Request #65081 - implemeting mb_scrub

ext/mbstring/mbstring.c

@@ -437,6 +437,11 @@ ZEND_BEGIN_ARG_INFO_EX(arginfo_mb_check_encoding, 0, 0, 0)
 	ZEND_ARG_INFO(0, encoding)
 ZEND_END_ARG_INFO()
 
+ZEND_BEGIN_ARG_INFO_EX(arginfo_mb_scrub, 0, 0, 1)
+	ZEND_ARG_INFO(0, str)
+	ZEND_ARG_INFO(0, encoding)
+ZEND_END_ARG_INFO()
+
 ZEND_BEGIN_ARG_INFO_EX(arginfo_mb_regex_encoding, 0, 0, 0)
 	ZEND_ARG_INFO(0, encoding)
 ZEND_END_ARG_INFO()
@@ -562,6 +567,7 @@ const zend_function_entry mbstring_functions[] = {
 	PHP_FE(mb_send_mail,			arginfo_mb_send_mail)
 	PHP_FE(mb_get_info,				arginfo_mb_get_info)
 	PHP_FE(mb_check_encoding,		arginfo_mb_check_encoding)
+	PHP_FE(mb_scrub,		arginfo_mb_scrub)
 #if HAVE_MBREGEX
 	PHP_MBREGEX_FUNCTION_ENTRIES
 #endif
@@ -4570,6 +4576,53 @@ PHP_FUNCTION(mb_check_encoding)
 }
 /* }}} */
 
+static inline char* php_mb_scrub(const char* str, size_t str_len, const char* enc)
+{
+	size_t ret_len;
+
+	return php_mb_convert_encoding(str, str_len, enc, enc, &ret_len);
+}
+
+/* {{{ proto bool mb_scrub([string str[, string encoding]]) */
+PHP_FUNCTION(mb_scrub)
+{
+	char* str;
+	size_t str_len;
+	char *enc = NULL;
+	size_t enc_len;
+
+	char *ret;
+
+#ifndef FAST_ZPP
+	if (zend_parse_parameters(ZEND_NUM_ARGS(), "s|s", &str, &str_len, &enc, &enc_len) == FAILURE) {
+		return;
+	}
+#else
+ZEND_PARSE_PARAMETERS_START(1, 2)
+		Z_PARAM_STRING(str, str_len)
+		Z_PARAM_OPTIONAL
+		Z_PARAM_STRING(enc, enc_len)
+ZEND_PARSE_PARAMETERS_END();
+#endif
+
+	if (enc == NULL) {
+		enc = (char *) MBSTRG(current_internal_encoding)->name;
+	} else if (!mbfl_is_support_encoding(enc)) {
+		php_error_docref(NULL, E_WARNING, "Unknown encoding \"%s\"", enc);
+		RETURN_FALSE; 
+	}
+
+	ret = php_mb_scrub(str, str_len, enc);
+
+	if (ret == NULL) {
+		RETURN_FALSE;
+	}
+
+	RETVAL_STRING(ret);
+	efree(ret);
+}
+/* }}} */
+
 /* {{{ php_mb_populate_current_detect_order_list */
 static void php_mb_populate_current_detect_order_list(void)
 {

ext/mbstring/mbstring.h

@@ -127,6 +127,7 @@ PHP_FUNCTION(mb_decode_numericentity);
 PHP_FUNCTION(mb_send_mail);
 PHP_FUNCTION(mb_get_info);
 PHP_FUNCTION(mb_check_encoding);
+PHP_FUNCTION(mb_scrub);
 
 MBSTRING_API char *php_mb_safe_strrchr_ex(const char *s, unsigned int c,
                                     size_t nbytes, const mbfl_encoding *enc);

ext/mbstring/tests/mb_scrub.phpt

@@ -0,0 +1,14 @@
+--TEST--
+mb_scrub()
+--SKIPIF--
+<?php extension_loaded('mbstring') or die('skip mbstring not available'); ?>
+--FILE--
+<?php
+var_dump(
+    "?" === mb_scrub("\x80"),
+    "?" === mb_scrub("\x80", 'UTF-8')
+);
+?>
+--EXPECT--
+bool(true)
+bool(true)