merged master (2012-09-11)
[lhc/web/wiklou.git] / includes / Content.php
1 <?php
2 /**
3 * A content object represents page content, e.g. the text to show on a page.
4 * Content objects have no knowledge about how they relate to wiki pages.
5 *
6 * @since 1.WD
7 */
8 interface Content {
9
10 /**
11 * @since WD.1
12 *
13 * @return string A string representing the content in a way useful for
14 * building a full text search index. If no useful representation exists,
15 * this method returns an empty string.
16 *
17 * @todo: test that this actually works
18 * @todo: make sure this also works with LuceneSearch / WikiSearch
19 */
20 public function getTextForSearchIndex( );
21
22 /**
23 * @since WD.1
24 *
25 * @return string The wikitext to include when another page includes this
26 * content, or false if the content is not includable in a wikitext page.
27 *
28 * @TODO: allow native handling, bypassing wikitext representation, like
29 * for includable special pages.
30 * @TODO: allow transclusion into other content models than Wikitext!
31 * @TODO: used in WikiPage and MessageCache to get message text. Not so
32 * nice. What should we use instead?!
33 */
34 public function getWikitextForTransclusion( );
35
36 /**
37 * Returns a textual representation of the content suitable for use in edit
38 * summaries and log messages.
39 *
40 * @since WD.1
41 *
42 * @param $maxlength int Maximum length of the summary text
43 * @return The summary text
44 */
45 public function getTextForSummary( $maxlength = 250 );
46
47 /**
48 * Returns native representation of the data. Interpretation depends on
49 * the data model used, as given by getDataModel().
50 *
51 * @since WD.1
52 *
53 * @return mixed The native representation of the content. Could be a
54 * string, a nested array structure, an object, a binary blob...
55 * anything, really.
56 *
57 * @NOTE: review all calls carefully, caller must be aware of content model!
58 */
59 public function getNativeData( );
60
61 /**
62 * Returns the content's nominal size in bogo-bytes.
63 *
64 * @return int
65 */
66 public function getSize( );
67
68 /**
69 * Returns the ID of the content model used by this Content object.
70 * Corresponds to the CONTENT_MODEL_XXX constants.
71 *
72 * @since WD.1
73 *
74 * @return String The model id
75 */
76 public function getModel();
77
78 /**
79 * Convenience method that returns the ContentHandler singleton for handling
80 * the content model that this Content object uses.
81 *
82 * Shorthand for ContentHandler::getForContent( $this )
83 *
84 * @since WD.1
85 *
86 * @return ContentHandler
87 */
88 public function getContentHandler();
89
90 /**
91 * Convenience method that returns the default serialization format for the
92 * content model that this Content object uses.
93 *
94 * Shorthand for $this->getContentHandler()->getDefaultFormat()
95 *
96 * @since WD.1
97 *
98 * @return String
99 */
100 public function getDefaultFormat();
101
102 /**
103 * Convenience method that returns the list of serialization formats
104 * supported for the content model that this Content object uses.
105 *
106 * Shorthand for $this->getContentHandler()->getSupportedFormats()
107 *
108 * @since WD.1
109 *
110 * @return Array of supported serialization formats
111 */
112 public function getSupportedFormats();
113
114 /**
115 * Returns true if $format is a supported serialization format for this
116 * Content object, false if it isn't.
117 *
118 * Note that this should always return true if $format is null, because null
119 * stands for the default serialization.
120 *
121 * Shorthand for $this->getContentHandler()->isSupportedFormat( $format )
122 *
123 * @since WD.1
124 *
125 * @param $format string The format to check
126 * @return bool Whether the format is supported
127 */
128 public function isSupportedFormat( $format );
129
130 /**
131 * Convenience method for serializing this Content object.
132 *
133 * Shorthand for $this->getContentHandler()->serializeContent( $this, $format )
134 *
135 * @since WD.1
136 *
137 * @param $format null|string The desired serialization format (or null for
138 * the default format).
139 * @return string Serialized form of this Content object
140 */
141 public function serialize( $format = null );
142
143 /**
144 * Returns true if this Content object represents empty content.
145 *
146 * @since WD.1
147 *
148 * @return bool Whether this Content object is empty
149 */
150 public function isEmpty();
151
152 /**
153 * Returns whether the content is valid. This is intended for local validity
154 * checks, not considering global consistency.
155 *
156 * Content needs to be valid before it can be saved.
157 *
158 * This default implementation always returns true.
159 *
160 * @since WD.1
161 *
162 * @return boolean
163 */
164 public function isValid();
165
166 /**
167 * Returns true if this Content objects is conceptually equivalent to the
168 * given Content object.
169 *
170 * Contract:
171 *
172 * - Will return false if $that is null.
173 * - Will return true if $that === $this.
174 * - Will return false if $that->getModelName() != $this->getModel().
175 * - Will return false if $that->getNativeData() is not equal to $this->getNativeData(),
176 * where the meaning of "equal" depends on the actual data model.
177 *
178 * Implementations should be careful to make equals() transitive and reflexive:
179 *
180 * - $a->equals( $b ) <=> $b->equals( $a )
181 * - $a->equals( $b ) && $b->equals( $c ) ==> $a->equals( $c )
182 *
183 * @since WD.1
184 *
185 * @param $that Content The Content object to compare to
186 * @return bool True if this Content object is equal to $that, false otherwise.
187 */
188 public function equals( Content $that = null );
189
190 /**
191 * Return a copy of this Content object. The following must be true for the
192 * object returned:
193 *
194 * if $copy = $original->copy()
195 *
196 * - get_class($original) === get_class($copy)
197 * - $original->getModel() === $copy->getModel()
198 * - $original->equals( $copy )
199 *
200 * If and only if the Content object is immutable, the copy() method can and
201 * should return $this. That is, $copy === $original may be true, but only
202 * for immutable content objects.
203 *
204 * @since WD.1
205 *
206 * @return Content. A copy of this object
207 */
208 public function copy( );
209
210 /**
211 * Returns true if this content is countable as a "real" wiki page, provided
212 * that it's also in a countable location (e.g. a current revision in the
213 * main namespace).
214 *
215 * @since WD.1
216 *
217 * @param $hasLinks Bool: If it is known whether this content contains
218 * links, provide this information here, to avoid redundant parsing to
219 * find out.
220 * @return boolean
221 */
222 public function isCountable( $hasLinks = null ) ;
223
224
225 /**
226 * Parse the Content object and generate a ParserOutput from the result.
227 * $result->getText() can be used to obtain the generated HTML. If no HTML
228 * is needed, $generateHtml can be set to false; in that case,
229 * $result->getText() may return null.
230 *
231 * @param $title Title The page title to use as a context for rendering
232 * @param $revId null|int The revision being rendered (optional)
233 * @param $options null|ParserOptions Any parser options
234 * @param $generateHtml Boolean Whether to generate HTML (default: true). If false,
235 * the result of calling getText() on the ParserOutput object returned by
236 * this method is undefined.
237 *
238 * @since WD.1
239 *
240 * @return ParserOutput
241 */
242 public function getParserOutput( Title $title,
243 $revId = null,
244 ParserOptions $options = null, $generateHtml = true );
245 # TODO: make RenderOutput and RenderOptions base classes
246
247 /**
248 * Returns a list of DataUpdate objects for recording information about this
249 * Content in some secondary data store. If the optional second argument,
250 * $old, is given, the updates may model only the changes that need to be
251 * made to replace information about the old content with information about
252 * the new content.
253 *
254 * This default implementation calls
255 * $this->getParserOutput( $content, $title, null, null, false ),
256 * and then calls getSecondaryDataUpdates( $title, $recursive ) on the
257 * resulting ParserOutput object.
258 *
259 * Subclasses may implement this to determine the necessary updates more
260 * efficiently, or make use of information about the old content.
261 *
262 * @param $title Title The context for determining the necessary updates
263 * @param $old Content|null An optional Content object representing the
264 * previous content, i.e. the content being replaced by this Content
265 * object.
266 * @param $recursive boolean Whether to include recursive updates (default:
267 * false).
268 * @param $parserOutput ParserOutput|null Optional ParserOutput object.
269 * Provide if you have one handy, to avoid re-parsing of the content.
270 *
271 * @return Array. A list of DataUpdate objects for putting information
272 * about this content object somewhere.
273 *
274 * @since WD.1
275 */
276 public function getSecondaryDataUpdates( Title $title,
277 Content $old = null,
278 $recursive = true, ParserOutput $parserOutput = null
279 );
280
281 /**
282 * Construct the redirect destination from this content and return an
283 * array of Titles, or null if this content doesn't represent a redirect.
284 * The last element in the array is the final destination after all redirects
285 * have been resolved (up to $wgMaxRedirects times).
286 *
287 * @since WD.1
288 *
289 * @return Array of Titles, with the destination last
290 */
291 public function getRedirectChain();
292
293 /**
294 * Construct the redirect destination from this content and return a Title,
295 * or null if this content doesn't represent a redirect.
296 * This will only return the immediate redirect target, useful for
297 * the redirect table and other checks that don't need full recursion.
298 *
299 * @since WD.1
300 *
301 * @return Title: The corresponding Title
302 */
303 public function getRedirectTarget();
304
305 /**
306 * Construct the redirect destination from this content and return the
307 * Title, or null if this content doesn't represent a redirect.
308 *
309 * This will recurse down $wgMaxRedirects times or until a non-redirect
310 * target is hit in order to provide (hopefully) the Title of the final
311 * destination instead of another redirect.
312 *
313 * There is usually no need to override the default behaviour, subclasses that
314 * want to implement redirects should override getRedirectTarget().
315 *
316 * @since WD.1
317 *
318 * @return Title
319 */
320 public function getUltimateRedirectTarget();
321
322 /**
323 * Returns whether this Content represents a redirect.
324 * Shorthand for getRedirectTarget() !== null.
325 *
326 * @since WD.1
327 *
328 * @return bool
329 */
330 public function isRedirect();
331
332 /**
333 * If this Content object is a redirect, this method updates the redirect target.
334 * Otherwise, it does nothing.
335 *
336 * @since WD.1
337 *
338 * @param Title $target the new redirect target
339 *
340 * @return Content a new Content object with the updated redirect (or $this if this Content object isn't a redirect)
341 */
342 public function updateRedirect( Title $target );
343
344 /**
345 * Returns the section with the given ID.
346 *
347 * @since WD.1
348 *
349 * @param $sectionId string The section's ID, given as a numeric string.
350 * The ID "0" retrieves the section before the first heading, "1" the
351 * text between the first heading (included) and the second heading
352 * (excluded), etc.
353 * @return Content|Boolean|null The section, or false if no such section
354 * exist, or null if sections are not supported.
355 */
356 public function getSection( $sectionId );
357
358 /**
359 * Replaces a section of the content and returns a Content object with the
360 * section replaced.
361 *
362 * @since WD.1
363 *
364 * @param $section Empty/null/false or a section number (0, 1, 2, T1, T2...), or "new"
365 * @param $with Content: new content of the section
366 * @param $sectionTitle String: new section's subject, only if $section is 'new'
367 * @return string Complete article text, or null if error
368 */
369 public function replaceSection( $section, Content $with, $sectionTitle = '' );
370
371 /**
372 * Returns a Content object with pre-save transformations applied (or this
373 * object if no transformations apply).
374 *
375 * @since WD.1
376 *
377 * @param $title Title
378 * @param $user User
379 * @param $popts null|ParserOptions
380 * @return Content
381 */
382 public function preSaveTransform( Title $title, User $user, ParserOptions $popts );
383
384 /**
385 * Returns a new WikitextContent object with the given section heading
386 * prepended, if supported. The default implementation just returns this
387 * Content object unmodified, ignoring the section header.
388 *
389 * @since WD.1
390 *
391 * @param $header string
392 * @return Content
393 */
394 public function addSectionHeader( $header );
395
396 /**
397 * Returns a Content object with preload transformations applied (or this
398 * object if no transformations apply).
399 *
400 * @since WD.1
401 *
402 * @param $title Title
403 * @param $popts null|ParserOptions
404 * @return Content
405 */
406 public function preloadTransform( Title $title, ParserOptions $popts );
407
408 /**
409 * Prepare Content for saving. Called before Content is saved by WikiPage::doEditContent() and in
410 * similar places.
411 *
412 * This may be used to check the content's consistency with global state. This function should
413 * NOT write any information to the database.
414 *
415 * Note that this method will usually be called inside the same transaction bracket that will be used
416 * to save the new revision.
417 *
418 * Note that this method is called before any update to the page table is performed. This means that
419 * $page may not yet know a page ID.
420 *
421 * @param WikiPage $page The page to be saved.
422 * @param int $flags bitfield for use with EDIT_XXX constants, see WikiPage::doEditContent()
423 * @param int $baseRevId the ID of the current revision
424 * @param User $user
425 *
426 * @return Status A status object indicating whether the content was successfully prepared for saving.
427 * If the returned status indicates an error, a rollback will be performed and the
428 * transaction aborted.
429 *
430 * @see see WikiPage::doEditContent()
431 */
432 public function prepareSave( WikiPage $page, $flags, $baseRevId, User $user );
433
434 /**
435 * Returns a list of updates to perform when this content is deleted.
436 * The necessary updates may be taken from the Content object, or depend on
437 * the current state of the database.
438 *
439 * @since WD.1
440 *
441 * @param $page \WikiPage the deleted page
442 * @param $parserOutput null|\ParserOutput optional parser output object
443 * for efficient access to meta-information about the content object.
444 * Provide if you have one handy.
445 *
446 * @return array A list of DataUpdate instances that will clean up the
447 * database after deletion.
448 */
449 public function getDeletionUpdates( WikiPage $page,
450 ParserOutput $parserOutput = null );
451
452 /**
453 * Returns true if this Content object matches the given magic word.
454 *
455 * @param MagicWord $word the magic word to match
456 *
457 * @return bool whether this Content object matches the given magic word.
458 */
459 public function matchMagicWord( MagicWord $word );
460
461 # TODO: handle ImagePage and CategoryPage
462 # TODO: make sure we cover lucene search / wikisearch.
463 # TODO: make sure ReplaceTemplates still works
464 # FUTURE: nice&sane integration of GeSHi syntax highlighting
465 # [11:59] <vvv> Hooks are ugly; make CodeHighlighter interface and a
466 # config to set the class which handles syntax highlighting
467 # [12:00] <vvv> And default it to a DummyHighlighter
468
469 # TODO: make sure we cover the external editor interface (does anyone actually use that?!)
470
471 # TODO: tie into API to provide contentModel for Revisions
472 # TODO: tie into API to provide serialized version and contentFormat for Revisions
473 # TODO: tie into API edit interface
474 # FUTURE: make EditForm plugin for EditPage
475
476 # FUTURE: special type for redirects?!
477 # FUTURE: MultipartMultipart < WikipageContent (Main + Links + X)
478 # FUTURE: LinksContent < LanguageLinksContent, CategoriesContent
479 }
480
481
482 /**
483 * A content object represents page content, e.g. the text to show on a page.
484 * Content objects have no knowledge about how they relate to Wiki pages.
485 *
486 * @since 1.WD
487 */
488 abstract class AbstractContent implements Content {
489
490 /**
491 * Name of the content model this Content object represents.
492 * Use with CONTENT_MODEL_XXX constants
493 *
494 * @var string $model_id
495 */
496 protected $model_id;
497
498 /**
499 * @param String $model_id
500 */
501 public function __construct( $model_id = null ) {
502 $this->model_id = $model_id;
503 }
504
505 /**
506 * @see Content::getModel()
507 */
508 public function getModel() {
509 return $this->model_id;
510 }
511
512 /**
513 * Throws an MWException if $model_id is not the id of the content model
514 * supported by this Content object.
515 *
516 * @param $model_id int the model to check
517 *
518 * @throws MWException
519 */
520 protected function checkModelID( $model_id ) {
521 if ( $model_id !== $this->model_id ) {
522 throw new MWException( "Bad content model: " .
523 "expected {$this->model_id} " .
524 "but got $model_id." );
525 }
526 }
527
528 /**
529 * @see Content::getContentHandler()
530 */
531 public function getContentHandler() {
532 return ContentHandler::getForContent( $this );
533 }
534
535 /**
536 * @see Content::getDefaultFormat()
537 */
538 public function getDefaultFormat() {
539 return $this->getContentHandler()->getDefaultFormat();
540 }
541
542 /**
543 * @see Content::getSupportedFormats()
544 */
545 public function getSupportedFormats() {
546 return $this->getContentHandler()->getSupportedFormats();
547 }
548
549 /**
550 * @see Content::isSupportedFormat()
551 */
552 public function isSupportedFormat( $format ) {
553 if ( !$format ) {
554 return true; // this means "use the default"
555 }
556
557 return $this->getContentHandler()->isSupportedFormat( $format );
558 }
559
560 /**
561 * Throws an MWException if $this->isSupportedFormat( $format ) doesn't
562 * return true.
563 *
564 * @param $format
565 * @throws MWException
566 */
567 protected function checkFormat( $format ) {
568 if ( !$this->isSupportedFormat( $format ) ) {
569 throw new MWException( "Format $format is not supported for content model " .
570 $this->getModel() );
571 }
572 }
573
574 /**
575 * @see Content::serialize
576 */
577 public function serialize( $format = null ) {
578 return $this->getContentHandler()->serializeContent( $this, $format );
579 }
580
581 /**
582 * @see Content::isEmpty()
583 */
584 public function isEmpty() {
585 return $this->getSize() == 0;
586 }
587
588 /**
589 * @see Content::isValid()
590 */
591 public function isValid() {
592 return true;
593 }
594
595 /**
596 * @see Content::equals()
597 */
598 public function equals( Content $that = null ) {
599 if ( is_null( $that ) ) {
600 return false;
601 }
602
603 if ( $that === $this ) {
604 return true;
605 }
606
607 if ( $that->getModel() !== $this->getModel() ) {
608 return false;
609 }
610
611 return $this->getNativeData() === $that->getNativeData();
612 }
613
614
615 /**
616 * Returns a list of DataUpdate objects for recording information about this
617 * Content in some secondary data store.
618 *
619 * This default implementation calls
620 * $this->getParserOutput( $content, $title, null, null, false ),
621 * and then calls getSecondaryDataUpdates( $title, $recursive ) on the
622 * resulting ParserOutput object.
623 *
624 * Subclasses may override this to determine the secondary data updates more
625 * efficiently, preferrably without the need to generate a parser output object.
626 *
627 * @see Content::getSecondaryDataUpdates()
628 *
629 * @param $title Title The context for determining the necessary updates
630 * @param $old Content|null An optional Content object representing the
631 * previous content, i.e. the content being replaced by this Content
632 * object.
633 * @param $recursive boolean Whether to include recursive updates (default:
634 * false).
635 * @param $parserOutput ParserOutput|null Optional ParserOutput object.
636 * Provide if you have one handy, to avoid re-parsing of the content.
637 *
638 * @return Array. A list of DataUpdate objects for putting information
639 * about this content object somewhere.
640 *
641 * @since WD.1
642 */
643 public function getSecondaryDataUpdates( Title $title,
644 Content $old = null,
645 $recursive = true, ParserOutput $parserOutput = null
646 ) {
647 if ( !$parserOutput ) {
648 $parserOutput = $this->getParserOutput( $title, null, null, false );
649 }
650
651 return $parserOutput->getSecondaryDataUpdates( $title, $recursive );
652 }
653
654
655 /**
656 * @see Content::getRedirectChain()
657 */
658 public function getRedirectChain() {
659 global $wgMaxRedirects;
660 $title = $this->getRedirectTarget();
661 if ( is_null( $title ) ) {
662 return null;
663 }
664 // recursive check to follow double redirects
665 $recurse = $wgMaxRedirects;
666 $titles = array( $title );
667 while ( --$recurse > 0 ) {
668 if ( $title->isRedirect() ) {
669 $page = WikiPage::factory( $title );
670 $newtitle = $page->getRedirectTarget();
671 } else {
672 break;
673 }
674 // Redirects to some special pages are not permitted
675 if ( $newtitle instanceOf Title && $newtitle->isValidRedirectTarget() ) {
676 // The new title passes the checks, so make that our current
677 // title so that further recursion can be checked
678 $title = $newtitle;
679 $titles[] = $newtitle;
680 } else {
681 break;
682 }
683 }
684 return $titles;
685 }
686
687 /**
688 * @see Content::getRedirectTarget()
689 */
690 public function getRedirectTarget() {
691 return null;
692 }
693
694 /**
695 * @see Content::getUltimateRedirectTarget()
696 * @note: migrated here from Title::newFromRedirectRecurse
697 */
698 public function getUltimateRedirectTarget() {
699 $titles = $this->getRedirectChain();
700 return $titles ? array_pop( $titles ) : null;
701 }
702
703 /**
704 * @see Content::isRedirect()
705 *
706 * @since WD.1
707 *
708 * @return bool
709 */
710 public function isRedirect() {
711 return $this->getRedirectTarget() !== null;
712 }
713
714 /**
715 * @see Content::updateRedirect()
716 *
717 * This default implementation always returns $this.
718 *
719 * @since WD.1
720 *
721 * @return Content $this
722 */
723 public function updateRedirect( Title $target ) {
724 return $this;
725 }
726
727 /**
728 * @see Content::getSection()
729 */
730 public function getSection( $sectionId ) {
731 return null;
732 }
733
734 /**
735 * @see Content::replaceSection()
736 */
737 public function replaceSection( $section, Content $with, $sectionTitle = '' ) {
738 return null;
739 }
740
741 /**
742 * @see Content::preSaveTransform()
743 */
744 public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) {
745 return $this;
746 }
747
748 /**
749 * @see Content::addSectionHeader()
750 */
751 public function addSectionHeader( $header ) {
752 return $this;
753 }
754
755 /**
756 * @see Content::preloadTransform()
757 */
758 public function preloadTransform( Title $title, ParserOptions $popts ) {
759 return $this;
760 }
761
762 /**
763 * @see Content::prepareSave()
764 */
765 public function prepareSave( WikiPage $page, $flags, $baseRevId, User $user ) {
766 if ( $this->isValid() ) {
767 return Status::newGood();
768 } else {
769 return Status::newFatal( "invalid-content-data" );
770 }
771 }
772
773 /**
774 * @see Content::getDeletionUpdates()
775 *
776 * @since WD.1
777 *
778 * @param $page \WikiPage the deleted page
779 * @param $parserOutput null|\ParserOutput optional parser output object
780 * for efficient access to meta-information about the content object.
781 * Provide if you have one handy.
782 *
783 * @return array A list of DataUpdate instances that will clean up the
784 * database after deletion.
785 */
786 public function getDeletionUpdates( WikiPage $page,
787 ParserOutput $parserOutput = null )
788 {
789 return array(
790 new LinksDeletionUpdate( $page ),
791 );
792 }
793
794 /**
795 * @see Content::matchMagicWord()
796 *
797 * This default implementation always returns false. Subclasses may override this to supply matching logic.
798 *
799 * @param MagicWord $word
800 *
801 * @return bool
802 */
803 public function matchMagicWord( MagicWord $word ) {
804 return false;
805 }
806 }
807
808 /**
809 * Content object implementation for representing flat text.
810 *
811 * TextContent instances are immutable
812 *
813 * @since WD.1
814 */
815 abstract class TextContent extends AbstractContent {
816
817 public function __construct( $text, $model_id = null ) {
818 parent::__construct( $model_id );
819
820 $this->mText = $text;
821 }
822
823 public function copy() {
824 return $this; # NOTE: this is ok since TextContent are immutable.
825 }
826
827 public function getTextForSummary( $maxlength = 250 ) {
828 global $wgContLang;
829
830 $text = $this->getNativeData();
831
832 $truncatedtext = $wgContLang->truncate(
833 preg_replace( "/[\n\r]/", ' ', $text ),
834 max( 0, $maxlength ) );
835
836 return $truncatedtext;
837 }
838
839 /**
840 * returns the text's size in bytes.
841 *
842 * @return int The size
843 */
844 public function getSize( ) {
845 $text = $this->getNativeData( );
846 return strlen( $text );
847 }
848
849 /**
850 * Returns true if this content is not a redirect, and $wgArticleCountMethod
851 * is "any".
852 *
853 * @param $hasLinks Bool: if it is known whether this content contains links,
854 * provide this information here, to avoid redundant parsing to find out.
855 *
856 * @return bool True if the content is countable
857 */
858 public function isCountable( $hasLinks = null ) {
859 global $wgArticleCountMethod;
860
861 if ( $this->isRedirect( ) ) {
862 return false;
863 }
864
865 if ( $wgArticleCountMethod === 'any' ) {
866 return true;
867 }
868
869 return false;
870 }
871
872 /**
873 * Returns the text represented by this Content object, as a string.
874 *
875 * @param the raw text
876 */
877 public function getNativeData( ) {
878 $text = $this->mText;
879 return $text;
880 }
881
882 /**
883 * Returns the text represented by this Content object, as a string.
884 *
885 * @param the raw text
886 */
887 public function getTextForSearchIndex( ) {
888 return $this->getNativeData();
889 }
890
891 /**
892 * Returns the text represented by this Content object, as a string.
893 *
894 * @param the raw text
895 */
896 public function getWikitextForTransclusion( ) {
897 return $this->getNativeData();
898 }
899
900 /**
901 * Diff this content object with another content object..
902 *
903 * @since WD.diff
904 *
905 * @param $that Content the other content object to compare this content object to
906 * @param $lang Language the language object to use for text segmentation.
907 * If not given, $wgContentLang is used.
908 *
909 * @return DiffResult a diff representing the changes that would have to be
910 * made to this content object to make it equal to $that.
911 */
912 public function diff( Content $that, Language $lang = null ) {
913 global $wgContLang;
914
915 $this->checkModelID( $that->getModel() );
916
917 # @todo: could implement this in DifferenceEngine and just delegate here?
918
919 if ( !$lang ) $lang = $wgContLang;
920
921 $otext = $this->getNativeData();
922 $ntext = $this->getNativeData();
923
924 # Note: Use native PHP diff, external engines don't give us abstract output
925 $ota = explode( "\n", $wgContLang->segmentForDiff( $otext ) );
926 $nta = explode( "\n", $wgContLang->segmentForDiff( $ntext ) );
927
928 $diff = new Diff( $ota, $nta );
929 return $diff;
930 }
931
932
933 /**
934 * Returns a generic ParserOutput object, wrapping the HTML returned by
935 * getHtml().
936 *
937 * @param $title Title Context title for parsing
938 * @param $revId int|null Revision ID (for {{REVISIONID}})
939 * @param $options ParserOptions|null Parser options
940 * @param $generateHtml bool Whether or not to generate HTML
941 *
942 * @return ParserOutput representing the HTML form of the text
943 */
944 public function getParserOutput( Title $title,
945 $revId = null,
946 ParserOptions $options = null, $generateHtml = true
947 ) {
948 # Generic implementation, relying on $this->getHtml()
949
950 if ( $generateHtml ) {
951 $html = $this->getHtml();
952 } else {
953 $html = '';
954 }
955
956 $po = new ParserOutput( $html );
957 return $po;
958 }
959
960 /**
961 * Generates an HTML version of the content, for display. Used by
962 * getParserOutput() to construct a ParserOutput object.
963 *
964 * This default implementation just calls getHighlightHtml(). Content
965 * models that have another mapping to HTML (as is the case for markup
966 * languages like wikitext) should override this method to generate the
967 * appropriate HTML.
968 *
969 * @return string An HTML representation of the content
970 */
971 protected function getHtml() {
972 return $this->getHighlightHtml();
973 }
974
975 /**
976 * Generates a syntax-highlighted version of the content, as HTML.
977 * Used by the default implementation of getHtml().
978 *
979 * @return string an HTML representation of the content's markup
980 */
981 protected function getHighlightHtml( ) {
982 # TODO: make Highlighter interface, use highlighter here, if available
983 return htmlspecialchars( $this->getNativeData() );
984 }
985 }
986
987 /**
988 * @since WD.1
989 */
990 class WikitextContent extends TextContent {
991
992 public function __construct( $text ) {
993 parent::__construct( $text, CONTENT_MODEL_WIKITEXT );
994 }
995
996 /**
997 * @see Content::getSection()
998 */
999 public function getSection( $section ) {
1000 global $wgParser;
1001
1002 $text = $this->getNativeData();
1003 $sect = $wgParser->getSection( $text, $section, false );
1004
1005 return new WikitextContent( $sect );
1006 }
1007
1008 /**
1009 * @see Content::replaceSection()
1010 */
1011 public function replaceSection( $section, Content $with, $sectionTitle = '' ) {
1012 wfProfileIn( __METHOD__ );
1013
1014 $myModelId = $this->getModel();
1015 $sectionModelId = $with->getModel();
1016
1017 if ( $sectionModelId != $myModelId ) {
1018 throw new MWException( "Incompatible content model for section: " .
1019 "document uses $myModelId but " .
1020 "section uses $sectionModelId." );
1021 }
1022
1023 $oldtext = $this->getNativeData();
1024 $text = $with->getNativeData();
1025
1026 if ( $section === '' ) {
1027 return $with; # XXX: copy first?
1028 } if ( $section == 'new' ) {
1029 # Inserting a new section
1030 $subject = $sectionTitle ? wfMessage( 'newsectionheaderdefaultlevel' )
1031 ->rawParams( $sectionTitle )->inContentLanguage()->text() . "\n\n" : '';
1032 if ( wfRunHooks( 'PlaceNewSection', array( $this, $oldtext, $subject, &$text ) ) ) {
1033 $text = strlen( trim( $oldtext ) ) > 0
1034 ? "{$oldtext}\n\n{$subject}{$text}"
1035 : "{$subject}{$text}";
1036 }
1037 } else {
1038 # Replacing an existing section; roll out the big guns
1039 global $wgParser;
1040
1041 $text = $wgParser->replaceSection( $oldtext, $section, $text );
1042 }
1043
1044 $newContent = new WikitextContent( $text );
1045
1046 wfProfileOut( __METHOD__ );
1047 return $newContent;
1048 }
1049
1050 /**
1051 * Returns a new WikitextContent object with the given section heading
1052 * prepended.
1053 *
1054 * @param $header string
1055 * @return Content
1056 */
1057 public function addSectionHeader( $header ) {
1058 $text = wfMessage( 'newsectionheaderdefaultlevel' )
1059 ->inContentLanguage()->params( $header )->text();
1060 $text .= "\n\n";
1061 $text .= $this->getNativeData();
1062
1063 return new WikitextContent( $text );
1064 }
1065
1066 /**
1067 * Returns a Content object with pre-save transformations applied using
1068 * Parser::preSaveTransform().
1069 *
1070 * @param $title Title
1071 * @param $user User
1072 * @param $popts ParserOptions
1073 * @return Content
1074 */
1075 public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) {
1076 global $wgParser;
1077
1078 $text = $this->getNativeData();
1079 $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts );
1080
1081 return new WikitextContent( $pst );
1082 }
1083
1084 /**
1085 * Returns a Content object with preload transformations applied (or this
1086 * object if no transformations apply).
1087 *
1088 * @param $title Title
1089 * @param $popts ParserOptions
1090 * @return Content
1091 */
1092 public function preloadTransform( Title $title, ParserOptions $popts ) {
1093 global $wgParser;
1094
1095 $text = $this->getNativeData();
1096 $plt = $wgParser->getPreloadText( $text, $title, $popts );
1097
1098 return new WikitextContent( $plt );
1099 }
1100
1101 /**
1102 * Implement redirect extraction for wikitext.
1103 *
1104 * @return null|Title
1105 *
1106 * @note: migrated here from Title::newFromRedirectInternal()
1107 *
1108 * @see Content::getRedirectTarget
1109 * @see AbstractContent::getRedirectTarget
1110 */
1111 public function getRedirectTarget() {
1112 global $wgMaxRedirects;
1113 if ( $wgMaxRedirects < 1 ) {
1114 // redirects are disabled, so quit early
1115 return null;
1116 }
1117 $redir = MagicWord::get( 'redirect' );
1118 $text = trim( $this->getNativeData() );
1119 if ( $redir->matchStartAndRemove( $text ) ) {
1120 // Extract the first link and see if it's usable
1121 // Ensure that it really does come directly after #REDIRECT
1122 // Some older redirects included a colon, so don't freak about that!
1123 $m = array();
1124 if ( preg_match( '!^\s*:?\s*\[{2}(.*?)(?:\|.*?)?\]{2}!', $text, $m ) ) {
1125 // Strip preceding colon used to "escape" categories, etc.
1126 // and URL-decode links
1127 if ( strpos( $m[1], '%' ) !== false ) {
1128 // Match behavior of inline link parsing here;
1129 $m[1] = rawurldecode( ltrim( $m[1], ':' ) );
1130 }
1131 $title = Title::newFromText( $m[1] );
1132 // If the title is a redirect to bad special pages or is invalid, return null
1133 if ( !$title instanceof Title || !$title->isValidRedirectTarget() ) {
1134 return null;
1135 }
1136 return $title;
1137 }
1138 }
1139 return null;
1140 }
1141
1142 /**
1143 * @see Content::updateRedirect()
1144 *
1145 * This implementation replaces the first link on the page with the given new target
1146 * if this Content object is a redirect. Otherwise, this method returns $this.
1147 *
1148 * @since WD.1
1149 *
1150 * @param Title $target
1151 *
1152 * @return Content a new Content object with the updated redirect (or $this if this Content object isn't a redirect)
1153 */
1154 public function updateRedirect( Title $target ) {
1155 if ( !$this->isRedirect() ) {
1156 return $this;
1157 }
1158
1159 # Fix the text
1160 # Remember that redirect pages can have categories, templates, etc.,
1161 # so the regex has to be fairly general
1162 $newText = preg_replace( '/ \[ \[ [^\]]* \] \] /x',
1163 '[[' . $target->getFullText() . ']]',
1164 $this->getNativeData(), 1 );
1165
1166 return new WikitextContent( $newText );
1167 }
1168
1169 /**
1170 * Returns true if this content is not a redirect, and this content's text
1171 * is countable according to the criteria defined by $wgArticleCountMethod.
1172 *
1173 * @param $hasLinks Bool if it is known whether this content contains
1174 * links, provide this information here, to avoid redundant parsing to
1175 * find out.
1176 * @param $title null|\Title
1177 *
1178 * @internal param \IContextSource $context context for parsing if necessary
1179 *
1180 * @return bool True if the content is countable
1181 */
1182 public function isCountable( $hasLinks = null, Title $title = null ) {
1183 global $wgArticleCountMethod;
1184
1185 if ( $this->isRedirect( ) ) {
1186 return false;
1187 }
1188
1189 $text = $this->getNativeData();
1190
1191 switch ( $wgArticleCountMethod ) {
1192 case 'any':
1193 return true;
1194 case 'comma':
1195 return strpos( $text, ',' ) !== false;
1196 case 'link':
1197 if ( $hasLinks === null ) { # not known, find out
1198 if ( !$title ) {
1199 $context = RequestContext::getMain();
1200 $title = $context->getTitle();
1201 }
1202
1203 $po = $this->getParserOutput( $title, null, null, false );
1204 $links = $po->getLinks();
1205 $hasLinks = !empty( $links );
1206 }
1207
1208 return $hasLinks;
1209 }
1210
1211 return false;
1212 }
1213
1214 public function getTextForSummary( $maxlength = 250 ) {
1215 $truncatedtext = parent::getTextForSummary( $maxlength );
1216
1217 # clean up unfinished links
1218 # XXX: make this optional? wasn't there in autosummary, but required for
1219 # deletion summary.
1220 $truncatedtext = preg_replace( '/\[\[([^\]]*)\]?$/', '$1', $truncatedtext );
1221
1222 return $truncatedtext;
1223 }
1224
1225
1226 /**
1227 * Returns a ParserOutput object resulting from parsing the content's text
1228 * using $wgParser.
1229 *
1230 * @since WD.1
1231 *
1232 * @param $content Content the content to render
1233 * @param $title \Title
1234 * @param $revId null
1235 * @param $options null|ParserOptions
1236 * @param $generateHtml bool
1237 *
1238 * @internal param \IContextSource|null $context
1239 * @return ParserOutput representing the HTML form of the text
1240 */
1241 public function getParserOutput( Title $title,
1242 $revId = null,
1243 ParserOptions $options = null, $generateHtml = true
1244 ) {
1245 global $wgParser;
1246
1247 if ( !$options ) {
1248 $options = new ParserOptions();
1249 }
1250
1251 $po = $wgParser->parse( $this->getNativeData(), $title, $options, true, true, $revId );
1252 return $po;
1253 }
1254
1255 protected function getHtml() {
1256 throw new MWException(
1257 "getHtml() not implemented for wikitext. "
1258 . "Use getParserOutput()->getText()."
1259 );
1260 }
1261
1262 /**
1263 * @see Content::matchMagicWord()
1264 *
1265 * This implementation calls $word->match() on the this TextContent object's text.
1266 *
1267 * @param MagicWord $word
1268 *
1269 * @return bool whether this Content object matches the given magic word.
1270 */
1271 public function matchMagicWord( MagicWord $word ) {
1272 return $word->match( $this->getNativeData() );
1273 }
1274 }
1275
1276 /**
1277 * Wrapper allowing us to handle a system message as a Content object. Note that this is generally *not* used
1278 * to represent content from the MediaWiki namespace, and that there is no MessageContentHandler. MessageContent
1279 * is just intended as glue for wrapping a message programatically.
1280 *
1281 * @since WD.1
1282 */
1283 class MessageContent extends AbstractContent {
1284
1285 /**
1286 * @var Message
1287 */
1288 protected $mMessage;
1289
1290 /**
1291 * @param Message|String $msg A Message object, or a message key
1292 * @param array|null $params An optional array of message parameters
1293 */
1294 public function __construct( $msg, $params = null ) {
1295 # XXX: messages may be wikitext, html or plain text! and maybe even something else entirely.
1296 parent::__construct( CONTENT_MODEL_WIKITEXT );
1297
1298 if ( is_string( $msg ) ) {
1299 $this->mMessage = wfMessage( $msg );
1300 } else {
1301 $this->mMessage = clone $msg;
1302 }
1303
1304 if ( $params ) {
1305 $this->mMessage = $this->mMessage->params( $params );
1306 }
1307 }
1308
1309 /**
1310 * Returns the message as rendered HTML
1311 *
1312 * @return string The message text, parsed into html
1313 */
1314 public function getHtml() {
1315 return $this->mMessage->parse();
1316 }
1317
1318 /**
1319 * Returns the message as rendered HTML
1320 *
1321 * @return string The message text, parsed into html
1322 */
1323 public function getWikitext() {
1324 return $this->mMessage->text();
1325 }
1326
1327 /**
1328 * Returns the message object, with any parameters already substituted.
1329 *
1330 * @return Message The message object.
1331 */
1332 public function getNativeData() {
1333 //NOTE: Message objects are mutable. Cloning here makes MessageContent immutable.
1334 return clone $this->mMessage;
1335 }
1336
1337 /**
1338 * @see Content::getTextForSearchIndex
1339 */
1340 public function getTextForSearchIndex() {
1341 return $this->mMessage->plain();
1342 }
1343
1344 /**
1345 * @see Content::getWikitextForTransclusion
1346 */
1347 public function getWikitextForTransclusion() {
1348 return $this->getWikitext();
1349 }
1350
1351 /**
1352 * @see Content::getTextForSummary
1353 */
1354 public function getTextForSummary( $maxlength = 250 ) {
1355 return substr( $this->mMessage->plain(), 0, $maxlength );
1356 }
1357
1358 /**
1359 * @see Content::getSize
1360 *
1361 * @return int
1362 */
1363 public function getSize() {
1364 return strlen( $this->mMessage->plain() );
1365 }
1366
1367 /**
1368 * @see Content::copy
1369 *
1370 * @return Content. A copy of this object
1371 */
1372 public function copy() {
1373 // MessageContent is immutable (because getNativeData() returns a clone of the Message object)
1374 return $this;
1375 }
1376
1377 /**
1378 * @see Content::isCountable
1379 *
1380 * @return bool false
1381 */
1382 public function isCountable( $hasLinks = null ) {
1383 return false;
1384 }
1385
1386 /**
1387 * @see Content::getParserOutput
1388 *
1389 * @return ParserOutput
1390 */
1391 public function getParserOutput(
1392 Title $title, $revId = null,
1393 ParserOptions $options = null, $generateHtml = true
1394 ) {
1395
1396 if ( $generateHtml ) {
1397 $html = $this->getHtml();
1398 } else {
1399 $html = '';
1400 }
1401
1402 $po = new ParserOutput( $html );
1403 return $po;
1404 }
1405 }
1406
1407 /**
1408 * @since WD.1
1409 */
1410 class JavaScriptContent extends TextContent {
1411 public function __construct( $text ) {
1412 parent::__construct( $text, CONTENT_MODEL_JAVASCRIPT );
1413 }
1414
1415 /**
1416 * Returns a Content object with pre-save transformations applied using
1417 * Parser::preSaveTransform().
1418 *
1419 * @param Title $title
1420 * @param User $user
1421 * @param ParserOptions $popts
1422 * @return Content
1423 */
1424 public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) {
1425 global $wgParser;
1426 // @todo: make pre-save transformation optional for script pages
1427 // See bug #32858
1428
1429 $text = $this->getNativeData();
1430 $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts );
1431
1432 return new JavaScriptContent( $pst );
1433 }
1434
1435
1436 protected function getHtml( ) {
1437 $html = "";
1438 $html .= "<pre class=\"mw-code mw-js\" dir=\"ltr\">\n";
1439 $html .= $this->getHighlightHtml( );
1440 $html .= "\n</pre>\n";
1441
1442 return $html;
1443 }
1444 }
1445
1446 /**
1447 * @since WD.1
1448 */
1449 class CssContent extends TextContent {
1450 public function __construct( $text ) {
1451 parent::__construct( $text, CONTENT_MODEL_CSS );
1452 }
1453
1454 /**
1455 * Returns a Content object with pre-save transformations applied using
1456 * Parser::preSaveTransform().
1457 *
1458 * @param $title Title
1459 * @param $user User
1460 * @param $popts ParserOptions
1461 * @return Content
1462 */
1463 public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) {
1464 global $wgParser;
1465 // @todo: make pre-save transformation optional for script pages
1466
1467 $text = $this->getNativeData();
1468 $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts );
1469
1470 return new CssContent( $pst );
1471 }
1472
1473
1474 protected function getHtml( ) {
1475 $html = "";
1476 $html .= "<pre class=\"mw-code mw-css\" dir=\"ltr\">\n";
1477 $html .= $this->getHighlightHtml( );
1478 $html .= "\n</pre>\n";
1479
1480 return $html;
1481 }
1482 }