[ticket/13614] Removed code that relies on phpbb_pcre_utf8_support()

Removed all calls to phpbb_pcre_utf8_support() as well as unreachable code.

PHPBB3-13614

phpBB/includes/acp/acp_bbcodes.php

@@ -416,8 +416,6 @@ class acp_bbcodes
 		// Allow unicode characters for URL|LOCAL_URL|RELATIVE_URL|INTTEXT tokens
 		$utf8 = preg_match('/(URL|LOCAL_URL|RELATIVE_URL|INTTEXT)/', $bbcode_match);
 
-		$utf8_pcre_properties = phpbb_pcre_utf8_support();
-
 		$fp_match = preg_quote($bbcode_match, '!');
 		$fp_replace = preg_replace('#^\[(.*?)\]#', '[$1:$uid]', $bbcode_match);
 		$fp_replace = preg_replace('#\[/(.*?)\]$#', '[/$1:$uid]', $fp_replace);
@@ -448,7 +446,7 @@ class acp_bbcodes
 				'!([a-zA-Z0-9-+.,_ ]+)!'	 =>	"$1"
 			),
 			'INTTEXT' => array(
-				($utf8_pcre_properties) ? '!([\p{L}\p{N}\-+,_. ]+)!u' : '!([a-zA-Z0-9\-+,_. ]+)!u'	 =>	"$1"
+				'!([\p{L}\p{N}\-+,_. ]+)!u'	 =>	"$1"
 			),
 			'IDENTIFIER' => array(
 				'!([a-zA-Z0-9-_]+)!'	 =>	"$1"
@@ -468,7 +466,7 @@ class acp_bbcodes
 			'EMAIL' => '(' . get_preg_expression('email') . ')',
 			'TEXT' => '(.*?)',
 			'SIMPLETEXT' => '([a-zA-Z0-9-+.,_ ]+)',
-			'INTTEXT' => ($utf8_pcre_properties) ? '([\p{L}\p{N}\-+,_. ]+)' : '([a-zA-Z0-9\-+,_. ]+)',
+			'INTTEXT' => '([\p{L}\p{N}\-+,_. ]+)',
 			'IDENTIFIER' => '([a-zA-Z0-9-_]+)',
 			'COLOR' => '([a-zA-Z]+|#[0-9abcdefABCDEF]+)',
 			'NUMBER' => '([0-9]+)',
@@ -476,7 +474,7 @@ class acp_bbcodes
 
 		$pad = 0;
 		$modifiers = 'i';
-		$modifiers .= ($utf8 && $utf8_pcre_properties) ? 'u' : '';
+		$modifiers .= ($utf8) ? 'u' : '';
 
 		if (preg_match_all('/\{(' . implode('|', array_keys($tokens)) . ')[0-9]*\}/i', $bbcode_match, $m))
 		{

phpBB/includes/functions.php

@@ -2811,31 +2811,19 @@ function get_preg_expression($mode)
 * Depends on whether installed PHP version supports unicode properties
 *
 * @param string	$word			word template to be replaced
-* @param bool	$use_unicode	whether or not to take advantage of PCRE supporting unicode
 *
 * @return string $preg_expr		regex to use with word censor
 */
-function get_censor_preg_expression($word, $use_unicode = true)
+function get_censor_preg_expression($word)
 {
 	// Unescape the asterisk to simplify further conversions
 	$word = str_replace('\*', '*', preg_quote($word, '#'));
 
-	if ($use_unicode && phpbb_pcre_utf8_support())
+	// Replace asterisk(s) inside the pattern, at the start and at the end of it with regexes
-	{
+	$word = preg_replace(array('#(?<=[\p{Nd}\p{L}_])\*+(?=[\p{Nd}\p{L}_])#iu', '#^\*+#', '#\*+$#'), array('([\x20]*?|[\p{Nd}\p{L}_-]*?)', '[\p{Nd}\p{L}_-]*?', '[\p{Nd}\p{L}_-]*?'), $word);
-		// Replace asterisk(s) inside the pattern, at the start and at the end of it with regexes
-		$word = preg_replace(array('#(?<=[\p{Nd}\p{L}_])\*+(?=[\p{Nd}\p{L}_])#iu', '#^\*+#', '#\*+$#'), array('([\x20]*?|[\p{Nd}\p{L}_-]*?)', '[\p{Nd}\p{L}_-]*?', '[\p{Nd}\p{L}_-]*?'), $word);
 
-		// Generate the final substitution
+	// Generate the final substitution
-		$preg_expr = '#(?<![\p{Nd}\p{L}_-])(' . $word . ')(?![\p{Nd}\p{L}_-])#iu';
+	$preg_expr = '#(?<![\p{Nd}\p{L}_-])(' . $word . ')(?![\p{Nd}\p{L}_-])#iu';
-	}
-	else
-	{
-		// Replace the asterisk inside the pattern, at the start and at the end of it with regexes
-		$word = preg_replace(array('#(?<=\S)\*+(?=\S)#iu', '#^\*+#', '#\*+$#'), array('(\x20*?\S*?)', '\S*?', '\S*?'), $word);
-
-		// Generate the final substitution
-		$preg_expr = '#(?<!\S)(' . $word . ')(?!\S)#iu';
-	}
 
 	return $preg_expr;
 }
@@ -4690,22 +4678,6 @@ function phpbb_user_session_handler()
 	return;
 }
 
-/**
-* Check if PCRE has UTF-8 support
-* PHP may not be linked with the bundled PCRE lib and instead with an older version
-*
-* @return bool	Returns true if PCRE (the regular expressions library) supports UTF-8 encoding
-*/
-function phpbb_pcre_utf8_support()
-{
-	static $utf8_pcre_properties = null;
-	if (is_null($utf8_pcre_properties))
-	{
-		$utf8_pcre_properties = (@preg_match('/\p{L}/u', 'a') !== false);
-	}
-	return $utf8_pcre_properties;
-}
-
 /**
 * Casts a numeric string $input to an appropriate numeric type (i.e. integer or float)
 *

phpBB/includes/functions_user.php

@@ -1647,89 +1647,37 @@ function validate_username($username, $allowed_username = false)
 		return 'INVALID_CHARS';
 	}
 
-	$mbstring = $pcre = false;
-
-	// generic UTF-8 character types supported?
-	if (phpbb_pcre_utf8_support())
-	{
-		$pcre = true;
-	}
-	else if (function_exists('mb_ereg_match'))
-	{
-		mb_regex_encoding('UTF-8');
-		$mbstring = true;
-	}
-
 	switch ($config['allow_name_chars'])
 	{
 		case 'USERNAME_CHARS_ANY':
-			$pcre = true;
 			$regex = '.+';
 		break;
 
 		case 'USERNAME_ALPHA_ONLY':
-			$pcre = true;
 			$regex = '[A-Za-z0-9]+';
 		break;
 
 		case 'USERNAME_ALPHA_SPACERS':
-			$pcre = true;
 			$regex = '[A-Za-z0-9-[\]_+ ]+';
 		break;
 
 		case 'USERNAME_LETTER_NUM':
-			if ($pcre)
+			$regex = '[\p{Lu}\p{Ll}\p{N}]+';
-			{
-				$regex = '[\p{Lu}\p{Ll}\p{N}]+';
-			}
-			else if ($mbstring)
-			{
-				$regex = '[[:upper:][:lower:][:digit:]]+';
-			}
-			else
-			{
-				$pcre = true;
-				$regex = '[a-zA-Z0-9]+';
-			}
 		break;
 
 		case 'USERNAME_LETTER_NUM_SPACERS':
-			if ($pcre)
+			$regex = '[-\]_+ [\p{Lu}\p{Ll}\p{N}]+';
-			{
-				$regex = '[-\]_+ [\p{Lu}\p{Ll}\p{N}]+';
-			}
-			else if ($mbstring)
-			{
-				$regex = '[-\]_+ \[[:upper:][:lower:][:digit:]]+';
-			}
-			else
-			{
-				$pcre = true;
-				$regex = '[-\]_+ [a-zA-Z0-9]+';
-			}
 		break;
 
 		case 'USERNAME_ASCII':
 		default:
-			$pcre = true;
 			$regex = '[\x01-\x7F]+';
 		break;
 	}
 
-	if ($pcre)
+	if (!preg_match('#^' . $regex . '$#u', $username))
 	{
-		if (!preg_match('#^' . $regex . '$#u', $username))
+		return 'INVALID_CHARS';
-		{
-			return 'INVALID_CHARS';
-		}
-	}
-	else if ($mbstring)
-	{
-		mb_ereg_search_init($username, '^' . $regex . '$');
-		if (!mb_ereg_search())
-		{
-			return 'INVALID_CHARS';
-		}
 	}
 
 	$sql = 'SELECT username
@@ -1784,35 +1732,10 @@ function validate_password($password)
 		return false;
 	}
 
-	$pcre = $mbstring = false;
+	$upp = '\p{Lu}';
-
+	$low = '\p{Ll}';
-	// generic UTF-8 character types supported?
+	$num = '\p{N}';
-	if (phpbb_pcre_utf8_support())
+	$sym = '[^\p{Lu}\p{Ll}\p{N}]';
-	{
-		$upp = '\p{Lu}';
-		$low = '\p{Ll}';
-		$num = '\p{N}';
-		$sym = '[^\p{Lu}\p{Ll}\p{N}]';
-		$pcre = true;
-	}
-	else if (function_exists('mb_ereg_match'))
-	{
-		mb_regex_encoding('UTF-8');
-		$upp = '[[:upper:]]';
-		$low = '[[:lower:]]';
-		$num = '[[:digit:]]';
-		$sym = '[^[:upper:][:lower:][:digit:]]';
-		$mbstring = true;
-	}
-	else
-	{
-		$upp = '[A-Z]';
-		$low = '[a-z]';
-		$num = '[0-9]';
-		$sym = '[^A-Za-z0-9]';
-		$pcre = true;
-	}
-
 	$chars = array();
 
 	switch ($config['pass_complex'])
@@ -1835,24 +1758,11 @@ function validate_password($password)
 			$chars[] = $upp;
 	}
 
-	if ($pcre)
+	foreach ($chars as $char)
 	{
-		foreach ($chars as $char)
+		if (!preg_match('#' . $char . '#u', $password))
 		{
-			if (!preg_match('#' . $char . '#u', $password))
+			return 'INVALID_CHARS';
-			{
-				return 'INVALID_CHARS';
-			}
-		}
-	}
-	else if ($mbstring)
-	{
-		foreach ($chars as $char)
-		{
-			if (mb_ereg($char, $password) === false)
-			{
-				return 'INVALID_CHARS';
-			}
 		}
 	}
 

tests/regex/censor_test.php

@@ -37,17 +37,7 @@ class phpbb_regex_censor_test extends phpbb_test_case
 	*/
 	public function test_censor_unicode($pattern, $subject)
 	{
-		$regex = get_censor_preg_expression($pattern, true);
+		$regex = get_censor_preg_expression($pattern);
-
-		$this->assertRegExp($regex, $subject);
-	}
-
-	/**
-	* @dataProvider censor_test_data
-	*/
-	public function test_censor_no_unicode($pattern, $subject)
-	{
-		$regex = get_censor_preg_expression($pattern, false);
 
 		$this->assertRegExp($regex, $subject);
 	}