Removed compile time dependency from ext/mbstring

NEWS

@@ -32,6 +32,7 @@ PHP                                                                        NEWS
   . Added multibyte suppport by default. Previosly php had to be compiled
     with --enable-zend-multibyte. Now it can be enabled or disabled throug
     zend.multibyte directive in php.ini (Dmitry)
+  . Removed compile time dependency from ext/mbstring (Dmitry)
   . Added scalar typehints to the parser and the reflection API. (Ilia, Derick)
   . Added support for Traits. (Stefan)
   . Added closure $this support back. (Stas)

Zend/zend_compile.c

@@ -200,9 +200,6 @@ void zend_init_compiler_data_structures(TSRMLS_D) /* {{{ */
 	CG(script_encoding_list) = NULL;
 	CG(script_encoding_list_size) = 0;
 	CG(internal_encoding) = NULL;
-	CG(encoding_detector) = NULL;
-	CG(encoding_converter) = NULL;
-	CG(encoding_oddlen) = NULL;
 	CG(encoding_declared) = 0;
 }
 /* }}} */

Zend/zend_globals.h

@@ -155,11 +155,6 @@ struct _zend_compiler_globals {
 
 	zend_encoding *internal_encoding;
 
-	/* multibyte utility functions */
-	zend_encoding_detector encoding_detector;
-	zend_encoding_converter encoding_converter;
-	zend_encoding_oddlen encoding_oddlen;
-
 #ifdef ZTS
 	zval ***static_members_table;
 	int last_static_member;

Zend/zend_multibyte.c

@@ -513,7 +513,36 @@ static zend_encoding *zend_encoding_table[] = {
 	NULL
 };
 
+static char* dummy_encoding_detector(const unsigned char *string, size_t length, char *list TSRMLS_DC)
+{
+	return NULL;
+}
+
+static int dummy_encoding_converter(unsigned char **to, size_t *to_length, const unsigned char *from, size_t from_length, const char *encoding_to, const char *encoding_from TSRMLS_DC)
+{
+	return -1;
+}
 
+static size_t dummy_encoding_oddlen(const unsigned char *string, size_t length, const char *encoding TSRMLS_DC)
+{
+	return 0;
+}
+
+static int dummy_encoding_list_checker(const char *encoding_list TSRMLS_DC)
+{
+	return 0;
+}
+
+static const char* dummy_get_internal_encoding(TSRMLS_D)
+{
+	return NULL;
+}
+
+ZEND_API zend_encoding_detector zend_multibyte_encoding_detector = dummy_encoding_detector;
+ZEND_API zend_encoding_converter zend_multibyte_encoding_converter = dummy_encoding_converter;
+ZEND_API zend_encoding_oddlen zend_multibyte_encoding_oddlen = dummy_encoding_oddlen;
+ZEND_API zend_encoding_list_checker zend_multibyte_check_encoding_list = dummy_encoding_list_checker;
+ZEND_API zend_encoding_name_getter zend_multibyte_get_internal_encoding = dummy_get_internal_encoding;
 
 ZEND_API int zend_multibyte_set_script_encoding(const char *encoding_list,
 size_t encoding_list_size TSRMLS_DC)
@@ -540,11 +569,13 @@ ZEND_API int zend_multibyte_set_internal_encoding(const char *encoding_name TSRM
 	return 0;
 }
 
-ZEND_API int zend_multibyte_set_functions(zend_encoding_detector encoding_detector, zend_encoding_converter encoding_converter, zend_encoding_oddlen encoding_oddlen TSRMLS_DC)
+ZEND_API int zend_multibyte_set_functions(zend_encoding_detector encoding_detector, zend_encoding_converter encoding_converter, zend_encoding_oddlen encoding_oddlen, zend_encoding_list_checker encoding_list_checker, zend_encoding_name_getter get_internal_encoding TSRMLS_DC)
 {
-	CG(encoding_detector) = encoding_detector;
-	CG(encoding_converter) = encoding_converter;
-	CG(encoding_oddlen) = encoding_oddlen;
+	zend_multibyte_encoding_detector = encoding_detector;
+	zend_multibyte_encoding_converter = encoding_converter;
+	zend_multibyte_encoding_oddlen = encoding_oddlen;
+	zend_multibyte_check_encoding_list = encoding_list_checker;
+	zend_multibyte_get_internal_encoding = get_internal_encoding;	
 	return 0;
 }
 
@@ -659,18 +690,16 @@ static size_t zend_multibyte_encoding_filter(unsigned char **to, size_t *to_leng
 {
 	size_t oddlen;
 
-	if (!CG(encoding_converter)) {
+	if (zend_multibyte_encoding_converter == dummy_encoding_converter) {
 		return 0;
 	}
 
-	if (CG(encoding_oddlen)) {
-		oddlen = CG(encoding_oddlen)(from, from_length, from_encoding TSRMLS_CC);
-		if (oddlen > 0) {
-			from_length -= oddlen;
-		}
+	oddlen = zend_multibyte_encoding_oddlen(from, from_length, from_encoding TSRMLS_CC);
+	if (oddlen > 0) {
+		from_length -= oddlen;
 	}
 
-	if (CG(encoding_converter)(to, to_length, from, from_length, to_encoding, from_encoding TSRMLS_CC) != 0) {
+	if (zend_multibyte_encoding_converter(to, to_length, from, from_length, to_encoding, from_encoding TSRMLS_CC) != 0) {
 		return 0;
 	}
 
@@ -1053,10 +1082,11 @@ static zend_encoding* zend_multibyte_find_script_encoding(zend_encoding *onetime
 	}
 
 	/* if multiple encodings specified, detect automagically */
-	if (CG(script_encoding_list_size) > 1 && CG(encoding_detector)) {
+	if (CG(script_encoding_list_size) > 1 && 
+	    zend_multibyte_encoding_detector != dummy_encoding_detector) {
 		list = zend_multibyte_assemble_encoding_list(CG(script_encoding_list),
 				CG(script_encoding_list_size));
-		name = CG(encoding_detector)(LANG_SCNG(script_org), 
+		name = zend_multibyte_encoding_detector(LANG_SCNG(script_org), 
 				LANG_SCNG(script_org_size), list TSRMLS_CC);
 		if (list) {
 			efree(list);

Zend/zend_multibyte.h

@@ -36,6 +36,10 @@ typedef int (*zend_encoding_converter)(unsigned char **to, size_t *to_length, co
 
 typedef size_t (*zend_encoding_oddlen)(const unsigned char *string, size_t length, const char *encoding TSRMLS_DC);
 
+typedef int (*zend_encoding_list_checker)(const char *encoding_list TSRMLS_DC);
+
+typedef const char* (*zend_encoding_name_getter)(TSRMLS_D);
+
 typedef struct _zend_encoding {
 	zend_encoding_filter input_filter;		/* escape input filter */
 	zend_encoding_filter output_filter;		/* escape output filter */
@@ -49,10 +53,18 @@ typedef struct _zend_encoding {
  * zend multibyte APIs
  */
 BEGIN_EXTERN_C()
+
+/* multibyte utility functions */
+ZEND_API extern zend_encoding_detector zend_multibyte_encoding_detector;
+ZEND_API extern zend_encoding_converter zend_multibyte_encoding_converter;
+ZEND_API extern zend_encoding_oddlen zend_multibyte_encoding_oddlen;
+ZEND_API extern zend_encoding_list_checker zend_multibyte_check_encoding_list;
+ZEND_API extern zend_encoding_name_getter zend_multibyte_get_internal_encoding;
+
 ZEND_API int zend_multibyte_set_script_encoding(const char *encoding_list,
 size_t encoding_list_size TSRMLS_DC);
 ZEND_API int zend_multibyte_set_internal_encoding(const char *encoding_name TSRMLS_DC);
-ZEND_API int zend_multibyte_set_functions(zend_encoding_detector encoding_detector, zend_encoding_converter encoding_converter, zend_encoding_oddlen encoding_oddlen TSRMLS_DC);
+ZEND_API int zend_multibyte_set_functions(zend_encoding_detector encoding_detector, zend_encoding_converter encoding_converter, zend_encoding_oddlen encoding_oddlen, zend_encoding_list_checker encoding_list_checker, zend_encoding_name_getter get_internal_encoding TSRMLS_DC);
 ZEND_API int zend_multibyte_set_filter(zend_encoding *onetime_encoding TSRMLS_DC);
 ZEND_API zend_encoding* zend_multibyte_fetch_encoding(const char *encoding_name);
 ZEND_API size_t zend_multibyte_script_encoding_filter(unsigned char **to, size_t 

ext/exif/exif.c

@@ -66,16 +66,6 @@
 #include "ext/standard/php_image.h"
 #include "ext/standard/info.h" 
 
-#if defined(PHP_WIN32) || (HAVE_MBSTRING && !defined(COMPILE_DL_MBSTRING))
-#define EXIF_USE_MBSTRING 1
-#else
-#define EXIF_USE_MBSTRING 0
-#endif
-
-#if EXIF_USE_MBSTRING
-#include "ext/mbstring/mbstring.h"
-#endif
-
 /* needed for ssize_t definition */
 #include <sys/types.h>
 
@@ -176,23 +166,19 @@ ZEND_DECLARE_MODULE_GLOBALS(exif)
 
 ZEND_INI_MH(OnUpdateEncode)
 {
-#if EXIF_USE_MBSTRING
-	if (new_value && strlen(new_value) && !php_mb_check_encoding_list(new_value TSRMLS_CC)) {
+	if (new_value && strlen(new_value) && !zend_multibyte_check_encoding_list(new_value TSRMLS_CC)) {
 		php_error_docref(NULL TSRMLS_CC, E_WARNING, "Illegal encoding ignored: '%s'", new_value);
 		return FAILURE;
 	}
-#endif
 	return OnUpdateString(entry, new_value, new_value_length, mh_arg1, mh_arg2, mh_arg3, stage TSRMLS_CC);
 }
 
 ZEND_INI_MH(OnUpdateDecode)
 {
-#if EXIF_USE_MBSTRING
-	if (!php_mb_check_encoding_list(new_value TSRMLS_CC)) {
+	if (!zend_multibyte_check_encoding_list(new_value TSRMLS_CC)) {
 		php_error_docref(NULL TSRMLS_CC, E_WARNING, "Illegal encoding ignored: '%s'", new_value);
 		return FAILURE;
 	}
-#endif
 	return OnUpdateString(entry, new_value, new_value_length, mh_arg1, mh_arg2, mh_arg3, stage TSRMLS_CC);
 }
 
@@ -224,7 +210,11 @@ static PHP_GINIT_FUNCTION(exif)
 PHP_MINIT_FUNCTION(exif)
 {
 	REGISTER_INI_ENTRIES();
-	REGISTER_LONG_CONSTANT("EXIF_USE_MBSTRING", EXIF_USE_MBSTRING, CONST_CS | CONST_PERSISTENT); 
+	if (zend_hash_exists(&module_registry, "mbstring", sizeof("mbstring"))) {
+		REGISTER_LONG_CONSTANT("EXIF_USE_MBSTRING", 1, CONST_CS | CONST_PERSISTENT); 
+	} else {
+		REGISTER_LONG_CONSTANT("EXIF_USE_MBSTRING", 0, CONST_CS | CONST_PERSISTENT); 
+	}
 	return SUCCESS;
 }
 /* }}} */
@@ -241,9 +231,7 @@ PHP_MSHUTDOWN_FUNCTION(exif)
 /* {{{ exif dependencies */
 static const zend_module_dep exif_module_deps[] = {
 	ZEND_MOD_REQUIRED("standard")
-#if EXIF_USE_MBSTRING
-	ZEND_MOD_REQUIRED("mbstring")
-#endif
+	ZEND_MOD_OPTIONAL("mbstring")
 	{NULL, NULL, NULL}
 };
 /* }}} */
@@ -2588,7 +2576,6 @@ static int exif_process_undefined(char **result, char *value, size_t byte_count
 
 /* {{{ exif_process_string_raw
  * Copy a string in Exif header to a character string returns length of allocated buffer if any. */
-#if !EXIF_USE_MBSTRING
 static int exif_process_string_raw(char **result, char *value, size_t byte_count) {
 	/* we cannot use strlcpy - here the problem is that we have to copy NUL
 	 * chars up to byte_count, we also have to add a single NUL character to
@@ -2602,7 +2589,6 @@ static int exif_process_string_raw(char **result, char *value, size_t byte_count
 	}
 	return 0;
 }
-#endif
 /* }}} */
 
 /* {{{ exif_process_string
@@ -2629,11 +2615,8 @@ static int exif_process_string(char **result, char *value, size_t byte_count TSR
 static int exif_process_user_comment(image_info_type *ImageInfo, char **pszInfoPtr, char **pszEncoding, char *szValuePtr, int ByteCount TSRMLS_DC)
 {
 	int   a;
-
-#if EXIF_USE_MBSTRING
 	char  *decode;
 	size_t len;;
-#endif
 
 	*pszEncoding = NULL;
 	/* Copy the comment */
@@ -2642,7 +2625,6 @@ static int exif_process_user_comment(image_info_type *ImageInfo, char **pszInfoP
 			*pszEncoding = estrdup((const char*)szValuePtr);
 			szValuePtr = szValuePtr+8;
 			ByteCount -= 8;
-#if EXIF_USE_MBSTRING
 			/* First try to detect BOM: ZERO WIDTH NOBREAK SPACE (FEFF 16) 
 			 * since we have no encoding support for the BOM yet we skip that.
 			 */
@@ -2659,34 +2641,38 @@ static int exif_process_user_comment(image_info_type *ImageInfo, char **pszInfoP
 			} else {
 				decode = ImageInfo->decode_unicode_le;
 			}
-			*pszInfoPtr = php_mb_convert_encoding(szValuePtr, ByteCount, ImageInfo->encode_unicode, decode, &len TSRMLS_CC);
+			if (zend_multibyte_encoding_converter(
+					pszInfoPtr, 
+					&len, 
+					szValuePtr,
+					ByteCount,
+					ImageInfo->encode_unicode,
+					decode
+					TSRMLS_DC) != 0) {
+				len = exif_process_string_raw(pszInfoPtr, szValuePtr, ByteCount);
+			}
 			return len;
-#else
-			return exif_process_string_raw(pszInfoPtr, szValuePtr, ByteCount);
-#endif
-		} else
-		if (!memcmp(szValuePtr, "ASCII\0\0\0", 8)) {
+		} else if (!memcmp(szValuePtr, "ASCII\0\0\0", 8)) {
 			*pszEncoding = estrdup((const char*)szValuePtr);
 			szValuePtr = szValuePtr+8;
 			ByteCount -= 8;
-		} else
-		if (!memcmp(szValuePtr, "JIS\0\0\0\0\0", 8)) {
+		} else if (!memcmp(szValuePtr, "JIS\0\0\0\0\0", 8)) {
 			/* JIS should be tanslated to MB or we leave it to the user - leave it to the user */
 			*pszEncoding = estrdup((const char*)szValuePtr);
 			szValuePtr = szValuePtr+8;
 			ByteCount -= 8;
-#if EXIF_USE_MBSTRING
-			if (ImageInfo->motorola_intel) {
-				*pszInfoPtr = php_mb_convert_encoding(szValuePtr, ByteCount, ImageInfo->encode_jis, ImageInfo->decode_jis_be, &len TSRMLS_CC);
-			} else {
-				*pszInfoPtr = php_mb_convert_encoding(szValuePtr, ByteCount, ImageInfo->encode_jis, ImageInfo->decode_jis_le, &len TSRMLS_CC);
+			if (zend_multibyte_encoding_converter(
+					pszInfoPtr, 
+					&len, 
+					szValuePtr,
+					ByteCount,
+					ImageInfo->encode_jis,
+					ImageInfo->motorola_intel ? ImageInfo->decode_jis_be : ImageInfo->decode_jis_le
+					TSRMLS_DC) != 0) {
+				len = exif_process_string_raw(pszInfoPtr, szValuePtr, ByteCount);
 			}
 			return len;
-#else
-			return exif_process_string_raw(pszInfoPtr, szValuePtr, ByteCount);
-#endif
-		} else
-		if (!memcmp(szValuePtr, "\0\0\0\0\0\0\0\0", 8)) {
+		} else if (!memcmp(szValuePtr, "\0\0\0\0\0\0\0\0", 8)) {
 			/* 8 NULL means undefined and should be ASCII... */
 			*pszEncoding = estrdup("UNDEFINED");
 			szValuePtr = szValuePtr+8;
@@ -2714,19 +2700,17 @@ static int exif_process_unicode(image_info_type *ImageInfo, xp_field_type *xp_fi
 	xp_field->tag = tag;	
 
 	/* Copy the comment */
-#if EXIF_USE_MBSTRING
-/*  What if MS supports big-endian with XP? */
-/*	if (ImageInfo->motorola_intel) {
-		xp_field->value = php_mb_convert_encoding(szValuePtr, ByteCount, ImageInfo->encode_unicode, ImageInfo->decode_unicode_be, &xp_field->size TSRMLS_CC);
-	} else {
-		xp_field->value = php_mb_convert_encoding(szValuePtr, ByteCount, ImageInfo->encode_unicode, ImageInfo->decode_unicode_le, &xp_field->size TSRMLS_CC);
-	}*/
-	xp_field->value = php_mb_convert_encoding(szValuePtr, ByteCount, ImageInfo->encode_unicode, ImageInfo->decode_unicode_le, &xp_field->size TSRMLS_CC);
-	return xp_field->size;
-#else
-	xp_field->size = exif_process_string_raw(&xp_field->value, szValuePtr, ByteCount);
+	if (zend_multibyte_encoding_converter(
+			&xp_field->value, 
+			&xp_field->size, 
+			szValuePtr,
+			ByteCount,
+			ImageInfo->encode_unicode,
+			ImageInfo->motorola_intel ? ImageInfo->decode_unicode_be : ImageInfo->decode_unicode_le
+			TSRMLS_DC) != 0) {
+		xp_field->size = exif_process_string_raw(&xp_field->value, szValuePtr, ByteCount);
+	}
 	return xp_field->size;
-#endif
 }
 /* }}} */