4 * Function library for anything related to formatting / validating / sanitizing
8 * Convert an integer (1337) to a string (3jk).
11 function yourls_int2string( $int, $chars = null ) {
13 $chars = yourls_get_shorturl_charset();
15 $base = strlen( $chars );
19 $i = fmod( $int, $base );
20 $str = $chars[ intval( $i ) ] . $str;
21 $int = ( $int - $i ) / $base;
24 return yourls_apply_filter( 'int2string', $str, $int, $chars );
28 * Convert a string (3jk) to an integer (1337)
31 function yourls_string2int( $string, $chars = null ) {
33 $chars = yourls_get_shorturl_charset();
35 $base = strlen( $chars );
37 $len = strlen( $string );
39 $arr = array_flip( str_split( $chars ) );
40 for( $i = 0; $i < $len; ++$i ) {
41 $integer += $arr[ $string[ $i ] ] * pow( $base, $len - $i -1 );
43 return yourls_apply_filter( 'string2int', $integer, $string, $chars );
47 * Return a unique(ish) hash for a string to be used as a valid HTML id
50 function yourls_string2htmlid( $string ) {
51 return yourls_apply_filter( 'string2htmlid', 'y'.abs( crc32( $string ) ) );
55 * Make sure a link keyword (ie "1fv" as in "site.com/1fv") is valid.
58 function yourls_sanitize_string( $string ) {
59 // make a regexp pattern with the shorturl charset, and remove everything but this
60 $pattern = yourls_make_regexp_pattern( yourls_get_shorturl_charset() );
61 $valid = substr( preg_replace( '![^'.$pattern.']!', '', $string ), 0, 199 );
63 return yourls_apply_filter( 'sanitize_string', $valid, $string );
67 * Alias function. I was always getting it wrong.
70 function yourls_sanitize_keyword( $keyword ) {
71 return yourls_sanitize_string( $keyword );
75 * Sanitize a page title. No HTML per W3C http://www.w3.org/TR/html401/struct/global.html#h-7.4.2
78 function yourls_sanitize_title( $unsafe_title ) {
79 $title = $unsafe_title;
80 $title = strip_tags( $title );
81 $title = preg_replace( "/\s+/", ' ', trim( $title ) );
82 return yourls_apply_filter( 'sanitize_title', $title, $unsafe_title );
86 * A few sanity checks on the URL. Used for redirection or DB. For display purpose, see yourls_esc_url()
88 * @param string $unsafe_url unsafe URL
89 * @param array $protocols Optional allowed protocols, default to global $yourls_allowedprotocols
90 * @return string Safe URL
92 function yourls_sanitize_url( $unsafe_url, $protocols = array() ) {
93 $url = yourls_esc_url( $unsafe_url, 'redirection', $protocols );
94 return yourls_apply_filter( 'sanitize_url', $url, $unsafe_url );
98 * Perform a replacement while a string is found, eg $subject = '%0%0%0DDD', $search ='%0D' -> $result =''
100 * Stolen from WP's _deep_replace
103 function yourls_deep_replace( $search, $subject ){
107 foreach( (array) $search as $val ) {
108 while( strpos( $subject, $val ) !== false ) {
110 $subject = str_replace( $val, '', $subject );
119 * Make sure an integer is a valid integer (PHP's intval() limits to too small numbers)
122 function yourls_sanitize_int( $in ) {
123 return ( substr( preg_replace( '/[^0-9]/', '', strval( $in ) ), 0, 20 ) );
127 * Escape a string or an array of strings before DB usage. ALWAYS escape before using in a SQL query. Thanks.
129 * @param string|array $data string or array of strings to be escaped
130 * @return string|array escaped data
132 function yourls_escape( $data ) {
133 if( is_array( $data ) ) {
134 foreach( $data as $k => $v ) {
135 if( is_array( $v ) ) {
136 $data[ $k ] = yourls_escape( $v );
138 $data[ $k ] = yourls_escape_real( $v );
142 $data = yourls_escape_real( $data );
149 * "Real" escape. This function should NOT be called directly. Use yourls_escape() instead.
151 * This function uses a "real" escape if possible, using PDO, MySQL or MySQLi functions,
152 * with a fallback to a "simple" addslashes
153 * If you're implementing a custom DB engine or a custom cache system, you can define an
154 * escape function using filter 'custom_escape_real'
157 * @param string $a string to be escaped
158 * @return string escaped string
160 function yourls_escape_real( $string ) {
162 if( isset( $ydb ) && ( $ydb instanceof ezSQLcore ) )
163 return $ydb->escape( $string );
165 // YOURLS DB classes have been bypassed by a custom DB engine or a custom cache layer
166 return yourls_apply_filter( 'custom_escape_real', addslashes( $string ), $string );
170 * Sanitize an IP address
173 function yourls_sanitize_ip( $ip ) {
174 return preg_replace( '/[^0-9a-fA-F:., ]/', '', $ip );
178 * Make sure a date is m(m)/d(d)/yyyy, return false otherwise
181 function yourls_sanitize_date( $date ) {
182 if( !preg_match( '!^\d{1,2}/\d{1,2}/\d{4}$!' , $date ) ) {
189 * Sanitize a date for SQL search. Return false if malformed input.
192 function yourls_sanitize_date_for_sql( $date ) {
193 if( !yourls_sanitize_date( $date ) )
195 return date( 'Y-m-d', strtotime( $date ) );
199 * Return trimmed string
202 function yourls_trim_long_string( $string, $length = 60, $append = '[...]' ) {
203 $newstring = $string;
204 if( function_exists( 'mb_substr' ) ) {
205 if ( mb_strlen( $newstring ) > $length ) {
206 $newstring = mb_substr( $newstring, 0, $length - mb_strlen( $append ), 'UTF-8' ) . $append;
209 if ( strlen( $newstring ) > $length ) {
210 $newstring = substr( $newstring, 0, $length - strlen( $append ) ) . $append;
213 return yourls_apply_filter( 'trim_long_string', $newstring, $string, $length, $append );
217 * Sanitize a version number (1.4.1-whatever -> 1.4.1)
220 function yourls_sanitize_version( $ver ) {
221 return preg_replace( '/[^0-9.]/', '', $ver );
225 * Sanitize a filename (no Win32 stuff)
228 function yourls_sanitize_filename( $file ) {
229 $file = str_replace( '\\', '/', $file ); // sanitize for Win32 installs
230 $file = preg_replace( '|/+|' ,'/', $file ); // remove any duplicate slash
235 * Check if a string seems to be UTF-8. Stolen from WP.
238 function yourls_seems_utf8( $str ) {
239 $length = strlen( $str );
240 for ( $i=0; $i < $length; $i++ ) {
241 $c = ord( $str[ $i ] );
242 if ( $c < 0x80 ) $n = 0; # 0bbbbbbb
243 elseif (($c & 0xE0) == 0xC0) $n=1; # 110bbbbb
244 elseif (($c & 0xF0) == 0xE0) $n=2; # 1110bbbb
245 elseif (($c & 0xF8) == 0xF0) $n=3; # 11110bbb
246 elseif (($c & 0xFC) == 0xF8) $n=4; # 111110bb
247 elseif (($c & 0xFE) == 0xFC) $n=5; # 1111110b
248 else return false; # Does not match any model
249 for ($j=0; $j<$n; $j++) { # n bytes matching 10bbbbbb follow ?
250 if ((++$i == $length) || ((ord($str[$i]) & 0xC0) != 0x80))
258 * Checks for invalid UTF8 in a string. Stolen from WP
262 * @param string $string The text which is to be checked.
263 * @param boolean $strip Optional. Whether to attempt to strip out invalid UTF8. Default is false.
264 * @return string The checked text.
266 function yourls_check_invalid_utf8( $string, $strip = false ) {
267 $string = (string) $string;
269 if ( 0 === strlen( $string ) ) {
273 // Check for support for utf8 in the installed PCRE library once and store the result in a static
275 if ( !isset( $utf8_pcre ) ) {
276 $utf8_pcre = @preg_match( '/^./u', 'a' );
278 // We can't demand utf8 in the PCRE installation, so just return the string in those cases
283 // preg_match fails when it encounters invalid UTF8 in $string
284 if ( 1 === @preg_match( '/^./us', $string ) ) {
288 // Attempt to strip the bad chars if requested (not recommended)
289 if ( $strip && function_exists( 'iconv' ) ) {
290 return iconv( 'utf-8', 'utf-8', $string );
297 * Converts a number of special characters into their HTML entities. Stolen from WP.
299 * Specifically deals with: &, <, >, ", and '.
301 * $quote_style can be set to ENT_COMPAT to encode " to
302 * ", or ENT_QUOTES to do both. Default is ENT_NOQUOTES where no quotes are encoded.
306 * @param string $string The text which is to be encoded.
307 * @param mixed $quote_style Optional. Converts double quotes if set to ENT_COMPAT, both single and double if set to ENT_QUOTES or none if set to ENT_NOQUOTES. Also compatible with old values; converting single quotes if set to 'single', double if set to 'double' or both if otherwise set. Default is ENT_NOQUOTES.
308 * @param boolean $double_encode Optional. Whether to encode existing html entities. Default is false.
309 * @return string The encoded text with HTML entities.
311 function yourls_specialchars( $string, $quote_style = ENT_NOQUOTES, $double_encode = false ) {
312 $string = (string) $string;
314 if ( 0 === strlen( $string ) )
317 // Don't bother if there are no specialchars - saves some processing
318 if ( ! preg_match( '/[&<>"\']/', $string ) )
321 // Account for the previous behaviour of the function when the $quote_style is not an accepted value
322 if ( empty( $quote_style ) )
323 $quote_style = ENT_NOQUOTES;
324 elseif ( ! in_array( $quote_style, array( 0, 2, 3, 'single', 'double' ), true ) )
325 $quote_style = ENT_QUOTES;
329 $_quote_style = $quote_style;
331 if ( $quote_style === 'double' ) {
332 $quote_style = ENT_COMPAT;
333 $_quote_style = ENT_COMPAT;
334 } elseif ( $quote_style === 'single' ) {
335 $quote_style = ENT_NOQUOTES;
338 // Handle double encoding ourselves
339 if ( $double_encode ) {
340 $string = @htmlspecialchars( $string, $quote_style, $charset );
342 // Decode & into &
343 $string = yourls_specialchars_decode( $string, $_quote_style );
345 // Guarantee every &entity; is valid or re-encode the &
346 $string = yourls_kses_normalize_entities( $string );
348 // Now re-encode everything except &entity;
349 $string = preg_split( '/(&#?x?[0-9a-z]+;)/i', $string, -1, PREG_SPLIT_DELIM_CAPTURE );
351 for ( $i = 0; $i < count( $string ); $i += 2 )
352 $string[$i] = @htmlspecialchars( $string[$i], $quote_style, $charset );
354 $string = implode( '', $string );
357 // Backwards compatibility
358 if ( 'single' === $_quote_style )
359 $string = str_replace( "'", ''', $string );
365 * Converts a number of HTML entities into their special characters. Stolen from WP.
367 * Specifically deals with: &, <, >, ", and '.
369 * $quote_style can be set to ENT_COMPAT to decode " entities,
370 * or ENT_QUOTES to do both " and '. Default is ENT_NOQUOTES where no quotes are decoded.
374 * @param string $string The text which is to be decoded.
375 * @param mixed $quote_style Optional. Converts double quotes if set to ENT_COMPAT, both single and double if set to ENT_QUOTES or none if set to ENT_NOQUOTES. Also compatible with old _wp_specialchars() values; converting single quotes if set to 'single', double if set to 'double' or both if otherwise set. Default is ENT_NOQUOTES.
376 * @return string The decoded text without HTML entities.
378 function yourls_specialchars_decode( $string, $quote_style = ENT_NOQUOTES ) {
379 $string = (string) $string;
381 if ( 0 === strlen( $string ) ) {
385 // Don't bother if there are no entities - saves a lot of processing
386 if ( strpos( $string, '&' ) === false ) {
390 // Match the previous behaviour of _wp_specialchars() when the $quote_style is not an accepted value
391 if ( empty( $quote_style ) ) {
392 $quote_style = ENT_NOQUOTES;
393 } elseif ( !in_array( $quote_style, array( 0, 2, 3, 'single', 'double' ), true ) ) {
394 $quote_style = ENT_QUOTES;
397 // More complete than get_html_translation_table( HTML_SPECIALCHARS )
398 $single = array( ''' => '\'', ''' => '\'' );
399 $single_preg = array( '/�*39;/' => ''', '/�*27;/i' => ''' );
400 $double = array( '"' => '"', '"' => '"', '"' => '"' );
401 $double_preg = array( '/�*34;/' => '"', '/�*22;/i' => '"' );
402 $others = array( '<' => '<', '<' => '<', '>' => '>', '>' => '>', '&' => '&', '&' => '&', '&' => '&' );
403 $others_preg = array( '/�*60;/' => '<', '/�*62;/' => '>', '/�*38;/' => '&', '/�*26;/i' => '&' );
405 if ( $quote_style === ENT_QUOTES ) {
406 $translation = array_merge( $single, $double, $others );
407 $translation_preg = array_merge( $single_preg, $double_preg, $others_preg );
408 } elseif ( $quote_style === ENT_COMPAT || $quote_style === 'double' ) {
409 $translation = array_merge( $double, $others );
410 $translation_preg = array_merge( $double_preg, $others_preg );
411 } elseif ( $quote_style === 'single' ) {
412 $translation = array_merge( $single, $others );
413 $translation_preg = array_merge( $single_preg, $others_preg );
414 } elseif ( $quote_style === ENT_NOQUOTES ) {
415 $translation = $others;
416 $translation_preg = $others_preg;
419 // Remove zero padding on numeric entities
420 $string = preg_replace( array_keys( $translation_preg ), array_values( $translation_preg ), $string );
422 // Replace characters according to translation table
423 return strtr( $string, $translation );
428 * Escaping for HTML blocks. Stolen from WP
432 * @param string $text
435 function yourls_esc_html( $text ) {
436 $safe_text = yourls_check_invalid_utf8( $text );
437 $safe_text = yourls_specialchars( $safe_text, ENT_QUOTES );
438 return yourls_apply_filter( 'esc_html', $safe_text, $text );
442 * Escaping for HTML attributes. Stolen from WP
446 * @param string $text
449 function yourls_esc_attr( $text ) {
450 $safe_text = yourls_check_invalid_utf8( $text );
451 $safe_text = yourls_specialchars( $safe_text, ENT_QUOTES );
452 return yourls_apply_filter( 'esc_attr', $safe_text, $text );
456 * Checks and cleans a URL before printing it. Stolen from WP.
458 * A number of characters are removed from the URL. If the URL is for displaying
459 * (the default behaviour) ampersands are also replaced.
463 * @param string $url The URL to be cleaned.
464 * @param string $context 'display' or something else. Use yourls_sanitize_url() for database or redirection usage.
465 * @param array $protocols Optional. Array of allowed protocols, defaults to global $yourls_allowedprotocols
466 * @return string The cleaned $url
468 function yourls_esc_url( $url, $context = 'display', $protocols = array() ) {
469 // make sure there's only one 'http://' at the beginning (prevents pasting a URL right after the default 'http://')
471 array( 'http://http://', 'http://https://' ),
472 array( 'http://', 'https://' ),
479 // make sure there's a protocol, add http:// if not
480 if ( ! yourls_get_protocol( $url ) )
481 $url = 'http://'.$url;
483 $original_url = $url;
485 // force scheme and domain to lowercase - see issues 591 and 1630
486 // We're not using parse_url() here because its opposite, http_build_url(), requires PECL. Plus, who doesn't love a neat Regexp? :)
487 if( preg_match( '!^([a-zA-Z0-9\+\.-]+:)(//)?(.*?@)?([^/#?]+)(.*)$!', $url, $matches ) ) {
488 list( $all, $scheme, $slashes, $userinfo, $domain, $rest ) = $matches;
489 $scheme = strtolower( $scheme );
490 // Domain to lowercase. On URN eg "urn:example:animal:ferret:nose" don't lowercase anything else
491 if( $slashes == '//' )
492 $domain = strtolower( $domain );
493 $url = $scheme . $slashes . $userinfo . $domain . $rest;
496 $url = preg_replace( '|[^a-z0-9-~+_.?#=!&;,/:%@$\|*\'()\\x80-\\xff]|i', '', $url );
497 // Previous regexp in YOURLS was '|[^a-z0-9-~+_.?\[\]\^#=!&;,/:%@$\|*`\'<>"()\\x80-\\xff\{\}]|i'
498 // TODO: check if that was it too destructive
499 $strip = array( '%0d', '%0a', '%0D', '%0A' );
500 $url = yourls_deep_replace( $strip, $url );
501 $url = str_replace( ';//', '://', $url );
503 // Replace ampersands and single quotes only when displaying.
504 if ( 'display' == $context ) {
505 $url = yourls_kses_normalize_entities( $url );
506 $url = str_replace( '&', '&', $url );
507 $url = str_replace( "'", ''', $url );
510 if ( ! is_array( $protocols ) or ! $protocols ) {
511 global $yourls_allowedprotocols;
512 $protocols = yourls_apply_filter( 'esc_url_protocols', $yourls_allowedprotocols );
513 // Note: $yourls_allowedprotocols is also globally filterable in functions-kses.php/yourls_kses_init()
516 if ( !yourls_is_allowed_protocol( $url, $protocols ) )
519 // I didn't use KSES function kses_bad_protocol() because it doesn't work the way I liked (returns //blah from illegal://blah)
521 return yourls_apply_filter( 'esc_url', $url, $original_url, $context );
525 * Escape single quotes, htmlspecialchar " < > &, and fix line endings. Stolen from WP.
527 * Escapes text strings for echoing in JS. It is intended to be used for inline JS
528 * (in a tag attribute, for example onclick="..."). Note that the strings have to
529 * be in single quotes. The filter 'js_escape' is also applied here.
533 * @param string $text The text to be escaped.
534 * @return string Escaped text.
536 function yourls_esc_js( $text ) {
537 $safe_text = yourls_check_invalid_utf8( $text );
538 $safe_text = yourls_specialchars( $safe_text, ENT_COMPAT );
539 $safe_text = preg_replace( '/&#(x)?0*(?(1)27|39);?/i', "'", stripslashes( $safe_text ) );
540 $safe_text = str_replace( "\r", '', $safe_text );
541 $safe_text = str_replace( "\n", '\\n', addslashes( $safe_text ) );
542 return yourls_apply_filter( 'esc_js', $safe_text, $text );
546 * Escaping for textarea values. Stolen from WP.
550 * @param string $text
553 function yourls_esc_textarea( $text ) {
554 $safe_text = htmlspecialchars( $text, ENT_QUOTES );
555 return yourls_apply_filter( 'esc_textarea', $safe_text, $text );
560 * PHP emulation of JS's encodeURI
562 * @link https://developer.mozilla.org/en/JavaScript/Reference/Global_Objects/encodeURI
566 function yourls_encodeURI( $url ) {
567 // Decode URL all the way
568 $result = yourls_rawurldecode_while_encoded( $url );
570 $result = strtr( rawurlencode( $result ), array (
571 '%3B' => ';', '%2C' => ',', '%2F' => '/', '%3F' => '?', '%3A' => ':', '%40' => '@',
572 '%26' => '&', '%3D' => '=', '%2B' => '+', '%24' => '$', '%21' => '!', '%2A' => '*',
573 '%27' => '\'', '%28' => '(', '%29' => ')', '%23' => '#',
576 // Known limit: this will most likely break IDN URLs such as http://www.académie-française.fr/
577 // To fully support IDN URLs, advocate use of a plugin.
578 return yourls_apply_filter( 'encodeURI', $result, $url );
582 * Adds backslashes before letters and before a number at the start of a string. Stolen from WP.
586 * @param string $string Value to which backslashes will be added.
587 * @return string String with backslashes inserted.
589 function yourls_backslashit($string) {
590 $string = preg_replace('/^([0-9])/', '\\\\\\\\\1', $string);
591 $string = preg_replace('/([a-z])/i', '\\\\\1', $string);
596 * Check if a string seems to be urlencoded
598 * We use rawurlencode instead of urlencode to avoid messing with '+'
601 * @param string $string
604 function yourls_is_rawurlencoded( $string ) {
605 return rawurldecode( $string ) != $string;
609 * rawurldecode a string till it's not encoded anymore
611 * Deals with multiple encoding (eg "%2521" => "%21" => "!").
612 * See https://github.com/YOURLS/YOURLS/issues/1303
615 * @param string $string
618 function yourls_rawurldecode_while_encoded( $string ) {
619 $string = rawurldecode( $string );
620 if( yourls_is_rawurlencoded( $string ) ) {
621 $string = yourls_rawurldecode_while_encoded( $string );
627 * Converts readable Javascript code into a valid bookmarklet link
629 * Uses https://github.com/ozh/bookmarkletgen
632 * @param string $code Javascript code
633 * @return string Bookmarklet link
635 function yourls_make_bookmarklet( $code ) {
636 if ( !class_exists( 'BookmarkletGen', false ) ) {
637 require_once YOURLS_INC . '/BookmarkletGen/BookmarkletGen.php';
640 $book = new BookmarkletGen;
641 return $book->crunch( $code );