reg.c

php-4.4.7学习linux时下载的源代码

/*
   +----------------------------------------------------------------------+
   | PHP Version 4                                                        |
   +----------------------------------------------------------------------+
   | Copyright (c) 1997-2007 The PHP Group                                |
   +----------------------------------------------------------------------+
   | This source file is subject to version 3.01 of the PHP license,      |
   | that is bundled with this package in the file LICENSE, and is        |
   | available through the world-wide-web at the following url:           |
   | http://www.php.net/license/3_01.txt                                  |
   | If you did not receive a copy of the PHP license and are unable to   |
   | obtain it through the world-wide-web, please send a note to          |
   | [e-mail] so we can mail you a copy immediately.               |
   +----------------------------------------------------------------------+
   | Authors: Rasmus Lerdorf <[e-mail]>                             |
   |          Jim Winstead <[e-mail]>                                 |
   |          Jaakko Hyv鋞ti <[e-mail]>                      | 
   +----------------------------------------------------------------------+
 */
/* $Id: reg.c,v 1.66.2.8.2.3 2007/01/01 09:46:48 sebastian Exp $ */

#include <stdio.h>
#include <ctype.h>
#include "php.h"
#include "php_string.h"
#include "reg.h"
#include "ext/standard/info.h"

ZEND_DECLARE_MODULE_GLOBALS(reg)

typedef struct {
	regex_t preg;
	int cflags;
} reg_cache;

static int reg_magic = 0;

/* {{{ _php_regcomp
 */
static int _php_regcomp(regex_t *preg, const char *pattern, int cflags)
{
	int r = 0;
	int patlen = strlen(pattern);
	reg_cache *rc = NULL;
	TSRMLS_FETCH();
	
	if(zend_hash_find(&REG(ht_rc), (char *) pattern, patlen+1, (void **) &rc) == SUCCESS
	   && rc->cflags == cflags) {
#ifdef HAVE_REGEX_T_RE_MAGIC
		/*
		 * We use a saved magic number to see whether cache is corrupted, and if it
		 * is, we flush it and compile the pattern from scratch.
		 */
		if (rc->preg.re_magic != reg_magic) {
			zend_hash_clean(&REG(ht_rc));
		} else {
			memcpy(preg, &rc->preg, sizeof(*preg));
			return r;
		}
	}

	r = regcomp(preg, pattern, cflags);
	if(!r) {
		reg_cache rcp;

		rcp.cflags = cflags;
		memcpy(&rcp.preg, preg, sizeof(*preg));
		/*
		 * Since we don't have access to the actual MAGIC1 definition in the private
		 * header file, we save the magic value immediately after compilation. Hopefully,
		 * it's good.
		 */
		if (!reg_magic) reg_magic = preg->re_magic;
		zend_hash_update(&REG(ht_rc), (char *) pattern, patlen+1,
						 (void *) &rcp, sizeof(rcp), NULL);
	}
#else
		memcpy(preg, &rc->preg, sizeof(*preg));
	} else {
		r = regcomp(preg, pattern, cflags);
		if(!r) {
			reg_cache rcp;

			rcp.cflags = cflags;
			memcpy(&rcp.preg, preg, sizeof(*preg));
			zend_hash_update(&REG(ht_rc), (char *) pattern, patlen+1,
							 (void *) &rcp, sizeof(rcp), NULL);
		}
	}
#endif
	return r;
}
/* }}} */

static void _free_reg_cache(reg_cache *rc) 
{
	regfree(&rc->preg);
}

#undef regfree
#define regfree(a);
#undef regcomp
#define regcomp(a, b, c) _php_regcomp(a, b, c)
	
static void php_reg_init_globals(zend_reg_globals *reg_globals TSRMLS_DC)
{
	zend_hash_init(&reg_globals->ht_rc, 0, NULL, (void (*)(void *)) _free_reg_cache, 1);
}

static void php_reg_destroy_globals(zend_reg_globals *reg_globals TSRMLS_DC)
{
	zend_hash_destroy(&reg_globals->ht_rc);
}

PHP_MINIT_FUNCTION(regex)
{
	ZEND_INIT_MODULE_GLOBALS(reg, php_reg_init_globals, php_reg_destroy_globals);
	return SUCCESS;
}

PHP_MSHUTDOWN_FUNCTION(regex)
{
#ifndef ZTS
	php_reg_destroy_globals(&reg_globals TSRMLS_CC);
#endif

	return SUCCESS;
}

PHP_MINFO_FUNCTION(regex)
{
#if HSREGEX
	php_info_print_table_row(2, "Regex Library", "Bundled library enabled");
#else
	php_info_print_table_row(2, "Regex Library", "System library enabled");
#endif
}


/* {{{ php_reg_eprint
 * php_reg_eprint - convert error number to name
 */
static void php_reg_eprint(int err, regex_t *re) {
	char *buf = NULL, *message = NULL;
	size_t len;
	size_t buf_len;

#ifdef REG_ITOA
	/* get the length of the message */
	buf_len = regerror(REG_ITOA | err, re, NULL, 0);
	if (buf_len) {
		buf = (char *)safe_emalloc(buf_len, sizeof(char), 0);
		if (!buf) return; /* fail silently */
		/* finally, get the error message */
		regerror(REG_ITOA | err, re, buf, buf_len);
	}
#else
	buf_len = 0;
#endif
	len = regerror(err, re, NULL, 0);
	if (len) {
		TSRMLS_FETCH();

		message = (char *)safe_emalloc((buf_len + len + 2), sizeof(char), 0);
		if (!message) {
			return; /* fail silently */
		}
		if (buf_len) {
			snprintf(message, buf_len, "%s: ", buf);
			buf_len += 1; /* so pointer math below works */
		}
		/* drop the message into place */
		regerror(err, re, message + buf_len, len);

		php_error_docref(NULL TSRMLS_CC, E_WARNING, "%s", message);
	}

	STR_FREE(buf);
	STR_FREE(message);
}
/* }}} */

/* {{{ php_ereg
 */
static void php_ereg(INTERNAL_FUNCTION_PARAMETERS, int icase)
{
	pval **regex,			/* Regular expression */
		**findin,		/* String to apply expression to */
		**array = NULL;		/* Optional register array */
	regex_t re;
	regmatch_t *subs;
	int err, match_len, string_len;
	uint i;
	int copts = 0;
	off_t start, end;
	char *buf = NULL;
	char *string = NULL;
	int   argc = ZEND_NUM_ARGS();
	
	if (argc < 2 || argc > 3 ||
	    zend_get_parameters_ex(argc, &regex, &findin, &array) == FAILURE) {
		WRONG_PARAM_COUNT;
	}

	if (icase)
		copts |= REG_ICASE;
	
	if (argc == 2)
		copts |= REG_NOSUB;

	/* compile the regular expression from the supplied regex */
	if (Z_TYPE_PP(regex) == IS_STRING) {
		err = regcomp(&re, Z_STRVAL_PP(regex), REG_EXTENDED | copts);
	} else {
		/* we convert numbers to integers and treat them as a string */
		if (Z_TYPE_PP(regex) == IS_DOUBLE)
			convert_to_long_ex(regex);	/* get rid of decimal places */
		convert_to_string_ex(regex);
		/* don't bother doing an extended regex with just a number */
		err = regcomp(&re, Z_STRVAL_PP(regex), copts);
	}

	if (err) {
		php_reg_eprint(err, &re);
		RETURN_FALSE;
	}

	/* make a copy of the string we're looking in */
	convert_to_string_ex(findin);
	string = estrndup(Z_STRVAL_PP(findin), Z_STRLEN_PP(findin));

	/* allocate storage for (sub-)expression-matches */
	subs = (regmatch_t *)ecalloc(sizeof(regmatch_t),re.re_nsub+1);
	
	/* actually execute the regular expression */
	err = regexec(&re, string, re.re_nsub+1, subs, 0);
	if (err && err != REG_NOMATCH) {
		php_reg_eprint(err, &re);
		regfree(&re);
		efree(subs);
		RETURN_FALSE;
	}
	match_len = 1;

	if (array && err != REG_NOMATCH) {
		match_len = (int) (subs[0].rm_eo - subs[0].rm_so);
		string_len = Z_STRLEN_PP(findin) + 1;

		buf = emalloc(string_len);

		zval_dtor(*array);	/* start with clean array */
		array_init(*array);

		for (i = 0; i <= re.re_nsub; i++) {
			start = subs[i].rm_so;
			end = subs[i].rm_eo;
			if (start != -1 && end > 0 && start < string_len && end < string_len && start < end) {
				add_index_stringl(*array, i, string+start, end-start, 1);
			} else {
				add_index_bool(*array, i, 0);
			}
		}
		efree(buf);
	}

	efree(subs);
	efree(string);
	if (err == REG_NOMATCH) {
		RETVAL_FALSE;
	} else {
		if (match_len == 0)
			match_len = 1;
		RETVAL_LONG(match_len);
	}
	regfree(&re);
}
/* }}} */

/* {{{ proto int ereg(string pattern, string string [, array registers])
   Regular expression match */
PHP_FUNCTION(ereg)
{
	php_ereg(INTERNAL_FUNCTION_PARAM_PASSTHRU, 0);
}
/* }}} */

/* {{{ proto int eregi(string pattern, string string [, array registers])
   Case-insensitive regular expression match */
PHP_FUNCTION(eregi)
{
	php_ereg(INTERNAL_FUNCTION_PARAM_PASSTHRU, 1);
}
/* }}} */

/* {{{ php_reg_replace
 * this is the meat and potatoes of regex replacement! */
PHPAPI char *php_reg_replace(const char *pattern, const char *replace, const char *string, int icase, int extended)
{
	regex_t re;
	regmatch_t *subs;

	char *buf,	/* buf is where we build the replaced string */
	     *nbuf,	/* nbuf is used when we grow the buffer */
		 *walkbuf; /* used to walk buf when replacing backrefs */
	const char *walk; /* used to walk replacement string for backrefs */
	int buf_len;
	int pos, tmp, string_len, new_l;
	int err, copts = 0;

	string_len = strlen(string);

	if (icase) {
		copts = REG_ICASE;
	}
	if (extended) {
		copts |= REG_EXTENDED;
	}

	err = regcomp(&re, pattern, copts);
	if (err) {
		php_reg_eprint(err, &re);
		return ((char *) -1);
	}


	/* allocate storage for (sub-)expression-matches */
	subs = (regmatch_t *)ecalloc(sizeof(regmatch_t),re.re_nsub+1);

	/* start with a buffer that is twice the size of the stringo
	   we're doing replacements in */
	buf_len = 2 * string_len + 1;
	buf = safe_emalloc(buf_len, sizeof(char), 0);

	err = pos = 0;
	buf[0] = '\0';