includes/diff/TextSlotDiffRenderer.php

   1 <?php
   2 /**
   3  * Renders a slot diff by doing a text diff on the native representation.
   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 DifferenceEngine
  22  */
  23
  24 use MediaWiki\Shell\Shell;
  25 use Wikimedia\Assert\Assert;
  26
  27 /**
  28  * Renders a slot diff by doing a text diff on the native representation.
  29  *
  30  * If you want to use this without content objects (to call getTextDiff() on some
  31  * non-content-related texts), obtain an instance with
  32  *     ContentHandler::getForModelID( CONTENT_MODEL_TEXT )
  33  *         ->getSlotDiffRenderer( RequestContext::getMain() )
  34  *
  35  * @ingroup DifferenceEngine
  36  */
  37 class TextSlotDiffRenderer extends SlotDiffRenderer {
  38
  39         /** Use the PHP diff implementation (DiffEngine). */
  40         const ENGINE_PHP = 'php';
  41
  42         /** Use the wikidiff2 PHP module. */
  43         const ENGINE_WIKIDIFF2 = 'wikidiff2';
  44
  45         /** Use an external executable. */
  46         const ENGINE_EXTERNAL = 'external';
  47
  48         /** @var IBufferingStatsdDataFactory|null */
  49         private $statsdDataFactory;
  50
  51         /** @var Language|null The language this content is in. */
  52         private $language;
  53
  54         /** @var string One of the ENGINE_* constants. */
  55         private $engine = self::ENGINE_PHP;
  56
  57         /** @var string Path to an executable to be used as the diff engine. */
  58         private $externalEngine;
  59
  60         /**
  61          * Convenience helper to use getTextDiff without an instance.
  62          * @param string $oldText
  63          * @param string $newText
  64          * @return string
  65          */
  66         public static function diff( $oldText, $newText ) {
  67                 /** @var TextSlotDiffRenderer $slotDiffRenderer */
  68                 $slotDiffRenderer = ContentHandler::getForModelID( CONTENT_MODEL_TEXT )
  69                         ->getSlotDiffRenderer( RequestContext::getMain() );
  70                 '@phan-var TextSlotDiffRenderer $slotDiffRenderer';
  71                 return $slotDiffRenderer->getTextDiff( $oldText, $newText );
  72         }
  73
  74         public function setStatsdDataFactory( IBufferingStatsdDataFactory $statsdDataFactory ) {
  75                 $this->statsdDataFactory = $statsdDataFactory;
  76         }
  77
  78         public function setLanguage( Language $language ) {
  79                 $this->language = $language;
  80         }
  81
  82         /**
  83          * Set which diff engine to use.
  84          * @param string $type One of the ENGINE_* constants.
  85          * @param string|null $executable Path to an external exectable, only when type is ENGINE_EXTERNAL.
  86          */
  87         public function setEngine( $type, $executable = null ) {
  88                 $engines = [ self::ENGINE_PHP, self::ENGINE_WIKIDIFF2, self::ENGINE_EXTERNAL ];
  89                 Assert::parameter( in_array( $type, $engines, true ), '$type',
  90                         'must be one of the TextSlotDiffRenderer::ENGINE_* constants' );
  91                 if ( $type === self::ENGINE_EXTERNAL ) {
  92                         Assert::parameter( is_string( $executable ) && is_executable( $executable ), '$executable',
  93                                 'must be a path to a valid executable' );
  94                 } else {
  95                         Assert::parameter( is_null( $executable ), '$executable',
  96                                 'must not be set unless $type is ENGINE_EXTERNAL' );
  97                 }
  98                 $this->engine = $type;
  99                 $this->externalEngine = $executable;
 100         }
 101
 102         /** @inheritDoc */
 103         public function getDiff( Content $oldContent = null, Content $newContent = null ) {
 104                 $this->normalizeContents( $oldContent, $newContent, TextContent::class );
 105
 106                 $oldText = $oldContent->serialize();
 107                 $newText = $newContent->serialize();
 108
 109                 return $this->getTextDiff( $oldText, $newText );
 110         }
 111
 112         /**
 113          * Diff the text representations of two content objects (or just two pieces of text in general).
 114          * @param string $oldText
 115          * @param string $newText
 116          * @return string HTML, one or more <tr> tags.
 117          */
 118         public function getTextDiff( $oldText, $newText ) {
 119                 Assert::parameterType( 'string', $oldText, '$oldText' );
 120                 Assert::parameterType( 'string', $newText, '$newText' );
 121
 122                 $diff = function () use ( $oldText, $newText ) {
 123                         $time = microtime( true );
 124
 125                         $result = $this->getTextDiffInternal( $oldText, $newText );
 126
 127                         $time = intval( ( microtime( true ) - $time ) * 1000 );
 128                         if ( $this->statsdDataFactory ) {
 129                                 $this->statsdDataFactory->timing( 'diff_time', $time );
 130                         }
 131
 132                         // TODO reimplement this using T142313
 133                         /*
 134                         // Log requests slower than 99th percentile
 135                         if ( $time > 100 && $this->mOldPage && $this->mNewPage ) {
 136                                 wfDebugLog( 'diff',
 137                                         "$time ms diff: {$this->mOldid} -> {$this->mNewid} {$this->mNewPage}" );
 138                         }
 139                         */
 140
 141                         return $result;
 142                 };
 143
 144                 /**
 145                  * @param Status $status
 146                  * @throws FatalError
 147                  */
 148                 $error = function ( $status ) {
 149                         throw new FatalError( $status->getWikiText() );
 150                 };
 151
 152                 // Use PoolCounter if the diff looks like it can be expensive
 153                 if ( strlen( $oldText ) + strlen( $newText ) > 20000 ) {
 154                         $work = new PoolCounterWorkViaCallback( 'diff',
 155                                 md5( $oldText ) . md5( $newText ),
 156                                 [ 'doWork' => $diff, 'error' => $error ]
 157                         );
 158                         return $work->execute();
 159                 }
 160
 161                 return $diff();
 162         }
 163
 164         /**
 165          * Diff the text representations of two content objects (or just two pieces of text in general).
 166          * This does the actual diffing, getTextDiff() wraps it with logging and resource limiting.
 167          * @param string $oldText
 168          * @param string $newText
 169          * @return string
 170          * @throws Exception
 171          */
 172         protected function getTextDiffInternal( $oldText, $newText ) {
 173                 // TODO move most of this into three parallel implementations of a text diff generator
 174                 // class, choose which one to use via dependecy injection
 175
 176                 $oldText = str_replace( "\r\n", "\n", $oldText );
 177                 $newText = str_replace( "\r\n", "\n", $newText );
 178
 179                 // Better external diff engine, the 2 may some day be dropped
 180                 // This one does the escaping and segmenting itself
 181                 if ( $this->engine === self::ENGINE_WIKIDIFF2 ) {
 182                         $wikidiff2Version = phpversion( 'wikidiff2' );
 183                         if (
 184                                 $wikidiff2Version !== false &&
 185                                 version_compare( $wikidiff2Version, '1.5.0', '>=' ) &&
 186                                 version_compare( $wikidiff2Version, '1.8.0', '<' )
 187                         ) {
 188                                 $text = wikidiff2_do_diff(
 189                                         $oldText,
 190                                         $newText,
 191                                         2,
 192                                         0
 193                                 );
 194                         } else {
 195                                 // Don't pass the 4th parameter introduced in version 1.5.0 and removed in version 1.8.0
 196                                 $text = wikidiff2_do_diff(
 197                                         $oldText,
 198                                         $newText,
 199                                         2
 200                                 );
 201                         }
 202
 203                         return $text;
 204                 } elseif ( $this->engine === self::ENGINE_EXTERNAL ) {
 205                         # Diff via the shell
 206                         $tmpDir = wfTempDir();
 207                         $tempName1 = tempnam( $tmpDir, 'diff_' );
 208                         $tempName2 = tempnam( $tmpDir, 'diff_' );
 209
 210                         $tempFile1 = fopen( $tempName1, "w" );
 211                         if ( !$tempFile1 ) {
 212                                 return false;
 213                         }
 214                         $tempFile2 = fopen( $tempName2, "w" );
 215                         if ( !$tempFile2 ) {
 216                                 return false;
 217                         }
 218                         fwrite( $tempFile1, $oldText );
 219                         fwrite( $tempFile2, $newText );
 220                         fclose( $tempFile1 );
 221                         fclose( $tempFile2 );
 222                         $cmd = [ $this->externalEngine, $tempName1, $tempName2 ];
 223                         $result = Shell::command( $cmd )
 224                                 ->execute();
 225                         $exitCode = $result->getExitCode();
 226                         if ( $exitCode !== 0 ) {
 227                                 throw new Exception( "External diff command returned code {$exitCode}. Stderr: "
 228                                                                          . wfEscapeWikiText( $result->getStderr() )
 229                                 );
 230                         }
 231                         $difftext = $result->getStdout();
 232                         unlink( $tempName1 );
 233                         unlink( $tempName2 );
 234
 235                         return $difftext;
 236                 } elseif ( $this->engine === self::ENGINE_PHP ) {
 237                         if ( $this->language ) {
 238                                 $oldText = $this->language->segmentForDiff( $oldText );
 239                                 $newText = $this->language->segmentForDiff( $newText );
 240                         }
 241                         $ota = explode( "\n", $oldText );
 242                         $nta = explode( "\n", $newText );
 243                         $diffs = new Diff( $ota, $nta );
 244                         $formatter = new TableDiffFormatter();
 245                         $difftext = $formatter->format( $diffs );
 246                         if ( $this->language ) {
 247                                 $difftext = $this->language->unsegmentForDiff( $difftext );
 248                         }
 249
 250                         return $difftext;
 251                 }
 252                 throw new LogicException( 'Invalid engine: ' . $this->engine );
 253         }
 254
 255 }