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