { + prev->next = node->next; + } + + delete node; + number--; + cursor.current_index = -1; + return 1; + } + prev = node; + node = node->next; + } + return 0; +} + +//********************************************************************* +// +int List::Remove(int position, int action /* = LIST_REMOVE_DESTROY */) +{ + Object *o = List::operator[](position); + if(action == LIST_REMOVE_DESTROY) delete o; + return List::Remove(o); +} + +//********************************************************************* +// Object *List::Get_Next() +// Return the next object in the list. +// +Object *List::Get_Next(ListCursor& cursor) const +{ + listnode *temp = cursor.current; + + if (cursor.current) + { + cursor.prev = cursor.current; + cursor.current = cursor.current->next; + if (cursor.current_index >= 0) + cursor.current_index++; + } + else + return 0; + return temp->object; +} + + +//********************************************************************* +// Object *List::Get_First() +// Return the first object in the list. +// +Object *List::Get_First() +{ + if (head) + return head->object; + else + return 0; +} + + +//********************************************************************* +// int List::Index(Object *obj) +// Return the index of an object in the list. +// +int List::Index(Object *obj) +{ + listnode *temp = head; + int index = 0; + + while (temp && temp->object != obj) + { + temp = temp->next; + index++; + } + if (index >= number) + return -1; + else + return index; +} + + +//********************************************************************* +// Object *List::Next(Object *prev) +// Return the next object in the list. Using this, the list will +// appear as a circular list. +// +Object *List::Next(Object *prev) +{ + listnode *node = head; + while (node) + { + if (node->object == prev) + { + node = node->next; + if (!node) + return head->object; + else + return node->object; + } + node = node->next; + } + + return 0; +} + + +//********************************************************************* +// Object *List::Previous(Object *next) +// Return the next object in the list. Using this, the list will +// appear as a circular list. +// +Object *List::Previous(Object *next) +{ + listnode *node = head; + listnode *prev = 0; + while (node) + { + if (node->object == next) + { + if (!prev) + return 0; + else + return prev->object; + } + prev = node; + node = node->next; + } + + return 0; +} + + +//********************************************************************* +// Return the nth object in the list. +// +const Object *List::Nth(ListCursor& cursor, int n) const +{ + if (n < 0 || n >= number) + return 0; + + listnode *temp = head; + + if (cursor.current_index == n) + return cursor.current->object; + + if (cursor.current && cursor.current_index >= 0 && n == cursor.current_index + 1) + { + cursor.prev = cursor.current; + cursor.current = cursor.current->next; + if (!cursor.current) + { + cursor.current_index = -1; + return 0; + } + cursor.current_index = n; + return cursor.current->object; + } + + for (int i = 0; temp && i < n; i++) + { + temp = temp->next; + } + + if (temp) + { + cursor.current_index = n; + cursor.current = temp; + return temp->object; + } + else + return 0; +} + + +//********************************************************************* +// Object *List::Last() +// Return the last object inserted. +// +Object *List::Last() +{ + if (tail) + { + return tail->object; + } + + return 0; +} + +//********************************************************************* +// +Object *List::Pop(int action /* = LIST_REMOVE_DESTROY */) +{ + Object *o = 0; + listnode *ln = head; + listnode *prev = 0; + + if (tail) { + if(action == LIST_REMOVE_DESTROY) { + delete tail->object; + } else { + o = tail->object; + } + if(head == tail) { + head = tail = 0; + } else { + + for (int i = 0; ln != tail; i++, ln = ln->next) + prev = ln; + tail = prev; + tail->next = 0; + } + } + + return o; +} + + +//********************************************************************* +// Object *List::Copy() const +// Return a deep copy of the list. +// +Object *List::Copy() const +{ + List *list = new List; + ListCursor cursor; + + Start_Get(cursor); + Object *obj; + while ((obj = Get_Next(cursor))) + { + list->Add(obj->Copy()); + } + return list; +} + + +//********************************************************************* +// List &List::operator=(List &list) +// Return a deep copy of the list. +// +List &List::operator=(List &list) +{ + Destroy(); + list.Start_Get(); + Object *obj; + while ((obj = list.Get_Next())) + { + Add(obj->Copy()); + } + return *this; +} + + +//********************************************************************* +// void AppendList(List &list) +// Move contents of other list to the end of this list, and empty the +// other list. +// +void List::AppendList(List &list) +{ + // Never mind an empty list or ourselves. + if (list.number == 0 || &list == this) + return; + + // Correct our pointers in head and tail. + if (tail) + { + // Link in other list. + tail->next = list.head; + + // Update members for added contents. + number += list.number; + tail = list.tail; + } + else + { + head = list.head; + tail = list.tail; + number = list.number; + } + + // Clear others members to be an empty list. + list.head = list.tail = 0; + list.cursor.current = 0; + list.cursor.current_index = -1; + list.number = 0; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/List.h b/debian/htdig/htdig-3.2.0b6/htlib/List.h new file mode 100644 index 00000000..2a3fcd28 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/List.h @@ -0,0 +1,190 @@ +// +// List.h +// +// List: A List class which holds objects of type Object. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: List.h,v 1.9 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _List_h_ +#define _List_h_ + +#include "Object.h" + +// +// Behaviour of the Remove method. See comment before method +// declaration for more information. +// +#define LIST_REMOVE_DESTROY 1 +#define LIST_REMOVE_RELEASE 2 + +class List; +class listnode; + +class ListCursor { + public: + ListCursor() { current = 0; prev = 0; current_index = -1; } + void Clear() { current = 0; prev = 0; current_index = -1; } + + // + // Support for the Start_Get and Get_Next routines + // + listnode *current; + listnode *prev; + int current_index; +}; + +class List : public Object +{ +public: + // + // Constructor/Destructor + // + List(); + virtual ~List(); + + // + // Insert at beginning of list. + // + virtual void Unshift(Object *o) { Insert(o, 0); } + // + // Remove from the beginning of the list and return the + // object. + // + virtual Object* Shift(int action = LIST_REMOVE_DESTROY) { + Object* o = Nth(0); + if(Remove(0, action) == NOTOK) return 0; + return o; + } + // + // Append an Object to the end of the list + // + virtual void Push(Object *o) { Add(o); } + // + // Remove the last object from the list and return it. + // + virtual Object *Pop(int action = LIST_REMOVE_DESTROY); + + // + // Add() will append an Object to the end of the list + // + virtual void Add(Object *); + + // + // Insert() will insert an object at the given position. If the + // position is larger than the number of objects in the list, the + // object is appended; no new objects are created between the end + // of the list and the given position. + // + virtual void Insert(Object *, int position); + + // + // Assign() will replace the object already at the given position + // with the new object. If there is no object at the position,the + // list is extended with nil objects until the position is reached + // and then the given object is put there. (This really makes the + // List analogous to a dynamic array...) + // + virtual void Assign(Object *, int position); + + // + // Find the given object in the list and remove it from the list. + // The object will NOT be deleted. If the object is not found, + // NOTOK will be returned, else OK. + // + virtual int Remove(Object *); + + // + // Remove object at position from the list. If action is + // LIST_REMOVE_DESTROY delete the object stored at position. + // If action is LIST_REMOVE_RELEASE the object is not deleted. + // If the object is not found, + // NOTOK will be returned, else OK. + // + virtual int Remove(int position, int action = LIST_REMOVE_DESTROY); + + // + // Release() will set the list to empty. This call will NOT + // delete objects that were in the list before this call. + // + virtual void Release(); + + // + // Destroy() will delete all the objects in the list. This is + // equivalent to calling the destructor + // + virtual void Destroy(); + + // + // List traversel + // + void Start_Get() { Start_Get(cursor); } + void Start_Get(ListCursor& cursor0) const { cursor0.current = head; cursor0.prev = 0; cursor0.current_index = -1;} + Object *Get_Next() { return Get_Next(cursor); } + Object *Get_Next(ListCursor& cursor) const; + Object *Get_First(); + Object *Next(Object *current); + Object *Previous(Object *current); + Object *Last(); + + // + // Direct access to list items. This can only be used to retrieve + // objects from the list. To assign new objects, use Insert(), + // Add(), or Assign(). + // + Object *operator[] (int n) { return Nth(n); } + const Object *operator[] (int n) const { return Nth(((List*)this)->cursor, n); } + const Object *Nth(ListCursor& cursor, int n) const; + const Object *Nth(int n) const { return Nth(((List*)this)->cursor, n); } + Object *Nth(int n) { return (Object*)((List*)this)->Nth(((List*)this)->cursor, n); } + + // + // Access to the number of elements + // + int Count() const { return number; } + + // + // Get the index number of an object. If the object is not found, + // returnes -1 + // + int Index(Object *); + + // + // Deep copy member function + // + Object *Copy() const; + + // + // Assignment + // + List &operator= (List *list) {return *this = *list;} + List &operator= (List &list); + + // Move one list to the end of another, emptying the other list. + void AppendList (List &list); + +protected: + // + // Pointers into the list + // + listnode *head; + listnode *tail; + + // + // For list traversal it is nice to know where we are... + // + ListCursor cursor; + + // + // Its nice to keep track of how many things we contain... + // + int number; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Makefile.am b/debian/htdig/htdig-3.2.0b6/htlib/Makefile.am new file mode 100644 index 00000000..15636583 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Makefile.am @@ -0,0 +1,68 @@ +include $(top_srcdir)/Makefile.config + +pkglib_LTLIBRARIES = libht.la + +libht_la_SOURCES = Configuration.cc Database.cc Dictionary.cc \ + DB2_db.cc IntObject.cc List.cc Object.cc \ + ParsedString.cc Queue.cc QuotedStringList.cc Stack.cc \ + String.cc StringList.cc StringMatch.cc String_fmt.cc \ + good_strtok.cc strcasecmp.cc \ + mktime.c strptime.cc timegm.c \ + getcwd.c memcpy.c memmove.c raise.c strerror.c \ + HtCodec.cc HtWordCodec.cc \ + HtVector.cc HtHeap.cc \ + HtPack.cc HtDateTime.cc HtRegex.cc HtRegexList.cc \ + HtRegexReplace.cc HtRegexReplaceList.cc \ + HtVectorGeneric.cc HtMaxMin.cc HtWordType.cc \ + myqsort.c \ + md5.cc mhash_md5.c + +libht_la_LIBADD=@LTLIBOBJS@ +# this includes regex.c + +libht_la_LDFLAGS = -release $(HTDIG_MAJOR_VERSION).$(HTDIG_MINOR_VERSION).$(HTDIG_MICRO_VERSION) ${extra_ldflags} + +pkginclude_HEADERS = \ + clib.h \ + Configuration.h \ + DB2_db.h \ + Database.h \ + Dictionary.h \ + HtCodec.h \ + HtDateTime.h \ + HtHeap.h \ + HtPack.h \ + HtRegex.h \ + HtRegexList.h \ + HtRegexReplace.h \ + HtRegexReplaceList.h \ + HtVector.h \ + HtWordCodec.h \ + IntObject.h \ + List.h \ + Object.h \ + ParsedString.h \ + Queue.h \ + QuotedStringList.h \ + Stack.h \ + StringList.h \ + StringMatch.h \ + good_strtok.h \ + htString.h \ + lib.h \ + gregex.h \ + HtVectorGeneric.h \ + HtVectorGenericCode.h \ + HtVector_int.h \ + HtVector_String.h \ + HtMaxMin.h \ + HtTime.h \ + HtWordType.h \ + HtRandom.h \ + ber.h \ + libdefs.h \ + myqsort.h \ + mhash_md5.h \ + md5.h + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Makefile.in b/debian/htdig/htdig-3.2.0b6/htlib/Makefile.in new file mode 100644 index 00000000..bbc58388 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Makefile.in @@ -0,0 +1,579 @@ +# Makefile.in generated by automake 1.7.9 from Makefile.am. +# @configure_input@ + +# Copyright 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003 +# Free Software Foundation, Inc. +# This Makefile.in is free software; the Free Software Foundation +# gives unlimited permission to copy and/or distribute it, +# with or without modifications, as long as this notice is preserved. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY, to the extent permitted by law; without +# even the implied warranty of MERCHANTABILITY or FITNESS FOR A +# PARTICULAR PURPOSE. + +@SET_MAKE@ + +# +# To compile with profiling do the following: +# +# make CFLAGS=-g CXXFLAGS=-g PROFILING=-p all +# + +srcdir = @srcdir@ +top_srcdir = @top_srcdir@ +VPATH = @srcdir@ +pkgdatadir = $(datadir)/@PACKAGE@ +pkglibdir = $(libdir)/@PACKAGE@ +pkgincludedir = $(includedir)/@PACKAGE@ +top_builddir = .. + +am__cd = CDPATH="$${ZSH_VERSION+.}$(PATH_SEPARATOR)" && cd +INSTALL = @INSTALL@ +install_sh_DATA = $(install_sh) -c -m 644 +install_sh_PROGRAM = $(install_sh) -c +install_sh_SCRIPT = $(install_sh) -c +INSTALL_HEADER = $(INSTALL_DATA) +transform = $(program_transform_name) +NORMAL_INSTALL = : +PRE_INSTALL = : +POST_INSTALL = : +NORMAL_UNINSTALL = : +PRE_UNINSTALL = : +POST_UNINSTALL = : +host_triplet = @host@ +ACLOCAL = @ACLOCAL@ +ALLOCA = @ALLOCA@ +AMDEP_FALSE = @AMDEP_FALSE@ +AMDEP_TRUE = @AMDEP_TRUE@ +AMTAR = @AMTAR@ +APACHE = @APACHE@ +APACHE_MODULES = @APACHE_MODULES@ +AR = @AR@ +AUTOCONF = @AUTOCONF@ +AUTOHEADER = @AUTOHEADER@ +AUTOMAKE = @AUTOMAKE@ +AWK = @AWK@ +CC = @CC@ +CCDEPMODE = @CCDEPMODE@ +CFLAGS = @CFLAGS@ +CGIBIN_DIR = @CGIBIN_DIR@ +COMMON_DIR = @COMMON_DIR@ +CONFIG_DIR = @CONFIG_DIR@ +CPP = @CPP@ +CPPFLAGS = @CPPFLAGS@ +CXX = @CXX@ +CXXCPP = @CXXCPP@ +CXXDEPMODE = @CXXDEPMODE@ +CXXFLAGS = @CXXFLAGS@ +CYGPATH_W = @CYGPATH_W@ +DATABASE_DIR = @DATABASE_DIR@ +DEFAULT_CONFIG_FILE = @DEFAULT_CONFIG_FILE@ +DEFS = @DEFS@ +DEPDIR = @DEPDIR@ +ECHO = @ECHO@ +ECHO_C = @ECHO_C@ +ECHO_N = @ECHO_N@ +ECHO_T = @ECHO_T@ +EGREP = @EGREP@ +EXEEXT = @EXEEXT@ +F77 = @F77@ +FFLAGS = @FFLAGS@ +FIND = @FIND@ +GUNZIP = @GUNZIP@ +HAVE_SSL = @HAVE_SSL@ +HTDIG_MAJOR_VERSION = @HTDIG_MAJOR_VERSION@ +HTDIG_MICRO_VERSION = @HTDIG_MICRO_VERSION@ +HTDIG_MINOR_VERSION = @HTDIG_MINOR_VERSION@ +IMAGE_DIR = @IMAGE_DIR@ +IMAGE_URL_PREFIX = @IMAGE_URL_PREFIX@ +INSTALL_DATA = @INSTALL_DATA@ +INSTALL_PROGRAM = @INSTALL_PROGRAM@ +INSTALL_SCRIPT = @INSTALL_SCRIPT@ +INSTALL_STRIP_PROGRAM = @INSTALL_STRIP_PROGRAM@ +LDFLAGS = @LDFLAGS@ +LEX = @LEX@ +LEXLIB = @LEXLIB@ +LEX_OUTPUT_ROOT = @LEX_OUTPUT_ROOT@ +LIBOBJS = @LIBOBJS@ +LIBS = @LIBS@ +LIBTOOL = @LIBTOOL@ +LN_S = @LN_S@ +LTLIBOBJS = @LTLIBOBJS@ +MAINT = @MAINT@ +MAINTAINER_MODE_FALSE = @MAINTAINER_MODE_FALSE@ +MAINTAINER_MODE_TRUE = @MAINTAINER_MODE_TRUE@ +MAKEINFO = @MAKEINFO@ +MV = @MV@ +OBJEXT = @OBJEXT@ +PACKAGE = @PACKAGE@ +PACKAGE_BUGREPORT = @PACKAGE_BUGREPORT@ +PACKAGE_NAME = @PACKAGE_NAME@ +PACKAGE_STRING = @PACKAGE_STRING@ +PACKAGE_TARNAME = @PACKAGE_TARNAME@ +PACKAGE_VERSION = @PACKAGE_VERSION@ +PATH_SEPARATOR = @PATH_SEPARATOR@ +PERL = @PERL@ +RANLIB = @RANLIB@ +RRDTOOL = @RRDTOOL@ +SEARCH_DIR = @SEARCH_DIR@ +SEARCH_FORM = @SEARCH_FORM@ +SED = @SED@ +SENDMAIL = @SENDMAIL@ +SET_MAKE = @SET_MAKE@ +SHELL = @SHELL@ +STRIP = @STRIP@ +TAR = @TAR@ +TESTS_FALSE = @TESTS_FALSE@ +TESTS_TRUE = @TESTS_TRUE@ +TIME = @TIME@ +TIMEV = @TIMEV@ +USER = @USER@ +VERSION = @VERSION@ +YACC = @YACC@ +ac_ct_AR = @ac_ct_AR@ +ac_ct_CC = @ac_ct_CC@ +ac_ct_CXX = @ac_ct_CXX@ +ac_ct_F77 = @ac_ct_F77@ +ac_ct_RANLIB = @ac_ct_RANLIB@ +ac_ct_STRIP = @ac_ct_STRIP@ +am__fastdepCC_FALSE = @am__fastdepCC_FALSE@ +am__fastdepCC_TRUE = @am__fastdepCC_TRUE@ +am__fastdepCXX_FALSE = @am__fastdepCXX_FALSE@ +am__fastdepCXX_TRUE = @am__fastdepCXX_TRUE@ +am__include = @am__include@ +am__leading_dot = @am__leading_dot@ +am__quote = @am__quote@ +bindir = @bindir@ +build = @build@ +build_alias = @build_alias@ +build_cpu = @build_cpu@ +build_os = @build_os@ +build_vendor = @build_vendor@ +datadir = @datadir@ +exec_prefix = @exec_prefix@ +extra_ldflags = @extra_ldflags@ +host = @host@ +host_alias = @host_alias@ +host_cpu = @host_cpu@ +host_os = @host_os@ +host_vendor = @host_vendor@ +includedir = @includedir@ +infodir = @infodir@ +install_sh = @install_sh@ +libdir = @libdir@ +libexecdir = @libexecdir@ +localstatedir = @localstatedir@ +mandir = @mandir@ +oldincludedir = @oldincludedir@ +prefix = @prefix@ +program_transform_name = @program_transform_name@ +sbindir = @sbindir@ +sharedstatedir = @sharedstatedir@ +subdirs = @subdirs@ +sysconfdir = @sysconfdir@ +target_alias = @target_alias@ + +AUTOMAKE_OPTIONS = foreign no-dependencies + +INCLUDES = -DDEFAULT_CONFIG_FILE=\"$(DEFAULT_CONFIG_FILE)\" \ + -I$(top_srcdir)/include -I$(top_srcdir)/htlib \ + -I$(top_srcdir)/htnet -I$(top_srcdir)/htcommon \ + -I$(top_srcdir)/htword \ + -I$(top_srcdir)/db -I$(top_builddir)/db \ + $(LOCAL_DEFINES) $(PROFILING) + + +HTLIBS = $(top_builddir)/htnet/libhtnet.la \ + $(top_builddir)/htcommon/libcommon.la \ + $(top_builddir)/htword/libhtword.la \ + $(top_builddir)/htlib/libht.la \ + $(top_builddir)/htcommon/libcommon.la \ + $(top_builddir)/htword/libhtword.la \ + $(top_builddir)/db/libhtdb.la \ + $(top_builddir)/htlib/libht.la + + +pkglib_LTLIBRARIES = libht.la + +libht_la_SOURCES = Configuration.cc Database.cc Dictionary.cc \ + DB2_db.cc IntObject.cc List.cc Object.cc \ + ParsedString.cc Queue.cc QuotedStringList.cc Stack.cc \ + String.cc StringList.cc StringMatch.cc String_fmt.cc \ + good_strtok.cc strcasecmp.cc \ + mktime.c strptime.cc timegm.c \ + getcwd.c memcpy.c memmove.c raise.c strerror.c \ + HtCodec.cc HtWordCodec.cc \ + HtVector.cc HtHeap.cc \ + HtPack.cc HtDateTime.cc HtRegex.cc HtRegexList.cc \ + HtRegexReplace.cc HtRegexReplaceList.cc \ + HtVectorGeneric.cc HtMaxMin.cc HtWordType.cc \ + myqsort.c \ + md5.cc mhash_md5.c + + +libht_la_LIBADD = @LTLIBOBJS@ + +# this includes regex.c +libht_la_LDFLAGS = -release $(HTDIG_MAJOR_VERSION).$(HTDIG_MINOR_VERSION).$(HTDIG_MICRO_VERSION) ${extra_ldflags} + +pkginclude_HEADERS = \ + clib.h \ + Configuration.h \ + DB2_db.h \ + Database.h \ + Dictionary.h \ + HtCodec.h \ + HtDateTime.h \ + HtHeap.h \ + HtPack.h \ + HtRegex.h \ + HtRegexList.h \ + HtRegexReplace.h \ + HtRegexReplaceList.h \ + HtVector.h \ + HtWordCodec.h \ + IntObject.h \ + List.h \ + Object.h \ + ParsedString.h \ + Queue.h \ + QuotedStringList.h \ + Stack.h \ + StringList.h \ + StringMatch.h \ + good_strtok.h \ + htString.h \ + lib.h \ + gregex.h \ + HtVectorGeneric.h \ + HtVectorGenericCode.h \ + HtVector_int.h \ + HtVector_String.h \ + HtMaxMin.h \ + HtTime.h \ + HtWordType.h \ + HtRandom.h \ + ber.h \ + libdefs.h \ + myqsort.h \ + mhash_md5.h \ + md5.h + +subdir = htlib +ACLOCAL_M4 = $(top_srcdir)/aclocal.m4 +mkinstalldirs = $(SHELL) $(top_srcdir)/mkinstalldirs +CONFIG_HEADER = $(top_builddir)/include/config.h +CONFIG_CLEAN_FILES = +LTLIBRARIES = $(pkglib_LTLIBRARIES) + +libht_la_DEPENDENCIES = @LTLIBOBJS@ +am_libht_la_OBJECTS = Configuration.lo Database.lo Dictionary.lo \ + DB2_db.lo IntObject.lo List.lo Object.lo ParsedString.lo \ + Queue.lo QuotedStringList.lo Stack.lo String.lo StringList.lo \ + StringMatch.lo String_fmt.lo good_strtok.lo strcasecmp.lo \ + mktime.lo strptime.lo timegm.lo getcwd.lo memcpy.lo memmove.lo \ + raise.lo strerror.lo HtCodec.lo HtWordCodec.lo HtVector.lo \ + HtHeap.lo HtPack.lo HtDateTime.lo HtRegex.lo HtRegexList.lo \ + HtRegexReplace.lo HtRegexReplaceList.lo HtVectorGeneric.lo \ + HtMaxMin.lo HtWordType.lo myqsort.lo md5.lo mhash_md5.lo +libht_la_OBJECTS = $(am_libht_la_OBJECTS) + +DEFAULT_INCLUDES = -I. -I$(srcdir) -I$(top_builddir)/include +depcomp = +am__depfiles_maybe = +COMPILE = $(CC) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) \ + $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) +LTCOMPILE = $(LIBTOOL) --mode=compile $(CC) $(DEFS) $(DEFAULT_INCLUDES) \ + $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CFLAGS) $(CFLAGS) +CCLD = $(CC) +LINK = $(LIBTOOL) --mode=link $(CCLD) $(AM_CFLAGS) $(CFLAGS) \ + $(AM_LDFLAGS) $(LDFLAGS) -o $@ +CXXCOMPILE = $(CXX) $(DEFS) $(DEFAULT_INCLUDES) $(INCLUDES) \ + $(AM_CPPFLAGS) $(CPPFLAGS) $(AM_CXXFLAGS) $(CXXFLAGS) +LTCXXCOMPILE = $(LIBTOOL) --mode=compile $(CXX) $(DEFS) \ + $(DEFAULT_INCLUDES) $(INCLUDES) $(AM_CPPFLAGS) $(CPPFLAGS) \ + $(AM_CXXFLAGS) $(CXXFLAGS) +CXXLD = $(CXX) +CXXLINK = $(LIBTOOL) --mode=link $(CXXLD) $(AM_CXXFLAGS) $(CXXFLAGS) \ + $(AM_LDFLAGS) $(LDFLAGS) -o $@ +DIST_SOURCES = $(libht_la_SOURCES) +HEADERS = $(pkginclude_HEADERS) + +DIST_COMMON = $(pkginclude_HEADERS) $(srcdir)/Makefile.in \ + $(top_srcdir)/Makefile.config Makefile.am malloc.c memcmp.c \ + realloc.c regex.c snprintf.c vsnprintf.c +SOURCES = $(libht_la_SOURCES) + +all: all-am + +.SUFFIXES: +.SUFFIXES: .c .cc .lo .o .obj +$(srcdir)/Makefile.in: @MAINTAINER_MODE_TRUE@ Makefile.am $(top_srcdir)/Makefile.config $(top_srcdir)/configure.in $(ACLOCAL_M4) + cd $(top_srcdir) && \ + $(AUTOMAKE) --foreign htlib/Makefile +Makefile: @MAINTAINER_MODE_TRUE@ $(srcdir)/Makefile.in $(top_builddir)/config.status + cd $(top_builddir) && $(SHELL) ./config.status $(subdir)/$@ $(am__depfiles_maybe) +pkglibLTLIBRARIES_INSTALL = $(INSTALL) +install-pkglibLTLIBRARIES: $(pkglib_LTLIBRARIES) + @$(NORMAL_INSTALL) + $(mkinstalldirs) $(DESTDIR)$(pkglibdir) + @list='$(pkglib_LTLIBRARIES)'; for p in $$list; do \ + if test -f $$p; then \ + f="`echo $$p | sed -e 's|^.*/||'`"; \ + echo " $(LIBTOOL) --mode=install $(pkglibLTLIBRARIES_INSTALL) $(INSTALL_STRIP_FLAG) $$p $(DESTDIR)$(pkglibdir)/$$f"; \ + $(LIBTOOL) --mode=install $(pkglibLTLIBRARIES_INSTALL) $(INSTALL_STRIP_FLAG) $$p $(DESTDIR)$(pkglibdir)/$$f; \ + else :; fi; \ + done + +uninstall-pkglibLTLIBRARIES: + @$(NORMAL_UNINSTALL) + @list='$(pkglib_LTLIBRARIES)'; for p in $$list; do \ + p="`echo $$p | sed -e 's|^.*/||'`"; \ + echo " $(LIBTOOL) --mode=uninstall rm -f $(DESTDIR)$(pkglibdir)/$$p"; \ + $(LIBTOOL) --mode=uninstall rm -f $(DESTDIR)$(pkglibdir)/$$p; \ + done + +clean-pkglibLTLIBRARIES: + -test -z "$(pkglib_LTLIBRARIES)" || rm -f $(pkglib_LTLIBRARIES) + @list='$(pkglib_LTLIBRARIES)'; for p in $$list; do \ + dir="`echo $$p | sed -e 's|/[^/]*$$||'`"; \ + test "$$dir" = "$$p" && dir=.; \ + echo "rm -f \"$${dir}/so_locations\""; \ + rm -f "$${dir}/so_locations"; \ + done +libht.la: $(libht_la_OBJECTS) $(libht_la_DEPENDENCIES) + $(CXXLINK) -rpath $(pkglibdir) $(libht_la_LDFLAGS) $(libht_la_OBJECTS) $(libht_la_LIBADD) $(LIBS) + +mostlyclean-compile: + -rm -f *.$(OBJEXT) core *.core + +distclean-compile: + -rm -f *.tab.c + +.c.o: + $(COMPILE) -c `test -f '$<' || echo '$(srcdir)/'`$< + +.c.obj: + $(COMPILE) -c `if test -f '$<'; then $(CYGPATH_W) '$<'; else $(CYGPATH_W) '$(srcdir)/$<'; fi` + +.c.lo: + $(LTCOMPILE) -c -o $@ `test -f '$<' || echo '$(srcdir)/'`$< + +.cc.o: + $(CXXCOMPILE) -c -o $@ `test -f '$<' || echo '$(srcdir)/'`$< + +.cc.obj: + $(CXXCOMPILE) -c -o $@ `if test -f '$<'; then $(CYGPATH_W) '$<'; else $(CYGPATH_W) '$(srcdir)/$<'; fi` + +.cc.lo: + $(LTCXXCOMPILE) -c -o $@ `test -f '$<' || echo '$(srcdir)/'`$< + +mostlyclean-libtool: + -rm -f *.lo + +clean-libtool: + -rm -rf .libs _libs + +distclean-libtool: + -rm -f libtool +uninstall-info-am: +pkgincludeHEADERS_INSTALL = $(INSTALL_HEADER) +install-pkgincludeHEADERS: $(pkginclude_HEADERS) + @$(NORMAL_INSTALL) + $(mkinstalldirs) $(DESTDIR)$(pkgincludedir) + @list='$(pkginclude_HEADERS)'; for p in $$list; do \ + if test -f "$$p"; then d=; else d="$(srcdir)/"; fi; \ + f="`echo $$p | sed -e 's|^.*/||'`"; \ + echo " $(pkgincludeHEADERS_INSTALL) $$d$$p $(DESTDIR)$(pkgincludedir)/$$f"; \ + $(pkgincludeHEADERS_INSTALL) $$d$$p $(DESTDIR)$(pkgincludedir)/$$f; \ + done + +uninstall-pkgincludeHEADERS: + @$(NORMAL_UNINSTALL) + @list='$(pkginclude_HEADERS)'; for p in $$list; do \ + f="`echo $$p | sed -e 's|^.*/||'`"; \ + echo " rm -f $(DESTDIR)$(pkgincludedir)/$$f"; \ + rm -f $(DESTDIR)$(pkgincludedir)/$$f; \ + done + +ETAGS = etags +ETAGSFLAGS = + +CTAGS = ctags +CTAGSFLAGS = + +tags: TAGS + +ID: $(HEADERS) $(SOURCES) $(LISP) $(TAGS_FILES) + list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ + unique=`for i in $$list; do \ + if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ + done | \ + $(AWK) ' { files[$$0] = 1; } \ + END { for (i in files) print i; }'`; \ + mkid -fID $$unique + +TAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ + $(TAGS_FILES) $(LISP) + tags=; \ + here=`pwd`; \ + list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ + unique=`for i in $$list; do \ + if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ + done | \ + $(AWK) ' { files[$$0] = 1; } \ + END { for (i in files) print i; }'`; \ + test -z "$(ETAGS_ARGS)$$tags$$unique" \ + || $(ETAGS) $(ETAGSFLAGS) $(AM_ETAGSFLAGS) $(ETAGS_ARGS) \ + $$tags $$unique + +ctags: CTAGS +CTAGS: $(HEADERS) $(SOURCES) $(TAGS_DEPENDENCIES) \ + $(TAGS_FILES) $(LISP) + tags=; \ + here=`pwd`; \ + list='$(SOURCES) $(HEADERS) $(LISP) $(TAGS_FILES)'; \ + unique=`for i in $$list; do \ + if test -f "$$i"; then echo $$i; else echo $(srcdir)/$$i; fi; \ + done | \ + $(AWK) ' { files[$$0] = 1; } \ + END { for (i in files) print i; }'`; \ + test -z "$(CTAGS_ARGS)$$tags$$unique" \ + || $(CTAGS) $(CTAGSFLAGS) $(AM_CTAGSFLAGS) $(CTAGS_ARGS) \ + $$tags $$unique + +GTAGS: + here=`$(am__cd) $(top_builddir) && pwd` \ + && cd $(top_srcdir) \ + && gtags -i $(GTAGS_ARGS) $$here + +distclean-tags: + -rm -f TAGS ID GTAGS GRTAGS GSYMS GPATH tags +DISTFILES = $(DIST_COMMON) $(DIST_SOURCES) $(TEXINFOS) $(EXTRA_DIST) + +top_distdir = .. +distdir = $(top_distdir)/$(PACKAGE)-$(VERSION) + +distdir: $(DISTFILES) + $(mkinstalldirs) $(distdir)/.. + @srcdirstrip=`echo "$(srcdir)" | sed 's|.|.|g'`; \ + topsrcdirstrip=`echo "$(top_srcdir)" | sed 's|.|.|g'`; \ + list='$(DISTFILES)'; for file in $$list; do \ + case $$file in \ + $(srcdir)/*) file=`echo "$$file" | sed "s|^$$srcdirstrip/||"`;; \ + $(top_srcdir)/*) file=`echo "$$file" | sed "s|^$$topsrcdirstrip/|$(top_builddir)/|"`;; \ + esac; \ + if test -f $$file || test -d $$file; then d=.; else d=$(srcdir); fi; \ + dir=`echo "$$file" | sed -e 's,/[^/]*$$,,'`; \ + if test "$$dir" != "$$file" && test "$$dir" != "."; then \ + dir="/$$dir"; \ + $(mkinstalldirs) "$(distdir)$$dir"; \ + else \ + dir=''; \ + fi; \ + if test -d $$d/$$file; then \ + if test -d $(srcdir)/$$file && test $$d != $(srcdir); then \ + cp -pR $(srcdir)/$$file $(distdir)$$dir || exit 1; \ + fi; \ + cp -pR $$d/$$file $(distdir)$$dir || exit 1; \ + else \ + test -f $(distdir)/$$file \ + || cp -p $$d/$$file $(distdir)/$$file \ + || exit 1; \ + fi; \ + done +check-am: all-am +check: check-am +all-am: Makefile $(LTLIBRARIES) $(HEADERS) + +installdirs: + $(mkinstalldirs) $(DESTDIR)$(pkglibdir) $(DESTDIR)$(pkgincludedir) +install: install-am +install-exec: install-exec-am +install-data: install-data-am +uninstall: uninstall-am + +install-am: all-am + @$(MAKE) $(AM_MAKEFLAGS) install-exec-am install-data-am + +installcheck: installcheck-am +install-strip: + $(MAKE) $(AM_MAKEFLAGS) INSTALL_PROGRAM="$(INSTALL_STRIP_PROGRAM)" \ + install_sh_PROGRAM="$(INSTALL_STRIP_PROGRAM)" INSTALL_STRIP_FLAG=-s \ + `test -z '$(STRIP)' || \ + echo "INSTALL_PROGRAM_ENV=STRIPPROG='$(STRIP)'"` install +mostlyclean-generic: + +clean-generic: + +distclean-generic: + -rm -f $(CONFIG_CLEAN_FILES) + +maintainer-clean-generic: + @echo "This command is intended for maintainers to use" + @echo "it deletes files that may require special tools to rebuild." +clean: clean-am + +clean-am: clean-generic clean-libtool clean-pkglibLTLIBRARIES \ + mostlyclean-am + +distclean: distclean-am + -rm -f Makefile +distclean-am: clean-am distclean-compile distclean-generic \ + distclean-libtool distclean-tags + +dvi: dvi-am + +dvi-am: + +info: info-am + +info-am: + +install-data-am: install-pkgincludeHEADERS + +install-exec-am: install-pkglibLTLIBRARIES + +install-info: install-info-am + +install-man: + +installcheck-am: + +maintainer-clean: maintainer-clean-am + -rm -f Makefile +maintainer-clean-am: distclean-am maintainer-clean-generic + +mostlyclean: mostlyclean-am + +mostlyclean-am: mostlyclean-compile mostlyclean-generic \ + mostlyclean-libtool + +pdf: pdf-am + +pdf-am: + +ps: ps-am + +ps-am: + +uninstall-am: uninstall-info-am uninstall-pkgincludeHEADERS \ + uninstall-pkglibLTLIBRARIES + +.PHONY: CTAGS GTAGS all all-am check check-am clean clean-generic \ + clean-libtool clean-pkglibLTLIBRARIES ctags distclean \ + distclean-compile distclean-generic distclean-libtool \ + distclean-tags distdir dvi dvi-am info info-am install \ + install-am install-data install-data-am install-exec \ + install-exec-am install-info install-info-am install-man \ + install-pkgincludeHEADERS install-pkglibLTLIBRARIES \ + install-strip installcheck installcheck-am installdirs \ + maintainer-clean maintainer-clean-generic mostlyclean \ + mostlyclean-compile mostlyclean-generic mostlyclean-libtool pdf \ + pdf-am ps ps-am tags uninstall uninstall-am uninstall-info-am \ + uninstall-pkgincludeHEADERS uninstall-pkglibLTLIBRARIES + +# Tell versions [3.59,3.63) of GNU make to not export all variables. +# Otherwise a system limit (for SysV at least) may be exceeded. +.NOEXPORT: diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Makefile.win32 b/debian/htdig/htdig-3.2.0b6/htlib/Makefile.win32 new file mode 100644 index 00000000..07ca32ba --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Makefile.win32 @@ -0,0 +1,34 @@ + +TARGET = $(LIBDIR)/libht$(LIBSFX) + +# ---------------------------------------------------------------------------- +# add new library members to this list + +# ---------------------------------------------------------------------------- + +include ../Makedefs.win32 + +CXXSRC = Configuration.cc Database.cc Dictionary.cc DB2_db.cc \ + IntObject.cc List.cc Object.cc ParsedString.cc Queue.cc \ + QuotedStringList.cc Stack.cc String.cc StringList.cc \ + StringMatch.cc String_fmt.cc good_strtok.cc strcasecmp.cc \ + strptime.cc HtCodec.cc HtWordCodec.cc HtVector.cc HtHeap.cc \ + HtPack.cc HtDateTime.cc HtRegex.cc HtRegexList.cc \ + HtRegexReplace.cc HtRegexReplaceList.cc HtVectorGeneric.cc \ + HtMaxMin.cc HtWordType.cc md5.cc + +SRC = mktime.c mhash_md5.c timegm.c getcwd.c memcmp.c memcpy.c \ + memmove.c raise.c strerror.c myqsort.c +ifdef WINDIR +SRC += regex_win32.c getopt_local.c +CXXSRC += filecopy.cc +endif + +CPPFLAGS += -DHAVE_CONFIG_H -I../db -I. -I../htword -I../htcommon + +$(TARGET): $(OBJDIRDEP) $(LIBDIRDEP) $(OBJS) + $(AR) $(ARFLAGS) $(OBJS) + +include ../Makerules.win32 + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Object.cc b/debian/htdig/htdig-3.2.0b6/htlib/Object.cc new file mode 100644 index 00000000..039947a3 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Object.cc @@ -0,0 +1,76 @@ +// +// Object.cc +// +// Object: This baseclass defines how an object should behave. +// This includes the ability to be put into a list +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Object.cc,v 1.6 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "Object.h" + +#include <stdio.h> + + +//*************************************************************************** +// Object::Object() +// +#ifdef NOINLINE +Object::Object() +{ +} + + +//*************************************************************************** +// Object::~Object() +// +Object::~Object() +{ +} + + +//*************************************************************************** +// int Object::compare(Object *) +// +int Object::compare(Object *) +{ + return 0; +} + + +//*************************************************************************** +// Object *Object::Copy() +// +Object *Object::Copy() +{ + return new Object; +} + + +//*************************************************************************** +// void Object::Serialize(String &) +// +void Object::Serialize(String &) +{ +} + + +//*************************************************************************** +// void Object::Deserialize(String &, int &) +// +void Object::Deserialize(String &, int &) +{ +} +#endif + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Object.h b/debian/htdig/htdig-3.2.0b6/htlib/Object.h new file mode 100644 index 00000000..29b28b81 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Object.h @@ -0,0 +1,47 @@ +// +// Object.h +// +// Object: This baseclass defines how an object should behave. +// This includes the ability to be put into a list +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Object.h,v 1.9 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _Object_h_ +#define _Object_h_ + +#include "lib.h" +#include <stdio.h> + +class String; + +class Object +{ +public: + // + // Constructor/Destructor + // + Object() {} + virtual ~Object() {} + + // + // To ensure a consistent comparison interface and to allow comparison + // of all kinds of different objects, we will define a comparison functions. + // + virtual int compare(const Object &) const { return 0;} + + // + // To allow a deep copy of data structures we will define a standard interface... + // This member will return a copy of itself, freshly allocated and deep copied. + // + virtual Object *Copy() const { fprintf(stderr, "Object::Copy: derived class does not implement Copy\n"); return new Object(); } +}; + + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/ParsedString.cc b/debian/htdig/htdig-3.2.0b6/htlib/ParsedString.cc new file mode 100644 index 00000000..3a774fb6 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/ParsedString.cc @@ -0,0 +1,202 @@ +// +// ParsedString.cc +// +// ParsedString: Contains a string. The string my contain $var, ${var}, $(var) +// `filename`. The get method will expand those using the +// dictionary given in argument. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: ParsedString.cc,v 1.9 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "ParsedString.h" + +#include <ctype.h> +#include <stdio.h> + + +//***************************************************************************** +// ParsedString::ParsedString() +// +ParsedString::ParsedString() +{ +} + + +//***************************************************************************** +// +ParsedString::ParsedString(const String& s) +{ + value = s; +} + + +//***************************************************************************** +// ParsedString::~ParsedString() +// +ParsedString::~ParsedString() +{ +} + + +//***************************************************************************** +// +void +ParsedString::set(const String& str) +{ + value = str; +} + + +//***************************************************************************** +// Return a fully parsed string. +// +// Allowed syntax: +// $var +// ${var} +// $(var) +// `filename` +// +// The filename can also contain variables +// +const String +ParsedString::get(const Dictionary &dict) const +{ + String variable; + String parsed; + ParsedString *temp; + const char *str = value.get(); + char delim = ' '; + int need_delim = 0; + + while (*str) + { + if (*str == '$') + { + // + // A dollar sign starts a variable. + // + str++; + need_delim = 1; + if (*str == '{') + delim = '}'; + else if (*str == '(') + delim = ')'; + else + need_delim = 0; + if (need_delim) + str++; + variable.trunc(); + while (isalnum(*str) || *str == '_' || *str == '-') + { + variable << *str++; + } + if (*str) + { + if (need_delim && *str == delim) + { + // + // Found end of variable + // + temp = (ParsedString *) dict[variable]; + if (temp) + parsed << temp->get(dict); + str++; + } + else if (need_delim) + { + // + // Error. Probably an illegal value in the name We'll + // assume the variable ended here. + // + temp = (ParsedString *) dict[variable]; + if (temp) + parsed << temp->get(dict); + } + else + { + // + // This variable didn't have a delimiter. + // + temp = (ParsedString *) dict[variable]; + if (temp) + parsed << temp->get(dict); + } + } + else + { + // + // End of string reached. We'll assume that this is also + // the end of the variable + // + temp = (ParsedString *) dict[variable]; + if (temp) + parsed << temp->get(dict); + } + } + else if (*str == '`') + { + // + // Back-quote delimits a filename which we need to insert + // + str++; + variable.trunc(); + while (*str && *str != '`') + { + variable << *str++; + } + if (*str == '`') + str++; + ParsedString filename(variable); + variable.trunc(); + getFileContents(variable, filename.get(dict)); + parsed << variable; + } + else if (*str == '\\') + { + // + // Backslash escapes the next character + // + str++; + if (*str) + parsed << *str++; + } + else + { + // + // Normal character + // + parsed << *str++; + } + } + return parsed; +} + + +void +ParsedString::getFileContents(String &str, const String& filename) const +{ + FILE *fl = fopen(filename, "r"); + char buffer[1000]; + + if (!fl) + return; + while (fgets(buffer, sizeof(buffer), fl)) + { + String s(buffer); + s.chop("\r\n\t "); + str << s << ' '; + } + str.chop(1); + fclose(fl); +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/ParsedString.h b/debian/htdig/htdig-3.2.0b6/htlib/ParsedString.h new file mode 100644 index 00000000..cd69f933 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/ParsedString.h @@ -0,0 +1,41 @@ +// +// ParsedString.h +// +// ParsedString: Contains a string. The string my contain $var, ${var}, $(var) +// `filename`. The get method will expand those using the +// dictionary given in argument. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: ParsedString.h,v 1.9 2004/05/28 13:15:21 lha Exp $ + +#ifndef _ParsedString_h_ +#define _ParsedString_h_ + +#include "Object.h" +#include "htString.h" +#include "Dictionary.h" + +class ParsedString : public Object +{ +public: + // + // Construction/Destruction + // + ParsedString(); + ParsedString(const String& s); + ~ParsedString(); + + void set(const String& s); + const String get(const Dictionary &d) const; +private: + String value; + + void getFileContents(String &str, const String& filename) const; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Queue.cc b/debian/htdig/htdig-3.2.0b6/htlib/Queue.cc new file mode 100644 index 00000000..2156df0f --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Queue.cc @@ -0,0 +1,112 @@ +// +// Queue.cc +// +// Queue: This class implements a linked list of objects. It itself is also an +// object +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Queue.cc,v 1.6 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "Queue.h" + +struct Queuenode +{ + Queuenode *next; + Object *obj; +}; + +//*************************************************************************** +// Queue::Queue() +// +Queue::Queue() +{ + head = tail = 0; + size = 0; +} + + +//*************************************************************************** +// Queue::~Queue() +// +Queue::~Queue() +{ + destroy(); +} + + +//*************************************************************************** +// void Queue::destroy() +// +void Queue::destroy() +{ + while (head) + { + Object *obj = pop(); + delete obj; + } + size = 0; + head = tail = 0; +} + + +//*************************************************************************** +// void Queue::push(Object *obj) +// Push an object onto the Queue. +// +void Queue::push(Object *obj) +{ + Queuenode *node = new Queuenode; + + node->obj = obj; + node->next = 0; + if (tail) + ((Queuenode *) tail)->next = node; + tail = node; + if (!head) + head = tail; + size++; +} + + +//*************************************************************************** +// Object *Queue::pop() +// Return the object at the head of the Queue and remove it +// +Object *Queue::pop() +{ + if (size == 0) + return 0; + + Queuenode *node = (Queuenode *) head; + Object *obj = node->obj; + head = (void *) node->next; + delete node; + size--; + + if (!head) + tail = 0; + return obj; +} + + +//*************************************************************************** +// Object *Queue::peek() +// Return the object at the top of the Queue. +// +Object *Queue::peek() +{ + if (size == 0) + return 0; + + return ((Queuenode *)head)->obj; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Queue.h b/debian/htdig/htdig-3.2.0b6/htlib/Queue.h new file mode 100644 index 00000000..da3066e4 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Queue.h @@ -0,0 +1,52 @@ +// +// Queue.h +// +// Queue: This class implements a linked list of objects. It itself is also an +// object +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Queue.h,v 1.6 2004/05/28 13:15:21 lha Exp $ +// +#ifndef _Queue_h_ +#define _Queue_h_ + +#include "Object.h" + +class Queue : public Object +{ +public: + // + // Constructors/Destructor + // + Queue(); + ~Queue(); + + // + // Queue access + // + void push(Object *obj); + Object *peek(); + Object *pop(); + int Size() {return size;} + + // + // Queue destruction + // + void destroy(); + +protected: + // + // These variables are to keep track of the linked list + // + void *head; + void *tail; + + int size; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/QuotedStringList.cc b/debian/htdig/htdig-3.2.0b6/htlib/QuotedStringList.cc new file mode 100644 index 00000000..2d6aa006 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/QuotedStringList.cc @@ -0,0 +1,90 @@ +// +// QuotedStringList.cc +// +// QuotedStringList: Fed with a string it will extract separator delimited +// words and store them in a list. The words may be +// delimited by " or ', hence the name. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: QuotedStringList.cc,v 1.7 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "QuotedStringList.h" + + +//***************************************************************************** +QuotedStringList::QuotedStringList() +{ +} + +//***************************************************************************** +int +QuotedStringList::Create(const char *str, const char *sep, int single) +{ + char quote = 0; + int quoted = 0; + String word; + + while (str && *str) + { + if (*str == '\\') + { + if (!str[1]) + break; + word << *++str; + } + else if (*str == quote) + { + quote = 0; + } + else if (!quote && (*str == '"' || *str == '\'')) + { + quote = *str; + quoted++; + } + else if (quote == 0 && strchr(sep, *str)) + { + Add(new String(word)); + word = 0; + quoted = 0; + if (!single) + { + while (strchr(sep, *str)) + str++; + str--; + } + } + else + word << *str; + str++; + } + + // + // Add the last word to the list + // + if (word.length() || quoted) + Add(new String(word)); + return Count(); +} + + +//***************************************************************************** +int +QuotedStringList::Create(const char *str, char sep, int single) +{ + char t[2] = "1"; + + t[0] = sep; + return Create(str, t, single); +} + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/QuotedStringList.h b/debian/htdig/htdig-3.2.0b6/htlib/QuotedStringList.h new file mode 100644 index 00000000..d869248d --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/QuotedStringList.h @@ -0,0 +1,47 @@ +// +// QuotedStringList.h +// +// QuotedStringList: Fed with a string it will extract separator delimited +// words and store them in a list. The words may be +// delimited by " or ', hence the name. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: QuotedStringList.h,v 1.7 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _QuotedStringList_h_ +#define _QuotedStringList_h_ + +#include "StringList.h" + +class QuotedStringList : public StringList +{ +public: + // + // Construction/Destruction + // + QuotedStringList(); + + // + // Creation of a String from a string or String + // + QuotedStringList(const char *str, char sep = '\t', int single = 0) { Create(str, sep, single); } + QuotedStringList(const String &str, char sep = '\t', int single = 0) { Create(str, sep, single); } + QuotedStringList(const char *str, const char *sep, int single = 0) { Create(str, sep, single); } + QuotedStringList(const String &str, const char *sep, int single = 0) { Create(str, sep, single); } + + int Create(const char *str, char sep = '\t', int single = 0); + int Create(const String &str, char sep = '\t', int single = 0) { return Create(str.get(), sep, single); } + int Create(const char *str, const char *sep, int single = 0); + int Create(const String &str, const char *sep, int single = 0) { return Create(str.get(), sep, single); } +private: +}; + +#endif + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Stack.cc b/debian/htdig/htdig-3.2.0b6/htlib/Stack.cc new file mode 100644 index 00000000..5d74519b --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Stack.cc @@ -0,0 +1,111 @@ +// +// Stack.cc +// +// Stack: This class implements a linked list of objects. It itself is also an +// object +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Stack.cc,v 1.5 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "Stack.h" + +struct stacknode +{ + stacknode *next; + Object *obj; +}; + +//*************************************************************************** +// Stack::Stack() +// +Stack::Stack() +{ + sp = 0; + size = 0; +} + + +//*************************************************************************** +// Stack::~Stack() +// +Stack::~Stack() +{ + while (sp) + { + Object *obj = pop(); + delete obj; + } +} + + +//*************************************************************************** +// void Stack::destroy() +// +void Stack::destroy() +{ + while (sp) + { + Object *obj = pop(); + delete obj; + } +} + + +//*************************************************************************** +// void Stack::push(Object *obj) +// PURPOSE: +// Push an object onto the stack. +// +void Stack::push(Object *obj) +{ + stacknode *node = new stacknode; + + node->obj = obj; + node->next = (stacknode *) sp; + sp = node; + size++; +} + + +//*************************************************************************** +// Object *Stack::pop() +// PURPOSE: +// Return the object at the top of the stack and remove it from the stack. +// +Object *Stack::pop() +{ + if (size == 0) + return 0; + + stacknode *node = (stacknode *) sp; + Object *obj = node->obj; + sp = (void *) node->next; + delete node; + size--; + + return obj; +} + + +//*************************************************************************** +// Object *Stack::peek() +// PURPOSE: +// Return the object at the top of the stack. +// +Object *Stack::peek() +{ + if (size == 0) + return 0; + + return ((stacknode *)sp)->obj; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/Stack.h b/debian/htdig/htdig-3.2.0b6/htlib/Stack.h new file mode 100644 index 00000000..82bbc8d2 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/Stack.h @@ -0,0 +1,52 @@ +// +// Stack.h +// +// Stack: This class implements a linked list of objects. It itself is also an +// object +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: Stack.h,v 1.6 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _Stack_h_ +#define _Stack_h_ + +#include "Object.h" + +class Stack : public Object +{ +public: + // + // Constructors/Destructor + // + Stack(); + ~Stack(); + + // + // Stack access + // + void push(Object *obj); + Object *peek(); + Object *pop(); + int Size() {return size;} + + // + // Stack destruction + // + void destroy(); + +protected: + // + // These variables are to keep track of the linked list + // + void *sp; + + int size; +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/String.cc b/debian/htdig/htdig-3.2.0b6/htlib/String.cc new file mode 100644 index 00000000..918a5881 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/String.cc @@ -0,0 +1,726 @@ +// +// String.cc +// +// String: (interface in htString.h) Just Another String class. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1995-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: String.cc,v 1.40 2004/05/28 13:15:21 lha Exp $ +// +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + + +#include "htString.h" +#include "Object.h" + +#ifndef _MSC_VER /* _WIN32 */ +#include <unistd.h> +#else +#include <io.h> +#endif + +#ifdef HAVE_STD +#include <iostream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <iostream.h> +#endif /* HAVE_STD */ + +#include <ctype.h> +#include <stdio.h> +#include <stdlib.h> + + +const int MinimumAllocationSize = 4; // Should be power of two. + +#ifdef NOINLINE +String::String() +{ + Length = Allocated = 0; + Data = 0; +} +#endif + +String::String(int init) +{ + Length = 0; + Allocated = init >= MinimumAllocationSize ? init : MinimumAllocationSize; + Data = new char[Allocated]; +} + +String::String(const char *s) +{ + Allocated = Length = 0; + Data = 0; + + int len; + if (s) + { + len = strlen(s); + copy(s, len, len); + } +} + +String::String(const char *s, int len) +{ + Allocated = Length = 0; + Data = 0; + if (s && len > 0) + copy(s, len, len); +} + +String::String(const String &s) +{ + Allocated = Length = 0; + Data = 0; + + if (s.length() > 0) + copy(s.Data, s.length(), s.length()); +} + +// +// This can be used for performance reasons if it is known the +// String will need to grow. +// +String::String(const String &s, int allocation_hint) +{ + Allocated = Length = 0; + Data = 0; + + if (s.length() != 0) + { + if (allocation_hint < s.length()) + allocation_hint = s.length(); + copy(s.Data, s.length(), allocation_hint); + } +} + +String::~String() +{ + if (Allocated) + delete [] Data; +} + +void String::operator = (const String &s) +{ + if (s.length() > 0) + { + allocate_space(s.length()); + Length = s.length(); + copy_data_from(s.Data, Length); + } + else + { + Length = 0; + } +} + +void String::operator = (const char *s) +{ + if (s) + { + int len = strlen(s); + allocate_fix_space(len); + Length = len; + copy_data_from(s, Length); + } + else + Length = 0; +} + +void String::append(const String &s) +{ + if (s.length() == 0) + return; + int new_len = Length + s.length(); + + reallocate_space(new_len); + copy_data_from(s.Data, s.length(), Length); + Length = new_len; +} + +void String::append(const char *s) +{ + if (!s) + return; + + append(s,strlen(s)); +} + +void String::append(const char *s, int slen) +{ + if (!s || !slen) + return; + +// if ( slen == 1 ) +// { +// append(*s); +// return; +// } + int new_len = Length + slen; + + if (new_len + 1 > Allocated) + reallocate_space(new_len); + copy_data_from(s, slen, Length); + Length = new_len; +} + +void String::append(char ch) +{ + int new_len = Length +1; + if (new_len + 1 > Allocated) + reallocate_space(new_len); + Data[Length] = ch; + Length = new_len; +} + +int String::compare(const String& obj) const +{ + int len; + int result; + const char *p1 = Data; + const char *p2 = obj.Data; + + len = Length; + result = 0; + + if (Length > obj.Length) + { + result = 1; + len = obj.Length; + } + else if (Length < obj.Length) + result = -1; + + while (len) + { + if (*p1 > *p2) + return 1; + if (*p1 < *p2) + return -1; + p1++; + p2++; + len--; + } + // + // Strings are equal up to the shortest length. + // The result depends upon the length difference. + // + return result; +} + +int String::nocase_compare(const String &s) const +{ + const char *p1 = get(); + const char *p2 = s.get(); + + return mystrcasecmp(p1, p2); +} + +int String::Write(int fd) const +{ + int left = Length; + char *wptr = Data; + + while (left) + { + int result = write(fd, wptr, left); + + if (result < 0) + return result; + + left -= result; + wptr += result; + } + return left; +} + +const char *String::get() const +{ + static const char *null = ""; + if (!Allocated) + return null; + Data[Length] = '\0'; // We always leave room for this. + return Data; +} + +char *String::get() +{ + static char *null = ""; + if (!Allocated) + return null; + Data[Length] = '\0'; // We always leave room for this. + return Data; +} + +char *String::new_char() const +{ + char *r; + if (!Allocated) + { + r = new char[1]; + *r = '\0'; + return r; + } + Data[Length] = '\0'; // We always leave room for this. + r = new char[Length + 1]; + strcpy(r, Data); + return r; +} + + +int String::as_integer(int def) const +{ + if (Length <= 0) + return def; + Data[Length] = '\0'; + return atoi(Data); +} + +double String::as_double(double def) const +{ + if (Length <= 0) + return def; + Data[Length] = '\0'; + return atof(Data); +} + +String String::sub(int start, int len) const +{ + if (start > Length) + return 0; + + if (len > Length - start) + len = Length - start; + + return String(Data + start, len); +} + +String String::sub(int start) const +{ + return sub(start, Length - start); +} + +int String::indexOf(const char *str) const +{ + char *c; + // + // Set the first char after string end to zero to prevent finding + // substrings including symbols after actual end of string + // + if (!Allocated) + return -1; + Data[Length] = '\0'; + + /* OLD CODE: for (i = 0; i < Length; i++) */ +#ifdef HAVE_STRSTR + if ((c = strstr(Data, str)) != NULL) + return(c -Data); +#else + int len = strlen(str); + int i; + for (i = 0; i <= Length-len; i++) + { + if (strncmp(&Data[i], str, len) == 0) + return i; + } +#endif + return -1; +} + +int String::indexOf(char ch) const +{ + int i; + for (i = 0; i < Length; i++) + { + if (Data[i] == ch) + return i; + } + return -1; +} + +int String::indexOf(char ch, int pos) const +{ + if (pos >= Length) + return -1; + for (int i = pos; i < Length; i++) + { + if (Data[i] == ch) + return i; + } + return -1; +} + +int String::lastIndexOf(char ch, int pos) const +{ + if (pos >= Length) + return -1; + while (pos >= 0) + { + if (Data[pos] == ch) + return pos; + pos--; + } + return -1; +} + +int String::lastIndexOf(char ch) const +{ + return lastIndexOf(ch, Length - 1); +} +#ifdef NOINLINE +String &String::operator << (const char *str) +{ + append(str); + return *this; +} + +String &String::operator << (char ch) +{ + append(&ch, 1); + return *this; +} +#endif + +String &String::operator << (int i) +{ + char str[20]; + sprintf(str, "%d", i); + append(str); + return *this; +} + +String &String::operator << (unsigned int i) +{ + char str[20]; + sprintf(str, "%u", i); + append(str); + return *this; +} + +String &String::operator << (long l) +{ + char str[20]; + sprintf(str, "%ld", l); + append(str); + return *this; +} + +String &String::operator << (const String &s) +{ + append(s.get(), s.length()); + return *this; +} + +char String::operator >> (char c) +{ + c = '\0'; + + if (Allocated && Length) + { + c = Data[Length - 1]; + Data[Length - 1] = '\0'; + Length--; + } + + return c; +} + +int String::lowercase() +{ + int converted = 0; + for (int i = 0; i < Length; i++) + { + if (isupper((unsigned char)Data[i])) { + Data[i] = tolower((unsigned char)Data[i]); + converted++; + } + } + return converted; +} + + +int String::uppercase() +{ + int converted = 0; + for (int i = 0; i < Length; i++) + { + if (islower((unsigned char)Data[i])) { + Data[i] = toupper((unsigned char)Data[i]); + converted++; + } + } + return converted; +} + + +void String::replace(char c1, char c2) +{ + for (int i = 0; i < Length; i++) + if (Data[i] == c1) + Data[i] = c2; +} + + +int String::remove(const char *chars) +{ + if (Length <= 0) + return 0; + + char *good, *bad; + int skipped = 0; + + good = bad = Data; + for (int i = 0; i < Length; i++) + { + if (strchr(chars, *bad)) + skipped++; + else + *good++ = *bad; + bad++; + } + Length -= skipped; + + return skipped; +} + +String &String::chop(int n) +{ + Length -= n; + if (Length < 0) + Length = 0; + return *this; +} + + +String &String::chop(char ch) +{ + while (Length > 0 && Data[Length - 1] == ch) + Length--; + return *this; +} + + +String &String::chop(const char *str) +{ + while (Length > 0 && strchr(str, Data[Length - 1])) + Length--; + return *this; +} + + +void String::Serialize(String &dest) +{ + dest.append((char *) &Length, sizeof(Length)); + dest.append(get(), Length); +} + + +void String::Deserialize(String &source, int &index) +{ + memcpy((char *) &Length, (char *) source.get() + index, sizeof(Length)); + index += sizeof(Length); + allocate_fix_space(Length); + copy_data_from(source.get() + index, Length); + index += Length; +} + + +//------------------------------------------------------------------------ +// Non member operators. +// +String operator + (const String &a, const String &b) +{ + String result(a, a.length() + b.length()); + + result.append(b); + return result; +} + +int operator == (const String &a, const String &b) +{ + if (a.Length != b.Length) + return 0; + + return a.compare(b) == 0; +} + +int operator != (const String &a, const String &b) +{ + return a.compare(b) != 0; +} + +int operator < (const String &a, const String &b) +{ + return a.compare(b) == -1; +} + +int operator > (const String &a, const String &b) +{ + return a.compare(b) == 1; +} + +int operator <= (const String &a, const String &b) +{ + return a.compare(b) <= 0; +} + +int operator >= (const String &a, const String &b) +{ + return a.compare(b) >= 0; +} + +#ifndef NOSTREAM +ostream &operator << (ostream &o, const String &s) +{ + o.write(s.Data, s.length()); + return o; +} +#endif /* NOSTREAM */ + +//------------------------------------------------------------------------ +// Private Methods. +// + +void String::copy_data_from(const char *s, int len, int dest_offset) +{ + memcpy(Data + dest_offset, s, len); +} + +void String::allocate_space(int len) +{ + len++; // In case we want to add a null. + + if (len <= Allocated) + return; + + if (Allocated) + delete [] Data; + + Allocated = MinimumAllocationSize; + while (Allocated < len) + Allocated <<= 1; + + Data = new char[Allocated]; +} + +void String::allocate_fix_space(int len) +{ + len++; // In case we want to add a null. + + if (len <= Allocated) + return; + + if (Allocated) + delete [] Data; + + Allocated = len; + if (Allocated < MinimumAllocationSize) + Allocated = MinimumAllocationSize; + Data = new char[Allocated]; +} + +void String::reallocate_space(int len) +{ + char *old_data = 0; + int old_data_len = 0; + + if (Allocated) + { + old_data = Data; + old_data_len = Length; + Allocated = 0; + } + allocate_space(len); + if (old_data) + { + copy_data_from(old_data, old_data_len); + delete [] old_data; + } +} + +void String::copy(const char *s, int len, int allocation_hint) +{ + if (len == 0 || allocation_hint == 0) + return; // We're not actually copying anything! + allocate_fix_space(allocation_hint); + Length = len; + copy_data_from(s, len); +} + +#ifndef NOSTREAM +void String::debug(ostream &o) +{ + o << "Length: " << Length << " Allocated: " << Allocated << + " Data: " << ((void*) Data) << " '" << *this << "'\n"; +} +#endif /* NOSTREAM */ + +int String::readLine(FILE *in) +{ + Length = 0; + allocate_fix_space(2048); + + while (fgets(Data + Length, Allocated - Length, in)) + { + Length += strlen(Data + Length); + if (Length == 0) + continue; + if (Data[Length - 1] == '\n') + { + // + // A full line has been read. Return it. + // + chop('\n'); + return 1; + } + if (Allocated > Length + 1) + { + // + // Not all available space filled. Probably EOF? + // + continue; + } + // + // Only a partial line was read. Increase available space in + // string and read some more. + // + reallocate_space(Allocated << 1); + } + chop('\n'); + + return Length > 0; +} + +#ifndef NOSTREAM +istream &operator >> (istream &in, String &line) +{ + line.Length = 0; + line.allocate_fix_space(2048); + + for (;;) + { + in.clear(); + in.getline(line.Data + line.Length, line.Allocated - line.Length); + line.Length += strlen(line.Data + line.Length); + // if read whole line, or eof, or read fewer chars than the max... + if (!in.fail() || in.eof() || line.Length + 1 < line.Allocated) + break; + // + // Only a partial line was read. Increase available space in + // string and read some more. + // + line.reallocate_space(line.Allocated << 1); + } + + return in; +} +#endif /* NOSTREAM */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/StringList.cc b/debian/htdig/htdig-3.2.0b6/htlib/StringList.cc new file mode 100644 index 00000000..28e03a4c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/StringList.cc @@ -0,0 +1,192 @@ +// +// StringList.cc +// +// StringList: Specialized List containing String objects. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: StringList.cc,v 1.14 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "StringList.h" +#include "htString.h" +#include "List.h" + +#include <stdlib.h> + + +//***************************************************************************** +// StringList::StringList() +// +StringList::StringList() +{ +} + +//***************************************************************************** +// int StringList::Create(const char *str, char *sep) +// +int StringList::Create(const char *str, const char *sep) +{ + String word; + + while (str && *str) + { + if (strchr(sep, *str)) + { + if (word.length()) + { + List::Add(new String(word)); + word = 0; + } + } + else + word << *str; + str++; + } + + // + // Add the last word to the list + // + if (word.length()) + List::Add(new String(word)); + return Count(); +} + + +//***************************************************************************** +// int StringList::Create(const char *str, char sep) +// +int StringList::Create(const char *str, char sep) +{ + String word; + + while (str && *str) + { + if (*str == sep) + { + if (word.length()) + { + List::Add(new String(word)); + word = 0; + } + } + else + word << *str; + str++; + } + + // + // Add the last word to the list + // + if (word.length()) + List::Add(new String(word)); + return Count(); +} + + +//***************************************************************************** +// char *StringList::operator [] (int n) +// +char *StringList::operator [] (int n) +{ + String *str = (String *) Nth(n); + if (str) + return str->get(); + else + return 0; +} + + +//***************************************************************************** +// void StringList::Add(const char *str) +// +void StringList::Add(const char *str) +{ + List::Add(new String(str)); +} + + +//***************************************************************************** +// void StringList::Assign(const char *str, int pos) +// +void StringList::Assign(const char *str, int pos) +{ + List::Assign(new String(str), pos); +} + +//***************************************************************************** +// void StringList::Insert(const char *str, int pos) +// +void StringList::Insert(const char *str, int pos) +{ + List::Insert(new String(str), pos); +} + +//***************************************************************************** +// static int StringCompare(const void *a, const void *b) +// +static int StringCompare(const void *a, const void *b) +{ + String *sa, *sb; + + sa = *((String **) a); + sb = *((String **) b); + + return strcmp(sa->get(), sb->get()); +} + + +//***************************************************************************** +// void StringList::Sort(int direction) +// +void StringList::Sort(int) +{ + String **array = new String*[Count()]; + int i; + int n = Count(); + + ListCursor cursor; + + Start_Get(cursor); + Object *obj; + for(i = 0; i < n && (obj = Get_Next(cursor)); i++) { + array[i] = (String*)obj; + } + + qsort((char *) array, (size_t) n, (size_t) sizeof(String *), + StringCompare); + + Release(); + + for (i = 0; i < n; i++) + { + List::Add(array[i]); + } + + delete array; +} + +//***************************************************************************** +// String StringList::Join(char sep) const +// +String StringList::Join(char sep) const +{ + String str; + int i; + + for (i=0; i < number; i++) + { + if (str.length()) + str.append(sep); + str.append(*((const String *) Nth(i))); + } + return str; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/StringList.h b/debian/htdig/htdig-3.2.0b6/htlib/StringList.h new file mode 100644 index 00000000..63a8dc54 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/StringList.h @@ -0,0 +1,73 @@ +// +// StringList.h +// +// StringList: Specialized List containing String objects. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: StringList.h,v 1.12 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _StringList_h_ +#define _StringList_h_ + +#include "Object.h" +#include "List.h" +#include "htString.h" + + +class StringList : public List +{ +public: + // + // Construction/Destruction + // + StringList(); + + // + // Creation of a String from a string or String + // + StringList(const char *str, char sep = '\t') { Create(str, sep); } + StringList(const String &str, char sep = '\t') { Create(str, sep); } + StringList(const char *str, const char *sep) { Create(str, sep); } + StringList(const String &str, const char *sep) { Create(str, sep); } + + int Create(const char *str, char sep = '\t'); + int Create(const String &str, char sep = '\t') { return Create(str.get(), sep); } + int Create(const char *str, const char *sep); + int Create(const String &str, const char *sep) { return Create(str.get(), sep); } + + // + // Standard List operations... + // + void Add(const char *); + void Add(String *obj) { List::Add(obj); } + void Insert(const char *, int pos); + void Insert(String *obj, int pos) { List::Insert(obj, pos); } + void Assign(const char *, int pos); + void Assign(String *obj, int pos) { List::Assign(obj, pos); } + + // + // Since we know we only store strings, we can reliably sort them. + // If direction is 1, the sort will be in descending order + // + void Sort(int direction = 0); + + // + // Join the Elements of the StringList together + // + String Join(char) const; + + // + // Getting at the parts of the StringList + // + char *operator [] (int n); + +private: +}; + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/StringMatch.cc b/debian/htdig/htdig-3.2.0b6/htlib/StringMatch.cc new file mode 100644 index 00000000..b1512cc3 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/StringMatch.cc @@ -0,0 +1,601 @@ +// +// StringMatch.cc +// +// StringMatch: This class provides an interface to a fairly specialized string +// lookup facility. It is intended to be used as a replace for any +// regualr expression matching when the pattern string is in the form: +// +// <string1>|<string2>|<string3>|... +// +// Just like regular expression routines, the pattern needs to be +// compiled before it can be used. This is done using the Pattern() +// member function. Once the pattern has been compiled, the member +// function Find() can be used to search for the pattern in a string. +// If a string has been found, the "which" and "length" parameters +// will be set to the string index and string length respectively. +// (The string index is counted starting from 0) The return value of +// Find() is the position at which the string was found or -1 if no +// strings could be found. If a case insensitive match needs to be +// performed, call the IgnoreCase() member function before calling +// Pattern(). This function will setup a character translation table +// which will convert all uppercase characters to lowercase. If some +// other translation is required, the TranslationTable() member +// function can be called to provide a custom table. This table needs +// to be 256 characters. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: StringMatch.cc,v 1.18 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "StringMatch.h" + +#include <string.h> +#include <ctype.h> + +#ifdef HAVE_STD +#include <fstream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <fstream.h> +#endif /* HAVE_STD */ + +// +// Entries in the state table can either be normal or final. +// Final states have an match index encoded in them. This number +// is shifted left by INDEX_SHIFT bits. +// +#define MATCH_INDEX_MASK 0xffff0000 +#define STATE_MASK 0x0000ffff +#define INDEX_SHIFT 16 + +//***************************************************************************** +// StringMatch::StringMatch() +// +StringMatch::StringMatch() +{ + // + // Clear out the state table pointers + // + for (int i = 0; i < 256; i++) + table[i] = 0; + local_alloc = 0; + trans = 0; +} + + +//***************************************************************************** +// StringMatch::~StringMatch() +// +StringMatch::~StringMatch() +{ + for (int i = 0; i < 256; i++) + delete [] table[i]; + if (local_alloc) + delete [] trans; +} + + +//***************************************************************************** +// void StringMatch::Pattern(char *pattern) +// Compile the given pattern into a state transition table +// +void +StringMatch::Pattern(char *pattern, char sep) +{ + if (!pattern || !*pattern) + { + // + // No pattern to compile... + // + return; + } + + // + // Allocate enough space in the state table to hold the worst case + // patterns... + // + int n = strlen(pattern); + + // ...but since the state table does not need an extra state + // for each string in the pattern, we can subtract the number + // of separators. Wins for small but numerous strings in + // the pattern. + char *tmpstr; + for (tmpstr = pattern; + (tmpstr = strchr(tmpstr, sep)) != NULL; + tmpstr++) // Pass the separator. + n--; + + int i; + + for (i = 0; i < 256; i++) + { + table[i] = new int[n]; + memset((unsigned char *) table[i], 0, n * sizeof(int)); + } + for (i = 0; i < n; i++) + table[0][i] = i; // "no-op" states for null char, to be ignored + + // + // Set up a standard case translation table if needed. + // + if (!trans) + { + trans = new unsigned char[256]; + for (i = 0; i < 256; i++) + { + trans[i] = (unsigned char)i; + } + local_alloc = 1; + } + + // + // Go though each of the patterns and build entries in the table. + // + int state = 0; + int totalStates = 0; + unsigned char previous = 0; + int previousState = 0; + int previousValue = 0; + int index = 1; + unsigned char chr; + + while ((unsigned char)*pattern) + { +#if 0 + if (totalStates > n) + { + cerr << "Fatal! Miscalculation of number of states" + << endl; + exit (2); + } +#endif + + chr = trans[(unsigned char)*pattern]; + if (chr == 0) + { + pattern++; + continue; + } + if (chr == sep) + { + // + // Next pattern + // + table[previous][previousState] = + previousValue | (index << INDEX_SHIFT); + index++; + state = 0; + // totalStates--; + } + else + { + previousValue = table[chr][state]; + previousState = state; + if (previousValue) + { + if (previousValue & MATCH_INDEX_MASK) + { + if (previousValue & STATE_MASK) + { + state = previousValue & STATE_MASK; + } + else + { + table[chr][state] |= ++totalStates; + state = totalStates; + } + } + else + { + state = previousValue & STATE_MASK; + } + } + else + { + table[chr][state] = ++totalStates; + state = totalStates; + } + } + previous = chr; + pattern++; + } + table[previous][previousState] = + previousValue | (index << INDEX_SHIFT); +} + + +//***************************************************************************** +// int StringMatch::FindFirst(const char *string, int &which, int &length) +// Attempt to find the first occurance of the previous compiled patterns. +// +int StringMatch::FindFirst(const char *string, int &which, int &length) +{ + which = -1; + length = -1; + + if (!table[0]) + return 0; + + int state = 0, new_state = 0; + int pos = 0; + int start_pos = 0; + + while ((unsigned char)string[pos]) + { + new_state = table[trans[(unsigned char)string[pos] & 0xff]][state]; + if (new_state) + { + if (state == 0) + { + // + // Keep track of where we started comparing so that we can + // come back to this point later if we didn't match anything + // + start_pos = pos; + } + } + else + { + // + // We came back to 0 state. This means we didn't match anything. + // + if (state) + { + // But we may already have a match, and are just being greedy. + if (which != -1) + return start_pos; + + pos = start_pos + 1; + } + else + pos++; + state = 0; + continue; + } + state = new_state; + if (state & MATCH_INDEX_MASK) + { + // + // Matched one of the patterns. + // Determine which and return. + // + which = ((unsigned int) (state & MATCH_INDEX_MASK) + >> INDEX_SHIFT) - 1; + length = pos - start_pos + 1; + state &= STATE_MASK; + + // Continue to find the longest, if there is one. + if (state == 0) + return start_pos; + } + pos++; + } + + // Maybe we were too greedy. + if (which != -1) + return start_pos; + + return -1; +} + + +//***************************************************************************** +// int StringMatch::Compare(const char *string, int &which, int &length) +// +int StringMatch::Compare(const char *string, int &which, int &length) +{ + which = -1; + length = -1; + + if (!table[0]) + return 0; + + int state = 0, new_state = 0; + int pos = 0; + int start_pos = 0; + + // + // Skip to at least the start of a word. + // + while ((unsigned char)string[pos]) + { + new_state = table[trans[string[pos]]][state]; + if (new_state) + { + if (state == 0) + { + start_pos = pos; + } + } + else + { + // We may already have a match, and are just being greedy. + if (which != -1) + return 1; + + return 0; + } + state = new_state; + if (state & MATCH_INDEX_MASK) + { + // + // Matched one of the patterns. + // + which = ((unsigned int) (state & MATCH_INDEX_MASK) + >> INDEX_SHIFT) - 1; + length = pos - start_pos + 1; + + // Continue to find the longest, if there is one. + state &= STATE_MASK; + if (state == 0) + return 1; + } + pos++; + } + + // Maybe we were too greedy. + if (which != -1) + return 1; + + return 0; +} + + +//***************************************************************************** +// int StringMatch::FindFirstWord(char *string) +// +int StringMatch::FindFirstWord(const char *string) +{ + int dummy; + return FindFirstWord(string, dummy, dummy); +} + + +//***************************************************************************** +// int StringMatch::CompareWord(const char *string) +// +int StringMatch::CompareWord(const char *string) +{ + int dummy; + return CompareWord(string, dummy, dummy); +} + + +//***************************************************************************** +// int StringMatch::FindFirstWord(char *string, int &which, int &length) +// Attempt to find the first occurance of the previous compiled patterns. +// +int StringMatch::FindFirstWord(const char *string, int &which, int &length) +{ + which = -1; + length = -1; + + int state = 0, new_state = 0; + int pos = 0; + int start_pos = 0; + int is_word = 1; + + // + // Skip to at least the start of a word. + // + while ((unsigned char)string[pos]) + { + new_state = table[trans[(unsigned char)string[pos]]][state]; + if (new_state) + { + if (state == 0) + { + start_pos = pos; + } + } + else + { + // + // We came back to 0 state. This means we didn't match anything. + // + if (state) + { + pos = start_pos + 1; + } + else + pos++; + state = 0; + continue; + } + state = new_state; + + if (state & MATCH_INDEX_MASK) + { + // + // Matched one of the patterns. + // + is_word = 1; + if (start_pos != 0) + { + if (HtIsStrictWordChar((unsigned char)string[start_pos - 1])) + is_word = 0; + } + if (HtIsStrictWordChar((unsigned char)string[pos + 1])) + is_word = 0; + if (is_word) + { + // + // Determine which and return. + // + which = ((unsigned int) (state & MATCH_INDEX_MASK) + >> INDEX_SHIFT) - 1; + length = pos - start_pos + 1; + return start_pos; + } + else + { + // + // Not at the end of word. Continue searching. + // + if (state & STATE_MASK) + { + state &= STATE_MASK; + } + else + { + pos = start_pos + 1; + state = 0; + } + } + } + pos++; + } + return -1; +} + + +//***************************************************************************** +// int StringMatch::CompareWord(const char *string, int &which, int &length) +// +int StringMatch::CompareWord(const char *string, int &which, int &length) +{ + which = -1; + length = -1; + + if (!table[0]) + return 0; + + int state = 0; + int position = 0; + + // + // Skip to at least the start of a word. + // + while ((unsigned char)string[position]) + { + state = table[trans[(unsigned char)string[position]]][state]; + if (state == 0) + { + return 0; + } + + if (state & MATCH_INDEX_MASK) + { + // + // Matched one of the patterns. See if it is a word. + // + int isWord = 1; + + if ((unsigned char)string[position + 1]) + { + if (HtIsStrictWordChar((unsigned char)string[position + 1])) + isWord = 0; + } + + if (isWord) + { + which = ((unsigned int) (state & MATCH_INDEX_MASK) + >> INDEX_SHIFT) - 1; + length = position + 1; + return 1; + } + else + { + // + // Not at the end of a word. Continue searching. + // + if ((state & STATE_MASK) != 0) + { + state &= STATE_MASK; + } + else + { + return 0; + } + } + } + position++; + } + return 0; +} + + +//***************************************************************************** +// void StringMatch::TranslationTable(char *table) +// +void StringMatch::TranslationTable(char *table) +{ + if (local_alloc) + delete [] trans; + trans = (unsigned char *) table; + local_alloc = 0; +} + + +//***************************************************************************** +// void StringMatch::IgnoreCase() +// Set up the case translation table to convert uppercase to lowercase +// +void StringMatch::IgnoreCase() +{ + if (!local_alloc || !trans) + { + trans = new unsigned char[256]; + for (int i = 0; i < 256; i++) + trans[i] = (unsigned char)i; + local_alloc = 1; + } + for (int i = 0; i < 256; i++) + if (isupper((unsigned char)i)) + trans[i] = tolower((unsigned char)i); +} + + +//***************************************************************************** +// void StringMatch::IgnorePunct(char *punct) +// Set up the character translation table to ignore punctuation +// +void StringMatch::IgnorePunct(char *punct) +{ + if (!local_alloc || !trans) + { + trans = new unsigned char[256]; + for (int i = 0; i < 256; i++) + trans[i] = (unsigned char)i; + local_alloc = 1; + } + if (punct) + for (int i = 0; punct[i]; i++) + trans[(unsigned char)punct[i]] = 0; + else + for (int i = 0; i < 256; i++) + if (HtIsWordChar(i) && !HtIsStrictWordChar(i)) + trans[i] = 0; +} + + +//***************************************************************************** +// int StringMatch::FindFirst(const char *source) +// +int StringMatch::FindFirst(const char *source) +{ + int dummy; + return FindFirst(source, dummy, dummy); +} + + +//***************************************************************************** +// int StringMatch::Compare(const char *source) +// +int StringMatch::Compare(const char *source) +{ + int dummy; + return Compare(source, dummy, dummy); +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/StringMatch.h b/debian/htdig/htdig-3.2.0b6/htlib/StringMatch.h new file mode 100644 index 00000000..d848fe5e --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/StringMatch.h @@ -0,0 +1,116 @@ +// +// StringMatch.h +// +// StringMatch: This class provides an interface to a fairly specialized string +// lookup facility. It is intended to be used as a replace for any +// regular expression matching when the pattern string is in the form: +// +// <string1>|<string2>|<string3>|... +// +// Just like regular expression routines, the pattern needs to be +// compiled before it can be used. This is done using the Pattern() +// member function. Once the pattern has been compiled, the member +// function Find() can be used to search for the pattern in a string. +// If a string has been found, the "which" and "length" parameters +// will be set to the string index and string length respectively. +// (The string index is counted starting from 0) The return value of +// Find() is the position at which the string was found or -1 if no +// strings could be found. If a case insensitive match needs to be +// performed, call the IgnoreCase() member function before calling +// Pattern(). This function will setup a character translation table +// which will convert all uppercase characters to lowercase. If some +// other translation is required, the TranslationTable() member +// function can be called to provide a custom table. This table needs +// to be 256 characters. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: StringMatch.h,v 1.13 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _StringMatch_h_ +#define _StringMatch_h_ + +#include "Object.h" +#include "HtWordType.h" + +class StringMatch : public Object +{ +public: + // + // Construction/Destruction + // + StringMatch(); + ~StringMatch(); + + // + // Set the pattern to search for. If given as a string needs to + // be in the form <string1>|<string2>|... If in the form of a + // List, it should be a list of String objects. + // + void Pattern(char *pattern, char sep = '|'); + + // + // Search for any of the strings in the pattern in the given + // string The return value is the offset in the source a pattern + // was found. In this case, the which variable will be set to the + // index of the pattern string and length will be set to the + // length of that pattern string. If none of the pattern strings + // could be found, the return value will be -1 + // + int FindFirst(const char *string, int &which, int &length); + int FindFirst(const char *string); + + int FindFirstWord(const char *string, int &which, int &length); + int FindFirstWord(const char *string); + + // + // If you are interested in matching instead of searching, use + // the following. Same parameters except that the return value will + // be 1 if there was a match, 0 if there was not. + // + int Compare(const char *string, int &which, int &length); + int Compare(const char *string); + + int CompareWord(const char *string, int &which, int &length); + int CompareWord(const char *string); + + // + // Provide a character translation table which will be applied to + // both the pattern and the input string. This table should be an + // array of 256 characters. If is the caller's responsibility to + // manage this table's allocation. The table should remain valid + // until this object has been destroyed. + // + void TranslationTable(char *table); + + // + // Build a local translation table which maps all uppercase + // characters to lowercase + // + void IgnoreCase(); + + // + // Build a local translation table which ignores all given punctuation + // characters + // + void IgnorePunct(char *punct = (char*)NULL); + + // + // Determine if there is a pattern associated with this Match object. + // + int hasPattern() {return table[0] != 0;} + +protected: + int *table[256]; + unsigned char *trans; + int local_alloc; +}; + +#endif + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/String_fmt.cc b/debian/htdig/htdig-3.2.0b6/htlib/String_fmt.cc new file mode 100644 index 00000000..753e23cb --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/String_fmt.cc @@ -0,0 +1,54 @@ +// +// String_fmt.cc +// +// String_fmt: Formatting functions for the String class. Those functions +// are also used in other files, they are not purely internal +// to the String class. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: String_fmt.cc,v 1.11 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "htString.h" + +#include <stdarg.h> +#include <stdio.h> + +#ifdef _MSC_VER /* _WIN32 */ +#define vsnprintf _vsnprintf +#endif + +static char buf[10000]; + +//***************************************************************************** +// char *form(char *fmt, ...) +// +char *form(const char *fmt, ...) +{ + va_list args; + va_start(args, fmt); + vsnprintf(buf, sizeof(buf), fmt, args); + va_end(args); + return buf; +} + + +//***************************************************************************** +// char *vform(char *fmt, va_list args) +// +char *vform(const char *fmt, va_list args) +{ + vsnprintf(buf, sizeof(buf), fmt, args); + return buf; +} + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/ber.h b/debian/htdig/htdig-3.2.0b6/htlib/ber.h new file mode 100644 index 00000000..fc0af86d --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/ber.h @@ -0,0 +1,85 @@ +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// + +#ifndef _ber_h +#define _ber_h + +#include <stdio.h> +#include <errno.h> + +typedef unsigned int ber_t; + +#define BER_MAX_BYTES (sizeof(ber_t) + 1) + +inline int ber_buf2value(const unsigned char* buf, int buf_len, ber_t& result) { + result = 0; + unsigned int bits = 0; + int length = 1; + while(*buf & 0x80) { + if(bits > sizeof(ber_t) * 8) return EINVAL; + result |= (*buf & 0x7f) << bits; + bits += 7; + buf++; + length++; + if(length > buf_len) return EINVAL; + } + result |= (*buf & 0x7f) << bits; + + return length; +} + +inline int ber_file2value(FILE* fp, ber_t& result) { + result = 0; + unsigned int bits = 0; + int c; + int length = 1; + while((c = fgetc(fp)) != EOF && (c & 0x80)) { + if(bits > sizeof(ber_t) * 8) return EINVAL; + result |= (c & 0x7f) << bits; + bits += 7; + length++; + } + + if(c == EOF) return EINVAL; + + result |= (c & 0x7f) << bits; + + return length; +} + +inline int ber_value2buf(unsigned char* buf, int buf_len, ber_t value) +{ + if(buf_len <= 0) return EINVAL; + int buf_idx = 0; + buf[buf_idx++] = (value & 0x7f); + while(value >>= 7) { + if(buf_idx >= buf_len) return EINVAL; + buf[buf_idx - 1] |= 0x80; + buf[buf_idx++] = (value & 0x7f); + } + return buf_idx; +} + +inline int ber_value2file(FILE* fp, ber_t value) +{ + int length = 1; + unsigned char current; + current = (value & 0x7f); + while(value >>= 7) { + current |= 0x80; + if(fputc(current, fp) == EOF) return EINVAL; + current = (value & 0x7f); + length++; + } + + if(fputc(current, fp) == EOF) return EINVAL; + + return length; +} + +#endif /* _ber_h */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/clib.h b/debian/htdig/htdig-3.2.0b6/htlib/clib.h new file mode 100644 index 00000000..04fc5403 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/clib.h @@ -0,0 +1,49 @@ +/* + Part of the ht://Dig package <http://www.htdig.org/> + Copyright (c) 1999-2004 The ht://Dig Group + For copyright details, see the file COPYING in your distribution + or the GNU Library General Public License (LGPL) version 2 or later + <http://www.gnu.org/copyleft/lgpl.html> +*/ +#ifndef _clib_h_ +#define _clib_h_ + +#include <sys/types.h> + +extern "C" +{ + +#ifndef HAVE_GETCWD +char *getcwd(char *, size_t); +#endif + +#ifndef HAVE_MEMCMP +int memcmp(const void *, const void *, size_t); +#endif + +#ifndef HAVE_MEMCPY +void *memcpy(void *, const void *, size_t); +#endif + +#ifndef HAVE_MEMMOVE +void *memmove(void *, const void *, size_t); +#endif + +#ifndef HAVE_RAISE +int raise (int); +#endif + +#ifndef HAVE_SNPRINTF +int snprintf(char *, size_t, const char *, ...); +#endif + +#ifndef HAVE_STRERROR +char *strerror(int); +#endif + +#ifndef HAVE_VSNPRINTF +int vsnprintf(char *, size_t, const char *, ...); +#endif +} + +#endif /* _clib_h_ */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/filecopy.cc b/debian/htdig/htdig-3.2.0b6/htlib/filecopy.cc new file mode 100644 index 00000000..4cd23e85 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/filecopy.cc @@ -0,0 +1,126 @@ +// +// filecopy.c +// +// Copies files from one file to another. +// Contains both Unix & Native Win32 Implementations +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 2003 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// Copyright (c) 2002 RightNow Technologies, Inc. +// Donated to The ht://Dig Group under LGPL License + +#include <stdio.h> + +#ifdef _WIN32 + +#define WIN32_LEAN_AND_MEAN +#include <windows.h> + +#else /* UNIX */ + +#include <stdio.h> +#include <sys/types.h> +#include <sys/stat.h> +#include <fcntl.h> +#include <unistd.h> +#include <errno.h> + +#endif /* _WIN32 | UNIX */ + + +#include "filecopy.h" + + + +//---------------------------------------------------------------------- +// int file_copy (char * from, char * to, char flags) +//---------------------------------------------------------------------- +// +// copy file 'from' -> 'to' +// +// set flags to FILECOPY_OVERWRITE_ON to overwrite the 'to' file if +// it exists +// +// set flags to FILECOPY_OVERWRITE_OFF to not overwrite the 'to' file +// if it exists +// +// returns 0/FALSE if unsucessful +// returns 1/TRUE if sucessful +// +// +#ifdef _WIN32 + +int file_copy (char * from, char * to, char flags) +{ + if (flags == FILECOPY_OVERWRITE_ON) + { + //overwrite + if (TRUE != (CopyFile(from , to, FALSE))) + return (FALSE); + } + else if (flags == FILECOPY_OVERWRITE_OFF) + { + //don't overwrite + if (TRUE != (CopyFile(from , to, TRUE))) + return (FALSE); + } + else //bad flag + { + return (FALSE); + } + + return (TRUE); +} + +#else //UNIX + +int file_copy (char * from, char * to, char flags) +{ + size_t nmemb; + //int nmemb; + FILE *ifp, *ofp; + char buf[BUFSIZ]; + + if (flags == FILECOPY_OVERWRITE_OFF) { + if (access(to, F_OK) == 0) { + //OUTLOG((FUNC, TRWRN, "file %s already exists\n", to)); + return(FALSE); + } + else if (errno != ENOENT) { + //OUTLOG((FUNC, TRERR, "access(%s, F_OK) failed\n", to)); + return(FALSE); + } + } + + if ((ifp=fopen(from, "r")) == NULL) { + //OUTLOG((FUNC, TRERR, "%s doesn't exist\n", from)); + return(FALSE); + } + + if ((ofp=fopen(to, "w+")) == NULL) { + //OUTLOG((FUNC, TRERR, "can't create %s\n", to)); + fclose(ifp); + return(FALSE); + } + + while ((nmemb=fread(buf, 1, sizeof(buf), ifp)) > 0) { + if (fwrite(buf, 1, nmemb, ofp) != nmemb) { + //OUTLOG((FUNC, TRERR, "fwrite failed\n")); + fclose(ifp); + fclose(ofp); + return(FALSE); + } + } + + fclose(ifp); + fclose(ofp); + + return (TRUE); +} + +#endif /* _WIN32 | UNIX */ + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/filecopy.h b/debian/htdig/htdig-3.2.0b6/htlib/filecopy.h new file mode 100644 index 00000000..c888e720 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/filecopy.h @@ -0,0 +1,40 @@ +// +// filecopy.h +// +// Copies files from one file to another. +// Contains both Unix & Native Win32 Implementations +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 2003 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// Copyright (c) 2002 RightNow Technologies, Inc. +// Donated to The ht://Dig Group under LGPL License + +#ifdef __cplusplus +//extern "C" { +#endif + +#ifndef FILECOPY_H + +#ifndef FALSE +#define FALSE 0 +#endif + +#ifndef TRUE +#define TRUE 1 +#endif + +#define FILECOPY_OVERWRITE_ON 1 +#define FILECOPY_OVERWRITE_OFF 2 + +int file_copy (char * from, char * to, char flags); + + +#ifdef __cplusplus +//} +#endif + +#endif /* FILECOPY_H */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/getcwd.c b/debian/htdig/htdig-3.2.0b6/htlib/getcwd.c new file mode 100644 index 00000000..019c96be --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/getcwd.c @@ -0,0 +1,278 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ +/* + * Copyright (c) 1989, 1991, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include <sys/types.h> +#include <sys/stat.h> + +#if HAVE_DIRENT_H +# include <dirent.h> +# define NAMLEN(dirent) strlen((dirent)->d_name) +#else +# define dirent direct +# define NAMLEN(dirent) (dirent)->d_namlen +# if HAVE_SYS_NDIR_H +# include <sys/ndir.h> +# endif +# if HAVE_SYS_DIR_H +# include <sys/dir.h> +# endif +# if HAVE_NDIR_H +# include <ndir.h> +# endif +#endif + +#include <errno.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> + +#ifndef _MSC_VER /* _WIN32 */ +#include <unistd.h> +#endif + +#ifndef HAVE_GETCWD + +#define ISDOT(dp) \ + (dp->d_name[0] == '.' && (dp->d_name[1] == '\0' || \ + (dp->d_name[1] == '.' && dp->d_name[2] == '\0'))) + +#ifndef dirfd +#define dirfd(dirp) ((dirp)->dd_fd) +#endif + +/* + * getcwd -- + * Get the current working directory. + * + * PUBLIC: #ifndef HAVE_GETCWD + * PUBLIC: char *getcwd __P((char *, size_t)); + * PUBLIC: #endif + */ +char * +getcwd(pt, size) + char *pt; + size_t size; +{ + register struct dirent *dp; + register DIR *dir; + register dev_t dev; + register ino_t ino; + register int first; + register char *bpt, *bup; + struct stat s; + dev_t root_dev; + ino_t root_ino; + size_t ptsize, upsize; + int ret, save_errno; + char *ept, *eup, *up; + + /* + * If no buffer specified by the user, allocate one as necessary. + * If a buffer is specified, the size has to be non-zero. The path + * is built from the end of the buffer backwards. + */ + if (pt) { + ptsize = 0; + if (!size) { + __os_set_errno(EINVAL); + return (NULL); + } + if (size == 1) { + __os_set_errno(ERANGE); + return (NULL); + } + ept = pt + size; + } else { + if ((ret = __os_malloc(ptsize = 1024 - 4, NULL, &pt)) != 0) { + __os_set_errno(ret); + return (NULL); + } + ept = pt + ptsize; + } + bpt = ept - 1; + *bpt = '\0'; + + /* + * Allocate bytes (1024 - malloc space) for the string of "../"'s. + * Should always be enough (it's 340 levels). If it's not, allocate + * as necessary. Special case the first stat, it's ".", not "..". + */ + if ((ret = __os_malloc(upsize = 1024 - 4, NULL, &up)) != 0) + goto err; + eup = up + 1024; + bup = up; + up[0] = '.'; + up[1] = '\0'; + + /* Save root values, so know when to stop. */ + if (stat("/", &s)) + goto err; + root_dev = s.st_dev; + root_ino = s.st_ino; + + __os_set_errno(0); /* XXX readdir has no error return. */ + + for (first = 1;; first = 0) { + /* Stat the current level. */ + if (lstat(up, &s)) + goto err; + + /* Save current node values. */ + ino = s.st_ino; + dev = s.st_dev; + + /* Check for reaching root. */ + if (root_dev == dev && root_ino == ino) { + *--bpt = PATH_SEPARATOR[0]; + /* + * It's unclear that it's a requirement to copy the + * path to the beginning of the buffer, but it's always + * been that way and stuff would probably break. + */ + bcopy(bpt, pt, ept - bpt); + __os_free(up, upsize); + return (pt); + } + + /* + * Build pointer to the parent directory, allocating memory + * as necessary. Max length is 3 for "../", the largest + * possible component name, plus a trailing NULL. + */ + if (bup + 3 + MAXNAMLEN + 1 >= eup) { + if (__os_realloc(upsize *= 2, NULL, &up) != 0) + goto err; + bup = up; + eup = up + upsize; + } + *bup++ = '.'; + *bup++ = '.'; + *bup = '\0'; + + /* Open and stat parent directory. */ + if (!(dir = opendir(up)) || fstat(dirfd(dir), &s)) + goto err; + + /* Add trailing slash for next directory. */ + *bup++ = PATH_SEPARATOR[0]; + + /* + * If it's a mount point, have to stat each element because + * the inode number in the directory is for the entry in the + * parent directory, not the inode number of the mounted file. + */ + save_errno = 0; + if (s.st_dev == dev) { + for (;;) { + if (!(dp = readdir(dir))) + goto notfound; + if (dp->d_fileno == ino) + break; + } + } else + for (;;) { + if (!(dp = readdir(dir))) + goto notfound; + if (ISDOT(dp)) + continue; + bcopy(dp->d_name, bup, dp->d_namlen + 1); + + /* Save the first error for later. */ + if (lstat(up, &s)) { + if (save_errno == 0) + save_errno = __os_get_errno(); + __os_set_errno(0); + continue; + } + if (s.st_dev == dev && s.st_ino == ino) + break; + } + + /* + * Check for length of the current name, preceding slash, + * leading slash. + */ + if (bpt - pt < dp->d_namlen + (first ? 1 : 2)) { + size_t len, off; + + if (!ptsize) { + __os_set_errno(ERANGE); + goto err; + } + off = bpt - pt; + len = ept - bpt; + if (__os_realloc(ptsize *= 2, NULL, &pt) != 0) + goto err; + bpt = pt + off; + ept = pt + ptsize; + bcopy(bpt, ept - len, len); + bpt = ept - len; + } + if (!first) + *--bpt = PATH_SEPARATOR[0]; + bpt -= dp->d_namlen; + bcopy(dp->d_name, bpt, dp->d_namlen); + (void)closedir(dir); + + /* Truncate any file name. */ + *bup = '\0'; + } + +notfound: + /* + * If readdir set errno, use it, not any saved error; otherwise, + * didn't find the current directory in its parent directory, set + * errno to ENOENT. + */ + if (__os_get_errno() == 0) + __os_set_errno(save_errno == 0 ? ENOENT : save_errno); + /* FALLTHROUGH */ +err: + if (ptsize) + __os_free(pt, ptsize); + __os_free(up, upsize); + return (NULL); +} +#endif /* HAVE_GETCWD */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/getopt_local.c b/debian/htdig/htdig-3.2.0b6/htlib/getopt_local.c new file mode 100644 index 00000000..3e73a009 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/getopt_local.c @@ -0,0 +1,122 @@ +/* getopt_local.c */ + +/* Public Domain getopt clone */ + +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 2003 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/* Added to HtDig code June 2003 by Neal Richter, RightNow Technologies */ + +/* +** This getopt behaves pretty much like you would expect. +** It does handle arguments like '-ab-' a little differently +** then normal; I think the -- 'stop option processing' should +** be treated like just another option, so that's what mine does. +** Other getopts seem to ignore the second '-' in '-ab-'. +** +** I hereby place this version of getopt in +** the public domain. Do with this what you will. +** I'm sure there is a nicer and faster version out there +** somewhere but I don't care! +** +** Robert Osborne, May 1991. +*/ + + +#include <stdio.h> + +#include "getopt_local.h" + +#ifdef GETOPT_LOCAL + +int optind = 1; +int opterr = 1; +char *optarg = (char *) 0; + +static char *next_arg = (char *) 0; + +#define NO_OPT 0 +#define OPT_PLAIN 1 +#define OPT_ARG 2 + + +/* ----- getopt -------------------------------------- Oct 23, 1999 21:48 --- + */ +int +getopt(int argc, char *argv[], char *optstring) +{ + int ret; + int which = NO_OPT; + + if (next_arg == (char*) 0) + { + if (argv[optind] == (char *) 0 || argv[optind][0] != '-') + return -1; + next_arg = &argv[optind][1]; + } + + if ((*next_arg == '\0') || (*next_arg == '-')) + { + optind++; + return -1; + } + + while(*optstring) + if (*next_arg == *optstring++) + which = (*optstring == ':') ? OPT_ARG : OPT_PLAIN; + + switch (which) + { + case NO_OPT: + case OPT_PLAIN: + ret = *next_arg++; + + if (*next_arg == '\0') + { + optind++; + next_arg = (char *)0; + } + + if (which == OPT_PLAIN) + return ret; + + if (opterr) + fprintf(stderr, "%s: illegal option -- %c\n", argv[0], ret); + + return '?'; + + case OPT_ARG: + ret = *next_arg++; + optind++; + + if (*next_arg != '\0') + { + optarg = next_arg; + next_arg = (char*) 0; + return ret; + } + + if (argv[optind] != (char*) 0) + { + optarg = argv[optind]; + optind++; + next_arg = (char*) 0; + return ret; + } + + next_arg = (char*) 0; + if (opterr) + fprintf(stderr, "%s: option requires an option -- %c\n", + argv[0], ret); + return '?'; + } + + return(-1); +} +#elif defined(_MSC_VER) /* _WIN32 */ +#error _MSC_VER but !GETOPT_LOCAL +#endif /* GETOPT_LOCAL */ + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/getopt_local.h b/debian/htdig/htdig-3.2.0b6/htlib/getopt_local.h new file mode 100644 index 00000000..731a21db --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/getopt_local.h @@ -0,0 +1,50 @@ +/* getopt_local.h */ + +/* Public Domain getopt clone */ + +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 2003 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/* Added to HtDig code June 2003 by Neal Richter, RightNow Technologies */ + +/* +** This getopt behaves pretty much like you would expect. +** It does handle arguments like '-ab-' a little differently +** then normal; I think the -- 'stop option processing' should +** be treated like just another option, so that's what mine does. +** Other getopts seem to ignore the second '-' in '-ab-'. +** +** I hereby place this version of getopt in +** the public domain. Do with this what you will. +** I'm sure there is a nicer and faster version out there +** somewhere but I don't care! +** +** Robert Osborne, May 1991. +*/ + + +#ifndef GETOPT_LOCAL_H +#define GETOPT_LOCAL_H + +#define GETOPT_LOCAL + +#ifdef __cplusplus +extern "C" { +#endif + +/* header for getopt_local.c */ + +extern int optind; +extern int opterr; +extern char *optarg; + +int getopt(int, char *[], char *); + +#ifdef __cplusplus +} +#endif + +#endif /* GETOPT_LOCAL_H */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/good_strtok.cc b/debian/htdig/htdig-3.2.0b6/htlib/good_strtok.cc new file mode 100644 index 00000000..ad537f56 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/good_strtok.cc @@ -0,0 +1,46 @@ +// +// good_strtok.cc +// +// good_strtok: The good_strtok() function is very similar to the +// standard strtok() library function, except that good_strtok() +// will only skip over 1 separator if it finds one. This is +// needed when parsing strings with empty fields. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: good_strtok.cc,v 1.7 2004/05/28 13:15:21 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "lib.h" + +// +// Perform the same function as the standard strtok() function except that +// multiple separators are NOT collapsed into one. +// +char *good_strtok(char *str, char term) +{ + static char *string; + + if (str) + { + string = str; + } + + if (string == NULL || *string == '\0') + return NULL; + + char *p = string; + while (*string && *string!=term) + string++; + if (*string) + *string++ = '\0'; + return p; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/good_strtok.h b/debian/htdig/htdig-3.2.0b6/htlib/good_strtok.h new file mode 100644 index 00000000..89e05e0b --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/good_strtok.h @@ -0,0 +1,24 @@ +// +// good_strtok.h +// +// good_strtok: The good_strtok() function is very similar to the +// standard strtok() library function, except that good_strtok() +// will only skip over 1 separator if it finds one. This is +// needed when parsing strings with empty fields. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: good_strtok.h,v 1.7 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _good_strtok_h_ +#define _good_strtok_h_ + +char *good_strtok(char *, char); + +#endif + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/gregex.h b/debian/htdig/htdig-3.2.0b6/htlib/gregex.h new file mode 100644 index 00000000..38c4b68f --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/gregex.h @@ -0,0 +1,568 @@ +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +/* Definitions for data structures and routines for the regular + expression library, version 0.12. + Copyright (C) 1985,1989-1993,1995-1998, 2000 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef _REGEX_H +#define _REGEX_H 1 + +/* Allow the use in C++ code. */ +#ifdef __cplusplus +extern "C" { +#endif + +/* POSIX says that <sys/types.h> must be included (by the caller) before + <regex.h>. */ + +#if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && defined VMS +/* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it + should be there. */ +# include <stddef.h> +#endif + +/* The following two types have to be signed and unsigned integer type + wide enough to hold a value of a pointer. For most ANSI compilers + ptrdiff_t and size_t should be likely OK. Still size of these two + types is 2 for Microsoft C. Ugh... */ +typedef long int s_reg_t; +typedef unsigned long int active_reg_t; + +/* The following bits are used to determine the regexp syntax we + recognize. The set/not-set meanings are chosen so that Emacs syntax + remains the value 0. The bits are given in alphabetical order, and + the definitions shifted by one from the previous bit; thus, when we + add or remove a bit, only one other definition need change. */ +typedef unsigned long int reg_syntax_t; + +/* If this bit is not set, then \ inside a bracket expression is literal. + If set, then such a \ quotes the following character. */ +#define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) + +/* If this bit is not set, then + and ? are operators, and \+ and \? are + literals. + If set, then \+ and \? are operators and + and ? are literals. */ +#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) + +/* If this bit is set, then character classes are supported. They are: + [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], + [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. + If not set, then character classes are not supported. */ +#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) + +/* If this bit is set, then ^ and $ are always anchors (outside bracket + expressions, of course). + If this bit is not set, then it depends: + ^ is an anchor if it is at the beginning of a regular + expression or after an open-group or an alternation operator; + $ is an anchor if it is at the end of a regular expression, or + before a close-group or an alternation operator. + + This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because + POSIX draft 11.2 says that * etc. in leading positions is undefined. + We already implemented a previous draft which made those constructs + invalid, though, so we haven't changed the code back. */ +#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) + +/* If this bit is set, then special characters are always special + regardless of where they are in the pattern. + If this bit is not set, then special characters are special only in + some contexts; otherwise they are ordinary. Specifically, + * + ? and intervals are only special when not after the beginning, + open-group, or alternation operator. */ +#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) + +/* If this bit is set, then *, +, ?, and { cannot be first in an re or + immediately after an alternation or begin-group operator. */ +#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) + +/* If this bit is set, then . matches newline. + If not set, then it doesn't. */ +#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) + +/* If this bit is set, then . doesn't match NUL. + If not set, then it does. */ +#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) + +/* If this bit is set, nonmatching lists [^...] do not match newline. + If not set, they do. */ +#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) + +/* If this bit is set, either \{...\} or {...} defines an + interval, depending on RE_NO_BK_BRACES. + If not set, \{, \}, {, and } are literals. */ +#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) + +/* If this bit is set, +, ? and | aren't recognized as operators. + If not set, they are. */ +#define RE_LIMITED_OPS (RE_INTERVALS << 1) + +/* If this bit is set, newline is an alternation operator. + If not set, newline is literal. */ +#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) + +/* If this bit is set, then `{...}' defines an interval, and \{ and \} + are literals. + If not set, then `\{...\}' defines an interval. */ +#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) + +/* If this bit is set, (...) defines a group, and \( and \) are literals. + If not set, \(...\) defines a group, and ( and ) are literals. */ +#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) + +/* If this bit is set, then \<digit> matches <digit>. + If not set, then \<digit> is a back-reference. */ +#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) + +/* If this bit is set, then | is an alternation operator, and \| is literal. + If not set, then \| is an alternation operator, and | is literal. */ +#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) + +/* If this bit is set, then an ending range point collating higher + than the starting range point, as in [z-a], is invalid. + If not set, then when ending range point collates higher than the + starting range point, the range is ignored. */ +#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) + +/* If this bit is set, then an unmatched ) is ordinary. + If not set, then an unmatched ) is invalid. */ +#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) + +/* If this bit is set, succeed as soon as we match the whole pattern, + without further backtracking. */ +#define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) + +/* If this bit is set, do not process the GNU regex operators. + If not set, then the GNU regex operators are recognized. */ +#define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) + +/* If this bit is set, turn on internal regex debugging. + If not set, and debugging was on, turn it off. + This only works if regex.c is compiled -DDEBUG. + We define this bit always, so that all that's needed to turn on + debugging is to recompile regex.c; the calling code can always have + this bit set, and it won't affect anything in the normal case. */ +#define RE_DEBUG (RE_NO_GNU_OPS << 1) + +/* This global variable defines the particular regexp syntax to use (for + some interfaces). When a regexp is compiled, the syntax used is + stored in the pattern buffer, so changing this does not affect + already-compiled regexps. */ +extern reg_syntax_t re_syntax_options; + +/* Define combinations of the above bits for the standard possibilities. + (The [[[ comments delimit what gets put into the Texinfo file, so + don't delete them!) */ +/* [[[begin syntaxes]]] */ +#define RE_SYNTAX_EMACS 0 + +#define RE_SYNTAX_AWK \ + (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ + | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ + | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GNU_AWK \ + ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ + & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS)) + +#define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ + | RE_INTERVALS | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GREP \ + (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ + | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ + | RE_NEWLINE_ALT) + +#define RE_SYNTAX_EGREP \ + (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ + | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ + | RE_NO_BK_VBAR) + +#define RE_SYNTAX_POSIX_EGREP \ + (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES) + +/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ +#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC + +#define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC + +/* Syntax bits common to both basic and extended POSIX regex syntax. */ +#define _RE_SYNTAX_POSIX_COMMON \ + (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ + | RE_INTERVALS | RE_NO_EMPTY_RANGES) + +#define RE_SYNTAX_POSIX_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) + +/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes + RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this + isn't minimal, since other operators, such as \`, aren't disabled. */ +#define RE_SYNTAX_POSIX_MINIMAL_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) + +#define RE_SYNTAX_POSIX_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ + | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) + +/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is + removed and RE_NO_BK_REFS is added. */ +#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) +/* [[[end syntaxes]]] */ + +/* Maximum number of duplicates an interval can allow. Some systems + (erroneously) define this in other header files, but we want our + value, so remove any previous define. */ +#ifdef RE_DUP_MAX +# undef RE_DUP_MAX +#endif +/* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */ +#define RE_DUP_MAX (0x7fff) + + +/* POSIX `cflags' bits (i.e., information for `regcomp'). */ + +/* If this bit is set, then use extended regular expression syntax. + If not set, then use basic regular expression syntax. */ +#define REG_EXTENDED 1 + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +#define REG_ICASE (REG_EXTENDED << 1) + +/* If this bit is set, then anchors do not match at newline + characters in the string. + If not set, then anchors do match at newlines. */ +#define REG_NEWLINE (REG_ICASE << 1) + +/* If this bit is set, then report only success or fail in regexec. + If not set, then returns differ between not matching and errors. */ +#define REG_NOSUB (REG_NEWLINE << 1) + + +/* POSIX `eflags' bits (i.e., information for regexec). */ + +/* If this bit is set, then the beginning-of-line operator doesn't match + the beginning of the string (presumably because it's not the + beginning of a line). + If not set, then the beginning-of-line operator does match the + beginning of the string. */ +#define REG_NOTBOL 1 + +/* Like REG_NOTBOL, except for the end-of-line. */ +#define REG_NOTEOL (1 << 1) + + +/* If any error codes are removed, changed, or added, update the + `re_error_msg' table in regex.c. */ +typedef enum +{ +#ifdef _XOPEN_SOURCE + REG_ENOSYS = -1, /* This will never happen for this implementation. */ +#endif + + REG_NOERROR = 0, /* Success. */ + REG_NOMATCH, /* Didn't find a match (for regexec). */ + + /* POSIX regcomp return error codes. (In the order listed in the + standard.) */ + REG_BADPAT, /* Invalid pattern. */ + REG_ECOLLATE, /* Not implemented. */ + REG_ECTYPE, /* Invalid character class name. */ + REG_EESCAPE, /* Trailing backslash. */ + REG_ESUBREG, /* Invalid back reference. */ + REG_EBRACK, /* Unmatched left bracket. */ + REG_EPAREN, /* Parenthesis imbalance. */ + REG_EBRACE, /* Unmatched \{. */ + REG_BADBR, /* Invalid contents of \{\}. */ + REG_ERANGE, /* Invalid range end. */ + REG_ESPACE, /* Ran out of memory. */ + REG_BADRPT, /* No preceding re for repetition op. */ + + /* Error codes we've added. */ + REG_EEND, /* Premature end. */ + REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ + REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ +} reg_errcode_t; + +/* This data structure represents a compiled pattern. Before calling + the pattern compiler, the fields `buffer', `allocated', `fastmap', + `translate', and `no_sub' can be set. After the pattern has been + compiled, the `re_nsub' field is available. All other fields are + private to the regex routines. */ + +#ifndef RE_TRANSLATE_TYPE +# define RE_TRANSLATE_TYPE char * +#endif + +struct re_pattern_buffer +{ +/* [[[begin pattern_buffer]]] */ + /* Space that holds the compiled pattern. It is declared as + `unsigned char *' because its elements are + sometimes used as array indexes. */ + unsigned char *buffer; + + /* Number of bytes to which `buffer' points. */ + unsigned long int allocated; + + /* Number of bytes actually used in `buffer'. */ + unsigned long int used; + + /* Syntax setting with which the pattern was compiled. */ + reg_syntax_t syntax; + + /* Pointer to a fastmap, if any, otherwise zero. re_search uses + the fastmap, if there is one, to skip over impossible + starting points for matches. */ + char *fastmap; + + /* Either a translate table to apply to all characters before + comparing them, or zero for no translation. The translation + is applied to a pattern when it is compiled and to a string + when it is matched. */ + RE_TRANSLATE_TYPE translate; + + /* Number of subexpressions found by the compiler. */ + size_t re_nsub; + + /* Zero if this pattern cannot match the empty string, one else. + Well, in truth it's used only in `re_search_2', to see + whether or not we should use the fastmap, so we don't set + this absolutely perfectly; see `re_compile_fastmap' (the + `duplicate' case). */ + unsigned can_be_null : 1; + + /* If REGS_UNALLOCATED, allocate space in the `regs' structure + for `max (RE_NREGS, re_nsub + 1)' groups. + If REGS_REALLOCATE, reallocate space if necessary. + If REGS_FIXED, use what's there. */ +#define REGS_UNALLOCATED 0 +#define REGS_REALLOCATE 1 +#define REGS_FIXED 2 + unsigned regs_allocated : 2; + + /* Set to zero when `regex_compile' compiles a pattern; set to one + by `re_compile_fastmap' if it updates the fastmap. */ + unsigned fastmap_accurate : 1; + + /* If set, `re_match_2' does not return information about + subexpressions. */ + unsigned no_sub : 1; + + /* If set, a beginning-of-line anchor doesn't match at the + beginning of the string. */ + unsigned not_bol : 1; + + /* Similarly for an end-of-line anchor. */ + unsigned not_eol : 1; + + /* If true, an anchor at a newline matches. */ + unsigned newline_anchor : 1; + +/* [[[end pattern_buffer]]] */ +}; + +typedef struct re_pattern_buffer regex_t; + +/* Type for byte offsets within the string. POSIX mandates this. */ +typedef int regoff_t; + + +/* This is the structure we store register match data in. See + regex.texinfo for a full description of what registers match. */ +struct re_registers +{ + unsigned num_regs; + regoff_t *start; + regoff_t *end; +}; + + +/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, + `re_match_2' returns information about at least this many registers + the first time a `regs' structure is passed. */ +#ifndef RE_NREGS +# define RE_NREGS 30 +#endif + + +/* POSIX specification for registers. Aside from the different names than + `re_registers', POSIX uses an array of structures, instead of a + structure of arrays. */ +typedef struct +{ + regoff_t rm_so; /* Byte offset from string's start to substring's start. */ + regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ +} regmatch_t; + +/* Declarations for routines. */ + +/* To avoid duplicating every routine declaration -- once with a + prototype (if we are ANSI), and once without (if we aren't) -- we + use the following macro to declare argument types. This + unfortunately clutters up the declarations a bit, but I think it's + worth it. */ + +#if defined(__STDC__) + +# define _RE_ARGS(args) args + +#else /* not __STDC__ */ + +# define _RE_ARGS(args) () + +#endif /* not __STDC__ */ + +/* Sets the current default syntax to SYNTAX, and return the old syntax. + You can also simply assign to the `re_syntax_options' variable. */ +extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax)); + +/* Compile the regular expression PATTERN, with length LENGTH + and syntax given by the global `re_syntax_options', into the buffer + BUFFER. Return NULL if successful, and an error string if not. */ +extern const char *re_compile_pattern + _RE_ARGS ((const char *pattern, size_t length, + struct re_pattern_buffer *buffer)); + + +/* Compile a fastmap for the compiled pattern in BUFFER; used to + accelerate searches. Return 0 if successful and -2 if was an + internal error. */ +extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); + + +/* Search in the string STRING (with length LENGTH) for the pattern + compiled into BUFFER. Start searching at position START, for RANGE + characters. Return the starting position of the match, -1 for no + match, or -2 for an internal error. Also return register + information in REGS (if REGS and BUFFER->no_sub are nonzero). */ +extern int re_search + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, int range, struct re_registers *regs)); + + +/* Like `re_search', but search in the concatenation of STRING1 and + STRING2. Also, stop searching at index START + STOP. */ +extern int re_search_2 + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, int range, struct re_registers *regs, int stop)); + + +/* Like `re_search', but return how many characters in STRING the regexp + in BUFFER matched, starting at position START. */ +extern int re_match + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, struct re_registers *regs)); + + +/* Relates to `re_match' as `re_search_2' relates to `re_search'. */ +extern int re_match_2 + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, struct re_registers *regs, int stop)); + + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using BUFFER and REGS will use this memory + for recording register information. STARTS and ENDS must be + allocated with malloc, and must each be at least `NUM_REGS * sizeof + (regoff_t)' bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ +extern void re_set_registers + _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs, + unsigned num_regs, regoff_t *starts, regoff_t *ends)); + +#if defined _REGEX_RE_COMP || defined _LIBC +# ifndef _CRAY +/* 4.2 bsd compatibility. */ +extern char *re_comp _RE_ARGS ((const char *)); +extern int re_exec _RE_ARGS ((const char *)); +# endif +#endif + +/* GCC 2.95 and later have "__restrict"; C99 compilers have + "restrict", and "configure" may have defined "restrict". */ +#ifndef __restrict +# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) +# if defined restrict || 199901L <= __STDC_VERSION__ +# define __restrict restrict +# else +# define __restrict +# endif +# endif +#endif +/* For now unconditionally define __restrict_arr to expand to nothing. + Ideally we would have a test for the compiler which allows defining + it to restrict. */ +#ifndef __restrict_arr +# define __restrict_arr +#endif + +/* POSIX compatibility. */ +extern int regcomp _RE_ARGS ((regex_t *__restrict __preg, + const char *__restrict __pattern, + int __cflags)); + +extern int regexec _RE_ARGS ((const regex_t *__restrict __preg, + const char *__restrict __string, size_t __nmatch, + regmatch_t __pmatch[__restrict_arr], + int __eflags)); + +extern size_t regerror _RE_ARGS ((int __errcode, const regex_t *__preg, + char *__errbuf, size_t __errbuf_size)); + +extern void regfree _RE_ARGS ((regex_t *__preg)); + + +#ifdef __cplusplus +} +#endif /* C++ */ + +#endif /* regex.h */ + +/* +Local variables: +make-backup-files: t +version-control: t +trim-versions-without-asking: nil +End: +*/ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/htString.h b/debian/htdig/htdig-3.2.0b6/htlib/htString.h new file mode 100644 index 00000000..592a072f --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/htString.h @@ -0,0 +1,260 @@ +// +// htString.h +// +// htString: (implementation in String.cc) Just Another String class. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: htString.h,v 1.24 2004/05/28 13:15:21 lha Exp $ +// +#ifndef __String_h +#define __String_h + +#include "Object.h" + +#include <stdarg.h> +#include <stdio.h> + +#ifdef HAVE_STD +#include <iostream> +#ifdef HAVE_NAMESPACES +using namespace std; +#endif +#else +#include <iostream.h> +#endif /* HAVE_STD */ + +class String : public Object +{ +public: + String() { Length = 0; Allocated = 0; Data = 0; } // Create an empty string + String(int init); // initial allocated length + String(const char *s); // from null terminated s + String(const char *s, int len); // from s with length len + String(const String &s); // Copy constructor + + // + // This can be used for performance reasons if it is known the + // String will need to grow. + // + String(const String &s, int allocation_hint); + + ~String(); + + inline int length() const; + char *get(); + const char *get() const; + operator char*() { return get(); } + operator const char*() const { return get(); } + + // + // Interpretation + // + int as_integer(int def = 0) const; + double as_double(double def = 0) const; + int empty() const { return length() == 0; } + + // + // If it is not posible to use the constructor with an initial + // allocation size, use the following member to set the size. + // + void allocate(int init) {reallocate_space(init);} + + // + // allocate space for a new char *, and copy the String in. + // + char *new_char() const; + + // + // Assignment + // + inline String& set(const char *s, int l) { trunc(); append(s, l); return *this; } + inline String& set(char *s) { trunc(); append(s, strlen(s)); return *this; } + void operator = (const String &s); + void operator = (const char *s); + inline void operator += (const String &s) { append(s); } + inline void operator += (const char *s) { append(s); } + + // + // Appending + // + inline String &operator << (const char *); + inline String &operator << (char); + inline String &operator << (unsigned char c) {return *this<<(char)c;} + String &operator << (int); + String &operator << (unsigned int); + String &operator << (long); + inline String &operator << (short i) {return *this<<(int)i;} + String &operator << (const String &); + String &operator << (const String *s) {return *this << *s;} + + // + // Access to specific characters + // + inline char &operator [] (int n); + inline char operator [] (int n) const; + inline char Nth (int n) { return (*this)[n]; } + inline char last() const { return Length > 0 ? Data[Length - 1] : '\0'; } + + // + // Removing + // + char operator >> (char c); + + // + // Comparison + // Return: + // 0 : 'this' is equal to 's'. + // -1 : 'this' is less than 's'. + // 1 : 'this' is greater than 's'. + // + int compare(const Object& s) const { return compare((const String&)s); } + int compare(const String& s) const; + int nocase_compare(const String &s) const; + + // + // Searching for parts + // + int lastIndexOf(char c) const; + int lastIndexOf(char c, int pos) const; + int indexOf(char c) const; + int indexOf(char c, int pos) const; + int indexOf(const char *) const; + int indexOf(const char *, int pos) const; + + // + // Manipulation + // + void append(const String &s); + void append(const char *s); + void append(const char *s, int n); + void append(char ch); + + inline String &trunc() { Length = 0; return *this; } + String &chop(int n = 1); + String &chop(char ch = '\n'); + String &chop(const char *str = (char *)"\r\n"); + + // + // SubStrings + // + // The string starting at postion 'start' and length 'len'. + // + String sub(int start, int len) const; + String sub(int start) const; + + // + // IO + // + int Write(int fd) const; + +#ifndef NOSTREAM + void debug(ostream &o); +#endif /* NOSTREAM */ + + // + // Non-member operators + // + friend String operator + (const String &a, const String &b); + friend int operator == (const String &a, const String &b); + friend int operator != (const String &a, const String &b); + friend int operator < (const String &a, const String &b); + friend int operator > (const String &a, const String &b); + friend int operator <= (const String &a, const String &b); + friend int operator >= (const String &a, const String &b); + +#ifndef NOSTREAM + friend ostream &operator << (ostream &o, const String &s); + + friend istream &operator >> (istream &in, String &line); +#endif /* NOSTREAM */ + + int readLine(FILE *in); + + int lowercase(); + int uppercase(); + + void replace(char c1, char c2); + int remove(const char *); + + Object *Copy() const { return new String(*this); } + + // + // Persistent storage support + // + void Serialize(String &); + void Deserialize(String &, int &); + +private: + int Length; // Current Length + int Allocated; // Total space allocated + char *Data; // The actual contents + + void copy_data_from(const char *s, int len, int dest_offset = 0); + void copy(const char *s, int len, int allocation_hint); + + // + // Possibly make Data bigger. + // + void reallocate_space(int len); + + // + // Allocate some space for the data. Delete Data if it + // has been allocated. + // + void allocate_space(int len); + // Allocate some space without rounding + void allocate_fix_space(int len); + + friend class StringIndex; +}; + +extern char *form(const char *, ...); +extern char *vform(const char *, va_list); + +// +// Inline methods. +// +inline String &String::operator << (const char *str) +{ + append(str); + return *this; +} + +inline String &String::operator << (char ch) +{ + append(ch); + return *this; +} + +inline int String::length() const +{ + return Length; +} + +inline char String::operator [] (int n) const +{ + if(n < 0) n = Length + n; + if(n >= Length || n < 0) return '\0'; + + return Data[n]; +} + +static char null = '\0'; + +inline char &String::operator [] (int n) +{ + if(n < 0) n = Length + n; + if(n >= Length || n < 0) return null; + + return Data[n]; +} + +// +// Non friend, non member operators +// +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/lib.h b/debian/htdig/htdig-3.2.0b6/htlib/lib.h new file mode 100644 index 00000000..d17bd7c4 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/lib.h @@ -0,0 +1,90 @@ +// +// lib.h +// +// lib: Contains typical declarations and header inclusions used by +// most sources in this directory. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: lib.h,v 1.16 2004/05/28 13:15:21 lha Exp $ +// + +#ifndef _lib_h +#define _lib_h + +#ifndef _MSC_VER /* _WIN32 */ +#include "clib.h" +#endif + +#include <string.h> + +#ifdef _MSC_VER /* _WIN32 */ +#include "dirent_local.h" +#define S_ISDIR(v) ((v)&_S_IFDIR) +#define S_ISREG(v) ((v)&_S_IFREG) +#else +#include <dirent.h> // for scandir +#endif + +#ifdef _MSC_VER /* _WIN32 */ +#include <io.h> +#include <stdlib.h> +#define S_IFIFO _S_IFIFO // pipe +#define S_IFBLK 0060000 // block special +#define S_IFLNK 0120000 // symbolic link +#define S_IFSOCK 0140000 // socket +#define S_IFWHT 0160000 // whiteout +#define R_OK 02 +#define popen _popen +#define pclose _pclose +#define lstat stat +#define readlink(x,y,z) {-1} +#define sleep(t) _sleep((t) * 1000) +#endif + +#if TIME_WITH_SYS_TIME +# include <sys/time.h> +# include <time.h> +#else +# if HAVE_SYS_TIME_H +# include <sys/time.h> +# else +# include <time.h> +# endif +#endif + +// +// Other defines used throughout the library +// +#define OK 0 +#define NOTOK (-1) + +// +// To get rid of inconsistencies between different machines we will ALWAYS +// use our own version of the following routines +// +int mystrcasecmp(const char *, const char *); +int mystrncasecmp(const char *, const char *, int); + +// +// The standard strstr() function is limited in that it does case-sensitive +// searches. This version will ignore case. +// +const char *mystrcasestr(const char *s, const char *pattern); + +// +// Too many problems with system strptime() functions... Just use our own +// version of it. +// +char *mystrptime(const char *buf, const char *fmt, struct tm *tm); + +// +// timegm() is quite rare, so provide our own. +// +extern "C" time_t Httimegm(struct tm *tm); + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/libdefs.h b/debian/htdig/htdig-3.2.0b6/htlib/libdefs.h new file mode 100644 index 00000000..e4f1d11c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/libdefs.h @@ -0,0 +1,109 @@ +#ifndef LIBDEFS_H +#define LIBDEFS_H + +/* + {{{ includes + */ + + +#ifdef _MSC_VER /* _WIN32 */ +#include <windows.h> +#endif + + +#include "htconfig.h" + + +#ifdef STDC_HEADERS +#include <string.h> +#include <stdlib.h> +#include <stdio.h> +#endif + +/* + }}} + */ +/* + {{{ typedefs + */ + +#if SIZEOF_UNSIGNED_LONG_INT == 8 +typedef unsigned long word64; +#define TIGER_64BIT +#elif SIZEOF_UNSIGNED_LONG_LONG_INT == 8 + +#ifndef _MSC_VER /* _WIN32 */ +typedef unsigned long long word64; +#else //ifdef _MSC_VER /* _WIN32 */ +typedef DWORD64 word64; +#endif +#else +#error "Cannot find a 64 bit integer in your system, sorry." +#endif + +#if SIZEOF_UNSIGNED_LONG_INT == 4 +typedef unsigned long word32; +#elif SIZEOF_UNSIGNED_INT == 4 +typedef unsigned int word32; +#else +#error "Cannot find a 32 bit integer in your system, sorry." +#endif + +#if SIZEOF_UNSIGNED_INT == 2 +typedef unsigned int word16; +#elif SIZEOF_UNSIGNED_SHORT_INT == 2 +typedef unsigned short word16; +#else +#error "Cannot find a 16 bit integer in your system, sorry." +#endif + +#if SIZEOF_UNSIGNED_CHAR == 1 +typedef unsigned char word8; +#else +#error "Cannot find an 8 bit char in your system, sorry." +#endif + +typedef word8 byte; +typedef word32 dword; + +/* + }}} + */ + +/* + {{{ macros and defines + */ + +#define RAND32 (word32) ((word32)rand() << 17 ^ (word32)rand() << 9 ^ rand()) + +#ifndef HAVE_MEMMOVE +#ifdef HAVE_BCOPY +#define memmove(d, s, n) bcopy ((s), (d), (n)) +#else +#error "Neither memmove nor bcopy exists on your system." +#endif +#endif + +#define ENCRYPT 0 +#define DECRYPT 1 + +/* + }}} + */ + +/* + {{{ prototypes + */ + +void Bzero(void *s, int n); + +word32 byteswap(word32 x); + +int BreakToThree(void *key, unsigned int keylen, + void *keyword1, void *keyword2, void *keyword3); + +/* + }}} + */ + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/malloc.c b/debian/htdig/htdig-3.2.0b6/htlib/malloc.c new file mode 100644 index 00000000..ec9a0d07 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/malloc.c @@ -0,0 +1,39 @@ + +/* Work around bug on some systems where malloc (0) fails. + Copyright (C) 1997, 1998 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* written by Jim Meyering */ + +#if HAVE_CONFIG_H +# include <config.h> +#endif +#undef malloc + +#include <stddef.h> + +char *malloc (); + +/* Allocate an N-byte block of memory from the heap. + If N is zero, allocate a 1-byte block. */ + +char * +rpl_malloc (size_t n) +{ + if (n == 0) + n = 1; + return malloc (n); +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/md5.cc b/debian/htdig/htdig-3.2.0b6/htlib/md5.cc new file mode 100644 index 00000000..999a4758 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/md5.cc @@ -0,0 +1,41 @@ +#include <stdio.h> +#include <stdlib.h> +#include <time.h> +extern "C" { +#include "mhash_md5.h" +} + +#define MD5_LENGTH 16 + +void md5(char *rhash, char *buf, int len, time_t *date, bool debug) +{ + + int i; + MD5_CTX *td; + unsigned char *hash; + + td = (MD5_CTX *)malloc(sizeof(MD5_CTX)); + MD5Init( td); + // td = mhash_init(MHASH_MD5); + + MD5Update(td,(unsigned char *) buf, len); + // mhash(td, buf, len); + + if (date) { + MD5Update(td,(unsigned char *)date, sizeof(*date)); + } + + hash = (unsigned char *)MD5Final(td); + // hash = (char *)mhash_end(td); + + memcpy(rhash,hash,MD5_LENGTH); + + if (debug) { + printf(" "); + for (i = 0; i < MD5_LENGTH; i++) { + printf("%.2x", hash[i]); + } + printf(" "); + } + delete td; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/md5.h b/debian/htdig/htdig-3.2.0b6/htlib/md5.h new file mode 100644 index 00000000..8e4c029a --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/md5.h @@ -0,0 +1,3 @@ +#define MD5_LENGTH 16 + +void md5(char *rhash, char *buf, int len, time_t *date, bool debug); diff --git a/debian/htdig/htdig-3.2.0b6/htlib/memcmp.c b/debian/htdig/htdig-3.2.0b6/htlib/memcmp.c new file mode 100644 index 00000000..55c327ec --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/memcmp.c @@ -0,0 +1,72 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ +/* + * Copyright (c) 1990, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include <sys/types.h> + +#ifndef HAVE_MEMCMP +/* + * memcmp -- + * + * PUBLIC: #ifndef HAVE_MEMCMP + * PUBLIC: int memcmp __P((const void *, const void *, size_t)); + * PUBLIC: #endif + */ +int +memcmp(s1, s2, n) + char *s1, *s2; + size_t n; +{ + if (n != 0) { + unsigned char *p1 = (unsigned char *)s1, + *p2 = (unsigned char *)s2; + do { + if (*p1++ != *p2++) + return (*--p1 - *--p2); + } while (--n != 0); + } + return (0); +} +#endif /* HAVE_MEMCMP */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/memcpy.c b/debian/htdig/htdig-3.2.0b6/htlib/memcpy.c new file mode 100644 index 00000000..22e07c04 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/memcpy.c @@ -0,0 +1,144 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ +/* + * Copyright (c) 1990, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#ifndef NO_SYSTEM_INCLUDES +#include <sys/types.h> +#endif + +#ifndef HAVE_MEMCPY + +/* + * sizeof(word) MUST BE A POWER OF TWO + * SO THAT wmask BELOW IS ALL ONES + */ +typedef int word; /* "word" used for optimal copy speed */ + +#undef wsize +#define wsize sizeof(word) +#undef wmask +#define wmask (wsize - 1) + +/* + * Copy a block of memory, handling overlap. + * This is the routine that actually implements + * (the portable versions of) bcopy, memcpy, and memmove. + */ +/* + * PUBLIC: #ifndef HAVE_MEMCPY + * PUBLIC: void *memcpy __P((void *, const void *, size_t)); + * PUBLIC: #endif + */ +void * +memcpy(dst0, src0, length) + void *dst0; + const void *src0; + register size_t length; +{ + register char *dst = dst0; + register const char *src = src0; + register size_t t; + + if (length == 0 || dst == src) /* nothing to do */ + goto done; + + /* + * Macros: loop-t-times; and loop-t-times, t>0 + */ +#undef TLOOP +#define TLOOP(s) if (t) TLOOP1(s) +#undef TLOOP1 +#define TLOOP1(s) do { s; } while (--t) + + if ((unsigned long)dst < (unsigned long)src) { + /* + * Copy forward. + */ + t = (int)src; /* only need low bits */ + if ((t | (int)dst) & wmask) { + /* + * Try to align operands. This cannot be done + * unless the low bits match. + */ + if ((t ^ (int)dst) & wmask || length < wsize) + t = length; + else + t = wsize - (t & wmask); + length -= t; + TLOOP1(*dst++ = *src++); + } + /* + * Copy whole words, then mop up any trailing bytes. + */ + t = length / wsize; + TLOOP(*(word *)dst = *(word *)src; src += wsize; dst += wsize); + t = length & wmask; + TLOOP(*dst++ = *src++); + } else { + /* + * Copy backwards. Otherwise essentially the same. + * Alignment works as before, except that it takes + * (t&wmask) bytes to align, not wsize-(t&wmask). + */ + src += length; + dst += length; + t = (int)src; + if ((t | (int)dst) & wmask) { + if ((t ^ (int)dst) & wmask || length <= wsize) + t = length; + else + t &= wmask; + length -= t; + TLOOP1(*--dst = *--src); + } + t = length / wsize; + TLOOP(src -= wsize; dst -= wsize; *(word *)dst = *(word *)src); + t = length & wmask; + TLOOP(*--dst = *--src); + } +done: + return (dst0); +} + +#endif /* HAVE_MEMCPY */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/memmove.c b/debian/htdig/htdig-3.2.0b6/htlib/memmove.c new file mode 100644 index 00000000..7615f63c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/memmove.c @@ -0,0 +1,143 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ +/* + * Copyright (c) 1990, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#ifndef NO_SYSTEM_INCLUDES +#include <sys/types.h> +#endif + +#ifndef HAVE_MEMMOVE +/* + * sizeof(word) MUST BE A POWER OF TWO + * SO THAT wmask BELOW IS ALL ONES + */ +typedef int word; /* "word" used for optimal copy speed */ + +#undef wsize +#define wsize sizeof(word) +#undef wmask +#define wmask (wsize - 1) + +/* + * Copy a block of memory, handling overlap. + * This is the routine that actually implements + * (the portable versions of) bcopy, memcpy, and memmove. + */ +/* + * PUBLIC: #ifndef HAVE_MEMMOVE + * PUBLIC: void *memmove __P((void *, const void *, size_t)); + * PUBLIC: #endif + */ +void * +memmove(dst0, src0, length) + void *dst0; + const void *src0; + register size_t length; +{ + register char *dst = dst0; + register const char *src = src0; + register size_t t; + + if (length == 0 || dst == src) /* nothing to do */ + goto done; + + /* + * Macros: loop-t-times; and loop-t-times, t>0 + */ +#undef TLOOP +#define TLOOP(s) if (t) TLOOP1(s) +#undef TLOOP1 +#define TLOOP1(s) do { s; } while (--t) + + if ((unsigned long)dst < (unsigned long)src) { + /* + * Copy forward. + */ + t = (int)src; /* only need low bits */ + if ((t | (int)dst) & wmask) { + /* + * Try to align operands. This cannot be done + * unless the low bits match. + */ + if ((t ^ (int)dst) & wmask || length < wsize) + t = length; + else + t = wsize - (t & wmask); + length -= t; + TLOOP1(*dst++ = *src++); + } + /* + * Copy whole words, then mop up any trailing bytes. + */ + t = length / wsize; + TLOOP(*(word *)dst = *(word *)src; src += wsize; dst += wsize); + t = length & wmask; + TLOOP(*dst++ = *src++); + } else { + /* + * Copy backwards. Otherwise essentially the same. + * Alignment works as before, except that it takes + * (t&wmask) bytes to align, not wsize-(t&wmask). + */ + src += length; + dst += length; + t = (int)src; + if ((t | (int)dst) & wmask) { + if ((t ^ (int)dst) & wmask || length <= wsize) + t = length; + else + t &= wmask; + length -= t; + TLOOP1(*--dst = *--src); + } + t = length / wsize; + TLOOP(src -= wsize; dst -= wsize; *(word *)dst = *(word *)src); + t = length & wmask; + TLOOP(*--dst = *--src); + } +done: + return (dst0); +} +#endif /* HAVE_MEMOVE */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/mhash_md5.c b/debian/htdig/htdig-3.2.0b6/htlib/mhash_md5.c new file mode 100644 index 00000000..2e0441a1 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/mhash_md5.c @@ -0,0 +1,534 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/* + Copyright (C) 1991-2, RSA Data Security, Inc. Created 1991. All + rights reserved. + + License to copy and use this software is granted provided that it + is identified as the "RSA Data Security, Inc. MD5 Message-Digest + Algorithm" in all material mentioning or referencing this software + or this function. + + License is also granted to make and use derivative works provided + that such works are identified as "derived from the RSA Data + Security, Inc. MD5 Message-Digest Algorithm" in all material + mentioning or referencing the derived work. + + RSA Data Security, Inc. makes no representations concerning either + the merchantability of this software or the suitability of this + software for any particular purpose. It is provided "as is" + without express or implied warranty of any kind. + + These notices must be retained in any copies of any part of this + documentation and/or software. + */ + +/* + This is a slightly modified version + */ + +#include "mhash_md5.h" + +/* + Constants for MD5Transform routine. + */ + + +#define S11 7 +#define S12 12 +#define S13 17 +#define S14 22 +#define S21 5 +#define S22 9 +#define S23 14 +#define S24 20 +#define S31 4 +#define S32 11 +#define S33 16 +#define S34 23 +#define S41 6 +#define S42 10 +#define S43 15 +#define S44 21 + +static void MD5Transform PROTO_LIST((word32[4], unsigned char[64])); +static void Encode PROTO_LIST + ((unsigned char *, word32 *, unsigned int)); +static void Decode PROTO_LIST + ((word32 *, unsigned char *, unsigned int)); +static void MD5_memcpy PROTO_LIST((POINTER, POINTER, unsigned int)); +static void MD5_memset PROTO_LIST((POINTER, int, unsigned int)); + +static unsigned char PADDING[64] = +{ + 0x80, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 +}; + +/* + F, G, H and I are basic MD5 functions. + */ +#define F(x, y, z) (((x) & (y)) | ((~x) & (z))) +#define G(x, y, z) (((x) & (z)) | ((y) & (~z))) +#define H(x, y, z) ((x) ^ (y) ^ (z)) +#define I(x, y, z) ((y) ^ ((x) | (~z))) + +/* + ROTATE_LEFT rotates x left n bits. + */ +#define ROTATE_LEFT(x, n) (((x) << (n)) | ((x) >> (32-(n)))) + +/* + FF, GG, HH, and II transformations for rounds 1, 2, 3, and 4. + Rotation is separate from addition to prevent recomputation. + */ +#define FF(a, b, c, d, x, s, ac) { \ + (a) += F ((b), (c), (d)) + (x) + (word32)(ac); \ + (a) = ROTATE_LEFT ((a), (s)); \ + (a) += (b); \ + } +#define GG(a, b, c, d, x, s, ac) { \ + (a) += G ((b), (c), (d)) + (x) + (word32)(ac); \ + (a) = ROTATE_LEFT ((a), (s)); \ + (a) += (b); \ + } +#define HH(a, b, c, d, x, s, ac) { \ + (a) += H ((b), (c), (d)) + (x) + (word32)(ac); \ + (a) = ROTATE_LEFT ((a), (s)); \ + (a) += (b); \ + } +#define II(a, b, c, d, x, s, ac) { \ + (a) += I ((b), (c), (d)) + (x) + (word32)(ac); \ + (a) = ROTATE_LEFT ((a), (s)); \ + (a) += (b); \ + } + +/* + MD5 initialization. Begins an MD5 operation, writing a new context. + */ +void +MD5Init(context) + MD5_CTX *context; /* + context + */ +{ + context->count[0] = context->count[1] = 0; + /* + Load magic initialization constants. + */ + context->state[0] = 0x67452301; + context->state[1] = 0xefcdab89; + context->state[2] = 0x98badcfe; + context->state[3] = 0x10325476; +} + +/* + MD5 block update operation. Continues an MD5 message-digest + operation, processing another message block, and updating the + context. + */ +void +MD5Update(context, input, inputLen) + MD5_CTX *context; /* + context + */ + const unsigned char *input; /* + input block + */ + unsigned int inputLen; /* + length of input block + */ +{ + unsigned int i, index, partLen; + + /* + Compute number of bytes mod 64 + */ + index = (unsigned int) ((context->count[0] >> 3) & 0x3F); + + /* + Update number of bits + */ + if ((context->count[0] += ((word32) inputLen << 3)) < ((word32) inputLen << 3)) + context->count[1]++; + context->count[1] += ((word32) inputLen >> 29); + + partLen = 64 - index; + + /* + Transform as many times as possible. + */ + if (inputLen >= partLen) { + MD5_memcpy + ((POINTER) & context->buffer[index], (POINTER) input, partLen); + MD5Transform(context->state, context->buffer); + + for (i = partLen; i + 63 < inputLen; i += 64) + MD5Transform(context->state, (unsigned char *) &input[i]); + + index = 0; + } + else + i = 0; + + /* + Buffer remaining input + */ + MD5_memcpy + ((POINTER) & context->buffer[index], (POINTER) & input[i], + inputLen - i); +} + +/* + MD5 finalization. Ends an MD5 message-digest operation, writing the + the message digest and zeroizing the context. + */ +void * +MD5Final(context) + MD5_CTX *context; /* + context + */ +{ + unsigned char bits[8]; + unsigned int index, padLen; + unsigned char *digest = malloc(16); + + /* + Save number of bits + */ + Encode(bits, context->count, 8); + + /* + Pad out to 56 mod 64. + */ + index = (unsigned int) ((context->count[0] >> 3) & 0x3f); + padLen = (index < 56) ? (56 - index) : (120 - index); + MD5Update(context, PADDING, padLen); + + /* + Append length (before padding) + */ + MD5Update(context, bits, 8); + + + + /* + Store state in digest + */ + Encode(digest, context->state, 16); + + /* + Zeroize sensitive information. + */ + MD5_memset((POINTER) context, 0, sizeof(*context)); + + return digest; +} + +/* + MD5 basic transformation. Transforms state based on block. + */ +static void +MD5Transform(state, block) + word32 state[4]; + unsigned char block[64]; +{ + word32 a = state[0], b = state[1], c = state[2], d = state[3], x[16]; + + Decode(x, block, 64); + + /* + Round 1 + */ + FF(a, b, c, d, x[0], S11, 0xd76aa478); /* + 1 + */ + FF(d, a, b, c, x[1], S12, 0xe8c7b756); /* + 2 + */ + FF(c, d, a, b, x[2], S13, 0x242070db); /* + 3 + */ + FF(b, c, d, a, x[3], S14, 0xc1bdceee); /* + 4 + */ + FF(a, b, c, d, x[4], S11, 0xf57c0faf); /* + 5 + */ + FF(d, a, b, c, x[5], S12, 0x4787c62a); /* + 6 + */ + FF(c, d, a, b, x[6], S13, 0xa8304613); /* + 7 + */ + FF(b, c, d, a, x[7], S14, 0xfd469501); /* + 8 + */ + FF(a, b, c, d, x[8], S11, 0x698098d8); /* + 9 + */ + FF(d, a, b, c, x[9], S12, 0x8b44f7af); /* + 10 + */ + FF(c, d, a, b, x[10], S13, 0xffff5bb1); /* + 11 + */ + FF(b, c, d, a, x[11], S14, 0x895cd7be); /* + 12 + */ + FF(a, b, c, d, x[12], S11, 0x6b901122); /* + 13 + */ + FF(d, a, b, c, x[13], S12, 0xfd987193); /* + 14 + */ + FF(c, d, a, b, x[14], S13, 0xa679438e); /* + 15 + */ + FF(b, c, d, a, x[15], S14, 0x49b40821); /* + 16 + */ + + /* + Round 2 + */ + GG(a, b, c, d, x[1], S21, 0xf61e2562); /* + 17 + */ + GG(d, a, b, c, x[6], S22, 0xc040b340); /* + 18 + */ + GG(c, d, a, b, x[11], S23, 0x265e5a51); /* + 19 + */ + GG(b, c, d, a, x[0], S24, 0xe9b6c7aa); /* + 20 + */ + GG(a, b, c, d, x[5], S21, 0xd62f105d); /* + 21 + */ + GG(d, a, b, c, x[10], S22, 0x2441453); /* + 22 + */ + GG(c, d, a, b, x[15], S23, 0xd8a1e681); /* + 23 + */ + GG(b, c, d, a, x[4], S24, 0xe7d3fbc8); /* + 24 + */ + GG(a, b, c, d, x[9], S21, 0x21e1cde6); /* + 25 + */ + GG(d, a, b, c, x[14], S22, 0xc33707d6); /* + 26 + */ + GG(c, d, a, b, x[3], S23, 0xf4d50d87); /* + 27 + */ + GG(b, c, d, a, x[8], S24, 0x455a14ed); /* + 28 + */ + GG(a, b, c, d, x[13], S21, 0xa9e3e905); /* + 29 + */ + GG(d, a, b, c, x[2], S22, 0xfcefa3f8); /* + 30 + */ + GG(c, d, a, b, x[7], S23, 0x676f02d9); /* + 31 + */ + GG(b, c, d, a, x[12], S24, 0x8d2a4c8a); /* + 32 + */ + + /* + Round 3 + */ + HH(a, b, c, d, x[5], S31, 0xfffa3942); /* + 33 + */ + HH(d, a, b, c, x[8], S32, 0x8771f681); /* + 34 + */ + HH(c, d, a, b, x[11], S33, 0x6d9d6122); /* + 35 + */ + HH(b, c, d, a, x[14], S34, 0xfde5380c); /* + 36 + */ + HH(a, b, c, d, x[1], S31, 0xa4beea44); /* + 37 + */ + HH(d, a, b, c, x[4], S32, 0x4bdecfa9); /* + 38 + */ + HH(c, d, a, b, x[7], S33, 0xf6bb4b60); /* + 39 + */ + HH(b, c, d, a, x[10], S34, 0xbebfbc70); /* + 40 + */ + HH(a, b, c, d, x[13], S31, 0x289b7ec6); /* + 41 + */ + HH(d, a, b, c, x[0], S32, 0xeaa127fa); /* + 42 + */ + HH(c, d, a, b, x[3], S33, 0xd4ef3085); /* + 43 + */ + HH(b, c, d, a, x[6], S34, 0x4881d05); /* + 44 + */ + HH(a, b, c, d, x[9], S31, 0xd9d4d039); /* + 45 + */ + HH(d, a, b, c, x[12], S32, 0xe6db99e5); /* + 46 + */ + HH(c, d, a, b, x[15], S33, 0x1fa27cf8); /* + 47 + */ + HH(b, c, d, a, x[2], S34, 0xc4ac5665); /* + 48 + */ + + /* + Round 4 + */ + II(a, b, c, d, x[0], S41, 0xf4292244); /* + 49 + */ + II(d, a, b, c, x[7], S42, 0x432aff97); /* + 50 + */ + II(c, d, a, b, x[14], S43, 0xab9423a7); /* + 51 + */ + II(b, c, d, a, x[5], S44, 0xfc93a039); /* + 52 + */ + II(a, b, c, d, x[12], S41, 0x655b59c3); /* + 53 + */ + II(d, a, b, c, x[3], S42, 0x8f0ccc92); /* + 54 + */ + II(c, d, a, b, x[10], S43, 0xffeff47d); /* + 55 + */ + II(b, c, d, a, x[1], S44, 0x85845dd1); /* + 56 + */ + II(a, b, c, d, x[8], S41, 0x6fa87e4f); /* + 57 + */ + II(d, a, b, c, x[15], S42, 0xfe2ce6e0); /* + 58 + */ + II(c, d, a, b, x[6], S43, 0xa3014314); /* + 59 + */ + II(b, c, d, a, x[13], S44, 0x4e0811a1); /* + 60 + */ + II(a, b, c, d, x[4], S41, 0xf7537e82); /* + 61 + */ + II(d, a, b, c, x[11], S42, 0xbd3af235); /* + 62 + */ + II(c, d, a, b, x[2], S43, 0x2ad7d2bb); /* + 63 + */ + II(b, c, d, a, x[9], S44, 0xeb86d391); /* + 64 + */ + + state[0] += a; + state[1] += b; + state[2] += c; + state[3] += d; + + /* + Zeroize sensitive information. + + + + */ + MD5_memset((POINTER) x, 0, sizeof(x)); +} + +/* + Encodes input (word32) into output (unsigned char). Assumes len is + a multiple of 4. + */ +static void +Encode(output, input, len) + unsigned char *output; + word32 *input; + unsigned int len; +{ + unsigned int i, j; + + for (i = 0, j = 0; j < len; i++, j += 4) { + output[j] = (unsigned char) (input[i] & 0xff); + output[j + 1] = (unsigned char) ((input[i] >> 8) & 0xff); + output[j + 2] = (unsigned char) ((input[i] >> 16) & 0xff); + output[j + 3] = (unsigned char) ((input[i] >> 24) & 0xff); + } +} + +/* + Decodes input (unsigned char) into output (word32). Assumes len is + a multiple of 4. + */ +static void +Decode(output, input, len) + word32 *output; + unsigned char *input; + unsigned int len; +{ + unsigned int i, j; + + for (i = 0, j = 0; j < len; i++, j += 4) + output[i] = ((word32) input[j]) | (((word32) input[j + 1]) << 8) | + (((word32) input[j + 2]) << 16) | (((word32) input[j + 3]) << 24); +} + +/* + Note: Replace "for loop" with standard memcpy if possible. + */ + +static void +MD5_memcpy(output, input, len) + POINTER output; + POINTER input; + unsigned int len; +{ + unsigned int i; + + for (i = 0; i < len; i++) + output[i] = input[i]; +} + +/* + Note: Replace "for loop" with standard memset if possible. + */ +static void +MD5_memset(output, value, len) + POINTER output; + int value; + unsigned int len; +{ + unsigned int i; + + for (i = 0; i < len; i++) + ((char *) output)[i] = (char) value; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/mhash_md5.h b/debian/htdig/htdig-3.2.0b6/htlib/mhash_md5.h new file mode 100644 index 00000000..a1cbd2fc --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/mhash_md5.h @@ -0,0 +1,86 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/* + MD5.H - header file for MD5C.C + */ + +/* + Copyright (C) 1991-2, RSA Data Security, Inc. Created 1991. All + rights reserved. + + License to copy and use this software is granted provided that it + is identified as the "RSA Data Security, Inc. MD5 Message-Digest + Algorithm" in all material mentioning or referencing this software + or this function. + + License is also granted to make and use derivative works provided + that such works are identified as "derived from the RSA Data + Security, Inc. MD5 Message-Digest Algorithm" in all material + mentioning or referencing the derived work. + + RSA Data Security, Inc. makes no representations concerning either + the merchantability of this software or the suitability of this + software for any particular purpose. It is provided "as is" + without express or implied warranty of any kind. + These notices must be retained in any copies of any part of this + documentation and/or software. + */ + +#ifndef MHASH_MD5_H +#define MHASH_MD5_H + +#include <libdefs.h> + +/* + PROTOTYPES should be set to one if and only if the compiler supports + function argument prototyping. + The following makes PROTOTYPES default to 0 if it has not already + been defined with C compiler flags. + */ +#ifndef PROTOTYPES +#define PROTOTYPES 1 +#endif + +/* + POINTER defines a generic pointer type + */ +typedef unsigned char *POINTER; + +/* + PROTO_LIST is defined depending on how PROTOTYPES is defined above. + If using PROTOTYPES, then PROTO_LIST returns the list, otherwise it + returns an empty list. + */ +#if PROTOTYPES +#define PROTO_LIST(list) list +#else +#define PROTO_LIST(list) () +#endif + + + +/* + MD5 context. + */ +typedef struct { + word32 state[4]; /* + state (ABCD) + */ + word32 count[2]; /* + number of bits, modulo 2^64 (lsb first) + */ + word8 buffer[64]; /* + input buffer + */ +} MD5_CTX; + +void MD5Init PROTO_LIST((MD5_CTX *)); +void MD5Update PROTO_LIST((MD5_CTX *, const unsigned char *, unsigned int)); +void *MD5Final PROTO_LIST((MD5_CTX *)); + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/mktime.c b/debian/htdig/htdig-3.2.0b6/htlib/mktime.c new file mode 100644 index 00000000..a59dab4c --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/mktime.c @@ -0,0 +1,535 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/* Convert a `struct tm' to a time_t value. + Copyright (C) 1993, 94, 95, 96, 97, 98, 99 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Paul Eggert (eggert@twinsun.com). + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* Define this to have a standalone program to test this implementation of + mktime. */ +/* #define DEBUG 1 */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif + +#ifdef _LIBC +# define HAVE_LIMITS_H 1 +# define STDC_HEADERS 1 +#endif + +/* Assume that leap seconds are possible, unless told otherwise. + If the host has a `zic' command with a `-L leapsecondfilename' option, + then it supports leap seconds; otherwise it probably doesn't. */ +#ifndef LEAP_SECONDS_POSSIBLE +# define LEAP_SECONDS_POSSIBLE 1 +#endif + +#include <sys/types.h> /* Some systems define `time_t' here. */ +#include <time.h> + +#if HAVE_LIMITS_H +# include <limits.h> +#endif + +#if DEBUG +# include <stdio.h> +# if STDC_HEADERS +# include <stdlib.h> +# endif +/* Make it work even if the system's libc has its own mktime routine. */ +# define mktime my_mktime +#endif /* DEBUG */ + +#ifndef __P +# if defined __GNUC__ || (defined __STDC__ && __STDC__) +# define __P(args) args +# else +# define __P(args) () +# endif /* GCC. */ +#endif /* Not __P. */ + +#ifndef CHAR_BIT +# define CHAR_BIT 8 +#endif + +/* The extra casts work around common compiler bugs. */ +#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +/* The outer cast is needed to work around a bug in Cray C 5.0.3.0. + It is necessary at least when t == time_t. */ +#define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) : (t) 0)) +#define TYPE_MAXIMUM(t) ((t) (~ (t) 0 - TYPE_MINIMUM (t))) + +#ifndef INT_MIN +# define INT_MIN TYPE_MINIMUM (int) +#endif +#ifndef INT_MAX +# define INT_MAX TYPE_MAXIMUM (int) +#endif + +#ifndef TIME_T_MIN +# define TIME_T_MIN TYPE_MINIMUM (time_t) +#endif +#ifndef TIME_T_MAX +# define TIME_T_MAX TYPE_MAXIMUM (time_t) +#endif + +#define TM_YEAR_BASE 1900 +#define EPOCH_YEAR 1970 + +#ifndef __isleap +/* Nonzero if YEAR is a leap year (every 4 years, + except every 100th isn't, and every 400th is). */ +# define __isleap(year) \ + ((year) % 4 == 0 && ((year) % 100 != 0 || (year) % 400 == 0)) +#endif + +/* How many days come before each month (0-12). */ +const unsigned short int __mon_yday[2][13] = + { + /* Normal years. */ + { 0, 31, 59, 90, 120, 151, 181, 212, 243, 273, 304, 334, 365 }, + /* Leap years. */ + { 0, 31, 60, 91, 121, 152, 182, 213, 244, 274, 305, 335, 366 } + }; + + +#ifdef _LIBC +# define my_mktime_localtime_r __localtime_r +#else +/* If we're a mktime substitute in a GNU program, then prefer + localtime to localtime_r, since many localtime_r implementations + are buggy. */ +static struct tm * +my_mktime_localtime_r (const time_t *t, struct tm *tp) +{ + struct tm *l = localtime (t); + if (! l) + return 0; + *tp = *l; + return tp; +} +#endif /* ! _LIBC */ + + +/* Yield the difference between (YEAR-YDAY HOUR:MIN:SEC) and (*TP), + measured in seconds, ignoring leap seconds. + YEAR uses the same numbering as TM->tm_year. + All values are in range, except possibly YEAR. + If TP is null, return a nonzero value. + If overflow occurs, yield the low order bits of the correct answer. */ +static time_t +ydhms_tm_diff (int year, int yday, int hour, int min, int sec, + const struct tm *tp) +{ + if (!tp) + return 1; + else + { + /* Compute intervening leap days correctly even if year is negative. + Take care to avoid int overflow. time_t overflow is OK, since + only the low order bits of the correct time_t answer are needed. + Don't convert to time_t until after all divisions are done, since + time_t might be unsigned. */ + int a4 = (year >> 2) + (TM_YEAR_BASE >> 2) - ! (year & 3); + int b4 = (tp->tm_year >> 2) + (TM_YEAR_BASE >> 2) - ! (tp->tm_year & 3); + int a100 = a4 / 25 - (a4 % 25 < 0); + int b100 = b4 / 25 - (b4 % 25 < 0); + int a400 = a100 >> 2; + int b400 = b100 >> 2; + int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400); + time_t years = year - (time_t) tp->tm_year; + time_t days = (365 * years + intervening_leap_days + + (yday - tp->tm_yday)); + return (60 * (60 * (24 * days + (hour - tp->tm_hour)) + + (min - tp->tm_min)) + + (sec - tp->tm_sec)); + } +} + +/* Use CONVERT to convert *T to a broken down time in *TP. + If *T is out of range for conversion, adjust it so that + it is the nearest in-range value and then convert that. */ +static struct tm * +ranged_convert (struct tm *(*convert) (const time_t *, struct tm *), + time_t *t, struct tm *tp) +{ + struct tm *r; + + if (! (r = (*convert) (t, tp)) && *t) + { + time_t bad = *t; + time_t ok = 0; + struct tm tm; + + /* BAD is a known unconvertible time_t, and OK is a known good one. + Use binary search to narrow the range between BAD and OK until + they differ by 1. */ + while (bad != ok + (bad < 0 ? -1 : 1)) + { + time_t mid = *t = (bad < 0 + ? bad + ((ok - bad) >> 1) + : ok + ((bad - ok) >> 1)); + if ((r = (*convert) (t, tp))) + { + tm = *r; + ok = mid; + } + else + bad = mid; + } + + if (!r && ok) + { + /* The last conversion attempt failed; + revert to the most recent successful attempt. */ + *t = ok; + *tp = tm; + r = tp; + } + } + + return r; +} + + +/* Convert *TP to a time_t value, inverting + the monotonic and mostly-unit-linear conversion function CONVERT. + Use *OFFSET to keep track of a guess at the offset of the result, + compared to what the result would be for UTC without leap seconds. + If *OFFSET's guess is correct, only one CONVERT call is needed. */ +time_t +__mktime_internal (struct tm *tp, + struct tm *(*convert) (const time_t *, struct tm *), + time_t *offset) +{ + time_t t, dt, t0, t1, t2; + struct tm tm; + + /* The maximum number of probes (calls to CONVERT) should be enough + to handle any combinations of time zone rule changes, solar time, + leap seconds, and oscillations around a spring-forward gap. + POSIX.1 prohibits leap seconds, but some hosts have them anyway. */ + int remaining_probes = 6; + + /* Time requested. Copy it in case CONVERT modifies *TP; this can + occur if TP is localtime's returned value and CONVERT is localtime. */ + int sec = tp->tm_sec; + int min = tp->tm_min; + int hour = tp->tm_hour; + int mday = tp->tm_mday; + int mon = tp->tm_mon; + int year_requested = tp->tm_year; + int isdst = tp->tm_isdst; + + /* Ensure that mon is in range, and set year accordingly. */ + int mon_remainder = mon % 12; + int negative_mon_remainder = mon_remainder < 0; + int mon_years = mon / 12 - negative_mon_remainder; + int year = year_requested + mon_years; + + /* The other values need not be in range: + the remaining code handles minor overflows correctly, + assuming int and time_t arithmetic wraps around. + Major overflows are caught at the end. */ + + /* Calculate day of year from year, month, and day of month. + The result need not be in range. */ + int yday = ((__mon_yday[__isleap (year + TM_YEAR_BASE)] + [mon_remainder + 12 * negative_mon_remainder]) + + mday - 1); + + int sec_requested = sec; +#if LEAP_SECONDS_POSSIBLE + /* Handle out-of-range seconds specially, + since ydhms_tm_diff assumes every minute has 60 seconds. */ + if (sec < 0) + sec = 0; + if (59 < sec) + sec = 59; +#endif + + /* Invert CONVERT by probing. First assume the same offset as last time. + Then repeatedly use the error to improve the guess. */ + + tm.tm_year = EPOCH_YEAR - TM_YEAR_BASE; + tm.tm_yday = tm.tm_hour = tm.tm_min = tm.tm_sec = 0; + t0 = ydhms_tm_diff (year, yday, hour, min, sec, &tm); + + for (t = t1 = t2 = t0 + *offset; + (dt = ydhms_tm_diff (year, yday, hour, min, sec, + ranged_convert (convert, &t, &tm))); + t1 = t2, t2 = t, t += dt) + if (t == t1 && t != t2 + && (isdst < 0 || tm.tm_isdst < 0 + || (isdst != 0) != (tm.tm_isdst != 0))) + /* We can't possibly find a match, as we are oscillating + between two values. The requested time probably falls + within a spring-forward gap of size DT. Follow the common + practice in this case, which is to return a time that is DT + away from the requested time, preferring a time whose + tm_isdst differs from the requested value. In practice, + this is more useful than returning -1. */ + break; + else if (--remaining_probes == 0) + return -1; + + /* If we have a match, check whether tm.tm_isdst has the requested + value, if any. */ + if (dt == 0 && isdst != tm.tm_isdst && 0 <= isdst && 0 <= tm.tm_isdst) + { + /* tm.tm_isdst has the wrong value. Look for a neighboring + time with the right value, and use its UTC offset. + Heuristic: probe the previous three calendar quarters (approximately), + looking for the desired isdst. This isn't perfect, + but it's good enough in practice. */ + int quarter = 7889238; /* seconds per average 1/4 Gregorian year */ + int i; + + /* If we're too close to the time_t limit, look in future quarters. */ + if (t < TIME_T_MIN + 3 * quarter) + quarter = -quarter; + + for (i = 1; i <= 3; i++) + { + time_t ot = t - i * quarter; + struct tm otm; + ranged_convert (convert, &ot, &otm); + if (otm.tm_isdst == isdst) + { + /* We found the desired tm_isdst. + Extrapolate back to the desired time. */ + t = ot + ydhms_tm_diff (year, yday, hour, min, sec, &otm); + ranged_convert (convert, &t, &tm); + break; + } + } + } + + *offset = t - t0; + +#if LEAP_SECONDS_POSSIBLE + if (sec_requested != tm.tm_sec) + { + /* Adjust time to reflect the tm_sec requested, not the normalized value. + Also, repair any damage from a false match due to a leap second. */ + t += sec_requested - sec + (sec == 0 && tm.tm_sec == 60); + if (! (*convert) (&t, &tm)) + return -1; + } +#endif + + if (TIME_T_MAX / INT_MAX / 366 / 24 / 60 / 60 < 3) + { + /* time_t isn't large enough to rule out overflows in ydhms_tm_diff, + so check for major overflows. A gross check suffices, + since if t has overflowed, it is off by a multiple of + TIME_T_MAX - TIME_T_MIN + 1. So ignore any component of + the difference that is bounded by a small value. */ + + double dyear = (double) year_requested + mon_years - tm.tm_year; + double dday = 366 * dyear + mday; + double dsec = 60 * (60 * (24 * dday + hour) + min) + sec_requested; + + /* On Irix4.0.5 cc, dividing TIME_T_MIN by 3 does not produce + correct results, ie., it erroneously gives a positive value + of 715827882. Setting a variable first then doing math on it + seems to work. (ghazi@caip.rutgers.edu) */ + + const time_t time_t_max = TIME_T_MAX; + const time_t time_t_min = TIME_T_MIN; + + if (time_t_max / 3 - time_t_min / 3 < (dsec < 0 ? - dsec : dsec)) + return -1; + } + + *tp = tm; + return t; +} + + +static time_t localtime_offset; + +/* Convert *TP to a time_t value. */ +time_t +mymktime (tp) + struct tm *tp; +{ +#ifdef _LIBC + /* POSIX.1 8.1.1 requires that whenever mktime() is called, the + time zone names contained in the external variable `tzname' shall + be set as if the tzset() function had been called. */ + __tzset (); +#endif + + return __mktime_internal (tp, my_mktime_localtime_r, &localtime_offset); +} + +#ifdef weak_alias +weak_alias (mktime, timelocal) +#endif + +#if DEBUG + +static int +not_equal_tm (a, b) + struct tm *a; + struct tm *b; +{ + return ((a->tm_sec ^ b->tm_sec) + | (a->tm_min ^ b->tm_min) + | (a->tm_hour ^ b->tm_hour) + | (a->tm_mday ^ b->tm_mday) + | (a->tm_mon ^ b->tm_mon) + | (a->tm_year ^ b->tm_year) + | (a->tm_mday ^ b->tm_mday) + | (a->tm_yday ^ b->tm_yday) + | (a->tm_isdst ^ b->tm_isdst)); +} + +static void +print_tm (tp) + struct tm *tp; +{ + if (tp) + printf ("%04d-%02d-%02d %02d:%02d:%02d yday %03d wday %d isdst %d", + tp->tm_year + TM_YEAR_BASE, tp->tm_mon + 1, tp->tm_mday, + tp->tm_hour, tp->tm_min, tp->tm_sec, + tp->tm_yday, tp->tm_wday, tp->tm_isdst); + else + printf ("0"); +} + +static int +check_result (tk, tmk, tl, lt) + time_t tk; + struct tm tmk; + time_t tl; + struct tm *lt; +{ + if (tk != tl || !lt || not_equal_tm (&tmk, lt)) + { + printf ("mktime ("); + print_tm (&tmk); + printf (")\nyields ("); + print_tm (lt); + printf (") == %ld, should be %ld\n", (long) tl, (long) tk); + return 1; + } + + return 0; +} + +int +main (argc, argv) + int argc; + char **argv; +{ + int status = 0; + struct tm tm, tmk, tml; + struct tm *lt; + time_t tk, tl; + char trailer; + + if ((argc == 3 || argc == 4) + && (sscanf (argv[1], "%d-%d-%d%c", + &tm.tm_year, &tm.tm_mon, &tm.tm_mday, &trailer) + == 3) + && (sscanf (argv[2], "%d:%d:%d%c", + &tm.tm_hour, &tm.tm_min, &tm.tm_sec, &trailer) + == 3)) + { + tm.tm_year -= TM_YEAR_BASE; + tm.tm_mon--; + tm.tm_isdst = argc == 3 ? -1 : atoi (argv[3]); + tmk = tm; + tl = mktime (&tmk); + lt = localtime (&tl); + if (lt) + { + tml = *lt; + lt = &tml; + } + printf ("mktime returns %ld == ", (long) tl); + print_tm (&tmk); + printf ("\n"); + status = check_result (tl, tmk, tl, lt); + } + else if (argc == 4 || (argc == 5 && strcmp (argv[4], "-") == 0)) + { + time_t from = atol (argv[1]); + time_t by = atol (argv[2]); + time_t to = atol (argv[3]); + + if (argc == 4) + for (tl = from; tl <= to; tl += by) + { + lt = localtime (&tl); + if (lt) + { + tmk = tml = *lt; + tk = mktime (&tmk); + status |= check_result (tk, tmk, tl, tml); + } + else + { + printf ("localtime (%ld) yields 0\n", (long) tl); + status = 1; + } + } + else + for (tl = from; tl <= to; tl += by) + { + /* Null benchmark. */ + lt = localtime (&tl); + if (lt) + { + tmk = tml = *lt; + tk = tl; + status |= check_result (tk, tmk, tl, tml); + } + else + { + printf ("localtime (%ld) yields 0\n", (long) tl); + status = 1; + } + } + } + else + printf ("Usage:\ +\t%s YYYY-MM-DD HH:MM:SS [ISDST] # Test given time.\n\ +\t%s FROM BY TO # Test values FROM, FROM+BY, ..., TO.\n\ +\t%s FROM BY TO - # Do not test those values (for benchmark).\n", + argv[0], argv[0], argv[0]); + + return status; +} + +#endif /* DEBUG */ + +/* +Local Variables: +compile-command: "gcc -DDEBUG -DHAVE_LIMITS_H -DSTDC_HEADERS -Wall -W -O -g mktime.c -o mktime" +End: +*/ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/myqsort.c b/debian/htdig/htdig-3.2.0b6/htlib/myqsort.c new file mode 100644 index 00000000..1931cfdd --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/myqsort.c @@ -0,0 +1,260 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/* Copyright (C) 1991, 1992, 1996, 1997 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Written by Douglas C. Schmidt (schmidt@ics.uci.edu). + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* + * Standard qsort function modified to add a user data argument to + * the comparison function. + */ +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include <stdlib.h> +#include <string.h> + +#include "myqsort.h" + +/* Byte-wise swap two items of size SIZE. */ +#define SWAP(a, b, size) \ + do \ + { \ + register size_t __size = (size); \ + register char *__a = (a), *__b = (b); \ + do \ + { \ + char __tmp = *__a; \ + *__a++ = *__b; \ + *__b++ = __tmp; \ + } while (--__size > 0); \ + } while (0) + +/* Discontinue quicksort algorithm when partition gets below this size. + This particular magic number was chosen to work best on a Sun 4/260. */ +#define MAX_THRESH 4 + +/* Stack node declarations used to store unfulfilled partition obligations. */ +typedef struct + { + char *lo; + char *hi; + } stack_node; + +/* The next 4 #defines implement a very fast in-line stack abstraction. */ +#define STACK_SIZE (8 * sizeof(unsigned long int)) +#define PUSH(low, high) ((void) ((top->lo = (low)), (top->hi = (high)), ++top)) +#define POP(low, high) ((void) (--top, (low = top->lo), (high = top->hi))) +#define STACK_NOT_EMPTY (stack < top) + + +/* Order size using quicksort. This implementation incorporates + four optimizations discussed in Sedgewick: + + 1. Non-recursive, using an explicit stack of pointer that store the + next array partition to sort. To save time, this maximum amount + of space required to store an array of MAX_INT is allocated on the + stack. Assuming a 32-bit integer, this needs only 32 * + sizeof(stack_node) == 136 bits. Pretty cheap, actually. + + 2. Chose the pivot element using a median-of-three decision tree. + This reduces the probability of selecting a bad pivot value and + eliminates certain extraneous comparisons. + + 3. Only quicksorts TOTAL_ELEMS / MAX_THRESH partitions, leaving + insertion sort to order the MAX_THRESH items within each partition. + This is a big win, since insertion sort is faster for small, mostly + sorted array segments. + + 4. The larger of the two sub-partitions is always pushed onto the + stack first, with the algorithm then concentrating on the + smaller partition. This *guarantees* no more than log (n) + stack size is needed (actually O(1) in this case)! */ + +void +myqsort(void *const pbase, size_t total_elems, size_t size, myqsort_cmp cmp, void *data) +{ + register char *base_ptr = (char *) pbase; + + /* Allocating SIZE bytes for a pivot buffer facilitates a better + algorithm below since we can do comparisons directly on the pivot. */ + char *pivot_buffer = (char *) malloc (size); + const size_t max_thresh = MAX_THRESH * size; + + if (total_elems == 0) { + /* Avoid lossage with unsigned arithmetic below. */ + free(pivot_buffer); + return; + } + + if (total_elems > MAX_THRESH) + { + char *lo = base_ptr; + char *hi = &lo[size * (total_elems - 1)]; + /* Largest size needed for 32-bit int!!! */ + stack_node stack[STACK_SIZE]; + stack_node *top = stack + 1; + + while (STACK_NOT_EMPTY) + { + char *left_ptr; + char *right_ptr; + + char *pivot = pivot_buffer; + + /* Select median value from among LO, MID, and HI. Rearrange + LO and HI so the three values are sorted. This lowers the + probability of picking a pathological pivot value and + skips a comparison for both the LEFT_PTR and RIGHT_PTR. */ + + char *mid = lo + size * ((hi - lo) / size >> 1); + + if ((*cmp) (data, (void *) mid, (void *) lo) < 0) + SWAP (mid, lo, size); + if ((*cmp) (data, (void *) hi, (void *) mid) < 0) + SWAP (mid, hi, size); + else + goto jump_over; + if ((*cmp) (data, (void *) mid, (void *) lo) < 0) + SWAP (mid, lo, size); + jump_over:; + memcpy (pivot, mid, size); + pivot = pivot_buffer; + + left_ptr = lo + size; + right_ptr = hi - size; + + /* Here's the famous ``collapse the walls'' section of quicksort. + Gotta like those tight inner loops! They are the main reason + that this algorithm runs much faster than others. */ + do + { + while ((*cmp) (data, (void *) left_ptr, (void *) pivot) < 0) + left_ptr += size; + + while ((*cmp) (data, (void *) pivot, (void *) right_ptr) < 0) + right_ptr -= size; + + if (left_ptr < right_ptr) + { + SWAP (left_ptr, right_ptr, size); + left_ptr += size; + right_ptr -= size; + } + else if (left_ptr == right_ptr) + { + left_ptr += size; + right_ptr -= size; + break; + } + } + while (left_ptr <= right_ptr); + + /* Set up pointers for next iteration. First determine whether + left and right partitions are below the threshold size. If so, + ignore one or both. Otherwise, push the larger partition's + bounds on the stack and continue sorting the smaller one. */ + + if ((size_t) (right_ptr - lo) <= max_thresh) + { + if ((size_t) (hi - left_ptr) <= max_thresh) + /* Ignore both small partitions. */ + POP (lo, hi); + else + /* Ignore small left partition. */ + lo = left_ptr; + } + else if ((size_t) (hi - left_ptr) <= max_thresh) + /* Ignore small right partition. */ + hi = right_ptr; + else if ((right_ptr - lo) > (hi - left_ptr)) + { + /* Push larger left partition indices. */ + PUSH (lo, right_ptr); + lo = left_ptr; + } + else + { + /* Push larger right partition indices. */ + PUSH (left_ptr, hi); + hi = right_ptr; + } + } + } + + /* Once the BASE_PTR array is partially sorted by quicksort the rest + is completely sorted using insertion sort, since this is efficient + for partitions below MAX_THRESH size. BASE_PTR points to the beginning + of the array to sort, and END_PTR points at the very last element in + the array (*not* one beyond it!). */ + +#define min(x, y) ((x) < (y) ? (x) : (y)) + + { + char *const end_ptr = &base_ptr[size * (total_elems - 1)]; + char *tmp_ptr = base_ptr; + char *thresh = min(end_ptr, base_ptr + max_thresh); + register char *run_ptr; + + /* Find smallest element in first threshold and place it at the + array's beginning. This is the smallest array element, + and the operation speeds up insertion sort's inner loop. */ + + for (run_ptr = tmp_ptr + size; run_ptr <= thresh; run_ptr += size) + if ((*cmp) (data, (void *) run_ptr, (void *) tmp_ptr) < 0) + tmp_ptr = run_ptr; + + if (tmp_ptr != base_ptr) + SWAP (tmp_ptr, base_ptr, size); + + /* Insertion sort, running from left-hand-side up to right-hand-side. */ + + run_ptr = base_ptr + size; + while ((run_ptr += size) <= end_ptr) + { + tmp_ptr = run_ptr - size; + while ((*cmp) (data, (void *) run_ptr, (void *) tmp_ptr) < 0) + tmp_ptr -= size; + + tmp_ptr += size; + if (tmp_ptr != run_ptr) + { + char *trav; + + trav = run_ptr + size; + while (--trav >= run_ptr) + { + char c = *trav; + char *hi, *lo; + + for (hi = lo = trav; (lo -= size) >= tmp_ptr; hi = lo) + *hi = *lo; + *hi = c; + } + } + } + } + + free(pivot_buffer); +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/myqsort.h b/debian/htdig/htdig-3.2.0b6/htlib/myqsort.h new file mode 100644 index 00000000..415324f6 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/myqsort.h @@ -0,0 +1,23 @@ +/* + * Part of the ht://Dig package <http://www.htdig.org/> + * Copyright (c) 1999-2004 The ht://Dig Group + * For copyright details, see the file COPYING in your distribution + * or the GNU Library General Public License (LGPL) version 2 or later + * <http://www.gnu.org/copyleft/lgpl.html> + */ +#ifndef _myqsort_h +#define _myqsort_h + +#ifdef __cplusplus +extern "C" { +#endif + +typedef int (*myqsort_cmp)(void *data, void *a, void *b); + +void myqsort(void *const pbase, size_t total_elems, size_t size, myqsort_cmp cmp, void *data); + +#ifdef __cplusplus +} +#endif + +#endif /* _myqsort_h */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/raise.c b/debian/htdig/htdig-3.2.0b6/htlib/raise.c new file mode 100644 index 00000000..e686822b --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/raise.c @@ -0,0 +1,39 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#ifndef HAVE_RAISE + +#ifndef NO_SYSTEM_INCLUDES +#include <signal.h> +#include <unistd.h> +#endif + +/* + * raise -- + * Send a signal to the current process. + * + * PUBLIC: #ifndef HAVE_RAISE + * PUBLIC: int raise __P((int)); + * PUBLIC: #endif + */ +int +raise(s) + int s; +{ + return (kill(getpid(), s)); +} +#endif /* HAVE_RAISE */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/realloc.c b/debian/htdig/htdig-3.2.0b6/htlib/realloc.c new file mode 100644 index 00000000..2d31766a --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/realloc.c @@ -0,0 +1,146 @@ +/* Change the size of a block allocated by `malloc'. + Copyright 1990, 1991, 1992, 1993 Free Software Foundation, Inc. + Written May 1989 by Mike Haertel. + +This library is free software; you can redistribute it and/or +modify it under the terms of the GNU Library General Public License as +published by the Free Software Foundation; either version 2 of the +License, or (at your option) any later version. + +This library is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +Library General Public License for more details. + +You should have received a copy of the GNU Library General Public +License along with this library; see the file COPYING.LIB. If +not, write to the Free Software Foundation, Inc., 675 Mass Ave, +Cambridge, MA 02139, USA. + + The author may be reached (Email) at the address mike@ai.mit.edu, + or (US mail) as Mike Haertel c/o Free Software Foundation. */ + +#ifndef _MALLOC_INTERNAL +#define _MALLOC_INTERNAL +#include <malloc.h> +#endif + +#define min(A, B) ((A) < (B) ? (A) : (B)) + +/* Debugging hook for realloc. */ +__ptr_t (*__realloc_hook) __P ((__ptr_t __ptr, size_t __size)); + +/* Resize the given region to the new size, returning a pointer + to the (possibly moved) region. This is optimized for speed; + some benchmarks seem to indicate that greater compactness is + achieved by unconditionally allocating and copying to a + new region. This module has incestuous knowledge of the + internals of both free and malloc. */ +__ptr_t +realloc (ptr, size) + __ptr_t ptr; + size_t size; +{ + __ptr_t result; + int type; + size_t block, blocks, oldlimit; + + if (size == 0) + { + free (ptr); + return malloc (0); + } + else if (ptr == NULL) + return malloc (size); + + if (__realloc_hook != NULL) + return (*__realloc_hook) (ptr, size); + + block = BLOCK (ptr); + + type = _heapinfo[block].busy.type; + switch (type) + { + case 0: + /* Maybe reallocate a large block to a small fragment. */ + if (size <= BLOCKSIZE / 2) + { + result = malloc (size); + if (result != NULL) + { + memcpy (result, ptr, size); + free (ptr); + return result; + } + } + + /* The new size is a large allocation as well; + see if we can hold it in place. */ + blocks = BLOCKIFY (size); + if (blocks < _heapinfo[block].busy.info.size) + { + /* The new size is smaller; return + excess memory to the free list. */ + _heapinfo[block + blocks].busy.type = 0; + _heapinfo[block + blocks].busy.info.size + = _heapinfo[block].busy.info.size - blocks; + _heapinfo[block].busy.info.size = blocks; + free (ADDRESS (block + blocks)); + result = ptr; + } + else if (blocks == _heapinfo[block].busy.info.size) + /* No size change necessary. */ + result = ptr; + else + { + /* Won't fit, so allocate a new region that will. + Free the old region first in case there is sufficient + adjacent free space to grow without moving. */ + blocks = _heapinfo[block].busy.info.size; + /* Prevent free from actually returning memory to the system. */ + oldlimit = _heaplimit; + _heaplimit = 0; + free (ptr); + _heaplimit = oldlimit; + result = malloc (size); + if (result == NULL) + { + /* Now we're really in trouble. We have to unfree + the thing we just freed. Unfortunately it might + have been coalesced with its neighbors. */ + if (_heapindex == block) + (void) malloc (blocks * BLOCKSIZE); + else + { + __ptr_t previous = malloc ((block - _heapindex) * BLOCKSIZE); + (void) malloc (blocks * BLOCKSIZE); + free (previous); + } + return NULL; + } + if (ptr != result) + memmove (result, ptr, blocks * BLOCKSIZE); + } + break; + + default: + /* Old size is a fragment; type is logarithm + to base two of the fragment size. */ + if (size > (size_t) (1 << (type - 1)) && size <= (size_t) (1 << type)) + /* The new size is the same kind of fragment. */ + result = ptr; + else + { + /* The new size is different; allocate a new space, + and copy the lesser of the new size and the old. */ + result = malloc (size); + if (result == NULL) + return NULL; + memcpy (result, ptr, min (size, (size_t) 1 << type)); + free (ptr); + } + break; + } + + return result; +} diff --git a/debian/htdig/htdig-3.2.0b6/htlib/regex.c b/debian/htdig/htdig-3.2.0b6/htlib/regex.c new file mode 100644 index 00000000..1a13901e --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/regex.c @@ -0,0 +1,7924 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> * +/* Copyright (c) 1999-2004 The ht://Dig Group * +/* For copyright details, see the file COPYING in your distribution * +/* or the GNU Library General Public License (LGPL) version 2 or later * +/* <http://www.gnu.org/copyleft/lgpl.html> * + + +/* Extended regular expression matching and search library, + version 0.12. + (Implements POSIX draft P1003.2/D11.2, except for some of the + internationalization features.) + Copyright (C) 1993-1999, 2000, 2001 Free Software Foundation, Inc. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +/* AIX requires this to be the first thing in the file. */ +#if defined _AIX && !defined REGEX_MALLOC + #pragma alloca +#endif + +#undef _GNU_SOURCE +#define _GNU_SOURCE + +#ifdef HAVE_CONFIG_H +# include "htconfig.h" +#endif + +#ifdef _MSC_VER /* _WIN32 */ +#define alloca _alloca +#endif + +#ifndef PARAMS +# if defined __GNUC__ || (defined __STDC__ && __STDC__) +# define PARAMS(args) args +# else +# define PARAMS(args) () +# endif /* GCC. */ +#endif /* Not PARAMS. */ + +#if defined STDC_HEADERS && !defined emacs +# include <stddef.h> +#else +/* We need this for `regex.h', and perhaps for the Emacs include files. */ +# include <sys/types.h> +#endif + +#define WIDE_CHAR_SUPPORT (HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_BTOWC) + +/* For platform which support the ISO C amendement 1 functionality we + support user defined character classes. */ +#if defined _LIBC || WIDE_CHAR_SUPPORT +/* Solaris 2.5 has a bug: <wchar.h> must be included before <wctype.h>. */ +# include <wchar.h> +# include <wctype.h> +#endif + +/* This is for multi byte string support. */ +#ifdef MBS_SUPPORT +# define CHAR_TYPE wchar_t +# define US_CHAR_TYPE wchar_t/* unsigned character type */ +# define COMPILED_BUFFER_VAR wc_buffer +# define OFFSET_ADDRESS_SIZE 1 /* the size which STORE_NUMBER macro use */ +# define CHAR_CLASS_SIZE ((__alignof__(wctype_t)+sizeof(wctype_t))/sizeof(CHAR_TYPE)+1) +# define PUT_CHAR(c) \ + do { \ + if (MC_CUR_MAX == 1) \ + putchar (c); \ + else \ + printf ("%C", (wint_t) c); /* Should we use wide stream?? */ \ + } while (0) +# define TRUE 1 +# define FALSE 0 +#else +# define CHAR_TYPE char +# define US_CHAR_TYPE unsigned char /* unsigned character type */ +# define COMPILED_BUFFER_VAR bufp->buffer +# define OFFSET_ADDRESS_SIZE 2 +# define PUT_CHAR(c) putchar (c) +#endif /* MBS_SUPPORT */ + +#ifdef _LIBC +/* We have to keep the namespace clean. */ +# define regfree(preg) __regfree (preg) +# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef) +# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags) +# define regerror(errcode, preg, errbuf, errbuf_size) \ + __regerror(errcode, preg, errbuf, errbuf_size) +# define re_set_registers(bu, re, nu, st, en) \ + __re_set_registers (bu, re, nu, st, en) +# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \ + __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) +# define re_match(bufp, string, size, pos, regs) \ + __re_match (bufp, string, size, pos, regs) +# define re_search(bufp, string, size, startpos, range, regs) \ + __re_search (bufp, string, size, startpos, range, regs) +# define re_compile_pattern(pattern, length, bufp) \ + __re_compile_pattern (pattern, length, bufp) +# define re_set_syntax(syntax) __re_set_syntax (syntax) +# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \ + __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop) +# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp) + +# define btowc __btowc + +/* We are also using some library internals. */ +# include <locale/localeinfo.h> +# include <locale/elem-hash.h> +# include <langinfo.h> +# include <locale/coll-lookup.h> +#endif + +/* This is for other GNU distributions with internationalized messages. */ +#if HAVE_LIBINTL_H || defined _LIBC +# include <libintl.h> +# ifdef _LIBC +# undef gettext +# define gettext(msgid) __dcgettext ("libc", msgid, LC_MESSAGES) +# endif +#else +# define gettext(msgid) (msgid) +#endif + +#ifndef gettext_noop +/* This define is so xgettext can find the internationalizable + strings. */ +# define gettext_noop(String) String +#endif + +/* The `emacs' switch turns on certain matching commands + that make sense only in Emacs. */ +#ifdef emacs + +# include "lisp.h" +# include "buffer.h" +# include "syntax.h" + +#else /* not emacs */ + +/* If we are not linking with Emacs proper, + we can't use the relocating allocator + even if config.h says that we can. */ +# undef REL_ALLOC + +# if defined STDC_HEADERS || defined _LIBC +# include <stdlib.h> +# else +char *malloc (); +char *realloc (); +# endif + +/* When used in Emacs's lib-src, we need to get bzero and bcopy somehow. + If nothing else has been done, use the method below. */ +# ifdef INHIBIT_STRING_HEADER +# if !(defined HAVE_BZERO && defined HAVE_BCOPY) +# if !defined bzero && !defined bcopy +# undef INHIBIT_STRING_HEADER +# endif +# endif +# endif + +/* This is the normal way of making sure we have a bcopy and a bzero. + This is used in most programs--a few other programs avoid this + by defining INHIBIT_STRING_HEADER. */ +# ifndef INHIBIT_STRING_HEADER +# if defined HAVE_STRING_H || defined STDC_HEADERS || defined _LIBC +# include <string.h> +# ifndef bzero +# ifndef _LIBC +# define bzero(s, n) (memset (s, '\0', n), (s)) +# else +# define bzero(s, n) __bzero (s, n) +# endif +# endif +# else +# include <strings.h> +# ifndef memcmp +# define memcmp(s1, s2, n) bcmp (s1, s2, n) +# endif +# ifndef memcpy +# define memcpy(d, s, n) (bcopy (s, d, n), (d)) +# endif +# endif +# endif + +/* Define the syntax stuff for \<, \>, etc. */ + +/* This must be nonzero for the wordchar and notwordchar pattern + commands in re_match_2. */ +# ifndef Sword +# define Sword 1 +# endif + +# ifdef SWITCH_ENUM_BUG +# define SWITCH_ENUM_CAST(x) ((int)(x)) +# else +# define SWITCH_ENUM_CAST(x) (x) +# endif + +#endif /* not emacs */ + +#if defined _LIBC || HAVE_LIMITS_H +# include <limits.h> +#endif + +#ifndef MB_LEN_MAX +# define MB_LEN_MAX 1 +#endif + +/* Get the interface, including the syntax bits. */ +#include <gregex.h> + +/* isalpha etc. are used for the character classes. */ +#include <ctype.h> + +/* Jim Meyering writes: + + "... Some ctype macros are valid only for character codes that + isascii says are ASCII (SGI's IRIX-4.0.5 is one such system --when + using /bin/cc or gcc but without giving an ansi option). So, all + ctype uses should be through macros like ISPRINT... If + STDC_HEADERS is defined, then autoconf has verified that the ctype + macros don't need to be guarded with references to isascii. ... + Defining isascii to 1 should let any compiler worth its salt + eliminate the && through constant folding." + Solaris defines some of these symbols so we must undefine them first. */ + +#undef ISASCII +#if defined STDC_HEADERS || (!defined isascii && !defined HAVE_ISASCII) +# define ISASCII(c) 1 +#else +# define ISASCII(c) isascii(c) +#endif + +#ifdef isblank +# define ISBLANK(c) (ISASCII (c) && isblank (c)) +#else +# define ISBLANK(c) ((c) == ' ' || (c) == '\t') +#endif +#ifdef isgraph +# define ISGRAPH(c) (ISASCII (c) && isgraph (c)) +#else +# define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c)) +#endif + +#undef ISPRINT +#define ISPRINT(c) (ISASCII (c) && isprint (c)) +#define ISDIGIT(c) (ISASCII (c) && isdigit (c)) +#define ISALNUM(c) (ISASCII (c) && isalnum (c)) +#define ISALPHA(c) (ISASCII (c) && isalpha (c)) +#define ISCNTRL(c) (ISASCII (c) && iscntrl (c)) +#define ISLOWER(c) (ISASCII (c) && islower (c)) +#define ISPUNCT(c) (ISASCII (c) && ispunct (c)) +#define ISSPACE(c) (ISASCII (c) && isspace (c)) +#define ISUPPER(c) (ISASCII (c) && isupper (c)) +#define ISXDIGIT(c) (ISASCII (c) && isxdigit (c)) + +#ifdef _tolower +# define TOLOWER(c) _tolower(c) +#else +# define TOLOWER(c) tolower(c) +#endif + +#ifndef NULL +# define NULL (void *)0 +#endif + +/* We remove any previous definition of `SIGN_EXTEND_CHAR', + since ours (we hope) works properly with all combinations of + machines, compilers, `char' and `unsigned char' argument types. + (Per Bothner suggested the basic approach.) */ +#undef SIGN_EXTEND_CHAR +#if __STDC__ +# define SIGN_EXTEND_CHAR(c) ((signed char) (c)) +#else /* not __STDC__ */ +/* As in Harbison and Steele. */ +# define SIGN_EXTEND_CHAR(c) ((((unsigned char) (c)) ^ 128) - 128) +#endif + +#ifndef emacs +/* How many characters in the character set. */ +# define CHAR_SET_SIZE 256 + +# ifdef SYNTAX_TABLE + +extern char *re_syntax_table; + +# else /* not SYNTAX_TABLE */ + +static char re_syntax_table[CHAR_SET_SIZE]; + +static void init_syntax_once PARAMS ((void)); + +static void +init_syntax_once () +{ + register int c; + static int done = 0; + + if (done) + return; + bzero (re_syntax_table, sizeof re_syntax_table); + + for (c = 0; c < CHAR_SET_SIZE; ++c) + if (ISALNUM (c)) + re_syntax_table[c] = Sword; + + re_syntax_table['_'] = Sword; + + done = 1; +} + +# endif /* not SYNTAX_TABLE */ + +# define SYNTAX(c) re_syntax_table[(unsigned char) (c)] + +#endif /* emacs */ + +/* Should we use malloc or alloca? If REGEX_MALLOC is not defined, we + use `alloca' instead of `malloc'. This is because using malloc in + re_search* or re_match* could cause memory leaks when C-g is used in + Emacs; also, malloc is slower and causes storage fragmentation. On + the other hand, malloc is more portable, and easier to debug. + + Because we sometimes use alloca, some routines have to be macros, + not functions -- `alloca'-allocated space disappears at the end of the + function it is called in. */ + +#ifdef REGEX_MALLOC + +# define REGEX_ALLOCATE malloc +# define REGEX_REALLOCATE(source, osize, nsize) realloc (source, nsize) +# define REGEX_FREE free + +#else /* not REGEX_MALLOC */ + +/* Emacs already defines alloca, sometimes. */ +# ifndef alloca + +/* Make alloca work the best possible way. */ +# ifdef __GNUC__ +# define alloca __builtin_alloca +# else /* not __GNUC__ */ +# if HAVE_ALLOCA_H +# include <alloca.h> +# endif /* HAVE_ALLOCA_H */ +# endif /* not __GNUC__ */ + +# endif /* not alloca */ + +# define REGEX_ALLOCATE alloca + +/* Assumes a `char *destination' variable. */ +# define REGEX_REALLOCATE(source, osize, nsize) \ + (destination = (char *) alloca (nsize), \ + memcpy (destination, source, osize)) + +/* No need to do anything to free, after alloca. */ +# define REGEX_FREE(arg) ((void)0) /* Do nothing! But inhibit gcc warning. */ + +#endif /* not REGEX_MALLOC */ + +/* Define how to allocate the failure stack. */ + +#if defined REL_ALLOC && defined REGEX_MALLOC + +# define REGEX_ALLOCATE_STACK(size) \ + r_alloc (&failure_stack_ptr, (size)) +# define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + r_re_alloc (&failure_stack_ptr, (nsize)) +# define REGEX_FREE_STACK(ptr) \ + r_alloc_free (&failure_stack_ptr) + +#else /* not using relocating allocator */ + +# ifdef REGEX_MALLOC + +# define REGEX_ALLOCATE_STACK malloc +# define REGEX_REALLOCATE_STACK(source, osize, nsize) realloc (source, nsize) +# define REGEX_FREE_STACK free + +# else /* not REGEX_MALLOC */ + +# define REGEX_ALLOCATE_STACK alloca + +# define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + REGEX_REALLOCATE (source, osize, nsize) +/* No need to explicitly free anything. */ +# define REGEX_FREE_STACK(arg) + +# endif /* not REGEX_MALLOC */ +#endif /* not using relocating allocator */ + + +/* True if `size1' is non-NULL and PTR is pointing anywhere inside + `string1' or just past its end. This works if PTR is NULL, which is + a good thing. */ +#define FIRST_STRING_P(ptr) \ + (size1 && string1 <= (ptr) && (ptr) <= string1 + size1) + +/* (Re)Allocate N items of type T using malloc, or fail. */ +#define TALLOC(n, t) ((t *) malloc ((n) * sizeof (t))) +#define RETALLOC(addr, n, t) ((addr) = (t *) realloc (addr, (n) * sizeof (t))) +#define RETALLOC_IF(addr, n, t) \ + if (addr) RETALLOC((addr), (n), t); else (addr) = TALLOC ((n), t) +#define REGEX_TALLOC(n, t) ((t *) REGEX_ALLOCATE ((n) * sizeof (t))) + +#define BYTEWIDTH 8 /* In bits. */ + +#define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) + +#undef MAX +#undef MIN +#define MAX(a, b) ((a) > (b) ? (a) : (b)) +#define MIN(a, b) ((a) < (b) ? (a) : (b)) + +typedef char boolean; +#define false 0 +#define true 1 + +static int re_match_2_internal PARAMS ((struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int pos, + struct re_registers *regs, + int stop)); + +/* These are the command codes that appear in compiled regular + expressions. Some opcodes are followed by argument bytes. A + command code can specify any interpretation whatsoever for its + arguments. Zero bytes may appear in the compiled regular expression. */ + +typedef enum +{ + no_op = 0, + + /* Succeed right away--no more backtracking. */ + succeed, + + /* Followed by one byte giving n, then by n literal bytes. */ + exactn, + +#ifdef MBS_SUPPORT + /* Same as exactn, but contains binary data. */ + exactn_bin, +#endif + + /* Matches any (more or less) character. */ + anychar, + + /* Matches any one char belonging to specified set. First + following byte is number of bitmap bytes. Then come bytes + for a bitmap saying which chars are in. Bits in each byte + are ordered low-bit-first. A character is in the set if its + bit is 1. A character too large to have a bit in the map is + automatically not in the set. */ + /* ifdef MBS_SUPPORT, following element is length of character + classes, length of collating symbols, length of equivalence + classes, length of character ranges, and length of characters. + Next, character class element, collating symbols elements, + equivalence class elements, range elements, and character + elements follow. + See regex_compile function. */ + charset, + + /* Same parameters as charset, but match any character that is + not one of those specified. */ + charset_not, + + /* Start remembering the text that is matched, for storing in a + register. Followed by one byte with the register number, in + the range 0 to one less than the pattern buffer's re_nsub + field. Then followed by one byte with the number of groups + inner to this one. (This last has to be part of the + start_memory only because we need it in the on_failure_jump + of re_match_2.) */ + start_memory, + + /* Stop remembering the text that is matched and store it in a + memory register. Followed by one byte with the register + number, in the range 0 to one less than `re_nsub' in the + pattern buffer, and one byte with the number of inner groups, + just like `start_memory'. (We need the number of inner + groups here because we don't have any easy way of finding the + corresponding start_memory when we're at a stop_memory.) */ + stop_memory, + + /* Match a duplicate of something remembered. Followed by one + byte containing the register number. */ + duplicate, + + /* Fail unless at beginning of line. */ + begline, + + /* Fail unless at end of line. */ + endline, + + /* Succeeds if at beginning of buffer (if emacs) or at beginning + of string to be matched (if not). */ + begbuf, + + /* Analogously, for end of buffer/string. */ + endbuf, + + /* Followed by two byte relative address to which to jump. */ + jump, + + /* Same as jump, but marks the end of an alternative. */ + jump_past_alt, + + /* Followed by two-byte relative address of place to resume at + in case of failure. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + on_failure_jump, + + /* Like on_failure_jump, but pushes a placeholder instead of the + current string position when executed. */ + on_failure_keep_string_jump, + + /* Throw away latest failure point and then jump to following + two-byte relative address. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + pop_failure_jump, + + /* Change to pop_failure_jump if know won't have to backtrack to + match; otherwise change to jump. This is used to jump + back to the beginning of a repeat. If what follows this jump + clearly won't match what the repeat does, such that we can be + sure that there is no use backtracking out of repetitions + already matched, then we change it to a pop_failure_jump. + Followed by two-byte address. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + maybe_pop_jump, + + /* Jump to following two-byte address, and push a dummy failure + point. This failure point will be thrown away if an attempt + is made to use it for a failure. A `+' construct makes this + before the first repeat. Also used as an intermediary kind + of jump when compiling an alternative. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + dummy_failure_jump, + + /* Push a dummy failure point and continue. Used at the end of + alternatives. */ + push_dummy_failure, + + /* Followed by two-byte relative address and two-byte number n. + After matching N times, jump to the address upon failure. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + succeed_n, + + /* Followed by two-byte relative address, and two-byte number n. + Jump to the address N times, then fail. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + jump_n, + + /* Set the following two-byte relative address to the + subsequent two-byte number. The address *includes* the two + bytes of number. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + set_number_at, + + wordchar, /* Matches any word-constituent character. */ + notwordchar, /* Matches any char that is not a word-constituent. */ + + wordbeg, /* Succeeds if at word beginning. */ + wordend, /* Succeeds if at word end. */ + + wordbound, /* Succeeds if at a word boundary. */ + notwordbound /* Succeeds if not at a word boundary. */ + +#ifdef emacs + ,before_dot, /* Succeeds if before point. */ + at_dot, /* Succeeds if at point. */ + after_dot, /* Succeeds if after point. */ + + /* Matches any character whose syntax is specified. Followed by + a byte which contains a syntax code, e.g., Sword. */ + syntaxspec, + + /* Matches any character whose syntax is not that specified. */ + notsyntaxspec +#endif /* emacs */ +} re_opcode_t; + +/* Common operations on the compiled pattern. */ + +/* Store NUMBER in two contiguous bytes starting at DESTINATION. */ +/* ifdef MBS_SUPPORT, we store NUMBER in 1 element. */ + +#ifdef MBS_SUPPORT +# define STORE_NUMBER(destination, number) \ + do { \ + *(destination) = (US_CHAR_TYPE)(number); \ + } while (0) +#else +# define STORE_NUMBER(destination, number) \ + do { \ + (destination)[0] = (number) & 0377; \ + (destination)[1] = (number) >> 8; \ + } while (0) +#endif /* MBS_SUPPORT */ + +/* Same as STORE_NUMBER, except increment DESTINATION to + the byte after where the number is stored. Therefore, DESTINATION + must be an lvalue. */ +/* ifdef MBS_SUPPORT, we store NUMBER in 1 element. */ + +#define STORE_NUMBER_AND_INCR(destination, number) \ + do { \ + STORE_NUMBER (destination, number); \ + (destination) += OFFSET_ADDRESS_SIZE; \ + } while (0) + +/* Put into DESTINATION a number stored in two contiguous bytes starting + at SOURCE. */ +/* ifdef MBS_SUPPORT, we store NUMBER in 1 element. */ + +#ifdef MBS_SUPPORT +# define EXTRACT_NUMBER(destination, source) \ + do { \ + (destination) = *(source); \ + } while (0) +#else +# define EXTRACT_NUMBER(destination, source) \ + do { \ + (destination) = *(source) & 0377; \ + (destination) += SIGN_EXTEND_CHAR (*((source) + 1)) << 8; \ + } while (0) +#endif + +#ifdef DEBUG +static void extract_number _RE_ARGS ((int *dest, US_CHAR_TYPE *source)); +static void +extract_number (dest, source) + int *dest; + US_CHAR_TYPE *source; +{ +#ifdef MBS_SUPPORT + *dest = *source; +#else + int temp = SIGN_EXTEND_CHAR (*(source + 1)); + *dest = *source & 0377; + *dest += temp << 8; +#endif +} + +# ifndef EXTRACT_MACROS /* To debug the macros. */ +# undef EXTRACT_NUMBER +# define EXTRACT_NUMBER(dest, src) extract_number (&dest, src) +# endif /* not EXTRACT_MACROS */ + +#endif /* DEBUG */ + +/* Same as EXTRACT_NUMBER, except increment SOURCE to after the number. + SOURCE must be an lvalue. */ + +#define EXTRACT_NUMBER_AND_INCR(destination, source) \ + do { \ + EXTRACT_NUMBER (destination, source); \ + (source) += OFFSET_ADDRESS_SIZE; \ + } while (0) + +#ifdef DEBUG +static void extract_number_and_incr _RE_ARGS ((int *destination, + US_CHAR_TYPE **source)); +static void +extract_number_and_incr (destination, source) + int *destination; + US_CHAR_TYPE **source; +{ + extract_number (destination, *source); + *source += OFFSET_ADDRESS_SIZE; +} + +# ifndef EXTRACT_MACROS +# undef EXTRACT_NUMBER_AND_INCR +# define EXTRACT_NUMBER_AND_INCR(dest, src) \ + extract_number_and_incr (&dest, &src) +# endif /* not EXTRACT_MACROS */ + +#endif /* DEBUG */ + +/* If DEBUG is defined, Regex prints many voluminous messages about what + it is doing (if the variable `debug' is nonzero). If linked with the + main program in `iregex.c', you can enter patterns and strings + interactively. And if linked with the main program in `main.c' and + the other test files, you can run the already-written tests. */ + +#ifdef DEBUG + +/* We use standard I/O for debugging. */ +# include <stdio.h> + +/* It is useful to test things that ``must'' be true when debugging. */ +# include <assert.h> + +static int debug; + +# define DEBUG_STATEMENT(e) e +# define DEBUG_PRINT1(x) if (debug) printf (x) +# define DEBUG_PRINT2(x1, x2) if (debug) printf (x1, x2) +# define DEBUG_PRINT3(x1, x2, x3) if (debug) printf (x1, x2, x3) +# define DEBUG_PRINT4(x1, x2, x3, x4) if (debug) printf (x1, x2, x3, x4) +# define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) \ + if (debug) print_partial_compiled_pattern (s, e) +# define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) \ + if (debug) print_double_string (w, s1, sz1, s2, sz2) + + +/* Print the fastmap in human-readable form. */ + +void +print_fastmap (fastmap) + char *fastmap; +{ + unsigned was_a_range = 0; + unsigned i = 0; + + while (i < (1 << BYTEWIDTH)) + { + if (fastmap[i++]) + { + was_a_range = 0; + putchar (i - 1); + while (i < (1 << BYTEWIDTH) && fastmap[i]) + { + was_a_range = 1; + i++; + } + if (was_a_range) + { + printf ("-"); + putchar (i - 1); + } + } + } + putchar ('\n'); +} + + +/* Print a compiled pattern string in human-readable form, starting at + the START pointer into it and ending just before the pointer END. */ + +void +print_partial_compiled_pattern (start, end) + US_CHAR_TYPE *start; + US_CHAR_TYPE *end; +{ + int mcnt, mcnt2; + US_CHAR_TYPE *p1; + US_CHAR_TYPE *p = start; + US_CHAR_TYPE *pend = end; + + if (start == NULL) + { + printf ("(null)\n"); + return; + } + + /* Loop over pattern commands. */ + while (p < pend) + { +#ifdef _LIBC + printf ("%td:\t", p - start); +#else + printf ("%ld:\t", (long int) (p - start)); +#endif + + switch ((re_opcode_t) *p++) + { + case no_op: + printf ("/no_op"); + break; + + case exactn: + mcnt = *p++; + printf ("/exactn/%d", mcnt); + do + { + putchar ('/'); + PUT_CHAR (*p++); + } + while (--mcnt); + break; + +#ifdef MBS_SUPPORT + case exactn_bin: + mcnt = *p++; + printf ("/exactn_bin/%d", mcnt); + do + { + printf("/%lx", (long int) *p++); + } + while (--mcnt); + break; +#endif /* MBS_SUPPORT */ + + case start_memory: + mcnt = *p++; + printf ("/start_memory/%d/%ld", mcnt, (long int) *p++); + break; + + case stop_memory: + mcnt = *p++; + printf ("/stop_memory/%d/%ld", mcnt, (long int) *p++); + break; + + case duplicate: + printf ("/duplicate/%ld", (long int) *p++); + break; + + case anychar: + printf ("/anychar"); + break; + + case charset: + case charset_not: + { +#ifdef MBS_SUPPORT + int i, length; + wchar_t *workp = p; + printf ("/charset [%s", + (re_opcode_t) *(workp - 1) == charset_not ? "^" : ""); + p += 5; + length = *workp++; /* the length of char_classes */ + for (i=0 ; i<length ; i++) + printf("[:%lx:]", (long int) *p++); + length = *workp++; /* the length of collating_symbol */ + for (i=0 ; i<length ;) + { + printf("[."); + while(*p != 0) + PUT_CHAR((i++,*p++)); + i++,p++; + printf(".]"); + } + length = *workp++; /* the length of equivalence_class */ + for (i=0 ; i<length ;) + { + printf("[="); + while(*p != 0) + PUT_CHAR((i++,*p++)); + i++,p++; + printf("=]"); + } + length = *workp++; /* the length of char_range */ + for (i=0 ; i<length ; i++) + { + wchar_t range_start = *p++; + wchar_t range_end = *p++; + if (MB_CUR_MAX == 1) + printf("%c-%c", (char) range_start, (char) range_end); + else + printf("%C-%C", (wint_t) range_start, (wint_t) range_end); + } + length = *workp++; /* the length of char */ + for (i=0 ; i<length ; i++) + if (MB_CUR_MAX == 1) + putchar (*p++); + else + printf("%C", (wint_t) *p++); + putchar (']'); +#else + register int c, last = -100; + register int in_range = 0; + + printf ("/charset [%s", + (re_opcode_t) *(p - 1) == charset_not ? "^" : ""); + + assert (p + *p < pend); + + for (c = 0; c < 256; c++) + if (c / 8 < *p + && (p[1 + (c/8)] & (1 << (c % 8)))) + { + /* Are we starting a range? */ + if (last + 1 == c && ! in_range) + { + putchar ('-'); + in_range = 1; + } + /* Have we broken a range? */ + else if (last + 1 != c && in_range) + { + putchar (last); + in_range = 0; + } + + if (! in_range) + putchar (c); + + last = c; + } + + if (in_range) + putchar (last); + + putchar (']'); + + p += 1 + *p; +#endif /* MBS_SUPPORT */ + } + break; + + case begline: + printf ("/begline"); + break; + + case endline: + printf ("/endline"); + break; + + case on_failure_jump: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/on_failure_jump to %td", p + mcnt - start); +#else + printf ("/on_failure_jump to %ld", (long int) (p + mcnt - start)); +#endif + break; + + case on_failure_keep_string_jump: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/on_failure_keep_string_jump to %td", p + mcnt - start); +#else + printf ("/on_failure_keep_string_jump to %ld", + (long int) (p + mcnt - start)); +#endif + break; + + case dummy_failure_jump: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/dummy_failure_jump to %td", p + mcnt - start); +#else + printf ("/dummy_failure_jump to %ld", (long int) (p + mcnt - start)); +#endif + break; + + case push_dummy_failure: + printf ("/push_dummy_failure"); + break; + + case maybe_pop_jump: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/maybe_pop_jump to %td", p + mcnt - start); +#else + printf ("/maybe_pop_jump to %ld", (long int) (p + mcnt - start)); +#endif + break; + + case pop_failure_jump: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/pop_failure_jump to %td", p + mcnt - start); +#else + printf ("/pop_failure_jump to %ld", (long int) (p + mcnt - start)); +#endif + break; + + case jump_past_alt: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/jump_past_alt to %td", p + mcnt - start); +#else + printf ("/jump_past_alt to %ld", (long int) (p + mcnt - start)); +#endif + break; + + case jump: + extract_number_and_incr (&mcnt, &p); +#ifdef _LIBC + printf ("/jump to %td", p + mcnt - start); +#else + printf ("/jump to %ld", (long int) (p + mcnt - start)); +#endif + break; + + case succeed_n: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); +#ifdef _LIBC + printf ("/succeed_n to %td, %d times", p1 - start, mcnt2); +#else + printf ("/succeed_n to %ld, %d times", + (long int) (p1 - start), mcnt2); +#endif + break; + + case jump_n: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/jump_n to %d, %d times", p1 - start, mcnt2); + break; + + case set_number_at: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); +#ifdef _LIBC + printf ("/set_number_at location %td to %d", p1 - start, mcnt2); +#else + printf ("/set_number_at location %ld to %d", + (long int) (p1 - start), mcnt2); +#endif + break; + + case wordbound: + printf ("/wordbound"); + break; + + case notwordbound: + printf ("/notwordbound"); + break; + + case wordbeg: + printf ("/wordbeg"); + break; + + case wordend: + printf ("/wordend"); + break; + +# ifdef emacs + case before_dot: + printf ("/before_dot"); + break; + + case at_dot: + printf ("/at_dot"); + break; + + case after_dot: + printf ("/after_dot"); + break; + + case syntaxspec: + printf ("/syntaxspec"); + mcnt = *p++; + printf ("/%d", mcnt); + break; + + case notsyntaxspec: + printf ("/notsyntaxspec"); + mcnt = *p++; + printf ("/%d", mcnt); + break; +# endif /* emacs */ + + case wordchar: + printf ("/wordchar"); + break; + + case notwordchar: + printf ("/notwordchar"); + break; + + case begbuf: + printf ("/begbuf"); + break; + + case endbuf: + printf ("/endbuf"); + break; + + default: + printf ("?%ld", (long int) *(p-1)); + } + + putchar ('\n'); + } + +#ifdef _LIBC + printf ("%td:\tend of pattern.\n", p - start); +#else + printf ("%ld:\tend of pattern.\n", (long int) (p - start)); +#endif +} + + +void +print_compiled_pattern (bufp) + struct re_pattern_buffer *bufp; +{ + US_CHAR_TYPE *buffer = (US_CHAR_TYPE*) bufp->buffer; + + print_partial_compiled_pattern (buffer, buffer + + bufp->used / sizeof(US_CHAR_TYPE)); + printf ("%ld bytes used/%ld bytes allocated.\n", + bufp->used, bufp->allocated); + + if (bufp->fastmap_accurate && bufp->fastmap) + { + printf ("fastmap: "); + print_fastmap (bufp->fastmap); + } + +#ifdef _LIBC + printf ("re_nsub: %Zd\t", bufp->re_nsub); +#else + printf ("re_nsub: %ld\t", (long int) bufp->re_nsub); +#endif + printf ("regs_alloc: %d\t", bufp->regs_allocated); + printf ("can_be_null: %d\t", bufp->can_be_null); + printf ("newline_anchor: %d\n", bufp->newline_anchor); + printf ("no_sub: %d\t", bufp->no_sub); + printf ("not_bol: %d\t", bufp->not_bol); + printf ("not_eol: %d\t", bufp->not_eol); + printf ("syntax: %lx\n", bufp->syntax); + /* Perhaps we should print the translate table? */ +} + + +void +print_double_string (where, string1, size1, string2, size2) + const CHAR_TYPE *where; + const CHAR_TYPE *string1; + const CHAR_TYPE *string2; + int size1; + int size2; +{ + int this_char; + + if (where == NULL) + printf ("(null)"); + else + { + if (FIRST_STRING_P (where)) + { + for (this_char = where - string1; this_char < size1; this_char++) + PUT_CHAR (string1[this_char]); + + where = string2; + } + + for (this_char = where - string2; this_char < size2; this_char++) + PUT_CHAR (string2[this_char]); + } +} + +void +printchar (c) + int c; +{ + putc (c, stderr); +} + +#else /* not DEBUG */ + +# undef assert +# define assert(e) + +# define DEBUG_STATEMENT(e) +# define DEBUG_PRINT1(x) +# define DEBUG_PRINT2(x1, x2) +# define DEBUG_PRINT3(x1, x2, x3) +# define DEBUG_PRINT4(x1, x2, x3, x4) +# define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) +# define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) + +#endif /* not DEBUG */ + +#ifdef MBS_SUPPORT +/* This convert a multibyte string to a wide character string. + And write their correspondances to offset_buffer(see below) + and write whether each wchar_t is binary data to is_binary. + This assume invalid multibyte sequences as binary data. + We assume offset_buffer and is_binary is already allocated + enough space. */ + +static size_t convert_mbs_to_wcs (CHAR_TYPE *dest, const unsigned char* src, + size_t len, int *offset_buffer, + char *is_binary); +static size_t +convert_mbs_to_wcs (dest, src, len, offset_buffer, is_binary) + CHAR_TYPE *dest; + const unsigned char* src; + size_t len; /* the length of multibyte string. */ + + /* It hold correspondances between src(char string) and + dest(wchar_t string) for optimization. + e.g. src = "xxxyzz" + dest = {'X', 'Y', 'Z'} + (each "xxx", "y" and "zz" represent one multibyte character + corresponding to 'X', 'Y' and 'Z'.) + offset_buffer = {0, 0+3("xxx"), 0+3+1("y"), 0+3+1+2("zz")} + = {0, 3, 4, 6} + */ + int *offset_buffer; + char *is_binary; +{ + wchar_t *pdest = dest; + const unsigned char *psrc = src; + size_t wc_count = 0; + + if (MB_CUR_MAX == 1) + { /* We don't need conversion. */ + for ( ; wc_count < len ; ++wc_count) + { + *pdest++ = *psrc++; + is_binary[wc_count] = FALSE; + offset_buffer[wc_count] = wc_count; + } + offset_buffer[wc_count] = wc_count; + } + else + { + /* We need conversion. */ + mbstate_t mbs; + int consumed; + size_t mb_remain = len; + size_t mb_count = 0; + + /* Initialize the conversion state. */ + memset (&mbs, 0, sizeof (mbstate_t)); + + offset_buffer[0] = 0; + for( ; mb_remain > 0 ; ++wc_count, ++pdest, mb_remain -= consumed, + psrc += consumed) + { + consumed = mbrtowc (pdest, psrc, mb_remain, &mbs); + + if (consumed <= 0) + /* failed to convert. maybe src contains binary data. + So we consume 1 byte manualy. */ + { + *pdest = *psrc; + consumed = 1; + is_binary[wc_count] = TRUE; + } + else + is_binary[wc_count] = FALSE; + /* In sjis encoding, we use yen sign as escape character in + place of reverse solidus. So we convert 0x5c(yen sign in + sjis) to not 0xa5(yen sign in UCS2) but 0x5c(reverse + solidus in UCS2). */ + if (consumed == 1 && (int) *psrc == 0x5c && (int) *pdest == 0xa5) + *pdest = (wchar_t) *psrc; + + offset_buffer[wc_count + 1] = mb_count += consumed; + } + } + + return wc_count; +} + +#endif /* MBS_SUPPORT */ + +/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can + also be assigned to arbitrarily: each pattern buffer stores its own + syntax, so it can be changed between regex compilations. */ +/* This has no initializer because initialized variables in Emacs + become read-only after dumping. */ +reg_syntax_t re_syntax_options; + + +/* Specify the precise syntax of regexps for compilation. This provides + for compatibility for various utilities which historically have + different, incompatible syntaxes. + + The argument SYNTAX is a bit mask comprised of the various bits + defined in regex.h. We return the old syntax. */ + +reg_syntax_t +re_set_syntax (syntax) + reg_syntax_t syntax; +{ + reg_syntax_t ret = re_syntax_options; + + re_syntax_options = syntax; +#ifdef DEBUG + if (syntax & RE_DEBUG) + debug = 1; + else if (debug) /* was on but now is not */ + debug = 0; +#endif /* DEBUG */ + return ret; +} +#ifdef _LIBC +weak_alias (__re_set_syntax, re_set_syntax) +#endif + +/* This table gives an error message for each of the error codes listed + in regex.h. Obviously the order here has to be same as there. + POSIX doesn't require that we do anything for REG_NOERROR, + but why not be nice? */ + +static const char re_error_msgid[] = + { +#define REG_NOERROR_IDX 0 + gettext_noop ("Success") /* REG_NOERROR */ + "\0" +#define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success") + gettext_noop ("No match") /* REG_NOMATCH */ + "\0" +#define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match") + gettext_noop ("Invalid regular expression") /* REG_BADPAT */ + "\0" +#define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression") + gettext_noop ("Invalid collation character") /* REG_ECOLLATE */ + "\0" +#define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character") + gettext_noop ("Invalid character class name") /* REG_ECTYPE */ + "\0" +#define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name") + gettext_noop ("Trailing backslash") /* REG_EESCAPE */ + "\0" +#define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash") + gettext_noop ("Invalid back reference") /* REG_ESUBREG */ + "\0" +#define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference") + gettext_noop ("Unmatched [ or [^") /* REG_EBRACK */ + "\0" +#define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^") + gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */ + "\0" +#define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(") + gettext_noop ("Unmatched \\{") /* REG_EBRACE */ + "\0" +#define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{") + gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */ + "\0" +#define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}") + gettext_noop ("Invalid range end") /* REG_ERANGE */ + "\0" +#define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end") + gettext_noop ("Memory exhausted") /* REG_ESPACE */ + "\0" +#define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted") + gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */ + "\0" +#define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression") + gettext_noop ("Premature end of regular expression") /* REG_EEND */ + "\0" +#define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression") + gettext_noop ("Regular expression too big") /* REG_ESIZE */ + "\0" +#define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big") + gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */ + }; + +static const size_t re_error_msgid_idx[] = + { + REG_NOERROR_IDX, + REG_NOMATCH_IDX, + REG_BADPAT_IDX, + REG_ECOLLATE_IDX, + REG_ECTYPE_IDX, + REG_EESCAPE_IDX, + REG_ESUBREG_IDX, + REG_EBRACK_IDX, + REG_EPAREN_IDX, + REG_EBRACE_IDX, + REG_BADBR_IDX, + REG_ERANGE_IDX, + REG_ESPACE_IDX, + REG_BADRPT_IDX, + REG_EEND_IDX, + REG_ESIZE_IDX, + REG_ERPAREN_IDX + }; + +/* Avoiding alloca during matching, to placate r_alloc. */ + +/* Define MATCH_MAY_ALLOCATE unless we need to make sure that the + searching and matching functions should not call alloca. On some + systems, alloca is implemented in terms of malloc, and if we're + using the relocating allocator routines, then malloc could cause a + relocation, which might (if the strings being searched are in the + ralloc heap) shift the data out from underneath the regexp + routines. + + Here's another reason to avoid allocation: Emacs + processes input from X in a signal handler; processing X input may + call malloc; if input arrives while a matching routine is calling + malloc, then we're scrod. But Emacs can't just block input while + calling matching routines; then we don't notice interrupts when + they come in. So, Emacs blocks input around all regexp calls + except the matching calls, which it leaves unprotected, in the + faith that they will not malloc. */ + +/* Normally, this is fine. */ +#define MATCH_MAY_ALLOCATE + +/* When using GNU C, we are not REALLY using the C alloca, no matter + what config.h may say. So don't take precautions for it. */ +#ifdef __GNUC__ +# undef C_ALLOCA +#endif + +/* The match routines may not allocate if (1) they would do it with malloc + and (2) it's not safe for them to use malloc. + Note that if REL_ALLOC is defined, matching would not use malloc for the + failure stack, but we would still use it for the register vectors; + so REL_ALLOC should not affect this. */ +#if (defined C_ALLOCA || defined REGEX_MALLOC) && defined emacs +# undef MATCH_MAY_ALLOCATE +#endif + + +/* Failure stack declarations and macros; both re_compile_fastmap and + re_match_2 use a failure stack. These have to be macros because of + REGEX_ALLOCATE_STACK. */ + + +/* Number of failure points for which to initially allocate space + when matching. If this number is exceeded, we allocate more + space, so it is not a hard limit. */ +#ifndef INIT_FAILURE_ALLOC +# define INIT_FAILURE_ALLOC 5 +#endif + +/* Roughly the maximum number of failure points on the stack. Would be + exactly that if always used MAX_FAILURE_ITEMS items each time we failed. + This is a variable only so users of regex can assign to it; we never + change it ourselves. */ + +#ifdef INT_IS_16BIT + +# if defined MATCH_MAY_ALLOCATE +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +long int re_max_failures = 4000; +# else +long int re_max_failures = 2000; +# endif + +union fail_stack_elt +{ + US_CHAR_TYPE *pointer; + long int integer; +}; + +typedef union fail_stack_elt fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned long int size; + unsigned long int avail; /* Offset of next open position. */ +} fail_stack_type; + +#else /* not INT_IS_16BIT */ + +# if defined MATCH_MAY_ALLOCATE +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +int re_max_failures = 4000; +# else +int re_max_failures = 2000; +# endif + +union fail_stack_elt +{ + US_CHAR_TYPE *pointer; + int integer; +}; + +typedef union fail_stack_elt fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} fail_stack_type; + +#endif /* INT_IS_16BIT */ + +#define FAIL_STACK_EMPTY() (fail_stack.avail == 0) +#define FAIL_STACK_PTR_EMPTY() (fail_stack_ptr->avail == 0) +#define FAIL_STACK_FULL() (fail_stack.avail == fail_stack.size) + + +/* Define macros to initialize and free the failure stack. + Do `return -2' if the alloc fails. */ + +#ifdef MATCH_MAY_ALLOCATE +# define INIT_FAIL_STACK() \ + do { \ + fail_stack.stack = (fail_stack_elt_t *) \ + REGEX_ALLOCATE_STACK (INIT_FAILURE_ALLOC * sizeof (fail_stack_elt_t)); \ + \ + if (fail_stack.stack == NULL) \ + return -2; \ + \ + fail_stack.size = INIT_FAILURE_ALLOC; \ + fail_stack.avail = 0; \ + } while (0) + +# define RESET_FAIL_STACK() REGEX_FREE_STACK (fail_stack.stack) +#else +# define INIT_FAIL_STACK() \ + do { \ + fail_stack.avail = 0; \ + } while (0) + +# define RESET_FAIL_STACK() +#endif + + +/* Double the size of FAIL_STACK, up to approximately `re_max_failures' items. + + Return 1 if succeeds, and 0 if either ran out of memory + allocating space for it or it was already too large. + + REGEX_REALLOCATE_STACK requires `destination' be declared. */ + +#define DOUBLE_FAIL_STACK(fail_stack) \ + ((fail_stack).size > (unsigned) (re_max_failures * MAX_FAILURE_ITEMS) \ + ? 0 \ + : ((fail_stack).stack = (fail_stack_elt_t *) \ + REGEX_REALLOCATE_STACK ((fail_stack).stack, \ + (fail_stack).size * sizeof (fail_stack_elt_t), \ + ((fail_stack).size << 1) * sizeof (fail_stack_elt_t)), \ + \ + (fail_stack).stack == NULL \ + ? 0 \ + : ((fail_stack).size <<= 1, \ + 1))) + + +/* Push pointer POINTER on FAIL_STACK. + Return 1 if was able to do so and 0 if ran out of memory allocating + space to do so. */ +#define PUSH_PATTERN_OP(POINTER, FAIL_STACK) \ + ((FAIL_STACK_FULL () \ + && !DOUBLE_FAIL_STACK (FAIL_STACK)) \ + ? 0 \ + : ((FAIL_STACK).stack[(FAIL_STACK).avail++].pointer = POINTER, \ + 1)) + +/* Push a pointer value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_POINTER(item) \ + fail_stack.stack[fail_stack.avail++].pointer = (US_CHAR_TYPE *) (item) + +/* This pushes an integer-valued item onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_INT(item) \ + fail_stack.stack[fail_stack.avail++].integer = (item) + +/* Push a fail_stack_elt_t value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_ELT(item) \ + fail_stack.stack[fail_stack.avail++] = (item) + +/* These three POP... operations complement the three PUSH... operations. + All assume that `fail_stack' is nonempty. */ +#define POP_FAILURE_POINTER() fail_stack.stack[--fail_stack.avail].pointer +#define POP_FAILURE_INT() fail_stack.stack[--fail_stack.avail].integer +#define POP_FAILURE_ELT() fail_stack.stack[--fail_stack.avail] + +/* Used to omit pushing failure point id's when we're not debugging. */ +#ifdef DEBUG +# define DEBUG_PUSH PUSH_FAILURE_INT +# define DEBUG_POP(item_addr) *(item_addr) = POP_FAILURE_INT () +#else +# define DEBUG_PUSH(item) +# define DEBUG_POP(item_addr) +#endif + + +/* Push the information about the state we will need + if we ever fail back to it. + + Requires variables fail_stack, regstart, regend, reg_info, and + num_regs_pushed be declared. DOUBLE_FAIL_STACK requires `destination' + be declared. + + Does `return FAILURE_CODE' if runs out of memory. */ + +#define PUSH_FAILURE_POINT(pattern_place, string_place, failure_code) \ + do { \ + char *destination; \ + /* Must be int, so when we don't save any registers, the arithmetic \ + of 0 + -1 isn't done as unsigned. */ \ + /* Can't be int, since there is not a shred of a guarantee that int \ + is wide enough to hold a value of something to which pointer can \ + be assigned */ \ + active_reg_t this_reg; \ + \ + DEBUG_STATEMENT (failure_id++); \ + DEBUG_STATEMENT (nfailure_points_pushed++); \ + DEBUG_PRINT2 ("\nPUSH_FAILURE_POINT #%u:\n", failure_id); \ + DEBUG_PRINT2 (" Before push, next avail: %d\n", (fail_stack).avail);\ + DEBUG_PRINT2 (" size: %d\n", (fail_stack).size);\ + \ + DEBUG_PRINT2 (" slots needed: %ld\n", NUM_FAILURE_ITEMS); \ + DEBUG_PRINT2 (" available: %d\n", REMAINING_AVAIL_SLOTS); \ + \ + /* Ensure we have enough space allocated for what we will push. */ \ + while (REMAINING_AVAIL_SLOTS < NUM_FAILURE_ITEMS) \ + { \ + if (!DOUBLE_FAIL_STACK (fail_stack)) \ + return failure_code; \ + \ + DEBUG_PRINT2 ("\n Doubled stack; size now: %d\n", \ + (fail_stack).size); \ + DEBUG_PRINT2 (" slots available: %d\n", REMAINING_AVAIL_SLOTS);\ + } \ + \ + /* Push the info, starting with the registers. */ \ + DEBUG_PRINT1 ("\n"); \ + \ + if (1) \ + for (this_reg = lowest_active_reg; this_reg <= highest_active_reg; \ + this_reg++) \ + { \ + DEBUG_PRINT2 (" Pushing reg: %lu\n", this_reg); \ + DEBUG_STATEMENT (num_regs_pushed++); \ + \ + DEBUG_PRINT2 (" start: %p\n", regstart[this_reg]); \ + PUSH_FAILURE_POINTER (regstart[this_reg]); \ + \ + DEBUG_PRINT2 (" end: %p\n", regend[this_reg]); \ + PUSH_FAILURE_POINTER (regend[this_reg]); \ + \ + DEBUG_PRINT2 (" info: %p\n ", \ + reg_info[this_reg].word.pointer); \ + DEBUG_PRINT2 (" match_null=%d", \ + REG_MATCH_NULL_STRING_P (reg_info[this_reg])); \ + DEBUG_PRINT2 (" active=%d", IS_ACTIVE (reg_info[this_reg])); \ + DEBUG_PRINT2 (" matched_something=%d", \ + MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT2 (" ever_matched=%d", \ + EVER_MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT1 ("\n"); \ + PUSH_FAILURE_ELT (reg_info[this_reg].word); \ + } \ + \ + DEBUG_PRINT2 (" Pushing low active reg: %ld\n", lowest_active_reg);\ + PUSH_FAILURE_INT (lowest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing high active reg: %ld\n", highest_active_reg);\ + PUSH_FAILURE_INT (highest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing pattern %p:\n", pattern_place); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pattern_place, pend); \ + PUSH_FAILURE_POINTER (pattern_place); \ + \ + DEBUG_PRINT2 (" Pushing string %p: `", string_place); \ + DEBUG_PRINT_DOUBLE_STRING (string_place, string1, size1, string2, \ + size2); \ + DEBUG_PRINT1 ("'\n"); \ + PUSH_FAILURE_POINTER (string_place); \ + \ + DEBUG_PRINT2 (" Pushing failure id: %u\n", failure_id); \ + DEBUG_PUSH (failure_id); \ + } while (0) + +/* This is the number of items that are pushed and popped on the stack + for each register. */ +#define NUM_REG_ITEMS 3 + +/* Individual items aside from the registers. */ +#ifdef DEBUG +# define NUM_NONREG_ITEMS 5 /* Includes failure point id. */ +#else +# define NUM_NONREG_ITEMS 4 +#endif + +/* We push at most this many items on the stack. */ +/* We used to use (num_regs - 1), which is the number of registers + this regexp will save; but that was changed to 5 + to avoid stack overflow for a regexp with lots of parens. */ +#define MAX_FAILURE_ITEMS (5 * NUM_REG_ITEMS + NUM_NONREG_ITEMS) + +/* We actually push this many items. */ +#define NUM_FAILURE_ITEMS \ + (((0 \ + ? 0 : highest_active_reg - lowest_active_reg + 1) \ + * NUM_REG_ITEMS) \ + + NUM_NONREG_ITEMS) + +/* How many items can still be added to the stack without overflowing it. */ +#define REMAINING_AVAIL_SLOTS ((fail_stack).size - (fail_stack).avail) + + +/* Pops what PUSH_FAIL_STACK pushes. + + We restore into the parameters, all of which should be lvalues: + STR -- the saved data position. + PAT -- the saved pattern position. + LOW_REG, HIGH_REG -- the highest and lowest active registers. + REGSTART, REGEND -- arrays of string positions. + REG_INFO -- array of information about each subexpression. + + Also assumes the variables `fail_stack' and (if debugging), `bufp', + `pend', `string1', `size1', `string2', and `size2'. */ +#define POP_FAILURE_POINT(str, pat, low_reg, high_reg, regstart, regend, reg_info)\ +{ \ + DEBUG_STATEMENT (unsigned failure_id;) \ + active_reg_t this_reg; \ + const US_CHAR_TYPE *string_temp; \ + \ + assert (!FAIL_STACK_EMPTY ()); \ + \ + /* Remove failure points and point to how many regs pushed. */ \ + DEBUG_PRINT1 ("POP_FAILURE_POINT:\n"); \ + DEBUG_PRINT2 (" Before pop, next avail: %d\n", fail_stack.avail); \ + DEBUG_PRINT2 (" size: %d\n", fail_stack.size); \ + \ + assert (fail_stack.avail >= NUM_NONREG_ITEMS); \ + \ + DEBUG_POP (&failure_id); \ + DEBUG_PRINT2 (" Popping failure id: %u\n", failure_id); \ + \ + /* If the saved string location is NULL, it came from an \ + on_failure_keep_string_jump opcode, and we want to throw away the \ + saved NULL, thus retaining our current position in the string. */ \ + string_temp = POP_FAILURE_POINTER (); \ + if (string_temp != NULL) \ + str = (const CHAR_TYPE *) string_temp; \ + \ + DEBUG_PRINT2 (" Popping string %p: `", str); \ + DEBUG_PRINT_DOUBLE_STRING (str, string1, size1, string2, size2); \ + DEBUG_PRINT1 ("'\n"); \ + \ + pat = (US_CHAR_TYPE *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" Popping pattern %p:\n", pat); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pat, pend); \ + \ + /* Restore register info. */ \ + high_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping high active reg: %ld\n", high_reg); \ + \ + low_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping low active reg: %ld\n", low_reg); \ + \ + if (1) \ + for (this_reg = high_reg; this_reg >= low_reg; this_reg--) \ + { \ + DEBUG_PRINT2 (" Popping reg: %ld\n", this_reg); \ + \ + reg_info[this_reg].word = POP_FAILURE_ELT (); \ + DEBUG_PRINT2 (" info: %p\n", \ + reg_info[this_reg].word.pointer); \ + \ + regend[this_reg] = (const CHAR_TYPE *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" end: %p\n", regend[this_reg]); \ + \ + regstart[this_reg] = (const CHAR_TYPE *) POP_FAILURE_POINTER ();\ + DEBUG_PRINT2 (" start: %p\n", regstart[this_reg]); \ + } \ + else \ + { \ + for (this_reg = highest_active_reg; this_reg > high_reg; this_reg--) \ + { \ + reg_info[this_reg].word.integer = 0; \ + regend[this_reg] = 0; \ + regstart[this_reg] = 0; \ + } \ + highest_active_reg = high_reg; \ + } \ + \ + set_regs_matched_done = 0; \ + DEBUG_STATEMENT (nfailure_points_popped++); \ +} /* POP_FAILURE_POINT */ + + +/* Structure for per-register (a.k.a. per-group) information. + Other register information, such as the + starting and ending positions (which are addresses), and the list of + inner groups (which is a bits list) are maintained in separate + variables. + + We are making a (strictly speaking) nonportable assumption here: that + the compiler will pack our bit fields into something that fits into + the type of `word', i.e., is something that fits into one item on the + failure stack. */ + + +/* Declarations and macros for re_match_2. */ + +typedef union +{ + fail_stack_elt_t word; + struct + { + /* This field is one if this group can match the empty string, + zero if not. If not yet determined, `MATCH_NULL_UNSET_VALUE'. */ +#define MATCH_NULL_UNSET_VALUE 3 + unsigned match_null_string_p : 2; + unsigned is_active : 1; + unsigned matched_something : 1; + unsigned ever_matched_something : 1; + } bits; +} register_info_type; + +#define REG_MATCH_NULL_STRING_P(R) ((R).bits.match_null_string_p) +#define IS_ACTIVE(R) ((R).bits.is_active) +#define MATCHED_SOMETHING(R) ((R).bits.matched_something) +#define EVER_MATCHED_SOMETHING(R) ((R).bits.ever_matched_something) + + +/* Call this when have matched a real character; it sets `matched' flags + for the subexpressions which we are currently inside. Also records + that those subexprs have matched. */ +#define SET_REGS_MATCHED() \ + do \ + { \ + if (!set_regs_matched_done) \ + { \ + active_reg_t r; \ + set_regs_matched_done = 1; \ + for (r = lowest_active_reg; r <= highest_active_reg; r++) \ + { \ + MATCHED_SOMETHING (reg_info[r]) \ + = EVER_MATCHED_SOMETHING (reg_info[r]) \ + = 1; \ + } \ + } \ + } \ + while (0) + +/* Registers are set to a sentinel when they haven't yet matched. */ +static CHAR_TYPE reg_unset_dummy; +#define REG_UNSET_VALUE (®_unset_dummy) +#define REG_UNSET(e) ((e) == REG_UNSET_VALUE) + +/* Subroutine declarations and macros for regex_compile. */ + +static reg_errcode_t regex_compile _RE_ARGS ((const char *pattern, size_t size, + reg_syntax_t syntax, + struct re_pattern_buffer *bufp)); +static void store_op1 _RE_ARGS ((re_opcode_t op, US_CHAR_TYPE *loc, int arg)); +static void store_op2 _RE_ARGS ((re_opcode_t op, US_CHAR_TYPE *loc, + int arg1, int arg2)); +static void insert_op1 _RE_ARGS ((re_opcode_t op, US_CHAR_TYPE *loc, + int arg, US_CHAR_TYPE *end)); +static void insert_op2 _RE_ARGS ((re_opcode_t op, US_CHAR_TYPE *loc, + int arg1, int arg2, US_CHAR_TYPE *end)); +static boolean at_begline_loc_p _RE_ARGS ((const CHAR_TYPE *pattern, + const CHAR_TYPE *p, + reg_syntax_t syntax)); +static boolean at_endline_loc_p _RE_ARGS ((const CHAR_TYPE *p, + const CHAR_TYPE *pend, + reg_syntax_t syntax)); +#ifdef MBS_SUPPORT +static reg_errcode_t compile_range _RE_ARGS ((CHAR_TYPE range_start, + const CHAR_TYPE **p_ptr, + const CHAR_TYPE *pend, + char *translate, + reg_syntax_t syntax, + US_CHAR_TYPE *b, + CHAR_TYPE *char_set)); +static void insert_space _RE_ARGS ((int num, CHAR_TYPE *loc, CHAR_TYPE *end)); +#else +static reg_errcode_t compile_range _RE_ARGS ((unsigned int range_start, + const CHAR_TYPE **p_ptr, + const CHAR_TYPE *pend, + char *translate, + reg_syntax_t syntax, + US_CHAR_TYPE *b)); +#endif /* MBS_SUPPORT */ + +/* Fetch the next character in the uncompiled pattern---translating it + if necessary. Also cast from a signed character in the constant + string passed to us by the user to an unsigned char that we can use + as an array index (in, e.g., `translate'). */ +/* ifdef MBS_SUPPORT, we translate only if character <= 0xff, + because it is impossible to allocate 4GB array for some encodings + which have 4 byte character_set like UCS4. */ +#ifndef PATFETCH +# ifdef MBS_SUPPORT +# define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (US_CHAR_TYPE) *p++; \ + if (translate && (c <= 0xff)) c = (US_CHAR_TYPE) translate[c]; \ + } while (0) +# else +# define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + if (translate) c = (unsigned char) translate[c]; \ + } while (0) +# endif /* MBS_SUPPORT */ +#endif + +/* Fetch the next character in the uncompiled pattern, with no + translation. */ +#define PATFETCH_RAW(c) \ + do {if (p == pend) return REG_EEND; \ + c = (US_CHAR_TYPE) *p++; \ + } while (0) + +/* Go backwards one character in the pattern. */ +#define PATUNFETCH p-- + + +/* If `translate' is non-null, return translate[D], else just D. We + cast the subscript to translate because some data is declared as + `char *', to avoid warnings when a string constant is passed. But + when we use a character as a subscript we must make it unsigned. */ +/* ifdef MBS_SUPPORT, we translate only if character <= 0xff, + because it is impossible to allocate 4GB array for some encodings + which have 4 byte character_set like UCS4. */ +#ifndef TRANSLATE +# ifdef MBS_SUPPORT +# define TRANSLATE(d) \ + ((translate && ((US_CHAR_TYPE) (d)) <= 0xff) \ + ? (char) translate[(unsigned char) (d)] : (d)) +#else +# define TRANSLATE(d) \ + (translate ? (char) translate[(unsigned char) (d)] : (d)) +# endif /* MBS_SUPPORT */ +#endif + + +/* Macros for outputting the compiled pattern into `buffer'. */ + +/* If the buffer isn't allocated when it comes in, use this. */ +#define INIT_BUF_SIZE (32 * sizeof(US_CHAR_TYPE)) + +/* Make sure we have at least N more bytes of space in buffer. */ +#ifdef MBS_SUPPORT +# define GET_BUFFER_SPACE(n) \ + while (((unsigned long)b - (unsigned long)COMPILED_BUFFER_VAR \ + + (n)*sizeof(CHAR_TYPE)) > bufp->allocated) \ + EXTEND_BUFFER () +#else +# define GET_BUFFER_SPACE(n) \ + while ((unsigned long) (b - bufp->buffer + (n)) > bufp->allocated) \ + EXTEND_BUFFER () +#endif /* MBS_SUPPORT */ + +/* Make sure we have one more byte of buffer space and then add C to it. */ +#define BUF_PUSH(c) \ + do { \ + GET_BUFFER_SPACE (1); \ + *b++ = (US_CHAR_TYPE) (c); \ + } while (0) + + +/* Ensure we have two more bytes of buffer space and then append C1 and C2. */ +#define BUF_PUSH_2(c1, c2) \ + do { \ + GET_BUFFER_SPACE (2); \ + *b++ = (US_CHAR_TYPE) (c1); \ + *b++ = (US_CHAR_TYPE) (c2); \ + } while (0) + + +/* As with BUF_PUSH_2, except for three bytes. */ +#define BUF_PUSH_3(c1, c2, c3) \ + do { \ + GET_BUFFER_SPACE (3); \ + *b++ = (US_CHAR_TYPE) (c1); \ + *b++ = (US_CHAR_TYPE) (c2); \ + *b++ = (US_CHAR_TYPE) (c3); \ + } while (0) + +/* Store a jump with opcode OP at LOC to location TO. We store a + relative address offset by the three bytes the jump itself occupies. */ +#define STORE_JUMP(op, loc, to) \ + store_op1 (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE))) + +/* Likewise, for a two-argument jump. */ +#define STORE_JUMP2(op, loc, to, arg) \ + store_op2 (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE)), arg) + +/* Like `STORE_JUMP', but for inserting. Assume `b' is the buffer end. */ +#define INSERT_JUMP(op, loc, to) \ + insert_op1 (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE)), b) + +/* Like `STORE_JUMP2', but for inserting. Assume `b' is the buffer end. */ +#define INSERT_JUMP2(op, loc, to, arg) \ + insert_op2 (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE)),\ + arg, b) + + +/* This is not an arbitrary limit: the arguments which represent offsets + into the pattern are two bytes long. So if 2^16 bytes turns out to + be too small, many things would have to change. */ +/* Any other compiler which, like MSC, has allocation limit below 2^16 + bytes will have to use approach similar to what was done below for + MSC and drop MAX_BUF_SIZE a bit. Otherwise you may end up + reallocating to 0 bytes. Such thing is not going to work too well. + You have been warned!! */ +#if defined _MSC_VER && !defined WIN32 +/* Microsoft C 16-bit versions limit malloc to approx 65512 bytes. + The REALLOC define eliminates a flurry of conversion warnings, + but is not required. */ +# define MAX_BUF_SIZE 65500L +# define REALLOC(p,s) realloc ((p), (size_t) (s)) +#else +# define MAX_BUF_SIZE (1L << 16) +# define REALLOC(p,s) realloc ((p), (s)) +#endif + +/* Extend the buffer by twice its current size via realloc and + reset the pointers that pointed into the old block to point to the + correct places in the new one. If extending the buffer results in it + being larger than MAX_BUF_SIZE, then flag memory exhausted. */ +#if __BOUNDED_POINTERS__ +# define SET_HIGH_BOUND(P) (__ptrhigh (P) = __ptrlow (P) + bufp->allocated) +# define MOVE_BUFFER_POINTER(P) \ + (__ptrlow (P) += incr, SET_HIGH_BOUND (P), __ptrvalue (P) += incr) +# define ELSE_EXTEND_BUFFER_HIGH_BOUND \ + else \ + { \ + SET_HIGH_BOUND (b); \ + SET_HIGH_BOUND (begalt); \ + if (fixup_alt_jump) \ + SET_HIGH_BOUND (fixup_alt_jump); \ + if (laststart) \ + SET_HIGH_BOUND (laststart); \ + if (pending_exact) \ + SET_HIGH_BOUND (pending_exact); \ + } +#else +# define MOVE_BUFFER_POINTER(P) (P) += incr +# define ELSE_EXTEND_BUFFER_HIGH_BOUND +#endif + +#ifdef MBS_SUPPORT +# define EXTEND_BUFFER() \ + do { \ + US_CHAR_TYPE *old_buffer = COMPILED_BUFFER_VAR; \ + int wchar_count; \ + if (bufp->allocated + sizeof(US_CHAR_TYPE) > MAX_BUF_SIZE) \ + return REG_ESIZE; \ + bufp->allocated <<= 1; \ + if (bufp->allocated > MAX_BUF_SIZE) \ + bufp->allocated = MAX_BUF_SIZE; \ + /* How many characters the new buffer can have? */ \ + wchar_count = bufp->allocated / sizeof(US_CHAR_TYPE); \ + if (wchar_count == 0) wchar_count = 1; \ + /* Truncate the buffer to CHAR_TYPE align. */ \ + bufp->allocated = wchar_count * sizeof(US_CHAR_TYPE); \ + RETALLOC (COMPILED_BUFFER_VAR, wchar_count, US_CHAR_TYPE); \ + bufp->buffer = (char*)COMPILED_BUFFER_VAR; \ + if (COMPILED_BUFFER_VAR == NULL) \ + return REG_ESPACE; \ + /* If the buffer moved, move all the pointers into it. */ \ + if (old_buffer != COMPILED_BUFFER_VAR) \ + { \ + int incr = COMPILED_BUFFER_VAR - old_buffer; \ + MOVE_BUFFER_POINTER (b); \ + MOVE_BUFFER_POINTER (begalt); \ + if (fixup_alt_jump) \ + MOVE_BUFFER_POINTER (fixup_alt_jump); \ + if (laststart) \ + MOVE_BUFFER_POINTER (laststart); \ + if (pending_exact) \ + MOVE_BUFFER_POINTER (pending_exact); \ + } \ + ELSE_EXTEND_BUFFER_HIGH_BOUND \ + } while (0) +#else +# define EXTEND_BUFFER() \ + do { \ + US_CHAR_TYPE *old_buffer = COMPILED_BUFFER_VAR; \ + if (bufp->allocated == MAX_BUF_SIZE) \ + return REG_ESIZE; \ + bufp->allocated <<= 1; \ + if (bufp->allocated > MAX_BUF_SIZE) \ + bufp->allocated = MAX_BUF_SIZE; \ + bufp->buffer = (US_CHAR_TYPE *) REALLOC (COMPILED_BUFFER_VAR, \ + bufp->allocated); \ + if (COMPILED_BUFFER_VAR == NULL) \ + return REG_ESPACE; \ + /* If the buffer moved, move all the pointers into it. */ \ + if (old_buffer != COMPILED_BUFFER_VAR) \ + { \ + int incr = COMPILED_BUFFER_VAR - old_buffer; \ + MOVE_BUFFER_POINTER (b); \ + MOVE_BUFFER_POINTER (begalt); \ + if (fixup_alt_jump) \ + MOVE_BUFFER_POINTER (fixup_alt_jump); \ + if (laststart) \ + MOVE_BUFFER_POINTER (laststart); \ + if (pending_exact) \ + MOVE_BUFFER_POINTER (pending_exact); \ + } \ + ELSE_EXTEND_BUFFER_HIGH_BOUND \ + } while (0) +#endif /* MBS_SUPPORT */ + +/* Since we have one byte reserved for the register number argument to + {start,stop}_memory, the maximum number of groups we can report + things about is what fits in that byte. */ +#define MAX_REGNUM 255 + +/* But patterns can have more than `MAX_REGNUM' registers. We just + ignore the excess. */ +typedef unsigned regnum_t; + + +/* Macros for the compile stack. */ + +/* Since offsets can go either forwards or backwards, this type needs to + be able to hold values from -(MAX_BUF_SIZE - 1) to MAX_BUF_SIZE - 1. */ +/* int may be not enough when sizeof(int) == 2. */ +typedef long pattern_offset_t; + +typedef struct +{ + pattern_offset_t begalt_offset; + pattern_offset_t fixup_alt_jump; + pattern_offset_t inner_group_offset; + pattern_offset_t laststart_offset; + regnum_t regnum; +} compile_stack_elt_t; + + +typedef struct +{ + compile_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} compile_stack_type; + + +#define INIT_COMPILE_STACK_SIZE 32 + +#define COMPILE_STACK_EMPTY (compile_stack.avail == 0) +#define COMPILE_STACK_FULL (compile_stack.avail == compile_stack.size) + +/* The next available element. */ +#define COMPILE_STACK_TOP (compile_stack.stack[compile_stack.avail]) + + +/* Set the bit for character C in a list. */ +#define SET_LIST_BIT(c) \ + (b[((unsigned char) (c)) / BYTEWIDTH] \ + |= 1 << (((unsigned char) c) % BYTEWIDTH)) + + +/* Get the next unsigned number in the uncompiled pattern. */ +#define GET_UNSIGNED_NUMBER(num) \ + { if (p != pend) \ + { \ + PATFETCH (c); \ + while ('0' <= c && c <= '9') \ + { \ + if (num < 0) \ + num = 0; \ + num = num * 10 + c - '0'; \ + if (p == pend) \ + break; \ + PATFETCH (c); \ + } \ + } \ + } + +#if defined _LIBC || WIDE_CHAR_SUPPORT +/* The GNU C library provides support for user-defined character classes + and the functions from ISO C amendement 1. */ +# ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +# else +/* This shouldn't happen but some implementation might still have this + problem. Use a reasonable default value. */ +# define CHAR_CLASS_MAX_LENGTH 256 +# endif + +# ifdef _LIBC +# define IS_CHAR_CLASS(string) __wctype (string) +# else +# define IS_CHAR_CLASS(string) wctype (string) +# endif +#else +# define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ + +# define IS_CHAR_CLASS(string) \ + (STREQ (string, "alpha") || STREQ (string, "upper") \ + || STREQ (string, "lower") || STREQ (string, "digit") \ + || STREQ (string, "alnum") || STREQ (string, "xdigit") \ + || STREQ (string, "space") || STREQ (string, "print") \ + || STREQ (string, "punct") || STREQ (string, "graph") \ + || STREQ (string, "cntrl") || STREQ (string, "blank")) +#endif + +#ifndef MATCH_MAY_ALLOCATE + +/* If we cannot allocate large objects within re_match_2_internal, + we make the fail stack and register vectors global. + The fail stack, we grow to the maximum size when a regexp + is compiled. + The register vectors, we adjust in size each time we + compile a regexp, according to the number of registers it needs. */ + +static fail_stack_type fail_stack; + +/* Size with which the following vectors are currently allocated. + That is so we can make them bigger as needed, + but never make them smaller. */ +static int regs_allocated_size; + +static const char ** regstart, ** regend; +static const char ** old_regstart, ** old_regend; +static const char **best_regstart, **best_regend; +static register_info_type *reg_info; +static const char **reg_dummy; +static register_info_type *reg_info_dummy; + +/* Make the register vectors big enough for NUM_REGS registers, + but don't make them smaller. */ + +static +regex_grow_registers (num_regs) + int num_regs; +{ + if (num_regs > regs_allocated_size) + { + RETALLOC_IF (regstart, num_regs, const char *); + RETALLOC_IF (regend, num_regs, const char *); + RETALLOC_IF (old_regstart, num_regs, const char *); + RETALLOC_IF (old_regend, num_regs, const char *); + RETALLOC_IF (best_regstart, num_regs, const char *); + RETALLOC_IF (best_regend, num_regs, const char *); + RETALLOC_IF (reg_info, num_regs, register_info_type); + RETALLOC_IF (reg_dummy, num_regs, const char *); + RETALLOC_IF (reg_info_dummy, num_regs, register_info_type); + + regs_allocated_size = num_regs; + } +} + +#endif /* not MATCH_MAY_ALLOCATE */ + +static boolean group_in_compile_stack _RE_ARGS ((compile_stack_type + compile_stack, + regnum_t regnum)); + +/* `regex_compile' compiles PATTERN (of length SIZE) according to SYNTAX. + Returns one of error codes defined in `regex.h', or zero for success. + + Assumes the `allocated' (and perhaps `buffer') and `translate' + fields are set in BUFP on entry. + + If it succeeds, results are put in BUFP (if it returns an error, the + contents of BUFP are undefined): + `buffer' is the compiled pattern; + `syntax' is set to SYNTAX; + `used' is set to the length of the compiled pattern; + `fastmap_accurate' is zero; + `re_nsub' is the number of subexpressions in PATTERN; + `not_bol' and `not_eol' are zero; + + The `fastmap' and `newline_anchor' fields are neither + examined nor set. */ + +/* Return, freeing storage we allocated. */ +#ifdef MBS_SUPPORT +# define FREE_STACK_RETURN(value) \ + return (free(pattern), free(mbs_offset), free(is_binary), free (compile_stack.stack), value) +#else +# define FREE_STACK_RETURN(value) \ + return (free (compile_stack.stack), value) +#endif /* MBS_SUPPORT */ + +static reg_errcode_t +#ifdef MBS_SUPPORT +regex_compile (cpattern, csize, syntax, bufp) + const char *cpattern; + size_t csize; +#else +regex_compile (pattern, size, syntax, bufp) + const char *pattern; + size_t size; +#endif /* MBS_SUPPORT */ + reg_syntax_t syntax; + struct re_pattern_buffer *bufp; +{ + /* We fetch characters from PATTERN here. Even though PATTERN is + `char *' (i.e., signed), we declare these variables as unsigned, so + they can be reliably used as array indices. */ + register US_CHAR_TYPE c, c1; + +#ifdef MBS_SUPPORT + /* A temporary space to keep wchar_t pattern and compiled pattern. */ + CHAR_TYPE *pattern, *COMPILED_BUFFER_VAR; + size_t size; + /* offset buffer for optimizatoin. See convert_mbs_to_wc. */ + int *mbs_offset = NULL; + /* It hold whether each wchar_t is binary data or not. */ + char *is_binary = NULL; + /* A flag whether exactn is handling binary data or not. */ + char is_exactn_bin = FALSE; +#endif /* MBS_SUPPORT */ + + /* A random temporary spot in PATTERN. */ + const CHAR_TYPE *p1; + + /* Points to the end of the buffer, where we should append. */ + register US_CHAR_TYPE *b; + + /* Keeps track of unclosed groups. */ + compile_stack_type compile_stack; + + /* Points to the current (ending) position in the pattern. */ +#ifdef MBS_SUPPORT + const CHAR_TYPE *p; + const CHAR_TYPE *pend; +#else + const CHAR_TYPE *p = pattern; + const CHAR_TYPE *pend = pattern + size; +#endif /* MBS_SUPPORT */ + + /* How to translate the characters in the pattern. */ + RE_TRANSLATE_TYPE translate = bufp->translate; + + /* Address of the count-byte of the most recently inserted `exactn' + command. This makes it possible to tell if a new exact-match + character can be added to that command or if the character requires + a new `exactn' command. */ + US_CHAR_TYPE *pending_exact = 0; + + /* Address of start of the most recently finished expression. + This tells, e.g., postfix * where to find the start of its + operand. Reset at the beginning of groups and alternatives. */ + US_CHAR_TYPE *laststart = 0; + + /* Address of beginning of regexp, or inside of last group. */ + US_CHAR_TYPE *begalt; + + /* Place in the uncompiled pattern (i.e., the {) to + which to go back if the interval is invalid. */ +#ifdef MBS_SUPPORT + const US_CHAR_TYPE *beg_interval; +#else + const char *beg_interval; +#endif /* MBS_SUPPORT */ + + /* Address of the place where a forward jump should go to the end of + the containing expression. Each alternative of an `or' -- except the + last -- ends with a forward jump of this sort. */ + US_CHAR_TYPE *fixup_alt_jump = 0; + + /* Counts open-groups as they are encountered. Remembered for the + matching close-group on the compile stack, so the same register + number is put in the stop_memory as the start_memory. */ + regnum_t regnum = 0; + +#ifdef MBS_SUPPORT + /* Initialize the wchar_t PATTERN and offset_buffer. */ + p = pend = pattern = TALLOC(csize, CHAR_TYPE); + mbs_offset = TALLOC(csize + 1, int); + is_binary = TALLOC(csize + 1, char); + if (pattern == NULL || mbs_offset == NULL || is_binary == NULL) + { + if (pattern) free(pattern); + if (mbs_offset) free(mbs_offset); + if (is_binary) free(is_binary); + return REG_ESPACE; + } + size = convert_mbs_to_wcs(pattern, cpattern, csize, mbs_offset, is_binary); + pend = p + size; + if (size < 0) + { + if (pattern) free(pattern); + if (mbs_offset) free(mbs_offset); + if (is_binary) free(is_binary); + return REG_BADPAT; + } +#endif + +#ifdef DEBUG + DEBUG_PRINT1 ("\nCompiling pattern: "); + if (debug) + { + unsigned debug_count; + + for (debug_count = 0; debug_count < size; debug_count++) + PUT_CHAR (pattern[debug_count]); + putchar ('\n'); + } +#endif /* DEBUG */ + + /* Initialize the compile stack. */ + compile_stack.stack = TALLOC (INIT_COMPILE_STACK_SIZE, compile_stack_elt_t); + if (compile_stack.stack == NULL) + { +#ifdef MBS_SUPPORT + if (pattern) free(pattern); + if (mbs_offset) free(mbs_offset); + if (is_binary) free(is_binary); +#endif + return REG_ESPACE; + } + + compile_stack.size = INIT_COMPILE_STACK_SIZE; + compile_stack.avail = 0; + + /* Initialize the pattern buffer. */ + bufp->syntax = syntax; + bufp->fastmap_accurate = 0; + bufp->not_bol = bufp->not_eol = 0; + + /* Set `used' to zero, so that if we return an error, the pattern + printer (for debugging) will think there's no pattern. We reset it + at the end. */ + bufp->used = 0; + + /* Always count groups, whether or not bufp->no_sub is set. */ + bufp->re_nsub = 0; + +#if !defined emacs && !defined SYNTAX_TABLE + /* Initialize the syntax table. */ + init_syntax_once (); +#endif + + if (bufp->allocated == 0) + { + if (bufp->buffer) + { /* If zero allocated, but buffer is non-null, try to realloc + enough space. This loses if buffer's address is bogus, but + that is the user's responsibility. */ +#ifdef MBS_SUPPORT + /* Free bufp->buffer and allocate an array for wchar_t pattern + buffer. */ + free(bufp->buffer); + COMPILED_BUFFER_VAR = TALLOC (INIT_BUF_SIZE/sizeof(US_CHAR_TYPE), + US_CHAR_TYPE); +#else + RETALLOC (COMPILED_BUFFER_VAR, INIT_BUF_SIZE, US_CHAR_TYPE); +#endif /* MBS_SUPPORT */ + } + else + { /* Caller did not allocate a buffer. Do it for them. */ + COMPILED_BUFFER_VAR = TALLOC (INIT_BUF_SIZE / sizeof(US_CHAR_TYPE), + US_CHAR_TYPE); + } + + if (!COMPILED_BUFFER_VAR) FREE_STACK_RETURN (REG_ESPACE); +#ifdef MBS_SUPPORT + bufp->buffer = (char*)COMPILED_BUFFER_VAR; +#endif /* MBS_SUPPORT */ + bufp->allocated = INIT_BUF_SIZE; + } +#ifdef MBS_SUPPORT + else + COMPILED_BUFFER_VAR = (US_CHAR_TYPE*) bufp->buffer; +#endif + + begalt = b = COMPILED_BUFFER_VAR; + + /* Loop through the uncompiled pattern until we're at the end. */ + while (p != pend) + { + PATFETCH (c); + + switch (c) + { + case '^': + { + if ( /* If at start of pattern, it's an operator. */ + p == pattern + 1 + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's come before. */ + || at_begline_loc_p (pattern, p, syntax)) + BUF_PUSH (begline); + else + goto normal_char; + } + break; + + + case '$': + { + if ( /* If at end of pattern, it's an operator. */ + p == pend + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's next. */ + || at_endline_loc_p (p, pend, syntax)) + BUF_PUSH (endline); + else + goto normal_char; + } + break; + + + case '+': + case '?': + if ((syntax & RE_BK_PLUS_QM) + || (syntax & RE_LIMITED_OPS)) + goto normal_char; + handle_plus: + case '*': + /* If there is no previous pattern... */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (!(syntax & RE_CONTEXT_INDEP_OPS)) + goto normal_char; + } + + { + /* Are we optimizing this jump? */ + boolean keep_string_p = false; + + /* 1 means zero (many) matches is allowed. */ + char zero_times_ok = 0, many_times_ok = 0; + + /* If there is a sequence of repetition chars, collapse it + down to just one (the right one). We can't combine + interval operators with these because of, e.g., `a{2}*', + which should only match an even number of `a's. */ + + for (;;) + { + zero_times_ok |= c != '+'; + many_times_ok |= c != '?'; + + if (p == pend) + break; + + PATFETCH (c); + + if (c == '*' + || (!(syntax & RE_BK_PLUS_QM) && (c == '+' || c == '?'))) + ; + + else if (syntax & RE_BK_PLUS_QM && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + if (!(c1 == '+' || c1 == '?')) + { + PATUNFETCH; + PATUNFETCH; + break; + } + + c = c1; + } + else + { + PATUNFETCH; + break; + } + + /* If we get here, we found another repeat character. */ + } + + /* Star, etc. applied to an empty pattern is equivalent + to an empty pattern. */ + if (!laststart) + break; + + /* Now we know whether or not zero matches is allowed + and also whether or not two or more matches is allowed. */ + if (many_times_ok) + { /* More than one repetition is allowed, so put in at the + end a backward relative jump from `b' to before the next + jump we're going to put in below (which jumps from + laststart to after this jump). + + But if we are at the `*' in the exact sequence `.*\n', + insert an unconditional jump backwards to the ., + instead of the beginning of the loop. This way we only + push a failure point once, instead of every time + through the loop. */ + assert (p - 1 > pattern); + + /* Allocate the space for the jump. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + + /* We know we are not at the first character of the pattern, + because laststart was nonzero. And we've already + incremented `p', by the way, to be the character after + the `*'. Do we have to do something analogous here + for null bytes, because of RE_DOT_NOT_NULL? */ + if (TRANSLATE (*(p - 2)) == TRANSLATE ('.') + && zero_times_ok + && p < pend && TRANSLATE (*p) == TRANSLATE ('\n') + && !(syntax & RE_DOT_NEWLINE)) + { /* We have .*\n. */ + STORE_JUMP (jump, b, laststart); + keep_string_p = true; + } + else + /* Anything else. */ + STORE_JUMP (maybe_pop_jump, b, laststart - + (1 + OFFSET_ADDRESS_SIZE)); + + /* We've added more stuff to the buffer. */ + b += 1 + OFFSET_ADDRESS_SIZE; + } + + /* On failure, jump from laststart to b + 3, which will be the + end of the buffer after this jump is inserted. */ + /* ifdef MBS_SUPPORT, 'b + 1 + OFFSET_ADDRESS_SIZE' instead of + 'b + 3'. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (keep_string_p ? on_failure_keep_string_jump + : on_failure_jump, + laststart, b + 1 + OFFSET_ADDRESS_SIZE); + pending_exact = 0; + b += 1 + OFFSET_ADDRESS_SIZE; + + if (!zero_times_ok) + { + /* At least one repetition is required, so insert a + `dummy_failure_jump' before the initial + `on_failure_jump' instruction of the loop. This + effects a skip over that instruction the first time + we hit that loop. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (dummy_failure_jump, laststart, laststart + + 2 + 2 * OFFSET_ADDRESS_SIZE); + b += 1 + OFFSET_ADDRESS_SIZE; + } + } + break; + + + case '.': + laststart = b; + BUF_PUSH (anychar); + break; + + + case '[': + { + boolean had_char_class = false; +#ifdef MBS_SUPPORT + CHAR_TYPE range_start = 0xffffffff; +#else + unsigned int range_start = 0xffffffff; +#endif + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + +#ifdef MBS_SUPPORT + /* We assume a charset(_not) structure as a wchar_t array. + charset[0] = (re_opcode_t) charset(_not) + charset[1] = l (= length of char_classes) + charset[2] = m (= length of collating_symbols) + charset[3] = n (= length of equivalence_classes) + charset[4] = o (= length of char_ranges) + charset[5] = p (= length of chars) + + charset[6] = char_class (wctype_t) + charset[6+CHAR_CLASS_SIZE] = char_class (wctype_t) + ... + charset[l+5] = char_class (wctype_t) + + charset[l+6] = collating_symbol (wchar_t) + ... + charset[l+m+5] = collating_symbol (wchar_t) + ifdef _LIBC we use the index if + _NL_COLLATE_SYMB_EXTRAMB instead of + wchar_t string. + + charset[l+m+6] = equivalence_classes (wchar_t) + ... + charset[l+m+n+5] = equivalence_classes (wchar_t) + ifdef _LIBC we use the index in + _NL_COLLATE_WEIGHT instead of + wchar_t string. + + charset[l+m+n+6] = range_start + charset[l+m+n+7] = range_end + ... + charset[l+m+n+2o+4] = range_start + charset[l+m+n+2o+5] = range_end + ifdef _LIBC we use the value looked up + in _NL_COLLATE_COLLSEQ instead of + wchar_t character. + + charset[l+m+n+2o+6] = char + ... + charset[l+m+n+2o+p+5] = char + + */ + + /* We need at least 6 spaces: the opcode, the length of + char_classes, the length of collating_symbols, the length of + equivalence_classes, the length of char_ranges, the length of + chars. */ + GET_BUFFER_SPACE (6); + + /* Save b as laststart. And We use laststart as the pointer + to the first element of the charset here. + In other words, laststart[i] indicates charset[i]. */ + laststart = b; + + /* We test `*p == '^' twice, instead of using an if + statement, so we only need one BUF_PUSH. */ + BUF_PUSH (*p == '^' ? charset_not : charset); + if (*p == '^') + p++; + + /* Push the length of char_classes, the length of + collating_symbols, the length of equivalence_classes, the + length of char_ranges and the length of chars. */ + BUF_PUSH_3 (0, 0, 0); + BUF_PUSH_2 (0, 0); + + /* Remember the first position in the bracket expression. */ + p1 = p; + + /* charset_not matches newline according to a syntax bit. */ + if ((re_opcode_t) b[-6] == charset_not + && (syntax & RE_HAT_LISTS_NOT_NEWLINE)) + { + BUF_PUSH('\n'); + laststart[5]++; /* Update the length of characters */ + } + + /* Read in characters and ranges, setting map bits. */ + for (;;) + { + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + PATFETCH (c); + + /* \ might escape characters inside [...] and [^...]. */ + if ((syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + BUF_PUSH(c1); + laststart[5]++; /* Update the length of chars */ + range_start = c1; + continue; + } + + /* Could be the end of the bracket expression. If it's + not (i.e., when the bracket expression is `[]' so + far), the ']' character bit gets set way below. */ + if (c == ']' && p != p1 + 1) + break; + + /* Look ahead to see if it's a range when the last thing + was a character class. */ + if (had_char_class && c == '-' && *p != ']') + FREE_STACK_RETURN (REG_ERANGE); + + /* Look ahead to see if it's a range when the last thing + was a character: if this is a hyphen not at the + beginning or the end of a list, then it's the range + operator. */ + if (c == '-' + && !(p - 2 >= pattern && p[-2] == '[') + && !(p - 3 >= pattern && p[-3] == '[' && p[-2] == '^') + && *p != ']') + { + reg_errcode_t ret; + /* Allocate the space for range_start and range_end. */ + GET_BUFFER_SPACE (2); + /* Update the pointer to indicate end of buffer. */ + b += 2; + ret = compile_range (range_start, &p, pend, translate, + syntax, b, laststart); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + else if (p[0] == '-' && p[1] != ']') + { /* This handles ranges made up of characters only. */ + reg_errcode_t ret; + + /* Move past the `-'. */ + PATFETCH (c1); + /* Allocate the space for range_start and range_end. */ + GET_BUFFER_SPACE (2); + /* Update the pointer to indicate end of buffer. */ + b += 2; + ret = compile_range (c, &p, pend, translate, syntax, b, + laststart); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + + /* See if we're at the beginning of a possible character + class. */ + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == ':') + { /* Leave room for the null. */ + char str[CHAR_CLASS_MAX_LENGTH + 1]; + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[:'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == ':' && *p == ']') || p == pend) + break; + if (c1 < CHAR_CLASS_MAX_LENGTH) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + /* If isn't a word bracketed by `[:' and `:]': + undo the ending character, the letters, and leave + the leading `:' and `[' (but store them as character). */ + if (c == ':' && *p == ']') + { + wctype_t wt; + uintptr_t alignedp; + + /* Query the character class as wctype_t. */ + wt = IS_CHAR_CLASS (str); + if (wt == 0) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + /* Allocate the space for character class. */ + GET_BUFFER_SPACE(CHAR_CLASS_SIZE); + /* Update the pointer to indicate end of buffer. */ + b += CHAR_CLASS_SIZE; + /* Move data which follow character classes + not to violate the data. */ + insert_space(CHAR_CLASS_SIZE, + laststart + 6 + laststart[1], + b - 1); + alignedp = ((uintptr_t)(laststart + 6 + laststart[1]) + + __alignof__(wctype_t) - 1) + & ~(uintptr_t)(__alignof__(wctype_t) - 1); + /* Store the character class. */ + *((wctype_t*)alignedp) = wt; + /* Update length of char_classes */ + laststart[1] += CHAR_CLASS_SIZE; + + had_char_class = true; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + BUF_PUSH ('['); + BUF_PUSH (':'); + laststart[5] += 2; /* Update the length of characters */ + range_start = ':'; + had_char_class = false; + } + } + else if (syntax & RE_CHAR_CLASSES && c == '[' && (*p == '=' + || *p == '.')) + { + CHAR_TYPE str[128]; /* Should be large enough. */ + CHAR_TYPE delim = *p; /* '=' or '.' */ +# ifdef _LIBC + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[=' or '[[.'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == delim && *p == ']') || p == pend) + break; + if (c1 < sizeof (str) - 1) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + if (c == delim && *p == ']' && str[0] != '\0') + { + unsigned int i, offset; + /* If we have no collation data we use the default + collation in which each character is in a class + by itself. It also means that ASCII is the + character set and therefore we cannot have character + with more than one byte in the multibyte + representation. */ + + /* If not defined _LIBC, we push the name and + `\0' for the sake of matching performance. */ + int datasize = c1 + 1; + +# ifdef _LIBC + int32_t idx = 0; + if (nrules == 0) +# endif + { + if (c1 != 1) + FREE_STACK_RETURN (REG_ECOLLATE); + } +# ifdef _LIBC + else + { + const int32_t *table; + const int32_t *weights; + const int32_t *extra; + const int32_t *indirect; + wint_t *cp; + + /* This #include defines a local function! */ +# include <locale/weightwc.h> + + if(delim == '=') + { + /* We push the index for equivalence class. */ + cp = (wint_t*)str; + + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_TABLEWC); + weights = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_WEIGHTWC); + extra = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_INDIRECTWC); + + idx = findidx ((const wint_t**)&cp); + if (idx == 0 || cp < (wint_t*) str + c1) + /* This is no valid character. */ + FREE_STACK_RETURN (REG_ECOLLATE); + + str[0] = (wchar_t)idx; + } + else /* delim == '.' */ + { + /* We push collation sequence value + for collating symbol. */ + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + char char_str[c1]; + + /* We have to convert the name to a single-byte + string. This is possible since the names + consist of ASCII characters and the internal + representation is UCS4. */ + for (i = 0; i < c1; ++i) + char_str[i] = str[i]; + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing table. */ + hash = elem_hash (char_str, c1); + + idx = 0; + elem = hash % table_size; + second = hash % (table_size - 2); + while (symb_table[2 * elem] != 0) + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && c1 == extra[symb_table[2 * elem + 1]] + && memcmp (str, + &extra[symb_table[2 * elem + 1] + + 1], c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + + if (symb_table[2 * elem] != 0) + { + /* Compute the index of the byte sequence + in the table. */ + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~4; + + str[0] = (wchar_t) idx + 4; + } + else if (symb_table[2 * elem] == 0 && c1 == 1) + { + /* No valid character. Match it as a + single byte character. */ + had_char_class = false; + BUF_PUSH(str[0]); + /* Update the length of characters */ + laststart[5]++; + range_start = str[0]; + + /* Throw away the ] at the end of the + collating symbol. */ + PATFETCH (c); + /* exit from the switch block. */ + continue; + } + else + FREE_STACK_RETURN (REG_ECOLLATE); + } + datasize = 1; + } +# endif + /* Throw away the ] at the end of the equivalence + class (or collating symbol). */ + PATFETCH (c); + + /* Allocate the space for the equivalence class + (or collating symbol) (and '\0' if needed). */ + GET_BUFFER_SPACE(datasize); + /* Update the pointer to indicate end of buffer. */ + b += datasize; + + if (delim == '=') + { /* equivalence class */ + /* Calculate the offset of char_ranges, + which is next to equivalence_classes. */ + offset = laststart[1] + laststart[2] + + laststart[3] +6; + /* Insert space. */ + insert_space(datasize, laststart + offset, b - 1); + + /* Write the equivalence_class and \0. */ + for (i = 0 ; i < datasize ; i++) + laststart[offset + i] = str[i]; + + /* Update the length of equivalence_classes. */ + laststart[3] += datasize; + had_char_class = true; + } + else /* delim == '.' */ + { /* collating symbol */ + /* Calculate the offset of the equivalence_classes, + which is next to collating_symbols. */ + offset = laststart[1] + laststart[2] + 6; + /* Insert space and write the collationg_symbol + and \0. */ + insert_space(datasize, laststart + offset, b-1); + for (i = 0 ; i < datasize ; i++) + laststart[offset + i] = str[i]; + + /* In re_match_2_internal if range_start < -1, we + assume -range_start is the offset of the + collating symbol which is specified as + the character of the range start. So we assign + -(laststart[1] + laststart[2] + 6) to + range_start. */ + range_start = -(laststart[1] + laststart[2] + 6); + /* Update the length of collating_symbol. */ + laststart[2] += datasize; + had_char_class = false; + } + } + else + { + c1++; + while (c1--) + PATUNFETCH; + BUF_PUSH ('['); + BUF_PUSH (delim); + laststart[5] += 2; /* Update the length of characters */ + range_start = delim; + had_char_class = false; + } + } + else + { + had_char_class = false; + BUF_PUSH(c); + laststart[5]++; /* Update the length of characters */ + range_start = c; + } + } + +#else /* not MBS_SUPPORT */ + /* Ensure that we have enough space to push a charset: the + opcode, the length count, and the bitset; 34 bytes in all. */ + GET_BUFFER_SPACE (34); + + laststart = b; + + /* We test `*p == '^' twice, instead of using an if + statement, so we only need one BUF_PUSH. */ + BUF_PUSH (*p == '^' ? charset_not : charset); + if (*p == '^') + p++; + + /* Remember the first position in the bracket expression. */ + p1 = p; + + /* Push the number of bytes in the bitmap. */ + BUF_PUSH ((1 << BYTEWIDTH) / BYTEWIDTH); + + /* Clear the whole map. */ + bzero (b, (1 << BYTEWIDTH) / BYTEWIDTH); + + /* charset_not matches newline according to a syntax bit. */ + if ((re_opcode_t) b[-2] == charset_not + && (syntax & RE_HAT_LISTS_NOT_NEWLINE)) + SET_LIST_BIT ('\n'); + + /* Read in characters and ranges, setting map bits. */ + for (;;) + { + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + PATFETCH (c); + + /* \ might escape characters inside [...] and [^...]. */ + if ((syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + SET_LIST_BIT (c1); + range_start = c1; + continue; + } + + /* Could be the end of the bracket expression. If it's + not (i.e., when the bracket expression is `[]' so + far), the ']' character bit gets set way below. */ + if (c == ']' && p != p1 + 1) + break; + + /* Look ahead to see if it's a range when the last thing + was a character class. */ + if (had_char_class && c == '-' && *p != ']') + FREE_STACK_RETURN (REG_ERANGE); + + /* Look ahead to see if it's a range when the last thing + was a character: if this is a hyphen not at the + beginning or the end of a list, then it's the range + operator. */ + if (c == '-' + && !(p - 2 >= pattern && p[-2] == '[') + && !(p - 3 >= pattern && p[-3] == '[' && p[-2] == '^') + && *p != ']') + { + reg_errcode_t ret + = compile_range (range_start, &p, pend, translate, + syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + + else if (p[0] == '-' && p[1] != ']') + { /* This handles ranges made up of characters only. */ + reg_errcode_t ret; + + /* Move past the `-'. */ + PATFETCH (c1); + + ret = compile_range (c, &p, pend, translate, syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + + /* See if we're at the beginning of a possible character + class. */ + + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == ':') + { /* Leave room for the null. */ + char str[CHAR_CLASS_MAX_LENGTH + 1]; + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[:'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == ':' && *p == ']') || p == pend) + break; + if (c1 < CHAR_CLASS_MAX_LENGTH) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + /* If isn't a word bracketed by `[:' and `:]': + undo the ending character, the letters, and leave + the leading `:' and `[' (but set bits for them). */ + if (c == ':' && *p == ']') + { +# if defined _LIBC || WIDE_CHAR_SUPPORT + boolean is_lower = STREQ (str, "lower"); + boolean is_upper = STREQ (str, "upper"); + wctype_t wt; + int ch; + + wt = IS_CHAR_CLASS (str); + if (wt == 0) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ++ch) + { +# ifdef _LIBC + if (__iswctype (__btowc (ch), wt)) + SET_LIST_BIT (ch); +# else + if (iswctype (btowc (ch), wt)) + SET_LIST_BIT (ch); +# endif + + if (translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + + had_char_class = true; +# else + int ch; + boolean is_alnum = STREQ (str, "alnum"); + boolean is_alpha = STREQ (str, "alpha"); + boolean is_blank = STREQ (str, "blank"); + boolean is_cntrl = STREQ (str, "cntrl"); + boolean is_digit = STREQ (str, "digit"); + boolean is_graph = STREQ (str, "graph"); + boolean is_lower = STREQ (str, "lower"); + boolean is_print = STREQ (str, "print"); + boolean is_punct = STREQ (str, "punct"); + boolean is_space = STREQ (str, "space"); + boolean is_upper = STREQ (str, "upper"); + boolean is_xdigit = STREQ (str, "xdigit"); + + if (!IS_CHAR_CLASS (str)) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ch++) + { + /* This was split into 3 if's to + avoid an arbitrary limit in some compiler. */ + if ( (is_alnum && ISALNUM (ch)) + || (is_alpha && ISALPHA (ch)) + || (is_blank && ISBLANK (ch)) + || (is_cntrl && ISCNTRL (ch))) + SET_LIST_BIT (ch); + if ( (is_digit && ISDIGIT (ch)) + || (is_graph && ISGRAPH (ch)) + || (is_lower && ISLOWER (ch)) + || (is_print && ISPRINT (ch))) + SET_LIST_BIT (ch); + if ( (is_punct && ISPUNCT (ch)) + || (is_space && ISSPACE (ch)) + || (is_upper && ISUPPER (ch)) + || (is_xdigit && ISXDIGIT (ch))) + SET_LIST_BIT (ch); + if ( translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + had_char_class = true; +# endif /* libc || wctype.h */ + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT (':'); + range_start = ':'; + had_char_class = false; + } + } + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == '=') + { + unsigned char str[MB_LEN_MAX + 1]; +# ifdef _LIBC + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[='. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == '=' && *p == ']') || p == pend) + break; + if (c1 < MB_LEN_MAX) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + if (c == '=' && *p == ']' && str[0] != '\0') + { + /* If we have no collation data we use the default + collation in which each character is in a class + by itself. It also means that ASCII is the + character set and therefore we cannot have character + with more than one byte in the multibyte + representation. */ +# ifdef _LIBC + if (nrules == 0) +# endif + { + if (c1 != 1) + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Set the bit for the character. */ + SET_LIST_BIT (str[0]); + } +# ifdef _LIBC + else + { + /* Try to match the byte sequence in `str' against + those known to the collate implementation. + First find out whether the bytes in `str' are + actually from exactly one character. */ + const int32_t *table; + const unsigned char *weights; + const unsigned char *extra; + const int32_t *indirect; + int32_t idx; + const unsigned char *cp = str; + int ch; + + /* This #include defines a local function! */ +# include <locale/weight.h> + + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); + + idx = findidx (&cp); + if (idx == 0 || cp < str + c1) + /* This is no valid character. */ + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Now we have to go throught the whole table + and find all characters which have the same + first level weight. + + XXX Note that this is not entirely correct. + we would have to match multibyte sequences + but this is not possible with the current + implementation. */ + for (ch = 1; ch < 256; ++ch) + /* XXX This test would have to be changed if we + would allow matching multibyte sequences. */ + if (table[ch] > 0) + { + int32_t idx2 = table[ch]; + size_t len = weights[idx2]; + + /* Test whether the lenghts match. */ + if (weights[idx] == len) + { + /* They do. New compare the bytes of + the weight. */ + size_t cnt = 0; + + while (cnt < len + && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + /* They match. Mark the character as + acceptable. */ + SET_LIST_BIT (ch); + } + } + } +# endif + had_char_class = true; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT ('='); + range_start = '='; + had_char_class = false; + } + } + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == '.') + { + unsigned char str[128]; /* Should be large enough. */ +# ifdef _LIBC + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[.'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == '.' && *p == ']') || p == pend) + break; + if (c1 < sizeof (str)) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + if (c == '.' && *p == ']' && str[0] != '\0') + { + /* If we have no collation data we use the default + collation in which each character is the name + for its own class which contains only the one + character. It also means that ASCII is the + character set and therefore we cannot have character + with more than one byte in the multibyte + representation. */ +# ifdef _LIBC + if (nrules == 0) +# endif + { + if (c1 != 1) + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Set the bit for the character. */ + SET_LIST_BIT (str[0]); + range_start = ((const unsigned char *) str)[0]; + } +# ifdef _LIBC + else + { + /* Try to match the byte sequence in `str' against + those known to the collate implementation. + First find out whether the bytes in `str' are + actually from exactly one character. */ + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing table. */ + hash = elem_hash (str, c1); + + idx = 0; + elem = hash % table_size; + second = hash % (table_size - 2); + while (symb_table[2 * elem] != 0) + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && c1 == extra[symb_table[2 * elem + 1]] + && memcmp (str, + &extra[symb_table[2 * elem + 1] + + 1], + c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + + if (symb_table[2 * elem] == 0) + /* This is no valid character. */ + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Now add the multibyte character(s) we found + to the accept list. + + XXX Note that this is not entirely correct. + we would have to match multibyte sequences + but this is not possible with the current + implementation. Also, we have to match + collating symbols, which expand to more than + one file, as a whole and not allow the + individual bytes. */ + c1 = extra[idx++]; + if (c1 == 1) + range_start = extra[idx]; + while (c1-- > 0) + { + SET_LIST_BIT (extra[idx]); + ++idx; + } + } +# endif + had_char_class = false; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT ('.'); + range_start = '.'; + had_char_class = false; + } + } + else + { + had_char_class = false; + SET_LIST_BIT (c); + range_start = c; + } + } + + /* Discard any (non)matching list bytes that are all 0 at the + end of the map. Decrease the map-length byte too. */ + while ((int) b[-1] > 0 && b[b[-1] - 1] == 0) + b[-1]--; + b += b[-1]; +#endif /* MBS_SUPPORT */ + } + break; + + + case '(': + if (syntax & RE_NO_BK_PARENS) + goto handle_open; + else + goto normal_char; + + + case ')': + if (syntax & RE_NO_BK_PARENS) + goto handle_close; + else + goto normal_char; + + + case '\n': + if (syntax & RE_NEWLINE_ALT) + goto handle_alt; + else + goto normal_char; + + + case '|': + if (syntax & RE_NO_BK_VBAR) + goto handle_alt; + else + goto normal_char; + + + case '{': + if (syntax & RE_INTERVALS && syntax & RE_NO_BK_BRACES) + goto handle_interval; + else + goto normal_char; + + + case '\\': + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + /* Do not translate the character after the \, so that we can + distinguish, e.g., \B from \b, even if we normally would + translate, e.g., B to b. */ + PATFETCH_RAW (c); + + switch (c) + { + case '(': + if (syntax & RE_NO_BK_PARENS) + goto normal_backslash; + + handle_open: + bufp->re_nsub++; + regnum++; + + if (COMPILE_STACK_FULL) + { + RETALLOC (compile_stack.stack, compile_stack.size << 1, + compile_stack_elt_t); + if (compile_stack.stack == NULL) return REG_ESPACE; + + compile_stack.size <<= 1; + } + + /* These are the values to restore when we hit end of this + group. They are all relative offsets, so that if the + whole pattern moves because of realloc, they will still + be valid. */ + COMPILE_STACK_TOP.begalt_offset = begalt - COMPILED_BUFFER_VAR; + COMPILE_STACK_TOP.fixup_alt_jump + = fixup_alt_jump ? fixup_alt_jump - COMPILED_BUFFER_VAR + 1 : 0; + COMPILE_STACK_TOP.laststart_offset = b - COMPILED_BUFFER_VAR; + COMPILE_STACK_TOP.regnum = regnum; + + /* We will eventually replace the 0 with the number of + groups inner to this one. But do not push a + start_memory for groups beyond the last one we can + represent in the compiled pattern. */ + if (regnum <= MAX_REGNUM) + { + COMPILE_STACK_TOP.inner_group_offset = b + - COMPILED_BUFFER_VAR + 2; + BUF_PUSH_3 (start_memory, regnum, 0); + } + + compile_stack.avail++; + + fixup_alt_jump = 0; + laststart = 0; + begalt = b; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + break; + + + case ')': + if (syntax & RE_NO_BK_PARENS) goto normal_backslash; + + if (COMPILE_STACK_EMPTY) + { + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_backslash; + else + FREE_STACK_RETURN (REG_ERPAREN); + } + + handle_close: + if (fixup_alt_jump) + { /* Push a dummy failure point at the end of the + alternative for a possible future + `pop_failure_jump' to pop. See comments at + `push_dummy_failure' in `re_match_2'. */ + BUF_PUSH (push_dummy_failure); + + /* We allocated space for this jump when we assigned + to `fixup_alt_jump', in the `handle_alt' case below. */ + STORE_JUMP (jump_past_alt, fixup_alt_jump, b - 1); + } + + /* See similar code for backslashed left paren above. */ + if (COMPILE_STACK_EMPTY) + { + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_char; + else + FREE_STACK_RETURN (REG_ERPAREN); + } + + /* Since we just checked for an empty stack above, this + ``can't happen''. */ + assert (compile_stack.avail != 0); + { + /* We don't just want to restore into `regnum', because + later groups should continue to be numbered higher, + as in `(ab)c(de)' -- the second group is #2. */ + regnum_t this_group_regnum; + + compile_stack.avail--; + begalt = COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.begalt_offset; + fixup_alt_jump + = COMPILE_STACK_TOP.fixup_alt_jump + ? COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.fixup_alt_jump - 1 + : 0; + laststart = COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.laststart_offset; + this_group_regnum = COMPILE_STACK_TOP.regnum; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + + /* We're at the end of the group, so now we know how many + groups were inside this one. */ + if (this_group_regnum <= MAX_REGNUM) + { + US_CHAR_TYPE *inner_group_loc + = COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.inner_group_offset; + + *inner_group_loc = regnum - this_group_regnum; + BUF_PUSH_3 (stop_memory, this_group_regnum, + regnum - this_group_regnum); + } + } + break; + + + case '|': /* `\|'. */ + if (syntax & RE_LIMITED_OPS || syntax & RE_NO_BK_VBAR) + goto normal_backslash; + handle_alt: + if (syntax & RE_LIMITED_OPS) + goto normal_char; + + /* Insert before the previous alternative a jump which + jumps to this alternative if the former fails. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (on_failure_jump, begalt, + b + 2 + 2 * OFFSET_ADDRESS_SIZE); + pending_exact = 0; + b += 1 + OFFSET_ADDRESS_SIZE; + + /* The alternative before this one has a jump after it + which gets executed if it gets matched. Adjust that + jump so it will jump to this alternative's analogous + jump (put in below, which in turn will jump to the next + (if any) alternative's such jump, etc.). The last such + jump jumps to the correct final destination. A picture: + _____ _____ + | | | | + | v | v + a | b | c + + If we are at `b', then fixup_alt_jump right now points to a + three-byte space after `a'. We'll put in the jump, set + fixup_alt_jump to right after `b', and leave behind three + bytes which we'll fill in when we get to after `c'. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + /* Mark and leave space for a jump after this alternative, + to be filled in later either by next alternative or + when know we're at the end of a series of alternatives. */ + fixup_alt_jump = b; + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + b += 1 + OFFSET_ADDRESS_SIZE; + + laststart = 0; + begalt = b; + break; + + + case '{': + /* If \{ is a literal. */ + if (!(syntax & RE_INTERVALS) + /* If we're at `\{' and it's not the open-interval + operator. */ + || (syntax & RE_NO_BK_BRACES)) + goto normal_backslash; + + handle_interval: + { + /* If got here, then the syntax allows intervals. */ + + /* At least (most) this many matches must be made. */ + int lower_bound = -1, upper_bound = -1; + beg_interval = p - 1; + + if (p == pend) + { + if (!(syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_EBRACE); + } + + GET_UNSIGNED_NUMBER (lower_bound); + + if (c == ',') + { + GET_UNSIGNED_NUMBER (upper_bound); + if ((!(syntax & RE_NO_BK_BRACES) && c != '\\') + || ((syntax & RE_NO_BK_BRACES) && c != '}')) + FREE_STACK_RETURN (REG_BADBR); + + if (upper_bound < 0) + upper_bound = RE_DUP_MAX; + } + else + /* Interval such as `{1}' => match exactly once. */ + upper_bound = lower_bound; + + if (lower_bound < 0 || upper_bound > RE_DUP_MAX + || lower_bound > upper_bound) + { + if (!(syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_BADBR); + } + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (c != '\\') FREE_STACK_RETURN (REG_EBRACE); + + PATFETCH (c); + } + + if (c != '}') + { + if (!(syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_BADBR); + } + + /* We just parsed a valid interval. */ + + /* If it's invalid to have no preceding re. */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (syntax & RE_CONTEXT_INDEP_OPS) + laststart = b; + else + goto unfetch_interval; + } + + /* If the upper bound is zero, don't want to succeed at + all; jump from `laststart' to `b + 3', which will be + the end of the buffer after we insert the jump. */ + /* ifdef MBS_SUPPORT, 'b + 1 + OFFSET_ADDRESS_SIZE' + instead of 'b + 3'. */ + if (upper_bound == 0) + { + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (jump, laststart, b + 1 + + OFFSET_ADDRESS_SIZE); + b += 1 + OFFSET_ADDRESS_SIZE; + } + + /* Otherwise, we have a nontrivial interval. When + we're all done, the pattern will look like: + set_number_at <jump count> <upper bound> + set_number_at <succeed_n count> <lower bound> + succeed_n <after jump addr> <succeed_n count> + <body of loop> + jump_n <succeed_n addr> <jump count> + (The upper bound and `jump_n' are omitted if + `upper_bound' is 1, though.) */ + else + { /* If the upper bound is > 1, we need to insert + more at the end of the loop. */ + unsigned nbytes = 2 + 4 * OFFSET_ADDRESS_SIZE + + (upper_bound > 1) * (2 + 4 * OFFSET_ADDRESS_SIZE); + + GET_BUFFER_SPACE (nbytes); + + /* Initialize lower bound of the `succeed_n', even + though it will be set during matching by its + attendant `set_number_at' (inserted next), + because `re_compile_fastmap' needs to know. + Jump to the `jump_n' we might insert below. */ + INSERT_JUMP2 (succeed_n, laststart, + b + 1 + 2 * OFFSET_ADDRESS_SIZE + + (upper_bound > 1) * (1 + 2 * OFFSET_ADDRESS_SIZE) + , lower_bound); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + + /* Code to initialize the lower bound. Insert + before the `succeed_n'. The `5' is the last two + bytes of this `set_number_at', plus 3 bytes of + the following `succeed_n'. */ + /* ifdef MBS_SUPPORT, The '1+2*OFFSET_ADDRESS_SIZE' + is the 'set_number_at', plus '1+OFFSET_ADDRESS_SIZE' + of the following `succeed_n'. */ + insert_op2 (set_number_at, laststart, 1 + + 2 * OFFSET_ADDRESS_SIZE, lower_bound, b); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + + if (upper_bound > 1) + { /* More than one repetition is allowed, so + append a backward jump to the `succeed_n' + that starts this interval. + + When we've reached this during matching, + we'll have matched the interval once, so + jump back only `upper_bound - 1' times. */ + STORE_JUMP2 (jump_n, b, laststart + + 2 * OFFSET_ADDRESS_SIZE + 1, + upper_bound - 1); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + + /* The location we want to set is the second + parameter of the `jump_n'; that is `b-2' as + an absolute address. `laststart' will be + the `set_number_at' we're about to insert; + `laststart+3' the number to set, the source + for the relative address. But we are + inserting into the middle of the pattern -- + so everything is getting moved up by 5. + Conclusion: (b - 2) - (laststart + 3) + 5, + i.e., b - laststart. + + We insert this at the beginning of the loop + so that if we fail during matching, we'll + reinitialize the bounds. */ + insert_op2 (set_number_at, laststart, b - laststart, + upper_bound - 1, b); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + } + } + pending_exact = 0; + beg_interval = NULL; + } + break; + + unfetch_interval: + /* If an invalid interval, match the characters as literals. */ + assert (beg_interval); + p = beg_interval; + beg_interval = NULL; + + /* normal_char and normal_backslash need `c'. */ + PATFETCH (c); + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (p > pattern && p[-1] == '\\') + goto normal_backslash; + } + goto normal_char; + +#ifdef emacs + /* There is no way to specify the before_dot and after_dot + operators. rms says this is ok. --karl */ + case '=': + BUF_PUSH (at_dot); + break; + + case 's': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (syntaxspec, syntax_spec_code[c]); + break; + + case 'S': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (notsyntaxspec, syntax_spec_code[c]); + break; +#endif /* emacs */ + + + case 'w': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (wordchar); + break; + + + case 'W': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (notwordchar); + break; + + + case '<': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbeg); + break; + + case '>': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordend); + break; + + case 'b': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbound); + break; + + case 'B': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (notwordbound); + break; + + case '`': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (begbuf); + break; + + case '\'': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (endbuf); + break; + + case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': + if (syntax & RE_NO_BK_REFS) + goto normal_char; + + c1 = c - '0'; + + if (c1 > regnum) + FREE_STACK_RETURN (REG_ESUBREG); + + /* Can't back reference to a subexpression if inside of it. */ + if (group_in_compile_stack (compile_stack, (regnum_t) c1)) + goto normal_char; + + laststart = b; + BUF_PUSH_2 (duplicate, c1); + break; + + + case '+': + case '?': + if (syntax & RE_BK_PLUS_QM) + goto handle_plus; + else + goto normal_backslash; + + default: + normal_backslash: + /* You might think it would be useful for \ to mean + not to translate; but if we don't translate it + it will never match anything. */ + c = TRANSLATE (c); + goto normal_char; + } + break; + + + default: + /* Expects the character in `c'. */ + normal_char: + /* If no exactn currently being built. */ + if (!pending_exact +#ifdef MBS_SUPPORT + /* If last exactn handle binary(or character) and + new exactn handle character(or binary). */ + || is_exactn_bin != is_binary[p - 1 - pattern] +#endif /* MBS_SUPPORT */ + + /* If last exactn not at current position. */ + || pending_exact + *pending_exact + 1 != b + + /* We have only one byte following the exactn for the count. */ + || *pending_exact == (1 << BYTEWIDTH) - 1 + + /* If followed by a repetition operator. */ + || *p == '*' || *p == '^' + || ((syntax & RE_BK_PLUS_QM) + ? *p == '\\' && (p[1] == '+' || p[1] == '?') + : (*p == '+' || *p == '?')) + || ((syntax & RE_INTERVALS) + && ((syntax & RE_NO_BK_BRACES) + ? *p == '{' + : (p[0] == '\\' && p[1] == '{')))) + { + /* Start building a new exactn. */ + + laststart = b; + +#ifdef MBS_SUPPORT + /* Is this exactn binary data or character? */ + is_exactn_bin = is_binary[p - 1 - pattern]; + if (is_exactn_bin) + BUF_PUSH_2 (exactn_bin, 0); + else + BUF_PUSH_2 (exactn, 0); +#else + BUF_PUSH_2 (exactn, 0); +#endif /* MBS_SUPPORT */ + pending_exact = b - 1; + } + + BUF_PUSH (c); + (*pending_exact)++; + break; + } /* switch (c) */ + } /* while p != pend */ + + + /* Through the pattern now. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + if (!COMPILE_STACK_EMPTY) + FREE_STACK_RETURN (REG_EPAREN); + + /* If we don't want backtracking, force success + the first time we reach the end of the compiled pattern. */ + if (syntax & RE_NO_POSIX_BACKTRACKING) + BUF_PUSH (succeed); + +#ifdef MBS_SUPPORT + free (pattern); + free (mbs_offset); + free (is_binary); +#endif + free (compile_stack.stack); + + /* We have succeeded; set the length of the buffer. */ +#ifdef MBS_SUPPORT + bufp->used = (uintptr_t) b - (uintptr_t) COMPILED_BUFFER_VAR; +#else + bufp->used = b - bufp->buffer; +#endif + +#ifdef DEBUG + if (debug) + { + DEBUG_PRINT1 ("\nCompiled pattern: \n"); + print_compiled_pattern (bufp); + } +#endif /* DEBUG */ + +#ifndef MATCH_MAY_ALLOCATE + /* Initialize the failure stack to the largest possible stack. This + isn't necessary unless we're trying to avoid calling alloca in + the search and match routines. */ + { + int num_regs = bufp->re_nsub + 1; + + /* Since DOUBLE_FAIL_STACK refuses to double only if the current size + is strictly greater than re_max_failures, the largest possible stack + is 2 * re_max_failures failure points. */ + if (fail_stack.size < (2 * re_max_failures * MAX_FAILURE_ITEMS)) + { + fail_stack.size = (2 * re_max_failures * MAX_FAILURE_ITEMS); + +# ifdef emacs + if (! fail_stack.stack) + fail_stack.stack + = (fail_stack_elt_t *) xmalloc (fail_stack.size + * sizeof (fail_stack_elt_t)); + else + fail_stack.stack + = (fail_stack_elt_t *) xrealloc (fail_stack.stack, + (fail_stack.size + * sizeof (fail_stack_elt_t))); +# else /* not emacs */ + if (! fail_stack.stack) + fail_stack.stack + = (fail_stack_elt_t *) malloc (fail_stack.size + * sizeof (fail_stack_elt_t)); + else + fail_stack.stack + = (fail_stack_elt_t *) realloc (fail_stack.stack, + (fail_stack.size + * sizeof (fail_stack_elt_t))); +# endif /* not emacs */ + } + + regex_grow_registers (num_regs); + } +#endif /* not MATCH_MAY_ALLOCATE */ + + return REG_NOERROR; +} /* regex_compile */ + +/* Subroutines for `regex_compile'. */ + +/* Store OP at LOC followed by two-byte integer parameter ARG. */ +/* ifdef MBS_SUPPORT, integer parameter is 1 wchar_t. */ + +static void +store_op1 (op, loc, arg) + re_opcode_t op; + US_CHAR_TYPE *loc; + int arg; +{ + *loc = (US_CHAR_TYPE) op; + STORE_NUMBER (loc + 1, arg); +} + + +/* Like `store_op1', but for two two-byte parameters ARG1 and ARG2. */ +/* ifdef MBS_SUPPORT, integer parameter is 1 wchar_t. */ + +static void +store_op2 (op, loc, arg1, arg2) + re_opcode_t op; + US_CHAR_TYPE *loc; + int arg1, arg2; +{ + *loc = (US_CHAR_TYPE) op; + STORE_NUMBER (loc + 1, arg1); + STORE_NUMBER (loc + 1 + OFFSET_ADDRESS_SIZE, arg2); +} + + +/* Copy the bytes from LOC to END to open up three bytes of space at LOC + for OP followed by two-byte integer parameter ARG. */ +/* ifdef MBS_SUPPORT, integer parameter is 1 wchar_t. */ + +static void +insert_op1 (op, loc, arg, end) + re_opcode_t op; + US_CHAR_TYPE *loc; + int arg; + US_CHAR_TYPE *end; +{ + register US_CHAR_TYPE *pfrom = end; + register US_CHAR_TYPE *pto = end + 1 + OFFSET_ADDRESS_SIZE; + + while (pfrom != loc) + *--pto = *--pfrom; + + store_op1 (op, loc, arg); +} + + +/* Like `insert_op1', but for two two-byte parameters ARG1 and ARG2. */ +/* ifdef MBS_SUPPORT, integer parameter is 1 wchar_t. */ + +static void +insert_op2 (op, loc, arg1, arg2, end) + re_opcode_t op; + US_CHAR_TYPE *loc; + int arg1, arg2; + US_CHAR_TYPE *end; +{ + register US_CHAR_TYPE *pfrom = end; + register US_CHAR_TYPE *pto = end + 1 + 2 * OFFSET_ADDRESS_SIZE; + + while (pfrom != loc) + *--pto = *--pfrom; + + store_op2 (op, loc, arg1, arg2); +} + + +/* P points to just after a ^ in PATTERN. Return true if that ^ comes + after an alternative or a begin-subexpression. We assume there is at + least one character before the ^. */ + +static boolean +at_begline_loc_p (pattern, p, syntax) + const CHAR_TYPE *pattern, *p; + reg_syntax_t syntax; +{ + const CHAR_TYPE *prev = p - 2; + boolean prev_prev_backslash = prev > pattern && prev[-1] == '\\'; + + return + /* After a subexpression? */ + (*prev == '(' && (syntax & RE_NO_BK_PARENS || prev_prev_backslash)) + /* After an alternative? */ + || (*prev == '|' && (syntax & RE_NO_BK_VBAR || prev_prev_backslash)); +} + + +/* The dual of at_begline_loc_p. This one is for $. We assume there is + at least one character after the $, i.e., `P < PEND'. */ + +static boolean +at_endline_loc_p (p, pend, syntax) + const CHAR_TYPE *p, *pend; + reg_syntax_t syntax; +{ + const CHAR_TYPE *next = p; + boolean next_backslash = *next == '\\'; + const CHAR_TYPE *next_next = p + 1 < pend ? p + 1 : 0; + + return + /* Before a subexpression? */ + (syntax & RE_NO_BK_PARENS ? *next == ')' + : next_backslash && next_next && *next_next == ')') + /* Before an alternative? */ + || (syntax & RE_NO_BK_VBAR ? *next == '|' + : next_backslash && next_next && *next_next == '|'); +} + + +/* Returns true if REGNUM is in one of COMPILE_STACK's elements and + false if it's not. */ + +static boolean +group_in_compile_stack (compile_stack, regnum) + compile_stack_type compile_stack; + regnum_t regnum; +{ + int this_element; + + for (this_element = compile_stack.avail - 1; + this_element >= 0; + this_element--) + if (compile_stack.stack[this_element].regnum == regnum) + return true; + + return false; +} + +#ifdef MBS_SUPPORT +/* This insert space, which size is "num", into the pattern at "loc". + "end" must point the end of the allocated buffer. */ +static void +insert_space (num, loc, end) + int num; + CHAR_TYPE *loc; + CHAR_TYPE *end; +{ + register CHAR_TYPE *pto = end; + register CHAR_TYPE *pfrom = end - num; + + while (pfrom >= loc) + *pto-- = *pfrom--; +} +#endif /* MBS_SUPPORT */ + +#ifdef MBS_SUPPORT +static reg_errcode_t +compile_range (range_start_char, p_ptr, pend, translate, syntax, b, + char_set) + CHAR_TYPE range_start_char; + const CHAR_TYPE **p_ptr, *pend; + CHAR_TYPE *char_set, *b; + RE_TRANSLATE_TYPE translate; + reg_syntax_t syntax; +{ + const CHAR_TYPE *p = *p_ptr; + CHAR_TYPE range_start, range_end; + reg_errcode_t ret; +# ifdef _LIBC + uint32_t nrules; + uint32_t start_val, end_val; +# endif + if (p == pend) + return REG_ERANGE; + +# ifdef _LIBC + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules != 0) + { + const char *collseq = (const char *) _NL_CURRENT(LC_COLLATE, + _NL_COLLATE_COLLSEQWC); + const unsigned char *extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + + if (range_start_char < -1) + { + /* range_start is a collating symbol. */ + int32_t *wextra; + /* Retreive the index and get collation sequence value. */ + wextra = (int32_t*)(extra + char_set[-range_start_char]); + start_val = wextra[1 + *wextra]; + } + else + start_val = collseq_table_lookup(collseq, TRANSLATE(range_start_char)); + + end_val = collseq_table_lookup (collseq, TRANSLATE (p[0])); + + /* Report an error if the range is empty and the syntax prohibits + this. */ + ret = ((syntax & RE_NO_EMPTY_RANGES) + && (start_val > end_val))? REG_ERANGE : REG_NOERROR; + + /* Insert space to the end of the char_ranges. */ + insert_space(2, b - char_set[5] - 2, b - 1); + *(b - char_set[5] - 2) = (wchar_t)start_val; + *(b - char_set[5] - 1) = (wchar_t)end_val; + char_set[4]++; /* ranges_index */ + } + else +# endif + { + range_start = (range_start_char >= 0)? TRANSLATE (range_start_char): + range_start_char; + range_end = TRANSLATE (p[0]); + /* Report an error if the range is empty and the syntax prohibits + this. */ + ret = ((syntax & RE_NO_EMPTY_RANGES) + && (range_start > range_end))? REG_ERANGE : REG_NOERROR; + + /* Insert space to the end of the char_ranges. */ + insert_space(2, b - char_set[5] - 2, b - 1); + *(b - char_set[5] - 2) = range_start; + *(b - char_set[5] - 1) = range_end; + char_set[4]++; /* ranges_index */ + } + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + return ret; +} +#else +/* Read the ending character of a range (in a bracket expression) from the + uncompiled pattern *P_PTR (which ends at PEND). We assume the + starting character is in `P[-2]'. (`P[-1]' is the character `-'.) + Then we set the translation of all bits between the starting and + ending characters (inclusive) in the compiled pattern B. + + Return an error code. + + We use these short variable names so we can use the same macros as + `regex_compile' itself. */ + +static reg_errcode_t +compile_range (range_start_char, p_ptr, pend, translate, syntax, b) + unsigned int range_start_char; + const char **p_ptr, *pend; + RE_TRANSLATE_TYPE translate; + reg_syntax_t syntax; + unsigned char *b; +{ + unsigned this_char; + const char *p = *p_ptr; + reg_errcode_t ret; +# if _LIBC + const unsigned char *collseq; + unsigned int start_colseq; + unsigned int end_colseq; +# else + unsigned end_char; +# endif + + if (p == pend) + return REG_ERANGE; + + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + /* Report an error if the range is empty and the syntax prohibits this. */ + ret = syntax & RE_NO_EMPTY_RANGES ? REG_ERANGE : REG_NOERROR; + +# if _LIBC + collseq = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_COLLSEQMB); + + start_colseq = collseq[(unsigned char) TRANSLATE (range_start_char)]; + end_colseq = collseq[(unsigned char) TRANSLATE (p[0])]; + for (this_char = 0; this_char <= (unsigned char) -1; ++this_char) + { + unsigned int this_colseq = collseq[(unsigned char) TRANSLATE (this_char)]; + + if (start_colseq <= this_colseq && this_colseq <= end_colseq) + { + SET_LIST_BIT (TRANSLATE (this_char)); + ret = REG_NOERROR; + } + } +# else + /* Here we see why `this_char' has to be larger than an `unsigned + char' -- we would otherwise go into an infinite loop, since all + characters <= 0xff. */ + range_start_char = TRANSLATE (range_start_char); + /* TRANSLATE(p[0]) is casted to char (not unsigned char) in TRANSLATE, + and some compilers cast it to int implicitly, so following for_loop + may fall to (almost) infinite loop. + e.g. If translate[p[0]] = 0xff, end_char may equals to 0xffffffff. + To avoid this, we cast p[0] to unsigned int and truncate it. */ + end_char = ((unsigned)TRANSLATE(p[0]) & ((1 << BYTEWIDTH) - 1)); + + for (this_char = range_start_char; this_char <= end_char; ++this_char) + { + SET_LIST_BIT (TRANSLATE (this_char)); + ret = REG_NOERROR; + } +# endif + + return ret; +} +#endif /* MBS_SUPPORT */ + +/* re_compile_fastmap computes a ``fastmap'' for the compiled pattern in + BUFP. A fastmap records which of the (1 << BYTEWIDTH) possible + characters can start a string that matches the pattern. This fastmap + is used by re_search to skip quickly over impossible starting points. + + The caller must supply the address of a (1 << BYTEWIDTH)-byte data + area as BUFP->fastmap. + + We set the `fastmap', `fastmap_accurate', and `can_be_null' fields in + the pattern buffer. + + Returns 0 if we succeed, -2 if an internal error. */ + +#ifdef MBS_SUPPORT +/* local function for re_compile_fastmap. + truncate wchar_t character to char. */ +static unsigned char truncate_wchar (CHAR_TYPE c); + +static unsigned char +truncate_wchar (c) + CHAR_TYPE c; +{ + unsigned char buf[MB_LEN_MAX]; + int retval = wctomb(buf, c); + return retval > 0 ? buf[0] : (unsigned char)c; +} +#endif /* MBS_SUPPORT */ + +int +re_compile_fastmap (bufp) + struct re_pattern_buffer *bufp; +{ + int j, k; +#ifdef MATCH_MAY_ALLOCATE + fail_stack_type fail_stack; +#endif +#ifndef REGEX_MALLOC + char *destination; +#endif + + register char *fastmap = bufp->fastmap; + +#ifdef MBS_SUPPORT + /* We need to cast pattern to (wchar_t*), because we casted this compiled + pattern to (char*) in regex_compile. */ + US_CHAR_TYPE *pattern = (US_CHAR_TYPE*)bufp->buffer; + register US_CHAR_TYPE *pend = (US_CHAR_TYPE*) (bufp->buffer + bufp->used); +#else + US_CHAR_TYPE *pattern = bufp->buffer; + register US_CHAR_TYPE *pend = pattern + bufp->used; +#endif /* MBS_SUPPORT */ + US_CHAR_TYPE *p = pattern; + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* Assume that each path through the pattern can be null until + proven otherwise. We set this false at the bottom of switch + statement, to which we get only if a particular path doesn't + match the empty string. */ + boolean path_can_be_null = true; + + /* We aren't doing a `succeed_n' to begin with. */ + boolean succeed_n_p = false; + + assert (fastmap != NULL && p != NULL); + + INIT_FAIL_STACK (); + bzero (fastmap, 1 << BYTEWIDTH); /* Assume nothing's valid. */ + bufp->fastmap_accurate = 1; /* It will be when we're done. */ + bufp->can_be_null = 0; + + while (1) + { + if (p == pend || *p == succeed) + { + /* We have reached the (effective) end of pattern. */ + if (!FAIL_STACK_EMPTY ()) + { + bufp->can_be_null |= path_can_be_null; + + /* Reset for next path. */ + path_can_be_null = true; + + p = fail_stack.stack[--fail_stack.avail].pointer; + + continue; + } + else + break; + } + + /* We should never be about to go beyond the end of the pattern. */ + assert (p < pend); + + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + + /* I guess the idea here is to simply not bother with a fastmap + if a backreference is used, since it's too hard to figure out + the fastmap for the corresponding group. Setting + `can_be_null' stops `re_search_2' from using the fastmap, so + that is all we do. */ + case duplicate: + bufp->can_be_null = 1; + goto done; + + + /* Following are the cases which match a character. These end + with `break'. */ + +#ifdef MBS_SUPPORT + case exactn: + fastmap[truncate_wchar(p[1])] = 1; + break; + case exactn_bin: + fastmap[p[1]] = 1; + break; +#else + case exactn: + fastmap[p[1]] = 1; + break; +#endif /* MBS_SUPPORT */ + + +#ifdef MBS_SUPPORT + /* It is hard to distinguish fastmap from (multi byte) characters + which depends on current locale. */ + case charset: + case charset_not: + case wordchar: + case notwordchar: + bufp->can_be_null = 1; + goto done; +#else + case charset: + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH))) + fastmap[j] = 1; + break; + + + case charset_not: + /* Chars beyond end of map must be allowed. */ + for (j = *p * BYTEWIDTH; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (!(p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH)))) + fastmap[j] = 1; + break; + + + case wordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == Sword) + fastmap[j] = 1; + break; + + + case notwordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != Sword) + fastmap[j] = 1; + break; +#endif + + case anychar: + { + int fastmap_newline = fastmap['\n']; + + /* `.' matches anything ... */ + for (j = 0; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + /* ... except perhaps newline. */ + if (!(bufp->syntax & RE_DOT_NEWLINE)) + fastmap['\n'] = fastmap_newline; + + /* Return if we have already set `can_be_null'; if we have, + then the fastmap is irrelevant. Something's wrong here. */ + else if (bufp->can_be_null) + goto done; + + /* Otherwise, have to check alternative paths. */ + break; + } + +#ifdef emacs + case syntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + case notsyntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + /* All cases after this match the empty string. These end with + `continue'. */ + + + case before_dot: + case at_dot: + case after_dot: + continue; +#endif /* emacs */ + + + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbound: + case notwordbound: + case wordbeg: + case wordend: + case push_dummy_failure: + continue; + + + case jump_n: + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case jump_past_alt: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + if (j > 0) + continue; + + /* Jump backward implies we just went through the body of a + loop and matched nothing. Opcode jumped to should be + `on_failure_jump' or `succeed_n'. Just treat it like an + ordinary jump. For a * loop, it has pushed its failure + point already; if so, discard that as redundant. */ + if ((re_opcode_t) *p != on_failure_jump + && (re_opcode_t) *p != succeed_n) + continue; + + p++; + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + + /* If what's on the stack is where we are now, pop it. */ + if (!FAIL_STACK_EMPTY () + && fail_stack.stack[fail_stack.avail - 1].pointer == p) + fail_stack.avail--; + + continue; + + + case on_failure_jump: + case on_failure_keep_string_jump: + handle_on_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + + /* For some patterns, e.g., `(a?)?', `p+j' here points to the + end of the pattern. We don't want to push such a point, + since when we restore it above, entering the switch will + increment `p' past the end of the pattern. We don't need + to push such a point since we obviously won't find any more + fastmap entries beyond `pend'. Such a pattern can match + the null string, though. */ + if (p + j < pend) + { + if (!PUSH_PATTERN_OP (p + j, fail_stack)) + { + RESET_FAIL_STACK (); + return -2; + } + } + else + bufp->can_be_null = 1; + + if (succeed_n_p) + { + EXTRACT_NUMBER_AND_INCR (k, p); /* Skip the n. */ + succeed_n_p = false; + } + + continue; + + + case succeed_n: + /* Get to the number of times to succeed. */ + p += OFFSET_ADDRESS_SIZE; + + /* Increment p past the n for when k != 0. */ + EXTRACT_NUMBER_AND_INCR (k, p); + if (k == 0) + { + p -= 2 * OFFSET_ADDRESS_SIZE; + succeed_n_p = true; /* Spaghetti code alert. */ + goto handle_on_failure_jump; + } + continue; + + + case set_number_at: + p += 2 * OFFSET_ADDRESS_SIZE; + continue; + + + case start_memory: + case stop_memory: + p += 2; + continue; + + + default: + abort (); /* We have listed all the cases. */ + } /* switch *p++ */ + + /* Getting here means we have found the possible starting + characters for one path of the pattern -- and that the empty + string does not match. We need not follow this path further. + Instead, look at the next alternative (remembered on the + stack), or quit if no more. The test at the top of the loop + does these things. */ + path_can_be_null = false; + p = pend; + } /* while p */ + + /* Set `can_be_null' for the last path (also the first path, if the + pattern is empty). */ + bufp->can_be_null |= path_can_be_null; + + done: + RESET_FAIL_STACK (); + return 0; +} /* re_compile_fastmap */ +#ifdef _LIBC +weak_alias (__re_compile_fastmap, re_compile_fastmap) +#endif + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use + this memory for recording register information. STARTS and ENDS + must be allocated using the malloc library routine, and must each + be at least NUM_REGS * sizeof (regoff_t) bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ + +void +re_set_registers (bufp, regs, num_regs, starts, ends) + struct re_pattern_buffer *bufp; + struct re_registers *regs; + unsigned num_regs; + regoff_t *starts, *ends; +{ + if (num_regs) + { + bufp->regs_allocated = REGS_REALLOCATE; + regs->num_regs = num_regs; + regs->start = starts; + regs->end = ends; + } + else + { + bufp->regs_allocated = REGS_UNALLOCATED; + regs->num_regs = 0; + regs->start = regs->end = (regoff_t *) 0; + } +} +#ifdef _LIBC +weak_alias (__re_set_registers, re_set_registers) +#endif + +/* Searching routines. */ + +/* Like re_search_2, below, but only one string is specified, and + doesn't let you say where to stop matching. */ + +int +re_search (bufp, string, size, startpos, range, regs) + struct re_pattern_buffer *bufp; + const char *string; + int size, startpos, range; + struct re_registers *regs; +{ + return re_search_2 (bufp, NULL, 0, string, size, startpos, range, + regs, size); +} +#ifdef _LIBC +weak_alias (__re_search, re_search) +#endif + + +/* Using the compiled pattern in BUFP->buffer, first tries to match the + virtual concatenation of STRING1 and STRING2, starting first at index + STARTPOS, then at STARTPOS + 1, and so on. + + STRING1 and STRING2 have length SIZE1 and SIZE2, respectively. + + RANGE is how far to scan while trying to match. RANGE = 0 means try + only at STARTPOS; in general, the last start tried is STARTPOS + + RANGE. + + In REGS, return the indices of the virtual concatenation of STRING1 + and STRING2 that matched the entire BUFP->buffer and its contained + subexpressions. + + Do not consider matching one past the index STOP in the virtual + concatenation of STRING1 and STRING2. + + We return either the position in the strings at which the match was + found, -1 if no match, or -2 if error (such as failure + stack overflow). */ + +int +re_search_2 (bufp, string1, size1, string2, size2, startpos, range, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + int size1, size2; + int startpos; + int range; + struct re_registers *regs; + int stop; +{ + int val; + register char *fastmap = bufp->fastmap; + register RE_TRANSLATE_TYPE translate = bufp->translate; + int total_size = size1 + size2; + int endpos = startpos + range; + + /* Check for out-of-range STARTPOS. */ + if (startpos < 0 || startpos > total_size) + return -1; + + /* Fix up RANGE if it might eventually take us outside + the virtual concatenation of STRING1 and STRING2. + Make sure we won't move STARTPOS below 0 or above TOTAL_SIZE. */ + if (endpos < 0) + range = 0 - startpos; + else if (endpos > total_size) + range = total_size - startpos; + + /* If the search isn't to be a backwards one, don't waste time in a + search for a pattern that must be anchored. */ + if (bufp->used > 0 && range > 0 + && ((re_opcode_t) bufp->buffer[0] == begbuf + /* `begline' is like `begbuf' if it cannot match at newlines. */ + || ((re_opcode_t) bufp->buffer[0] == begline + && !bufp->newline_anchor))) + { + if (startpos > 0) + return -1; + else + range = 1; + } + +#ifdef emacs + /* In a forward search for something that starts with \=. + don't keep searching past point. */ + if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == at_dot && range > 0) + { + range = PT - startpos; + if (range <= 0) + return -1; + } +#endif /* emacs */ + + /* Update the fastmap now if not correct already. */ + if (fastmap && !bufp->fastmap_accurate) + if (re_compile_fastmap (bufp) == -2) + return -2; + + /* Loop through the string, looking for a place to start matching. */ + for (;;) + { + /* If a fastmap is supplied, skip quickly over characters that + cannot be the start of a match. If the pattern can match the + null string, however, we don't need to skip characters; we want + the first null string. */ + if (fastmap && startpos < total_size && !bufp->can_be_null) + { + if (range > 0) /* Searching forwards. */ + { + register const char *d; + register int lim = 0; + int irange = range; + + if (startpos < size1 && startpos + range >= size1) + lim = range - (size1 - startpos); + + d = (startpos >= size1 ? string2 - size1 : string1) + startpos; + + /* Written out as an if-else to avoid testing `translate' + inside the loop. */ + if (translate) + while (range > lim + && !fastmap[(unsigned char) + translate[(unsigned char) *d++]]) + range--; + else + while (range > lim && !fastmap[(unsigned char) *d++]) + range--; + + startpos += irange - range; + } + else /* Searching backwards. */ + { + register CHAR_TYPE c = (size1 == 0 || startpos >= size1 + ? string2[startpos - size1] + : string1[startpos]); + + if (!fastmap[(unsigned char) TRANSLATE (c)]) + goto advance; + } + } + + /* If can't match the null string, and that's all we have left, fail. */ + if (range >= 0 && startpos == total_size && fastmap + && !bufp->can_be_null) + return -1; + + val = re_match_2_internal (bufp, string1, size1, string2, size2, + startpos, regs, stop); +#ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +#endif + + if (val >= 0) + return startpos; + + if (val == -2) + return -2; + + advance: + if (!range) + break; + else if (range > 0) + { + range--; + startpos++; + } + else + { + range++; + startpos--; + } + } + return -1; +} /* re_search_2 */ +#ifdef _LIBC +weak_alias (__re_search_2, re_search_2) +#endif + +#ifdef MBS_SUPPORT +/* This converts PTR, a pointer into one of the search wchar_t strings + `string1' and `string2' into an multibyte string offset from the + beginning of that string. We use mbs_offset to optimize. + See convert_mbs_to_wcs. */ +# define POINTER_TO_OFFSET(ptr) \ + (FIRST_STRING_P (ptr) \ + ? ((regoff_t)(mbs_offset1 != NULL? mbs_offset1[(ptr)-string1] : 0)) \ + : ((regoff_t)((mbs_offset2 != NULL? mbs_offset2[(ptr)-string2] : 0) \ + + csize1))) +#else +/* This converts PTR, a pointer into one of the search strings `string1' + and `string2' into an offset from the beginning of that string. */ +# define POINTER_TO_OFFSET(ptr) \ + (FIRST_STRING_P (ptr) \ + ? ((regoff_t) ((ptr) - string1)) \ + : ((regoff_t) ((ptr) - string2 + size1))) +#endif /* MBS_SUPPORT */ + +/* Macros for dealing with the split strings in re_match_2. */ + +#define MATCHING_IN_FIRST_STRING (dend == end_match_1) + +/* Call before fetching a character with *d. This switches over to + string2 if necessary. */ +#define PREFETCH() \ + while (d == dend) \ + { \ + /* End of string2 => fail. */ \ + if (dend == end_match_2) \ + goto fail; \ + /* End of string1 => advance to string2. */ \ + d = string2; \ + dend = end_match_2; \ + } + + +/* Test if at very beginning or at very end of the virtual concatenation + of `string1' and `string2'. If only one string, it's `string2'. */ +#define AT_STRINGS_BEG(d) ((d) == (size1 ? string1 : string2) || !size2) +#define AT_STRINGS_END(d) ((d) == end2) + + +/* Test if D points to a character which is word-constituent. We have + two special cases to check for: if past the end of string1, look at + the first character in string2; and if before the beginning of + string2, look at the last character in string1. */ +#ifdef MBS_SUPPORT +/* Use internationalized API instead of SYNTAX. */ +# define WORDCHAR_P(d) \ + (iswalnum ((wint_t)((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d))) != 0) +#else +# define WORDCHAR_P(d) \ + (SYNTAX ((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d)) \ + == Sword) +#endif /* MBS_SUPPORT */ + +/* Disabled due to a compiler bug -- see comment at case wordbound */ +#if 0 +/* Test if the character before D and the one at D differ with respect + to being word-constituent. */ +#define AT_WORD_BOUNDARY(d) \ + (AT_STRINGS_BEG (d) || AT_STRINGS_END (d) \ + || WORDCHAR_P (d - 1) != WORDCHAR_P (d)) +#endif + +/* Free everything we malloc. */ +#ifdef MATCH_MAY_ALLOCATE +# define FREE_VAR(var) if (var) REGEX_FREE (var); var = NULL +# ifdef MBS_SUPPORT +# define FREE_VARIABLES() \ + do { \ + REGEX_FREE_STACK (fail_stack.stack); \ + FREE_VAR (regstart); \ + FREE_VAR (regend); \ + FREE_VAR (old_regstart); \ + FREE_VAR (old_regend); \ + FREE_VAR (best_regstart); \ + FREE_VAR (best_regend); \ + FREE_VAR (reg_info); \ + FREE_VAR (reg_dummy); \ + FREE_VAR (reg_info_dummy); \ + FREE_VAR (string1); \ + FREE_VAR (string2); \ + FREE_VAR (mbs_offset1); \ + FREE_VAR (mbs_offset2); \ + } while (0) +# else /* not MBS_SUPPORT */ +# define FREE_VARIABLES() \ + do { \ + REGEX_FREE_STACK (fail_stack.stack); \ + FREE_VAR (regstart); \ + FREE_VAR (regend); \ + FREE_VAR (old_regstart); \ + FREE_VAR (old_regend); \ + FREE_VAR (best_regstart); \ + FREE_VAR (best_regend); \ + FREE_VAR (reg_info); \ + FREE_VAR (reg_dummy); \ + FREE_VAR (reg_info_dummy); \ + } while (0) +# endif /* MBS_SUPPORT */ +#else +# define FREE_VAR(var) if (var) free (var); var = NULL +# ifdef MBS_SUPPORT +# define FREE_VARIABLES() \ + do { \ + FREE_VAR (string1); \ + FREE_VAR (string2); \ + FREE_VAR (mbs_offset1); \ + FREE_VAR (mbs_offset2); \ + } while (0) +# else +# define FREE_VARIABLES() ((void)0) /* Do nothing! But inhibit gcc warning. */ +# endif /* MBS_SUPPORT */ +#endif /* not MATCH_MAY_ALLOCATE */ + +/* These values must meet several constraints. They must not be valid + register values; since we have a limit of 255 registers (because + we use only one byte in the pattern for the register number), we can + use numbers larger than 255. They must differ by 1, because of + NUM_FAILURE_ITEMS above. And the value for the lowest register must + be larger than the value for the highest register, so we do not try + to actually save any registers when none are active. */ +#define NO_HIGHEST_ACTIVE_REG (1 << BYTEWIDTH) +#define NO_LOWEST_ACTIVE_REG (NO_HIGHEST_ACTIVE_REG + 1) + +/* Matching routines. */ + +#ifndef emacs /* Emacs never uses this. */ +/* re_match is like re_match_2 except it takes only a single string. */ + +int +re_match (bufp, string, size, pos, regs) + struct re_pattern_buffer *bufp; + const char *string; + int size, pos; + struct re_registers *regs; +{ + int result = re_match_2_internal (bufp, NULL, 0, string, size, + pos, regs, size); +# ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +# endif + return result; +} +# ifdef _LIBC +weak_alias (__re_match, re_match) +# endif +#endif /* not emacs */ + +static boolean group_match_null_string_p _RE_ARGS ((US_CHAR_TYPE **p, + US_CHAR_TYPE *end, + register_info_type *reg_info)); +static boolean alt_match_null_string_p _RE_ARGS ((US_CHAR_TYPE *p, + US_CHAR_TYPE *end, + register_info_type *reg_info)); +static boolean common_op_match_null_string_p _RE_ARGS ((US_CHAR_TYPE **p, + US_CHAR_TYPE *end, + register_info_type *reg_info)); +static int bcmp_translate _RE_ARGS ((const CHAR_TYPE *s1, const CHAR_TYPE *s2, + int len, char *translate)); + +/* re_match_2 matches the compiled pattern in BUFP against the + the (virtual) concatenation of STRING1 and STRING2 (of length SIZE1 + and SIZE2, respectively). We start matching at POS, and stop + matching at STOP. + + If REGS is non-null and the `no_sub' field of BUFP is nonzero, we + store offsets for the substring each group matched in REGS. See the + documentation for exactly how many groups we fill. + + We return -1 if no match, -2 if an internal error (such as the + failure stack overflowing). Otherwise, we return the length of the + matched substring. */ + +int +re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + int size1, size2; + int pos; + struct re_registers *regs; + int stop; +{ + int result = re_match_2_internal (bufp, string1, size1, string2, size2, + pos, regs, stop); +#ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +#endif + return result; +} +#ifdef _LIBC +weak_alias (__re_match_2, re_match_2) +#endif + +#ifdef MBS_SUPPORT + +static int count_mbs_length PARAMS ((int *, int)); + +/* This check the substring (from 0, to length) of the multibyte string, + to which offset_buffer correspond. And count how many wchar_t_characters + the substring occupy. We use offset_buffer to optimization. + See convert_mbs_to_wcs. */ + +static int +count_mbs_length(offset_buffer, length) + int *offset_buffer; + int length; +{ + int wcs_size; + + /* Check whether the size is valid. */ + if (length < 0) + return -1; + + if (offset_buffer == NULL) + return 0; + + for (wcs_size = 0 ; offset_buffer[wcs_size] != -1 ; wcs_size++) + { + if (offset_buffer[wcs_size] == length) + return wcs_size; + if (offset_buffer[wcs_size] > length) + /* It is a fragment of a wide character. */ + return -1; + } + + /* We reached at the sentinel. */ + return -1; +} +#endif /* MBS_SUPPORT */ + +/* This is a separate function so that we can force an alloca cleanup + afterwards. */ +static int +#ifdef MBS_SUPPORT +re_match_2_internal (bufp, cstring1, csize1, cstring2, csize2, pos, regs, stop) + struct re_pattern_buffer *bufp; + const char *cstring1, *cstring2; + int csize1, csize2; +#else +re_match_2_internal (bufp, string1, size1, string2, size2, pos, regs, stop) + struct re_pattern_buffer *bufp; + const char *string1, *string2; + int size1, size2; +#endif + int pos; + struct re_registers *regs; + int stop; +{ + /* General temporaries. */ + int mcnt; + US_CHAR_TYPE *p1; +#ifdef MBS_SUPPORT + /* We need wchar_t* buffers correspond to string1, string2. */ + CHAR_TYPE *string1 = NULL, *string2 = NULL; + /* We need the size of wchar_t buffers correspond to csize1, csize2. */ + int size1 = 0, size2 = 0; + /* offset buffer for optimizatoin. See convert_mbs_to_wc. */ + int *mbs_offset1 = NULL, *mbs_offset2 = NULL; + /* They hold whether each wchar_t is binary data or not. */ + char *is_binary = NULL; +#endif /* MBS_SUPPORT */ + + /* Just past the end of the corresponding string. */ + const CHAR_TYPE *end1, *end2; + + /* Pointers into string1 and string2, just past the last characters in + each to consider matching. */ + const CHAR_TYPE *end_match_1, *end_match_2; + + /* Where we are in the data, and the end of the current string. */ + const CHAR_TYPE *d, *dend; + + /* Where we are in the pattern, and the end of the pattern. */ +#ifdef MBS_SUPPORT + US_CHAR_TYPE *pattern, *p; + register US_CHAR_TYPE *pend; +#else + US_CHAR_TYPE *p = bufp->buffer; + register US_CHAR_TYPE *pend = p + bufp->used; +#endif /* MBS_SUPPORT */ + + /* Mark the opcode just after a start_memory, so we can test for an + empty subpattern when we get to the stop_memory. */ + US_CHAR_TYPE *just_past_start_mem = 0; + + /* We use this to map every character in the string. */ + RE_TRANSLATE_TYPE translate = bufp->translate; + + /* Failure point stack. Each place that can handle a failure further + down the line pushes a failure point on this stack. It consists of + restart, regend, and reg_info for all registers corresponding to + the subexpressions we're currently inside, plus the number of such + registers, and, finally, two char *'s. The first char * is where + to resume scanning the pattern; the second one is where to resume + scanning the strings. If the latter is zero, the failure point is + a ``dummy''; if a failure happens and the failure point is a dummy, + it gets discarded and the next next one is tried. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + fail_stack_type fail_stack; +#endif +#ifdef DEBUG + static unsigned failure_id; + unsigned nfailure_points_pushed = 0, nfailure_points_popped = 0; +#endif + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* We fill all the registers internally, independent of what we + return, for use in backreferences. The number here includes + an element for register zero. */ + size_t num_regs = bufp->re_nsub + 1; + + /* The currently active registers. */ + active_reg_t lowest_active_reg = NO_LOWEST_ACTIVE_REG; + active_reg_t highest_active_reg = NO_HIGHEST_ACTIVE_REG; + + /* Information on the contents of registers. These are pointers into + the input strings; they record just what was matched (on this + attempt) by a subexpression part of the pattern, that is, the + regnum-th regstart pointer points to where in the pattern we began + matching and the regnum-th regend points to right after where we + stopped matching the regnum-th subexpression. (The zeroth register + keeps track of what the whole pattern matches.) */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_TYPE **regstart, **regend; +#endif + + /* If a group that's operated upon by a repetition operator fails to + match anything, then the register for its start will need to be + restored because it will have been set to wherever in the string we + are when we last see its open-group operator. Similarly for a + register's end. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_TYPE **old_regstart, **old_regend; +#endif + + /* The is_active field of reg_info helps us keep track of which (possibly + nested) subexpressions we are currently in. The matched_something + field of reg_info[reg_num] helps us tell whether or not we have + matched any of the pattern so far this time through the reg_num-th + subexpression. These two fields get reset each time through any + loop their register is in. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + register_info_type *reg_info; +#endif + + /* The following record the register info as found in the above + variables when we find a match better than any we've seen before. + This happens as we backtrack through the failure points, which in + turn happens only if we have not yet matched the entire string. */ + unsigned best_regs_set = false; +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_TYPE **best_regstart, **best_regend; +#endif + + /* Logically, this is `best_regend[0]'. But we don't want to have to + allocate space for that if we're not allocating space for anything + else (see below). Also, we never need info about register 0 for + any of the other register vectors, and it seems rather a kludge to + treat `best_regend' differently than the rest. So we keep track of + the end of the best match so far in a separate variable. We + initialize this to NULL so that when we backtrack the first time + and need to test it, it's not garbage. */ + const CHAR_TYPE *match_end = NULL; + + /* This helps SET_REGS_MATCHED avoid doing redundant work. */ + int set_regs_matched_done = 0; + + /* Used when we pop values we don't care about. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_TYPE **reg_dummy; + register_info_type *reg_info_dummy; +#endif + +#ifdef DEBUG + /* Counts the total number of registers pushed. */ + unsigned num_regs_pushed = 0; +#endif + + DEBUG_PRINT1 ("\n\nEntering re_match_2.\n"); + + INIT_FAIL_STACK (); + +#ifdef MATCH_MAY_ALLOCATE + /* Do not bother to initialize all the register variables if there are + no groups in the pattern, as it takes a fair amount of time. If + there are groups, we include space for register 0 (the whole + pattern), even though we never use it, since it simplifies the + array indexing. We should fix this. */ + if (bufp->re_nsub) + { + regstart = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + regend = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + old_regstart = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + old_regend = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + best_regstart = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + best_regend = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + reg_info = REGEX_TALLOC (num_regs, register_info_type); + reg_dummy = REGEX_TALLOC (num_regs, const CHAR_TYPE *); + reg_info_dummy = REGEX_TALLOC (num_regs, register_info_type); + + if (!(regstart && regend && old_regstart && old_regend && reg_info + && best_regstart && best_regend && reg_dummy && reg_info_dummy)) + { + FREE_VARIABLES (); + return -2; + } + } + else + { + /* We must initialize all our variables to NULL, so that + `FREE_VARIABLES' doesn't try to free them. */ + regstart = regend = old_regstart = old_regend = best_regstart + = best_regend = reg_dummy = NULL; + reg_info = reg_info_dummy = (register_info_type *) NULL; + } +#endif /* MATCH_MAY_ALLOCATE */ + + /* The starting position is bogus. */ +#ifdef MBS_SUPPORT + if (pos < 0 || pos > csize1 + csize2) +#else + if (pos < 0 || pos > size1 + size2) +#endif + { + FREE_VARIABLES (); + return -1; + } + +#ifdef MBS_SUPPORT + /* Allocate wchar_t array for string1 and string2 and + fill them with converted string. */ + if (csize1 != 0) + { + string1 = REGEX_TALLOC (csize1 + 1, CHAR_TYPE); + mbs_offset1 = REGEX_TALLOC (csize1 + 1, int); + is_binary = REGEX_TALLOC (csize1 + 1, char); + if (!string1 || !mbs_offset1 || !is_binary) + { + FREE_VAR (string1); + FREE_VAR (mbs_offset1); + FREE_VAR (is_binary); + return -2; + } + size1 = convert_mbs_to_wcs(string1, cstring1, csize1, + mbs_offset1, is_binary); + string1[size1] = L'\0'; /* for a sentinel */ + FREE_VAR (is_binary); + } + if (csize2 != 0) + { + string2 = REGEX_TALLOC (csize2 + 1, CHAR_TYPE); + mbs_offset2 = REGEX_TALLOC (csize2 + 1, int); + is_binary = REGEX_TALLOC (csize2 + 1, char); + if (!string2 || !mbs_offset2 || !is_binary) + { + FREE_VAR (string1); + FREE_VAR (mbs_offset1); + FREE_VAR (string2); + FREE_VAR (mbs_offset2); + FREE_VAR (is_binary); + return -2; + } + size2 = convert_mbs_to_wcs(string2, cstring2, csize2, + mbs_offset2, is_binary); + string2[size2] = L'\0'; /* for a sentinel */ + FREE_VAR (is_binary); + } + + /* We need to cast pattern to (wchar_t*), because we casted this compiled + pattern to (char*) in regex_compile. */ + p = pattern = (CHAR_TYPE*)bufp->buffer; + pend = (CHAR_TYPE*)(bufp->buffer + bufp->used); + +#endif /* MBS_SUPPORT */ + + /* Initialize subexpression text positions to -1 to mark ones that no + start_memory/stop_memory has been seen for. Also initialize the + register information struct. */ + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = regend[mcnt] + = old_regstart[mcnt] = old_regend[mcnt] = REG_UNSET_VALUE; + + REG_MATCH_NULL_STRING_P (reg_info[mcnt]) = MATCH_NULL_UNSET_VALUE; + IS_ACTIVE (reg_info[mcnt]) = 0; + MATCHED_SOMETHING (reg_info[mcnt]) = 0; + EVER_MATCHED_SOMETHING (reg_info[mcnt]) = 0; + } + + /* We move `string1' into `string2' if the latter's empty -- but not if + `string1' is null. */ + if (size2 == 0 && string1 != NULL) + { + string2 = string1; + size2 = size1; + string1 = 0; + size1 = 0; + } + end1 = string1 + size1; + end2 = string2 + size2; + + /* Compute where to stop matching, within the two strings. */ +#ifdef MBS_SUPPORT + if (stop <= csize1) + { + mcnt = count_mbs_length(mbs_offset1, stop); + end_match_1 = string1 + mcnt; + end_match_2 = string2; + } + else + { + end_match_1 = end1; + mcnt = count_mbs_length(mbs_offset2, stop-csize1); + end_match_2 = string2 + mcnt; + } + if (mcnt < 0) + { /* count_mbs_length return error. */ + FREE_VARIABLES (); + return -1; + } +#else + if (stop <= size1) + { + end_match_1 = string1 + stop; + end_match_2 = string2; + } + else + { + end_match_1 = end1; + end_match_2 = string2 + stop - size1; + } +#endif /* MBS_SUPPORT */ + + /* `p' scans through the pattern as `d' scans through the data. + `dend' is the end of the input string that `d' points within. `d' + is advanced into the following input string whenever necessary, but + this happens before fetching; therefore, at the beginning of the + loop, `d' can be pointing at the end of a string, but it cannot + equal `string2'. */ +#ifdef MBS_SUPPORT + if (size1 > 0 && pos <= csize1) + { + mcnt = count_mbs_length(mbs_offset1, pos); + d = string1 + mcnt; + dend = end_match_1; + } + else + { + mcnt = count_mbs_length(mbs_offset2, pos-csize1); + d = string2 + mcnt; + dend = end_match_2; + } + + if (mcnt < 0) + { /* count_mbs_length return error. */ + FREE_VARIABLES (); + return -1; + } +#else + if (size1 > 0 && pos <= size1) + { + d = string1 + pos; + dend = end_match_1; + } + else + { + d = string2 + pos - size1; + dend = end_match_2; + } +#endif /* MBS_SUPPORT */ + + DEBUG_PRINT1 ("The compiled pattern is:\n"); + DEBUG_PRINT_COMPILED_PATTERN (bufp, p, pend); + DEBUG_PRINT1 ("The string to match is: `"); + DEBUG_PRINT_DOUBLE_STRING (d, string1, size1, string2, size2); + DEBUG_PRINT1 ("'\n"); + + /* This loops over pattern commands. It exits by returning from the + function if the match is complete, or it drops through if the match + fails at this starting point in the input data. */ + for (;;) + { +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + + if (p == pend) + { /* End of pattern means we might have succeeded. */ + DEBUG_PRINT1 ("end of pattern ... "); + + /* If we haven't matched the entire string, and we want the + longest match, try backtracking. */ + if (d != end_match_2) + { + /* 1 if this match ends in the same string (string1 or string2) + as the best previous match. */ + boolean same_str_p = (FIRST_STRING_P (match_end) + == MATCHING_IN_FIRST_STRING); + /* 1 if this match is the best seen so far. */ + boolean best_match_p; + + /* AIX compiler got confused when this was combined + with the previous declaration. */ + if (same_str_p) + best_match_p = d > match_end; + else + best_match_p = !MATCHING_IN_FIRST_STRING; + + DEBUG_PRINT1 ("backtracking.\n"); + + if (!FAIL_STACK_EMPTY ()) + { /* More failure points to try. */ + + /* If exceeds best match so far, save it. */ + if (!best_regs_set || best_match_p) + { + best_regs_set = true; + match_end = d; + + DEBUG_PRINT1 ("\nSAVING match as best so far.\n"); + + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + best_regstart[mcnt] = regstart[mcnt]; + best_regend[mcnt] = regend[mcnt]; + } + } + goto fail; + } + + /* If no failure points, don't restore garbage. And if + last match is real best match, don't restore second + best one. */ + else if (best_regs_set && !best_match_p) + { + restore_best_regs: + /* Restore best match. It may happen that `dend == + end_match_1' while the restored d is in string2. + For example, the pattern `x.*y.*z' against the + strings `x-' and `y-z-', if the two strings are + not consecutive in memory. */ + DEBUG_PRINT1 ("Restoring best registers.\n"); + + d = match_end; + dend = ((d >= string1 && d <= end1) + ? end_match_1 : end_match_2); + + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = best_regstart[mcnt]; + regend[mcnt] = best_regend[mcnt]; + } + } + } /* d != end_match_2 */ + + succeed_label: + DEBUG_PRINT1 ("Accepting match.\n"); + /* If caller wants register contents data back, do it. */ + if (regs && !bufp->no_sub) + { + /* Have the register data arrays been allocated? */ + if (bufp->regs_allocated == REGS_UNALLOCATED) + { /* No. So allocate them with malloc. We need one + extra element beyond `num_regs' for the `-1' marker + GNU code uses. */ + regs->num_regs = MAX (RE_NREGS, num_regs + 1); + regs->start = TALLOC (regs->num_regs, regoff_t); + regs->end = TALLOC (regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + { + FREE_VARIABLES (); + return -2; + } + bufp->regs_allocated = REGS_REALLOCATE; + } + else if (bufp->regs_allocated == REGS_REALLOCATE) + { /* Yes. If we need more elements than were already + allocated, reallocate them. If we need fewer, just + leave it alone. */ + if (regs->num_regs < num_regs + 1) + { + regs->num_regs = num_regs + 1; + RETALLOC (regs->start, regs->num_regs, regoff_t); + RETALLOC (regs->end, regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + { + FREE_VARIABLES (); + return -2; + } + } + } + else + { + /* These braces fend off a "empty body in an else-statement" + warning under GCC when assert expands to nothing. */ + assert (bufp->regs_allocated == REGS_FIXED); + } + + /* Convert the pointer data in `regstart' and `regend' to + indices. Register zero has to be set differently, + since we haven't kept track of any info for it. */ + if (regs->num_regs > 0) + { + regs->start[0] = pos; +#ifdef MBS_SUPPORT + if (MATCHING_IN_FIRST_STRING) + regs->end[0] = mbs_offset1 != NULL ? + mbs_offset1[d-string1] : 0; + else + regs->end[0] = csize1 + (mbs_offset2 != NULL ? + mbs_offset2[d-string2] : 0); +#else + regs->end[0] = (MATCHING_IN_FIRST_STRING + ? ((regoff_t) (d - string1)) + : ((regoff_t) (d - string2 + size1))); +#endif /* MBS_SUPPORT */ + } + + /* Go through the first `min (num_regs, regs->num_regs)' + registers, since that is all we initialized. */ + for (mcnt = 1; (unsigned) mcnt < MIN (num_regs, regs->num_regs); + mcnt++) + { + if (REG_UNSET (regstart[mcnt]) || REG_UNSET (regend[mcnt])) + regs->start[mcnt] = regs->end[mcnt] = -1; + else + { + regs->start[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regstart[mcnt]); + regs->end[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regend[mcnt]); + } + } + + /* If the regs structure we return has more elements than + were in the pattern, set the extra elements to -1. If + we (re)allocated the registers, this is the case, + because we always allocate enough to have at least one + -1 at the end. */ + for (mcnt = num_regs; (unsigned) mcnt < regs->num_regs; mcnt++) + regs->start[mcnt] = regs->end[mcnt] = -1; + } /* regs && !bufp->no_sub */ + + DEBUG_PRINT4 ("%u failure points pushed, %u popped (%u remain).\n", + nfailure_points_pushed, nfailure_points_popped, + nfailure_points_pushed - nfailure_points_popped); + DEBUG_PRINT2 ("%u registers pushed.\n", num_regs_pushed); + +#ifdef MBS_SUPPORT + if (MATCHING_IN_FIRST_STRING) + mcnt = mbs_offset1 != NULL ? mbs_offset1[d-string1] : 0; + else + mcnt = (mbs_offset2 != NULL ? mbs_offset2[d-string2] : 0) + + csize1; + mcnt -= pos; +#else + mcnt = d - pos - (MATCHING_IN_FIRST_STRING + ? string1 + : string2 - size1); +#endif /* MBS_SUPPORT */ + + DEBUG_PRINT2 ("Returning %d from re_match_2.\n", mcnt); + + FREE_VARIABLES (); + return mcnt; + } + + /* Otherwise match next pattern command. */ + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + /* Ignore these. Used to ignore the n of succeed_n's which + currently have n == 0. */ + case no_op: + DEBUG_PRINT1 ("EXECUTING no_op.\n"); + break; + + case succeed: + DEBUG_PRINT1 ("EXECUTING succeed.\n"); + goto succeed_label; + + /* Match the next n pattern characters exactly. The following + byte in the pattern defines n, and the n bytes after that + are the characters to match. */ + case exactn: +#ifdef MBS_SUPPORT + case exactn_bin: +#endif + mcnt = *p++; + DEBUG_PRINT2 ("EXECUTING exactn %d.\n", mcnt); + + /* This is written out as an if-else so we don't waste time + testing `translate' inside the loop. */ + if (translate) + { + do + { + PREFETCH (); +#ifdef MBS_SUPPORT + if (*d <= 0xff) + { + if ((US_CHAR_TYPE) translate[(unsigned char) *d++] + != (US_CHAR_TYPE) *p++) + goto fail; + } + else + { + if (*d++ != (CHAR_TYPE) *p++) + goto fail; + } +#else + if ((US_CHAR_TYPE) translate[(unsigned char) *d++] + != (US_CHAR_TYPE) *p++) + goto fail; +#endif /* MBS_SUPPORT */ + } + while (--mcnt); + } + else + { + do + { + PREFETCH (); + if (*d++ != (CHAR_TYPE) *p++) goto fail; + } + while (--mcnt); + } + SET_REGS_MATCHED (); + break; + + + /* Match any character except possibly a newline or a null. */ + case anychar: + DEBUG_PRINT1 ("EXECUTING anychar.\n"); + + PREFETCH (); + + if ((!(bufp->syntax & RE_DOT_NEWLINE) && TRANSLATE (*d) == '\n') + || (bufp->syntax & RE_DOT_NOT_NULL && TRANSLATE (*d) == '\000')) + goto fail; + + SET_REGS_MATCHED (); + DEBUG_PRINT2 (" Matched `%ld'.\n", (long int) *d); + d++; + break; + + + case charset: + case charset_not: + { + register US_CHAR_TYPE c; +#ifdef MBS_SUPPORT + unsigned int i, char_class_length, coll_symbol_length, + equiv_class_length, ranges_length, chars_length, length; + CHAR_TYPE *workp, *workp2, *charset_top; +#define WORK_BUFFER_SIZE 128 + CHAR_TYPE str_buf[WORK_BUFFER_SIZE]; +# ifdef _LIBC + uint32_t nrules; +# endif /* _LIBC */ +#endif /* MBS_SUPPORT */ + boolean not = (re_opcode_t) *(p - 1) == charset_not; + + DEBUG_PRINT2 ("EXECUTING charset%s.\n", not ? "_not" : ""); + PREFETCH (); + c = TRANSLATE (*d); /* The character to match. */ +#ifdef MBS_SUPPORT +# ifdef _LIBC + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif /* _LIBC */ + charset_top = p - 1; + char_class_length = *p++; + coll_symbol_length = *p++; + equiv_class_length = *p++; + ranges_length = *p++; + chars_length = *p++; + /* p points charset[6], so the address of the next instruction + (charset[l+m+n+2o+k+p']) equals p[l+m+n+2*o+p'], + where l=length of char_classes, m=length of collating_symbol, + n=equivalence_class, o=length of char_range, + p'=length of character. */ + workp = p; + /* Update p to indicate the next instruction. */ + p += char_class_length + coll_symbol_length+ equiv_class_length + + 2*ranges_length + chars_length; + + /* match with char_class? */ + for (i = 0; i < char_class_length ; i += CHAR_CLASS_SIZE) + { + wctype_t wctype; + uintptr_t alignedp = ((uintptr_t)workp + + __alignof__(wctype_t) - 1) + & ~(uintptr_t)(__alignof__(wctype_t) - 1); + wctype = *((wctype_t*)alignedp); + workp += CHAR_CLASS_SIZE; + if (iswctype((wint_t)c, wctype)) + goto char_set_matched; + } + + /* match with collating_symbol? */ +# ifdef _LIBC + if (nrules != 0) + { + const unsigned char *extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + + for (workp2 = workp + coll_symbol_length ; workp < workp2 ; + workp++) + { + int32_t *wextra; + wextra = (int32_t*)(extra + *workp++); + for (i = 0; i < *wextra; ++i) + if (TRANSLATE(d[i]) != wextra[1 + i]) + break; + + if (i == *wextra) + { + /* Update d, however d will be incremented at + char_set_matched:, we decrement d here. */ + d += i - 1; + goto char_set_matched; + } + } + } + else /* (nrules == 0) */ +# endif + /* If we can't look up collation data, we use wcscoll + instead. */ + { + for (workp2 = workp + coll_symbol_length ; workp < workp2 ;) + { + const CHAR_TYPE *backup_d = d, *backup_dend = dend; + length = wcslen(workp); + + /* If wcscoll(the collating symbol, whole string) > 0, + any substring of the string never match with the + collating symbol. */ + if (wcscoll(workp, d) > 0) + { + workp += length + 1; + continue; + } + + /* First, we compare the collating symbol with + the first character of the string. + If it don't match, we add the next character to + the compare buffer in turn. */ + for (i = 0 ; i < WORK_BUFFER_SIZE-1 ; i++, d++) + { + int match; + if (d == dend) + { + if (dend == end_match_2) + break; + d = string2; + dend = end_match_2; + } + + /* add next character to the compare buffer. */ + str_buf[i] = TRANSLATE(*d); + str_buf[i+1] = '\0'; + + match = wcscoll(workp, str_buf); + if (match == 0) + goto char_set_matched; + + if (match < 0) + /* (str_buf > workp) indicate (str_buf + X > workp), + because for all X (str_buf + X > str_buf). + So we don't need continue this loop. */ + break; + + /* Otherwise(str_buf < workp), + (str_buf+next_character) may equals (workp). + So we continue this loop. */ + } + /* not matched */ + d = backup_d; + dend = backup_dend; + workp += length + 1; + } + } + /* match with equivalence_class? */ +# ifdef _LIBC + if (nrules != 0) + { + const CHAR_TYPE *backup_d = d, *backup_dend = dend; + /* Try to match the equivalence class against + those known to the collate implementation. */ + const int32_t *table; + const int32_t *weights; + const int32_t *extra; + const int32_t *indirect; + int32_t idx, idx2; + wint_t *cp; + size_t len; + + /* This #include defines a local function! */ +# include <locale/weightwc.h> + + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); + weights = (const wint_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); + extra = (const wint_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); + + /* Write 1 collating element to str_buf, and + get its index. */ + idx2 = 0; + + for (i = 0 ; idx2 == 0 && i < WORK_BUFFER_SIZE - 1; i++) + { + cp = (wint_t*)str_buf; + if (d == dend) + { + if (dend == end_match_2) + break; + d = string2; + dend = end_match_2; + } + str_buf[i] = TRANSLATE(*(d+i)); + str_buf[i+1] = '\0'; /* sentinel */ + idx2 = findidx ((const wint_t**)&cp); + } + + /* Update d, however d will be incremented at + char_set_matched:, we decrement d here. */ + d = backup_d + ((wchar_t*)cp - (wchar_t*)str_buf - 1); + if (d >= dend) + { + if (dend == end_match_2) + d = dend; + else + { + d = string2; + dend = end_match_2; + } + } + + len = weights[idx2]; + + for (workp2 = workp + equiv_class_length ; workp < workp2 ; + workp++) + { + idx = (int32_t)*workp; + /* We already checked idx != 0 in regex_compile. */ + + if (idx2 != 0 && len == weights[idx]) + { + int cnt = 0; + while (cnt < len && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + goto char_set_matched; + } + } + /* not matched */ + d = backup_d; + dend = backup_dend; + } + else /* (nrules == 0) */ +# endif + /* If we can't look up collation data, we use wcscoll + instead. */ + { + for (workp2 = workp + equiv_class_length ; workp < workp2 ;) + { + const CHAR_TYPE *backup_d = d, *backup_dend = dend; + length = wcslen(workp); + + /* If wcscoll(the collating symbol, whole string) > 0, + any substring of the string never match with the + collating symbol. */ + if (wcscoll(workp, d) > 0) + { + workp += length + 1; + break; + } + + /* First, we compare the equivalence class with + the first character of the string. + If it don't match, we add the next character to + the compare buffer in turn. */ + for (i = 0 ; i < WORK_BUFFER_SIZE - 1 ; i++, d++) + { + int match; + if (d == dend) + { + if (dend == end_match_2) + break; + d = string2; + dend = end_match_2; + } + + /* add next character to the compare buffer. */ + str_buf[i] = TRANSLATE(*d); + str_buf[i+1] = '\0'; + + match = wcscoll(workp, str_buf); + + if (match == 0) + goto char_set_matched; + + if (match < 0) + /* (str_buf > workp) indicate (str_buf + X > workp), + because for all X (str_buf + X > str_buf). + So we don't need continue this loop. */ + break; + + /* Otherwise(str_buf < workp), + (str_buf+next_character) may equals (workp). + So we continue this loop. */ + } + /* not matched */ + d = backup_d; + dend = backup_dend; + workp += length + 1; + } + } + + /* match with char_range? */ +#ifdef _LIBC + if (nrules != 0) + { + uint32_t collseqval; + const char *collseq = (const char *) + _NL_CURRENT(LC_COLLATE, _NL_COLLATE_COLLSEQWC); + + collseqval = collseq_table_lookup (collseq, c); + + for (; workp < p - chars_length ;) + { + uint32_t start_val, end_val; + + /* We already compute the collation sequence value + of the characters (or collating symbols). */ + start_val = (uint32_t) *workp++; /* range_start */ + end_val = (uint32_t) *workp++; /* range_end */ + + if (start_val <= collseqval && collseqval <= end_val) + goto char_set_matched; + } + } + else +#endif + { + /* We set range_start_char at str_buf[0], range_end_char + at str_buf[4], and compared char at str_buf[2]. */ + str_buf[1] = 0; + str_buf[2] = c; + str_buf[3] = 0; + str_buf[5] = 0; + for (; workp < p - chars_length ;) + { + wchar_t *range_start_char, *range_end_char; + + /* match if (range_start_char <= c <= range_end_char). */ + + /* If range_start(or end) < 0, we assume -range_start(end) + is the offset of the collating symbol which is specified + as the character of the range start(end). */ + + /* range_start */ + if (*workp < 0) + range_start_char = charset_top - (*workp++); + else + { + str_buf[0] = *workp++; + range_start_char = str_buf; + } + + /* range_end */ + if (*workp < 0) + range_end_char = charset_top - (*workp++); + else + { + str_buf[4] = *workp++; + range_end_char = str_buf + 4; + } + + if (wcscoll(range_start_char, str_buf+2) <= 0 && + wcscoll(str_buf+2, range_end_char) <= 0) + + goto char_set_matched; + } + } + + /* match with char? */ + for (; workp < p ; workp++) + if (c == *workp) + goto char_set_matched; + + not = !not; + + char_set_matched: + if (not) goto fail; +#else + /* Cast to `unsigned' instead of `unsigned char' in case the + bit list is a full 32 bytes long. */ + if (c < (unsigned) (*p * BYTEWIDTH) + && p[1 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + p += 1 + *p; + + if (!not) goto fail; +#undef WORK_BUFFER_SIZE +#endif /* MBS_SUPPORT */ + SET_REGS_MATCHED (); + d++; + break; + } + + + /* The beginning of a group is represented by start_memory. + The arguments are the register number in the next byte, and the + number of groups inner to this one in the next. The text + matched within the group is recorded (in the internal + registers data structure) under the register number. */ + case start_memory: + DEBUG_PRINT3 ("EXECUTING start_memory %ld (%ld):\n", + (long int) *p, (long int) p[1]); + + /* Find out if this group can match the empty string. */ + p1 = p; /* To send to group_match_null_string_p. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[*p]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[*p]) + = group_match_null_string_p (&p1, pend, reg_info); + + /* Save the position in the string where we were the last time + we were at this open-group operator in case the group is + operated upon by a repetition operator, e.g., with `(a*)*b' + against `ab'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regstart[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regstart[*p]) ? d : regstart[*p] + : regstart[*p]; + DEBUG_PRINT2 (" old_regstart: %d\n", + POINTER_TO_OFFSET (old_regstart[*p])); + + regstart[*p] = d; + DEBUG_PRINT2 (" regstart: %d\n", POINTER_TO_OFFSET (regstart[*p])); + + IS_ACTIVE (reg_info[*p]) = 1; + MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 0; + + /* This is the new highest active register. */ + highest_active_reg = *p; + + /* If nothing was active before, this is the new lowest active + register. */ + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *p; + + /* Move past the register number and inner group count. */ + p += 2; + just_past_start_mem = p; + + break; + + + /* The stop_memory opcode represents the end of a group. Its + arguments are the same as start_memory's: the register + number, and the number of inner groups. */ + case stop_memory: + DEBUG_PRINT3 ("EXECUTING stop_memory %ld (%ld):\n", + (long int) *p, (long int) p[1]); + + /* We need to save the string position the last time we were at + this close-group operator in case the group is operated + upon by a repetition operator, e.g., with `((a*)*(b*)*)*' + against `aba'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regend[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regend[*p]) ? d : regend[*p] + : regend[*p]; + DEBUG_PRINT2 (" old_regend: %d\n", + POINTER_TO_OFFSET (old_regend[*p])); + + regend[*p] = d; + DEBUG_PRINT2 (" regend: %d\n", POINTER_TO_OFFSET (regend[*p])); + + /* This register isn't active anymore. */ + IS_ACTIVE (reg_info[*p]) = 0; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 0; + + /* If this was the only register active, nothing is active + anymore. */ + if (lowest_active_reg == highest_active_reg) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + { /* We must scan for the new highest active register, since + it isn't necessarily one less than now: consider + (a(b)c(d(e)f)g). When group 3 ends, after the f), the + new highest active register is 1. */ + US_CHAR_TYPE r = *p - 1; + while (r > 0 && !IS_ACTIVE (reg_info[r])) + r--; + + /* If we end up at register zero, that means that we saved + the registers as the result of an `on_failure_jump', not + a `start_memory', and we jumped to past the innermost + `stop_memory'. For example, in ((.)*) we save + registers 1 and 2 as a result of the *, but when we pop + back to the second ), we are at the stop_memory 1. + Thus, nothing is active. */ + if (r == 0) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + highest_active_reg = r; + } + + /* If just failed to match something this time around with a + group that's operated on by a repetition operator, try to + force exit from the ``loop'', and restore the register + information for this group that we had before trying this + last match. */ + if ((!MATCHED_SOMETHING (reg_info[*p]) + || just_past_start_mem == p - 1) + && (p + 2) < pend) + { + boolean is_a_jump_n = false; + + p1 = p + 2; + mcnt = 0; + switch ((re_opcode_t) *p1++) + { + case jump_n: + is_a_jump_n = true; + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (is_a_jump_n) + p1 += OFFSET_ADDRESS_SIZE; + break; + + default: + /* do nothing */ ; + } + p1 += mcnt; + + /* If the next operation is a jump backwards in the pattern + to an on_failure_jump right before the start_memory + corresponding to this stop_memory, exit from the loop + by forcing a failure after pushing on the stack the + on_failure_jump's jump in the pattern, and d. */ + if (mcnt < 0 && (re_opcode_t) *p1 == on_failure_jump + && (re_opcode_t) p1[1+OFFSET_ADDRESS_SIZE] == start_memory + && p1[2+OFFSET_ADDRESS_SIZE] == *p) + { + /* If this group ever matched anything, then restore + what its registers were before trying this last + failed match, e.g., with `(a*)*b' against `ab' for + regstart[1], and, e.g., with `((a*)*(b*)*)*' + against `aba' for regend[3]. + + Also restore the registers for inner groups for, + e.g., `((a*)(b*))*' against `aba' (register 3 would + otherwise get trashed). */ + + if (EVER_MATCHED_SOMETHING (reg_info[*p])) + { + unsigned r; + + EVER_MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Restore this and inner groups' (if any) registers. */ + for (r = *p; r < (unsigned) *p + (unsigned) *(p + 1); + r++) + { + regstart[r] = old_regstart[r]; + + /* xx why this test? */ + if (old_regend[r] >= regstart[r]) + regend[r] = old_regend[r]; + } + } + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + PUSH_FAILURE_POINT (p1 + mcnt, d, -2); + + goto fail; + } + } + + /* Move past the register number and the inner group count. */ + p += 2; + break; + + + /* \<digit> has been turned into a `duplicate' command which is + followed by the numeric value of <digit> as the register number. */ + case duplicate: + { + register const CHAR_TYPE *d2, *dend2; + int regno = *p++; /* Get which register to match against. */ + DEBUG_PRINT2 ("EXECUTING duplicate %d.\n", regno); + + /* Can't back reference a group which we've never matched. */ + if (REG_UNSET (regstart[regno]) || REG_UNSET (regend[regno])) + goto fail; + + /* Where in input to try to start matching. */ + d2 = regstart[regno]; + + /* Where to stop matching; if both the place to start and + the place to stop matching are in the same string, then + set to the place to stop, otherwise, for now have to use + the end of the first string. */ + + dend2 = ((FIRST_STRING_P (regstart[regno]) + == FIRST_STRING_P (regend[regno])) + ? regend[regno] : end_match_1); + for (;;) + { + /* If necessary, advance to next segment in register + contents. */ + while (d2 == dend2) + { + if (dend2 == end_match_2) break; + if (dend2 == regend[regno]) break; + + /* End of string1 => advance to string2. */ + d2 = string2; + dend2 = regend[regno]; + } + /* At end of register contents => success */ + if (d2 == dend2) break; + + /* If necessary, advance to next segment in data. */ + PREFETCH (); + + /* How many characters left in this segment to match. */ + mcnt = dend - d; + + /* Want how many consecutive characters we can match in + one shot, so, if necessary, adjust the count. */ + if (mcnt > dend2 - d2) + mcnt = dend2 - d2; + + /* Compare that many; failure if mismatch, else move + past them. */ + if (translate + ? bcmp_translate (d, d2, mcnt, translate) + : memcmp (d, d2, mcnt*sizeof(US_CHAR_TYPE))) + goto fail; + d += mcnt, d2 += mcnt; + + /* Do this because we've match some characters. */ + SET_REGS_MATCHED (); + } + } + break; + + + /* begline matches the empty string at the beginning of the string + (unless `not_bol' is set in `bufp'), and, if + `newline_anchor' is set, after newlines. */ + case begline: + DEBUG_PRINT1 ("EXECUTING begline.\n"); + + if (AT_STRINGS_BEG (d)) + { + if (!bufp->not_bol) break; + } + else if (d[-1] == '\n' && bufp->newline_anchor) + { + break; + } + /* In all other cases, we fail. */ + goto fail; + + + /* endline is the dual of begline. */ + case endline: + DEBUG_PRINT1 ("EXECUTING endline.\n"); + + if (AT_STRINGS_END (d)) + { + if (!bufp->not_eol) break; + } + + /* We have to ``prefetch'' the next character. */ + else if ((d == end1 ? *string2 : *d) == '\n' + && bufp->newline_anchor) + { + break; + } + goto fail; + + + /* Match at the very beginning of the data. */ + case begbuf: + DEBUG_PRINT1 ("EXECUTING begbuf.\n"); + if (AT_STRINGS_BEG (d)) + break; + goto fail; + + + /* Match at the very end of the data. */ + case endbuf: + DEBUG_PRINT1 ("EXECUTING endbuf.\n"); + if (AT_STRINGS_END (d)) + break; + goto fail; + + + /* on_failure_keep_string_jump is used to optimize `.*\n'. It + pushes NULL as the value for the string on the stack. Then + `pop_failure_point' will keep the current value for the + string, instead of restoring it. To see why, consider + matching `foo\nbar' against `.*\n'. The .* matches the foo; + then the . fails against the \n. But the next thing we want + to do is match the \n against the \n; if we restored the + string value, we would be back at the foo. + + Because this is used only in specific cases, we don't need to + check all the things that `on_failure_jump' does, to make + sure the right things get saved on the stack. Hence we don't + share its code. The only reason to push anything on the + stack at all is that otherwise we would have to change + `anychar's code to do something besides goto fail in this + case; that seems worse than this. */ + case on_failure_keep_string_jump: + DEBUG_PRINT1 ("EXECUTING on_failure_keep_string_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p):\n", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x):\n", mcnt, p + mcnt); +#endif + + PUSH_FAILURE_POINT (p + mcnt, NULL, -2); + break; + + + /* Uses of on_failure_jump: + + Each alternative starts with an on_failure_jump that points + to the beginning of the next alternative. Each alternative + except the last ends with a jump that in effect jumps past + the rest of the alternatives. (They really jump to the + ending jump of the following alternative, because tensioning + these jumps is a hassle.) + + Repeats start with an on_failure_jump that points past both + the repetition text and either the following jump or + pop_failure_jump back to this on_failure_jump. */ + case on_failure_jump: + on_failure: + DEBUG_PRINT1 ("EXECUTING on_failure_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p)", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x)", mcnt, p + mcnt); +#endif + + /* If this on_failure_jump comes right before a group (i.e., + the original * applied to a group), save the information + for that group and all inner ones, so that if we fail back + to this point, the group's information will be correct. + For example, in \(a*\)*\1, we need the preceding group, + and in \(zz\(a*\)b*\)\2, we need the inner group. */ + + /* We can't use `p' to check ahead because we push + a failure point to `p + mcnt' after we do this. */ + p1 = p; + + /* We need to skip no_op's before we look for the + start_memory in case this on_failure_jump is happening as + the result of a completed succeed_n, as in \(a\)\{1,3\}b\1 + against aba. */ + while (p1 < pend && (re_opcode_t) *p1 == no_op) + p1++; + + if (p1 < pend && (re_opcode_t) *p1 == start_memory) + { + /* We have a new highest active register now. This will + get reset at the start_memory we are about to get to, + but we will have saved all the registers relevant to + this repetition op, as described above. */ + highest_active_reg = *(p1 + 1) + *(p1 + 2); + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *(p1 + 1); + } + + DEBUG_PRINT1 (":\n"); + PUSH_FAILURE_POINT (p + mcnt, d, -2); + break; + + + /* A smart repeat ends with `maybe_pop_jump'. + We change it to either `pop_failure_jump' or `jump'. */ + case maybe_pop_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT2 ("EXECUTING maybe_pop_jump %d.\n", mcnt); + { + register US_CHAR_TYPE *p2 = p; + + /* Compare the beginning of the repeat with what in the + pattern follows its end. If we can establish that there + is nothing that they would both match, i.e., that we + would have to backtrack because of (as in, e.g., `a*a') + then we can change to pop_failure_jump, because we'll + never have to backtrack. + + This is not true in the case of alternatives: in + `(a|ab)*' we do need to backtrack to the `ab' alternative + (e.g., if the string was `ab'). But instead of trying to + detect that here, the alternative has put on a dummy + failure point which is what we will end up popping. */ + + /* Skip over open/close-group commands. + If what follows this loop is a ...+ construct, + look at what begins its body, since we will have to + match at least one of that. */ + while (1) + { + if (p2 + 2 < pend + && ((re_opcode_t) *p2 == stop_memory + || (re_opcode_t) *p2 == start_memory)) + p2 += 3; + else if (p2 + 2 + 2 * OFFSET_ADDRESS_SIZE < pend + && (re_opcode_t) *p2 == dummy_failure_jump) + p2 += 2 + 2 * OFFSET_ADDRESS_SIZE; + else + break; + } + + p1 = p + mcnt; + /* p1[0] ... p1[2] are the `on_failure_jump' corresponding + to the `maybe_finalize_jump' of this case. Examine what + follows. */ + + /* If we're at the end of the pattern, we can change. */ + if (p2 == pend) + { + /* Consider what happens when matching ":\(.*\)" + against ":/". I don't really understand this code + yet. */ + p[-(1+OFFSET_ADDRESS_SIZE)] = (US_CHAR_TYPE) + pop_failure_jump; + DEBUG_PRINT1 + (" End of pattern: change to `pop_failure_jump'.\n"); + } + + else if ((re_opcode_t) *p2 == exactn +#ifdef MBS_SUPPORT + || (re_opcode_t) *p2 == exactn_bin +#endif + || (bufp->newline_anchor && (re_opcode_t) *p2 == endline)) + { + register US_CHAR_TYPE c + = *p2 == (US_CHAR_TYPE) endline ? '\n' : p2[2]; + + if (((re_opcode_t) p1[1+OFFSET_ADDRESS_SIZE] == exactn +#ifdef MBS_SUPPORT + || (re_opcode_t) p1[1+OFFSET_ADDRESS_SIZE] == exactn_bin +#endif + ) && p1[3+OFFSET_ADDRESS_SIZE] != c) + { + p[-(1+OFFSET_ADDRESS_SIZE)] = (US_CHAR_TYPE) + pop_failure_jump; +#ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + DEBUG_PRINT3 (" %C != %C => pop_failure_jump.\n", + (wint_t) c, + (wint_t) p1[3+OFFSET_ADDRESS_SIZE]); + else +#endif + DEBUG_PRINT3 (" %c != %c => pop_failure_jump.\n", + (char) c, + (char) p1[3+OFFSET_ADDRESS_SIZE]); + } + +#ifndef MBS_SUPPORT + else if ((re_opcode_t) p1[3] == charset + || (re_opcode_t) p1[3] == charset_not) + { + int not = (re_opcode_t) p1[3] == charset_not; + + if (c < (unsigned) (p1[4] * BYTEWIDTH) + && p1[5 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + /* `not' is equal to 1 if c would match, which means + that we can't change to pop_failure_jump. */ + if (!not) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } +#endif /* not MBS_SUPPORT */ + } +#ifndef MBS_SUPPORT + else if ((re_opcode_t) *p2 == charset) + { + /* We win if the first character of the loop is not part + of the charset. */ + if ((re_opcode_t) p1[3] == exactn + && ! ((int) p2[1] * BYTEWIDTH > (int) p1[5] + && (p2[2 + p1[5] / BYTEWIDTH] + & (1 << (p1[5] % BYTEWIDTH))))) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + + else if ((re_opcode_t) p1[3] == charset_not) + { + int idx; + /* We win if the charset_not inside the loop + lists every character listed in the charset after. */ + for (idx = 0; idx < (int) p2[1]; idx++) + if (! (p2[2 + idx] == 0 + || (idx < (int) p1[4] + && ((p2[2 + idx] & ~ p1[5 + idx]) == 0)))) + break; + + if (idx == p2[1]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + else if ((re_opcode_t) p1[3] == charset) + { + int idx; + /* We win if the charset inside the loop + has no overlap with the one after the loop. */ + for (idx = 0; + idx < (int) p2[1] && idx < (int) p1[4]; + idx++) + if ((p2[2 + idx] & p1[5 + idx]) != 0) + break; + + if (idx == p2[1] || idx == p1[4]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + } +#endif /* not MBS_SUPPORT */ + } + p -= OFFSET_ADDRESS_SIZE; /* Point at relative address again. */ + if ((re_opcode_t) p[-1] != pop_failure_jump) + { + p[-1] = (US_CHAR_TYPE) jump; + DEBUG_PRINT1 (" Match => jump.\n"); + goto unconditional_jump; + } + /* Note fall through. */ + + + /* The end of a simple repeat has a pop_failure_jump back to + its matching on_failure_jump, where the latter will push a + failure point. The pop_failure_jump takes off failure + points put on by this pop_failure_jump's matching + on_failure_jump; we got through the pattern to here from the + matching on_failure_jump, so didn't fail. */ + case pop_failure_jump: + { + /* We need to pass separate storage for the lowest and + highest registers, even though we don't care about the + actual values. Otherwise, we will restore only one + register from the stack, since lowest will == highest in + `pop_failure_point'. */ + active_reg_t dummy_low_reg, dummy_high_reg; + US_CHAR_TYPE *pdummy = NULL; + const CHAR_TYPE *sdummy = NULL; + + DEBUG_PRINT1 ("EXECUTING pop_failure_jump.\n"); + POP_FAILURE_POINT (sdummy, pdummy, + dummy_low_reg, dummy_high_reg, + reg_dummy, reg_dummy, reg_info_dummy); + } + /* Note fall through. */ + + unconditional_jump: +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + /* Note fall through. */ + + /* Unconditionally jump (without popping any failure points). */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); /* Get the amount to jump. */ + DEBUG_PRINT2 ("EXECUTING jump %d ", mcnt); + p += mcnt; /* Do the jump. */ +#ifdef _LIBC + DEBUG_PRINT2 ("(to %p).\n", p); +#else + DEBUG_PRINT2 ("(to 0x%x).\n", p); +#endif + break; + + + /* We need this opcode so we can detect where alternatives end + in `group_match_null_string_p' et al. */ + case jump_past_alt: + DEBUG_PRINT1 ("EXECUTING jump_past_alt.\n"); + goto unconditional_jump; + + + /* Normally, the on_failure_jump pushes a failure point, which + then gets popped at pop_failure_jump. We will end up at + pop_failure_jump, also, and with a pattern of, say, `a+', we + are skipping over the on_failure_jump, so we have to push + something meaningless for pop_failure_jump to pop. */ + case dummy_failure_jump: + DEBUG_PRINT1 ("EXECUTING dummy_failure_jump.\n"); + /* It doesn't matter what we push for the string here. What + the code at `fail' tests is the value for the pattern. */ + PUSH_FAILURE_POINT (NULL, NULL, -2); + goto unconditional_jump; + + + /* At the end of an alternative, we need to push a dummy failure + point in case we are followed by a `pop_failure_jump', because + we don't want the failure point for the alternative to be + popped. For example, matching `(a|ab)*' against `aab' + requires that we match the `ab' alternative. */ + case push_dummy_failure: + DEBUG_PRINT1 ("EXECUTING push_dummy_failure.\n"); + /* See comments just above at `dummy_failure_jump' about the + two zeroes. */ + PUSH_FAILURE_POINT (NULL, NULL, -2); + break; + + /* Have to succeed matching what follows at least n times. + After that, handle like `on_failure_jump'. */ + case succeed_n: + EXTRACT_NUMBER (mcnt, p + OFFSET_ADDRESS_SIZE); + DEBUG_PRINT2 ("EXECUTING succeed_n %d.\n", mcnt); + + assert (mcnt >= 0); + /* Originally, this is how many times we HAVE to succeed. */ + if (mcnt > 0) + { + mcnt--; + p += OFFSET_ADDRESS_SIZE; + STORE_NUMBER_AND_INCR (p, mcnt); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p - OFFSET_ADDRESS_SIZE + , mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p - OFFSET_ADDRESS_SIZE + , mcnt); +#endif + } + else if (mcnt == 0) + { +#ifdef _LIBC + DEBUG_PRINT2 (" Setting two bytes from %p to no_op.\n", + p + OFFSET_ADDRESS_SIZE); +#else + DEBUG_PRINT2 (" Setting two bytes from 0x%x to no_op.\n", + p + OFFSET_ADDRESS_SIZE); +#endif /* _LIBC */ + +#ifdef MBS_SUPPORT + p[1] = (US_CHAR_TYPE) no_op; +#else + p[2] = (US_CHAR_TYPE) no_op; + p[3] = (US_CHAR_TYPE) no_op; +#endif /* MBS_SUPPORT */ + goto on_failure; + } + break; + + case jump_n: + EXTRACT_NUMBER (mcnt, p + OFFSET_ADDRESS_SIZE); + DEBUG_PRINT2 ("EXECUTING jump_n %d.\n", mcnt); + + /* Originally, this is how many times we CAN jump. */ + if (mcnt) + { + mcnt--; + STORE_NUMBER (p + OFFSET_ADDRESS_SIZE, mcnt); + +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p + OFFSET_ADDRESS_SIZE, + mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p + OFFSET_ADDRESS_SIZE, + mcnt); +#endif /* _LIBC */ + goto unconditional_jump; + } + /* If don't have to jump any more, skip over the rest of command. */ + else + p += 2 * OFFSET_ADDRESS_SIZE; + break; + + case set_number_at: + { + DEBUG_PRINT1 ("EXECUTING set_number_at.\n"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); + p1 = p + mcnt; + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p1, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p1, mcnt); +#endif + STORE_NUMBER (p1, mcnt); + break; + } + +#if 0 + /* The DEC Alpha C compiler 3.x generates incorrect code for the + test WORDCHAR_P (d - 1) != WORDCHAR_P (d) in the expansion of + AT_WORD_BOUNDARY, so this code is disabled. Expanding the + macro and introducing temporary variables works around the bug. */ + + case wordbound: + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + break; + goto fail; + + case notwordbound: + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + goto fail; + break; +#else + case wordbound: + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + break; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + break; + goto fail; + } + + case notwordbound: + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + goto fail; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + goto fail; + break; + } +#endif + + case wordbeg: + DEBUG_PRINT1 ("EXECUTING wordbeg.\n"); + if (WORDCHAR_P (d) && (AT_STRINGS_BEG (d) || !WORDCHAR_P (d - 1))) + break; + goto fail; + + case wordend: + DEBUG_PRINT1 ("EXECUTING wordend.\n"); + if (!AT_STRINGS_BEG (d) && WORDCHAR_P (d - 1) + && (!WORDCHAR_P (d) || AT_STRINGS_END (d))) + break; + goto fail; + +#ifdef emacs + case before_dot: + DEBUG_PRINT1 ("EXECUTING before_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) >= point) + goto fail; + break; + + case at_dot: + DEBUG_PRINT1 ("EXECUTING at_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) != point) + goto fail; + break; + + case after_dot: + DEBUG_PRINT1 ("EXECUTING after_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) <= point) + goto fail; + break; + + case syntaxspec: + DEBUG_PRINT2 ("EXECUTING syntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchsyntax; + + case wordchar: + DEBUG_PRINT1 ("EXECUTING Emacs wordchar.\n"); + mcnt = (int) Sword; + matchsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) != (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + break; + + case notsyntaxspec: + DEBUG_PRINT2 ("EXECUTING notsyntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchnotsyntax; + + case notwordchar: + DEBUG_PRINT1 ("EXECUTING Emacs notwordchar.\n"); + mcnt = (int) Sword; + matchnotsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) == (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + break; + +#else /* not emacs */ + case wordchar: + DEBUG_PRINT1 ("EXECUTING non-Emacs wordchar.\n"); + PREFETCH (); + if (!WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + break; + + case notwordchar: + DEBUG_PRINT1 ("EXECUTING non-Emacs notwordchar.\n"); + PREFETCH (); + if (WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + break; +#endif /* not emacs */ + + default: + abort (); + } + continue; /* Successfully executed one pattern command; keep going. */ + + + /* We goto here if a matching operation fails. */ + fail: + if (!FAIL_STACK_EMPTY ()) + { /* A restart point is known. Restore to that state. */ + DEBUG_PRINT1 ("\nFAIL:\n"); + POP_FAILURE_POINT (d, p, + lowest_active_reg, highest_active_reg, + regstart, regend, reg_info); + + /* If this failure point is a dummy, try the next one. */ + if (!p) + goto fail; + + /* If we failed to the end of the pattern, don't examine *p. */ + assert (p <= pend); + if (p < pend) + { + boolean is_a_jump_n = false; + + /* If failed to a backwards jump that's part of a repetition + loop, need to pop this failure point and use the next one. */ + switch ((re_opcode_t) *p) + { + case jump_n: + is_a_jump_n = true; + case maybe_pop_jump: + case pop_failure_jump: + case jump: + p1 = p + 1; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + + if ((is_a_jump_n && (re_opcode_t) *p1 == succeed_n) + || (!is_a_jump_n + && (re_opcode_t) *p1 == on_failure_jump)) + goto fail; + break; + default: + /* do nothing */ ; + } + } + + if (d >= string1 && d <= end1) + dend = end_match_1; + } + else + break; /* Matching at this starting point really fails. */ + } /* for (;;) */ + + if (best_regs_set) + goto restore_best_regs; + + FREE_VARIABLES (); + + return -1; /* Failure to match. */ +} /* re_match_2 */ + +/* Subroutine definitions for re_match_2. */ + + +/* We are passed P pointing to a register number after a start_memory. + + Return true if the pattern up to the corresponding stop_memory can + match the empty string, and false otherwise. + + If we find the matching stop_memory, sets P to point to one past its number. + Otherwise, sets P to an undefined byte less than or equal to END. + + We don't handle duplicates properly (yet). */ + +static boolean +group_match_null_string_p (p, end, reg_info) + US_CHAR_TYPE **p, *end; + register_info_type *reg_info; +{ + int mcnt; + /* Point to after the args to the start_memory. */ + US_CHAR_TYPE *p1 = *p + 2; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and return true or + false, as appropriate, when we get to one that can't, or to the + matching stop_memory. */ + + switch ((re_opcode_t) *p1) + { + /* Could be either a loop or a series of alternatives. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + /* If the next operation is not a jump backwards in the + pattern. */ + + if (mcnt >= 0) + { + /* Go through the on_failure_jumps of the alternatives, + seeing if any of the alternatives cannot match nothing. + The last alternative starts with only a jump, + whereas the rest start with on_failure_jump and end + with a jump, e.g., here is the pattern for `a|b|c': + + /on_failure_jump/0/6/exactn/1/a/jump_past_alt/0/6 + /on_failure_jump/0/6/exactn/1/b/jump_past_alt/0/3 + /exactn/1/c + + So, we have to first go through the first (n-1) + alternatives and then deal with the last one separately. */ + + + /* Deal with the first (n-1) alternatives, which start + with an on_failure_jump (see above) that jumps to right + past a jump_past_alt. */ + + while ((re_opcode_t) p1[mcnt-(1+OFFSET_ADDRESS_SIZE)] == + jump_past_alt) + { + /* `mcnt' holds how many bytes long the alternative + is, including the ending `jump_past_alt' and + its number. */ + + if (!alt_match_null_string_p (p1, p1 + mcnt - + (1 + OFFSET_ADDRESS_SIZE), + reg_info)) + return false; + + /* Move to right after this alternative, including the + jump_past_alt. */ + p1 += mcnt; + + /* Break if it's the beginning of an n-th alternative + that doesn't begin with an on_failure_jump. */ + if ((re_opcode_t) *p1 != on_failure_jump) + break; + + /* Still have to check that it's not an n-th + alternative that starts with an on_failure_jump. */ + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if ((re_opcode_t) p1[mcnt-(1+OFFSET_ADDRESS_SIZE)] != + jump_past_alt) + { + /* Get to the beginning of the n-th alternative. */ + p1 -= 1 + OFFSET_ADDRESS_SIZE; + break; + } + } + + /* Deal with the last alternative: go back and get number + of the `jump_past_alt' just before it. `mcnt' contains + the length of the alternative. */ + EXTRACT_NUMBER (mcnt, p1 - OFFSET_ADDRESS_SIZE); + + if (!alt_match_null_string_p (p1, p1 + mcnt, reg_info)) + return false; + + p1 += mcnt; /* Get past the n-th alternative. */ + } /* if mcnt > 0 */ + break; + + + case stop_memory: + assert (p1[1] == **p); + *p = p1 + 2; + return true; + + + default: + if (!common_op_match_null_string_p (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return false; +} /* group_match_null_string_p */ + + +/* Similar to group_match_null_string_p, but doesn't deal with alternatives: + It expects P to be the first byte of a single alternative and END one + byte past the last. The alternative can contain groups. */ + +static boolean +alt_match_null_string_p (p, end, reg_info) + US_CHAR_TYPE *p, *end; + register_info_type *reg_info; +{ + int mcnt; + US_CHAR_TYPE *p1 = p; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and break when we get + to one that can't. */ + + switch ((re_opcode_t) *p1) + { + /* It's a loop. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + break; + + default: + if (!common_op_match_null_string_p (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return true; +} /* alt_match_null_string_p */ + + +/* Deals with the ops common to group_match_null_string_p and + alt_match_null_string_p. + + Sets P to one after the op and its arguments, if any. */ + +static boolean +common_op_match_null_string_p (p, end, reg_info) + US_CHAR_TYPE **p, *end; + register_info_type *reg_info; +{ + int mcnt; + boolean ret; + int reg_no; + US_CHAR_TYPE *p1 = *p; + + switch ((re_opcode_t) *p1++) + { + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbeg: + case wordend: + case wordbound: + case notwordbound: +#ifdef emacs + case before_dot: + case at_dot: + case after_dot: +#endif + break; + + case start_memory: + reg_no = *p1; + assert (reg_no > 0 && reg_no <= MAX_REGNUM); + ret = group_match_null_string_p (&p1, end, reg_info); + + /* Have to set this here in case we're checking a group which + contains a group and a back reference to it. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[reg_no]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[reg_no]) = ret; + + if (!ret) + return false; + break; + + /* If this is an optimized succeed_n for zero times, make the jump. */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (mcnt >= 0) + p1 += mcnt; + else + return false; + break; + + case succeed_n: + /* Get to the number of times to succeed. */ + p1 += OFFSET_ADDRESS_SIZE; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + if (mcnt == 0) + { + p1 -= 2 * OFFSET_ADDRESS_SIZE; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + } + else + return false; + break; + + case duplicate: + if (!REG_MATCH_NULL_STRING_P (reg_info[*p1])) + return false; + break; + + case set_number_at: + p1 += 2 * OFFSET_ADDRESS_SIZE; + + default: + /* All other opcodes mean we cannot match the empty string. */ + return false; + } + + *p = p1; + return true; +} /* common_op_match_null_string_p */ + + +/* Return zero if TRANSLATE[S1] and TRANSLATE[S2] are identical for LEN + bytes; nonzero otherwise. */ + +static int +bcmp_translate (s1, s2, len, translate) + const CHAR_TYPE *s1, *s2; + register int len; + RE_TRANSLATE_TYPE translate; +{ + register const US_CHAR_TYPE *p1 = (const US_CHAR_TYPE *) s1; + register const US_CHAR_TYPE *p2 = (const US_CHAR_TYPE *) s2; + while (len) + { +#ifdef MBS_SUPPORT + if (((*p1<=0xff)?translate[*p1++]:*p1++) + != ((*p2<=0xff)?translate[*p2++]:*p2++)) + return 1; +#else + if (translate[*p1++] != translate[*p2++]) return 1; +#endif /* MBS_SUPPORT */ + len--; + } + return 0; +} + +/* Entry points for GNU code. */ + +/* re_compile_pattern is the GNU regular expression compiler: it + compiles PATTERN (of length SIZE) and puts the result in BUFP. + Returns 0 if the pattern was valid, otherwise an error string. + + Assumes the `allocated' (and perhaps `buffer') and `translate' fields + are set in BUFP on entry. + + We call regex_compile to do the actual compilation. */ + +const char * +re_compile_pattern (pattern, length, bufp) + const char *pattern; + size_t length; + struct re_pattern_buffer *bufp; +{ + reg_errcode_t ret; + + /* GNU code is written to assume at least RE_NREGS registers will be set + (and at least one extra will be -1). */ + bufp->regs_allocated = REGS_UNALLOCATED; + + /* And GNU code determines whether or not to get register information + by passing null for the REGS argument to re_match, etc., not by + setting no_sub. */ + bufp->no_sub = 0; + + /* Match anchors at newline. */ + bufp->newline_anchor = 1; + + ret = regex_compile (pattern, length, re_syntax_options, bufp); + + if (!ret) + return NULL; + return gettext (re_error_msgid + re_error_msgid_idx[(int) ret]); +} +#ifdef _LIBC +weak_alias (__re_compile_pattern, re_compile_pattern) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +#ifdef _LIBC +/* Make these definitions weak in libc, so POSIX programs can redefine + these names if they don't use our functions, and still use + regcomp/regexec below without link errors. */ +weak_function +#endif +re_comp (s) + const char *s; +{ + reg_errcode_t ret; + + if (!s) + { + if (!re_comp_buf.buffer) + return gettext ("No previous regular expression"); + return 0; + } + + if (!re_comp_buf.buffer) + { + re_comp_buf.buffer = (unsigned char *) malloc (200); + if (re_comp_buf.buffer == NULL) + return (char *) gettext (re_error_msgid + + re_error_msgid_idx[(int) REG_ESPACE]); + re_comp_buf.allocated = 200; + + re_comp_buf.fastmap = (char *) malloc (1 << BYTEWIDTH); + if (re_comp_buf.fastmap == NULL) + return (char *) gettext (re_error_msgid + + re_error_msgid_idx[(int) REG_ESPACE]); + } + + /* Since `re_exec' always passes NULL for the `regs' argument, we + don't need to initialize the pattern buffer fields which affect it. */ + + /* Match anchors at newlines. */ + re_comp_buf.newline_anchor = 1; + + ret = regex_compile (s, strlen (s), re_syntax_options, &re_comp_buf); + + if (!ret) + return NULL; + + /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */ + return (char *) gettext (re_error_msgid + re_error_msgid_idx[(int) ret]); +} + + +int +#ifdef _LIBC +weak_function +#endif +re_exec (s) + const char *s; +{ + const int len = strlen (s); + return + 0 <= re_search (&re_comp_buf, s, len, 0, len, (struct re_registers *) 0); +} + +#endif /* _REGEX_RE_COMP */ + +/* POSIX.2 functions. Don't define these for Emacs. */ + +#ifndef emacs + +/* regcomp takes a regular expression as a string and compiles it. + + PREG is a regex_t *. We do not expect any fields to be initialized, + since POSIX says we shouldn't. Thus, we set + + `buffer' to the compiled pattern; + `used' to the length of the compiled pattern; + `syntax' to RE_SYNTAX_POSIX_EXTENDED if the + REG_EXTENDED bit in CFLAGS is set; otherwise, to + RE_SYNTAX_POSIX_BASIC; + `newline_anchor' to REG_NEWLINE being set in CFLAGS; + `fastmap' to an allocated space for the fastmap; + `fastmap_accurate' to zero; + `re_nsub' to the number of subexpressions in PATTERN. + + PATTERN is the address of the pattern string. + + CFLAGS is a series of bits which affect compilation. + + If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we + use POSIX basic syntax. + + If REG_NEWLINE is set, then . and [^...] don't match newline. + Also, regexec will try a match beginning after every newline. + + If REG_ICASE is set, then we considers upper- and lowercase + versions of letters to be equivalent when matching. + + If REG_NOSUB is set, then when PREG is passed to regexec, that + routine will report only success or failure, and nothing about the + registers. + + It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for + the return codes and their meanings.) */ + +int +regcomp (preg, pattern, cflags) + regex_t *preg; + const char *pattern; + int cflags; +{ + reg_errcode_t ret; + reg_syntax_t syntax + = (cflags & REG_EXTENDED) ? + RE_SYNTAX_POSIX_EXTENDED : RE_SYNTAX_POSIX_BASIC; + + /* regex_compile will allocate the space for the compiled pattern. */ + preg->buffer = 0; + preg->allocated = 0; + preg->used = 0; + + /* Try to allocate space for the fastmap. */ + preg->fastmap = (char *) malloc (1 << BYTEWIDTH); + + if (cflags & REG_ICASE) + { + unsigned i; + + preg->translate + = (RE_TRANSLATE_TYPE) malloc (CHAR_SET_SIZE + * sizeof (*(RE_TRANSLATE_TYPE)0)); + if (preg->translate == NULL) + return (int) REG_ESPACE; + + /* Map uppercase characters to corresponding lowercase ones. */ + for (i = 0; i < CHAR_SET_SIZE; i++) + preg->translate[i] = ISUPPER (i) ? TOLOWER (i) : i; + } + else + preg->translate = NULL; + + /* If REG_NEWLINE is set, newlines are treated differently. */ + if (cflags & REG_NEWLINE) + { /* REG_NEWLINE implies neither . nor [^...] match newline. */ + syntax &= ~RE_DOT_NEWLINE; + syntax |= RE_HAT_LISTS_NOT_NEWLINE; + /* It also changes the matching behavior. */ + preg->newline_anchor = 1; + } + else + preg->newline_anchor = 0; + + preg->no_sub = !!(cflags & REG_NOSUB); + + /* POSIX says a null character in the pattern terminates it, so we + can use strlen here in compiling the pattern. */ + ret = regex_compile (pattern, strlen (pattern), syntax, preg); + + /* POSIX doesn't distinguish between an unmatched open-group and an + unmatched close-group: both are REG_EPAREN. */ + if (ret == REG_ERPAREN) ret = REG_EPAREN; + + if (ret == REG_NOERROR && preg->fastmap) + { + /* Compute the fastmap now, since regexec cannot modify the pattern + buffer. */ + if (re_compile_fastmap (preg) == -2) + { + /* Some error occurred while computing the fastmap, just forget + about it. */ + free (preg->fastmap); + preg->fastmap = NULL; + } + } + + return (int) ret; +} +#ifdef _LIBC +weak_alias (__regcomp, regcomp) +#endif + + +/* regexec searches for a given pattern, specified by PREG, in the + string STRING. + + If NMATCH is zero or REG_NOSUB was set in the cflags argument to + `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at + least NMATCH elements, and we set them to the offsets of the + corresponding matched substrings. + + EFLAGS specifies `execution flags' which affect matching: if + REG_NOTBOL is set, then ^ does not match at the beginning of the + string; if REG_NOTEOL is set, then $ does not match at the end. + + We return 0 if we find a match and REG_NOMATCH if not. */ + +int +regexec (preg, string, nmatch, pmatch, eflags) + const regex_t *preg; + const char *string; + size_t nmatch; + regmatch_t pmatch[]; + int eflags; +{ + int ret; + struct re_registers regs; + regex_t private_preg; + int len = strlen (string); + boolean want_reg_info = !preg->no_sub && nmatch > 0; + + private_preg = *preg; + + private_preg.not_bol = !!(eflags & REG_NOTBOL); + private_preg.not_eol = !!(eflags & REG_NOTEOL); + + /* The user has told us exactly how many registers to return + information about, via `nmatch'. We have to pass that on to the + matching routines. */ + private_preg.regs_allocated = REGS_FIXED; + + if (want_reg_info) + { + regs.num_regs = nmatch; + regs.start = TALLOC (nmatch * 2, regoff_t); + if (regs.start == NULL) + return (int) REG_NOMATCH; + regs.end = regs.start + nmatch; + } + + /* Perform the searching operation. */ + ret = re_search (&private_preg, string, len, + /* start: */ 0, /* range: */ len, + want_reg_info ? ®s : (struct re_registers *) 0); + + /* Copy the register information to the POSIX structure. */ + if (want_reg_info) + { + if (ret >= 0) + { + unsigned r; + + for (r = 0; r < nmatch; r++) + { + pmatch[r].rm_so = regs.start[r]; + pmatch[r].rm_eo = regs.end[r]; + } + } + + /* If we needed the temporary register info, free the space now. */ + free (regs.start); + } + + /* We want zero return to mean success, unlike `re_search'. */ + return ret >= 0 ? (int) REG_NOERROR : (int) REG_NOMATCH; +} +#ifdef _LIBC +weak_alias (__regexec, regexec) +#endif + + +/* Returns a message corresponding to an error code, ERRCODE, returned + from either regcomp or regexec. We don't use PREG here. */ + +size_t +regerror (errcode, preg, errbuf, errbuf_size) + int errcode; + const regex_t *preg; + char *errbuf; + size_t errbuf_size; +{ + const char *msg; + size_t msg_size; + + if (errcode < 0 + || errcode >= (int) (sizeof (re_error_msgid_idx) + / sizeof (re_error_msgid_idx[0]))) + /* Only error codes returned by the rest of the code should be passed + to this routine. If we are given anything else, or if other regex + code generates an invalid error code, then the program has a bug. + Dump core so we can fix it. */ + abort (); + + msg = gettext (re_error_msgid + re_error_msgid_idx[errcode]); + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (errbuf_size != 0) + { + if (msg_size > errbuf_size) + { +#if defined HAVE_MEMPCPY || defined _LIBC + *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0'; +#else + memcpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; +#endif + } + else + memcpy (errbuf, msg, msg_size); + } + + return msg_size; +} +#ifdef _LIBC +weak_alias (__regerror, regerror) +#endif + + +/* Free dynamically allocated space used by PREG. */ + +void +regfree (preg) + regex_t *preg; +{ + if (preg->buffer != NULL) + free (preg->buffer); + preg->buffer = NULL; + + preg->allocated = 0; + preg->used = 0; + + if (preg->fastmap != NULL) + free (preg->fastmap); + preg->fastmap = NULL; + preg->fastmap_accurate = 0; + + if (preg->translate != NULL) + free (preg->translate); + preg->translate = NULL; +} +#ifdef _LIBC +weak_alias (__regfree, regfree) +#endif + +#endif /* not emacs */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/regex_win32.c b/debian/htdig/htdig-3.2.0b6/htlib/regex_win32.c new file mode 100644 index 00000000..566b1fc0 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/regex_win32.c @@ -0,0 +1,5742 @@ +/* LGPLd GNU regex for Native WIN32 */ + +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 2003 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/* Added June 2003 Neal Richter, RightNow Technologies */ + +/* note that this version is significantly different from the original */ +/* version 0.12 GNU source code. It compiles and works on Native WIN32. */ + +/* Extended regular expression matching and search library, + version 0.12. + (Implements POSIX draft P1003.2/D11.2, except for some of the + internationalization features.) + + Copyright (C) 1993, 1994, 1995, 1996, 1997 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#if defined(_WIN32) +#pragma warning(disable: 4018 4101) +#endif + +/* AIX requires this to be the first thing in the file. */ +#if defined (_AIX) && !defined (REGEX_MALLOC) +#pragma alloca +#endif + +#undef _GNU_SOURCE +#define _GNU_SOURCE + +#if defined(LINUX) +#define STDC_HEADERS +#endif + +#if defined(STDC_HEADERS) && !defined(emacs) +#include <stddef.h> +#else +/* We need this for `regex.h', and perhaps for the Emacs include files. */ +#include <sys/types.h> +#endif + +/* For platform which support the ISO C amendement 1 functionality we + support user defined character classes. */ +#if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) +# include <wctype.h> +# include <wchar.h> +#endif + +/* This is for other GNU distributions with internationalized messages. */ +#if HAVE_LIBINTL_H || defined (_LIBC) +# include <libintl.h> +#else +# define gettext(msgid) (msgid) +#endif + +#ifndef gettext_noop +/* This define is so xgettext can find the internationalizable + strings. */ +#define gettext_noop(String) String +#endif + +/* The `emacs' switch turns on certain matching commands + that make sense only in Emacs. */ +#ifdef emacs + +#include "lisp.h" +#include "buffer.h" +#include "syntax.h" + +#else /* not emacs */ + +/* If we are not linking with Emacs proper, + we can't use the relocating allocator + even if config.h says that we can. */ +#undef REL_ALLOC + +#if defined (STDC_HEADERS) || defined (_LIBC) || defined(_WIN32) +#include <stdlib.h> +#else +char *malloc (); +char *realloc (); +void free(); +#endif + +/* When used in Emacs's lib-src, we need to get bzero and bcopy somehow. + If nothing else has been done, use the method below. */ +#ifdef INHIBIT_STRING_HEADER +#if !(defined (HAVE_BZERO) && defined (HAVE_BCOPY)) +#if !defined (bzero) && !defined (bcopy) +#undef INHIBIT_STRING_HEADER +#endif +#endif +#endif + +#include <string.h> + +/* This is the normal way of making sure we have a bcopy and a bzero. + This is used in most programs--a few other programs avoid this + by defining INHIBIT_STRING_HEADER. */ +#ifndef INHIBIT_STRING_HEADER +#if defined (HAVE_STRING_H) || defined (STDC_HEADERS) || defined (_LIBC) || defined (_WIN32) +#ifndef bcmp +#define bcmp(s1, s2, n) memcmp ((s1), (s2), (n)) +#endif +#ifndef bcopy +#define bcopy(s, d, n) memcpy ((d), (s), (n)) +#endif +#ifndef bzero +#define bzero(s, n) memset ((s), 0, (n)) +#endif +#else +#include <strings.h> +#endif +#endif + +/* Define the syntax stuff for \<, \>, etc. */ + +/* This must be nonzero for the wordchar and notwordchar pattern + commands in re_match_2. */ +#ifndef Sword +#define Sword 1 +#endif + +#ifdef SWITCH_ENUM_BUG +#define SWITCH_ENUM_CAST(x) ((int)(x)) +#else +#define SWITCH_ENUM_CAST(x) (x) +#endif + +#ifdef SYNTAX_TABLE + +extern char *re_syntax_table; + +#else /* not SYNTAX_TABLE */ + +/* How many characters in the character set. */ +#define CHAR_SET_SIZE 256 + +static char re_syntax_table[CHAR_SET_SIZE]; + +static void +init_syntax_once () +{ + register int c; + static int done = 0; + + if (done) + return; + + bzero (re_syntax_table, sizeof re_syntax_table); + + for (c = 'a'; c <= 'z'; c++) + re_syntax_table[c] = Sword; + + for (c = 'A'; c <= 'Z'; c++) + re_syntax_table[c] = Sword; + + for (c = '0'; c <= '9'; c++) + re_syntax_table[c] = Sword; + + re_syntax_table['_'] = Sword; + + done = 1; +} + +#endif /* not SYNTAX_TABLE */ + +#define SYNTAX(c) re_syntax_table[c] + +#endif /* not emacs */ + +/* Get the interface, including the syntax bits. */ +/* #include "regex.h" */ +#include "regex_win32.h" + +/* isalpha etc. are used for the character classes. */ +#include <ctype.h> + +/* Jim Meyering writes: + + "... Some ctype macros are valid only for character codes that + isascii says are ASCII (SGI's IRIX-4.0.5 is one such system --when + using /bin/cc or gcc but without giving an ansi option). So, all + ctype uses should be through macros like ISPRINT... If + STDC_HEADERS is defined, then autoconf has verified that the ctype + macros don't need to be guarded with references to isascii. ... + Defining isascii to 1 should let any compiler worth its salt + eliminate the && through constant folding." */ + +#if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII)) +#define ISASCII(c) 1 +#else +#define ISASCII(c) isascii(c) +#endif + +#ifdef isblank +#define ISBLANK(c) (ISASCII (c) && isblank (c)) +#else +#define ISBLANK(c) ((c) == ' ' || (c) == '\t') +#endif +#ifdef isgraph +#define ISGRAPH(c) (ISASCII (c) && isgraph (c)) +#else +#define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c)) +#endif + +#define ISPRINT(c) (ISASCII (c) && isprint (c)) +#define ISDIGIT(c) (ISASCII (c) && isdigit (c)) +#define ISALNUM(c) (ISASCII (c) && isalnum (c)) +#define ISALPHA(c) (ISASCII (c) && isalpha (c)) +#define ISCNTRL(c) (ISASCII (c) && iscntrl (c)) +#define ISLOWER(c) (ISASCII (c) && islower (c)) +#define ISPUNCT(c) (ISASCII (c) && ispunct (c)) +#define ISSPACE(c) (ISASCII (c) && isspace (c)) +#define ISUPPER(c) (ISASCII (c) && isupper (c)) +#define ISXDIGIT(c) (ISASCII (c) && isxdigit (c)) + +#ifndef NULL +#define NULL (void *)0 +#endif + +/* We remove any previous definition of `SIGN_EXTEND_CHAR', + since ours (we hope) works properly with all combinations of + machines, compilers, `char' and `unsigned char' argument types. + (Per Bothner suggested the basic approach.) */ +#undef SIGN_EXTEND_CHAR +#if __STDC__ +#define SIGN_EXTEND_CHAR(c) ((signed char) (c)) +#else /* not __STDC__ */ +/* As in Harbison and Steele. */ +#define SIGN_EXTEND_CHAR(c) ((((unsigned char) (c)) ^ 128) - 128) +#endif + +/* Should we use malloc or alloca? If REGEX_MALLOC is not defined, we + use `alloca' instead of `malloc'. This is because using malloc in + re_search* or re_match* could cause memory leaks when C-g is used in + Emacs; also, malloc is slower and causes storage fragmentation. On + the other hand, malloc is more portable, and easier to debug. + + Because we sometimes use alloca, some routines have to be macros, + not functions -- `alloca'-allocated space disappears at the end of the + function it is called in. */ + +#if defined(REGEX_MALLOC) || defined(_WIN32) + +#define REGEX_ALLOCATE malloc +#define REGEX_REALLOCATE(source, osize, nsize) realloc (source, nsize) +#define REGEX_FREE free +#define REGEX_MALLOC + +#else /* not REGEX_MALLOC */ + +/* Emacs already defines alloca, sometimes. */ +#ifndef alloca + +/* Make alloca work the best possible way. */ +#ifdef __GNUC__ +#define alloca __builtin_alloca +#else /* not __GNUC__ */ +#if HAVE_ALLOCA_H +#include <alloca.h> +#else /* not __GNUC__ or HAVE_ALLOCA_H */ +#if 0 /* It is a bad idea to declare alloca. We always cast the result. */ +#ifndef _AIX /* Already did AIX, up at the top. */ +char *alloca (); +#endif /* not _AIX */ +#endif +#endif /* not HAVE_ALLOCA_H */ +#endif /* not __GNUC__ */ + +#endif /* not alloca */ + +#define REGEX_ALLOCATE alloca + +/* Assumes a `char *destination' variable. */ +#define REGEX_REALLOCATE(source, osize, nsize) \ + (destination = (char *) alloca (nsize), \ + bcopy (source, destination, osize), \ + destination) + +/* No need to do anything to free, after alloca. */ +#define REGEX_FREE(arg) ((void)0) /* Do nothing! But inhibit gcc warning. */ + +#endif /* not REGEX_MALLOC */ + +/* Define how to allocate the failure stack. */ + +#if defined (REL_ALLOC) && defined (REGEX_MALLOC) + +#define REGEX_ALLOCATE_STACK(size) \ + r_alloc (&failure_stack_ptr, (size)) +#define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + r_re_alloc (&failure_stack_ptr, (nsize)) +#define REGEX_FREE_STACK(ptr) \ + r_alloc_free (&failure_stack_ptr) + +#else /* not using relocating allocator */ + +#ifdef REGEX_MALLOC + +#define REGEX_ALLOCATE_STACK malloc +#define REGEX_REALLOCATE_STACK(source, osize, nsize) realloc (source, nsize) +#define REGEX_FREE_STACK free + +#else /* not REGEX_MALLOC */ + +#define REGEX_ALLOCATE_STACK alloca + +#define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + REGEX_REALLOCATE (source, osize, nsize) +/* No need to explicitly free anything. */ +#define REGEX_FREE_STACK(arg) + +#endif /* not REGEX_MALLOC */ +#endif /* not using relocating allocator */ + + +/* True if `size1' is non-NULL and PTR is pointing anywhere inside + `string1' or just past its end. This works if PTR is NULL, which is + a good thing. */ +#define FIRST_STRING_P(ptr) \ + (size1 && string1 <= (ptr) && (ptr) <= string1 + size1) + +/* (Re)Allocate N items of type T using malloc, or fail. */ +#define TALLOC(n, t) ((t *) malloc ((n) * sizeof (t))) +#define RETALLOC(addr, n, t) ((addr) = (t *) realloc (addr, (n) * sizeof (t))) +#define RETALLOC_IF(addr, n, t) \ + if (addr) RETALLOC((addr), (n), t); else (addr) = TALLOC ((n), t) +#define REGEX_TALLOC(n, t) ((t *) REGEX_ALLOCATE ((n) * sizeof (t))) + +#define BYTEWIDTH 8 /* In bits. */ + +#define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) + +#undef MAX +#undef MIN +#define MAX(a, b) ((a) > (b) ? (a) : (b)) +#define MIN(a, b) ((a) < (b) ? (a) : (b)) + +typedef char boolean; +#define false 0 +#define true 1 + +static int +re_match_2_internal(struct re_pattern_buffer *bufp, + const char *string1, + int size1, + const char *string2, + int size2, + int pos, + struct re_registers *regs, + int stop); + +/* These are the command codes that appear in compiled regular + expressions. Some opcodes are followed by argument bytes. A + command code can specify any interpretation whatsoever for its + arguments. Zero bytes may appear in the compiled regular expression. */ + +typedef enum +{ + no_op = 0, + + /* Succeed right away--no more backtracking. */ + succeed, + + /* Followed by one byte giving n, then by n literal bytes. */ + exactn, + + /* Matches any (more or less) character. */ + anychar, + + /* Matches any one char belonging to specified set. First + following byte is number of bitmap bytes. Then come bytes + for a bitmap saying which chars are in. Bits in each byte + are ordered low-bit-first. A character is in the set if its + bit is 1. A character too large to have a bit in the map is + automatically not in the set. */ + charset, + + /* Same parameters as charset, but match any character that is + not one of those specified. */ + charset_not, + + /* Start remembering the text that is matched, for storing in a + register. Followed by one byte with the register number, in + the range 0 to one less than the pattern buffer's re_nsub + field. Then followed by one byte with the number of groups + inner to this one. (This last has to be part of the + start_memory only because we need it in the on_failure_jump + of re_match_2.) */ + start_memory, + + /* Stop remembering the text that is matched and store it in a + memory register. Followed by one byte with the register + number, in the range 0 to one less than `re_nsub' in the + pattern buffer, and one byte with the number of inner groups, + just like `start_memory'. (We need the number of inner + groups here because we don't have any easy way of finding the + corresponding start_memory when we're at a stop_memory.) */ + stop_memory, + + /* Match a duplicate of something remembered. Followed by one + byte containing the register number. */ + duplicate, + + /* Fail unless at beginning of line. */ + begline, + + /* Fail unless at end of line. */ + endline, + + /* Succeeds if at beginning of buffer (if emacs) or at beginning + of string to be matched (if not). */ + begbuf, + + /* Analogously, for end of buffer/string. */ + endbuf, + + /* Followed by two byte relative address to which to jump. */ + jump, + + /* Same as jump, but marks the end of an alternative. */ + jump_past_alt, + + /* Followed by two-byte relative address of place to resume at + in case of failure. */ + on_failure_jump, + + /* Like on_failure_jump, but pushes a placeholder instead of the + current string position when executed. */ + on_failure_keep_string_jump, + + /* Throw away latest failure point and then jump to following + two-byte relative address. */ + pop_failure_jump, + + /* Change to pop_failure_jump if know won't have to backtrack to + match; otherwise change to jump. This is used to jump + back to the beginning of a repeat. If what follows this jump + clearly won't match what the repeat does, such that we can be + sure that there is no use backtracking out of repetitions + already matched, then we change it to a pop_failure_jump. + Followed by two-byte address. */ + maybe_pop_jump, + + /* Jump to following two-byte address, and push a dummy failure + point. This failure point will be thrown away if an attempt + is made to use it for a failure. A `+' construct makes this + before the first repeat. Also used as an intermediary kind + of jump when compiling an alternative. */ + dummy_failure_jump, + + /* Push a dummy failure point and continue. Used at the end of + alternatives. */ + push_dummy_failure, + + /* Followed by two-byte relative address and two-byte number n. + After matching N times, jump to the address upon failure. */ + succeed_n, + + /* Followed by two-byte relative address, and two-byte number n. + Jump to the address N times, then fail. */ + jump_n, + + /* Set the following two-byte relative address to the + subsequent two-byte number. The address *includes* the two + bytes of number. */ + set_number_at, + + wordchar, /* Matches any word-constituent character. */ + notwordchar, /* Matches any char that is not a word-constituent. */ + + wordbeg, /* Succeeds if at word beginning. */ + wordend, /* Succeeds if at word end. */ + + wordbound, /* Succeeds if at a word boundary. */ + notwordbound /* Succeeds if not at a word boundary. */ + +#ifdef emacs + ,before_dot, /* Succeeds if before point. */ + at_dot, /* Succeeds if at point. */ + after_dot, /* Succeeds if after point. */ + + /* Matches any character whose syntax is specified. Followed by + a byte which contains a syntax code, e.g., Sword. */ + syntaxspec, + + /* Matches any character whose syntax is not that specified. */ + notsyntaxspec +#endif /* emacs */ +} re_opcode_t; + +/* Common operations on the compiled pattern. */ + +/* Store NUMBER in two contiguous bytes starting at DESTINATION. */ + +#define STORE_NUMBER(destination, number) \ + do { \ + (destination)[0] = (number) & 0377; \ + (destination)[1] = (number) >> 8; \ + } while (0) + +/* Same as STORE_NUMBER, except increment DESTINATION to + the byte after where the number is stored. Therefore, DESTINATION + must be an lvalue. */ + +#define STORE_NUMBER_AND_INCR(destination, number) \ + do { \ + STORE_NUMBER (destination, number); \ + (destination) += 2; \ + } while (0) + +/* Put into DESTINATION a number stored in two contiguous bytes starting + at SOURCE. */ + +#define EXTRACT_NUMBER(destination, source) \ + do { \ + (destination) = *(source) & 0377; \ + (destination) += SIGN_EXTEND_CHAR (*((source) + 1)) << 8; \ + } while (0) + +#ifdef DEBUG +static void extract_number _RE_ARGS ((int *dest, unsigned char *source)); +static void +extract_number (dest, source) + int *dest; + unsigned char *source; +{ + int temp = SIGN_EXTEND_CHAR (*(source + 1)); + *dest = *source & 0377; + *dest += temp << 8; +} + +#ifndef EXTRACT_MACROS /* To debug the macros. */ +#undef EXTRACT_NUMBER +#define EXTRACT_NUMBER(dest, src) extract_number (&dest, src) +#endif /* not EXTRACT_MACROS */ + +#endif /* DEBUG */ + +/* Same as EXTRACT_NUMBER, except increment SOURCE to after the number. + SOURCE must be an lvalue. */ + +#define EXTRACT_NUMBER_AND_INCR(destination, source) \ + do { \ + EXTRACT_NUMBER (destination, source); \ + (source) += 2; \ + } while (0) + +#ifdef DEBUG +static void extract_number_and_incr _RE_ARGS ((int *destination, + unsigned char **source)); +static void +extract_number_and_incr (destination, source) + int *destination; + unsigned char **source; +{ + extract_number (destination, *source); + *source += 2; +} + +#ifndef EXTRACT_MACROS +#undef EXTRACT_NUMBER_AND_INCR +#define EXTRACT_NUMBER_AND_INCR(dest, src) \ + extract_number_and_incr (&dest, &src) +#endif /* not EXTRACT_MACROS */ + +#endif /* DEBUG */ + +/* If DEBUG is defined, Regex prints many voluminous messages about what + it is doing (if the variable `debug' is nonzero). If linked with the + main program in `iregex.c', you can enter patterns and strings + interactively. And if linked with the main program in `main.c' and + the other test files, you can run the already-written tests. */ + +#ifdef DEBUG + +/* We use standard I/O for debugging. */ +#include <stdio.h> + +/* It is useful to test things that ``must'' be true when debugging. */ +#include <assert.h> + +static int debug = 0; + +#define DEBUG_STATEMENT(e) e +#define DEBUG_PRINT1(x) if (debug) printf (x) +#define DEBUG_PRINT2(x1, x2) if (debug) printf (x1, x2) +#define DEBUG_PRINT3(x1, x2, x3) if (debug) printf (x1, x2, x3) +#define DEBUG_PRINT4(x1, x2, x3, x4) if (debug) printf (x1, x2, x3, x4) +#define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) \ + if (debug) print_partial_compiled_pattern (s, e) +#define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) \ + if (debug) print_double_string (w, s1, sz1, s2, sz2) + + +/* Print the fastmap in human-readable form. */ + +void +print_fastmap (fastmap) + char *fastmap; +{ + unsigned was_a_range = 0; + unsigned i = 0; + + while (i < (1 << BYTEWIDTH)) + { + if (fastmap[i++]) + { + was_a_range = 0; + putchar (i - 1); + while (i < (1 << BYTEWIDTH) && fastmap[i]) + { + was_a_range = 1; + i++; + } + if (was_a_range) + { + printf ("-"); + putchar (i - 1); + } + } + } + putchar ('\n'); +} + + +/* Print a compiled pattern string in human-readable form, starting at + the START pointer into it and ending just before the pointer END. */ + +void +print_partial_compiled_pattern (start, end) + unsigned char *start; + unsigned char *end; +{ + int mcnt, mcnt2; + unsigned char *p1; + unsigned char *p = start; + unsigned char *pend = end; + + if (start == NULL) + { + printf ("(null)\n"); + return; + } + + /* Loop over pattern commands. */ + while (p < pend) + { + printf ("%d:\t", p - start); + + switch ((re_opcode_t) *p++) + { + case no_op: + printf ("/no_op"); + break; + + case exactn: + mcnt = *p++; + printf ("/exactn/%d", mcnt); + do + { + putchar ('/'); + putchar (*p++); + } + while (--mcnt); + break; + + case start_memory: + mcnt = *p++; + printf ("/start_memory/%d/%d", mcnt, *p++); + break; + + case stop_memory: + mcnt = *p++; + printf ("/stop_memory/%d/%d", mcnt, *p++); + break; + + case duplicate: + printf ("/duplicate/%d", *p++); + break; + + case anychar: + printf ("/anychar"); + break; + + case charset: + case charset_not: + { + register int c, last = -100; + register int in_range = 0; + + printf ("/charset [%s", + (re_opcode_t) *(p - 1) == charset_not ? "^" : ""); + + assert (p + *p < pend); + + for (c = 0; c < 256; c++) + if (c / 8 < *p + && (p[1 + (c/8)] & (1 << (c % 8)))) + { + /* Are we starting a range? */ + if (last + 1 == c && ! in_range) + { + putchar ('-'); + in_range = 1; + } + /* Have we broken a range? */ + else if (last + 1 != c && in_range) + { + putchar (last); + in_range = 0; + } + + if (! in_range) + putchar (c); + + last = c; + } + + if (in_range) + putchar (last); + + putchar (']'); + + p += 1 + *p; + } + break; + + case begline: + printf ("/begline"); + break; + + case endline: + printf ("/endline"); + break; + + case on_failure_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/on_failure_jump to %d", p + mcnt - start); + break; + + case on_failure_keep_string_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/on_failure_keep_string_jump to %d", p + mcnt - start); + break; + + case dummy_failure_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/dummy_failure_jump to %d", p + mcnt - start); + break; + + case push_dummy_failure: + printf ("/push_dummy_failure"); + break; + + case maybe_pop_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/maybe_pop_jump to %d", p + mcnt - start); + break; + + case pop_failure_jump: + extract_number_and_incr (&mcnt, &p); + printf ("/pop_failure_jump to %d", p + mcnt - start); + break; + + case jump_past_alt: + extract_number_and_incr (&mcnt, &p); + printf ("/jump_past_alt to %d", p + mcnt - start); + break; + + case jump: + extract_number_and_incr (&mcnt, &p); + printf ("/jump to %d", p + mcnt - start); + break; + + case succeed_n: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/succeed_n to %d, %d times", p1 - start, mcnt2); + break; + + case jump_n: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/jump_n to %d, %d times", p1 - start, mcnt2); + break; + + case set_number_at: + extract_number_and_incr (&mcnt, &p); + p1 = p + mcnt; + extract_number_and_incr (&mcnt2, &p); + printf ("/set_number_at location %d to %d", p1 - start, mcnt2); + break; + + case wordbound: + printf ("/wordbound"); + break; + + case notwordbound: + printf ("/notwordbound"); + break; + + case wordbeg: + printf ("/wordbeg"); + break; + + case wordend: + printf ("/wordend"); + +#ifdef emacs + case before_dot: + printf ("/before_dot"); + break; + + case at_dot: + printf ("/at_dot"); + break; + + case after_dot: + printf ("/after_dot"); + break; + + case syntaxspec: + printf ("/syntaxspec"); + mcnt = *p++; + printf ("/%d", mcnt); + break; + + case notsyntaxspec: + printf ("/notsyntaxspec"); + mcnt = *p++; + printf ("/%d", mcnt); + break; +#endif /* emacs */ + + case wordchar: + printf ("/wordchar"); + break; + + case notwordchar: + printf ("/notwordchar"); + break; + + case begbuf: + printf ("/begbuf"); + break; + + case endbuf: + printf ("/endbuf"); + break; + + default: + printf ("?%d", *(p-1)); + } + + putchar ('\n'); + } + + printf ("%d:\tend of pattern.\n", p - start); +} + + +void +print_compiled_pattern (bufp) + struct re_pattern_buffer *bufp; +{ + unsigned char *buffer = bufp->buffer; + + print_partial_compiled_pattern (buffer, buffer + bufp->used); + printf ("%ld bytes used/%ld bytes allocated.\n", + bufp->used, bufp->allocated); + + if (bufp->fastmap_accurate && bufp->fastmap) + { + printf ("fastmap: "); + print_fastmap (bufp->fastmap); + } + + printf ("re_nsub: %d\t", bufp->re_nsub); + printf ("regs_alloc: %d\t", bufp->regs_allocated); + printf ("can_be_null: %d\t", bufp->can_be_null); + printf ("newline_anchor: %d\n", bufp->newline_anchor); + printf ("no_sub: %d\t", bufp->no_sub); + printf ("not_bol: %d\t", bufp->not_bol); + printf ("not_eol: %d\t", bufp->not_eol); + printf ("syntax: %lx\n", bufp->syntax); + /* Perhaps we should print the translate table? */ +} + + +void +print_double_string (where, string1, size1, string2, size2) + const char *where; + const char *string1; + const char *string2; + int size1; + int size2; +{ + int this_char; + + if (where == NULL) + printf ("(null)"); + else + { + if (FIRST_STRING_P (where)) + { + for (this_char = where - string1; this_char < size1; this_char++) + putchar (string1[this_char]); + + where = string2; + } + + for (this_char = where - string2; this_char < size2; this_char++) + putchar (string2[this_char]); + } +} + +void +printchar (c) + int c; +{ + putc (c, stderr); +} + +#else /* not DEBUG */ + +#undef assert +#define assert(e) + +#define DEBUG_STATEMENT(e) +#define DEBUG_PRINT1(x) +#define DEBUG_PRINT2(x1, x2) +#define DEBUG_PRINT3(x1, x2, x3) +#define DEBUG_PRINT4(x1, x2, x3, x4) +#define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) +#define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) + +#endif /* not DEBUG */ + +/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can + also be assigned to arbitrarily: each pattern buffer stores its own + syntax, so it can be changed between regex compilations. */ +/* This has no initializer because initialized variables in Emacs + become read-only after dumping. */ +reg_syntax_t re_syntax_options; + + +/* Specify the precise syntax of regexps for compilation. This provides + for compatibility for various utilities which historically have + different, incompatible syntaxes. + + The argument SYNTAX is a bit mask comprised of the various bits + defined in regex.h. We return the old syntax. */ + +reg_syntax_t +re_set_syntax(reg_syntax_t syntax) +{ + reg_syntax_t ret = re_syntax_options; + + re_syntax_options = syntax; +#ifdef DEBUG + if (syntax & RE_DEBUG) + debug = 1; + else if (debug) /* was on but now is not */ + debug = 0; +#endif /* DEBUG */ + return ret; +} + +/* This table gives an error message for each of the error codes listed + in regex.h. Obviously the order here has to be same as there. + POSIX doesn't require that we do anything for REG_NOERROR, + but why not be nice? */ + +static const char *re_error_msgid[] = +{ + gettext_noop ("Success"), /* REG_NOERROR */ + gettext_noop ("No match"), /* REG_NOMATCH */ + gettext_noop ("Invalid regular expression"), /* REG_BADPAT */ + gettext_noop ("Invalid collation character"), /* REG_ECOLLATE */ + gettext_noop ("Invalid character class name"), /* REG_ECTYPE */ + gettext_noop ("Trailing backslash"), /* REG_EESCAPE */ + gettext_noop ("Invalid back reference"), /* REG_ESUBREG */ + gettext_noop ("Unmatched [ or [^"), /* REG_EBRACK */ + gettext_noop ("Unmatched ( or \\("), /* REG_EPAREN */ + gettext_noop ("Unmatched \\{"), /* REG_EBRACE */ + gettext_noop ("Invalid content of \\{\\}"), /* REG_BADBR */ + gettext_noop ("Invalid range end"), /* REG_ERANGE */ + gettext_noop ("Memory exhausted"), /* REG_ESPACE */ + gettext_noop ("Invalid preceding regular expression"), /* REG_BADRPT */ + gettext_noop ("Premature end of regular expression"), /* REG_EEND */ + gettext_noop ("Regular expression too big"), /* REG_ESIZE */ + gettext_noop ("Unmatched ) or \\)"), /* REG_ERPAREN */ +}; + +/* Avoiding alloca during matching, to placate r_alloc. */ + +/* Define MATCH_MAY_ALLOCATE unless we need to make sure that the + searching and matching functions should not call alloca. On some + systems, alloca is implemented in terms of malloc, and if we're + using the relocating allocator routines, then malloc could cause a + relocation, which might (if the strings being searched are in the + ralloc heap) shift the data out from underneath the regexp + routines. + + Here's another reason to avoid allocation: Emacs + processes input from X in a signal handler; processing X input may + call malloc; if input arrives while a matching routine is calling + malloc, then we're scrod. But Emacs can't just block input while + calling matching routines; then we don't notice interrupts when + they come in. So, Emacs blocks input around all regexp calls + except the matching calls, which it leaves unprotected, in the + faith that they will not malloc. */ + +/* Normally, this is fine. */ +#define MATCH_MAY_ALLOCATE + +/* When using GNU C, we are not REALLY using the C alloca, no matter + what config.h may say. So don't take precautions for it. */ +#ifdef __GNUC__ +#undef C_ALLOCA +#endif + +/* The match routines may not allocate if (1) they would do it with malloc + and (2) it's not safe for them to use malloc. + Note that if REL_ALLOC is defined, matching would not use malloc for the + failure stack, but we would still use it for the register vectors; + so REL_ALLOC should not affect this. */ +#if (defined (C_ALLOCA) || defined (REGEX_MALLOC)) && defined (emacs) +#undef MATCH_MAY_ALLOCATE +#endif + + +/* Failure stack declarations and macros; both re_compile_fastmap and + re_match_2 use a failure stack. These have to be macros because of + REGEX_ALLOCATE_STACK. */ + + +/* Number of failure points for which to initially allocate space + when matching. If this number is exceeded, we allocate more + space, so it is not a hard limit. */ +#ifndef INIT_FAILURE_ALLOC +#define INIT_FAILURE_ALLOC 5 +#endif + +/* Roughly the maximum number of failure points on the stack. Would be + exactly that if always used MAX_FAILURE_ITEMS items each time we failed. + This is a variable only so users of regex can assign to it; we never + change it ourselves. */ + +#ifdef INT_IS_16BIT + +#if defined (MATCH_MAY_ALLOCATE) +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +static long int re_max_failures = 4000; +#else +static long int re_max_failures = 2000; +#endif + +union fail_stack_elt +{ + unsigned char *pointer; + long int integer; +}; + +typedef union fail_stack_elt fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned long int size; + unsigned long int avail; /* Offset of next open position. */ +} fail_stack_type; + +#else /* not INT_IS_16BIT */ + +#if defined (MATCH_MAY_ALLOCATE) +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +static int re_max_failures = 20000; +#else +static int re_max_failures = 2000; +#endif + +union fail_stack_elt +{ + unsigned char *pointer; + int integer; +}; + +typedef union fail_stack_elt fail_stack_elt_t; + +typedef struct +{ + fail_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} fail_stack_type; + +#endif /* INT_IS_16BIT */ + +#define FAIL_STACK_EMPTY() (fail_stack.avail == 0) +#define FAIL_STACK_PTR_EMPTY() (fail_stack_ptr->avail == 0) +#define FAIL_STACK_FULL() (fail_stack.avail == fail_stack.size) + + +/* Define macros to initialize and free the failure stack. + Do `return -2' if the alloc fails. */ + +#ifdef MATCH_MAY_ALLOCATE +#define INIT_FAIL_STACK() \ + do { \ + fail_stack.stack = (fail_stack_elt_t *) \ + REGEX_ALLOCATE_STACK (INIT_FAILURE_ALLOC * sizeof (fail_stack_elt_t)); \ + \ + if (fail_stack.stack == NULL) \ + return -2; \ + \ + fail_stack.size = INIT_FAILURE_ALLOC; \ + fail_stack.avail = 0; \ + } while (0) + +#define RESET_FAIL_STACK() REGEX_FREE_STACK (fail_stack.stack) +#else +#define INIT_FAIL_STACK() \ + do { \ + fail_stack.avail = 0; \ + } while (0) + +#define RESET_FAIL_STACK() +#endif + + +/* Double the size of FAIL_STACK, up to approximately `re_max_failures' items. + + Return 1 if succeeds, and 0 if either ran out of memory + allocating space for it or it was already too large. + + REGEX_REALLOCATE_STACK requires `destination' be declared. */ + +#define DOUBLE_FAIL_STACK(fail_stack) \ + ((fail_stack).size > (unsigned) (re_max_failures * MAX_FAILURE_ITEMS) \ + ? 0 \ + : ((fail_stack).stack = (fail_stack_elt_t *) \ + REGEX_REALLOCATE_STACK ((fail_stack).stack, \ + (fail_stack).size * sizeof (fail_stack_elt_t), \ + ((fail_stack).size << 1) * sizeof (fail_stack_elt_t)), \ + \ + (fail_stack).stack == NULL \ + ? 0 \ + : ((fail_stack).size <<= 1, \ + 1))) + + +/* Push pointer POINTER on FAIL_STACK. + Return 1 if was able to do so and 0 if ran out of memory allocating + space to do so. */ +#define PUSH_PATTERN_OP(POINTER, FAIL_STACK) \ + ((FAIL_STACK_FULL () \ + && !DOUBLE_FAIL_STACK (FAIL_STACK)) \ + ? 0 \ + : ((FAIL_STACK).stack[(FAIL_STACK).avail++].pointer = POINTER, \ + 1)) + +/* Push a pointer value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_POINTER(item) \ + fail_stack.stack[fail_stack.avail++].pointer = (unsigned char *) (item) + +/* This pushes an integer-valued item onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_INT(item) \ + fail_stack.stack[fail_stack.avail++].integer = (item) + +/* Push a fail_stack_elt_t value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +#define PUSH_FAILURE_ELT(item) \ + fail_stack.stack[fail_stack.avail++] = (item) + +/* These three POP... operations complement the three PUSH... operations. + All assume that `fail_stack' is nonempty. */ +#define POP_FAILURE_POINTER() fail_stack.stack[--fail_stack.avail].pointer +#define POP_FAILURE_INT() fail_stack.stack[--fail_stack.avail].integer +#define POP_FAILURE_ELT() fail_stack.stack[--fail_stack.avail] + +/* Used to omit pushing failure point id's when we're not debugging. */ +#ifdef DEBUG +#define DEBUG_PUSH PUSH_FAILURE_INT +#define DEBUG_POP(item_addr) (item_addr)->integer = POP_FAILURE_INT () +#else +#define DEBUG_PUSH(item) +#define DEBUG_POP(item_addr) +#endif + + +/* Push the information about the state we will need + if we ever fail back to it. + + Requires variables fail_stack, regstart, regend, reg_info, and + num_regs be declared. DOUBLE_FAIL_STACK requires `destination' be + declared. + + Does `return FAILURE_CODE' if runs out of memory. */ + +#define PUSH_FAILURE_POINT(pattern_place, string_place, failure_code) \ + do { \ + char *destination; \ + /* Must be int, so when we don't save any registers, the arithmetic \ + of 0 + -1 isn't done as unsigned. */ \ + /* Can't be int, since there is not a shred of a guarantee that int \ + is wide enough to hold a value of something to which pointer can \ + be assigned */ \ + s_reg_t this_reg; \ + \ + DEBUG_STATEMENT (failure_id++); \ + DEBUG_STATEMENT (nfailure_points_pushed++); \ + DEBUG_PRINT2 ("\nPUSH_FAILURE_POINT #%u:\n", failure_id); \ + DEBUG_PRINT2 (" Before push, next avail: %d\n", (fail_stack).avail);\ + DEBUG_PRINT2 (" size: %d\n", (fail_stack).size);\ + \ + DEBUG_PRINT2 (" slots needed: %d\n", NUM_FAILURE_ITEMS); \ + DEBUG_PRINT2 (" available: %d\n", REMAINING_AVAIL_SLOTS); \ + \ + /* Ensure we have enough space allocated for what we will push. */ \ + while (REMAINING_AVAIL_SLOTS < NUM_FAILURE_ITEMS) \ + { \ + if (!DOUBLE_FAIL_STACK (fail_stack)) \ + return failure_code; \ + \ + DEBUG_PRINT2 ("\n Doubled stack; size now: %d\n", \ + (fail_stack).size); \ + DEBUG_PRINT2 (" slots available: %d\n", REMAINING_AVAIL_SLOTS);\ + } \ + \ + /* Push the info, starting with the registers. */ \ + DEBUG_PRINT1 ("\n"); \ + \ + if (1) \ + for (this_reg = lowest_active_reg; this_reg <= highest_active_reg; \ + this_reg++) \ + { \ + DEBUG_PRINT2 (" Pushing reg: %d\n", this_reg); \ + DEBUG_STATEMENT (num_regs_pushed++); \ + \ + DEBUG_PRINT2 (" start: 0x%x\n", regstart[this_reg]); \ + PUSH_FAILURE_POINTER (regstart[this_reg]); \ + \ + DEBUG_PRINT2 (" end: 0x%x\n", regend[this_reg]); \ + PUSH_FAILURE_POINTER (regend[this_reg]); \ + \ + DEBUG_PRINT2 (" info: 0x%x\n ", reg_info[this_reg]); \ + DEBUG_PRINT2 (" match_null=%d", \ + REG_MATCH_NULL_STRING_P (reg_info[this_reg])); \ + DEBUG_PRINT2 (" active=%d", IS_ACTIVE (reg_info[this_reg])); \ + DEBUG_PRINT2 (" matched_something=%d", \ + MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT2 (" ever_matched=%d", \ + EVER_MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT1 ("\n"); \ + PUSH_FAILURE_ELT (reg_info[this_reg].word); \ + } \ + \ + DEBUG_PRINT2 (" Pushing low active reg: %d\n", lowest_active_reg);\ + PUSH_FAILURE_INT (lowest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing high active reg: %d\n", highest_active_reg);\ + PUSH_FAILURE_INT (highest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing pattern 0x%x:\n", pattern_place); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pattern_place, pend); \ + PUSH_FAILURE_POINTER (pattern_place); \ + \ + DEBUG_PRINT2 (" Pushing string 0x%x: `", string_place); \ + DEBUG_PRINT_DOUBLE_STRING (string_place, string1, size1, string2, \ + size2); \ + DEBUG_PRINT1 ("'\n"); \ + PUSH_FAILURE_POINTER (string_place); \ + \ + DEBUG_PRINT2 (" Pushing failure id: %u\n", failure_id); \ + DEBUG_PUSH (failure_id); \ + } while (0) + +/* This is the number of items that are pushed and popped on the stack + for each register. */ +#define NUM_REG_ITEMS 3 + +/* Individual items aside from the registers. */ +#ifdef DEBUG +#define NUM_NONREG_ITEMS 5 /* Includes failure point id. */ +#else +#define NUM_NONREG_ITEMS 4 +#endif + +/* We push at most this many items on the stack. */ +/* We used to use (num_regs - 1), which is the number of registers + this regexp will save; but that was changed to 5 + to avoid stack overflow for a regexp with lots of parens. */ +#define MAX_FAILURE_ITEMS (5 * NUM_REG_ITEMS + NUM_NONREG_ITEMS) + +/* We actually push this many items. */ +#define NUM_FAILURE_ITEMS \ + (((0 \ + ? 0 : highest_active_reg - lowest_active_reg + 1) \ + * NUM_REG_ITEMS) \ + + NUM_NONREG_ITEMS) + +/* How many items can still be added to the stack without overflowing it. */ +#define REMAINING_AVAIL_SLOTS ((fail_stack).size - (fail_stack).avail) + + +/* Pops what PUSH_FAIL_STACK pushes. + + We restore into the parameters, all of which should be lvalues: + STR -- the saved data position. + PAT -- the saved pattern position. + LOW_REG, HIGH_REG -- the highest and lowest active registers. + REGSTART, REGEND -- arrays of string positions. + REG_INFO -- array of information about each subexpression. + + Also assumes the variables `fail_stack' and (if debugging), `bufp', + `pend', `string1', `size1', `string2', and `size2'. */ + +#define POP_FAILURE_POINT(str, pat, low_reg, high_reg, regstart, regend, reg_info)\ +{ \ + DEBUG_STATEMENT (fail_stack_elt_t failure_id;) \ + s_reg_t this_reg; \ + const unsigned char *string_temp; \ + \ + assert (!FAIL_STACK_EMPTY ()); \ + \ + /* Remove failure points and point to how many regs pushed. */ \ + DEBUG_PRINT1 ("POP_FAILURE_POINT:\n"); \ + DEBUG_PRINT2 (" Before pop, next avail: %d\n", fail_stack.avail); \ + DEBUG_PRINT2 (" size: %d\n", fail_stack.size); \ + \ + assert (fail_stack.avail >= NUM_NONREG_ITEMS); \ + \ + DEBUG_POP (&failure_id); \ + DEBUG_PRINT2 (" Popping failure id: %u\n", failure_id); \ + \ + /* If the saved string location is NULL, it came from an \ + on_failure_keep_string_jump opcode, and we want to throw away the \ + saved NULL, thus retaining our current position in the string. */ \ + string_temp = POP_FAILURE_POINTER (); \ + if (string_temp != NULL) \ + str = (const char *) string_temp; \ + \ + DEBUG_PRINT2 (" Popping string 0x%x: `", str); \ + DEBUG_PRINT_DOUBLE_STRING (str, string1, size1, string2, size2); \ + DEBUG_PRINT1 ("'\n"); \ + \ + pat = (unsigned char *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" Popping pattern 0x%x:\n", pat); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pat, pend); \ + \ + /* Restore register info. */ \ + high_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping high active reg: %d\n", high_reg); \ + \ + low_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping low active reg: %d\n", low_reg); \ + \ + if (1) \ + for (this_reg = high_reg; this_reg >= low_reg; this_reg--) \ + { \ + DEBUG_PRINT2 (" Popping reg: %d\n", this_reg); \ + \ + reg_info[this_reg].word = POP_FAILURE_ELT (); \ + DEBUG_PRINT2 (" info: 0x%x\n", reg_info[this_reg]); \ + \ + regend[this_reg] = (const char *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" end: 0x%x\n", regend[this_reg]); \ + \ + regstart[this_reg] = (const char *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" start: 0x%x\n", regstart[this_reg]); \ + } \ + else \ + { \ + for (this_reg = highest_active_reg; this_reg > high_reg; this_reg--) \ + { \ + reg_info[this_reg].word.integer = 0; \ + regend[this_reg] = 0; \ + regstart[this_reg] = 0; \ + } \ + highest_active_reg = high_reg; \ + } \ + \ + set_regs_matched_done = 0; \ + DEBUG_STATEMENT (nfailure_points_popped++); \ +} /* POP_FAILURE_POINT */ + + + +/* Structure for per-register (a.k.a. per-group) information. + Other register information, such as the + starting and ending positions (which are addresses), and the list of + inner groups (which is a bits list) are maintained in separate + variables. + + We are making a (strictly speaking) nonportable assumption here: that + the compiler will pack our bit fields into something that fits into + the type of `word', i.e., is something that fits into one item on the + failure stack. */ + + +/* Declarations and macros for re_match_2. */ + +typedef union +{ + fail_stack_elt_t word; + struct + { + /* This field is one if this group can match the empty string, + zero if not. If not yet determined, `MATCH_NULL_UNSET_VALUE'. */ +#define MATCH_NULL_UNSET_VALUE 3 + unsigned match_null_string_p : 2; + unsigned is_active : 1; + unsigned matched_something : 1; + unsigned ever_matched_something : 1; + } bits; +} register_info_type; + +#define REG_MATCH_NULL_STRING_P(R) ((R).bits.match_null_string_p) +#define IS_ACTIVE(R) ((R).bits.is_active) +#define MATCHED_SOMETHING(R) ((R).bits.matched_something) +#define EVER_MATCHED_SOMETHING(R) ((R).bits.ever_matched_something) + + +/* Call this when have matched a real character; it sets `matched' flags + for the subexpressions which we are currently inside. Also records + that those subexprs have matched. */ +#define SET_REGS_MATCHED() \ + do \ + { \ + if (!set_regs_matched_done) \ + { \ + active_reg_t r; \ + set_regs_matched_done = 1; \ + for (r = lowest_active_reg; r <= highest_active_reg; r++) \ + { \ + MATCHED_SOMETHING (reg_info[r]) \ + = EVER_MATCHED_SOMETHING (reg_info[r]) \ + = 1; \ + } \ + } \ + } \ + while (0) + +/* Registers are set to a sentinel when they haven't yet matched. */ +static char reg_unset_dummy; +#define REG_UNSET_VALUE (®_unset_dummy) +#define REG_UNSET(e) ((e) == REG_UNSET_VALUE) + +/* Subroutine declarations and macros for regex_compile. */ + +static reg_errcode_t regex_compile _RE_ARGS ((const char *pattern, size_t size, + reg_syntax_t syntax, + struct re_pattern_buffer *bufp)); +static void store_op1 _RE_ARGS ((re_opcode_t op, unsigned char *loc, int arg)); +static void store_op2 _RE_ARGS ((re_opcode_t op, unsigned char *loc, + int arg1, int arg2)); +static void insert_op1 _RE_ARGS ((re_opcode_t op, unsigned char *loc, + int arg, unsigned char *end)); +static void insert_op2 _RE_ARGS ((re_opcode_t op, unsigned char *loc, + int arg1, int arg2, unsigned char *end)); +static boolean at_begline_loc_p _RE_ARGS ((const char *pattern, const char *p, + reg_syntax_t syntax)); +static boolean at_endline_loc_p _RE_ARGS ((const char *p, const char *pend, + reg_syntax_t syntax)); +static reg_errcode_t compile_range _RE_ARGS ((const char **p_ptr, + const char *pend, + char *translate, + reg_syntax_t syntax, + unsigned char *b)); + +/* Fetch the next character in the uncompiled pattern---translating it + if necessary. Also cast from a signed character in the constant + string passed to us by the user to an unsigned char that we can use + as an array index (in, e.g., `translate'). */ +#ifndef PATFETCH +#define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + if (translate) c = (unsigned char) translate[c]; \ + } while (0) +#endif + +/* Fetch the next character in the uncompiled pattern, with no + translation. */ +#define PATFETCH_RAW(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + } while (0) + +/* Go backwards one character in the pattern. */ +#define PATUNFETCH p-- + + +/* If `translate' is non-null, return translate[D], else just D. We + cast the subscript to translate because some data is declared as + `char *', to avoid warnings when a string constant is passed. But + when we use a character as a subscript we must make it unsigned. */ +#ifndef TRANSLATE +#define TRANSLATE(d) \ + (translate ? (char) translate[(unsigned char) (d)] : (d)) +#endif + + +/* Macros for outputting the compiled pattern into `buffer'. */ + +/* If the buffer isn't allocated when it comes in, use this. */ +#define INIT_BUF_SIZE 32 + +/* Make sure we have at least N more bytes of space in buffer. */ +#define GET_BUFFER_SPACE(n) \ + while ((unsigned long) (b - bufp->buffer + (n)) > bufp->allocated) \ + EXTEND_BUFFER () + +/* Make sure we have one more byte of buffer space and then add C to it. */ +#define BUF_PUSH(c) \ + do { \ + GET_BUFFER_SPACE (1); \ + *b++ = (unsigned char) (c); \ + } while (0) + + +/* Ensure we have two more bytes of buffer space and then append C1 and C2. */ +#define BUF_PUSH_2(c1, c2) \ + do { \ + GET_BUFFER_SPACE (2); \ + *b++ = (unsigned char) (c1); \ + *b++ = (unsigned char) (c2); \ + } while (0) + + +/* As with BUF_PUSH_2, except for three bytes. */ +#define BUF_PUSH_3(c1, c2, c3) \ + do { \ + GET_BUFFER_SPACE (3); \ + *b++ = (unsigned char) (c1); \ + *b++ = (unsigned char) (c2); \ + *b++ = (unsigned char) (c3); \ + } while (0) + + +/* Store a jump with opcode OP at LOC to location TO. We store a + relative address offset by the three bytes the jump itself occupies. */ +#define STORE_JUMP(op, loc, to) \ + store_op1 (op, loc, (int) ((to) - (loc) - 3)) + +/* Likewise, for a two-argument jump. */ +#define STORE_JUMP2(op, loc, to, arg) \ + store_op2 (op, loc, (int) ((to) - (loc) - 3), arg) + +/* Like `STORE_JUMP', but for inserting. Assume `b' is the buffer end. */ +#define INSERT_JUMP(op, loc, to) \ + insert_op1 (op, loc, (int) ((to) - (loc) - 3), b) + +/* Like `STORE_JUMP2', but for inserting. Assume `b' is the buffer end. */ +#define INSERT_JUMP2(op, loc, to, arg) \ + insert_op2 (op, loc, (int) ((to) - (loc) - 3), arg, b) + + +/* This is not an arbitrary limit: the arguments which represent offsets + into the pattern are two bytes long. So if 2^16 bytes turns out to + be too small, many things would have to change. */ +/* Any other compiler which, like MSC, has allocation limit below 2^16 + bytes will have to use approach similar to what was done below for + MSC and drop MAX_BUF_SIZE a bit. Otherwise you may end up + reallocating to 0 bytes. Such thing is not going to work too well. + You have been warned!! */ +#if defined(_MSC_VER) && !defined(_WIN32) +/* Microsoft C 16-bit versions limit malloc to approx 65512 bytes. + The REALLOC define eliminates a flurry of conversion warnings, + but is not required. */ +#define MAX_BUF_SIZE 65500L +#define REALLOC(p,s) realloc ((p), (size_t) (s)) +#else +#define MAX_BUF_SIZE (1L << 16) +#define REALLOC(p,s) realloc ((p), (s)) +#endif + +/* Extend the buffer by twice its current size via realloc and + reset the pointers that pointed into the old block to point to the + correct places in the new one. If extending the buffer results in it + being larger than MAX_BUF_SIZE, then flag memory exhausted. */ +#define EXTEND_BUFFER() \ + do { \ + unsigned char *old_buffer = bufp->buffer; \ + if (bufp->allocated == MAX_BUF_SIZE) \ + return REG_ESIZE; \ + bufp->allocated <<= 1; \ + if (bufp->allocated > MAX_BUF_SIZE) \ + bufp->allocated = MAX_BUF_SIZE; \ + bufp->buffer = (unsigned char *) REALLOC (bufp->buffer, bufp->allocated);\ + if (bufp->buffer == NULL) \ + return REG_ESPACE; \ + /* If the buffer moved, move all the pointers into it. */ \ + if (old_buffer != bufp->buffer) \ + { \ + b = (b - old_buffer) + bufp->buffer; \ + begalt = (begalt - old_buffer) + bufp->buffer; \ + if (fixup_alt_jump) \ + fixup_alt_jump = (fixup_alt_jump - old_buffer) + bufp->buffer;\ + if (laststart) \ + laststart = (laststart - old_buffer) + bufp->buffer; \ + if (pending_exact) \ + pending_exact = (pending_exact - old_buffer) + bufp->buffer; \ + } \ + } while (0) + + +/* Since we have one byte reserved for the register number argument to + {start,stop}_memory, the maximum number of groups we can report + things about is what fits in that byte. */ +#define MAX_REGNUM 255 + +/* But patterns can have more than `MAX_REGNUM' registers. We just + ignore the excess. */ +typedef unsigned regnum_t; + + +/* Macros for the compile stack. */ + +/* Since offsets can go either forwards or backwards, this type needs to + be able to hold values from -(MAX_BUF_SIZE - 1) to MAX_BUF_SIZE - 1. */ +/* int may be not enough when sizeof(int) == 2. */ +typedef long pattern_offset_t; + +typedef struct +{ + pattern_offset_t begalt_offset; + pattern_offset_t fixup_alt_jump; + pattern_offset_t inner_group_offset; + pattern_offset_t laststart_offset; + regnum_t regnum; +} compile_stack_elt_t; + + +typedef struct +{ + compile_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} compile_stack_type; + + +#define INIT_COMPILE_STACK_SIZE 32 + +#define COMPILE_STACK_EMPTY (compile_stack.avail == 0) +#define COMPILE_STACK_FULL (compile_stack.avail == compile_stack.size) + +/* The next available element. */ +#define COMPILE_STACK_TOP (compile_stack.stack[compile_stack.avail]) + + +/* Set the bit for character C in a list. */ +#define SET_LIST_BIT(c) \ + (b[((unsigned char) (c)) / BYTEWIDTH] \ + |= 1 << (((unsigned char) c) % BYTEWIDTH)) + + +/* Get the next unsigned number in the uncompiled pattern. */ +#define GET_UNSIGNED_NUMBER(num) \ + { if (p != pend) \ + { \ + PATFETCH (c); \ + while (ISDIGIT (c)) \ + { \ + if (num < 0) \ + num = 0; \ + num = num * 10 + c - '0'; \ + if (p == pend) \ + break; \ + PATFETCH (c); \ + } \ + } \ + } + +#if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) +/* The GNU C library provides support for user-defined character classes + and the functions from ISO C amendement 1. */ +# ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +# else +/* This shouldn't happen but some implementation might still have this + problem. Use a reasonable default value. */ +# define CHAR_CLASS_MAX_LENGTH 256 +# endif + +# define IS_CHAR_CLASS(string) wctype (string) +#else +# define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ + +# define IS_CHAR_CLASS(string) \ + (STREQ (string, "alpha") || STREQ (string, "upper") \ + || STREQ (string, "lower") || STREQ (string, "digit") \ + || STREQ (string, "alnum") || STREQ (string, "xdigit") \ + || STREQ (string, "space") || STREQ (string, "print") \ + || STREQ (string, "punct") || STREQ (string, "graph") \ + || STREQ (string, "cntrl") || STREQ (string, "blank")) +#endif + +#ifndef MATCH_MAY_ALLOCATE + +/* If we cannot allocate large objects within re_match_2_internal, + we make the fail stack and register vectors global. + The fail stack, we grow to the maximum size when a regexp + is compiled. + The register vectors, we adjust in size each time we + compile a regexp, according to the number of registers it needs. */ + +static fail_stack_type fail_stack; + +/* Size with which the following vectors are currently allocated. + That is so we can make them bigger as needed, + but never make them smaller. */ +static int regs_allocated_size; + +static const char ** regstart, ** regend; +static const char ** old_regstart, ** old_regend; +static const char **best_regstart, **best_regend; +static register_info_type *reg_info; +static const char **reg_dummy; +static register_info_type *reg_info_dummy; + +/* Make the register vectors big enough for NUM_REGS registers, + but don't make them smaller. */ + +static +regex_grow_registers (num_regs) + int num_regs; +{ + if (num_regs > regs_allocated_size) + { + RETALLOC_IF (regstart, num_regs, const char *); + RETALLOC_IF (regend, num_regs, const char *); + RETALLOC_IF (old_regstart, num_regs, const char *); + RETALLOC_IF (old_regend, num_regs, const char *); + RETALLOC_IF (best_regstart, num_regs, const char *); + RETALLOC_IF (best_regend, num_regs, const char *); + RETALLOC_IF (reg_info, num_regs, register_info_type); + RETALLOC_IF (reg_dummy, num_regs, const char *); + RETALLOC_IF (reg_info_dummy, num_regs, register_info_type); + + regs_allocated_size = num_regs; + } +} + +#endif /* not MATCH_MAY_ALLOCATE */ + +static boolean group_in_compile_stack _RE_ARGS ((compile_stack_type + compile_stack, + regnum_t regnum)); + +/* `regex_compile' compiles PATTERN (of length SIZE) according to SYNTAX. + Returns one of error codes defined in `regex.h', or zero for success. + + Assumes the `allocated' (and perhaps `buffer') and `translate' + fields are set in BUFP on entry. + + If it succeeds, results are put in BUFP (if it returns an error, the + contents of BUFP are undefined): + `buffer' is the compiled pattern; + `syntax' is set to SYNTAX; + `used' is set to the length of the compiled pattern; + `fastmap_accurate' is zero; + `re_nsub' is the number of subexpressions in PATTERN; + `not_bol' and `not_eol' are zero; + + The `fastmap' and `newline_anchor' fields are neither + examined nor set. */ + +/* Return, freeing storage we allocated. */ +#define FREE_STACK_RETURN(value) \ + return (free (compile_stack.stack), value) + +static reg_errcode_t +regex_compile (const char *pattern, + size_t size, + reg_syntax_t syntax, + struct re_pattern_buffer *bufp) +{ + /* We fetch characters from PATTERN here. Even though PATTERN is + `char *' (i.e., signed), we declare these variables as unsigned, so + they can be reliably used as array indices. */ + register unsigned char c, c1; + + /* A random temporary spot in PATTERN. */ + const char *p1; + + /* Points to the end of the buffer, where we should append. */ + register unsigned char *b; + + /* Keeps track of unclosed groups. */ + compile_stack_type compile_stack; + + /* Points to the current (ending) position in the pattern. */ + const char *p = pattern; + const char *pend = pattern + size; + + /* How to translate the characters in the pattern. */ + RE_TRANSLATE_TYPE translate = bufp->translate; + + /* Address of the count-byte of the most recently inserted `exactn' + command. This makes it possible to tell if a new exact-match + character can be added to that command or if the character requires + a new `exactn' command. */ + unsigned char *pending_exact = 0; + + /* Address of start of the most recently finished expression. + This tells, e.g., postfix * where to find the start of its + operand. Reset at the beginning of groups and alternatives. */ + unsigned char *laststart = 0; + + /* Address of beginning of regexp, or inside of last group. */ + unsigned char *begalt; + + /* Place in the uncompiled pattern (i.e., the {) to + which to go back if the interval is invalid. */ + const char *beg_interval; + + /* Address of the place where a forward jump should go to the end of + the containing expression. Each alternative of an `or' -- except the + last -- ends with a forward jump of this sort. */ + unsigned char *fixup_alt_jump = 0; + + /* Counts open-groups as they are encountered. Remembered for the + matching close-group on the compile stack, so the same register + number is put in the stop_memory as the start_memory. */ + regnum_t regnum = 0; + +#ifdef DEBUG + DEBUG_PRINT1 ("\nCompiling pattern: "); + if (debug) + { + unsigned debug_count; + + for (debug_count = 0; debug_count < size; debug_count++) + putchar (pattern[debug_count]); + putchar ('\n'); + } +#endif /* DEBUG */ + + /* Initialize the compile stack. */ + compile_stack.stack = TALLOC (INIT_COMPILE_STACK_SIZE, compile_stack_elt_t); + if (compile_stack.stack == NULL) + return REG_ESPACE; + + compile_stack.size = INIT_COMPILE_STACK_SIZE; + compile_stack.avail = 0; + + /* Initialize the pattern buffer. */ + bufp->syntax = syntax; + bufp->fastmap_accurate = 0; + bufp->not_bol = bufp->not_eol = 0; + + /* Set `used' to zero, so that if we return an error, the pattern + printer (for debugging) will think there's no pattern. We reset it + at the end. */ + bufp->used = 0; + + /* Always count groups, whether or not bufp->no_sub is set. */ + bufp->re_nsub = 0; + +#if !defined (emacs) && !defined (SYNTAX_TABLE) + /* Initialize the syntax table. */ + init_syntax_once (); +#endif + + if (bufp->allocated == 0) + { + if (bufp->buffer) + { /* If zero allocated, but buffer is non-null, try to realloc + enough space. This loses if buffer's address is bogus, but + that is the user's responsibility. */ + RETALLOC (bufp->buffer, INIT_BUF_SIZE, unsigned char); + } + else + { /* Caller did not allocate a buffer. Do it for them. */ + bufp->buffer = TALLOC (INIT_BUF_SIZE, unsigned char); + } + if (!bufp->buffer) FREE_STACK_RETURN (REG_ESPACE); + + bufp->allocated = INIT_BUF_SIZE; + } + + begalt = b = bufp->buffer; + + /* Loop through the uncompiled pattern until we're at the end. */ + while (p != pend) + { + PATFETCH (c); + + switch (c) + { + case '^': + { + if ( /* If at start of pattern, it's an operator. */ + p == pattern + 1 + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's come before. */ + || at_begline_loc_p (pattern, p, syntax)) + BUF_PUSH (begline); + else + goto normal_char; + } + break; + + + case '$': + { + if ( /* If at end of pattern, it's an operator. */ + p == pend + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's next. */ + || at_endline_loc_p (p, pend, syntax)) + BUF_PUSH (endline); + else + goto normal_char; + } + break; + + + case '+': + case '?': + if ((syntax & RE_BK_PLUS_QM) + || (syntax & RE_LIMITED_OPS)) + goto normal_char; + handle_plus: + case '*': + /* If there is no previous pattern... */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (!(syntax & RE_CONTEXT_INDEP_OPS)) + goto normal_char; + } + + { + /* Are we optimizing this jump? */ + boolean keep_string_p = false; + + /* 1 means zero (many) matches is allowed. */ + char zero_times_ok = 0, many_times_ok = 0; + + /* If there is a sequence of repetition chars, collapse it + down to just one (the right one). We can't combine + interval operators with these because of, e.g., `a{2}*', + which should only match an even number of `a's. */ + + for (;;) + { + zero_times_ok |= c != '+'; + many_times_ok |= c != '?'; + + if (p == pend) + break; + + PATFETCH (c); + + if (c == '*' + || (!(syntax & RE_BK_PLUS_QM) && (c == '+' || c == '?'))) + ; + + else if (syntax & RE_BK_PLUS_QM && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + if (!(c1 == '+' || c1 == '?')) + { + PATUNFETCH; + PATUNFETCH; + break; + } + + c = c1; + } + else + { + PATUNFETCH; + break; + } + + /* If we get here, we found another repeat character. */ + } + + /* Star, etc. applied to an empty pattern is equivalent + to an empty pattern. */ + if (!laststart) + break; + + /* Now we know whether or not zero matches is allowed + and also whether or not two or more matches is allowed. */ + if (many_times_ok) + { /* More than one repetition is allowed, so put in at the + end a backward relative jump from `b' to before the next + jump we're going to put in below (which jumps from + laststart to after this jump). + + But if we are at the `*' in the exact sequence `.*\n', + insert an unconditional jump backwards to the ., + instead of the beginning of the loop. This way we only + push a failure point once, instead of every time + through the loop. */ + assert (p - 1 > pattern); + + /* Allocate the space for the jump. */ + GET_BUFFER_SPACE (3); + + /* We know we are not at the first character of the pattern, + because laststart was nonzero. And we've already + incremented `p', by the way, to be the character after + the `*'. Do we have to do something analogous here + for null bytes, because of RE_DOT_NOT_NULL? */ + if (TRANSLATE (*(p - 2)) == TRANSLATE ('.') + && zero_times_ok + && p < pend && TRANSLATE (*p) == TRANSLATE ('\n') + && !(syntax & RE_DOT_NEWLINE)) + { /* We have .*\n. */ + STORE_JUMP (jump, b, laststart); + keep_string_p = true; + } + else + /* Anything else. */ + STORE_JUMP (maybe_pop_jump, b, laststart - 3); + + /* We've added more stuff to the buffer. */ + b += 3; + } + + /* On failure, jump from laststart to b + 3, which will be the + end of the buffer after this jump is inserted. */ + GET_BUFFER_SPACE (3); + INSERT_JUMP (keep_string_p ? on_failure_keep_string_jump + : on_failure_jump, + laststart, b + 3); + pending_exact = 0; + b += 3; + + if (!zero_times_ok) + { + /* At least one repetition is required, so insert a + `dummy_failure_jump' before the initial + `on_failure_jump' instruction of the loop. This + effects a skip over that instruction the first time + we hit that loop. */ + GET_BUFFER_SPACE (3); + INSERT_JUMP (dummy_failure_jump, laststart, laststart + 6); + b += 3; + } + } + break; + + + case '.': + laststart = b; + BUF_PUSH (anychar); + break; + + + case '[': + { + boolean had_char_class = false; + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + /* Ensure that we have enough space to push a charset: the + opcode, the length count, and the bitset; 34 bytes in all. */ + GET_BUFFER_SPACE (34); + + laststart = b; + + /* We test `*p == '^' twice, instead of using an if + statement, so we only need one BUF_PUSH. */ + BUF_PUSH (*p == '^' ? charset_not : charset); + if (*p == '^') + p++; + + /* Remember the first position in the bracket expression. */ + p1 = p; + + /* Push the number of bytes in the bitmap. */ + BUF_PUSH ((1 << BYTEWIDTH) / BYTEWIDTH); + + /* Clear the whole map. */ + bzero (b, (1 << BYTEWIDTH) / BYTEWIDTH); + + /* charset_not matches newline according to a syntax bit. */ + if ((re_opcode_t) b[-2] == charset_not + && (syntax & RE_HAT_LISTS_NOT_NEWLINE)) + SET_LIST_BIT ('\n'); + + /* Read in characters and ranges, setting map bits. */ + for (;;) + { + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + PATFETCH (c); + + /* \ might escape characters inside [...] and [^...]. */ + if ((syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + SET_LIST_BIT (c1); + continue; + } + + /* Could be the end of the bracket expression. If it's + not (i.e., when the bracket expression is `[]' so + far), the ']' character bit gets set way below. */ + if (c == ']' && p != p1 + 1) + break; + + /* Look ahead to see if it's a range when the last thing + was a character class. */ + if (had_char_class && c == '-' && *p != ']') + FREE_STACK_RETURN (REG_ERANGE); + + /* Look ahead to see if it's a range when the last thing + was a character: if this is a hyphen not at the + beginning or the end of a list, then it's the range + operator. */ + if (c == '-' + && !(p - 2 >= pattern && p[-2] == '[') + && !(p - 3 >= pattern && p[-3] == '[' && p[-2] == '^') + && *p != ']') + { + reg_errcode_t ret + = compile_range (&p, pend, translate, syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + } + + else if (p[0] == '-' && p[1] != ']') + { /* This handles ranges made up of characters only. */ + reg_errcode_t ret; + + /* Move past the `-'. */ + PATFETCH (c1); + + ret = compile_range (&p, pend, translate, syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + } + + /* See if we're at the beginning of a possible character + class. */ + + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == ':') + { /* Leave room for the null. */ + char str[CHAR_CLASS_MAX_LENGTH + 1]; + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[:'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if (c == ':' || c == ']' || p == pend + || c1 == CHAR_CLASS_MAX_LENGTH) + break; + str[c1++] = c; + } + str[c1] = '\0'; + + /* If isn't a word bracketed by `[:' and:`]': + undo the ending character, the letters, and leave + the leading `:' and `[' (but set bits for them). */ + if (c == ':' && *p == ']') + { +#if defined _LIBC || (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H) + boolean is_lower = STREQ (str, "lower"); + boolean is_upper = STREQ (str, "upper"); + wctype_t wt; + int ch; + + wt = wctype (str); + if (wt == 0) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ++ch) + { + if (iswctype (btowc (ch), wt)) + SET_LIST_BIT (ch); + + if (translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + + had_char_class = true; +#else + int ch; + boolean is_alnum = STREQ (str, "alnum"); + boolean is_alpha = STREQ (str, "alpha"); + boolean is_blank = STREQ (str, "blank"); + boolean is_cntrl = STREQ (str, "cntrl"); + boolean is_digit = STREQ (str, "digit"); + boolean is_graph = STREQ (str, "graph"); + boolean is_lower = STREQ (str, "lower"); + boolean is_print = STREQ (str, "print"); + boolean is_punct = STREQ (str, "punct"); + boolean is_space = STREQ (str, "space"); + boolean is_upper = STREQ (str, "upper"); + boolean is_xdigit = STREQ (str, "xdigit"); + + if (!IS_CHAR_CLASS (str)) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ch++) + { + /* This was split into 3 if's to + avoid an arbitrary limit in some compiler. */ + if ( (is_alnum && ISALNUM (ch)) + || (is_alpha && ISALPHA (ch)) + || (is_blank && ISBLANK (ch)) + || (is_cntrl && ISCNTRL (ch))) + SET_LIST_BIT (ch); + if ( (is_digit && ISDIGIT (ch)) + || (is_graph && ISGRAPH (ch)) + || (is_lower && ISLOWER (ch)) + || (is_print && ISPRINT (ch))) + SET_LIST_BIT (ch); + if ( (is_punct && ISPUNCT (ch)) + || (is_space && ISSPACE (ch)) + || (is_upper && ISUPPER (ch)) + || (is_xdigit && ISXDIGIT (ch))) + SET_LIST_BIT (ch); + if ( translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + had_char_class = true; +#endif /* libc || wctype.h */ + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT (':'); + had_char_class = false; + } + } + else + { + had_char_class = false; + SET_LIST_BIT (c); + } + } + + /* Discard any (non)matching list bytes that are all 0 at the + end of the map. Decrease the map-length byte too. */ + while ((int) b[-1] > 0 && b[b[-1] - 1] == 0) + b[-1]--; + b += b[-1]; + } + break; + + + case '(': + if (syntax & RE_NO_BK_PARENS) + goto handle_open; + else + goto normal_char; + + + case ')': + if (syntax & RE_NO_BK_PARENS) + goto handle_close; + else + goto normal_char; + + + case '\n': + if (syntax & RE_NEWLINE_ALT) + goto handle_alt; + else + goto normal_char; + + + case '|': + if (syntax & RE_NO_BK_VBAR) + goto handle_alt; + else + goto normal_char; + + + case '{': + if (syntax & RE_INTERVALS && syntax & RE_NO_BK_BRACES) + goto handle_interval; + else + goto normal_char; + + + case '\\': + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + /* Do not translate the character after the \, so that we can + distinguish, e.g., \B from \b, even if we normally would + translate, e.g., B to b. */ + PATFETCH_RAW (c); + + switch (c) + { + case '(': + if (syntax & RE_NO_BK_PARENS) + goto normal_backslash; + + handle_open: + bufp->re_nsub++; + regnum++; + + if (COMPILE_STACK_FULL) + { + RETALLOC (compile_stack.stack, compile_stack.size << 1, + compile_stack_elt_t); + if (compile_stack.stack == NULL) return REG_ESPACE; + + compile_stack.size <<= 1; + } + + /* These are the values to restore when we hit end of this + group. They are all relative offsets, so that if the + whole pattern moves because of realloc, they will still + be valid. */ + COMPILE_STACK_TOP.begalt_offset = begalt - bufp->buffer; + COMPILE_STACK_TOP.fixup_alt_jump + = fixup_alt_jump ? fixup_alt_jump - bufp->buffer + 1 : 0; + COMPILE_STACK_TOP.laststart_offset = b - bufp->buffer; + COMPILE_STACK_TOP.regnum = regnum; + + /* We will eventually replace the 0 with the number of + groups inner to this one. But do not push a + start_memory for groups beyond the last one we can + represent in the compiled pattern. */ + if (regnum <= MAX_REGNUM) + { + COMPILE_STACK_TOP.inner_group_offset = b - bufp->buffer + 2; + BUF_PUSH_3 (start_memory, regnum, 0); + } + + compile_stack.avail++; + + fixup_alt_jump = 0; + laststart = 0; + begalt = b; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + break; + + + case ')': + if (syntax & RE_NO_BK_PARENS) goto normal_backslash; + + if (COMPILE_STACK_EMPTY) { + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_backslash; + else + FREE_STACK_RETURN (REG_ERPAREN); + } + handle_close: + if (fixup_alt_jump) + { /* Push a dummy failure point at the end of the + alternative for a possible future + `pop_failure_jump' to pop. See comments at + `push_dummy_failure' in `re_match_2'. */ + BUF_PUSH (push_dummy_failure); + + /* We allocated space for this jump when we assigned + to `fixup_alt_jump', in the `handle_alt' case below. */ + STORE_JUMP (jump_past_alt, fixup_alt_jump, b - 1); + } + + /* See similar code for backslashed left paren above. */ + if (COMPILE_STACK_EMPTY) { + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_char; + else + FREE_STACK_RETURN (REG_ERPAREN); + } + /* Since we just checked for an empty stack above, this + ``can't happen''. */ + assert (compile_stack.avail != 0); + { + /* We don't just want to restore into `regnum', because + later groups should continue to be numbered higher, + as in `(ab)c(de)' -- the second group is #2. */ + regnum_t this_group_regnum; + + compile_stack.avail--; + begalt = bufp->buffer + COMPILE_STACK_TOP.begalt_offset; + fixup_alt_jump + = COMPILE_STACK_TOP.fixup_alt_jump + ? bufp->buffer + COMPILE_STACK_TOP.fixup_alt_jump - 1 + : 0; + laststart = bufp->buffer + COMPILE_STACK_TOP.laststart_offset; + this_group_regnum = COMPILE_STACK_TOP.regnum; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + + /* We're at the end of the group, so now we know how many + groups were inside this one. */ + if (this_group_regnum <= MAX_REGNUM) + { + unsigned char *inner_group_loc + = bufp->buffer + COMPILE_STACK_TOP.inner_group_offset; + + *inner_group_loc = regnum - this_group_regnum; + BUF_PUSH_3 (stop_memory, this_group_regnum, + regnum - this_group_regnum); + } + } + break; + + + case '|': /* `\|'. */ + if (syntax & RE_LIMITED_OPS || syntax & RE_NO_BK_VBAR) + goto normal_backslash; + handle_alt: + if (syntax & RE_LIMITED_OPS) + goto normal_char; + + /* Insert before the previous alternative a jump which + jumps to this alternative if the former fails. */ + GET_BUFFER_SPACE (3); + INSERT_JUMP (on_failure_jump, begalt, b + 6); + pending_exact = 0; + b += 3; + + /* The alternative before this one has a jump after it + which gets executed if it gets matched. Adjust that + jump so it will jump to this alternative's analogous + jump (put in below, which in turn will jump to the next + (if any) alternative's such jump, etc.). The last such + jump jumps to the correct final destination. A picture: + _____ _____ + | | | | + | v | v + a | b | c + + If we are at `b', then fixup_alt_jump right now points to a + three-byte space after `a'. We'll put in the jump, set + fixup_alt_jump to right after `b', and leave behind three + bytes which we'll fill in when we get to after `c'. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + /* Mark and leave space for a jump after this alternative, + to be filled in later either by next alternative or + when know we're at the end of a series of alternatives. */ + fixup_alt_jump = b; + GET_BUFFER_SPACE (3); + b += 3; + + laststart = 0; + begalt = b; + break; + + + case '{': + /* If \{ is a literal. */ + if (!(syntax & RE_INTERVALS) + /* If we're at `\{' and it's not the open-interval + operator. */ + || ((syntax & RE_INTERVALS) && (syntax & RE_NO_BK_BRACES)) + || (p - 2 == pattern && p == pend)) + goto normal_backslash; + + handle_interval: + { + /* If got here, then the syntax allows intervals. */ + + /* At least (most) this many matches must be made. */ + int lower_bound = -1, upper_bound = -1; + + beg_interval = p - 1; + + if (p == pend) + { + if (syntax & RE_NO_BK_BRACES) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_EBRACE); + } + + GET_UNSIGNED_NUMBER (lower_bound); + + if (c == ',') + { + GET_UNSIGNED_NUMBER (upper_bound); + if (upper_bound < 0) upper_bound = RE_DUP_MAX; + } + else + /* Interval such as `{1}' => match exactly once. */ + upper_bound = lower_bound; + + if (lower_bound < 0 || upper_bound > RE_DUP_MAX + || lower_bound > upper_bound) + { + if (syntax & RE_NO_BK_BRACES) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_BADBR); + } + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (c != '\\') FREE_STACK_RETURN (REG_EBRACE); + + PATFETCH (c); + } + + if (c != '}') + { + if (syntax & RE_NO_BK_BRACES) + goto unfetch_interval; + else + FREE_STACK_RETURN (REG_BADBR); + } + + /* We just parsed a valid interval. */ + + /* If it's invalid to have no preceding re. */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (syntax & RE_CONTEXT_INDEP_OPS) + laststart = b; + else + goto unfetch_interval; + } + + /* If the upper bound is zero, don't want to succeed at + all; jump from `laststart' to `b + 3', which will be + the end of the buffer after we insert the jump. */ + if (upper_bound == 0) + { + GET_BUFFER_SPACE (3); + INSERT_JUMP (jump, laststart, b + 3); + b += 3; + } + + /* Otherwise, we have a nontrivial interval. When + we're all done, the pattern will look like: + set_number_at <jump count> <upper bound> + set_number_at <succeed_n count> <lower bound> + succeed_n <after jump addr> <succeed_n count> + <body of loop> + jump_n <succeed_n addr> <jump count> + (The upper bound and `jump_n' are omitted if + `upper_bound' is 1, though.) */ + else + { /* If the upper bound is > 1, we need to insert + more at the end of the loop. */ + unsigned nbytes = 10 + (upper_bound > 1) * 10; + + GET_BUFFER_SPACE (nbytes); + + /* Initialize lower bound of the `succeed_n', even + though it will be set during matching by its + attendant `set_number_at' (inserted next), + because `re_compile_fastmap' needs to know. + Jump to the `jump_n' we might insert below. */ + INSERT_JUMP2 (succeed_n, laststart, + b + 5 + (upper_bound > 1) * 5, + lower_bound); + b += 5; + + /* Code to initialize the lower bound. Insert + before the `succeed_n'. The `5' is the last two + bytes of this `set_number_at', plus 3 bytes of + the following `succeed_n'. */ + insert_op2 (set_number_at, laststart, 5, lower_bound, b); + b += 5; + + if (upper_bound > 1) + { /* More than one repetition is allowed, so + append a backward jump to the `succeed_n' + that starts this interval. + + When we've reached this during matching, + we'll have matched the interval once, so + jump back only `upper_bound - 1' times. */ + STORE_JUMP2 (jump_n, b, laststart + 5, + upper_bound - 1); + b += 5; + + /* The location we want to set is the second + parameter of the `jump_n'; that is `b-2' as + an absolute address. `laststart' will be + the `set_number_at' we're about to insert; + `laststart+3' the number to set, the source + for the relative address. But we are + inserting into the middle of the pattern -- + so everything is getting moved up by 5. + Conclusion: (b - 2) - (laststart + 3) + 5, + i.e., b - laststart. + + We insert this at the beginning of the loop + so that if we fail during matching, we'll + reinitialize the bounds. */ + insert_op2 (set_number_at, laststart, b - laststart, + upper_bound - 1, b); + b += 5; + } + } + pending_exact = 0; + beg_interval = NULL; + } + break; + + unfetch_interval: + /* If an invalid interval, match the characters as literals. */ + assert (beg_interval); + p = beg_interval; + beg_interval = NULL; + + /* normal_char and normal_backslash need `c'. */ + PATFETCH (c); + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (p > pattern && p[-1] == '\\') + goto normal_backslash; + } + goto normal_char; + +#ifdef emacs + /* There is no way to specify the before_dot and after_dot + operators. rms says this is ok. --karl */ + case '=': + BUF_PUSH (at_dot); + break; + + case 's': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (syntaxspec, syntax_spec_code[c]); + break; + + case 'S': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (notsyntaxspec, syntax_spec_code[c]); + break; +#endif /* emacs */ + + + case 'w': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (wordchar); + break; + + + case 'W': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (notwordchar); + break; + + + case '<': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbeg); + break; + + case '>': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordend); + break; + + case 'b': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbound); + break; + + case 'B': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (notwordbound); + break; + + case '`': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (begbuf); + break; + + case '\'': + if (re_syntax_options & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (endbuf); + break; + + case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': + if (syntax & RE_NO_BK_REFS) + goto normal_char; + + c1 = c - '0'; + + if (c1 > regnum) + FREE_STACK_RETURN (REG_ESUBREG); + + /* Can't back reference to a subexpression if inside of it. */ + if (group_in_compile_stack (compile_stack, (regnum_t) c1)) + goto normal_char; + + laststart = b; + BUF_PUSH_2 (duplicate, c1); + break; + + + case '+': + case '?': + if (syntax & RE_BK_PLUS_QM) + goto handle_plus; + else + goto normal_backslash; + + default: + normal_backslash: + /* You might think it would be useful for \ to mean + not to translate; but if we don't translate it + it will never match anything. */ + c = TRANSLATE (c); + goto normal_char; + } + break; + + + default: + /* Expects the character in `c'. */ + normal_char: + /* If no exactn currently being built. */ + if (!pending_exact + + /* If last exactn not at current position. */ + || pending_exact + *pending_exact + 1 != b + + /* We have only one byte following the exactn for the count. */ + || *pending_exact == (1 << BYTEWIDTH) - 1 + + /* If followed by a repetition operator. */ + || *p == '*' || *p == '^' + || ((syntax & RE_BK_PLUS_QM) + ? *p == '\\' && (p[1] == '+' || p[1] == '?') + : (*p == '+' || *p == '?')) + || ((syntax & RE_INTERVALS) + && ((syntax & RE_NO_BK_BRACES) + ? *p == '{' + : (p[0] == '\\' && p[1] == '{')))) + { + /* Start building a new exactn. */ + + laststart = b; + + BUF_PUSH_2 (exactn, 0); + pending_exact = b - 1; + } + + BUF_PUSH (c); + (*pending_exact)++; + break; + } /* switch (c) */ + } /* while p != pend */ + + + /* Through the pattern now. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + if (!COMPILE_STACK_EMPTY) + FREE_STACK_RETURN (REG_EPAREN); + + /* If we don't want backtracking, force success + the first time we reach the end of the compiled pattern. */ + if (syntax & RE_NO_POSIX_BACKTRACKING) + BUF_PUSH (succeed); + + free (compile_stack.stack); + + /* We have succeeded; set the length of the buffer. */ + bufp->used = b - bufp->buffer; + +#ifdef DEBUG + if (debug) + { + DEBUG_PRINT1 ("\nCompiled pattern: \n"); + print_compiled_pattern (bufp); + } +#endif /* DEBUG */ + +#ifndef MATCH_MAY_ALLOCATE + /* Initialize the failure stack to the largest possible stack. This + isn't necessary unless we're trying to avoid calling alloca in + the search and match routines. */ + { + int num_regs = bufp->re_nsub + 1; + + /* Since DOUBLE_FAIL_STACK refuses to double only if the current size + is strictly greater than re_max_failures, the largest possible stack + is 2 * re_max_failures failure points. */ + if (fail_stack.size < (2 * re_max_failures * MAX_FAILURE_ITEMS)) + { + fail_stack.size = (2 * re_max_failures * MAX_FAILURE_ITEMS); + +#ifdef emacs + if (! fail_stack.stack) + fail_stack.stack + = (fail_stack_elt_t *) xmalloc (fail_stack.size + * sizeof (fail_stack_elt_t)); + else + fail_stack.stack + = (fail_stack_elt_t *) xrealloc (fail_stack.stack, + (fail_stack.size + * sizeof (fail_stack_elt_t))); +#else /* not emacs */ + if (! fail_stack.stack) + fail_stack.stack + = (fail_stack_elt_t *) malloc (fail_stack.size + * sizeof (fail_stack_elt_t)); + else + fail_stack.stack + = (fail_stack_elt_t *) realloc (fail_stack.stack, + (fail_stack.size + * sizeof (fail_stack_elt_t))); +#endif /* not emacs */ + } + + regex_grow_registers (num_regs); + } +#endif /* not MATCH_MAY_ALLOCATE */ + + return REG_NOERROR; +} /* regex_compile */ + +/* Subroutines for `regex_compile'. */ + +/* Store OP at LOC followed by two-byte integer parameter ARG. */ + +static void +store_op1 (re_opcode_t op, + unsigned char *loc, + int arg) +{ + *loc = (unsigned char) op; + STORE_NUMBER (loc + 1, arg); +} + + +/* Like `store_op1', but for two two-byte parameters ARG1 and ARG2. */ + +static void +store_op2(re_opcode_t op, + unsigned char *loc, + int arg1, + int arg2) +{ + *loc = (unsigned char) op; + STORE_NUMBER (loc + 1, arg1); + STORE_NUMBER (loc + 3, arg2); +} + + +/* Copy the bytes from LOC to END to open up three bytes of space at LOC + for OP followed by two-byte integer parameter ARG. */ + +static void +insert_op1(re_opcode_t op, + unsigned char *loc, + int arg, + unsigned char *end) +{ + register unsigned char *pfrom = end; + register unsigned char *pto = end + 3; + + while (pfrom != loc) + *--pto = *--pfrom; + + store_op1 (op, loc, arg); +} + + +/* Like `insert_op1', but for two two-byte parameters ARG1 and ARG2. */ + +static void +insert_op2(re_opcode_t op, + unsigned char *loc, + int arg1, + int arg2, + unsigned char *end) +{ + register unsigned char *pfrom = end; + register unsigned char *pto = end + 5; + + while (pfrom != loc) + *--pto = *--pfrom; + + store_op2 (op, loc, arg1, arg2); +} + + +/* P points to just after a ^ in PATTERN. Return true if that ^ comes + after an alternative or a begin-subexpression. We assume there is at + least one character before the ^. */ + +static boolean +at_begline_loc_p(const char *pattern, + const char *p, + reg_syntax_t syntax) +{ + const char *prev = p - 2; + boolean prev_prev_backslash = prev > pattern && prev[-1] == '\\'; + + return + /* After a subexpression? */ + (*prev == '(' && (syntax & RE_NO_BK_PARENS || prev_prev_backslash)) + /* After an alternative? */ + || (*prev == '|' && (syntax & RE_NO_BK_VBAR || prev_prev_backslash)); +} + + +/* The dual of at_begline_loc_p. This one is for $. We assume there is + at least one character after the $, i.e., `P < PEND'. */ + +static boolean +at_endline_loc_p(const char *p, + const char *pend, + reg_syntax_t syntax) +{ + const char *next = p; + boolean next_backslash = *next == '\\'; + const char *next_next = p + 1 < pend ? p + 1 : 0; + + return + /* Before a subexpression? */ + (syntax & RE_NO_BK_PARENS ? *next == ')' + : next_backslash && next_next && *next_next == ')') + /* Before an alternative? */ + || (syntax & RE_NO_BK_VBAR ? *next == '|' + : next_backslash && next_next && *next_next == '|'); +} + + +/* Returns true if REGNUM is in one of COMPILE_STACK's elements and + false if it's not. */ + +static boolean +group_in_compile_stack(compile_stack_type compile_stack, + regnum_t regnum) +{ + int this_element; + + for (this_element = compile_stack.avail - 1; + this_element >= 0; + this_element--) + if (compile_stack.stack[this_element].regnum == regnum) + return true; + + return false; +} + + +/* Read the ending character of a range (in a bracket expression) from the + uncompiled pattern *P_PTR (which ends at PEND). We assume the + starting character is in `P[-2]'. (`P[-1]' is the character `-'.) + Then we set the translation of all bits between the starting and + ending characters (inclusive) in the compiled pattern B. + + Return an error code. + + We use these short variable names so we can use the same macros as + `regex_compile' itself. */ + +static reg_errcode_t +compile_range(const char **p_ptr, + const char *pend, + RE_TRANSLATE_TYPE translate, + reg_syntax_t syntax, + unsigned char *b) +{ + unsigned this_char; + + const char *p = *p_ptr; + unsigned int range_start, range_end; + + if (p == pend) + return REG_ERANGE; + + /* Even though the pattern is a signed `char *', we need to fetch + with unsigned char *'s; if the high bit of the pattern character + is set, the range endpoints will be negative if we fetch using a + signed char *. + + We also want to fetch the endpoints without translating them; the + appropriate translation is done in the bit-setting loop below. */ + /* The SVR4 compiler on the 3B2 had trouble with unsigned const char *. */ + range_start = ((const unsigned char *) p)[-2]; + range_end = ((const unsigned char *) p)[0]; + + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + /* If the start is after the end, the range is empty. */ + if (range_start > range_end) + return syntax & RE_NO_EMPTY_RANGES ? REG_ERANGE : REG_NOERROR; + + /* Here we see why `this_char' has to be larger than an `unsigned + char' -- the range is inclusive, so if `range_end' == 0xff + (assuming 8-bit characters), we would otherwise go into an infinite + loop, since all characters <= 0xff. */ + for (this_char = range_start; this_char <= range_end; this_char++) + { + SET_LIST_BIT (TRANSLATE (this_char)); + } + + return REG_NOERROR; +} + +/* re_compile_fastmap computes a ``fastmap'' for the compiled pattern in + BUFP. A fastmap records which of the (1 << BYTEWIDTH) possible + characters can start a string that matches the pattern. This fastmap + is used by re_search to skip quickly over impossible starting points. + + The caller must supply the address of a (1 << BYTEWIDTH)-byte data + area as BUFP->fastmap. + + We set the `fastmap', `fastmap_accurate', and `can_be_null' fields in + the pattern buffer. + + Returns 0 if we succeed, -2 if an internal error. */ + +int +re_compile_fastmap(struct re_pattern_buffer *bufp) +{ + int j, k; +#ifdef MATCH_MAY_ALLOCATE + fail_stack_type fail_stack; +#endif +#ifndef REGEX_MALLOC + char *destination; +#endif + + register char *fastmap = bufp->fastmap; + unsigned char *pattern = bufp->buffer; + unsigned char *p = pattern; + register unsigned char *pend = pattern + bufp->used; + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* Assume that each path through the pattern can be null until + proven otherwise. We set this false at the bottom of switch + statement, to which we get only if a particular path doesn't + match the empty string. */ + boolean path_can_be_null = true; + + /* We aren't doing a `succeed_n' to begin with. */ + boolean succeed_n_p = false; + + assert (fastmap != NULL && p != NULL); + + INIT_FAIL_STACK (); + bzero (fastmap, 1 << BYTEWIDTH); /* Assume nothing's valid. */ + bufp->fastmap_accurate = 1; /* It will be when we're done. */ + bufp->can_be_null = 0; + + while (1) + { + if (p == pend || *p == succeed) + { + /* We have reached the (effective) end of pattern. */ + if (!FAIL_STACK_EMPTY ()) + { + bufp->can_be_null |= path_can_be_null; + + /* Reset for next path. */ + path_can_be_null = true; + + p = fail_stack.stack[--fail_stack.avail].pointer; + + continue; + } + else + break; + } + + /* We should never be about to go beyond the end of the pattern. */ + assert (p < pend); + + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + + /* I guess the idea here is to simply not bother with a fastmap + if a backreference is used, since it's too hard to figure out + the fastmap for the corresponding group. Setting + `can_be_null' stops `re_search_2' from using the fastmap, so + that is all we do. */ + case duplicate: + bufp->can_be_null = 1; + goto done; + + + /* Following are the cases which match a character. These end + with `break'. */ + + case exactn: + fastmap[p[1]] = 1; + break; + + + case charset: + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH))) + fastmap[j] = 1; + break; + + + case charset_not: + /* Chars beyond end of map must be allowed. */ + for (j = *p * BYTEWIDTH; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (!(p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH)))) + fastmap[j] = 1; + break; + + + case wordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == Sword) + fastmap[j] = 1; + break; + + + case notwordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != Sword) + fastmap[j] = 1; + break; + + + case anychar: + { + int fastmap_newline = fastmap['\n']; + + /* `.' matches anything ... */ + for (j = 0; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + /* ... except perhaps newline. */ + if (!(bufp->syntax & RE_DOT_NEWLINE)) + fastmap['\n'] = fastmap_newline; + + /* Return if we have already set `can_be_null'; if we have, + then the fastmap is irrelevant. Something's wrong here. */ + else if (bufp->can_be_null) + goto done; + + /* Otherwise, have to check alternative paths. */ + break; + } + +#ifdef emacs + case syntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + case notsyntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + /* All cases after this match the empty string. These end with + `continue'. */ + + + case before_dot: + case at_dot: + case after_dot: + continue; +#endif /* emacs */ + + + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbound: + case notwordbound: + case wordbeg: + case wordend: + case push_dummy_failure: + continue; + + + case jump_n: + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case jump_past_alt: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + if (j > 0) + continue; + + /* Jump backward implies we just went through the body of a + loop and matched nothing. Opcode jumped to should be + `on_failure_jump' or `succeed_n'. Just treat it like an + ordinary jump. For a * loop, it has pushed its failure + point already; if so, discard that as redundant. */ + if ((re_opcode_t) *p != on_failure_jump + && (re_opcode_t) *p != succeed_n) + continue; + + p++; + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + + /* If what's on the stack is where we are now, pop it. */ + if (!FAIL_STACK_EMPTY () + && fail_stack.stack[fail_stack.avail - 1].pointer == p) + fail_stack.avail--; + + continue; + + + case on_failure_jump: + case on_failure_keep_string_jump: + handle_on_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + + /* For some patterns, e.g., `(a?)?', `p+j' here points to the + end of the pattern. We don't want to push such a point, + since when we restore it above, entering the switch will + increment `p' past the end of the pattern. We don't need + to push such a point since we obviously won't find any more + fastmap entries beyond `pend'. Such a pattern can match + the null string, though. */ + if (p + j < pend) + { + if (!PUSH_PATTERN_OP (p + j, fail_stack)) + { + RESET_FAIL_STACK (); + return -2; + } + } + else + bufp->can_be_null = 1; + + if (succeed_n_p) + { + EXTRACT_NUMBER_AND_INCR (k, p); /* Skip the n. */ + succeed_n_p = false; + } + + continue; + + + case succeed_n: + /* Get to the number of times to succeed. */ + p += 2; + + /* Increment p past the n for when k != 0. */ + EXTRACT_NUMBER_AND_INCR (k, p); + if (k == 0) + { + p -= 4; + succeed_n_p = true; /* Spaghetti code alert. */ + goto handle_on_failure_jump; + } + continue; + + + case set_number_at: + p += 4; + continue; + + + case start_memory: + case stop_memory: + p += 2; + continue; + + + default: + abort (); /* We have listed all the cases. */ + } /* switch *p++ */ + + /* Getting here means we have found the possible starting + characters for one path of the pattern -- and that the empty + string does not match. We need not follow this path further. + Instead, look at the next alternative (remembered on the + stack), or quit if no more. The test at the top of the loop + does these things. */ + path_can_be_null = false; + p = pend; + } /* while p */ + + /* Set `can_be_null' for the last path (also the first path, if the + pattern is empty). */ + bufp->can_be_null |= path_can_be_null; + + done: + RESET_FAIL_STACK (); + return 0; +} /* re_compile_fastmap */ + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use + this memory for recording register information. STARTS and ENDS + must be allocated using the malloc library routine, and must each + be at least NUM_REGS * sizeof (regoff_t) bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ + +void +re_set_registers(struct re_pattern_buffer *bufp, + struct re_registers *regs, + unsigned num_regs, + regoff_t *starts, + regoff_t *ends) +{ + if (num_regs) + { + bufp->regs_allocated = REGS_REALLOCATE; + regs->num_regs = num_regs; + regs->start = starts; + regs->end = ends; + } + else + { + bufp->regs_allocated = REGS_UNALLOCATED; + regs->num_regs = 0; + regs->start = regs->end = (regoff_t *) 0; + } +} + +/* Searching routines. */ + +/* Like re_search_2, below, but only one string is specified, and + doesn't let you say where to stop matching. */ + +int +re_search(struct re_pattern_buffer *bufp, + const char *string, + int size, + int startpos, + int range, + struct re_registers *regs) +{ + return re_search_2 (bufp, NULL, 0, string, size, startpos, range, + regs, size); +} + + +/* Using the compiled pattern in BUFP->buffer, first tries to match the + virtual concatenation of STRING1 and STRING2, starting first at index + STARTPOS, then at STARTPOS + 1, and so on. + + STRING1 and STRING2 have length SIZE1 and SIZE2, respectively. + + RANGE is how far to scan while trying to match. RANGE = 0 means try + only at STARTPOS; in general, the last start tried is STARTPOS + + RANGE. + + In REGS, return the indices of the virtual concatenation of STRING1 + and STRING2 that matched the entire BUFP->buffer and its contained + subexpressions. + + Do not consider matching one past the index STOP in the virtual + concatenation of STRING1 and STRING2. + + We return either the position in the strings at which the match was + found, -1 if no match, or -2 if error (such as failure + stack overflow). */ + +int +re_search_2(struct re_pattern_buffer *bufp, + const char *string1, + int size1, + const char *string2, + int size2, + int startpos, + int range, + struct re_registers *regs, + int stop) +{ + int val; + register char *fastmap = bufp->fastmap; + register RE_TRANSLATE_TYPE translate = bufp->translate; + int total_size = size1 + size2; + int endpos = startpos + range; + + /* Check for out-of-range STARTPOS. */ + if (startpos < 0 || startpos > total_size) + return -1; + + /* Fix up RANGE if it might eventually take us outside + the virtual concatenation of STRING1 and STRING2. + Make sure we won't move STARTPOS below 0 or above TOTAL_SIZE. */ + if (endpos < 0) + range = 0 - startpos; + else if (endpos > total_size) + range = total_size - startpos; + + /* If the search isn't to be a backwards one, don't waste time in a + search for a pattern that must be anchored. */ + if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == begbuf && range > 0) + { + if (startpos > 0) + return -1; + else + range = 1; + } + +#ifdef emacs + /* In a forward search for something that starts with \=. + don't keep searching past point. */ + if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == at_dot && range > 0) + { + range = PT - startpos; + if (range <= 0) + return -1; + } +#endif /* emacs */ + + /* Update the fastmap now if not correct already. */ + if (fastmap && !bufp->fastmap_accurate) + if (re_compile_fastmap (bufp) == -2) + return -2; + + /* Loop through the string, looking for a place to start matching. */ + for (;;) + { + /* If a fastmap is supplied, skip quickly over characters that + cannot be the start of a match. If the pattern can match the + null string, however, we don't need to skip characters; we want + the first null string. */ + if (fastmap && startpos < total_size && !bufp->can_be_null) + { + if (range > 0) /* Searching forwards. */ + { + register const char *d; + register int lim = 0; + int irange = range; + + if (startpos < size1 && startpos + range >= size1) + lim = range - (size1 - startpos); + + d = (startpos >= size1 ? string2 - size1 : string1) + startpos; + + /* Written out as an if-else to avoid testing `translate' + inside the loop. */ + if (translate) + while (range > lim + && !fastmap[(unsigned char) + translate[(unsigned char) *d++]]) + range--; + else + while (range > lim && !fastmap[(unsigned char) *d++]) + range--; + + startpos += irange - range; + } + else /* Searching backwards. */ + { + register char c = (size1 == 0 || startpos >= size1 + ? string2[startpos - size1] + : string1[startpos]); + + if (!fastmap[(unsigned char) TRANSLATE (c)]) + goto advance; + } + } + + /* If can't match the null string, and that's all we have left, fail. */ + if (range >= 0 && startpos == total_size && fastmap + && !bufp->can_be_null) + return -1; + + val = re_match_2_internal (bufp, string1, size1, string2, size2, + startpos, regs, stop); +#ifndef REGEX_MALLOC +#ifdef C_ALLOCA + alloca (0); +#endif +#endif + + if (val >= 0) + return startpos; + + if (val == -2) + return -2; + + advance: + if (!range) + break; + else if (range > 0) + { + range--; + startpos++; + } + else + { + range++; + startpos--; + } + } + return -1; +} /* re_search_2 */ + +/* This converts PTR, a pointer into one of the search strings `string1' + and `string2' into an offset from the beginning of that string. */ +#define POINTER_TO_OFFSET(ptr) \ + (FIRST_STRING_P (ptr) \ + ? ((regoff_t) ((ptr) - string1)) \ + : ((regoff_t) ((ptr) - string2 + size1))) + +/* Macros for dealing with the split strings in re_match_2. */ + +#define MATCHING_IN_FIRST_STRING (dend == end_match_1) + +/* Call before fetching a character with *d. This switches over to + string2 if necessary. */ +#define PREFETCH() \ + while (d == dend) \ + { \ + /* End of string2 => fail. */ \ + if (dend == end_match_2) \ + goto fail; \ + /* End of string1 => advance to string2. */ \ + d = string2; \ + dend = end_match_2; \ + } + + +/* Test if at very beginning or at very end of the virtual concatenation + of `string1' and `string2'. If only one string, it's `string2'. */ +#define AT_STRINGS_BEG(d) ((d) == (size1 ? string1 : string2) || !size2) +#define AT_STRINGS_END(d) ((d) == end2) + + +/* Test if D points to a character which is word-constituent. We have + two special cases to check for: if past the end of string1, look at + the first character in string2; and if before the beginning of + string2, look at the last character in string1. */ +#define WORDCHAR_P(d) \ + (SYNTAX ((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d)) \ + == Sword) + +/* Disabled due to a compiler bug -- see comment at case wordbound */ +#if 0 +/* Test if the character before D and the one at D differ with respect + to being word-constituent. */ +#define AT_WORD_BOUNDARY(d) \ + (AT_STRINGS_BEG (d) || AT_STRINGS_END (d) \ + || WORDCHAR_P (d - 1) != WORDCHAR_P (d)) +#endif + +/* Free everything we malloc. */ +#ifdef MATCH_MAY_ALLOCATE +#define FREE_VAR(var) if (var) REGEX_FREE (var); var = NULL +#define FREE_VARIABLES() \ + do { \ + REGEX_FREE_STACK (fail_stack.stack); \ + FREE_VAR ((void*) regstart); \ + FREE_VAR ((void*) regend); \ + FREE_VAR ((void*) old_regstart); \ + FREE_VAR ((void*) old_regend); \ + FREE_VAR ((void*) best_regstart); \ + FREE_VAR ((void*) best_regend); \ + FREE_VAR ((void*) reg_info); \ + FREE_VAR ((void*) reg_dummy); \ + FREE_VAR ((void*) reg_info_dummy); \ + } while (0) +#else +#define FREE_VARIABLES() ((void)0) /* Do nothing! But inhibit gcc warning. */ +#endif /* not MATCH_MAY_ALLOCATE */ + +/* These values must meet several constraints. They must not be valid + register values; since we have a limit of 255 registers (because + we use only one byte in the pattern for the register number), we can + use numbers larger than 255. They must differ by 1, because of + NUM_FAILURE_ITEMS above. And the value for the lowest register must + be larger than the value for the highest register, so we do not try + to actually save any registers when none are active. */ +#define NO_HIGHEST_ACTIVE_REG (1 << BYTEWIDTH) +#define NO_LOWEST_ACTIVE_REG (NO_HIGHEST_ACTIVE_REG + 1) + +/* Matching routines. */ + +#ifndef emacs /* Emacs never uses this. */ +/* re_match is like re_match_2 except it takes only a single string. */ + +int +re_match(struct re_pattern_buffer *bufp, + const char *string, + int size, + int pos, + struct re_registers *regs) +{ + int result = re_match_2_internal (bufp, NULL, 0, string, size, + pos, regs, size); +#ifndef REGEX_MALLOC +#ifdef C_ALLOCA + alloca (0); +#endif +#endif + return result; +} +#endif /* not emacs */ + +static boolean group_match_null_string_p _RE_ARGS ((unsigned char **p, + unsigned char *end, + register_info_type *reg_info)); +static boolean alt_match_null_string_p _RE_ARGS ((unsigned char *p, + unsigned char *end, + register_info_type *reg_info)); +static boolean common_op_match_null_string_p _RE_ARGS ((unsigned char **p, + unsigned char *end, + register_info_type *reg_info)); +static int bcmp_translate _RE_ARGS ((const char *s1, const char *s2, + int len, char *translate)); + +/* re_match_2 matches the compiled pattern in BUFP against the + the (virtual) concatenation of STRING1 and STRING2 (of length SIZE1 + and SIZE2, respectively). We start matching at POS, and stop + matching at STOP. + + If REGS is non-null and the `no_sub' field of BUFP is nonzero, we + store offsets for the substring each group matched in REGS. See the + documentation for exactly how many groups we fill. + + We return -1 if no match, -2 if an internal error (such as the + failure stack overflowing). Otherwise, we return the length of the + matched substring. */ + +int +re_match_2(struct re_pattern_buffer *bufp, + const char *string1, + int size1, + const char *string2, + int size2, + int pos, + struct re_registers *regs, + int stop) +{ + int result = re_match_2_internal (bufp, string1, size1, string2, size2, + pos, regs, stop); +#ifndef REGEX_MALLOC +#ifdef C_ALLOCA + alloca (0); +#endif +#endif + return result; +} + +/* This is a separate function so that we can force an alloca cleanup + afterwards. */ +static int +re_match_2_internal(struct re_pattern_buffer *bufp, + const char *string1, + int size1, + const char *string2, + int size2, + int pos, + struct re_registers *regs, + int stop) +{ + /* General temporaries. */ + int mcnt; + unsigned char *p1; + + /* Just past the end of the corresponding string. */ + const char *end1, *end2; + + /* Pointers into string1 and string2, just past the last characters in + each to consider matching. */ + const char *end_match_1, *end_match_2; + + /* Where we are in the data, and the end of the current string. */ + const char *d, *dend; + + /* Where we are in the pattern, and the end of the pattern. */ + unsigned char *p = bufp->buffer; + register unsigned char *pend = p + bufp->used; + + /* Mark the opcode just after a start_memory, so we can test for an + empty subpattern when we get to the stop_memory. */ + unsigned char *just_past_start_mem = 0; + + /* We use this to map every character in the string. */ + RE_TRANSLATE_TYPE translate = bufp->translate; + + /* Failure point stack. Each place that can handle a failure further + down the line pushes a failure point on this stack. It consists of + restart, regend, and reg_info for all registers corresponding to + the subexpressions we're currently inside, plus the number of such + registers, and, finally, two char *'s. The first char * is where + to resume scanning the pattern; the second one is where to resume + scanning the strings. If the latter is zero, the failure point is + a ``dummy''; if a failure happens and the failure point is a dummy, + it gets discarded and the next next one is tried. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + fail_stack_type fail_stack; +#endif +#ifdef DEBUG + static unsigned failure_id = 0; + unsigned nfailure_points_pushed = 0, nfailure_points_popped = 0; +#endif + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* We fill all the registers internally, independent of what we + return, for use in backreferences. The number here includes + an element for register zero. */ + size_t num_regs = bufp->re_nsub + 1; + + /* The currently active registers. */ + active_reg_t lowest_active_reg = NO_LOWEST_ACTIVE_REG; + active_reg_t highest_active_reg = NO_HIGHEST_ACTIVE_REG; + + /* Information on the contents of registers. These are pointers into + the input strings; they record just what was matched (on this + attempt) by a subexpression part of the pattern, that is, the + regnum-th regstart pointer points to where in the pattern we began + matching and the regnum-th regend points to right after where we + stopped matching the regnum-th subexpression. (The zeroth register + keeps track of what the whole pattern matches.) */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **regstart, **regend; +#endif + + /* If a group that's operated upon by a repetition operator fails to + match anything, then the register for its start will need to be + restored because it will have been set to wherever in the string we + are when we last see its open-group operator. Similarly for a + register's end. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **old_regstart, **old_regend; +#endif + + /* The is_active field of reg_info helps us keep track of which (possibly + nested) subexpressions we are currently in. The matched_something + field of reg_info[reg_num] helps us tell whether or not we have + matched any of the pattern so far this time through the reg_num-th + subexpression. These two fields get reset each time through any + loop their register is in. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + register_info_type *reg_info; +#endif + + /* The following record the register info as found in the above + variables when we find a match better than any we've seen before. + This happens as we backtrack through the failure points, which in + turn happens only if we have not yet matched the entire string. */ + unsigned best_regs_set = false; +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **best_regstart, **best_regend; +#endif + + /* Logically, this is `best_regend[0]'. But we don't want to have to + allocate space for that if we're not allocating space for anything + else (see below). Also, we never need info about register 0 for + any of the other register vectors, and it seems rather a kludge to + treat `best_regend' differently than the rest. So we keep track of + the end of the best match so far in a separate variable. We + initialize this to NULL so that when we backtrack the first time + and need to test it, it's not garbage. */ + const char *match_end = NULL; + + /* This helps SET_REGS_MATCHED avoid doing redundant work. */ + int set_regs_matched_done = 0; + + /* Used when we pop values we don't care about. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const char **reg_dummy; + register_info_type *reg_info_dummy; +#endif + +#ifdef DEBUG + /* Counts the total number of registers pushed. */ + unsigned num_regs_pushed = 0; +#endif + + DEBUG_PRINT1 ("\n\nEntering re_match_2.\n"); + + INIT_FAIL_STACK (); + +#ifdef MATCH_MAY_ALLOCATE + /* Do not bother to initialize all the register variables if there are + no groups in the pattern, as it takes a fair amount of time. If + there are groups, we include space for register 0 (the whole + pattern), even though we never use it, since it simplifies the + array indexing. We should fix this. */ + if (bufp->re_nsub) + { + regstart = REGEX_TALLOC (num_regs, const char *); + regend = REGEX_TALLOC (num_regs, const char *); + old_regstart = REGEX_TALLOC (num_regs, const char *); + old_regend = REGEX_TALLOC (num_regs, const char *); + best_regstart = REGEX_TALLOC (num_regs, const char *); + best_regend = REGEX_TALLOC (num_regs, const char *); + reg_info = REGEX_TALLOC (num_regs, register_info_type); + reg_dummy = REGEX_TALLOC (num_regs, const char *); + reg_info_dummy = REGEX_TALLOC (num_regs, register_info_type); + + if (!(regstart && regend && old_regstart && old_regend && reg_info + && best_regstart && best_regend && reg_dummy && reg_info_dummy)) + { + FREE_VARIABLES (); + return -2; + } + } + else + { + /* We must initialize all our variables to NULL, so that + `FREE_VARIABLES' doesn't try to free them. */ + regstart = regend = old_regstart = old_regend = best_regstart + = best_regend = reg_dummy = NULL; + reg_info = reg_info_dummy = (register_info_type *) NULL; + } +#endif /* MATCH_MAY_ALLOCATE */ + + /* The starting position is bogus. */ + if (pos < 0 || pos > size1 + size2) + { + FREE_VARIABLES (); + return -1; + } + + /* Initialize subexpression text positions to -1 to mark ones that no + start_memory/stop_memory has been seen for. Also initialize the + register information struct. */ + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = regend[mcnt] + = old_regstart[mcnt] = old_regend[mcnt] = REG_UNSET_VALUE; + + REG_MATCH_NULL_STRING_P (reg_info[mcnt]) = MATCH_NULL_UNSET_VALUE; + IS_ACTIVE (reg_info[mcnt]) = 0; + MATCHED_SOMETHING (reg_info[mcnt]) = 0; + EVER_MATCHED_SOMETHING (reg_info[mcnt]) = 0; + } + + /* We move `string1' into `string2' if the latter's empty -- but not if + `string1' is null. */ + if (size2 == 0 && string1 != NULL) + { + string2 = string1; + size2 = size1; + string1 = 0; + size1 = 0; + } + end1 = string1 + size1; + end2 = string2 + size2; + + /* Compute where to stop matching, within the two strings. */ + if (stop <= size1) + { + end_match_1 = string1 + stop; + end_match_2 = string2; + } + else + { + end_match_1 = end1; + end_match_2 = string2 + stop - size1; + } + + /* `p' scans through the pattern as `d' scans through the data. + `dend' is the end of the input string that `d' points within. `d' + is advanced into the following input string whenever necessary, but + this happens before fetching; therefore, at the beginning of the + loop, `d' can be pointing at the end of a string, but it cannot + equal `string2'. */ + if (size1 > 0 && pos <= size1) + { + d = string1 + pos; + dend = end_match_1; + } + else + { + d = string2 + pos - size1; + dend = end_match_2; + } + + DEBUG_PRINT1 ("The compiled pattern is:\n"); + DEBUG_PRINT_COMPILED_PATTERN (bufp, p, pend); + DEBUG_PRINT1 ("The string to match is: `"); + DEBUG_PRINT_DOUBLE_STRING (d, string1, size1, string2, size2); + DEBUG_PRINT1 ("'\n"); + + /* This loops over pattern commands. It exits by returning from the + function if the match is complete, or it drops through if the match + fails at this starting point in the input data. */ + for (;;) + { +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + + if (p == pend) + { /* End of pattern means we might have succeeded. */ + DEBUG_PRINT1 ("end of pattern ... "); + + /* If we haven't matched the entire string, and we want the + longest match, try backtracking. */ + if (d != end_match_2) + { + /* 1 if this match ends in the same string (string1 or string2) + as the best previous match. */ + boolean same_str_p = (FIRST_STRING_P (match_end) + == MATCHING_IN_FIRST_STRING); + /* 1 if this match is the best seen so far. */ + boolean best_match_p; + + /* AIX compiler got confused when this was combined + with the previous declaration. */ + if (same_str_p) + best_match_p = d > match_end; + else + best_match_p = !MATCHING_IN_FIRST_STRING; + + DEBUG_PRINT1 ("backtracking.\n"); + + if (!FAIL_STACK_EMPTY ()) + { /* More failure points to try. */ + + /* If exceeds best match so far, save it. */ + if (!best_regs_set || best_match_p) + { + best_regs_set = true; + match_end = d; + + DEBUG_PRINT1 ("\nSAVING match as best so far.\n"); + + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + best_regstart[mcnt] = regstart[mcnt]; + best_regend[mcnt] = regend[mcnt]; + } + } + goto fail; + } + + /* If no failure points, don't restore garbage. And if + last match is real best match, don't restore second + best one. */ + else if (best_regs_set && !best_match_p) + { + restore_best_regs: + /* Restore best match. It may happen that `dend == + end_match_1' while the restored d is in string2. + For example, the pattern `x.*y.*z' against the + strings `x-' and `y-z-', if the two strings are + not consecutive in memory. */ + DEBUG_PRINT1 ("Restoring best registers.\n"); + + d = match_end; + dend = ((d >= string1 && d <= end1) + ? end_match_1 : end_match_2); + + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = best_regstart[mcnt]; + regend[mcnt] = best_regend[mcnt]; + } + } + } /* d != end_match_2 */ + + succeed_label: + DEBUG_PRINT1 ("Accepting match.\n"); + + /* If caller wants register contents data back, do it. */ + if (regs && !bufp->no_sub) + { + /* Have the register data arrays been allocated? */ + if (bufp->regs_allocated == REGS_UNALLOCATED) + { /* No. So allocate them with malloc. We need one + extra element beyond `num_regs' for the `-1' marker + GNU code uses. */ + regs->num_regs = MAX (RE_NREGS, num_regs + 1); + regs->start = TALLOC (regs->num_regs, regoff_t); + regs->end = TALLOC (regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + { + FREE_VARIABLES (); + return -2; + } + bufp->regs_allocated = REGS_REALLOCATE; + } + else if (bufp->regs_allocated == REGS_REALLOCATE) + { /* Yes. If we need more elements than were already + allocated, reallocate them. If we need fewer, just + leave it alone. */ + if (regs->num_regs < num_regs + 1) + { + regs->num_regs = num_regs + 1; + RETALLOC (regs->start, regs->num_regs, regoff_t); + RETALLOC (regs->end, regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + { + FREE_VARIABLES (); + return -2; + } + } + } + else + { + /* These braces fend off a "empty body in an else-statement" + warning under GCC when assert expands to nothing. */ + assert (bufp->regs_allocated == REGS_FIXED); + } + + /* Convert the pointer data in `regstart' and `regend' to + indices. Register zero has to be set differently, + since we haven't kept track of any info for it. */ + if (regs->num_regs > 0) + { + regs->start[0] = pos; + regs->end[0] = (MATCHING_IN_FIRST_STRING + ? ((regoff_t) (d - string1)) + : ((regoff_t) (d - string2 + size1))); + } + + /* Go through the first `min (num_regs, regs->num_regs)' + registers, since that is all we initialized. */ + for (mcnt = 1; (unsigned) mcnt < MIN (num_regs, regs->num_regs); + mcnt++) + { + if (REG_UNSET (regstart[mcnt]) || REG_UNSET (regend[mcnt])) + regs->start[mcnt] = regs->end[mcnt] = -1; + else + { + regs->start[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regstart[mcnt]); + regs->end[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regend[mcnt]); + } + } + + /* If the regs structure we return has more elements than + were in the pattern, set the extra elements to -1. If + we (re)allocated the registers, this is the case, + because we always allocate enough to have at least one + -1 at the end. */ + for (mcnt = num_regs; (unsigned) mcnt < regs->num_regs; mcnt++) + regs->start[mcnt] = regs->end[mcnt] = -1; + } /* regs && !bufp->no_sub */ + + DEBUG_PRINT4 ("%u failure points pushed, %u popped (%u remain).\n", + nfailure_points_pushed, nfailure_points_popped, + nfailure_points_pushed - nfailure_points_popped); + DEBUG_PRINT2 ("%u registers pushed.\n", num_regs_pushed); + + mcnt = d - pos - (MATCHING_IN_FIRST_STRING + ? string1 + : string2 - size1); + + DEBUG_PRINT2 ("Returning %d from re_match_2.\n", mcnt); + + FREE_VARIABLES (); + return mcnt; + } + + /* Otherwise match next pattern command. */ + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + /* Ignore these. Used to ignore the n of succeed_n's which + currently have n == 0. */ + case no_op: + DEBUG_PRINT1 ("EXECUTING no_op.\n"); + break; + + case succeed: + DEBUG_PRINT1 ("EXECUTING succeed.\n"); + goto succeed_label; + + /* Match the next n pattern characters exactly. The following + byte in the pattern defines n, and the n bytes after that + are the characters to match. */ + case exactn: + mcnt = *p++; + DEBUG_PRINT2 ("EXECUTING exactn %d.\n", mcnt); + + /* This is written out as an if-else so we don't waste time + testing `translate' inside the loop. */ + if (translate) + { + do + { + PREFETCH (); + if ((unsigned char) translate[(unsigned char) *d++] + != (unsigned char) *p++) + goto fail; + } + while (--mcnt); + } + else + { + do + { + PREFETCH (); + if (*d++ != (char) *p++) goto fail; + } + while (--mcnt); + } + SET_REGS_MATCHED (); + break; + + + /* Match any character except possibly a newline or a null. */ + case anychar: + DEBUG_PRINT1 ("EXECUTING anychar.\n"); + + PREFETCH (); + + if ((!(bufp->syntax & RE_DOT_NEWLINE) && TRANSLATE (*d) == '\n') + || (bufp->syntax & RE_DOT_NOT_NULL && TRANSLATE (*d) == '\000')) + goto fail; + + SET_REGS_MATCHED (); + DEBUG_PRINT2 (" Matched `%d'.\n", *d); + d++; + break; + + + case charset: + case charset_not: + { + register unsigned char c; + boolean not = (re_opcode_t) *(p - 1) == charset_not; + + DEBUG_PRINT2 ("EXECUTING charset%s.\n", not ? "_not" : ""); + + PREFETCH (); + c = TRANSLATE (*d); /* The character to match. */ + + /* Cast to `unsigned' instead of `unsigned char' in case the + bit list is a full 32 bytes long. */ + if (c < (unsigned) (*p * BYTEWIDTH) + && p[1 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + p += 1 + *p; + + if (!not) goto fail; + + SET_REGS_MATCHED (); + d++; + break; + } + + + /* The beginning of a group is represented by start_memory. + The arguments are the register number in the next byte, and the + number of groups inner to this one in the next. The text + matched within the group is recorded (in the internal + registers data structure) under the register number. */ + case start_memory: + DEBUG_PRINT3 ("EXECUTING start_memory %d (%d):\n", *p, p[1]); + + /* Find out if this group can match the empty string. */ + p1 = p; /* To send to group_match_null_string_p. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[*p]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[*p]) + = group_match_null_string_p (&p1, pend, reg_info); + + /* Save the position in the string where we were the last time + we were at this open-group operator in case the group is + operated upon by a repetition operator, e.g., with `(a*)*b' + against `ab'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regstart[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regstart[*p]) ? d : regstart[*p] + : regstart[*p]; + DEBUG_PRINT2 (" old_regstart: %d\n", + POINTER_TO_OFFSET (old_regstart[*p])); + + regstart[*p] = d; + DEBUG_PRINT2 (" regstart: %d\n", POINTER_TO_OFFSET (regstart[*p])); + + IS_ACTIVE (reg_info[*p]) = 1; + MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 0; + + /* This is the new highest active register. */ + highest_active_reg = *p; + + /* If nothing was active before, this is the new lowest active + register. */ + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *p; + + /* Move past the register number and inner group count. */ + p += 2; + just_past_start_mem = p; + + break; + + + /* The stop_memory opcode represents the end of a group. Its + arguments are the same as start_memory's: the register + number, and the number of inner groups. */ + case stop_memory: + DEBUG_PRINT3 ("EXECUTING stop_memory %d (%d):\n", *p, p[1]); + + /* We need to save the string position the last time we were at + this close-group operator in case the group is operated + upon by a repetition operator, e.g., with `((a*)*(b*)*)*' + against `aba'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regend[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regend[*p]) ? d : regend[*p] + : regend[*p]; + DEBUG_PRINT2 (" old_regend: %d\n", + POINTER_TO_OFFSET (old_regend[*p])); + + regend[*p] = d; + DEBUG_PRINT2 (" regend: %d\n", POINTER_TO_OFFSET (regend[*p])); + + /* This register isn't active anymore. */ + IS_ACTIVE (reg_info[*p]) = 0; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 0; + + /* If this was the only register active, nothing is active + anymore. */ + if (lowest_active_reg == highest_active_reg) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + { /* We must scan for the new highest active register, since + it isn't necessarily one less than now: consider + (a(b)c(d(e)f)g). When group 3 ends, after the f), the + new highest active register is 1. */ + unsigned char r = *p - 1; + while (r > 0 && !IS_ACTIVE (reg_info[r])) + r--; + + /* If we end up at register zero, that means that we saved + the registers as the result of an `on_failure_jump', not + a `start_memory', and we jumped to past the innermost + `stop_memory'. For example, in ((.)*) we save + registers 1 and 2 as a result of the *, but when we pop + back to the second ), we are at the stop_memory 1. + Thus, nothing is active. */ + if (r == 0) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + highest_active_reg = r; + } + + /* If just failed to match something this time around with a + group that's operated on by a repetition operator, try to + force exit from the ``loop'', and restore the register + information for this group that we had before trying this + last match. */ + if ((!MATCHED_SOMETHING (reg_info[*p]) + || just_past_start_mem == p - 1) + && (p + 2) < pend) + { + boolean is_a_jump_n = false; + + p1 = p + 2; + mcnt = 0; + switch ((re_opcode_t) *p1++) + { + case jump_n: + is_a_jump_n = true; + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (is_a_jump_n) + p1 += 2; + break; + + default: + /* do nothing */ ; + } + p1 += mcnt; + + /* If the next operation is a jump backwards in the pattern + to an on_failure_jump right before the start_memory + corresponding to this stop_memory, exit from the loop + by forcing a failure after pushing on the stack the + on_failure_jump's jump in the pattern, and d. */ + if (mcnt < 0 && (re_opcode_t) *p1 == on_failure_jump + && (re_opcode_t) p1[3] == start_memory && p1[4] == *p) + { + /* If this group ever matched anything, then restore + what its registers were before trying this last + failed match, e.g., with `(a*)*b' against `ab' for + regstart[1], and, e.g., with `((a*)*(b*)*)*' + against `aba' for regend[3]. + + Also restore the registers for inner groups for, + e.g., `((a*)(b*))*' against `aba' (register 3 would + otherwise get trashed). */ + + if (EVER_MATCHED_SOMETHING (reg_info[*p])) + { + unsigned r; + + EVER_MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Restore this and inner groups' (if any) registers. */ + for (r = *p; r < (unsigned) *p + (unsigned) *(p + 1); + r++) + { + regstart[r] = old_regstart[r]; + + /* xx why this test? */ + if (old_regend[r] >= regstart[r]) + regend[r] = old_regend[r]; + } + } + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + PUSH_FAILURE_POINT (p1 + mcnt, d, -2); + + goto fail; + } + } + + /* Move past the register number and the inner group count. */ + p += 2; + break; + + + /* \<digit> has been turned into a `duplicate' command which is + followed by the numeric value of <digit> as the register number. */ + case duplicate: + { + register const char *d2, *dend2; + int regno = *p++; /* Get which register to match against. */ + DEBUG_PRINT2 ("EXECUTING duplicate %d.\n", regno); + + /* Can't back reference a group which we've never matched. */ + if (REG_UNSET (regstart[regno]) || REG_UNSET (regend[regno])) + goto fail; + + /* Where in input to try to start matching. */ + d2 = regstart[regno]; + + /* Where to stop matching; if both the place to start and + the place to stop matching are in the same string, then + set to the place to stop, otherwise, for now have to use + the end of the first string. */ + + dend2 = ((FIRST_STRING_P (regstart[regno]) + == FIRST_STRING_P (regend[regno])) + ? regend[regno] : end_match_1); + for (;;) + { + /* If necessary, advance to next segment in register + contents. */ + while (d2 == dend2) + { + if (dend2 == end_match_2) break; + if (dend2 == regend[regno]) break; + + /* End of string1 => advance to string2. */ + d2 = string2; + dend2 = regend[regno]; + } + /* At end of register contents => success */ + if (d2 == dend2) break; + + /* If necessary, advance to next segment in data. */ + PREFETCH (); + + /* How many characters left in this segment to match. */ + mcnt = dend - d; + + /* Want how many consecutive characters we can match in + one shot, so, if necessary, adjust the count. */ + if (mcnt > dend2 - d2) + mcnt = dend2 - d2; + + /* Compare that many; failure if mismatch, else move + past them. */ + if (translate + ? bcmp_translate (d, d2, mcnt, translate) + : bcmp (d, d2, mcnt)) + goto fail; + d += mcnt, d2 += mcnt; + + /* Do this because we've match some characters. */ + SET_REGS_MATCHED (); + } + } + break; + + + /* begline matches the empty string at the beginning of the string + (unless `not_bol' is set in `bufp'), and, if + `newline_anchor' is set, after newlines. */ + case begline: + DEBUG_PRINT1 ("EXECUTING begline.\n"); + + if (AT_STRINGS_BEG (d)) + { + if (!bufp->not_bol) break; + } + else if (d[-1] == '\n' && bufp->newline_anchor) + { + break; + } + /* In all other cases, we fail. */ + goto fail; + + + /* endline is the dual of begline. */ + case endline: + DEBUG_PRINT1 ("EXECUTING endline.\n"); + + if (AT_STRINGS_END (d)) + { + if (!bufp->not_eol) break; + } + + /* We have to ``prefetch'' the next character. */ + else if ((d == end1 ? *string2 : *d) == '\n' + && bufp->newline_anchor) + { + break; + } + goto fail; + + + /* Match at the very beginning of the data. */ + case begbuf: + DEBUG_PRINT1 ("EXECUTING begbuf.\n"); + if (AT_STRINGS_BEG (d)) + break; + goto fail; + + + /* Match at the very end of the data. */ + case endbuf: + DEBUG_PRINT1 ("EXECUTING endbuf.\n"); + if (AT_STRINGS_END (d)) + break; + goto fail; + + + /* on_failure_keep_string_jump is used to optimize `.*\n'. It + pushes NULL as the value for the string on the stack. Then + `pop_failure_point' will keep the current value for the + string, instead of restoring it. To see why, consider + matching `foo\nbar' against `.*\n'. The .* matches the foo; + then the . fails against the \n. But the next thing we want + to do is match the \n against the \n; if we restored the + string value, we would be back at the foo. + + Because this is used only in specific cases, we don't need to + check all the things that `on_failure_jump' does, to make + sure the right things get saved on the stack. Hence we don't + share its code. The only reason to push anything on the + stack at all is that otherwise we would have to change + `anychar's code to do something besides goto fail in this + case; that seems worse than this. */ + case on_failure_keep_string_jump: + DEBUG_PRINT1 ("EXECUTING on_failure_keep_string_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p):\n", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x):\n", mcnt, p + mcnt); +#endif + + PUSH_FAILURE_POINT (p + mcnt, NULL, -2); + break; + + + /* Uses of on_failure_jump: + + Each alternative starts with an on_failure_jump that points + to the beginning of the next alternative. Each alternative + except the last ends with a jump that in effect jumps past + the rest of the alternatives. (They really jump to the + ending jump of the following alternative, because tensioning + these jumps is a hassle.) + + Repeats start with an on_failure_jump that points past both + the repetition text and either the following jump or + pop_failure_jump back to this on_failure_jump. */ + case on_failure_jump: + on_failure: + DEBUG_PRINT1 ("EXECUTING on_failure_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p)", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x)", mcnt, p + mcnt); +#endif + + /* If this on_failure_jump comes right before a group (i.e., + the original * applied to a group), save the information + for that group and all inner ones, so that if we fail back + to this point, the group's information will be correct. + For example, in \(a*\)*\1, we need the preceding group, + and in \(zz\(a*\)b*\)\2, we need the inner group. */ + + /* We can't use `p' to check ahead because we push + a failure point to `p + mcnt' after we do this. */ + p1 = p; + + /* We need to skip no_op's before we look for the + start_memory in case this on_failure_jump is happening as + the result of a completed succeed_n, as in \(a\)\{1,3\}b\1 + against aba. */ + while (p1 < pend && (re_opcode_t) *p1 == no_op) + p1++; + + if (p1 < pend && (re_opcode_t) *p1 == start_memory) + { + /* We have a new highest active register now. This will + get reset at the start_memory we are about to get to, + but we will have saved all the registers relevant to + this repetition op, as described above. */ + highest_active_reg = *(p1 + 1) + *(p1 + 2); + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *(p1 + 1); + } + + DEBUG_PRINT1 (":\n"); + PUSH_FAILURE_POINT (p + mcnt, d, -2); + break; + + + /* A smart repeat ends with `maybe_pop_jump'. + We change it to either `pop_failure_jump' or `jump'. */ + case maybe_pop_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT2 ("EXECUTING maybe_pop_jump %d.\n", mcnt); + { + register unsigned char *p2 = p; + + /* Compare the beginning of the repeat with what in the + pattern follows its end. If we can establish that there + is nothing that they would both match, i.e., that we + would have to backtrack because of (as in, e.g., `a*a') + then we can change to pop_failure_jump, because we'll + never have to backtrack. + + This is not true in the case of alternatives: in + `(a|ab)*' we do need to backtrack to the `ab' alternative + (e.g., if the string was `ab'). But instead of trying to + detect that here, the alternative has put on a dummy + failure point which is what we will end up popping. */ + + /* Skip over open/close-group commands. + If what follows this loop is a ...+ construct, + look at what begins its body, since we will have to + match at least one of that. */ + while (1) + { + if (p2 + 2 < pend + && ((re_opcode_t) *p2 == stop_memory + || (re_opcode_t) *p2 == start_memory)) + p2 += 3; + else if (p2 + 6 < pend + && (re_opcode_t) *p2 == dummy_failure_jump) + p2 += 6; + else + break; + } + + p1 = p + mcnt; + /* p1[0] ... p1[2] are the `on_failure_jump' corresponding + to the `maybe_finalize_jump' of this case. Examine what + follows. */ + + /* If we're at the end of the pattern, we can change. */ + if (p2 == pend) + { + /* Consider what happens when matching ":\(.*\)" + against ":/". I don't really understand this code + yet. */ + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 + (" End of pattern: change to `pop_failure_jump'.\n"); + } + + else if ((re_opcode_t) *p2 == exactn + || (bufp->newline_anchor && (re_opcode_t) *p2 == endline)) + { + register unsigned char c + = *p2 == (unsigned char) endline ? '\n' : p2[2]; + + if ((re_opcode_t) p1[3] == exactn && p1[5] != c) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT3 (" %c != %c => pop_failure_jump.\n", + c, p1[5]); + } + + else if ((re_opcode_t) p1[3] == charset + || (re_opcode_t) p1[3] == charset_not) + { + int not = (re_opcode_t) p1[3] == charset_not; + + if (c < (unsigned char) (p1[4] * BYTEWIDTH) + && p1[5 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + /* `not' is equal to 1 if c would match, which means + that we can't change to pop_failure_jump. */ + if (!not) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + } + else if ((re_opcode_t) *p2 == charset) + { +#ifdef DEBUG + register unsigned char c + = *p2 == (unsigned char) endline ? '\n' : p2[2]; +#endif + +#if 0 + if ((re_opcode_t) p1[3] == exactn + && ! ((int) p2[1] * BYTEWIDTH > (int) p1[5] + && (p2[2 + p1[5] / BYTEWIDTH] + & (1 << (p1[5] % BYTEWIDTH))))) +#else + if ((re_opcode_t) p1[3] == exactn + && ! ((int) p2[1] * BYTEWIDTH > (int) p1[4] + && (p2[2 + p1[4] / BYTEWIDTH] + & (1 << (p1[4] % BYTEWIDTH))))) +#endif + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT3 (" %c != %c => pop_failure_jump.\n", + c, p1[5]); + } + + else if ((re_opcode_t) p1[3] == charset_not) + { + int idx; + /* We win if the charset_not inside the loop + lists every character listed in the charset after. */ + for (idx = 0; idx < (int) p2[1]; idx++) + if (! (p2[2 + idx] == 0 + || (idx < (int) p1[4] + && ((p2[2 + idx] & ~ p1[5 + idx]) == 0)))) + break; + + if (idx == p2[1]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + else if ((re_opcode_t) p1[3] == charset) + { + int idx; + /* We win if the charset inside the loop + has no overlap with the one after the loop. */ + for (idx = 0; + idx < (int) p2[1] && idx < (int) p1[4]; + idx++) + if ((p2[2 + idx] & p1[5 + idx]) != 0) + break; + + if (idx == p2[1] || idx == p1[4]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + } + } + p -= 2; /* Point at relative address again. */ + if ((re_opcode_t) p[-1] != pop_failure_jump) + { + p[-1] = (unsigned char) jump; + DEBUG_PRINT1 (" Match => jump.\n"); + goto unconditional_jump; + } + /* Note fall through. */ + + + /* The end of a simple repeat has a pop_failure_jump back to + its matching on_failure_jump, where the latter will push a + failure point. The pop_failure_jump takes off failure + points put on by this pop_failure_jump's matching + on_failure_jump; we got through the pattern to here from the + matching on_failure_jump, so didn't fail. */ + case pop_failure_jump: + { + /* We need to pass separate storage for the lowest and + highest registers, even though we don't care about the + actual values. Otherwise, we will restore only one + register from the stack, since lowest will == highest in + `pop_failure_point'. */ + active_reg_t dummy_low_reg, dummy_high_reg; + unsigned char *pdummy; + const char *sdummy; + + DEBUG_PRINT1 ("EXECUTING pop_failure_jump.\n"); + POP_FAILURE_POINT (sdummy, pdummy, + dummy_low_reg, dummy_high_reg, + reg_dummy, reg_dummy, reg_info_dummy); + } + /* Note fall through. */ + + unconditional_jump: +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + /* Note fall through. */ + + /* Unconditionally jump (without popping any failure points). */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p); /* Get the amount to jump. */ + DEBUG_PRINT2 ("EXECUTING jump %d ", mcnt); + p += mcnt; /* Do the jump. */ +#ifdef _LIBC + DEBUG_PRINT2 ("(to %p).\n", p); +#else + DEBUG_PRINT2 ("(to 0x%x).\n", p); +#endif + break; + + + /* We need this opcode so we can detect where alternatives end + in `group_match_null_string_p' et al. */ + case jump_past_alt: + DEBUG_PRINT1 ("EXECUTING jump_past_alt.\n"); + goto unconditional_jump; + + + /* Normally, the on_failure_jump pushes a failure point, which + then gets popped at pop_failure_jump. We will end up at + pop_failure_jump, also, and with a pattern of, say, `a+', we + are skipping over the on_failure_jump, so we have to push + something meaningless for pop_failure_jump to pop. */ + case dummy_failure_jump: + DEBUG_PRINT1 ("EXECUTING dummy_failure_jump.\n"); + /* It doesn't matter what we push for the string here. What + the code at `fail' tests is the value for the pattern. */ + PUSH_FAILURE_POINT (0, 0, -2); + goto unconditional_jump; + + + /* At the end of an alternative, we need to push a dummy failure + point in case we are followed by a `pop_failure_jump', because + we don't want the failure point for the alternative to be + popped. For example, matching `(a|ab)*' against `aab' + requires that we match the `ab' alternative. */ + case push_dummy_failure: + DEBUG_PRINT1 ("EXECUTING push_dummy_failure.\n"); + /* See comments just above at `dummy_failure_jump' about the + two zeroes. */ + PUSH_FAILURE_POINT (0, 0, -2); + break; + + /* Have to succeed matching what follows at least n times. + After that, handle like `on_failure_jump'. */ + case succeed_n: + EXTRACT_NUMBER (mcnt, p + 2); + DEBUG_PRINT2 ("EXECUTING succeed_n %d.\n", mcnt); + + assert (mcnt >= 0); + /* Originally, this is how many times we HAVE to succeed. */ + if (mcnt > 0) + { + mcnt--; + p += 2; + STORE_NUMBER_AND_INCR (p, mcnt); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p - 2, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p - 2, mcnt); +#endif + } + else if (mcnt == 0) + { +#ifdef _LIBC + DEBUG_PRINT2 (" Setting two bytes from %p to no_op.\n", p+2); +#else + DEBUG_PRINT2 (" Setting two bytes from 0x%x to no_op.\n", p+2); +#endif + p[2] = (unsigned char) no_op; + p[3] = (unsigned char) no_op; + goto on_failure; + } + break; + + case jump_n: + EXTRACT_NUMBER (mcnt, p + 2); + DEBUG_PRINT2 ("EXECUTING jump_n %d.\n", mcnt); + + /* Originally, this is how many times we CAN jump. */ + if (mcnt) + { + mcnt--; + STORE_NUMBER (p + 2, mcnt); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p + 2, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p + 2, mcnt); +#endif + goto unconditional_jump; + } + /* If don't have to jump any more, skip over the rest of command. */ + else + p += 4; + break; + + case set_number_at: + { + DEBUG_PRINT1 ("EXECUTING set_number_at.\n"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); + p1 = p + mcnt; + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p1, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p1, mcnt); +#endif + STORE_NUMBER (p1, mcnt); + break; + } + +#if 0 + /* The DEC Alpha C compiler 3.x generates incorrect code for the + test WORDCHAR_P (d - 1) != WORDCHAR_P (d) in the expansion of + AT_WORD_BOUNDARY, so this code is disabled. Expanding the + macro and introducing temporary variables works around the bug. */ + + case wordbound: + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + break; + goto fail; + + case notwordbound: + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + goto fail; + break; +#else + case wordbound: + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + break; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + break; + goto fail; + } + + case notwordbound: + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + goto fail; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + goto fail; + break; + } +#endif + + case wordbeg: + DEBUG_PRINT1 ("EXECUTING wordbeg.\n"); + if (WORDCHAR_P (d) && (AT_STRINGS_BEG (d) || !WORDCHAR_P (d - 1))) + break; + goto fail; + + case wordend: + DEBUG_PRINT1 ("EXECUTING wordend.\n"); + if (!AT_STRINGS_BEG (d) && WORDCHAR_P (d - 1) + && (!WORDCHAR_P (d) || AT_STRINGS_END (d))) + break; + goto fail; + +#ifdef emacs + case before_dot: + DEBUG_PRINT1 ("EXECUTING before_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) >= point) + goto fail; + break; + + case at_dot: + DEBUG_PRINT1 ("EXECUTING at_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) != point) + goto fail; + break; + + case after_dot: + DEBUG_PRINT1 ("EXECUTING after_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) <= point) + goto fail; + break; + + case syntaxspec: + DEBUG_PRINT2 ("EXECUTING syntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchsyntax; + + case wordchar: + DEBUG_PRINT1 ("EXECUTING Emacs wordchar.\n"); + mcnt = (int) Sword; + matchsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) != (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + break; + + case notsyntaxspec: + DEBUG_PRINT2 ("EXECUTING notsyntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchnotsyntax; + + case notwordchar: + DEBUG_PRINT1 ("EXECUTING Emacs notwordchar.\n"); + mcnt = (int) Sword; + matchnotsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) == (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + break; + +#else /* not emacs */ + case wordchar: + DEBUG_PRINT1 ("EXECUTING non-Emacs wordchar.\n"); + PREFETCH (); + if (!WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + break; + + case notwordchar: + DEBUG_PRINT1 ("EXECUTING non-Emacs notwordchar.\n"); + PREFETCH (); + if (WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + break; +#endif /* not emacs */ + + default: + abort (); + } + continue; /* Successfully executed one pattern command; keep going. */ + + + /* We goto here if a matching operation fails. */ + fail: + if (!FAIL_STACK_EMPTY ()) + { /* A restart point is known. Restore to that state. */ + DEBUG_PRINT1 ("\nFAIL:\n"); + POP_FAILURE_POINT (d, p, + lowest_active_reg, highest_active_reg, + regstart, regend, reg_info); + + /* If this failure point is a dummy, try the next one. */ + if (!p) + goto fail; + + /* If we failed to the end of the pattern, don't examine *p. */ + assert (p <= pend); + if (p < pend) + { + boolean is_a_jump_n = false; + + /* If failed to a backwards jump that's part of a repetition + loop, need to pop this failure point and use the next one. */ + switch ((re_opcode_t) *p) + { + case jump_n: + is_a_jump_n = true; + case maybe_pop_jump: + case pop_failure_jump: + case jump: + p1 = p + 1; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + + if ((is_a_jump_n && (re_opcode_t) *p1 == succeed_n) + || (!is_a_jump_n + && (re_opcode_t) *p1 == on_failure_jump)) + goto fail; + break; + default: + /* do nothing */ ; + } + } + + if (d >= string1 && d <= end1) + dend = end_match_1; + } + else + break; /* Matching at this starting point really fails. */ + } /* for (;;) */ + + if (best_regs_set) + goto restore_best_regs; + + FREE_VARIABLES (); + + return -1; /* Failure to match. */ +} /* re_match_2 */ + +/* Subroutine definitions for re_match_2. */ + + +/* We are passed P pointing to a register number after a start_memory. + + Return true if the pattern up to the corresponding stop_memory can + match the empty string, and false otherwise. + + If we find the matching stop_memory, sets P to point to one past its number. + Otherwise, sets P to an undefined byte less than or equal to END. + + We don't handle duplicates properly (yet). */ + +static boolean +group_match_null_string_p(unsigned char **p, + unsigned char *end, + register_info_type *reg_info) +{ + int mcnt; + /* Point to after the args to the start_memory. */ + unsigned char *p1 = *p + 2; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and return true or + false, as appropriate, when we get to one that can't, or to the + matching stop_memory. */ + + switch ((re_opcode_t) *p1) + { + /* Could be either a loop or a series of alternatives. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + /* If the next operation is not a jump backwards in the + pattern. */ + + if (mcnt >= 0) + { + /* Go through the on_failure_jumps of the alternatives, + seeing if any of the alternatives cannot match nothing. + The last alternative starts with only a jump, + whereas the rest start with on_failure_jump and end + with a jump, e.g., here is the pattern for `a|b|c': + + /on_failure_jump/0/6/exactn/1/a/jump_past_alt/0/6 + /on_failure_jump/0/6/exactn/1/b/jump_past_alt/0/3 + /exactn/1/c + + So, we have to first go through the first (n-1) + alternatives and then deal with the last one separately. */ + + + /* Deal with the first (n-1) alternatives, which start + with an on_failure_jump (see above) that jumps to right + past a jump_past_alt. */ + + while ((re_opcode_t) p1[mcnt-3] == jump_past_alt) + { + /* `mcnt' holds how many bytes long the alternative + is, including the ending `jump_past_alt' and + its number. */ + + if (!alt_match_null_string_p (p1, p1 + mcnt - 3, + reg_info)) + return false; + + /* Move to right after this alternative, including the + jump_past_alt. */ + p1 += mcnt; + + /* Break if it's the beginning of an n-th alternative + that doesn't begin with an on_failure_jump. */ + if ((re_opcode_t) *p1 != on_failure_jump) + break; + + /* Still have to check that it's not an n-th + alternative that starts with an on_failure_jump. */ + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if ((re_opcode_t) p1[mcnt-3] != jump_past_alt) + { + /* Get to the beginning of the n-th alternative. */ + p1 -= 3; + break; + } + } + + /* Deal with the last alternative: go back and get number + of the `jump_past_alt' just before it. `mcnt' contains + the length of the alternative. */ + EXTRACT_NUMBER (mcnt, p1 - 2); + + if (!alt_match_null_string_p (p1, p1 + mcnt, reg_info)) + return false; + + p1 += mcnt; /* Get past the n-th alternative. */ + } /* if mcnt > 0 */ + break; + + + case stop_memory: + assert (p1[1] == **p); + *p = p1 + 2; + return true; + + + default: + if (!common_op_match_null_string_p (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return false; +} /* group_match_null_string_p */ + + +/* Similar to group_match_null_string_p, but doesn't deal with alternatives: + It expects P to be the first byte of a single alternative and END one + byte past the last. The alternative can contain groups. */ + +static boolean +alt_match_null_string_p(unsigned char *p, + unsigned char *end, + register_info_type *reg_info) +{ + int mcnt; + unsigned char *p1 = p; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and break when we get + to one that can't. */ + + switch ((re_opcode_t) *p1) + { + /* It's a loop. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + break; + + default: + if (!common_op_match_null_string_p (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return true; +} /* alt_match_null_string_p */ + + +/* Deals with the ops common to group_match_null_string_p and + alt_match_null_string_p. + + Sets P to one after the op and its arguments, if any. */ + +static boolean +common_op_match_null_string_p(unsigned char **p, + unsigned char *end, + register_info_type *reg_info) +{ + int mcnt; + boolean ret; + int reg_no; + unsigned char *p1 = *p; + + switch ((re_opcode_t) *p1++) + { + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbeg: + case wordend: + case wordbound: + case notwordbound: +#ifdef emacs + case before_dot: + case at_dot: + case after_dot: +#endif + break; + + case start_memory: + reg_no = *p1; + assert (reg_no > 0 && reg_no <= MAX_REGNUM); + ret = group_match_null_string_p (&p1, end, reg_info); + + /* Have to set this here in case we're checking a group which + contains a group and a back reference to it. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[reg_no]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[reg_no]) = ret; + + if (!ret) + return false; + break; + + /* If this is an optimized succeed_n for zero times, make the jump. */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (mcnt >= 0) + p1 += mcnt; + else + return false; + break; + + case succeed_n: + /* Get to the number of times to succeed. */ + p1 += 2; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + if (mcnt == 0) + { + p1 -= 4; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + } + else + return false; + break; + + case duplicate: + if (!REG_MATCH_NULL_STRING_P (reg_info[*p1])) + return false; + break; + + case set_number_at: + p1 += 4; + + default: + /* All other opcodes mean we cannot match the empty string. */ + return false; + } + + *p = p1; + return true; +} /* common_op_match_null_string_p */ + + +/* Return zero if TRANSLATE[S1] and TRANSLATE[S2] are identical for LEN + bytes; nonzero otherwise. */ + +static int +bcmp_translate(const char *s1, + const char *s2, + register int len, + RE_TRANSLATE_TYPE translate) +{ + register const unsigned char *p1 = (const unsigned char *) s1; + register const unsigned char *p2 = (const unsigned char *) s2; + while (len) + { + if (translate[*p1++] != translate[*p2++]) return 1; + len--; + } + return 0; +} + +/* Entry points for GNU code. */ + +/* re_compile_pattern is the GNU regular expression compiler: it + compiles PATTERN (of length SIZE) and puts the result in BUFP. + Returns 0 if the pattern was valid, otherwise an error string. + + Assumes the `allocated' (and perhaps `buffer') and `translate' fields + are set in BUFP on entry. + + We call regex_compile to do the actual compilation. */ + +const char * +re_compile_pattern(const char *pattern, + size_t length, + struct re_pattern_buffer *bufp) +{ + reg_errcode_t ret; + + /* GNU code is written to assume at least RE_NREGS registers will be set + (and at least one extra will be -1). */ + bufp->regs_allocated = REGS_UNALLOCATED; + + /* And GNU code determines whether or not to get register information + by passing null for the REGS argument to re_match, etc., not by + setting no_sub. */ + bufp->no_sub = 0; + + /* Match anchors at newline. */ + bufp->newline_anchor = 1; + + ret = regex_compile (pattern, length, re_syntax_options, bufp); + + if (!ret) + return NULL; + return gettext (re_error_msgid[(int) ret]); +} + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined (_REGEX_RE_COMP) || defined (_LIBC) + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +#ifdef _LIBC +/* Make these definitions weak in libc, so POSIX programs can redefine + these names if they don't use our functions, and still use + regcomp/regexec below without link errors. */ +weak_function +#endif +re_comp (s) + const char *s; +{ + reg_errcode_t ret; + + if (!s) + { + if (!re_comp_buf.buffer) + return gettext ("No previous regular expression"); + return 0; + } + + if (!re_comp_buf.buffer) + { + re_comp_buf.buffer = (unsigned char *) malloc (200); + if (re_comp_buf.buffer == NULL) + return gettext (re_error_msgid[(int) REG_ESPACE]); + re_comp_buf.allocated = 200; + + re_comp_buf.fastmap = (char *) malloc (1 << BYTEWIDTH); + if (re_comp_buf.fastmap == NULL) + return gettext (re_error_msgid[(int) REG_ESPACE]); + } + + /* Since `re_exec' always passes NULL for the `regs' argument, we + don't need to initialize the pattern buffer fields which affect it. */ + + /* Match anchors at newlines. */ + re_comp_buf.newline_anchor = 1; + + ret = regex_compile (s, strlen (s), re_syntax_options, &re_comp_buf); + + if (!ret) + return NULL; + + /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */ + return (char *) gettext (re_error_msgid[(int) ret]); +} + + +int +#ifdef _LIBC +weak_function +#endif +re_exec (s) + const char *s; +{ + const int len = strlen (s); + return + 0 <= re_search (&re_comp_buf, s, len, 0, len, (struct re_registers *) 0); +} + +#endif /* _REGEX_RE_COMP */ + +/* POSIX.2 functions. Don't define these for Emacs. */ + +#ifndef emacs + +/* regcomp takes a regular expression as a string and compiles it. + + PREG is a regex_t *. We do not expect any fields to be initialized, + since POSIX says we shouldn't. Thus, we set + + `buffer' to the compiled pattern; + `used' to the length of the compiled pattern; + `syntax' to RE_SYNTAX_POSIX_EXTENDED if the + REG_EXTENDED bit in CFLAGS is set; otherwise, to + RE_SYNTAX_POSIX_BASIC; + `newline_anchor' to REG_NEWLINE being set in CFLAGS; + `fastmap' and `fastmap_accurate' to zero; + `re_nsub' to the number of subexpressions in PATTERN. + + PATTERN is the address of the pattern string. + + CFLAGS is a series of bits which affect compilation. + + If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we + use POSIX basic syntax. + + If REG_NEWLINE is set, then . and [^...] don't match newline. + Also, regexec will try a match beginning after every newline. + + If REG_ICASE is set, then we considers upper- and lowercase + versions of letters to be equivalent when matching. + + If REG_NOSUB is set, then when PREG is passed to regexec, that + routine will report only success or failure, and nothing about the + registers. + + It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for + the return codes and their meanings.) */ + +int +regcomp(regex_t *preg, + const char *pattern, + int cflags) +{ + reg_errcode_t ret; + reg_syntax_t syntax + = (cflags & REG_EXTENDED) ? + RE_SYNTAX_POSIX_EXTENDED : RE_SYNTAX_POSIX_BASIC; + + /* regex_compile will allocate the space for the compiled pattern. */ + preg->buffer = 0; + preg->allocated = 0; + preg->used = 0; + + /* Don't bother to use a fastmap when searching. This simplifies the + REG_NEWLINE case: if we used a fastmap, we'd have to put all the + characters after newlines into the fastmap. This way, we just try + every character. */ + preg->fastmap = 0; + + if (cflags & REG_ICASE) + { + unsigned i; + + preg->translate + = (RE_TRANSLATE_TYPE) malloc (CHAR_SET_SIZE + * sizeof (*(RE_TRANSLATE_TYPE)0)); + if (preg->translate == NULL) + return (int) REG_ESPACE; + + /* Map uppercase characters to corresponding lowercase ones. */ + for (i = 0; i < CHAR_SET_SIZE; i++) + preg->translate[i] = ISUPPER (i) ? tolower (i) : i; + } + else + preg->translate = NULL; + + /* If REG_NEWLINE is set, newlines are treated differently. */ + if (cflags & REG_NEWLINE) + { /* REG_NEWLINE implies neither . nor [^...] match newline. */ + syntax &= ~RE_DOT_NEWLINE; + syntax |= RE_HAT_LISTS_NOT_NEWLINE; + /* It also changes the matching behavior. */ + preg->newline_anchor = 1; + } + else + preg->newline_anchor = 0; + + preg->no_sub = !!(cflags & REG_NOSUB); + + /* POSIX says a null character in the pattern terminates it, so we + can use strlen here in compiling the pattern. */ + ret = regex_compile (pattern, strlen (pattern), syntax, preg); + + /* POSIX doesn't distinguish between an unmatched open-group and an + unmatched close-group: both are REG_EPAREN. */ + if (ret == REG_ERPAREN) ret = REG_EPAREN; + + return (int) ret; +} + + +/* regexec searches for a given pattern, specified by PREG, in the + string STRING. + + If NMATCH is zero or REG_NOSUB was set in the cflags argument to + `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at + least NMATCH elements, and we set them to the offsets of the + corresponding matched substrings. + + EFLAGS specifies `execution flags' which affect matching: if + REG_NOTBOL is set, then ^ does not match at the beginning of the + string; if REG_NOTEOL is set, then $ does not match at the end. + + We return 0 if we find a match and REG_NOMATCH if not. */ + +int +regexec(const regex_t *preg, + const char *string, + size_t nmatch, + regmatch_t pmatch[], + int eflags) +{ + int ret; + struct re_registers regs; + regex_t private_preg; + int len = strlen (string); + boolean want_reg_info = !preg->no_sub && nmatch > 0; + + private_preg = *preg; + + private_preg.not_bol = !!(eflags & REG_NOTBOL); + private_preg.not_eol = !!(eflags & REG_NOTEOL); + + /* The user has told us exactly how many registers to return + information about, via `nmatch'. We have to pass that on to the + matching routines. */ + private_preg.regs_allocated = REGS_FIXED; + + if (want_reg_info) + { + regs.num_regs = nmatch; + regs.start = TALLOC (nmatch, regoff_t); + regs.end = TALLOC (nmatch, regoff_t); + if (regs.start == NULL || regs.end == NULL) + return (int) REG_NOMATCH; + } + + /* Perform the searching operation. */ + ret = re_search (&private_preg, string, len, + /* start: */ 0, /* range: */ len, + want_reg_info ? ®s : (struct re_registers *) 0); + + /* Copy the register information to the POSIX structure. */ + if (want_reg_info) + { + if (ret >= 0) + { + unsigned r; + + for (r = 0; r < nmatch; r++) + { + pmatch[r].rm_so = regs.start[r]; + pmatch[r].rm_eo = regs.end[r]; + } + } + + /* If we needed the temporary register info, free the space now. */ + free (regs.start); + free (regs.end); + } + + /* We want zero return to mean success, unlike `re_search'. */ + return ret >= 0 ? (int) REG_NOERROR : (int) REG_NOMATCH; +} + + +/* Returns a message corresponding to an error code, ERRCODE, returned + from either regcomp or regexec. We don't use PREG here. */ + +size_t +regerror(int errcode, + const regex_t *preg, + char *errbuf, + size_t errbuf_size) +{ + const char *msg; + size_t msg_size; + + if (errcode < 0 + || errcode >= (int) (sizeof (re_error_msgid) + / sizeof (re_error_msgid[0]))) + /* Only error codes returned by the rest of the code should be passed + to this routine. If we are given anything else, or if other regex + code generates an invalid error code, then the program has a bug. + Dump core so we can fix it. */ + abort (); + + msg = gettext (re_error_msgid[errcode]); + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (errbuf_size != 0) + { + if (msg_size > errbuf_size) + { + strncpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; + } + else + strcpy (errbuf, msg); + } + + return msg_size; +} + + +/* Free dynamically allocated space used by PREG. */ + +void +regfree(regex_t *preg) +{ + if (preg->buffer != NULL) + free (preg->buffer); + preg->buffer = NULL; + + preg->allocated = 0; + preg->used = 0; + + if (preg->fastmap != NULL) + free (preg->fastmap); + preg->fastmap = NULL; + preg->fastmap_accurate = 0; + + if (preg->translate != NULL) + free (preg->translate); + preg->translate = NULL; +} + +#endif /* not emacs */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/regex_win32.h b/debian/htdig/htdig-3.2.0b6/htlib/regex_win32.h new file mode 100644 index 00000000..8471cba0 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/regex_win32.h @@ -0,0 +1,548 @@ +/* LGPLd GNU regex for Native WIN32 */ + +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 2003 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + +/* Added June 2003 Neal Richter, RightNow Technologies */ + +/* note that this version is significantly different from the original */ +/* version 0.12 GNU source code. It compiles and works on Native WIN32. */ + +/* Definitions for data structures and routines for the regular + expression library, version 0.12. + Copyright (C) 1985,89,90,91,92,93,95,96,97 Free Software Foundation, Inc. + + This file is part of the GNU C Library. Its master source is NOT part of + the C library, however. The master source lives in /gd/gnu/lib. + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + +#ifndef __REGEXP_LIBRARY_H__ +#define __REGEXP_LIBRARY_H__ + +/* Allow the use in C++ code. */ +#ifdef __cplusplus +extern "C" { +#endif + +#include <stdlib.h> + /* POSIX says that <sys/types.h> must be included (by the caller) before + <regex.h>. */ + +#if !defined (_POSIX_C_SOURCE) && !defined (_POSIX_SOURCE) && defined (VMS) + /* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it + should be there. */ +#include <stddef.h> +#endif + + /* The following two types have to be signed and unsigned integer type + wide enough to hold a value of a pointer. For most ANSI compilers + ptrdiff_t and size_t should be likely OK. Still size of these two + types is 2 for Microsoft C. Ugh... */ + typedef long int s_reg_t; + typedef unsigned long int active_reg_t; + + /* The following bits are used to determine the regexp syntax we + recognize. The set/not-set meanings are chosen so that Emacs syntax + remains the value 0. The bits are given in alphabetical order, and + the definitions shifted by one from the previous bit; thus, when we + add or remove a bit, only one other definition need change. */ + typedef unsigned long int reg_syntax_t; + + /* If this bit is not set, then \ inside a bracket expression is literal. + If set, then such a \ quotes the following character. */ +#define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) + + /* If this bit is not set, then + and ? are operators, and \+ and \? are + literals. + If set, then \+ and \? are operators and + and ? are literals. */ +#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) + + /* If this bit is set, then character classes are supported. They are: + [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], + [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. + If not set, then character classes are not supported. */ +#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) + + /* If this bit is set, then ^ and $ are always anchors (outside bracket + expressions, of course). + If this bit is not set, then it depends: + ^ is an anchor if it is at the beginning of a regular + expression or after an open-group or an alternation operator; + $ is an anchor if it is at the end of a regular expression, or + before a close-group or an alternation operator. + + This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because + POSIX draft 11.2 says that * etc. in leading positions is undefined. + We already implemented a previous draft which made those constructs + invalid, though, so we haven't changed the code back. */ +#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) + + /* If this bit is set, then special characters are always special + regardless of where they are in the pattern. + If this bit is not set, then special characters are special only in + some contexts; otherwise they are ordinary. Specifically, + * + ? and intervals are only special when not after the beginning, + open-group, or alternation operator. */ +#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) + + /* If this bit is set, then *, +, ?, and { cannot be first in an re or + immediately after an alternation or begin-group operator. */ +#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) + + /* If this bit is set, then . matches newline. + If not set, then it doesn't. */ +#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) + + /* If this bit is set, then . doesn't match NUL. + If not set, then it does. */ +#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) + + /* If this bit is set, nonmatching lists [^...] do not match newline. + If not set, they do. */ +#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) + + /* If this bit is set, either \{...\} or {...} defines an + interval, depending on RE_NO_BK_BRACES. + If not set, \{, \}, {, and } are literals. */ +#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) + + /* If this bit is set, +, ? and | aren't recognized as operators. + If not set, they are. */ +#define RE_LIMITED_OPS (RE_INTERVALS << 1) + + /* If this bit is set, newline is an alternation operator. + If not set, newline is literal. */ +#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) + + /* If this bit is set, then `{...}' defines an interval, and \{ and \} + are literals. + If not set, then `\{...\}' defines an interval. */ +#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) + + /* If this bit is set, (...) defines a group, and \( and \) are literals. + If not set, \(...\) defines a group, and ( and ) are literals. */ +#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) + + /* If this bit is set, then \<digit> matches <digit>. + If not set, then \<digit> is a back-reference. */ +#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) + + /* If this bit is set, then | is an alternation operator, and \| is literal. + If not set, then \| is an alternation operator, and | is literal. */ +#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) + + /* If this bit is set, then an ending range point collating higher + than the starting range point, as in [z-a], is invalid. + If not set, then when ending range point collates higher than the + starting range point, the range is ignored. */ +#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) + + /* If this bit is set, then an unmatched ) is ordinary. + If not set, then an unmatched ) is invalid. */ +#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) + + /* If this bit is set, succeed as soon as we match the whole pattern, + without further backtracking. */ +#define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) + + /* If this bit is set, do not process the GNU regex operators. + If not set, then the GNU regex operators are recognized. */ +#define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) + + /* If this bit is set, turn on internal regex debugging. + If not set, and debugging was on, turn it off. + This only works if regex.c is compiled -DDEBUG. + We define this bit always, so that all that's needed to turn on + debugging is to recompile regex.c; the calling code can always have + this bit set, and it won't affect anything in the normal case. */ +#define RE_DEBUG (RE_NO_GNU_OPS << 1) + + /* This global variable defines the particular regexp syntax to use (for + some interfaces). When a regexp is compiled, the syntax used is + stored in the pattern buffer, so changing this does not affect + already-compiled regexps. */ + extern reg_syntax_t re_syntax_options; + + /* Define combinations of the above bits for the standard possibilities. + (The [[[ comments delimit what gets put into the Texinfo file, so + don't delete them!) */ + /* [[[begin syntaxes]]] */ +#define RE_SYNTAX_EMACS 0 + +#define RE_SYNTAX_AWK \ + (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ + | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ + | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GNU_AWK \ + ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ + & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS)) + +#define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ + | RE_INTERVALS | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GREP \ + (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ + | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ + | RE_NEWLINE_ALT) + +#define RE_SYNTAX_EGREP \ + (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ + | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ + | RE_NO_BK_VBAR) + +#define RE_SYNTAX_POSIX_EGREP \ + (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES) + + /* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ +#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC + +#define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC + + /* Syntax bits common to both basic and extended POSIX regex syntax. */ +#define _RE_SYNTAX_POSIX_COMMON \ + (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ + | RE_INTERVALS | RE_NO_EMPTY_RANGES) + +#define RE_SYNTAX_POSIX_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) + + /* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes + RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this + isn't minimal, since other operators, such as \`, aren't disabled. */ +#define RE_SYNTAX_POSIX_MINIMAL_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) + +#define RE_SYNTAX_POSIX_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ + | RE_UNMATCHED_RIGHT_PAREN_ORD) + + /* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS + replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */ +#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) + /* [[[end syntaxes]]] */ + + /* Maximum number of duplicates an interval can allow. Some systems + (erroneously) define this in other header files, but we want our + value, so remove any previous define. */ +#ifdef RE_DUP_MAX +#undef RE_DUP_MAX +#endif + /* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */ +#define RE_DUP_MAX (0x7fff) + + + /* POSIX `cflags' bits (i.e., information for `regcomp'). */ + + /* If this bit is set, then use extended regular expression syntax. + If not set, then use basic regular expression syntax. */ +#define REG_EXTENDED 1 + + /* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +#define REG_ICASE (REG_EXTENDED << 1) + + /* If this bit is set, then anchors do not match at newline + characters in the string. + If not set, then anchors do match at newlines. */ +#define REG_NEWLINE (REG_ICASE << 1) + + /* If this bit is set, then report only success or fail in regexec. + If not set, then returns differ between not matching and errors. */ +#define REG_NOSUB (REG_NEWLINE << 1) + + + /* POSIX `eflags' bits (i.e., information for regexec). */ + + /* If this bit is set, then the beginning-of-line operator doesn't match + the beginning of the string (presumably because it's not the + beginning of a line). + If not set, then the beginning-of-line operator does match the + beginning of the string. */ +#define REG_NOTBOL 1 + + /* Like REG_NOTBOL, except for the end-of-line. */ +#define REG_NOTEOL (1 << 1) + + + /* If any error codes are removed, changed, or added, update the + `re_error_msg' table in regex.c. */ + typedef enum + { + REG_NOERROR = 0, /* Success. */ + REG_NOMATCH, /* Didn't find a match (for regexec). */ + + /* POSIX regcomp return error codes. (In the order listed in the + standard.) */ + REG_BADPAT, /* Invalid pattern. */ + REG_ECOLLATE, /* Not implemented. */ + REG_ECTYPE, /* Invalid character class name. */ + REG_EESCAPE, /* Trailing backslash. */ + REG_ESUBREG, /* Invalid back reference. */ + REG_EBRACK, /* Unmatched left bracket. */ + REG_EPAREN, /* Parenthesis imbalance. */ + REG_EBRACE, /* Unmatched \{. */ + REG_BADBR, /* Invalid contents of \{\}. */ + REG_ERANGE, /* Invalid range end. */ + REG_ESPACE, /* Ran out of memory. */ + REG_BADRPT, /* No preceding re for repetition op. */ + + /* Error codes we've added. */ + REG_EEND, /* Premature end. */ + REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ + REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ + } reg_errcode_t; + + /* This data structure represents a compiled pattern. Before calling + the pattern compiler, the fields `buffer', `allocated', `fastmap', + `translate', and `no_sub' can be set. After the pattern has been + compiled, the `re_nsub' field is available. All other fields are + private to the regex routines. */ + +#ifndef RE_TRANSLATE_TYPE +#define RE_TRANSLATE_TYPE char * +#endif + + struct re_pattern_buffer + { + /* [[[begin pattern_buffer]]] */ + /* Space that holds the compiled pattern. It is declared as + `unsigned char *' because its elements are + sometimes used as array indexes. */ + unsigned char *buffer; + + /* Number of bytes to which `buffer' points. */ + unsigned long int allocated; + + /* Number of bytes actually used in `buffer'. */ + unsigned long int used; + + /* Syntax setting with which the pattern was compiled. */ + reg_syntax_t syntax; + + /* Pointer to a fastmap, if any, otherwise zero. re_search uses + the fastmap, if there is one, to skip over impossible + starting points for matches. */ + char *fastmap; + + /* Either a translate table to apply to all characters before + comparing them, or zero for no translation. The translation + is applied to a pattern when it is compiled and to a string + when it is matched. */ + RE_TRANSLATE_TYPE translate; + + /* Number of subexpressions found by the compiler. */ + size_t re_nsub; + + /* Zero if this pattern cannot match the empty string, one else. + Well, in truth it's used only in `re_search_2', to see + whether or not we should use the fastmap, so we don't set + this absolutely perfectly; see `re_compile_fastmap' (the + `duplicate' case). */ + unsigned can_be_null : 1; + + /* If REGS_UNALLOCATED, allocate space in the `regs' structure + for `max (RE_NREGS, re_nsub + 1)' groups. + If REGS_REALLOCATE, reallocate space if necessary. + If REGS_FIXED, use what's there. */ +#define REGS_UNALLOCATED 0 +#define REGS_REALLOCATE 1 +#define REGS_FIXED 2 + unsigned regs_allocated : 2; + + /* Set to zero when `regex_compile' compiles a pattern; set to one + by `re_compile_fastmap' if it updates the fastmap. */ + unsigned fastmap_accurate : 1; + + /* If set, `re_match_2' does not return information about + subexpressions. */ + unsigned no_sub : 1; + + /* If set, a beginning-of-line anchor doesn't match at the + beginning of the string. */ + unsigned not_bol : 1; + + /* Similarly for an end-of-line anchor. */ + unsigned not_eol : 1; + + /* If true, an anchor at a newline matches. */ + unsigned newline_anchor : 1; + + /* [[[end pattern_buffer]]] */ + }; + + typedef struct re_pattern_buffer regex_t; + + /* Type for byte offsets within the string. POSIX mandates this. */ + typedef int regoff_t; + + + /* This is the structure we store register match data in. See + regex.texinfo for a full description of what registers match. */ + struct re_registers + { + unsigned num_regs; + regoff_t *start; + regoff_t *end; + }; + + + /* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, + `re_match_2' returns information about at least this many registers + the first time a `regs' structure is passed. */ +#ifndef RE_NREGS +#define RE_NREGS 30 +#endif + + + /* POSIX specification for registers. Aside from the different names than + `re_registers', POSIX uses an array of structures, instead of a + structure of arrays. */ + typedef struct + { + regoff_t rm_so; /* Byte offset from string's start to substring's start. */ + regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ + } regmatch_t; + + /* Declarations for routines. */ + + /* To avoid duplicating every routine declaration -- once with a + prototype (if we are ANSI), and once without (if we aren't) -- we + use the following macro to declare argument types. This + unfortunately clutters up the declarations a bit, but I think it's + worth it. */ + +#if defined(__STDC__) || defined(_WIN32) +#define _RE_ARGS(args) args + +#else /* not __STDC__ */ + +#define _RE_ARGS(args) () + +#endif /* not __STDC__ */ + + /* Sets the current default syntax to SYNTAX, and return the old syntax. + You can also simply assign to the `re_syntax_options' variable. */ + extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax)); + + /* Compile the regular expression PATTERN, with length LENGTH + and syntax given by the global `re_syntax_options', into the buffer + BUFFER. Return NULL if successful, and an error string if not. */ + extern const char *re_compile_pattern + _RE_ARGS ((const char *pattern, size_t length, + struct re_pattern_buffer *buffer)); + + + /* Compile a fastmap for the compiled pattern in BUFFER; used to + accelerate searches. Return 0 if successful and -2 if was an + internal error. */ + extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer)); + + + /* Search in the string STRING (with length LENGTH) for the pattern + compiled into BUFFER. Start searching at position START, for RANGE + characters. Return the starting position of the match, -1 for no + match, or -2 for an internal error. Also return register + information in REGS (if REGS and BUFFER->no_sub are nonzero). */ + extern int re_search + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, int range, struct re_registers *regs)); + + + /* Like `re_search', but search in the concatenation of STRING1 and + STRING2. Also, stop searching at index START + STOP. */ + extern int re_search_2 + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, int range, struct re_registers *regs, int stop)); + + + /* Like `re_search', but return how many characters in STRING the regexp + in BUFFER matched, starting at position START. */ + extern int re_match + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string, + int length, int start, struct re_registers *regs)); + + + /* Relates to `re_match' as `re_search_2' relates to `re_search'. */ + extern int re_match_2 + _RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, struct re_registers *regs, int stop)); + + + /* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using BUFFER and REGS will use this memory + for recording register information. STARTS and ENDS must be + allocated with malloc, and must each be at least `NUM_REGS * sizeof + (regoff_t)' bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ + extern void re_set_registers + _RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs, + unsigned num_regs, regoff_t *starts, regoff_t *ends)); + +#ifdef _REGEX_RE_COMP +#ifndef _CRAY + /* 4.2 bsd compatibility. */ + extern char *re_comp _RE_ARGS ((const char *)); + extern int re_exec _RE_ARGS ((const char *)); +#endif +#endif + + /* POSIX compatibility. */ + int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags)); + int regexec + _RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch, + regmatch_t pmatch[], int eflags)); + size_t regerror + _RE_ARGS ((int errcode, const regex_t *preg, char *errbuf, + size_t errbuf_size)); + void regfree _RE_ARGS ((regex_t *preg)); + + +#ifdef __cplusplus +} +#endif /* C++ */ + +#endif /* not __REGEXP_LIBRARY_H__ */ + +/* + Local variables: + make-backup-files: t + version-control: t + trim-versions-without-asking: nil + End: +*/ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/snprintf.c b/debian/htdig/htdig-3.2.0b6/htlib/snprintf.c new file mode 100644 index 00000000..7986023d --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/snprintf.c @@ -0,0 +1,75 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#ifndef HAVE_SNPRINTF + +#ifndef NO_SYSTEM_INCLUDES +#include <sys/types.h> + +#include <stdio.h> +#ifdef __STDC__ +#include <stdarg.h> +#else +#include <varargs.h> +#endif +#endif + +/* + * snprintf -- + * Bounded version of sprintf. + * + * PUBLIC: #ifndef HAVE_SNPRINTF + * PUBLIC: #ifdef __STDC__ + * PUBLIC: int snprintf __P((char *, size_t, const char *, ...)); + * PUBLIC: #else + * PUBLIC: int snprintf(); + * PUBLIC: #endif + * PUBLIC: #endif + */ +int +#ifdef __STDC__ +snprintf(char *str, size_t n, const char *fmt, ...) +#else +snprintf(str, n, fmt, va_alist) + char *str; + size_t n; + const char *fmt; + va_dcl +#endif +{ + va_list ap; + int rval; + + n = 0; +#ifdef __STDC__ + va_start(ap, fmt); +#else + va_start(ap); +#endif +#ifdef SPRINTF_RET_CHARPNT + (void)vsprintf(str, fmt, ap); + va_end(ap); + return (strlen(str)); +#else + rval = vsprintf(str, fmt, ap); + va_end(ap); + return (rval); +#endif +} +#endif /* HAVE_SNPRINTF */ + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/strcasecmp.cc b/debian/htdig/htdig-3.2.0b6/htlib/strcasecmp.cc new file mode 100644 index 00000000..1dfa74e2 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/strcasecmp.cc @@ -0,0 +1,101 @@ +// +// strcasecmp.cc +// +// strcasecmp: replacement of the strcasecmp functions for architectures that do +// not have it. +// +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// +// $Id: strcasecmp.cc,v 1.10 2004/05/28 13:15:22 lha Exp $ +// + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#include "lib.h" +#include <ctype.h> + +//***************************************************************************** +// +int mystrcasecmp(const char *str1, const char *str2) +{ + if (!str1 && !str2) + return 0; + if (!str1) + return 1; + if (!str2) + return -1; + while (*str1 && + *str2 && + tolower((unsigned char)*str1) == tolower((unsigned char)*str2)) + { + str1++; + str2++; + } + + return tolower((unsigned char)*str1) - tolower((unsigned char)*str2); +} + + +//#define tolower(ch) (isupper(ch) ? (ch) + 'a' - 'A' : (ch)) +//***************************************************************************** +// +int mystrncasecmp(const char *str1, const char *str2, int n) +{ + if (!str1 && !str2) + return 0; + if (!str1) + return 1; + if (!str2) + return -1; + if (n < 0) + return 0; + while (n && + *str1 && + *str2 && + tolower((unsigned char)*str1) == tolower((unsigned char)*str2)) + { + str1++; + str2++; + n--; + } + + return n == 0 ? 0 : + tolower((unsigned char)*str1) - tolower((unsigned char)*str2); +} + + +//***************************************************************************** +// char *strdup(char *str) +// +char *strdup(char *str) +{ + char *p = new char[strlen(str) + 1]; + strcpy(p, str); + return p; +} + + +//***************************************************************************** +// char *mystrcasestr(const char *s, const char *pattern) +// +const char * +mystrcasestr(const char *s, const char *pattern) +{ + int length = strlen(pattern); + + while (*s) + { + if (mystrncasecmp(s, pattern, length) == 0) + return s; + s++; + } + return 0; +} + + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/strerror.c b/debian/htdig/htdig-3.2.0b6/htlib/strerror.c new file mode 100644 index 00000000..761dd7b0 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/strerror.c @@ -0,0 +1,86 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ +/* + * Copyright (c) 1988, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#ifndef HAVE_STRERROR + +/* + * strerror -- + * Return the string associated with an errno. + * + * PUBLIC: #ifndef HAVE_STRERROR + * PUBLIC: char *strerror __P((int)); + * PUBLIC: #endif + */ +char * +strerror(num) + int num; +{ + extern int sys_nerr; + extern char *sys_errlist[]; +#undef UPREFIX +#define UPREFIX "Unknown error: " + static char ebuf[40] = UPREFIX; /* 64-bit number + slop */ + int errnum; + char *p, *t, tmp[40]; + + errnum = num; /* convert to unsigned */ + if (errnum < sys_nerr) + return(sys_errlist[errnum]); + + /* Do this by hand, so we don't include stdio(3). */ + t = tmp; + do { + *t++ = "0123456789"[errnum % 10]; + } while (errnum /= 10); + for (p = ebuf + sizeof(UPREFIX) - 1;;) { + *p++ = *--t; + if (t <= tmp) + break; + } + return(ebuf); +} + +#endif /* HAVE_STRERROR */ diff --git a/debian/htdig/htdig-3.2.0b6/htlib/strptime.cc b/debian/htdig/htdig-3.2.0b6/htlib/strptime.cc new file mode 100644 index 00000000..5cd28751 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/strptime.cc @@ -0,0 +1,377 @@ +// Part of the ht://Dig package <http://www.htdig.org/> +// Copyright (c) 1999-2004 The ht://Dig Group +// For copyright details, see the file COPYING in your distribution +// or the GNU Library General Public License (LGPL) version 2 or later +// <http://www.gnu.org/copyleft/lgpl.html> +// + +/* + * Copyright (c) 1994 Powerdog Industries. All rights reserved. + * + * Redistribution and use in source and binary forms, without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer + * in the documentation and/or other materials provided with the + * distribution. + * 3. All advertising materials mentioning features or use of this + * software must display the following acknowledgement: + * This product includes software developed by Powerdog Industries. + * 4. The name of Powerdog Industries may not be used to endorse or + * promote products derived from this software without specific prior + * written permission. + * + * THIS SOFTWARE IS PROVIDED BY POWERDOG INDUSTRIES ``AS IS'' AND ANY + * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR + * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE POWERDOG INDUSTRIES BE + * LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR + * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF + * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR + * BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, + * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE + * OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, + * EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + */ + +#ifndef lint +/* +static char copyright[] = +"@(#) Copyright (c) 1994 Powerdog Industries. All rights reserved."; +static char sccsid[] = "@(#)strptime.c 1.0 (Powerdog) 94/03/27"; +*/ +#endif /* not lint */ + +#include <lib.h> + +#include <time.h> +#include <ctype.h> +#include <locale.h> +#include <string.h> + +#define asizeof(a) ((int)(sizeof (a) / sizeof ((a)[0]))) + +struct mydtconv { + char *abbrev_month_names[12]; + char *month_names[12]; + char *abbrev_weekday_names[7]; + char *weekday_names[7]; + char *time_format; + char *sdate_format; + char *dtime_format; + char *am_string; + char *pm_string; + char *ldate_format; +}; + +static struct mydtconv En_US = { + { "Jan", "Feb", "Mar", "Apr", "May", "Jun", + "Jul", "Aug", "Sep", "Oct", "Nov", "Dec" }, + { "January", "February", "March", "April", + "May", "June", "July", "August", + "September", "October", "November", "December" }, + { "Sun", "Mon", "Tue", "Wed", "Thu", "Fri", "Sat" }, + { "Sunday", "Monday", "Tuesday", "Wednesday", + "Thursday", "Friday", "Saturday" }, + "%H:%M:%S", + "%m/%d/%y", + "%a %b %e %T %Z %Y", + "AM", + "PM", + "%A, %B, %e, %Y" +}; + +char * +mystrptime(const char *buf, const char *fmt, struct tm *tm) +{ + char c; + const char *ptr; + int i, len = 0; + + ptr = fmt; + while (*ptr != 0) + { + if (*buf == 0) + break; + + c = *ptr++; + + if (c != '%') + { + if (isspace(c)) + while (*buf != 0 && isspace(*buf)) + buf++; + else if (c != *buf++) + return 0; + continue; + } + + c = *ptr++; + switch (c) + { + case 0: + case '%': + if (*buf++ != '%') + return 0; + break; + + case 'C': + buf = mystrptime(buf, En_US.ldate_format, tm); + if (buf == 0) + return 0; + break; + + case 'c': + buf = mystrptime(buf, "%x %X", tm); + if (buf == 0) + return 0; + break; + + case 'D': + buf = mystrptime(buf, "%m/%d/%y", tm); + if (buf == 0) + return 0; + break; + + case 'R': + buf = mystrptime(buf, "%H:%M", tm); + if (buf == 0) + return 0; + break; + + case 'r': + buf = mystrptime(buf, "%I:%M:%S %p", tm); + if (buf == 0) + return 0; + break; + + case 'T': + buf = mystrptime(buf, "%H:%M:%S", tm); + if (buf == 0) + return 0; + break; + + case 'X': + buf = mystrptime(buf, En_US.time_format, tm); + if (buf == 0) + return 0; + break; + + case 'x': + buf = mystrptime(buf, En_US.sdate_format, tm); + if (buf == 0) + return 0; + break; + + case 'j': + if (!isdigit(*buf)) + return 0; + + for (i = 0; *buf != 0 && isdigit(*buf); buf++) + { + i *= 10; + i += *buf - '0'; + } + if (i > 365) + return 0; + + tm->tm_yday = i; + break; + + case 'M': + case 'S': + if (*buf == 0 || isspace(*buf)) + break; + + if (!isdigit(*buf)) + return 0; + + for (i = 0; *buf != 0 && isdigit(*buf); buf++) + { + i *= 10; + i += *buf - '0'; + } + if (i > 59) + return 0; + + if (c == 'M') + tm->tm_min = i; + else + tm->tm_sec = i; + + if (*buf != 0 && isspace(*buf)) + while (*ptr != 0 && !isspace(*ptr)) + ptr++; + break; + + case 'H': + case 'I': + case 'k': + case 'l': + if (!isdigit(*buf)) + return 0; + + for (i = 0; *buf != 0 && isdigit(*buf); buf++) + { + i *= 10; + i += *buf - '0'; + } + if (c == 'H' || c == 'k') { + if (i > 23) + return 0; + } else if (i > 11) + return 0; + + tm->tm_hour = i; + + if (*buf != 0 && isspace(*buf)) + while (*ptr != 0 && !isspace(*ptr)) + ptr++; + break; + + case 'p': + len = strlen(En_US.am_string); + if (mystrncasecmp(buf, En_US.am_string, len) == 0) + { + if (tm->tm_hour > 12) + return 0; + if (tm->tm_hour == 12) + tm->tm_hour = 0; + buf += len; + break; + } + + len = strlen(En_US.pm_string); + if (mystrncasecmp(buf, En_US.pm_string, len) == 0) + { + if (tm->tm_hour > 12) + return 0; + if (tm->tm_hour != 12) + tm->tm_hour += 12; + buf += len; + break; + } + + return 0; + + case 'A': + case 'a': + for (i = 0; i < asizeof(En_US.weekday_names); i++) + { + len = strlen(En_US.weekday_names[i]); + if (mystrncasecmp(buf, + En_US.weekday_names[i], + len) == 0) + break; + + len = strlen(En_US.abbrev_weekday_names[i]); + if (mystrncasecmp(buf, + En_US.abbrev_weekday_names[i], + len) == 0) + break; + } + if (i == asizeof(En_US.weekday_names)) + return 0; + + tm->tm_wday = i; + buf += len; + break; + + case 'd': + case 'e': + if (!isdigit(*buf)) + return 0; + + for (i = 0; *buf != 0 && isdigit(*buf); buf++) + { + i *= 10; + i += *buf - '0'; + } + if (i > 31) + return 0; + + tm->tm_mday = i; + + if (*buf != 0 && isspace(*buf)) + while (*ptr != 0 && !isspace(*ptr)) + ptr++; + break; + + case 'B': + case 'b': + case 'h': + for (i = 0; i < asizeof(En_US.month_names); i++) + { + len = strlen(En_US.month_names[i]); + if (mystrncasecmp(buf, + En_US.month_names[i], + len) == 0) + break; + + len = strlen(En_US.abbrev_month_names[i]); + if (mystrncasecmp(buf, + En_US.abbrev_month_names[i], + len) == 0) + break; + } + if (i == asizeof(En_US.month_names)) + return 0; + + tm->tm_mon = i; + buf += len; + break; + + case 'm': + if (!isdigit(*buf)) + return 0; + + for (i = 0; *buf != 0 && isdigit(*buf); buf++) + { + i *= 10; + i += *buf - '0'; + } + if (i < 1 || i > 12) + return 0; + + tm->tm_mon = i - 1; + + if (*buf != 0 && isspace(*buf)) + while (*ptr != 0 && !isspace(*ptr)) + ptr++; + break; + + case 'Y': + case 'y': + if (*buf == 0 || isspace(*buf)) + break; + + if (!isdigit(*buf)) + return 0; + + for (i = 0; *buf != 0 && isdigit(*buf); buf++) + { + i *= 10; + i += *buf - '0'; + } + if (c == 'y' && i < 69) /* Unix Epoch pivot year */ + i += 100; + if (c == 'Y') + i -= 1900; + if (i < 0) + return 0; + + tm->tm_year = i; + + if (*buf != 0 && isspace(*buf)) + while (*ptr != 0 && !isspace(*ptr)) + ptr++; + break; + } + } + + return (char *) buf; +} + diff --git a/debian/htdig/htdig-3.2.0b6/htlib/timegm.c b/debian/htdig/htdig-3.2.0b6/htlib/timegm.c new file mode 100644 index 00000000..67588d93 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/timegm.c @@ -0,0 +1,142 @@ +/* + timegm.cc + + timegm: Portable version of timegm (mytimegm) for ht://Dig + Based on a version from the GNU C Library + and a previous implementation for ht://Dig + + Part of the ht://Dig package <http://www.htdig.org/> + Copyright (c) 1999-2004 The ht://Dig Group + For copyright details, see the file COPYING in your distribution + or the GNU Library General Public License (LGPL) version 2 or later + <http://www.gnu.org/copyleft/lgpl.html> + + $Id: timegm.c,v 1.18 2004/05/28 13:15:22 lha Exp $ +*/ + +/* Copyright (C) 1993, 1994, 1995, 1996, 1997 Free Software Foundation, Inc. + This file is part of the GNU C Library. + Contributed by Paul Eggert (eggert@twinsun.com). + + The GNU C Library is free software; you can redistribute it and/or + modify it under the terms of the GNU Library General Public License as + published by the Free Software Foundation; either version 2 of the + License, or (at your option) any later version. + + The GNU C Library is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + Library General Public License for more details. + + You should have received a copy of the GNU Library General Public + License along with the GNU C Library; see the file COPYING.LIB. If not, + write to the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. */ + + +/* #define TEST_TIMEGM */ + +#include <time.h> +#ifdef TEST_TIMEGM +#include <stdio.h> +#include <stdlib.h> +#endif + +static struct tm *my_mktime_gmtime_r (const time_t *t, struct tm *tp); + +static struct tm *my_mktime_gmtime_r (const time_t *t, struct tm *tp) +{ + struct tm *l = gmtime (t); + if (! l) + return 0; + *tp = *l; + return tp; +} + +time_t __mktime_internal(struct tm *, + struct tm *(*) (const time_t *, struct tm *), + time_t *); + +time_t Httimegm(tmp) +struct tm *tmp; +{ + static time_t gmtime_offset; + tmp->tm_isdst = 0; + return __mktime_internal (tmp, my_mktime_gmtime_r, &gmtime_offset); +} + +#ifdef TEST_TIMEGM + +void parse_time(char *s, struct tm *tm) +{ + sscanf(s, "%d.%d.%d %d:%d:%d", + &tm->tm_year, &tm->tm_mon, &tm->tm_mday, + &tm->tm_hour, &tm->tm_min, &tm->tm_sec); + tm->tm_year -= 1900; + tm->tm_mon--; +} + +void print_time(struct tm *tm) +{ + fprintf(stderr, "%04d.%02d.%02d %02d:%02d:%02d", + tm->tm_year+1900, tm->tm_mon+1, tm->tm_mday, + tm->tm_hour, tm->tm_min, tm->tm_sec); +} + +int time_equal(struct tm *tm1, struct tm *tm2) +{ + return ((tm1->tm_year == tm2->tm_year) && + (tm1->tm_mon == tm2->tm_mon) && + (tm1->tm_mday == tm2->tm_mday) && + (tm1->tm_hour == tm2->tm_hour) && + (tm1->tm_min == tm2->tm_min) && + (tm1->tm_sec == tm2->tm_sec)); +} + +int main(void) +{ + char *test_dates[] = + { + "1970.01.01 00:00:00", + "1970.01.01 00:00:01", + "1972.02.05 23:59:59", + "1972.02.28 00:59:59", + "1972.02.28 23:59:59", + "1972.02.29 00:00:00", + "1972.03.01 13:00:04", + "1973.03.01 12:00:00", + "1980.01.01 00:00:05", + "1984.12.31 23:00:00", + "1997.06.05 17:55:35", + "1999.12.31 23:00:00", + "2000.01.01 00:00:05", + "2000.02.28 23:00:05", + "2000.02.29 23:00:05", + "2000.03.01 00:00:05", + "2007.06.05 17:55:35", + "2038.01.19 03:14:07", + 0 + }; + int i, ok = 1; + struct tm orig, *conv; + time_t t; + + for (i = 0; (test_dates[i]); i++) + { + parse_time(test_dates[i], &orig); + t = Httimegm(&orig); + conv = gmtime(&t); + if (!time_equal(&orig, conv)) + { + fprintf(stderr, "timegm() test failed!\n Original: "); + print_time(&orig); + fprintf(stderr, "\n Converted: "); + print_time(conv); + fprintf(stderr, "\n time_t: %ld\n", (long) t); + ok = 0; + } + } + exit(ok ? 0 : 1); +} + +#endif diff --git a/debian/htdig/htdig-3.2.0b6/htlib/vsnprintf.c b/debian/htdig/htdig-3.2.0b6/htlib/vsnprintf.c new file mode 100644 index 00000000..dac21c23 --- /dev/null +++ b/debian/htdig/htdig-3.2.0b6/htlib/vsnprintf.c @@ -0,0 +1,58 @@ +/* Part of the ht://Dig package <http://www.htdig.org/> */ +/* Copyright (c) 1999-2004 The ht://Dig Group */ +/* For copyright details, see the file COPYING in your distribution */ +/* or the GNU Library General Public License (LGPL) version 2 or later */ +/* <http://www.gnu.org/copyleft/lgpl.html> */ + + +/*- + * See the file LICENSE for redistribution information. + * + * Copyright (c) 1996, 1997, 1998, 1999 + * Sleepycat Software. All rights reserved. + */ + +#ifdef HAVE_CONFIG_H +#include "htconfig.h" +#endif /* HAVE_CONFIG_H */ + +#ifndef HAVE_VSNPRINTF + +#ifndef NO_SYSTEM_INCLUDES +#include <sys/types.h> + +#include <stdio.h> +#ifdef __STDC__ +#include <stdarg.h> +#else +#include <varargs.h> +#endif +#endif + +/* + * vsnprintf -- + * Bounded version of vsprintf. + * + * PUBLIC: #ifndef HAVE_VSNPRINTF + * PUBLIC: int vsnprintf(); + * PUBLIC: #endif + */ + +int +vsnprintf(str, n, fmt, ap) + char *str; + size_t n; + const char *fmt; + va_list ap; +{ + n = 0; + +#ifdef SPRINTF_RET_CHARPNT + (void)vsprintf(str, fmt, ap); + return (strlen(str)); +#else + return (vsprintf(str, fmt, ap)); +#endif +} +#endif /* HAVE_VSNPRINTF */ + |