includes/libs/CSSJanus.php

   1 <?php
   2 /**
   3  * This program is free software; you can redistribute it and/or modify
   4  * it under the terms of the GNU General Public License as published by
   5  * the Free Software Foundation; either version 2 of the License, or
   6  * (at your option) any later version.
   7  *
   8  * This program is distributed in the hope that it will be useful,
   9  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  10  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  11  * GNU General Public License for more details.
  12  *
  13  * You should have received a copy of the GNU General Public License along
  14  * with this program; if not, write to the Free Software Foundation, Inc.,
  15  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  16  * http://www.gnu.org/copyleft/gpl.html
  17  *
  18  */
  19
  20 /**
  21  * This is a PHP port of CSSJanus, a utility that transforms CSS style sheets
  22  * written for LTR to RTL.
  23  *
  24  * The original Python version of CSSJanus is Copyright 2008 by Google Inc. and
  25  * is distributed under the Apache license. This PHP port is Copyright 2010 by
  26  * Roan Kattouw and is dual-licensed under the GPL (as in the comment above) and
  27  * the Apache (as in the original code) licenses.
  28  *
  29  * Original code: http://code.google.com/p/cssjanus/source/browse/trunk/cssjanus.py
  30  * License of original code: http://code.google.com/p/cssjanus/source/browse/trunk/LICENSE
  31  * @author Roan Kattouw
  32  *
  33  */
  34 class CSSJanus {
  35         // Patterns defined as null are built dynamically by buildPatterns()
  36         private static $patterns = array(
  37                 'tmpToken' => '`TMP`',
  38                 'nonAscii' => '[\200-\377]',
  39                 'unicode' => '(?:(?:\\[0-9a-f]{1,6})(?:\r\n|\s)?)',
  40                 'num' => '(?:[0-9]*\.[0-9]+|[0-9]+)',
  41                 'unit' => '(?:em|ex|px|cm|mm|in|pt|pc|deg|rad|grad|ms|s|hz|khz|%)',
  42                 'body_selector' => 'body\s*{\s*',
  43                 'direction' => 'direction\s*:\s*',
  44                 'escape' => null,
  45                 'nmstart' => null,
  46                 'nmchar' => null,
  47                 'ident' => null,
  48                 'quantity' => null,
  49                 'possibly_negative_quantity' => null,
  50                 'color' => null,
  51                 'url_special_chars' => '[!#$%&*-~]',
  52                 'valid_after_uri_chars' => '[\'\"]?\s*',
  53                 'url_chars' => null,
  54                 'lookahead_not_open_brace' => null,
  55                 'lookahead_not_closing_paren' => null,
  56                 'lookahead_for_closing_paren' => null,
  57                 'lookbehind_not_letter' => '(?<![a-zA-Z])',
  58                 'chars_within_selector' => '[^\}]*?',
  59                 'noflip_annotation' => '\/\*\s*@noflip\s*\*\/',
  60                 'noflip_single' => null,
  61                 'noflip_class' => null,
  62                 'comment' => '/\/\*[^*]*\*+([^\/*][^*]*\*+)*\//',
  63                 'direction_ltr' => null,
  64                 'direction_rtl' => null,
  65                 'left' => null,
  66                 'right' => null,
  67                 'left_in_url' => null,
  68                 'right_in_url' => null,
  69                 'ltr_in_url' => null,
  70                 'rtl_in_url' => null,
  71                 'cursor_east' => null,
  72                 'cursor_west' => null,
  73                 'four_notation_quantity' => null,
  74                 'four_notation_color' => null,
  75                 'bg_horizontal_percentage' => null,
  76                 'bg_horizontal_percentage_x' => null,
  77         );
  78
  79         /**
  80          * Build patterns we can't define above because they depend on other patterns.
  81          */
  82         private static function buildPatterns() {
  83                 if ( !is_null( self::$patterns['escape'] ) ) {
  84                         // Patterns have already been built
  85                         return;
  86                 }
  87
  88                 $patterns =& self::$patterns;
  89                 $patterns['escape'] = "(?:{$patterns['unicode']}|\\[^\r\n\f0-9a-f])";
  90                 $patterns['nmstart'] = "(?:[_a-z]|{$patterns['nonAscii']}|{$patterns['escape']})";
  91                 $patterns['nmchar'] = "(?:[_a-z0-9-]|{$patterns['nonAscii']}|{$patterns['escape']})";
  92                 $patterns['ident'] = "-?{$patterns['nmstart']}{$patterns['nmchar']}*";
  93                 $patterns['quantity'] = "{$patterns['num']}(?:\s*{$patterns['unit']}|{$patterns['ident']})?";
  94                 $patterns['possibly_negative_quantity'] = "((?:-?{$patterns['quantity']})|(?:inherit|auto))";
  95                 $patterns['color'] = "(#?{$patterns['nmchar']}+)";
  96                 $patterns['url_chars'] = "(?:{$patterns['url_special_chars']}|{$patterns['nonAscii']}|{$patterns['escape']})*";
  97                 $patterns['lookahead_not_open_brace'] = "(?!({$patterns['nmchar']}|\r?\n|\s|#|\:|\.|\,|\+|>)*?{)";
  98                 $patterns['lookahead_not_closing_paren'] = "(?!{$patterns['url_chars']}?{$patterns['valid_after_uri_chars']}\))";
  99                 $patterns['lookahead_for_closing_paren'] = "(?={$patterns['url_chars']}?{$patterns['valid_after_uri_chars']}\))";
 100                 $patterns['noflip_single'] = "/({$patterns['noflip_annotation']}{$patterns['lookahead_not_open_brace']}[^;}]+;?)/i";
 101                 $patterns['noflip_class'] = "/({$patterns['noflip_annotation']}{$patterns['chars_within_selector']}})/i";
 102                 $patterns['direction_ltr'] = "/({$patterns['direction']})ltr/i";
 103                 $patterns['direction_rtl'] = "/({$patterns['direction']})rtl/i";
 104                 $patterns['left'] = "/{$patterns['lookbehind_not_letter']}(left){$patterns['lookahead_not_closing_paren']}{$patterns['lookahead_not_open_brace']}/i";
 105                 $patterns['right'] = "/{$patterns['lookbehind_not_letter']}(right){$patterns['lookahead_not_closing_paren']}{$patterns['lookahead_not_open_brace']}/i";
 106                 $patterns['left_in_url'] = "/{$patterns['lookbehind_not_letter']}(left){$patterns['lookahead_for_closing_paren']}/i";
 107                 $patterns['right_in_url'] = "/{$patterns['lookbehind_not_letter']}(right){$patterns['lookahead_for_closing_paren']}/i";
 108                 $patterns['ltr_in_url'] = "/{$patterns['lookbehind_not_letter']}(ltr){$patterns['lookahead_for_closing_paren']}/i";
 109                 $patterns['rtl_in_url'] = "/{$patterns['lookbehind_not_letter']}(rtl){$patterns['lookahead_for_closing_paren']}/i";
 110                 $patterns['cursor_east'] = "/{$patterns['lookbehind_not_letter']}([ns]?)e-resize/";
 111                 $patterns['cursor_west'] = "/{$patterns['lookbehind_not_letter']}([ns]?)w-resize/";
 112                 $patterns['four_notation_quantity'] = "/{$patterns['possibly_negative_quantity']}(\s+){$patterns['possibly_negative_quantity']}(\s+){$patterns['possibly_negative_quantity']}(\s+){$patterns['possibly_negative_quantity']}/i";
 113                 $patterns['four_notation_color'] = "/(-color\s*:\s*){$patterns['color']}(\s+){$patterns['color']}(\s+){$patterns['color']}(\s+){$patterns['color']}/i";
 114                 // The two regexes below are parenthesized differently then in the original implementation to make the
 115                 // callback's job more straightforward
 116                 $patterns['bg_horizontal_percentage'] = "/(background(?:-position)?\s*:\s*[^%]*?)({$patterns['num']})(%\s*(?:{$patterns['quantity']}|{$patterns['ident']}))/";
 117                 $patterns['bg_horizontal_percentage_x'] = "/(background-position-x\s*:\s*)({$patterns['num']})(%)/";
 118         }
 119
 120         /**
 121          * Transform an LTR stylesheet to RTL
 122          * @param $css String: stylesheet to transform
 123          * @param $swapLtrRtlInURL Boolean: If true, swap 'ltr' and 'rtl' in URLs
 124          * @param $swapLeftRightInURL Boolean: If true, swap 'left' and 'right' in URLs
 125          * @return Transformed stylesheet
 126          */
 127         public static function transform( $css, $swapLtrRtlInURL = false, $swapLeftRightInURL = false ) {
 128                 // We wrap tokens in ` , not ~ like the original implementation does.
 129                 // This was done because ` is not a legal character in CSS and can only
 130                 // occur in URLs, where we escape it to %60 before inserting our tokens.
 131                 $css = str_replace( '`', '%60', $css );
 132
 133                 self::buildPatterns();
 134
 135                 // Tokenize single line rules with /* @noflip */
 136                 $noFlipSingle = new CSSJanus_Tokenizer( self::$patterns['noflip_single'], '`NOFLIP_SINGLE`' );
 137                 $css = $noFlipSingle->tokenize( $css );
 138
 139                 // Tokenize class rules with /* @noflip */
 140                 $noFlipClass = new CSSJanus_Tokenizer( self::$patterns['noflip_class'], '`NOFLIP_CLASS`' );
 141                 $css = $noFlipClass->tokenize( $css );
 142
 143                 // Tokenize comments
 144                 $comments = new CSSJanus_Tokenizer( self::$patterns['comment'], '`C`' );
 145                 $css = $comments->tokenize( $css );
 146
 147                 // LTR->RTL fixes start here
 148                 $css = self::fixDirection( $css );
 149                 if ( $swapLtrRtlInURL ) {
 150                         $css = self::fixLtrRtlInURL( $css );
 151                 }
 152
 153                 if ( $swapLeftRightInURL ) {
 154                         $css = self::fixLeftRightInURL( $css );
 155                 }
 156                 $css = self::fixLeftAndRight( $css );
 157                 $css = self::fixCursorProperties( $css );
 158                 $css = self::fixFourPartNotation( $css );
 159                 $css = self::fixBackgroundPosition( $css );
 160
 161                 // Detokenize stuff we tokenized before
 162                 $css = $comments->detokenize( $css );
 163                 $css = $noFlipClass->detokenize( $css );
 164                 $css = $noFlipSingle->detokenize( $css );
 165
 166                 return $css;
 167         }
 168
 169         /**
 170          * Replace direction: ltr; with direction: rtl; and vice versa.
 171          *
 172          * The original implementation only does this inside body selectors
 173          * and misses "body\n{\ndirection:ltr;\n}". This function does not have
 174          * these problems.
 175          *
 176          * See http://code.google.com/p/cssjanus/issues/detail?id=15 and
 177          * TODO: URL
 178          */
 179         private static function fixDirection( $css ) {
 180                 $css = preg_replace( self::$patterns['direction_ltr'],
 181                         '$1' . self::$patterns['tmpToken'], $css );
 182                 $css = preg_replace( self::$patterns['direction_rtl'], '$1ltr', $css );
 183                 $css = str_replace( self::$patterns['tmpToken'], 'rtl', $css );
 184
 185                 return $css;
 186         }
 187
 188         /**
 189          * Replace 'ltr' with 'rtl' and vice versa in background URLs
 190          */
 191         private static function fixLtrRtlInURL( $css ) {
 192                 $css = preg_replace( self::$patterns['ltr_in_url'], self::$patterns['tmpToken'], $css );
 193                 $css = preg_replace( self::$patterns['rtl_in_url'], 'ltr', $css );
 194                 $css = str_replace( self::$patterns['tmpToken'], 'rtl', $css );
 195
 196                 return $css;
 197         }
 198
 199         /**
 200          * Replace 'left' with 'right' and vice versa in background URLs
 201          */
 202         private static function fixLeftRightInURL( $css ) {
 203                 $css = preg_replace( self::$patterns['left_in_url'], self::$patterns['tmpToken'], $css );
 204                 $css = preg_replace( self::$patterns['right_in_url'], 'left', $css );
 205                 $css = str_replace( self::$patterns['tmpToken'], 'right', $css );
 206
 207                 return $css;
 208         }
 209
 210         /**
 211          * Flip rules like left: , padding-right: , etc.
 212          */
 213         private static function fixLeftAndRight( $css ) {
 214                 $css = preg_replace( self::$patterns['left'], self::$patterns['tmpToken'], $css );
 215                 $css = preg_replace( self::$patterns['right'], 'left', $css );
 216                 $css = str_replace( self::$patterns['tmpToken'], 'right', $css );
 217
 218                 return $css;
 219         }
 220
 221         /**
 222          * Flip East and West in rules like cursor: nw-resize;
 223          */
 224         private static function fixCursorProperties( $css ) {
 225                 $css = preg_replace( self::$patterns['cursor_east'],
 226                         '$1' . self::$patterns['tmpToken'], $css );
 227                 $css = preg_replace( self::$patterns['cursor_west'], '$1e-resize', $css );
 228                 $css = str_replace( self::$patterns['tmpToken'], 'w-resize', $css );
 229
 230                 return $css;
 231         }
 232
 233         /**
 234          * Swap the second and fourth parts in four-part notation rules like
 235          * padding: 1px 2px 3px 4px;
 236          *
 237          * Unlike the original implementation, this function doesn't suffer from
 238          * the bug where whitespace is not preserved when flipping four-part rules
 239          * and four-part color rules with multiple whitespace characters between
 240          * colors are not recognized.
 241          * See http://code.google.com/p/cssjanus/issues/detail?id=16
 242          */
 243         private static function fixFourPartNotation( $css ) {
 244                 $css = preg_replace( self::$patterns['four_notation_quantity'], '$1$2$7$4$5$6$3', $css );
 245                 $css = preg_replace( self::$patterns['four_notation_color'], '$1$2$3$8$5$6$7$4', $css );
 246
 247                 return $css;
 248         }
 249
 250         /**
 251          * Flip horizontal background percentages.
 252          */
 253         private static function fixBackgroundPosition( $css ) {
 254                 $css = preg_replace_callback( self::$patterns['bg_horizontal_percentage'],
 255                         array( 'self', 'calculateNewBackgroundPosition' ), $css );
 256                 $css = preg_replace_callback( self::$patterns['bg_horizontal_percentage_x'],
 257                         array( 'self', 'calculateNewBackgroundPosition' ), $css );
 258
 259                 return $css;
 260         }
 261
 262         /**
 263          * Callback for calculateNewBackgroundPosition()
 264          */
 265         private static function calculateNewBackgroundPosition( $matches ) {
 266                 return $matches[1] . ( 100 - $matches[2] ) . $matches[3];
 267         }
 268 }
 269
 270 /**
 271  * Utility class used by CSSJanus that tokenizes and untokenizes things we want
 272  * to protect from being janused.
 273  * @author Roan Kattouw
 274  */
 275 class CSSJanus_Tokenizer {
 276         private $regex, $token;
 277         private $originals;
 278
 279         /**
 280          * Constructor
 281          * @param $regex string Regular expression whose matches to replace by a token.
 282          * @param $token string Token
 283          */
 284         public function __construct( $regex, $token ) {
 285                 $this->regex = $regex;
 286                 $this->token = $token;
 287                 $this->originals = array();
 288         }
 289
 290         /**
 291          * Replace all occurrences of $regex in $str with a token and remember
 292          * the original strings.
 293          * @param $str String to tokenize
 294          * @return string Tokenized string
 295          */
 296         public function tokenize( $str ) {
 297                 return preg_replace_callback( $this->regex, array( $this, 'tokenizeCallback' ), $str );
 298         }
 299
 300         private function tokenizeCallback( $matches ) {
 301                 $this->originals[] = $matches[0];
 302                 return $this->token;
 303         }
 304
 305         /**
 306          * Replace tokens with their originals. If multiple strings were tokenized, it's important they be
 307          * detokenized in exactly the SAME ORDER.
 308          * @param $str String: previously run through tokenize()
 309          * @return string Original string
 310          */
 311         public function detokenize( $str ) {
 312                 // PHP has no function to replace only the first occurrence or to
 313                 // replace occurrences of the same string with different values,
 314                 // so we use preg_replace_callback() even though we don't really need a regex
 315                 return preg_replace_callback( '/' . preg_quote( $this->token, '/' ) . '/',
 316                         array( $this, 'detokenizeCallback' ), $str );
 317         }
 318
 319         private function detokenizeCallback( $matches ) {
 320                 $retval = current( $this->originals );
 321                 next( $this->originals );
 322
 323                 return $retval;
 324         }
 325 }