Use uint32_t for 'illegal_substchar' codepoint in mbstring

alexdowad · alexdowad · commit 3ce888a83739 · 2022-10-05T10:02:02.000+09:00
This value is a wchar, so the best type for it is uint32_t.
diff --git a/ext/mbstring/libmbfl/mbfl/mbfilter.c b/ext/mbstring/libmbfl/mbfl/mbfilter.c
@@ -154,7 +154,7 @@ void mbfl_buffer_converter_illegal_mode(mbfl_buffer_converter *convd, int mode)
 	}
 }
 
-void mbfl_buffer_converter_illegal_substchar(mbfl_buffer_converter *convd, int substchar)
+void mbfl_buffer_converter_illegal_substchar(mbfl_buffer_converter *convd, uint32_t substchar)
 {
 	if (convd->filter2) {
 		convd->filter2->illegal_substchar = substchar;
@@ -387,7 +387,7 @@ mbfl_convert_encoding(
 
 	if (filter2 != NULL) {
 		filter2->illegal_mode = MBFL_OUTPUTFILTER_ILLEGAL_MODE_CHAR;
-		filter2->illegal_substchar = 0x3f;		/* '?' */
+		filter2->illegal_substchar = '?';
 	}
 
 	mbfl_memory_device_init(&device, string->len, (string->len >> 2) + 8);
diff --git a/ext/mbstring/libmbfl/mbfl/mbfilter.h b/ext/mbstring/libmbfl/mbfl/mbfilter.h
@@ -142,7 +142,7 @@ struct _mbfl_buffer_converter {
 MBFLAPI extern mbfl_buffer_converter * mbfl_buffer_converter_new(const mbfl_encoding *from, const mbfl_encoding *to, size_t buf_initsz);
 MBFLAPI extern void mbfl_buffer_converter_delete(mbfl_buffer_converter *convd);
 MBFLAPI extern void mbfl_buffer_converter_illegal_mode(mbfl_buffer_converter *convd, int mode);
-MBFLAPI extern void mbfl_buffer_converter_illegal_substchar(mbfl_buffer_converter *convd, int substchar);
+MBFLAPI extern void mbfl_buffer_converter_illegal_substchar(mbfl_buffer_converter *convd, uint32_t substchar);
 MBFLAPI extern size_t mbfl_buffer_converter_feed(mbfl_buffer_converter *convd, mbfl_string *string);
 MBFLAPI extern void mbfl_buffer_converter_flush(mbfl_buffer_converter *convd);
 MBFLAPI extern mbfl_string * mbfl_buffer_converter_result(mbfl_buffer_converter *convd, mbfl_string *result);
diff --git a/ext/mbstring/libmbfl/mbfl/mbfl_convert.c b/ext/mbstring/libmbfl/mbfl/mbfl_convert.c
@@ -243,7 +243,7 @@ int mbfl_filt_conv_illegal_output(int c, mbfl_convert_filter *filter)
 	unsigned int w = c;
 	int ret = 0;
 	int mode_backup = filter->illegal_mode;
-	int substchar_backup = filter->illegal_substchar;
+	uint32_t substchar_backup = filter->illegal_substchar;
 
 	/* The used substitution character may not be supported by the target character encoding.
 	 * If that happens, first try to use "?" instead and if that also fails, silently drop the
diff --git a/ext/mbstring/libmbfl/mbfl/mbfl_convert.h b/ext/mbstring/libmbfl/mbfl/mbfl_convert.h
@@ -57,7 +57,7 @@ struct _mbfl_convert_filter {
 	const mbfl_encoding *from;
 	const mbfl_encoding *to;
 	int illegal_mode;
-	int illegal_substchar;
+	uint32_t illegal_substchar;
 	size_t num_illegalchar;
 	void *opaque;
 };
diff --git a/ext/mbstring/mbstring.c b/ext/mbstring/mbstring.c
@@ -883,8 +883,8 @@ static PHP_INI_MH(OnUpdate_mbstring_substitute_character)
 	} else {
 		MBSTRG(filter_illegal_mode) = MBFL_OUTPUTFILTER_ILLEGAL_MODE_CHAR;
 		MBSTRG(current_filter_illegal_mode) = MBFL_OUTPUTFILTER_ILLEGAL_MODE_CHAR;
-		MBSTRG(filter_illegal_substchar) = 0x3f;	/* '?' */
-		MBSTRG(current_filter_illegal_substchar) = 0x3f;	/* '?' */
+		MBSTRG(filter_illegal_substchar) = '?';
+		MBSTRG(current_filter_illegal_substchar) = '?';
 	}
 
 	return SUCCESS;
@@ -1015,9 +1015,9 @@ ZEND_TSRMLS_CACHE_UPDATE();
 	mbstring_globals->default_detect_order_list = (enum mbfl_no_encoding *) php_mb_default_identify_list_neut;
 	mbstring_globals->default_detect_order_list_size = sizeof(php_mb_default_identify_list_neut) / sizeof(php_mb_default_identify_list_neut[0]);
 	mbstring_globals->filter_illegal_mode = MBFL_OUTPUTFILTER_ILLEGAL_MODE_CHAR;
-	mbstring_globals->filter_illegal_substchar = 0x3f;	/* '?' */
+	mbstring_globals->filter_illegal_substchar = '?';
 	mbstring_globals->current_filter_illegal_mode = MBFL_OUTPUTFILTER_ILLEGAL_MODE_CHAR;
-	mbstring_globals->current_filter_illegal_substchar = 0x3f;	/* '?' */
+	mbstring_globals->current_filter_illegal_substchar = '?';
 	mbstring_globals->illegalchars = 0;
 	mbstring_globals->encoding_translation = 0;
 	mbstring_globals->strict_detection = 0;
diff --git a/ext/mbstring/mbstring.h b/ext/mbstring/mbstring.h
@@ -87,9 +87,9 @@ ZEND_BEGIN_MODULE_GLOBALS(mbstring)
 	enum mbfl_no_encoding *default_detect_order_list;
 	size_t default_detect_order_list_size;
 	int filter_illegal_mode;
-	int filter_illegal_substchar;
+	uint32_t filter_illegal_substchar;
 	int current_filter_illegal_mode;
-	int current_filter_illegal_substchar;
+	uint32_t current_filter_illegal_substchar;
 	enum mbfl_no_language language;
 	bool encoding_translation;
 	bool strict_detection;
diff --git a/ext/mbstring/php_unicode.c b/ext/mbstring/php_unicode.c
@@ -238,7 +238,7 @@ static uint32_t *emit_special_casing_sequence(uint32_t w, uint32_t *out)
 	return out;
 }
 
-MBSTRING_API zend_string *php_unicode_convert_case(php_case_mode case_mode, const char *srcstr, size_t in_len, const mbfl_encoding *src_encoding, int illegal_mode, int illegal_substchar)
+MBSTRING_API zend_string *php_unicode_convert_case(php_case_mode case_mode, const char *srcstr, size_t in_len, const mbfl_encoding *src_encoding, int illegal_mode, uint32_t illegal_substchar)
 {
 	/* A Unicode codepoint can expand out to up to 3 codepoints when uppercased, lowercased, or title cased
 	 * See http://www.unicode.org/Public/UNIDATA/SpecialCasing.txt */
diff --git a/ext/mbstring/php_unicode.h b/ext/mbstring/php_unicode.h
@@ -91,7 +91,7 @@ typedef enum {
 
 MBSTRING_API zend_string *php_unicode_convert_case(
 		php_case_mode case_mode, const char *srcstr, size_t srclen,
-		const mbfl_encoding *src_encoding, int illegal_mode, int illegal_substchar);
+		const mbfl_encoding *src_encoding, int illegal_mode, uint32_t illegal_substchar);
 
 /* Optimize the common ASCII case for lower/upper */
 

Original file line number	Diff line number	Diff line change
`@@ -154,7 +154,7 @@ void mbfl_buffer_converter_illegal_mode(mbfl_buffer_converter *convd, int mode)`
`154`	`154`	`}`
`155`	`155`	`}`
`156`	`156`
`157`		`-void mbfl_buffer_converter_illegal_substchar(mbfl_buffer_converter *convd, int substchar)`
	`157`	`+void mbfl_buffer_converter_illegal_substchar(mbfl_buffer_converter *convd, uint32_t substchar)`
`158`	`158`	`{`
`159`	`159`	`if (convd->filter2) {`
`160`	`160`	`convd->filter2->illegal_substchar = substchar;`
`@@ -387,7 +387,7 @@ mbfl_convert_encoding(`
`387`	`387`
`388`	`388`	`if (filter2 != NULL) {`
`389`	`389`	`filter2->illegal_mode = MBFL_OUTPUTFILTER_ILLEGAL_MODE_CHAR;`
`390`		`- filter2->illegal_substchar = 0x3f; /* '?' */`
	`390`	`+ filter2->illegal_substchar = '?';`
`391`	`391`	`}`
`392`	`392`
`393`	`393`	`mbfl_memory_device_init(&device, string->len, (string->len >> 2) + 8);`
Original file line number	Diff line number	Diff line change
`@@ -238,7 +238,7 @@ static uint32_t emit_special_casing_sequence(uint32_t w, uint32_t out)`
`238`	`238`	`return out;`
`239`	`239`	`}`
`240`	`240`
`241`		`-MBSTRING_API zend_string php_unicode_convert_case(php_case_mode case_mode, const char srcstr, size_t in_len, const mbfl_encoding *src_encoding, int illegal_mode, int illegal_substchar)`
	`241`	`+MBSTRING_API zend_string php_unicode_convert_case(php_case_mode case_mode, const char srcstr, size_t in_len, const mbfl_encoding *src_encoding, int illegal_mode, uint32_t illegal_substchar)`
`242`	`242`	`{`
`243`	`243`	`/* A Unicode codepoint can expand out to up to 3 codepoints when uppercased, lowercased, or title cased`
`244`	`244`	`* See http://www.unicode.org/Public/UNIDATA/SpecialCasing.txt */`