[RFC] Add mb_ucfirst and mb_lcfirst functions (#13161)

2026-03-24 00:02:20 +01:00 · 2024-03-21 01:25:19 +09:00
parent 5430ecd3db
commit 4d51bfa270
6 changed files with 168 additions and 1 deletions
--- a/1
+++ b/1
@@ -90,6 +90,7 @@ PHP                                                                        NEWS

 - MBString:
  . Added mb_trim, mb_ltrim and mb_rtrim. (Yuya Hamada)
+  . Added mb_ucfirst and mb_lcfirst. (Yuya Hamada)

 - MySQLnd:
  . Fixed bug GH-13440 (PDO quote bottleneck). (nielsdos)
--- a/2
+++ b/2
@@ -431,6 +431,8 @@ PHP 8.4 UPGRADE NOTES
 - MBString:
  . Added mb_trim, mb_ltrim and mb_rtrim functions.
    RFC: https://wiki.php.net/rfc/mb_trim
+  . Added mb_ucfirst and mb_lcfirst functions.
+    RFC: https://wiki.php.net/rfc/mb_ucfirst

 - Opcache:
  . If JIT is enabled, PHP will now exit with a fatal error on startup in case
--- a/ext/mbstring/mbstring.c
+++ b/ext/mbstring/mbstring.c
@@ -2951,6 +2951,50 @@ PHP_FUNCTION(mb_strtolower)
 	RETURN_STR(mbstring_convert_case(PHP_UNICODE_CASE_LOWER, ZSTR_VAL(str), ZSTR_LEN(str), enc));
 }

+static void php_mb_ulcfirst(INTERNAL_FUNCTION_PARAMETERS, php_case_mode mode)
+{
+	zend_string *str, *from_encoding = NULL;
+
+	ZEND_PARSE_PARAMETERS_START(1, 2)
+		Z_PARAM_STR(str)
+		Z_PARAM_OPTIONAL
+		Z_PARAM_STR_OR_NULL(from_encoding)
+	ZEND_PARSE_PARAMETERS_END();
+
+	const mbfl_encoding *enc = php_mb_get_encoding(from_encoding, 2);
+	if (!enc) {
+		RETURN_THROWS();
+	}
+
+	zend_string *first = mb_get_substr(str, 0, 1, enc);
+	zend_string *head = mbstring_convert_case(mode, ZSTR_VAL(first), ZSTR_LEN(first), enc);
+
+	if (zend_string_equals(first, head)) {
+		zend_string_release_ex(first, false);
+		zend_string_release_ex(head, false);
+		RETURN_STR(zend_string_copy(str));
+	}
+
+	zend_string *second = mb_get_substr(str, 1, MBFL_SUBSTR_UNTIL_END, enc);
+	zend_string *retval = zend_string_concat2(ZSTR_VAL(head), ZSTR_LEN(head), ZSTR_VAL(second), ZSTR_LEN(second));
+
+	zend_string_release_ex(first, false);
+	zend_string_release_ex(head, false);
+	zend_string_release_ex(second, false);
+
+	RETVAL_STR(retval);
+}
+
+PHP_FUNCTION(mb_ucfirst)
+{
+	php_mb_ulcfirst(INTERNAL_FUNCTION_PARAM_PASSTHRU, PHP_UNICODE_CASE_TITLE);
+}
+
+PHP_FUNCTION(mb_lcfirst)
+{
+	php_mb_ulcfirst(INTERNAL_FUNCTION_PARAM_PASSTHRU, PHP_UNICODE_CASE_LOWER);
+}
+
 typedef enum {
 	MB_LTRIM = 1,
 	MB_RTRIM = 2,
--- a/ext/mbstring/mbstring.stub.php
+++ b/ext/mbstring/mbstring.stub.php
@@ -135,6 +135,10 @@ function mb_strtoupper(string $string, ?string $encoding = null): string {}
 /** @refcount 1 */
 function mb_strtolower(string $string, ?string $encoding = null): string {}

+function mb_ucfirst(string $string, ?string $encoding = null): string {}
+
+function mb_lcfirst(string $string, ?string $encoding = null): string {}
+
 function mb_trim(string $string, string $characters = " \f\n\r\t\v\x00\u{00A0}\u{1680}\u{2000}\u{2001}\u{2002}\u{2003}\u{2004}\u{2005}\u{2006}\u{2007}\u{2008}\u{2009}\u{200A}\u{2028}\u{2029}\u{202F}\u{205F}\u{3000}\u{0085}\u{180E}", ?string $encoding = null): string {}

 function mb_ltrim(string $string, string $characters = " \f\n\r\t\v\x00\u{00A0}\u{1680}\u{2000}\u{2001}\u{2002}\u{2003}\u{2004}\u{2005}\u{2006}\u{2007}\u{2008}\u{2009}\u{200A}\u{2028}\u{2029}\u{202F}\u{205F}\u{3000}\u{0085}\u{180E}", ?string $encoding = null): string {}
--- a/ext/mbstring/mbstring_arginfo.h
+++ b/ext/mbstring/mbstring_arginfo.h
@@ -1,5 +1,5 @@
 /* This is a generated file, edit the .stub.php file instead.
- * Stub hash: 0fe7418224f3a0ab65a06bf215ebcb09ce632488 */
+ * Stub hash: ea642b9010bc38a3b13710662fef48663d4385e1 */

 ZEND_BEGIN_ARG_WITH_RETURN_TYPE_MASK_EX(arginfo_mb_language, 0, 0, MAY_BE_STRING|MAY_BE_BOOL)
 	ZEND_ARG_TYPE_INFO_WITH_DEFAULT_VALUE(0, language, IS_STRING, 1, "null")
@@ -118,6 +118,10 @@ ZEND_END_ARG_INFO()

 #define arginfo_mb_strtolower arginfo_mb_strtoupper

+#define arginfo_mb_ucfirst arginfo_mb_strtoupper
+
+#define arginfo_mb_lcfirst arginfo_mb_strtoupper
+
 ZEND_BEGIN_ARG_WITH_RETURN_TYPE_INFO_EX(arginfo_mb_trim, 0, 1, IS_STRING, 0)
 	ZEND_ARG_TYPE_INFO(0, string, IS_STRING, 0)
 	ZEND_ARG_TYPE_INFO_WITH_DEFAULT_VALUE(0, characters, IS_STRING, 0, "\" \\f\\n\\r\\t\\v\\x00                 　᠎\"")
@@ -348,6 +352,8 @@ ZEND_FUNCTION(mb_convert_encoding);
 ZEND_FUNCTION(mb_convert_case);
 ZEND_FUNCTION(mb_strtoupper);
 ZEND_FUNCTION(mb_strtolower);
+ZEND_FUNCTION(mb_ucfirst);
+ZEND_FUNCTION(mb_lcfirst);
 ZEND_FUNCTION(mb_trim);
 ZEND_FUNCTION(mb_ltrim);
 ZEND_FUNCTION(mb_rtrim);
@@ -445,6 +451,8 @@ static const zend_function_entry ext_functions[] = {
 	ZEND_FE(mb_convert_case, arginfo_mb_convert_case)
 	ZEND_FE(mb_strtoupper, arginfo_mb_strtoupper)
 	ZEND_FE(mb_strtolower, arginfo_mb_strtolower)
+	ZEND_FE(mb_ucfirst, arginfo_mb_ucfirst)
+	ZEND_FE(mb_lcfirst, arginfo_mb_lcfirst)
 	ZEND_FE(mb_trim, arginfo_mb_trim)
 	ZEND_FE(mb_ltrim, arginfo_mb_ltrim)
 	ZEND_FE(mb_rtrim, arginfo_mb_rtrim)
--- a/ext/mbstring/tests/mb_ucfirst_lcfirst.phpt
+++ b/ext/mbstring/tests/mb_ucfirst_lcfirst.phpt
@@ -0,0 +1,108 @@
+--TEST--
+mb_ucfirst(), mb_lcfirst functions tests
+--EXTENSIONS--
+mbstring
+--FILE--
+<?php
+mb_internal_encoding("UTF-8");
+
+function test_ascii_mb_ucfirst() {
+	for ($i = 0; $i < 128; $i++) {
+		if ($i >= 97 && $i <= 122) { /* a to z */
+			if (mb_ucfirst(chr($i)) !== chr($i - (97 - 65))) {
+				echo "mb_ucfirst compare failed: " . chr($i) . "\n";
+			}
+		} else {
+			if (mb_ucfirst(chr($i)) !== chr($i)) {
+				echo "mb_ucfirst compare failed: " . chr($i) . "\n";
+			}
+		}
+	}
+	echo "Done mb_ucfirst\n";
+}
+
+function test_ascii_mb_lcfirst() {
+	for ($i = 0; $i < 128; $i++) {
+		if ($i >= 65 && $i <= 90) { /* A to Z */
+			if (mb_lcfirst(chr($i)) !== chr($i + (97 - 65))) {
+				echo "mb_lcfirst compare failed: " . chr($i) . "\n";
+				return;
+			}
+		} else {
+			if (mb_lcfirst(chr($i)) !== chr($i)) {
+				echo "mb_lcfirst compare failed: " . chr($i) . "\n";
+				return;
+			}
+		}
+	}
+	echo "Done mb_lcfirst\n";
+}
+
+echo "== Empty String ==\n";
+var_dump(mb_ucfirst(""));
+var_dump(mb_lcfirst(""));
+echo "== ASCII ==\n";
+test_ascii_mb_ucfirst();
+test_ascii_mb_lcfirst();
+echo "== mb_ucfirst ==\n";
+var_dump(mb_ucfirst("ａｂ"));
+var_dump(mb_ucfirst("ＡＢＳ"));
+var_dump(mb_ucfirst("đắt quá!"));
+var_dump(mb_ucfirst("აბგ"));
+var_dump(mb_ucfirst("ǉ"));
+echo "== mb_lcfirst ==\n";
+var_dump(mb_lcfirst("ＡＢＳ"));
+var_dump(mb_lcfirst("Xin chào"));
+var_dump(mb_lcfirst("Đẹp quá!"));
+echo "== SJIS ==\n";
+var_dump(bin2hex(mb_ucfirst(mb_convert_encoding("ｅｂｉ", "SJIS", "UTF-8"), "SJIS")));
+var_dump(bin2hex(mb_lcfirst(mb_convert_encoding("ＥＢＩ", "SJIS", "UTF-8"), "SJIS")));
+var_dump(bin2hex(mb_ucfirst(hex2bin("8471"), "SJIS"))); /* б */
+var_dump(bin2hex(mb_lcfirst(hex2bin("8441"), "SJIS"))); /* Б */
+var_dump(bin2hex(mb_ucfirst(hex2bin("83bf"), "SJIS"))); /* α */
+var_dump(bin2hex(mb_lcfirst(hex2bin("839f"), "SJIS"))); /* Α */
+var_dump(bin2hex(mb_lcfirst(hex2bin("82a0"), "SJIS"))); /* あ */
+var_dump(bin2hex(mb_ucfirst(hex2bin("83bf8471"), "SJIS")));
+var_dump(bin2hex(mb_lcfirst(hex2bin("839f8441"), "SJIS")));
+echo "== EUC-JP ==\n";
+var_dump(bin2hex(mb_ucfirst(hex2bin("a6d8"), "EUC-JP"))); /* Ω */
+var_dump(bin2hex(mb_lcfirst(hex2bin("a6b8"), "EUC-JP"))); /* ω */
+var_dump(bin2hex(mb_ucfirst(hex2bin("a4a2a4a2"), "EUC-JP"))); /* あ */
+echo "== Longer strings ==\n";
+var_dump(mb_ucfirst("э" . str_repeat("A", 65536)) === "Э" . str_repeat("A", 65536));
+var_dump(mb_lcfirst("Э" . str_repeat("A", 65536)) === "э" . str_repeat("A", 65536));
+?>
+--EXPECT--
+== Empty String ==
+string(0) ""
+string(0) ""
+== ASCII ==
+Done mb_ucfirst
+Done mb_lcfirst
+== mb_ucfirst ==
+string(6) "Ａｂ"
+string(9) "ＡＢＳ"
+string(12) "Đắt quá!"
+string(9) "აბგ"
+string(2) "ǈ"
+== mb_lcfirst ==
+string(9) "ａＢＳ"
+string(9) "xin chào"
+string(12) "đẹp quá!"
+== SJIS ==
+string(12) "826482828289"
+string(12) "828582618268"
+string(4) "8441"
+string(4) "8471"
+string(4) "839f"
+string(4) "83bf"
+string(4) "82a0"
+string(8) "839f8471"
+string(8) "83bf8441"
+== EUC-JP ==
+string(4) "a6b8"
+string(4) "a6d8"
+string(8) "a4a2a4a2"
+== Longer strings ==
+bool(true)
+bool(true)