3 * Renders a slot diff by doing a text diff on the native representation.
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.
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.
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
21 * @ingroup DifferenceEngine
24 use MediaWiki\Shell\Shell
;
25 use Wikimedia\Assert\Assert
;
28 * Renders a slot diff by doing a text diff on the native representation.
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() )
35 * @ingroup DifferenceEngine
37 class TextSlotDiffRenderer
extends SlotDiffRenderer
{
39 /** Use the PHP diff implementation (DiffEngine). */
40 const ENGINE_PHP
= 'php';
42 /** Use the wikidiff2 PHP module. */
43 const ENGINE_WIKIDIFF2
= 'wikidiff2';
45 /** Use an external executable. */
46 const ENGINE_EXTERNAL
= 'external';
48 /** @var IBufferingStatsdDataFactory|null */
49 private $statsdDataFactory;
51 /** @var Language|null The language this content is in. */
55 * Number of paragraph moves the algorithm should attempt to detect.
56 * Only used with the wikidiff2 engine.
58 * @see $wgWikiDiff2MovedParagraphDetectionCutoff
60 private $wikiDiff2MovedParagraphDetectionCutoff = 0;
62 /** @var string One of the ENGINE_* constants. */
63 private $engine = self
::ENGINE_PHP
;
65 /** @var string Path to an executable to be used as the diff engine. */
66 private $externalEngine;
69 * Convenience helper to use getTextDiff without an instance.
70 * @param string $oldText
71 * @param string $newText
74 public static function diff( $oldText, $newText ) {
75 /** @var $slotDiffRenderer TextSlotDiffRenderer */
76 $slotDiffRenderer = ContentHandler
::getForModelID( CONTENT_MODEL_TEXT
)
77 ->getSlotDiffRenderer( RequestContext
::getMain() );
78 return $slotDiffRenderer->getTextDiff( $oldText, $newText );
81 public function setStatsdDataFactory( IBufferingStatsdDataFactory
$statsdDataFactory ) {
82 $this->statsdDataFactory
= $statsdDataFactory;
85 public function setLanguage( Language
$language ) {
86 $this->language
= $language;
90 * @see $wgWikiDiff2MovedParagraphDetectionCutoff
92 public function setWikiDiff2MovedParagraphDetectionCutoff( $cutoff ) {
93 Assert
::parameterType( 'integer', $cutoff, '$cutoff' );
94 $this->wikiDiff2MovedParagraphDetectionCutoff
= $cutoff;
98 * Set which diff engine to use.
99 * @param string $type One of the ENGINE_* constants.
100 * @param string|null $executable Path to an external exectable, only when type is ENGINE_EXTERNAL.
102 public function setEngine( $type, $executable = null ) {
103 $engines = [ self
::ENGINE_PHP
, self
::ENGINE_WIKIDIFF2
, self
::ENGINE_EXTERNAL
];
104 Assert
::parameter( in_array( $type, $engines, true ), '$type',
105 'must be one of the TextSlotDiffRenderer::ENGINE_* constants' );
106 if ( $type === self
::ENGINE_EXTERNAL
) {
107 Assert
::parameter( is_string( $executable ) && is_executable( $executable ), '$executable',
108 'must be a path to a valid executable' );
110 Assert
::parameter( is_null( $executable ), '$executable',
111 'must not be set unless $type is ENGINE_EXTERNAL' );
113 $this->engine
= $type;
114 $this->externalEngine
= $executable;
118 public function getDiff( Content
$oldContent = null, Content
$newContent = null ) {
119 if ( !$oldContent && !$newContent ) {
120 throw new InvalidArgumentException( '$oldContent and $newContent cannot both be null' );
121 } elseif ( $oldContent && !( $oldContent instanceof TextContent
) ) {
122 throw new InvalidArgumentException( __CLASS__
. ' does not handle ' . get_class( $oldContent ) );
123 } elseif ( $newContent && !( $newContent instanceof TextContent
) ) {
124 throw new InvalidArgumentException( __CLASS__
. ' does not handle ' . get_class( $newContent ) );
127 if ( !$oldContent ) {
128 $oldContent = $newContent->getContentHandler()->makeEmptyContent();
129 } elseif ( !$newContent ) {
130 $newContent = $oldContent->getContentHandler()->makeEmptyContent();
133 $oldText = $oldContent->serialize();
134 $newText = $newContent->serialize();
136 return $this->getTextDiff( $oldText, $newText );
140 * Diff the text representations of two content objects (or just two pieces of text in general).
141 * @param string $oldText
142 * @param string $newText
145 public function getTextDiff( $oldText, $newText ) {
146 Assert
::parameterType( 'string', $oldText, '$oldText' );
147 Assert
::parameterType( 'string', $newText, '$newText' );
149 $diff = function () use ( $oldText, $newText ) {
150 $time = microtime( true );
152 $result = $this->getTextDiffInternal( $oldText, $newText );
154 $time = intval( ( microtime( true ) - $time ) * 1000 );
155 if ( $this->statsdDataFactory
) {
156 $this->statsdDataFactory
->timing( 'diff_time', $time );
159 // TODO reimplement this using T142313
161 // Log requests slower than 99th percentile
162 if ( $time > 100 && $this->mOldPage && $this->mNewPage ) {
164 "$time ms diff: {$this->mOldid} -> {$this->mNewid} {$this->mNewPage}" );
172 * @param Status $status
175 $error = function ( $status ) {
176 throw new FatalError( $status->getWikiText() );
179 // Use PoolCounter if the diff looks like it can be expensive
180 if ( strlen( $oldText ) +
strlen( $newText ) > 20000 ) {
181 $work = new PoolCounterWorkViaCallback( 'diff',
182 md5( $oldText ) . md5( $newText ),
183 [ 'doWork' => $diff, 'error' => $error ]
185 return $work->execute();
192 * Diff the text representations of two content objects (or just two pieces of text in general).
193 * This does the actual diffing, getTextDiff() wraps it with logging and resource limiting.
194 * @param string $oldText
195 * @param string $newText
199 protected function getTextDiffInternal( $oldText, $newText ) {
200 // TODO move most of this into three parallel implementations of a text diff generator
201 // class, choose which one to use via dependecy injection
203 $oldText = str_replace( "\r\n", "\n", $oldText );
204 $newText = str_replace( "\r\n", "\n", $newText );
206 // Better external diff engine, the 2 may some day be dropped
207 // This one does the escaping and segmenting itself
208 if ( $this->engine
=== self
::ENGINE_WIKIDIFF2
) {
209 $wikidiff2Version = phpversion( 'wikidiff2' );
211 $wikidiff2Version !== false &&
212 version_compare( $wikidiff2Version, '1.5.0', '>=' ) &&
213 version_compare( $wikidiff2Version, '1.8.0', '<' )
215 $text = wikidiff2_do_diff(
219 $this->wikiDiff2MovedParagraphDetectionCutoff
222 // Don't pass the 4th parameter introduced in version 1.5.0 and removed in version 1.8.0
223 $text = wikidiff2_do_diff(
229 // Log a warning in case the configuration value is set to not silently ignore it
230 if ( $this->wikiDiff2MovedParagraphDetectionCutoff
> 0 ) {
231 wfLogWarning( '$wgWikiDiff2MovedParagraphDetectionCutoff is set but has no
232 effect since the used version of WikiDiff2 does not support it.' );
237 } elseif ( $this->engine
=== self
::ENGINE_EXTERNAL
) {
239 $tmpDir = wfTempDir();
240 $tempName1 = tempnam( $tmpDir, 'diff_' );
241 $tempName2 = tempnam( $tmpDir, 'diff_' );
243 $tempFile1 = fopen( $tempName1, "w" );
247 $tempFile2 = fopen( $tempName2, "w" );
251 fwrite( $tempFile1, $oldText );
252 fwrite( $tempFile2, $newText );
253 fclose( $tempFile1 );
254 fclose( $tempFile2 );
255 $cmd = [ $this->externalEngine
, $tempName1, $tempName2 ];
256 $result = Shell
::command( $cmd )
258 $exitCode = $result->getExitCode();
259 if ( $exitCode !== 0 ) {
260 throw new Exception( "External diff command returned code {$exitCode}. Stderr: "
261 . wfEscapeWikiText( $result->getStderr() )
264 $difftext = $result->getStdout();
265 unlink( $tempName1 );
266 unlink( $tempName2 );
269 } elseif ( $this->engine
=== self
::ENGINE_PHP
) {
270 if ( $this->language
) {
271 $oldText = $this->language
->segmentForDiff( $oldText );
272 $newText = $this->language
->segmentForDiff( $newText );
274 $ota = explode( "\n", $oldText );
275 $nta = explode( "\n", $newText );
276 $diffs = new Diff( $ota, $nta );
277 $formatter = new TableDiffFormatter();
278 $difftext = $formatter->format( $diffs );
279 if ( $this->language
) {
280 $difftext = $this->language
->unsegmentForDiff( $difftext );
285 throw new LogicException( 'Invalid engine: ' . $this->engine
);