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 $data;
  30         protected $regex;
  31
  32         protected $parser;
  33
  34         protected $circularRefGuard;
  35         protected $depth = 0;
  36         protected $highestDepth = 0;
  37         protected $expandSize = 0;
  38
  39         protected $depthLimit = 20;
  40         protected $sizeLimit = 5000000;
  41
  42         /**
  43          * @param Parser|null $parser
  44          * @param array $options
  45          */
  46         public function __construct( Parser $parser = null, $options = [] ) {
  47                 $this->data = [
  48                         'nowiki' => [],
  49                         'general' => []
  50                 ];
  51                 $this->regex = '/' . Parser::MARKER_PREFIX . "([^\x7f<>&'\"]+)" . Parser::MARKER_SUFFIX . '/';
  52                 $this->circularRefGuard = [];
  53                 $this->parser = $parser;
  54
  55                 if ( isset( $options['depthLimit'] ) ) {
  56                         $this->depthLimit = $options['depthLimit'];
  57                 }
  58                 if ( isset( $options['sizeLimit'] ) ) {
  59                         $this->sizeLimit = $options['sizeLimit'];
  60                 }
  61         }
  62
  63         /**
  64          * Add a nowiki strip item
  65          * @param string $marker
  66          * @param string $value
  67          */
  68         public function addNoWiki( $marker, $value ) {
  69                 $this->addItem( 'nowiki', $marker, $value );
  70         }
  71
  72         /**
  73          * @param string $marker
  74          * @param string $value
  75          */
  76         public function addGeneral( $marker, $value ) {
  77                 $this->addItem( 'general', $marker, $value );
  78         }
  79
  80         /**
  81          * @throws MWException
  82          * @param string $type
  83          * @param string $marker
  84          * @param string $value
  85          */
  86         protected function addItem( $type, $marker, $value ) {
  87                 if ( !preg_match( $this->regex, $marker, $m ) ) {
  88                         throw new MWException( "Invalid marker: $marker" );
  89                 }
  90
  91                 $this->data[$type][$m[1]] = $value;
  92         }
  93
  94         /**
  95          * @param string $text
  96          * @return mixed
  97          */
  98         public function unstripGeneral( $text ) {
  99                 return $this->unstripType( 'general', $text );
 100         }
 101
 102         /**
 103          * @param string $text
 104          * @return mixed
 105          */
 106         public function unstripNoWiki( $text ) {
 107                 return $this->unstripType( 'nowiki', $text );
 108         }
 109
 110         /**
 111          * @param string $text
 112          * @return mixed
 113          */
 114         public function unstripBoth( $text ) {
 115                 $text = $this->unstripType( 'general', $text );
 116                 $text = $this->unstripType( 'nowiki', $text );
 117                 return $text;
 118         }
 119
 120         /**
 121          * @param string $type
 122          * @param string $text
 123          * @return mixed
 124          */
 125         protected function unstripType( $type, $text ) {
 126                 // Shortcut
 127                 if ( !count( $this->data[$type] ) ) {
 128                         return $text;
 129                 }
 130
 131                 $callback = function ( $m ) use ( $type ) {
 132                         $marker = $m[1];
 133                         if ( isset( $this->data[$type][$marker] ) ) {
 134                                 if ( isset( $this->circularRefGuard[$marker] ) ) {
 135                                         return $this->getWarning( 'parser-unstrip-loop-warning' );
 136                                 }
 137
 138                                 if ( $this->depth > $this->highestDepth ) {
 139                                         $this->highestDepth = $this->depth;
 140                                 }
 141                                 if ( $this->depth >= $this->depthLimit ) {
 142                                         return $this->getLimitationWarning( 'unstrip-depth', $this->depthLimit );
 143                                 }
 144
 145                                 $value = $this->data[$type][$marker];
 146                                 if ( $value instanceof Closure ) {
 147                                         $value = $value();
 148                                 }
 149
 150                                 $this->expandSize += strlen( $value );
 151                                 if ( $this->expandSize > $this->sizeLimit ) {
 152                                         return $this->getLimitationWarning( 'unstrip-size', $this->sizeLimit );
 153                                 }
 154
 155                                 $this->circularRefGuard[$marker] = true;
 156                                 $this->depth++;
 157                                 $ret = $this->unstripType( $type, $value );
 158                                 $this->depth--;
 159                                 unset( $this->circularRefGuard[$marker] );
 160
 161                                 return $ret;
 162                         } else {
 163                                 return $m[0];
 164                         }
 165                 };
 166
 167                 $text = preg_replace_callback( $this->regex, $callback, $text );
 168                 return $text;
 169         }
 170
 171         /**
 172          * Get warning HTML and register a limitation warning with the parser
 173          *
 174          * @param string $type
 175          * @param int $max
 176          * @return string
 177          */
 178         private function getLimitationWarning( $type, $max = '' ) {
 179                 if ( $this->parser ) {
 180                         $this->parser->limitationWarn( $type, $max );
 181                 }
 182                 return $this->getWarning( "$type-warning", $max );
 183         }
 184
 185         /**
 186          * Get warning HTML
 187          *
 188          * @param string $message
 189          * @param int $max
 190          * @return string
 191          */
 192         private function getWarning( $message, $max = '' ) {
 193                 return '<span class="error">' .
 194                         wfMessage( $message )
 195                                 ->numParams( $max )->inContentLanguage()->text() .
 196                         '</span>';
 197         }
 198
 199         /**
 200          * Get an array of parameters to pass to ParserOutput::setLimitReportData()
 201          *
 202          * @unstable Should only be called by Parser
 203          * @return array
 204          */
 205         public function getLimitReport() {
 206                 return [
 207                         [ 'limitreport-unstrip-depth',
 208                                 [
 209                                         $this->highestDepth,
 210                                         $this->depthLimit
 211                                 ],
 212                         ],
 213                         [ 'limitreport-unstrip-size',
 214                                 [
 215                                         $this->expandSize,
 216                                         $this->sizeLimit
 217                                 ],
 218                         ]
 219                 ];
 220         }
 221
 222         /**
 223          * Get a StripState object which is sufficient to unstrip the given text.
 224          * It will contain the minimum subset of strip items necessary.
 225          *
 226          * @deprecated since 1.31
 227          * @param string $text
 228          * @return StripState
 229          */
 230         public function getSubState( $text ) {
 231                 wfDeprecated( __METHOD__, '1.31' );
 232
 233                 $subState = new StripState;
 234                 $pos = 0;
 235                 while ( true ) {
 236                         $startPos = strpos( $text, Parser::MARKER_PREFIX, $pos );
 237                         $endPos = strpos( $text, Parser::MARKER_SUFFIX, $pos );
 238                         if ( $startPos === false || $endPos === false ) {
 239                                 break;
 240                         }
 241
 242                         $endPos += strlen( Parser::MARKER_SUFFIX );
 243                         $marker = substr( $text, $startPos, $endPos - $startPos );
 244                         if ( !preg_match( $this->regex, $marker, $m ) ) {
 245                                 continue;
 246                         }
 247
 248                         $key = $m[1];
 249                         if ( isset( $this->data['nowiki'][$key] ) ) {
 250                                 $subState->data['nowiki'][$key] = $this->data['nowiki'][$key];
 251                         } elseif ( isset( $this->data['general'][$key] ) ) {
 252                                 $subState->data['general'][$key] = $this->data['general'][$key];
 253                         }
 254                         $pos = $endPos;
 255                 }
 256                 return $subState;
 257         }
 258
 259         /**
 260          * Merge another StripState object into this one. The strip marker keys
 261          * will not be preserved. The strings in the $texts array will have their
 262          * strip markers rewritten, the resulting array of strings will be returned.
 263          *
 264          * @deprecated since 1.31
 265          * @param StripState $otherState
 266          * @param array $texts
 267          * @return array
 268          */
 269         public function merge( $otherState, $texts ) {
 270                 wfDeprecated( __METHOD__, '1.31' );
 271
 272                 $mergePrefix = wfRandomString( 16 );
 273
 274                 foreach ( $otherState->data as $type => $items ) {
 275                         foreach ( $items as $key => $value ) {
 276                                 $this->data[$type]["$mergePrefix-$key"] = $value;
 277                         }
 278                 }
 279
 280                 $callback = function ( $m ) use ( $mergePrefix ) {
 281                         $key = $m[1];
 282                         return Parser::MARKER_PREFIX . $mergePrefix . '-' . $key . Parser::MARKER_SUFFIX;
 283                 };
 284                 $texts = preg_replace_callback( $otherState->regex, $callback, $texts );
 285                 return $texts;
 286         }
 287
 288         /**
 289          * Remove any strip markers found in the given text.
 290          *
 291          * @param string $text
 292          * @return string
 293          */
 294         public function killMarkers( $text ) {
 295                 return preg_replace( $this->regex, '', $text );
 296         }
 297 }