GH-18344 add Locale::addLikelySubtags/Locale::minimizeSubtags support. (#18487)

from a minimized locale, addLikelySubtags augments it with likely
subtags so no changes is the locale is already maximized e.g.
`en_Latn_US`, minimizeSubtags on the other hand does the opposite
operation.

NEWS

@@ -94,6 +94,8 @@ PHP                                                                        NEWS
     (David Carlier)
   . Added null bytes presence in locale inputs for Locale class. (David Carlier)
   . Added grapheme_levenshtein() function. (Yuya Hamada)
+  . Added Locale::addLikelySubtags/Locale::minimizeSubtags to handle
+    adding/removing likely subtags to a locale. (David Carlier)
 
 - MySQLi:
   . Fixed bugs GH-17900 and GH-8084 (calling mysqli::__construct twice).

UPGRADING

@@ -178,6 +178,8 @@ PHP 8.5 UPGRADE NOTES
     NumberFormatter::CURRENCY_PLURAL, NumberFormatter::CASH_CURRENCY,
     and NumberFormatter::CURRENCY_STANDARD for various currency-related
     number formats.
+  . Added Locale::addLikelySubtags and Locale::minimizeSubtags to
+    handle likely tags on a given locale.
 
 - XSL:
   . The $namespace argument of XSLTProcessor::getParameter(),

ext/intl/locale/locale.stub.php

@@ -138,4 +138,14 @@ class Locale
      * @alias locale_is_right_to_left
      */
     public static function isRightToLeft(string $locale): bool {}
+
+    /**
+     * @alias locale_add_likely_subtags
+     */
+    public static function addLikelySubtags(string $locale): string|false {}
+
+    /**
+     * @alias locale_minimize_subtags
+     */
+    public static function minimizeSubtags(string $locale): string|false {}
 }

ext/intl/locale/locale_arginfo.h

@@ -1,5 +1,5 @@
 /* This is a generated file, edit the .stub.php file instead.
- * Stub hash: f09cfd61f3e20576c7f6d5da17a6d9c009d6ab64 */
+ * Stub hash: ff1f75bd34a52f57210734e2f5e29efb87566137 */
 
 ZEND_BEGIN_ARG_WITH_TENTATIVE_RETURN_TYPE_INFO_EX(arginfo_class_Locale_getDefault, 0, 0, IS_STRING, 0)
 ZEND_END_ARG_INFO()
@@ -66,6 +66,12 @@ ZEND_BEGIN_ARG_WITH_RETURN_TYPE_INFO_EX(arginfo_class_Locale_isRightToLeft, 0, 1
 	ZEND_ARG_TYPE_INFO(0, locale, IS_STRING, 0)
 ZEND_END_ARG_INFO()
 
+ZEND_BEGIN_ARG_WITH_RETURN_TYPE_MASK_EX(arginfo_class_Locale_addLikelySubtags, 0, 1, MAY_BE_STRING|MAY_BE_FALSE)
+	ZEND_ARG_TYPE_INFO(0, locale, IS_STRING, 0)
+ZEND_END_ARG_INFO()
+
+#define arginfo_class_Locale_minimizeSubtags arginfo_class_Locale_addLikelySubtags
+
 ZEND_FUNCTION(locale_get_default);
 ZEND_FUNCTION(locale_set_default);
 ZEND_FUNCTION(locale_get_primary_language);
@@ -85,6 +91,8 @@ ZEND_FUNCTION(locale_lookup);
 ZEND_FUNCTION(locale_canonicalize);
 ZEND_FUNCTION(locale_accept_from_http);
 ZEND_FUNCTION(locale_is_right_to_left);
+ZEND_FUNCTION(locale_add_likely_subtags);
+ZEND_FUNCTION(locale_minimize_subtags);
 
 static const zend_function_entry class_Locale_methods[] = {
 	ZEND_RAW_FENTRY("getDefault", zif_locale_get_default, arginfo_class_Locale_getDefault, ZEND_ACC_PUBLIC|ZEND_ACC_STATIC, NULL, NULL)
@@ -106,6 +114,8 @@ static const zend_function_entry class_Locale_methods[] = {
 	ZEND_RAW_FENTRY("canonicalize", zif_locale_canonicalize, arginfo_class_Locale_canonicalize, ZEND_ACC_PUBLIC|ZEND_ACC_STATIC, NULL, NULL)
 	ZEND_RAW_FENTRY("acceptFromHttp", zif_locale_accept_from_http, arginfo_class_Locale_acceptFromHttp, ZEND_ACC_PUBLIC|ZEND_ACC_STATIC, NULL, NULL)
 	ZEND_RAW_FENTRY("isRightToLeft", zif_locale_is_right_to_left, arginfo_class_Locale_isRightToLeft, ZEND_ACC_PUBLIC|ZEND_ACC_STATIC, NULL, NULL)
+	ZEND_RAW_FENTRY("addLikelySubtags", zif_locale_add_likely_subtags, arginfo_class_Locale_addLikelySubtags, ZEND_ACC_PUBLIC|ZEND_ACC_STATIC, NULL, NULL)
+	ZEND_RAW_FENTRY("minimizeSubtags", zif_locale_minimize_subtags, arginfo_class_Locale_minimizeSubtags, ZEND_ACC_PUBLIC|ZEND_ACC_STATIC, NULL, NULL)
 	ZEND_FE_END
 };
 

ext/intl/locale/locale_methods.c

@@ -1639,3 +1639,49 @@ PHP_FUNCTION(locale_is_right_to_left)
 
 	RETURN_BOOL(uloc_isRightToLeft(locale));
 }
+
+PHP_FUNCTION(locale_add_likely_subtags)
+{
+	char *locale, maximized_locale[ULOC_FULLNAME_CAPACITY];
+	UErrorCode status = 0;
+	size_t locale_len;
+
+	ZEND_PARSE_PARAMETERS_START(1, 1)
+		Z_PARAM_PATH(locale, locale_len)
+	ZEND_PARSE_PARAMETERS_END();
+
+	if (!locale_len) {
+		locale = (char *)intl_locale_get_default();
+	}
+
+	int32_t maximized_locale_len = uloc_addLikelySubtags(locale, maximized_locale, sizeof(maximized_locale), &status);
+	INTL_CHECK_STATUS(status, "locale_add_likely_subtags: invalid locale");
+	if (maximized_locale_len < 0) {
+		RETURN_FALSE;
+	}
+
+	RETURN_STRINGL(maximized_locale, maximized_locale_len);
+}
+
+PHP_FUNCTION(locale_minimize_subtags)
+{
+	char *locale, minimized_locale[ULOC_FULLNAME_CAPACITY];
+	UErrorCode status = 0;
+	size_t locale_len;
+
+	ZEND_PARSE_PARAMETERS_START(1, 1)
+		Z_PARAM_PATH(locale, locale_len)
+	ZEND_PARSE_PARAMETERS_END();
+
+	if (!locale_len) {
+		locale = (char *)intl_locale_get_default();
+	}
+
+	int32_t minimized_locale_len = uloc_minimizeSubtags(locale, minimized_locale, sizeof(minimized_locale), &status);
+	INTL_CHECK_STATUS(status, "locale_minimize_subtags: invalid locale");
+	if (minimized_locale_len < 0) {
+		RETURN_FALSE;
+	}
+
+	RETURN_STRINGL(minimized_locale, minimized_locale_len);
+}

ext/intl/php_intl.stub.php

@@ -505,6 +505,10 @@ function locale_accept_from_http(string $header): string|false {}
 
 function locale_is_right_to_left(string $locale): bool {}
 
+function locale_add_likely_subtags(string $locale): string|false {}
+
+function locale_minimize_subtags(string $locale): string|false {}
+
 /* msgformat */
 
 function msgfmt_create(string $locale, string $pattern): ?MessageFormatter {}

ext/intl/php_intl_arginfo.h

@@ -1,5 +1,5 @@
 /* This is a generated file, edit the .stub.php file instead.
- * Stub hash: 70b621ef9169fd3b913347adc0baf3626584a2c3 */
+ * Stub hash: 0d5b028a1ab8f35e8ee1b51ce3141b6ef782af28 */
 
 ZEND_BEGIN_ARG_WITH_RETURN_OBJ_INFO_EX(arginfo_intlcal_create_instance, 0, 0, IntlCalendar, 1)
 	ZEND_ARG_INFO_WITH_DEFAULT_VALUE(0, timezone, "null")
@@ -578,6 +578,12 @@ ZEND_BEGIN_ARG_WITH_RETURN_TYPE_INFO_EX(arginfo_locale_is_right_to_left, 0, 1, _
 	ZEND_ARG_TYPE_INFO(0, locale, IS_STRING, 0)
 ZEND_END_ARG_INFO()
 
+ZEND_BEGIN_ARG_WITH_RETURN_TYPE_MASK_EX(arginfo_locale_add_likely_subtags, 0, 1, MAY_BE_STRING|MAY_BE_FALSE)
+	ZEND_ARG_TYPE_INFO(0, locale, IS_STRING, 0)
+ZEND_END_ARG_INFO()
+
+#define arginfo_locale_minimize_subtags arginfo_locale_add_likely_subtags
+
 ZEND_BEGIN_ARG_WITH_RETURN_OBJ_INFO_EX(arginfo_msgfmt_create, 0, 2, MessageFormatter, 1)
 	ZEND_ARG_TYPE_INFO(0, locale, IS_STRING, 0)
 	ZEND_ARG_TYPE_INFO(0, pattern, IS_STRING, 0)
@@ -934,6 +940,8 @@ ZEND_FUNCTION(locale_canonicalize);
 ZEND_FUNCTION(locale_lookup);
 ZEND_FUNCTION(locale_accept_from_http);
 ZEND_FUNCTION(locale_is_right_to_left);
+ZEND_FUNCTION(locale_add_likely_subtags);
+ZEND_FUNCTION(locale_minimize_subtags);
 ZEND_FUNCTION(msgfmt_create);
 ZEND_FUNCTION(msgfmt_format);
 ZEND_FUNCTION(msgfmt_format_message);
@@ -1123,6 +1131,8 @@ static const zend_function_entry ext_functions[] = {
 	ZEND_FE(locale_lookup, arginfo_locale_lookup)
 	ZEND_FE(locale_accept_from_http, arginfo_locale_accept_from_http)
 	ZEND_FE(locale_is_right_to_left, arginfo_locale_is_right_to_left)
+	ZEND_FE(locale_add_likely_subtags, arginfo_locale_add_likely_subtags)
+	ZEND_FE(locale_minimize_subtags, arginfo_locale_minimize_subtags)
 	ZEND_FE(msgfmt_create, arginfo_msgfmt_create)
 	ZEND_FE(msgfmt_format, arginfo_msgfmt_format)
 	ZEND_FE(msgfmt_format_message, arginfo_msgfmt_format_message)

ext/intl/tests/locale_subtags.phpt

@@ -0,0 +1,35 @@
+--TEST--
+Locale::addLikelySubtags/Locale::minimizeSubtags usage
+--EXTENSIONS--
+intl
+--FILE--
+<?php
+$locale = "en";
+$max = Locale::addLikelySubtags($locale);
+$min = Locale::minimizeSubtags($max);
+var_dump($min === $locale);
+var_dump($max !== $locale && strlen($max) > strlen($locale));
+var_dump(Locale::addLikelySubtags($max) === $max);
+var_dump(Locale::minimizeSubtags($locale) === $locale);
+var_dump(Locale::addLikelySubtags("%%%invalid%%%locale%%%"));
+var_dump(intl_get_error_message());
+var_dump(Locale::minimizeSubtags("%%%Invalid%%%maximized%%%locale%%%"));
+var_dump(intl_get_error_message());
+var_dump(Locale::addLikelySubTags(str_repeat($locale, 1024)));
+var_dump(intl_get_error_message());
+var_dump(Locale::minimizeSubTags(str_repeat($max, 1024)));
+var_dump(intl_get_error_message());
+?>
+--EXPECTF--
+bool(true)
+bool(true)
+bool(true)
+bool(true)
+bool(false)
+string(67) "locale_add_likely_subtags: invalid locale: U_ILLEGAL_ARGUMENT_ERROR"
+bool(false)
+string(65) "locale_minimize_subtags: invalid locale: U_ILLEGAL_ARGUMENT_ERROR"
+bool(false)
+string(%d) "locale_add_likely_subtags: invalid locale: %s"
+bool(false)
+string(%d) "locale_minimize_subtags: invalid locale: %s"