[pull] master from php:master

NEWS

@@ -28,6 +28,17 @@ PHP                                                                        NEWS
 - FPM:
   . Fixed failed debug assertion when php_admin_value setting fails. (ilutov)
 
+- Filter:
+  . Fixed bug GH-16993 (filter_var_array with FILTER_VALIDATE_INT|FILTER_NULL_ON_FAILURE
+    should emit warning for invalid filter usage). (alexandre-daubois)
+
+- Intl:
+  . Added grapheme_strpos(), grapheme_stripos(), grapheme_strrpos(),
+    grapheme_strripos(), grapheme_strstr(), grapheme_stristr() and
+    grapheme_levenshtein() functions add $locale parameter (Yuya Hamada).
+  . Fixed bug GH-11952 (Fix locale strings canonicalization for IntlDateFormatter
+    and NumberFormatter). (alexandre-daubois)
+
 - Opcache:
   . Fixed bug GH-19486 (Incorrect opline after deoptimization). (Arnaud)
   . Fixed bug GH-19601 (Wrong JIT stack setup on aarch64/clang). (Arnaud)

UPGRADING

@@ -565,6 +565,10 @@ PHP 8.5 UPGRADE NOTES
     TransLiterator::getErrorCode(), and TransLiterator::getErrorMessage()
     have dropped the false from the return type union. Returning false
     was actually never possible.
+  . grapheme_strpos(), grapheme_stripos(), grapheme_strrpos(),
+    grapheme_strripos(), grapheme_strstr(), grapheme_stristr() and
+    grapheme_levenshtein() functions add $locale parameter.
+    RFC: https://wiki.php.net/rfc/grapheme_add_locale_for_case_insensitive
 
 - LDAP:
   . ldap_get_option() now accepts a NULL connection, as ldap_set_option(),

ext/filter/filter.c

@@ -627,6 +627,14 @@ static void php_filter_array_handler(zval *input, HashTable *op_ht, zend_long op
 				zval nval;
 				ZVAL_DEREF(tmp);
 				ZVAL_DUP(&nval, tmp);
+
+				if (Z_TYPE_P(arg_elm) != IS_ARRAY) {
+					zend_long filter_id = zval_get_long(arg_elm);
+					if (!PHP_FILTER_ID_EXISTS(filter_id)) {
+						php_error_docref(NULL, E_WARNING, "Unknown filter with ID " ZEND_LONG_FMT, filter_id);
+					}
+				}
+
 				php_filter_call(&nval, -1,
 					Z_TYPE_P(arg_elm) == IS_ARRAY ? Z_ARRVAL_P(arg_elm) : NULL,
 					Z_TYPE_P(arg_elm) == IS_ARRAY ? 0 : zval_get_long(arg_elm),

ext/filter/tests/039.phpt

@@ -125,12 +125,18 @@ array(1) {
   ["var_name"]=>
   NULL
 }
+
+Warning: filter_var_array(): Unknown filter with ID -1 in %s on line %d
 array(1) {
   ["var_name"]=>
   string(0) ""
 }
 -- (5)
+
+Warning: filter_var_array(): Unknown filter with ID -1 in %s on line %d
 filter_var_array(): Argument #2 ($options) cannot contain empty keys
+
+Warning: filter_var_array(): Unknown filter with ID 0 in %s on line %d
 filter_var_array(): Argument #2 ($options) cannot contain empty keys
 
 Warning: filter_var_array(): Unknown filter with ID -1 in %s on line %d

ext/filter/tests/gh16993.phpt

@@ -0,0 +1,24 @@
+--TEST--
+GH-16993: filter_var_array should emit warning for unknown filters
+--EXTENSIONS--
+filter
+--FILE--
+<?php
+$data = ['test' => '42'];
+
+$result = filter_var_array($data, ['test' => FILTER_VALIDATE_INT|FILTER_NULL_ON_FAILURE]);
+var_dump($result);
+
+$result = filter_var_array($data, ['test' => ['filter' => FILTER_VALIDATE_INT, 'flags' => FILTER_NULL_ON_FAILURE]]);
+var_dump($result);
+?>
+--EXPECTF--
+Warning: filter_var_array(): Unknown filter with ID 134217985 in %s on line %d
+array(1) {
+  ["test"]=>
+  string(2) "42"
+}
+array(1) {
+  ["test"]=>
+  int(42)
+}

ext/intl/dateformat/dateformat_create.cpp

@@ -22,6 +22,7 @@
 extern "C" {
 #include <unicode/ustring.h>
 #include <unicode/udat.h>
+#include <unicode/uloc.h>
 
 #include "php_intl.h"
 #include "dateformat_create.h"
@@ -106,7 +107,12 @@ static zend_result datefmt_ctor(INTERNAL_FUNCTION_PARAMETERS)
 	if (locale_len == 0) {
 		locale_str = (char *) intl_locale_get_default();
 	}
-	locale = Locale::createFromName(locale_str);
+
+	char* canonicalized_locale = canonicalize_locale_string(locale_str);
+	const char* final_locale = canonicalized_locale ? canonicalized_locale : locale_str;
+	const char* stored_locale = canonicalized_locale ? canonicalized_locale : locale_str;
+
+	locale = Locale::createFromName(final_locale);
 	/* get*Name accessors being set does not preclude being bogus */
 	if (locale.isBogus() || ((locale_len == 1 && locale_str[0] != 'C') || (locale_len > 1 && strlen(locale.getISO3Language()) == 0))) {
         zend_argument_value_error(1, "\"%s\" is invalid", locale_str);
@@ -144,7 +150,7 @@ static zend_result datefmt_ctor(INTERNAL_FUNCTION_PARAMETERS)
 	}
 
 	DATE_FORMAT_OBJECT(dfo) = udat_open((UDateFormatStyle)time_type,
-			(UDateFormatStyle)date_type, locale_str, NULL, 0, svalue,
+			(UDateFormatStyle)date_type, final_locale, NULL, 0, svalue,
 			slength, &INTL_DATA_ERROR_CODE(dfo));
 
 	if (pattern_str && pattern_str_len > 0) {
@@ -176,9 +182,13 @@ static zend_result datefmt_ctor(INTERNAL_FUNCTION_PARAMETERS)
 	dfo->date_type			= date_type;
 	dfo->time_type			= time_type;
 	dfo->calendar			= calendar_type;
-	dfo->requested_locale	= estrdup(locale_str);
+	/* Store the canonicalized locale, or fallback to original if canonicalization failed */
+	dfo->requested_locale	= estrdup(stored_locale);
 
 error:
+	if (canonicalized_locale) {
+		efree(canonicalized_locale);
+	}
 	if (svalue) {
 		efree(svalue);
 	}

ext/intl/formatter/formatter_main.c

@@ -64,12 +64,18 @@ static int numfmt_ctor(INTERNAL_FUNCTION_PARAMETERS)
 		return FAILURE;
 	}
 
-	/* Create an ICU number formatter. */
-	FORMATTER_OBJECT(nfo) = unum_open(style, spattern, spattern_len, locale, NULL, &INTL_DATA_ERROR_CODE(nfo));
+	char* canonicalized_locale = canonicalize_locale_string(locale);
+	const char* final_locale = canonicalized_locale ? canonicalized_locale : locale;
 
-	if(spattern) {
+	FORMATTER_OBJECT(nfo) = unum_open(style, spattern, spattern_len, final_locale, NULL, &INTL_DATA_ERROR_CODE(nfo));
+
+	if (spattern) {
 		efree(spattern);
 	}
+
+	if (canonicalized_locale) {
+		efree(canonicalized_locale);
+	}
 
 	INTL_CTOR_CHECK_STATUS(nfo, "number formatter creation failed");
 	return SUCCESS;

ext/intl/grapheme/grapheme_string.c

@@ -81,19 +81,20 @@ PHP_FUNCTION(grapheme_strlen)
 /* {{{ Find position of first occurrence of a string within another */
 PHP_FUNCTION(grapheme_strpos)
 {
-	char *haystack, *needle;
-	size_t haystack_len, needle_len;
+	char *haystack, *needle, *locale = "";
+	size_t haystack_len, needle_len, locale_len = 0;
 	const char *found;
 	zend_long loffset = 0;
 	int32_t offset = 0;
 	size_t noffset = 0;
 	zend_long ret_pos;
 
-	ZEND_PARSE_PARAMETERS_START(2, 3)
+	ZEND_PARSE_PARAMETERS_START(2, 4)
 		Z_PARAM_STRING(haystack, haystack_len)
 		Z_PARAM_STRING(needle, needle_len)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_LONG(loffset)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if ( OUTSIDE_STRING(loffset, haystack_len) ) {
@@ -121,7 +122,7 @@ PHP_FUNCTION(grapheme_strpos)
 	}
 
 	/* do utf16 part of the strpos */
-	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, 0 /* fIgnoreCase */, 0 /* last */ );
+	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, /* fIgnoreCase */ 0, /* last */ 0, locale);
 
 	if ( ret_pos >= 0 ) {
 		RETURN_LONG(ret_pos);
@@ -134,19 +135,20 @@ PHP_FUNCTION(grapheme_strpos)
 /* {{{ Find position of first occurrence of a string within another, ignoring case differences */
 PHP_FUNCTION(grapheme_stripos)
 {
-	char *haystack, *needle;
-	size_t haystack_len, needle_len;
+	char *haystack, *needle, *locale = "";
+	size_t haystack_len, needle_len, locale_len = 0;
 	const char *found;
 	zend_long loffset = 0;
 	int32_t offset = 0;
 	zend_long ret_pos;
 	int is_ascii;
 
-	ZEND_PARSE_PARAMETERS_START(2, 3)
+	ZEND_PARSE_PARAMETERS_START(2, 4)
 		Z_PARAM_STRING(haystack, haystack_len)
 		Z_PARAM_STRING(needle, needle_len)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_LONG(loffset)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if ( OUTSIDE_STRING(loffset, haystack_len) ) {
@@ -185,7 +187,7 @@ PHP_FUNCTION(grapheme_stripos)
 	}
 
 	/* do utf16 part of the strpos */
-	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, 1 /* fIgnoreCase */, 0 /*last */ );
+	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, /* fIgnoreCase */ 1, /*last */ 0, locale);
 
 	if ( ret_pos >= 0 ) {
 		RETURN_LONG(ret_pos);
@@ -200,17 +202,19 @@ PHP_FUNCTION(grapheme_stripos)
 PHP_FUNCTION(grapheme_strrpos)
 {
 	char *haystack, *needle;
-	size_t haystack_len, needle_len;
+	char *locale = "";
+	size_t haystack_len, needle_len, locale_len = 0;
 	zend_long loffset = 0;
 	int32_t offset = 0;
 	zend_long ret_pos;
 	int is_ascii;
 
-	ZEND_PARSE_PARAMETERS_START(2, 3)
+	ZEND_PARSE_PARAMETERS_START(2, 4)
 		Z_PARAM_STRING(haystack, haystack_len)
 		Z_PARAM_STRING(needle, needle_len)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_LONG(loffset)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if ( OUTSIDE_STRING(loffset, haystack_len) ) {
@@ -242,7 +246,7 @@ PHP_FUNCTION(grapheme_strrpos)
 		/* else we need to continue via utf16 */
 	}
 
-	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, 0 /* f_ignore_case */, 1/* last */);
+	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, /* f_ignore_case */ 0, /* last */ 1, locale);
 
 	if ( ret_pos >= 0 ) {
 		RETURN_LONG(ret_pos);
@@ -257,18 +261,19 @@ PHP_FUNCTION(grapheme_strrpos)
 /* {{{ Find position of last occurrence of a string within another, ignoring case */
 PHP_FUNCTION(grapheme_strripos)
 {
-	char *haystack, *needle;
-	size_t haystack_len, needle_len;
+	char *haystack, *needle, *locale = "";
+	size_t haystack_len, needle_len, locale_len = 0;
 	zend_long loffset = 0;
 	int32_t offset = 0;
 	zend_long ret_pos;
 	int is_ascii;
 
-	ZEND_PARSE_PARAMETERS_START(2, 3)
+	ZEND_PARSE_PARAMETERS_START(2, 4)
 		Z_PARAM_STRING(haystack, haystack_len)
 		Z_PARAM_STRING(needle, needle_len)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_LONG(loffset)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if ( OUTSIDE_STRING(loffset, haystack_len) ) {
@@ -309,7 +314,7 @@ PHP_FUNCTION(grapheme_strripos)
 		/* else we need to continue via utf16 */
 	}
 
-	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL,  1 /* f_ignore_case */, 1 /*last */);
+	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, offset, NULL, /* f_ignore_case */ 1, /*last */ 1, locale);
 
 	if ( ret_pos >= 0 ) {
 		RETURN_LONG(ret_pos);
@@ -324,10 +329,10 @@ PHP_FUNCTION(grapheme_strripos)
 /* {{{ Returns part of a string */
 PHP_FUNCTION(grapheme_substr)
 {
-	char *str;
+	char *str, *locale = "";
 	zend_string *u8_sub_str;
 	UChar *ustr;
-	size_t str_len;
+	size_t str_len, locale_len = 0;
 	int32_t ustr_len;
 	zend_long lstart = 0, length = 0;
 	int32_t start = 0;
@@ -339,11 +344,12 @@ PHP_FUNCTION(grapheme_substr)
 	int32_t (*iter_func)(UBreakIterator *);
 	bool no_length = true;
 
-	ZEND_PARSE_PARAMETERS_START(2, 3)
+	ZEND_PARSE_PARAMETERS_START(2, 4)
 		Z_PARAM_STRING(str, str_len)
 		Z_PARAM_LONG(lstart)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_LONG_OR_NULL(length, no_length)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if (lstart < INT32_MIN || lstart > INT32_MAX) {
@@ -537,17 +543,18 @@ PHP_FUNCTION(grapheme_substr)
 /* {{{	strstr_common_handler */
 static void strstr_common_handler(INTERNAL_FUNCTION_PARAMETERS, int f_ignore_case)
 {
-	char *haystack, *needle;
+	char *haystack, *needle, *locale = "";
 	const char *found;
-	size_t haystack_len, needle_len;
+	size_t haystack_len, needle_len, locale_len = 0;
 	int32_t ret_pos, uchar_pos;
 	bool part = false;
 
-	ZEND_PARSE_PARAMETERS_START(2, 3)
+	ZEND_PARSE_PARAMETERS_START(2, 4)
 		Z_PARAM_STRING(haystack, haystack_len)
 		Z_PARAM_STRING(needle, needle_len)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_BOOL(part)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if ( !f_ignore_case ) {
@@ -574,7 +581,7 @@ static void strstr_common_handler(INTERNAL_FUNCTION_PARAMETERS, int f_ignore_cas
 	}
 
 	/* need to work in utf16 */
-	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, 0, &uchar_pos, f_ignore_case, 0 /*last */ );
+	ret_pos = grapheme_strpos_utf16(haystack, haystack_len, needle, needle_len, 0, &uchar_pos, f_ignore_case, /* last */ 0, locale);
 
 	if ( ret_pos < 0 ) {
 		RETURN_FALSE;
@@ -919,14 +926,17 @@ PHP_FUNCTION(grapheme_levenshtein)
 	zend_long cost_ins = 1;
 	zend_long cost_rep = 1;
 	zend_long cost_del = 1;
+	char *locale = "";
+	size_t locale_len = 0;
 
-	ZEND_PARSE_PARAMETERS_START(2, 5)
+	ZEND_PARSE_PARAMETERS_START(2, 6)
 		Z_PARAM_STR(string1)
 		Z_PARAM_STR(string2)
 		Z_PARAM_OPTIONAL
 		Z_PARAM_LONG(cost_ins)
 		Z_PARAM_LONG(cost_rep)
 		Z_PARAM_LONG(cost_del)
+		Z_PARAM_PATH(locale, locale_len)
 	ZEND_PARSE_PARAMETERS_END();
 
 	if (cost_ins <= 0 || cost_ins > UINT_MAX / 4) {
@@ -1043,7 +1053,7 @@ PHP_FUNCTION(grapheme_levenshtein)
 		RETVAL_FALSE;
 		goto out_bi2;
 	}
-	UCollator *collator = ucol_open("", &ustatus);
+	UCollator *collator = ucol_open(locale, &ustatus);
 	if (U_FAILURE(ustatus)) {
 		intl_error_set_code(NULL, ustatus);
 

ext/intl/grapheme/grapheme_util.c

@@ -94,7 +94,7 @@ void grapheme_substr_ascii(char *str, size_t str_len, int32_t f, int32_t l, char
 
 
 /* {{{ grapheme_strpos_utf16 - strrpos using utf16*/
-int32_t grapheme_strpos_utf16(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset, int32_t *puchar_pos, int f_ignore_case, int last)
+int32_t grapheme_strpos_utf16(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset, int32_t *puchar_pos, int f_ignore_case, int last, const char* locale)
 {
 	UChar *uhaystack = NULL, *uneedle = NULL;
 	int32_t uhaystack_len = 0, uneedle_len = 0, char_pos, ret_pos, offset_pos = 0;
@@ -136,7 +136,7 @@ int32_t grapheme_strpos_utf16(char *haystack, size_t haystack_len, char *needle,
 	}
 
 	status = U_ZERO_ERROR;
-	src = usearch_open(uneedle, uneedle_len, uhaystack, uhaystack_len, "", bi, &status);
+	src = usearch_open(uneedle, uneedle_len, uhaystack, uhaystack_len, locale, bi, &status);
 	STRPOS_CHECK_STATUS(status, "Error creating search object");
 
 	if(f_ignore_case) {

ext/intl/grapheme/grapheme_util.h

@@ -25,8 +25,8 @@ zend_long grapheme_ascii_check(const unsigned char *day, size_t len);
 void grapheme_substr_ascii(char *str, size_t str_len, int32_t f, int32_t l, char **sub_str, int32_t *sub_str_len);
 zend_long grapheme_strrpos_ascii(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset);
 
-int32_t grapheme_strrpos_utf16(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset, int f_ignore_case);
-int32_t grapheme_strpos_utf16(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset, int *puchar_pos, int f_ignore_case, int last);
+int32_t grapheme_strrpos_utf16(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset, int f_ignore_case, const char *locale);
+int32_t grapheme_strpos_utf16(char *haystack, size_t haystack_len, char *needle, size_t needle_len, int32_t offset, int *puchar_pos, int f_ignore_case, int last, const char *locale);
 
 int32_t grapheme_split_string(const UChar *text, int32_t text_length, int boundary_array[], int boundary_array_len );