X-Git-Url: https://git.mxchange.org/?a=blobdiff_plain;f=src%2FUtil%2FStrings.php;h=2405fbababcfdbcd8d6788f4ed28040c6b954c4f;hb=1d6f5c33a1a7c538d4e529d22701fd735855da15;hp=40109e67eef69c56c9a4c1269c395cd5983e0ac4;hpb=ffc406d8195871a6580c78f1cc42ff0b7deeba02;p=friendica.git diff --git a/src/Util/Strings.php b/src/Util/Strings.php index 40109e67ee..2405fbabab 100644 --- a/src/Util/Strings.php +++ b/src/Util/Strings.php @@ -1,405 +1,564 @@ . + * */ namespace Friendica\Util; +use Friendica\Content\ContactSelector; +use Friendica\Core\Logger; +use Friendica\Core\System; +use ParagonIE\ConstantTime\Base64; + /** - * @brief This class handles string functions + * This class handles string functions */ class Strings { - /** - * @brief Generates a pseudo-random string of hexadecimal characters - * - * @param int $size - * @return string - */ - public static function getRandomHex($size = 64) - { - $byte_size = ceil($size / 2); - - $bytes = random_bytes($byte_size); - - $return = substr(bin2hex($bytes), 0, $size); - - return $return; - } - - /** - * This is our primary input filter. - * - * The high bit hack only involved some old IE browser, forget which (IE5/Mac?) - * that had an XSS attack vector due to stripping the high-bit on an 8-bit character - * after cleansing, and angle chars with the high bit set could get through as markup. - * - * This is now disabled because it was interfering with some legitimate unicode sequences - * and hopefully there aren't a lot of those browsers left. - * - * Use this on any text input where angle chars are not valid or permitted - * They will be replaced with safer brackets. This may be filtered further - * if these are not allowed either. - * - * @param string $string Input string - * @return string Filtered string - */ - public static function removeTags($string) - { - return str_replace(["<", ">"], ['[', ']'], $string); - } - - /** - * @brief Use this on "body" or "content" input where angle chars shouldn't be removed, - * and allow them to be safely displayed. - * @param string $string - * - * @return string - */ - public static function escapeTags($string) - { - return htmlspecialchars($string, ENT_COMPAT, 'UTF-8', false); - } - - /** - * @brief Generate a string that's random, but usually pronounceable. Used to generate initial passwords - * - * @param int $len length - * - * @return string - */ - public static function getRandomName($len) - { - if ($len <= 0) { - return ''; - } - - $vowels = ['a', 'a', 'ai', 'au', 'e', 'e', 'e', 'ee', 'ea', 'i', 'ie', 'o', 'ou', 'u']; - - if (mt_rand(0, 5) == 4) { - $vowels[] = 'y'; - } - - $cons = [ - 'b', 'bl', 'br', - 'c', 'ch', 'cl', 'cr', - 'd', 'dr', - 'f', 'fl', 'fr', - 'g', 'gh', 'gl', 'gr', - 'h', - 'j', - 'k', 'kh', 'kl', 'kr', - 'l', - 'm', - 'n', - 'p', 'ph', 'pl', 'pr', - 'qu', - 'r', 'rh', - 's' ,'sc', 'sh', 'sm', 'sp', 'st', - 't', 'th', 'tr', - 'v', - 'w', 'wh', - 'x', - 'z', 'zh' - ]; - - $midcons = ['ck', 'ct', 'gn', 'ld', 'lf', 'lm', 'lt', 'mb', 'mm', 'mn', 'mp', - 'nd', 'ng', 'nk', 'nt', 'rn', 'rp', 'rt']; - - $noend = ['bl', 'br', 'cl', 'cr', 'dr', 'fl', 'fr', 'gl', 'gr', - 'kh', 'kl', 'kr', 'mn', 'pl', 'pr', 'rh', 'tr', 'qu', 'wh', 'q']; - - $start = mt_rand(0, 2); - if ($start == 0) { - $table = $vowels; - } else { - $table = $cons; - } - - $word = ''; - - for ($x = 0; $x < $len; $x ++) { - $r = mt_rand(0, count($table) - 1); - $word .= $table[$r]; - - if ($table == $vowels) { - $table = array_merge($cons, $midcons); - } else { - $table = $vowels; - } - - } - - $word = substr($word, 0, $len); - - foreach ($noend as $noe) { - $noelen = strlen($noe); - if ((strlen($word) > $noelen) && (substr($word, -$noelen) == $noe)) { - $word = self::getRandomName($len); - break; - } - } - - return $word; - } - - /** - * @brief translate and format the networkname of a contact - * - * @param string $network Networkname of the contact (e.g. dfrn, rss and so on) - * @param string $url The contact url - * - * @return string Formatted network name - */ - public static function formatNetworkName($network, $url = 0) - { - if ($network != "") { - if ($url != "") { - $network_name = ''.ContactSelector::networkToName($network, $url).""; - } else { - $network_name = ContactSelector::networkToName($network); - } - - return $network_name; - } - } - - /** - * @brief Remove intentation from a text - * - * @param string $text String to be transformed. - * @param string $chr Optional. Indentation tag. Default tab (\t). - * @param int $count Optional. Default null. - * - * @return string Transformed string. - */ - public static function deindent($text, $chr = "[\t ]", $count = NULL) - { - $lines = explode("\n", $text); - - if (is_null($count)) { - $m = []; - $k = 0; - while ($k < count($lines) && strlen($lines[$k]) == 0) { - $k++; - } - preg_match("|^" . $chr . "*|", $lines[$k], $m); - $count = strlen($m[0]); - } - - for ($k = 0; $k < count($lines); $k++) { - $lines[$k] = preg_replace("|^" . $chr . "{" . $count . "}|", "", $lines[$k]); - } - - return implode("\n", $lines); - } - - /** - * @brief Get byte size returned in a Data Measurement (KB, MB, GB) - * - * @param int $bytes The number of bytes to be measured - * @param int $precision Optional. Default 2. - * - * @return string Size with measured units. - */ - public static function formatBytes($bytes, $precision = 2) - { - $units = ['B', 'KB', 'MB', 'GB', 'TB']; - $bytes = max($bytes, 0); - $pow = floor(($bytes ? log($bytes) : 0) / log(1024)); - $pow = min($pow, count($units) - 1); - $bytes /= pow(1024, $pow); - - return round($bytes, $precision) . ' ' . $units[$pow]; - } - - /** - * @brief Protect percent characters in sprintf calls - * - * @param string $s String to transform. - * - * @return string Transformed string. - */ - public static function protectSprintf($s) - { - return str_replace('%', '%%', $s); - } - - /** - * @brief Base64 Encode URL and translate +/ to -_ Optionally strip padding. - * - * @param string $s URL to encode - * @param boolean $strip_padding Optional. Default false - * - * @return string Encoded URL - */ - public static function base64UrlEncode($s, $strip_padding = false) - { - $s = strtr(base64_encode($s), '+/', '-_'); - - if ($strip_padding) { - $s = str_replace('=', '', $s); - } - - return $s; - } - - /** - * @brief Decode Base64 Encoded URL and translate -_ to +/ - * @param string $s URL to decode - * - * @return string Decoded URL - */ - public static function base64UrlDecode($s) - { - if (is_array($s)) { - Logger::log('base64url_decode: illegal input: ' . print_r(debug_backtrace(), true)); - return $s; - } - - /* - * // Placeholder for new rev of salmon which strips base64 padding. - * // PHP base64_decode handles the un-padded input without requiring this step - * // Uncomment if you find you need it. - * - * $l = strlen($s); - * if (!strpos($s,'=')) { - * $m = $l % 4; - * if ($m == 2) - * $s .= '=='; - * if ($m == 3) - * $s .= '='; - * } - * - */ - - return base64_decode(strtr($s, '-_', '+/')); - } - - /** - * @brief Pull out all #hashtags and @person tags from $string. - * - * We also get @person@domain.com - which would make - * the regex quite complicated as tags can also - * end a sentence. So we'll run through our results - * and strip the period from any tags which end with one. - * Returns array of tags found, or empty array. - * - * @param string $string Post content - * - * @return array List of tag and person names - */ - public static function getTags($string) - { - $ret = []; - - // Convert hashtag links to hashtags - $string = preg_replace('/#\[url\=([^\[\]]*)\](.*?)\[\/url\]/ism', '#$2', $string); - - // ignore anything in a code block - $string = preg_replace('/\[code\](.*?)\[\/code\]/sm', '', $string); - - // Force line feeds at bbtags - $string = str_replace(['[', ']'], ["\n[", "]\n"], $string); - - // ignore anything in a bbtag - $string = preg_replace('/\[(.*?)\]/sm', '', $string); - - // Match full names against @tags including the space between first and last - // We will look these up afterward to see if they are full names or not recognisable. - - if (preg_match_all('/(@[^ \x0D\x0A,:?]+ [^ \x0D\x0A@,:?]+)([ \x0D\x0A@,:?]|$)/', $string, $matches)) { - foreach ($matches[1] as $match) { - if (strstr($match, ']')) { - // we might be inside a bbcode color tag - leave it alone - continue; - } - - if (substr($match, -1, 1) === '.') { - $ret[] = substr($match, 0, -1); - } else { - $ret[] = $match; - } - } - } - - // Otherwise pull out single word tags. These can be @nickname, @first_last - // and #hash tags. - - if (preg_match_all('/([!#@][^\^ \x0D\x0A,;:?]+)([ \x0D\x0A,;:?]|$)/', $string, $matches)) { - foreach ($matches[1] as $match) { - if (strstr($match, ']')) { - // we might be inside a bbcode color tag - leave it alone - continue; - } - if (substr($match, -1, 1) === '.') { - $match = substr($match,0,-1); - } - // ignore strictly numeric tags like #1 - if ((strpos($match, '#') === 0) && ctype_digit(substr($match, 1))) { - continue; - } - // try not to catch url fragments - if (strpos($string, $match) && preg_match('/[a-zA-z0-9\/]/', substr($string, strpos($string, $match) - 1, 1))) { - continue; - } - $ret[] = $match; - } - } - - return $ret; - } - - /** - * @brief Check for a valid email string - * - * @param string $email_address Email address to be evaluated. - * - * @return boolean Value indicating whether or not the string is a valid email address. - */ - public static function isValidEmail($email_address) - { - return preg_match('/^[_a-zA-Z0-9\-\+]+(\.[_a-zA-Z0-9\-\+]+)*@[a-zA-Z0-9-]+(\.[a-zA-Z0-9-]+)+$/', $email_address); - } - - /** - * @brief Normalize url - * - * @param string $url URL to be normalized. - * - * @return string Normalized URL. - */ - public static function normaliseLink($url) - { - $ret = str_replace(['https:', '//www.'], ['http:', '//'], $url); - return rtrim($ret, '/'); - } - - /** - * @brief Normalize OpenID identity - * - * @param string $s OpenID Identity - * - * @return string normalized OpenId Identity - */ - function normaliseOpenID($s) - { - return trim(str_replace(['http://', 'https://'], ['', ''], $s), '/'); - } - - /** - * @brief Compare two URLs to see if they are the same, but ignore - * slight but hopefully insignificant differences such as if one - * is https and the other isn't, or if one is www.something and - * the other isn't - and also ignore case differences. - * - * @param string $a first url - * @param string $b second url - * @return boolean True if the URLs match, otherwise False - * - */ - public static function compareLink($a, $b) - { - return (strcasecmp(self::normaliseLink($a), self::normaliseLink($b)) === 0); - } + /** + * Generates a pseudo-random string of hexadecimal characters + * + * @param int $size Size of string (default: 64) + * + * @return string Pseudo-random string + * @throws \Exception + */ + public static function getRandomHex(int $size = 64): string + { + $byte_size = ceil($size / 2); + + $bytes = random_bytes($byte_size); + + $return = substr(bin2hex($bytes), 0, $size); + + return $return; + } + + /** + * Checks, if the given string is a valid hexadecimal code + * + * @param string $hexCode + * @return bool + */ + public static function isHex(string $hexCode): bool + { + return !empty($hexCode) ? @preg_match("/^[a-f0-9]{2,}$/i", $hexCode) && !(strlen($hexCode) & 1) : false; + } + + /** + * Use this on "body" or "content" input where angle chars shouldn't be removed, + * and allow them to be safely displayed. + * @param string $string + * + * @return string + */ + public static function escapeHtml($string) + { + return htmlspecialchars($string, ENT_COMPAT, 'UTF-8', false); + } + + /** + * Generate a string that's random, but usually pronounceable. Used to generate initial passwords + * + * @param int $len length + * @return string + */ + public static function getRandomName(int $len): string + { + if ($len <= 0) { + return ''; + } + + $vowels = ['a', 'a', 'ai', 'au', 'e', 'e', 'e', 'ee', 'ea', 'i', 'ie', 'o', 'ou', 'u']; + + if (mt_rand(0, 5) == 4) { + $vowels[] = 'y'; + } + + $cons = [ + 'b', 'bl', 'br', + 'c', 'ch', 'cl', 'cr', + 'd', 'dr', + 'f', 'fl', 'fr', + 'g', 'gh', 'gl', 'gr', + 'h', + 'j', + 'k', 'kh', 'kl', 'kr', + 'l', + 'm', + 'n', + 'p', 'ph', 'pl', 'pr', + 'qu', + 'r', 'rh', + 's', 'sc', 'sh', 'sm', 'sp', 'st', + 't', 'th', 'tr', + 'v', + 'w', 'wh', + 'x', + 'z', 'zh' + ]; + + $midcons = [ + 'ck', 'ct', 'gn', 'ld', 'lf', 'lm', 'lt', 'mb', 'mm', 'mn', 'mp', + 'nd', 'ng', 'nk', 'nt', 'rn', 'rp', 'rt' + ]; + + $noend = [ + 'bl', 'br', 'cl', 'cr', 'dr', 'fl', 'fr', 'gl', 'gr', + 'kh', 'kl', 'kr', 'mn', 'pl', 'pr', 'rh', 'tr', 'qu', 'wh', 'q' + ]; + + $start = mt_rand(0, 2); + if ($start == 0) { + $table = $vowels; + } else { + $table = $cons; + } + + $word = ''; + + for ($x = 0; $x < $len; $x++) { + $r = mt_rand(0, count($table) - 1); + $word .= $table[$r]; + + if ($table == $vowels) { + $table = array_merge($cons, $midcons); + } else { + $table = $vowels; + } + } + + $word = substr($word, 0, $len); + + foreach ($noend as $noe) { + $noelen = strlen($noe); + if ((strlen($word) > $noelen) && (substr($word, -$noelen) == $noe)) { + $word = self::getRandomName($len); + break; + } + } + + return $word; + } + + /** + * Translate and format the network name of a contact + * + * @param string $network Network name of the contact (e.g. dfrn, rss and so on) + * @param string $url The contact url + * + * @return string Formatted network name + * @throws \Friendica\Network\HTTPException\InternalServerErrorException + */ + public static function formatNetworkName(string $network, string $url = ''): string + { + if ($network != '') { + if ($url != '') { + $network_name = '' . ContactSelector::networkToName($network, $url) . ''; + } else { + $network_name = ContactSelector::networkToName($network); + } + + return $network_name; + } + + return ''; + } + + /** + * Remove indentation from a text + * + * @param string $text String to be transformed. + * @param string $chr Optional. Indentation tag. Default tab (\t). + * @param int $count Optional. Default null. + * + * @return string Transformed string. + */ + public static function deindent(string $text, string $chr = "[\t ]", int $count = null): string + { + $lines = explode("\n", $text); + + if (is_null($count)) { + $m = []; + $k = 0; + while ($k < count($lines) && strlen($lines[$k]) == 0) { + $k++; + } + preg_match("|^" . $chr . "*|", $lines[$k], $m); + $count = strlen($m[0]); + } + + for ($k = 0; $k < count($lines); $k++) { + $lines[$k] = preg_replace("|^" . $chr . "{" . $count . "}|", "", $lines[$k]); + } + + return implode("\n", $lines); + } + + /** + * Get byte size returned in a Data Measurement (KB, MB, GB) + * + * @param int $bytes The number of bytes to be measured + * @param int $precision Optional. Default 2. + * + * @return string Size with measured units. + */ + public static function formatBytes(int $bytes, int $precision = 2): string + { + // If this method is called for an infinite (== unlimited) amount of bytes: + if ($bytes == INF) { + return INF; + } + + $units = ['B', 'KiB', 'MiB', 'GiB', 'TiB']; + $bytes = max($bytes, 0); + $pow = floor(($bytes ? log($bytes) : 0) / log(1024)); + $pow = min($pow, count($units) - 1); + $bytes /= pow(1024, $pow); + + return round($bytes, $precision) . ' ' . $units[$pow]; + } + + /** + * Protect percent characters in sprintf calls + * + * @param string $s String to transform. + * @return string Transformed string. + */ + public static function protectSprintf(string $s): string + { + return str_replace('%', '%%', $s); + } + + /** + * Base64 Encode URL and translate +/ to -_ Optionally strip padding. + * + * @param string $s URL to encode + * @param boolean $strip_padding Optional. Default false + * @return string Encoded URL + * @see https://web.archive.org/web/20160506073138/http://salmon-protocol.googlecode.com:80/svn/trunk/draft-panzer-magicsig-01.html#params + */ + public static function base64UrlEncode(string $s, bool $strip_padding = false): string + { + if ($strip_padding) { + $s = Base64::encodeUnpadded($s); + } else { + $s = Base64::encode($s); + } + + return strtr($s, '+/', '-_'); + } + + /** + * Decode Base64 Encoded URL and translate -_ to +/ + * + * @param string $s URL to decode + * @return string Decoded URL + * @throws \Exception + * @see https://web.archive.org/web/20160506073138/http://salmon-protocol.googlecode.com:80/svn/trunk/draft-panzer-magicsig-01.html#params + */ + public static function base64UrlDecode(string $s): string + { + return Base64::decode(strtr($s, '-_', '+/')); + } + + /** + * Normalize url + * + * @param string $url URL to be normalized. + * @return string Normalized URL. + */ + public static function normaliseLink(string $url): string + { + $ret = str_replace(['https:', '//www.'], ['http:', '//'], $url); + return rtrim($ret, '/'); + } + + /** + * Normalize OpenID identity + * + * @param string $s OpenID Identity + * @return string normalized OpenId Identity + */ + public static function normaliseOpenID(string $s): string + { + return trim(str_replace(['http://', 'https://'], ['', ''], $s), '/'); + } + + /** + * Compare two URLs to see if they are the same, but ignore + * slight but hopefully insignificant differences such as if one + * is https and the other isn't, or if one is www.something and + * the other isn't - and also ignore case differences. + * + * @param string $a first url + * @param string $b second url + * @return boolean True if the URLs match, otherwise False + * + */ + public static function compareLink(string $a, string $b): bool + { + return (strcasecmp(self::normaliseLink($a), self::normaliseLink($b)) === 0); + } + + /** + * Ensures the provided URI has its query string punctuation in order. + * + * @param string $uri + * @return string + */ + public static function ensureQueryParameter(string $uri): string + { + if (strpos($uri, '?') === false && ($pos = strpos($uri, '&')) !== false) { + $uri = substr($uri, 0, $pos) . '?' . substr($uri, $pos + 1); + } + + return $uri; + } + + /** + * Check if the trimmed provided string is starting with one of the provided characters + * + * @param string $string + * @param array $chars + * + * @return bool + */ + public static function startsWithChars(string $string, array $chars): bool + { + $return = in_array(substr(trim($string), 0, 1), $chars); + + return $return; + } + + /** + * Check if the first string starts with the second + * + * @see http://maettig.com/code/php/php-performance-benchmarks.php#startswith + * @param string $string + * @param string $start + * @return bool + */ + public static function startsWith(string $string, string $start): bool + { + $return = substr_compare($string, $start, 0, strlen($start)) === 0; + + return $return; + } + + /** + * Checks if the first string ends with the second + * + * @see http://maettig.com/code/php/php-performance-benchmarks.php#endswith + * @param string $string + * @param string $end + * + * @return bool + */ + public static function endsWith(string $string, string $end): bool + { + return (substr_compare($string, $end, -strlen($end)) === 0); + } + + /** + * Returns the regular expression string to match URLs in a given text + * + * @return string + */ + public static function autoLinkRegEx(): string + { + return '@ +(??«»“”‘’.] # Domain can\'t start with a . + [^/\s\xA0`!()\[\]{};:\'",<>?«»“”‘’]+ # Domain can\'t end with a . + \. + [^/\s\xA0`!()\[\]{};:\'".,<>?«»“”‘’]+/? # Followed by a slash + ) + (?: # One or more: + [^\s\xA0()<>]+ # Run of non-space, non-()<> + | # or + \(([^\s\xA0()<>]+|(\([^\s()<>]+\)))*\) # balanced parens, up to 2 levels + | # or + [^\s\xA0`!()\[\]{};:\'".,<>?«»“”‘’] # not a space or one of these punct chars + )*'; + } + + /** + * Ensures a single path item doesn't contain any path-traversing characters + * + * @param string $pathItem + * + * @see https://stackoverflow.com/a/46097713 + * @return string + */ + public static function sanitizeFilePathItem(string $pathItem): string + { + $pathItem = str_replace('/', '_', $pathItem); + $pathItem = str_replace('\\', '_', $pathItem); + $pathItem = str_replace(DIRECTORY_SEPARATOR, '_', $pathItem); // In case it does not equal the standard values + + return $pathItem; + } + + /** + * Multi-byte safe implementation of substr_replace where $start and $length are character offset and count rather + * than byte offset and counts. + * + * Depends on mbstring, use default encoding. + * + * @param string $string + * @param string $replacement + * @param int $start + * @param int|null $length + * + * @return string + * @see substr_replace() + */ + public static function substringReplace(string $string, string $replacement, int $start, int $length = null): string + { + $string_length = mb_strlen($string); + + $length = $length ?? $string_length; + + if ($start < 0) { + $start = max(0, $string_length + $start); + } else if ($start > $string_length) { + $start = $string_length; + } + + if ($length < 0) { + $length = max(0, $string_length - $start + $length); + } else if ($length > $string_length) { + $length = $string_length; + } + + if (($start + $length) > $string_length) { + $length = $string_length - $start; + } + + return mb_substr($string, 0, $start) . $replacement . mb_substr($string, $start + $length, $string_length - $start - $length); + } + + /** + * Perform a custom function on a text after having escaped blocks matched by the provided regular expressions. + * Only full matches are used, capturing group are ignored. + * + * To change the provided text, the callback function needs to return it and this function will return the modified + * version as well after having restored the escaped blocks. + * + * @param string $text + * @param string $regex + * @param callable $callback + * + * @return string + */ + public static function performWithEscapedBlocks(string $text, string $regex, callable $callback): string + { + // Enables nested use + $executionId = random_int(PHP_INT_MAX / 10, PHP_INT_MAX); + + $blocks = []; + + $return = preg_replace_callback($regex, + function ($matches) use ($executionId, &$blocks) { + $return = '«block-' . $executionId . '-' . count($blocks) . '»'; + + $blocks[] = $matches[0]; + + return $return; + }, + $text + ); + + if (is_null($return)) { + Logger::notice('Received null value from preg_replace_callback', ['text' => $text, 'regex' => $regex, 'blocks' => $blocks, 'executionId' => $executionId, 'callstack' => System::callstack(10)]); + } + + $text = $callback($return ?? $text) ?? ''; + + // Restore code blocks + $text = preg_replace_callback('/«block-' . $executionId . '-([0-9]+)»/iU', + function ($matches) use ($blocks) { + $return = $matches[0]; + if (isset($blocks[intval($matches[1])])) { + $return = $blocks[$matches[1]]; + } + return $return; + }, + $text + ); + + return $text; + } + + /** + * This function converts a PHP's shorhand notation string for file sizes in to an integer number of total bytes. + * For example: The string for shorthand notation of '2M' (which is 2,097,152 Bytes) is converted to 2097152 + * @see https://www.php.net/manual/en/faq.using.php#faq.using.shorthandbytes + * @param string $shorthand + * @return int + */ + public static function getBytesFromShorthand(string $shorthand): int + { + $shorthand = trim($shorthand); + + if (is_numeric($shorthand)) { + return $shorthand; + } + + $last = strtolower($shorthand[strlen($shorthand)-1]); + $shorthand = substr($shorthand, 0, -1); + + switch($last) { + case 'g': + $shorthand *= 1024; + case 'm': + $shorthand *= 1024; + case 'k': + $shorthand *= 1024; + } + + return $shorthand; + } + }