includes/parser/StripState.php

   1 <?php
   2 /**
   3  * Holder for stripped items when parsing wiki markup.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Parser
  22  */
  23
  24 /**
  25  * @todo document, briefly.
  26  * @ingroup Parser
  27  */
  28 class StripState {
  29         protected $prefix;
  30         protected $data;
  31         protected $regex;
  32
  33         protected $circularRefGuard;
  34         protected $recursionLevel = 0;
  35
  36         const UNSTRIP_RECURSION_LIMIT = 20;
  37
  38         /**
  39          * @param string|null $prefix
  40          * @since 1.26 The prefix argument should be omitted, as the strip marker
  41          *  prefix string is now a constant.
  42          */
  43         public function __construct( $prefix = null ) {
  44                 if ( $prefix !== null ) {
  45                         wfDeprecated( __METHOD__ . ' with called with $prefix argument' .
  46                                 ' (call with no arguments instead)', '1.26' );
  47                 }
  48                 $this->data = [
  49                         'nowiki' => [],
  50                         'general' => []
  51                 ];
  52                 $this->regex = '/' . Parser::MARKER_PREFIX . "([^\x7f<>&'\"]+)" . Parser::MARKER_SUFFIX . '/';
  53                 $this->circularRefGuard = [];
  54         }
  55
  56         /**
  57          * Add a nowiki strip item
  58          * @param string $marker
  59          * @param string $value
  60          */
  61         public function addNoWiki( $marker, $value ) {
  62                 $this->addItem( 'nowiki', $marker, $value );
  63         }
  64
  65         /**
  66          * @param string $marker
  67          * @param string $value
  68          */
  69         public function addGeneral( $marker, $value ) {
  70                 $this->addItem( 'general', $marker, $value );
  71         }
  72
  73         /**
  74          * @throws MWException
  75          * @param string $type
  76          * @param string $marker
  77          * @param string $value
  78          */
  79         protected function addItem( $type, $marker, $value ) {
  80                 if ( !preg_match( $this->regex, $marker, $m ) ) {
  81                         throw new MWException( "Invalid marker: $marker" );
  82                 }
  83
  84                 $this->data[$type][$m[1]] = $value;
  85         }
  86
  87         /**
  88          * @param string $text
  89          * @return mixed
  90          */
  91         public function unstripGeneral( $text ) {
  92                 return $this->unstripType( 'general', $text );
  93         }
  94
  95         /**
  96          * @param string $text
  97          * @return mixed
  98          */
  99         public function unstripNoWiki( $text ) {
 100                 return $this->unstripType( 'nowiki', $text );
 101         }
 102
 103         /**
 104          * @param string $text
 105          * @return mixed
 106          */
 107         public function unstripBoth( $text ) {
 108                 $text = $this->unstripType( 'general', $text );
 109                 $text = $this->unstripType( 'nowiki', $text );
 110                 return $text;
 111         }
 112
 113         /**
 114          * @param string $type
 115          * @param string $text
 116          * @return mixed
 117          */
 118         protected function unstripType( $type, $text ) {
 119                 // Shortcut
 120                 if ( !count( $this->data[$type] ) ) {
 121                         return $text;
 122                 }
 123
 124                 $callback = function ( $m ) use ( $type ) {
 125                         $marker = $m[1];
 126                         if ( isset( $this->data[$type][$marker] ) ) {
 127                                 if ( isset( $this->circularRefGuard[$marker] ) ) {
 128                                         return '<span class="error">'
 129                                                 . wfMessage( 'parser-unstrip-loop-warning' )->inContentLanguage()->text()
 130                                                 . '</span>';
 131                                 }
 132                                 if ( $this->recursionLevel >= self::UNSTRIP_RECURSION_LIMIT ) {
 133                                         return '<span class="error">' .
 134                                                 wfMessage( 'parser-unstrip-recursion-limit' )
 135                                                         ->numParams( self::UNSTRIP_RECURSION_LIMIT )->inContentLanguage()->text() .
 136                                                 '</span>';
 137                                 }
 138                                 $this->circularRefGuard[$marker] = true;
 139                                 $this->recursionLevel++;
 140                                 $value = $this->data[$type][$marker];
 141                                 if ( $value instanceof Closure ) {
 142                                         $value = $value();
 143                                 }
 144                                 $ret = $this->unstripType( $type, $value );
 145                                 $this->recursionLevel--;
 146                                 unset( $this->circularRefGuard[$marker] );
 147                                 return $ret;
 148                         } else {
 149                                 return $m[0];
 150                         }
 151                 };
 152
 153                 $text = preg_replace_callback( $this->regex, $callback, $text );
 154                 return $text;
 155         }
 156
 157         /**
 158          * Get a StripState object which is sufficient to unstrip the given text.
 159          * It will contain the minimum subset of strip items necessary.
 160          *
 161          * @param string $text
 162          *
 163          * @return StripState
 164          */
 165         public function getSubState( $text ) {
 166                 $subState = new StripState();
 167                 $pos = 0;
 168                 while ( true ) {
 169                         $startPos = strpos( $text, Parser::MARKER_PREFIX, $pos );
 170                         $endPos = strpos( $text, Parser::MARKER_SUFFIX, $pos );
 171                         if ( $startPos === false || $endPos === false ) {
 172                                 break;
 173                         }
 174
 175                         $endPos += strlen( Parser::MARKER_SUFFIX );
 176                         $marker = substr( $text, $startPos, $endPos - $startPos );
 177                         if ( !preg_match( $this->regex, $marker, $m ) ) {
 178                                 continue;
 179                         }
 180
 181                         $key = $m[1];
 182                         if ( isset( $this->data['nowiki'][$key] ) ) {
 183                                 $subState->data['nowiki'][$key] = $this->data['nowiki'][$key];
 184                         } elseif ( isset( $this->data['general'][$key] ) ) {
 185                                 $subState->data['general'][$key] = $this->data['general'][$key];
 186                         }
 187                         $pos = $endPos;
 188                 }
 189                 return $subState;
 190         }
 191
 192         /**
 193          * Merge another StripState object into this one. The strip marker keys
 194          * will not be preserved. The strings in the $texts array will have their
 195          * strip markers rewritten, the resulting array of strings will be returned.
 196          *
 197          * @param StripState $otherState
 198          * @param array $texts
 199          * @return array
 200          */
 201         public function merge( $otherState, $texts ) {
 202                 $mergePrefix = wfRandomString( 16 );
 203
 204                 foreach ( $otherState->data as $type => $items ) {
 205                         foreach ( $items as $key => $value ) {
 206                                 $this->data[$type]["$mergePrefix-$key"] = $value;
 207                         }
 208                 }
 209
 210                 $callback = function ( $m ) use ( $mergePrefix ) {
 211                         $key = $m[1];
 212                         return Parser::MARKER_PREFIX . $mergePrefix . '-' . $key . Parser::MARKER_SUFFIX;
 213                 };
 214                 $texts = preg_replace_callback( $otherState->regex, $callback, $texts );
 215                 return $texts;
 216         }
 217
 218         /**
 219          * Remove any strip markers found in the given text.
 220          *
 221          * @param string $text
 222          * @return string
 223          */
 224         public function killMarkers( $text ) {
 225                 return preg_replace( $this->regex, '', $text );
 226         }
 227 }