From: Christoph M. Becker <cmb@php.net>
Date: Tue, 23 Jun 2015 17:41:02 +0000 (+0200)
Subject: Merge branch 'PHP-5.6'
X-Git-Tag: php-7.0.0beta1~12^2~49^2~20
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=ca049e0ae97d5a176214ef3c97962093de1a7f49;p=php

Merge branch 'PHP-5.6'

* PHP-5.6:
  updated NEWS
  Fixed Bug #53823 (preg_replace: * qualifier on unicode replace garbles the string)
---

ca049e0ae97d5a176214ef3c97962093de1a7f49
diff --cc ext/pcre/php_pcre.c
index 418859f5d2,f7afc47458..59a0aa569f
--- a/ext/pcre/php_pcre.c
+++ b/ext/pcre/php_pcre.c
@@@ -233,9 -225,28 +233,28 @@@ static char **make_subpats_table(int nu
  }
  /* }}} */
  
+ /* {{{ static calculate_unit_length */
+ /* Calculates the byte length of the next character. Assumes valid UTF-8 for PCRE_UTF8. */
+ static zend_always_inline int calculate_unit_length(pcre_cache_entry *pce, char *start)
+ {
+ 	int unit_len;
+ 
+ 	if (pce->compile_options & PCRE_UTF8) {
+ 		char *end = start;
+ 
+ 		/* skip continuation bytes */
+ 		while ((*++end & 0xC0) == 0x80);
+ 		unit_len = end - start;
+ 	} else {
+ 		unit_len = 1;
+ 	}
+ 	return unit_len;
+ }
+ /* }}} */
+ 
  /* {{{ pcre_get_compiled_regex_cache
   */
 -PHPAPI pcre_cache_entry* pcre_get_compiled_regex_cache(char *regex, int regex_len TSRMLS_DC)
 +PHPAPI pcre_cache_entry* pcre_get_compiled_regex_cache(zend_string *regex)
  {
  	pcre				*re = NULL;
  	pcre_extra			*extra;
@@@ -854,8 -799,10 +873,10 @@@ PHPAPI void php_pcre_match_impl(pcre_ca
  			   the start offset, and continue. Fudge the offset values
  			   to achieve this, unless we're already at the end of the string. */
  			if (g_notempty != 0 && start_offset < subject_len) {
+ 				int unit_len = calculate_unit_length(pce, subject + start_offset);
+ 				
 -				offsets[0] = start_offset;
 -				offsets[1] = start_offset + unit_len;
 +				offsets[0] = (int)start_offset;
- 				offsets[1] = (int)(start_offset + 1);
++				offsets[1] = (int)(start_offset + unit_len);
  			} else
  				break;
  		} else {
@@@ -1247,29 -1261,25 +1268,31 @@@ PHPAPI zend_string *php_pcre_replace_im
  			   the start offset, and continue. Fudge the offset values
  			   to achieve this, unless we're already at the end of the string. */
  			if (g_notempty != 0 && start_offset < subject_len) {
+ 				int unit_len = calculate_unit_length(pce, piece);
+ 
  				offsets[0] = start_offset;
- 				offsets[1] = start_offset + 1;
- 				memcpy(&result->val[result_len], piece, 1);
- 				result_len++;
+ 				offsets[1] = start_offset + unit_len;
 -				memcpy(&result[*result_len], piece, unit_len);
 -				*result_len += unit_len;
++				memcpy(&result->val[result_len], piece, unit_len);
++				result_len += unit_len;
  			} else {
 -				new_len = *result_len + subject_len - start_offset;
 -				if (new_len + 1 > alloc_len) {
 -					alloc_len = new_len + 1; /* now we know exactly how long it is */
 -					new_buf = safe_emalloc(alloc_len, sizeof(char), 0);
 -					memcpy(new_buf, result, *result_len);
 -					efree(result);
 -					result = new_buf;
 +				if (!result && subject_str) {
 +					result = zend_string_copy(subject_str);
 +					break;
 +				}
 +				new_len = result_len + subject_len - start_offset;
 +				if (new_len > alloc_len) {
 +					alloc_len = new_len; /* now we know exactly how long it is */
 +					if (NULL != result) {
 +						result = zend_string_realloc(result, alloc_len, 0);
 +					} else {
 +						result = zend_string_alloc(alloc_len, 0);
 +					}
  				}
  				/* stick that last bit of string on our output */
 -				memcpy(&result[*result_len], piece, subject_len - start_offset);
 -				*result_len += subject_len - start_offset;
 -				result[*result_len] = '\0';
 +				memcpy(&result->val[result_len], piece, subject_len - start_offset);
 +				result_len += subject_len - start_offset;
 +				result->val[result_len] = '\0';
 +				result->len = result_len;
  				break;
  			}
  		} else {