4 * A content object represents page content, e.g. the text to show on a page.
5 * Content objects have no knowledge about how they relate to Wiki pages.
9 abstract class Content
{
12 * Name of the content model this Content object represents.
13 * Use with CONTENT_MODEL_XXX constants
15 * @var String $model_id
22 * @return String a string representing the content in a way useful for building a full text search index.
23 * If no useful representation exists, this method returns an empty string.
25 public abstract function getTextForSearchIndex( );
30 * @return String the wikitext to include when another page includes this content, or false if the content is not
31 * includable in a wikitext page.
33 * @TODO: allow native handling, bypassing wikitext representation, like for includable special pages.
34 * @TODO: use in parser, etc!
36 public abstract function getWikitextForTransclusion( );
39 * Returns a textual representation of the content suitable for use in edit summaries and log messages.
43 * @param int $maxlength maximum length of the summary text
44 * @return String the summary text
46 public abstract function getTextForSummary( $maxlength = 250 );
49 * Returns native represenation of the data. Interpretation depends on the data model used,
50 * as given by getDataModel().
54 * @return mixed the native representation of the content. Could be a string, a nested array
55 * structure, an object, a binary blob... anything, really.
57 * @NOTE: review all calls carefully, caller must be aware of content model!
59 public abstract function getNativeData( );
62 * returns the content's nominal size in bogo-bytes.
66 public abstract function getSize( );
69 * @param int $model_id
71 public function __construct( $model_id = null ) {
72 $this->model_id
= $model_id;
76 * Returns the id of the content model used by this content objects.
77 * Corresponds to the CONTENT_MODEL_XXX constants.
81 * @return int the model id
83 public function getModel() {
84 return $this->model_id
;
88 * Throws an MWException if $model_id is not the id of the content model
89 * supported by this Content object.
91 * @param int $model_id the model to check
93 protected function checkModelID( $model_id ) {
94 if ( $model_id !== $this->model_id
) {
95 $model_name = ContentHandler
::getContentModelName( $model_id );
96 $own_model_name = ContentHandler
::getContentModelName( $this->model_id
);
98 throw new MWException( "Bad content model: expected {$this->model_id} ($own_model_name) but got found $model_id ($model_name)." );
103 * Conveniance method that returns the ContentHandler singleton for handling the content
104 * model this Content object uses.
106 * Shorthand for ContentHandler::getForContent( $this )
110 * @return ContentHandler
112 public function getContentHandler() {
113 return ContentHandler
::getForContent( $this );
117 * Conveniance method that returns the default serialization format for the content model
118 * model this Content object uses.
120 * Shorthand for $this->getContentHandler()->getDefaultFormat()
124 * @return ContentHandler
126 public function getDefaultFormat() {
127 return $this->getContentHandler()->getDefaultFormat();
131 * Conveniance method that returns the list of serialization formats supported
132 * for the content model model this Content object uses.
134 * Shorthand for $this->getContentHandler()->getSupportedFormats()
138 * @return array of supported serialization formats
140 public function getSupportedFormats() {
141 return $this->getContentHandler()->getSupportedFormats();
145 * Returns true if $format is a supported serialization format for this Content object,
148 * Note that this will always return true if $format is null, because null stands for the
149 * default serialization.
151 * Shorthand for $this->getContentHandler()->isSupportedFormat( $format )
155 * @param String $format the format to check
156 * @return bool whether the format is supported
158 public function isSupportedFormat( $format ) {
160 return true; // this means "use the default"
163 return $this->getContentHandler()->isSupportedFormat( $format );
167 * Throws an MWException if $this->isSupportedFormat( $format ) doesn't return true.
170 * @throws MWException
172 protected function checkFormat( $format ) {
173 if ( !$this->isSupportedFormat( $format ) ) {
174 throw new MWException( "Format $format is not supported for content model " . $this->getModel() );
179 * Conveniance method for serializing this Content object.
181 * Shorthand for $this->getContentHandler()->serializeContent( $this, $format )
185 * @param null|String $format the desired serialization format (or null for the default format).
186 * @return String serialized form of this Content object
188 public function serialize( $format = null ) {
189 return $this->getContentHandler()->serializeContent( $this, $format );
193 * Returns true if this Content object represents empty content.
197 * @return bool whether this Content object is empty
199 public function isEmpty() {
200 return $this->getSize() == 0;
204 * Returns if the content is valid. This is intended for local validity checks, not considering global consistency.
205 * It needs to be valid before it can be saved.
207 * This default implementation always returns true.
213 public function isValid() {
218 * Diff this content object with another content object..
224 public abstract function diff( Content
$that );
227 * Returns true if this Content objects is conceptually equivalent to the given Content object.
229 * Will returns false if $that is null.
230 * Will return true if $that === $this.
231 * Will return false if $that->getModleName() != $this->getModel().
232 * Will return false if $that->getNativeData() is not equal to $this->getNativeData(),
233 * where the meaning of "equal" depends on the actual data model.
235 * Implementations should be careful to make equals() transitive and reflexive:
237 * * $a->equals( $b ) <=> $b->equals( $b )
238 * * $a->equals( $b ) && $b->equals( $c ) ==> $a->equals( $c )
242 * @param Content $that the Content object to compare to
243 * @return bool true if this Content object is euqual to $that, false otherwise.
245 public function equals( Content
$that = null ) {
246 if ( is_null( $that ) ){
250 if ( $that === $this ) {
254 if ( $that->getModel() !== $this->getModel() ) {
258 return $this->getNativeData() === $that->getNativeData();
262 * Return a copy of this Content object. The following must be true for the object returned
263 * if $copy = $original->copy()
265 * * get_class($original) === get_class($copy)
266 * * $original->getModel() === $copy->getModel()
267 * * $original->equals( $copy )
269 * If and only if the Content object is imutable, the copy() method can and should
270 * return $this. That is, $copy === $original may be true, but only for imutable content
275 * @return Content. A copy of this object
277 public abstract function copy( );
280 * Returns true if this content is countable as a "real" wiki page, provided
281 * that it's also in a countable location (e.g. a current revision in the main namespace).
285 * @param $hasLinks Bool: if it is known whether this content contains links, provide this information here,
286 * to avoid redundant parsing to find out.
289 public abstract function isCountable( $hasLinks = null ) ;
292 * @param Title $title
294 * @param null|ParserOptions $options
295 * @param Boolean $generateHtml whether to generate Html (default: true). If false,
296 * the result of calling getText() on the ParserOutput object returned by
297 * this method is undefined.
301 * @return ParserOutput
303 public abstract function getParserOutput( Title
$title, $revId = null, ParserOptions
$options = null, $generateHtml = true );
306 * Returns a list of DataUpdate objects for recording information about this Content in some secondary
307 * data store. If the optional second argument, $old, is given, the updates may model only the changes that
308 * need to be made to replace information about the old content with infomration about the new content.
310 * This default implementation calls $this->getParserOutput( $title, null, null, false ), and then
311 * calls getSecondaryDataUpdates( $title, $recursive ) on the resulting ParserOutput object.
313 * Subclasses may implement this to determine the necessary updates more efficiently, or make use of information
314 * about the old content.
316 * @param Title $title the context for determining the necessary updates
317 * @param Content|null $old a Content object representing the previous content, i.e. the content being
318 * replaced by this Content object.
319 * @param bool $recursive whether to include recursive updates (default: false).
321 * @return Array. A list of DataUpdate objects for putting information about this content object somewhere.
325 public function getSecondaryDataUpdates( Title
$title, Content
$old = null, $recursive = false ) {
326 $po = $this->getParserOutput( $title, null, null, false );
327 return $po->getSecondaryDataUpdates( $title, $recursive );
331 * Construct the redirect destination from this content and return an
332 * array of Titles, or null if this content doesn't represent a redirect.
333 * The last element in the array is the final destination after all redirects
334 * have been resolved (up to $wgMaxRedirects times).
338 * @return Array of Titles, with the destination last
340 public function getRedirectChain() {
345 * Construct the redirect destination from this content and return an
346 * array of Titles, or null if this content doesn't represent a redirect.
347 * This will only return the immediate redirect target, useful for
348 * the redirect table and other checks that don't need full recursion.
352 * @return Title: The corresponding Title
354 public function getRedirectTarget() {
359 * Construct the redirect destination from this content and return the
360 * Title, or null if this content doesn't represent a redirect.
361 * This will recurse down $wgMaxRedirects times or until a non-redirect target is hit
362 * in order to provide (hopefully) the Title of the final destination instead of another redirect.
368 public function getUltimateRedirectTarget() {
377 public function isRedirect() {
378 return $this->getRedirectTarget() !== null;
382 * Returns the section with the given id.
384 * The default implementation returns null.
388 * @param String $sectionId the section's id, given as a numeric string. The id "0" retrieves the section before
389 * the first heading, "1" the text between the first heading (inluded) and the second heading (excluded), etc.
390 * @return Content|Boolean|null the section, or false if no such section exist, or null if sections are not supported
392 public function getSection( $sectionId ) {
397 * Replaces a section of the content and returns a Content object with the section replaced.
401 * @param $section empty/null/false or a section number (0, 1, 2, T1, T2...), or "new"
402 * @param $with Content: new content of the section
403 * @param $sectionTitle String: new section's subject, only if $section is 'new'
404 * @return string Complete article text, or null if error
406 public function replaceSection( $section, Content
$with, $sectionTitle = '' ) {
411 * Returns a Content object with pre-save transformations applied (or this object if no transformations apply).
415 * @param Title $title
417 * @param null|ParserOptions $popts
420 public function preSaveTransform( Title
$title, User
$user, ParserOptions
$popts ) {
425 * Returns a new WikitextContent object with the given section heading prepended, if supported.
426 * The default implementation just returns this Content object unmodified, ignoring the section header.
430 * @param $header String
433 public function addSectionHeader( $header ) {
438 * Returns a Content object with preload transformations applied (or this object if no transformations apply).
442 * @param Title $title
443 * @param null|ParserOptions $popts
446 public function preloadTransform( Title
$title, ParserOptions
$popts ) {
450 # TODO: handle ImagePage and CategoryPage
451 # TODO: make sure we cover lucene search / wikisearch.
452 # TODO: make sure ReplaceTemplates still works
453 # FUTURE: nice&sane integration of GeSHi syntax highlighting
454 # [11:59] <vvv> Hooks are ugly; make CodeHighlighter interface and a config to set the class which handles syntax highlighting
455 # [12:00] <vvv> And default it to a DummyHighlighter
457 # TODO: make sure we cover the external editor interface (does anyone actually use that?!)
459 # TODO: tie into API to provide contentModel for Revisions
460 # TODO: tie into API to provide serialized version and contentFormat for Revisions
461 # TODO: tie into API edit interface
462 # FUTURE: make EditForm plugin for EditPage
464 # FUTURE: special type for redirects?!
465 # FUTURE: MultipartMultipart < WikipageContent (Main + Links + X)
466 # FUTURE: LinksContent < LanguageLinksContent, CategoriesContent
469 * Content object implementation for representing flat text.
471 * TextContent instances are imutable
475 abstract class TextContent
extends Content
{
477 public function __construct( $text, $model_id = null ) {
478 parent
::__construct( $model_id );
480 $this->mText
= $text;
483 public function copy() {
484 return $this; #NOTE: this is ok since TextContent are imutable.
487 public function getTextForSummary( $maxlength = 250 ) {
490 $text = $this->getNativeData();
492 $truncatedtext = $wgContLang->truncate(
493 preg_replace( "/[\n\r]/", ' ', $text ),
494 max( 0, $maxlength ) );
496 return $truncatedtext;
500 * returns the text's size in bytes.
502 * @return int the size
504 public function getSize( ) {
505 $text = $this->getNativeData( );
506 return strlen( $text );
510 * Returns true if this content is not a redirect, and $wgArticleCountMethod is "any".
512 * @param $hasLinks Bool: if it is known whether this content contains links, provide this information here,
513 * to avoid redundant parsing to find out.
515 * @return bool true if the content is countable
517 public function isCountable( $hasLinks = null ) {
518 global $wgArticleCountMethod;
520 if ( $this->isRedirect( ) ) {
524 if ( $wgArticleCountMethod === 'any' ) {
532 * Returns the text represented by this Content object, as a string.
534 * @return String the raw text
536 public function getNativeData( ) {
537 $text = $this->mText
;
542 * Returns the text represented by this Content object, as a string.
544 * @return String the raw text
546 public function getTextForSearchIndex( ) {
547 return $this->getNativeData();
551 * Returns the text represented by this Content object, as a string.
553 * @return String the raw text
555 public function getWikitextForTransclusion( ) {
556 return $this->getNativeData();
560 * Returns a generic ParserOutput object, wrapping the HTML returned by getHtml().
562 * @return ParserOutput representing the HTML form of the text
564 public function getParserOutput( Title
$title, $revId = null, ParserOptions
$options = null, $generateHtml = true ) {
565 # generic implementation, relying on $this->getHtml()
567 if ( $generateHtml ) $html = $this->getHtml( $options );
570 $po = new ParserOutput( $html );
575 protected abstract function getHtml( );
578 * Diff this content object with another content object..
582 * @param Content $that the other content object to compare this content object to
583 * @param Language $lang the language object to use for text segmentation. If not given, $wgContentLang is used.
585 * @return DiffResult a diff representing the changes that would have to be made to this content object
586 * to make it equal to $that.
588 public function diff( Content
$that, Language
$lang = null ) {
591 $this->checkModelID( $that->getModel() );
593 #@todo: could implement this in DifferenceEngine and just delegate here?
595 if ( !$lang ) $lang = $wgContLang;
597 $otext = $this->getNativeData();
598 $ntext = $this->getNativeData();
600 # Note: Use native PHP diff, external engines don't give us abstract output
601 $ota = explode( "\n", $wgContLang->segmentForDiff( $otext ) );
602 $nta = explode( "\n", $wgContLang->segmentForDiff( $ntext ) );
604 $diffs = new Diff( $ota, $nta );
614 class WikitextContent
extends TextContent
{
616 public function __construct( $text ) {
617 parent
::__construct($text, CONTENT_MODEL_WIKITEXT
);
620 protected function getHtml( ) {
621 throw new MWException( "getHtml() not implemented for wikitext. Use getParserOutput()->getText()." );
625 * Returns a ParserOutput object resulting from parsing the content's text using $wgParser.
629 * @param IContextSource|null $context
631 * @param null|ParserOptions $options
632 * @param bool $generateHtml
634 * @return ParserOutput representing the HTML form of the text
636 public function getParserOutput( Title
$title, $revId = null, ParserOptions
$options = null, $generateHtml = true ) {
640 $options = new ParserOptions();
643 $po = $wgParser->parse( $this->mText
, $title, $options, true, true, $revId );
649 * Returns the section with the given id.
651 * @param String $sectionId the section's id
652 * @return Content|false|null the section, or false if no such section exist, or null if sections are not supported
654 public function getSection( $section ) {
657 $text = $this->getNativeData();
658 $sect = $wgParser->getSection( $text, $section, false );
660 return new WikitextContent( $sect );
664 * Replaces a section in the wikitext
666 * @param $section empty/null/false or a section number (0, 1, 2, T1, T2...), or "new"
667 * @param $with Content: new content of the section
668 * @param $sectionTitle String: new section's subject, only if $section is 'new'
669 * @return Content Complete article content, or null if error
671 public function replaceSection( $section, Content
$with, $sectionTitle = '' ) {
672 wfProfileIn( __METHOD__
);
674 $myModelId = $this->getModel();
675 $sectionModelId = $with->getModel();
677 if ( $sectionModelId != $myModelId ) {
678 $myModelName = ContentHandler
::getContentModelName( $myModelId );
679 $sectionModelName = ContentHandler
::getContentModelName( $sectionModelId );
681 throw new MWException( "Incompatible content model for section: document uses $myModelId ($myModelName), "
682 . "section uses $sectionModelId ($sectionModelName)." );
685 $oldtext = $this->getNativeData();
686 $text = $with->getNativeData();
688 if ( $section === '' ) {
689 return $with; #XXX: copy first?
690 } if ( $section == 'new' ) {
691 # Inserting a new section
692 $subject = $sectionTitle ?
wfMsgForContent( 'newsectionheaderdefaultlevel', $sectionTitle ) . "\n\n" : '';
693 if ( wfRunHooks( 'PlaceNewSection', array( $this, $oldtext, $subject, &$text ) ) ) {
694 $text = strlen( trim( $oldtext ) ) > 0
695 ?
"{$oldtext}\n\n{$subject}{$text}"
696 : "{$subject}{$text}";
699 # Replacing an existing section; roll out the big guns
702 $text = $wgParser->replaceSection( $oldtext, $section, $text );
705 $newContent = new WikitextContent( $text );
707 wfProfileOut( __METHOD__
);
712 * Returns a new WikitextContent object with the given section heading prepended.
714 * @param $header String
717 public function addSectionHeader( $header ) {
718 $text = wfMsgForContent( 'newsectionheaderdefaultlevel', $header ) . "\n\n" . $this->getNativeData();
720 return new WikitextContent( $text );
724 * Returns a Content object with pre-save transformations applied (or this object if no transformations apply).
726 * @param Title $title
728 * @param ParserOptions $popts
731 public function preSaveTransform( Title
$title, User
$user, ParserOptions
$popts ) {
732 global $wgParser, $wgConteLang;
734 $text = $this->getNativeData();
735 $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts );
737 return new WikitextContent( $pst );
741 * Returns a Content object with preload transformations applied (or this object if no transformations apply).
743 * @param Title $title
744 * @param ParserOptions $popts
747 public function preloadTransform( Title
$title, ParserOptions
$popts ) {
748 global $wgParser, $wgConteLang;
750 $text = $this->getNativeData();
751 $plt = $wgParser->getPreloadText( $text, $title, $popts );
753 return new WikitextContent( $plt );
756 public function getRedirectChain() {
757 $text = $this->getNativeData();
758 return Title
::newFromRedirectArray( $text );
761 public function getRedirectTarget() {
762 $text = $this->getNativeData();
763 return Title
::newFromRedirect( $text );
766 public function getUltimateRedirectTarget() {
767 $text = $this->getNativeData();
768 return Title
::newFromRedirectRecurse( $text );
772 * Returns true if this content is not a redirect, and this content's text is countable according to
773 * the criteria defiend by $wgArticleCountMethod.
775 * @param Bool $hasLinks if it is known whether this content contains links, provide this information here,
776 * to avoid redundant parsing to find out.
777 * @param IContextSource $context context for parsing if necessary
779 * @return bool true if the content is countable
781 public function isCountable( $hasLinks = null, Title
$title = null ) {
782 global $wgArticleCountMethod, $wgRequest;
784 if ( $this->isRedirect( ) ) {
788 $text = $this->getNativeData();
790 switch ( $wgArticleCountMethod ) {
794 return strpos( $text, ',' ) !== false;
796 if ( $hasLinks === null ) { # not known, find out
798 $context = RequestContext
::getMain();
799 $title = $context->getTitle();
802 $po = $this->getParserOutput( $title, null, null, false );
803 $links = $po->getLinks();
804 $hasLinks = !empty( $links );
811 public function getTextForSummary( $maxlength = 250 ) {
812 $truncatedtext = parent
::getTextForSummary( $maxlength );
814 #clean up unfinished links
815 #XXX: make this optional? wasn't there in autosummary, but required for deletion summary.
816 $truncatedtext = preg_replace( '/\[\[([^\]]*)\]?$/', '$1', $truncatedtext );
818 return $truncatedtext;
826 class MessageContent
extends TextContent
{
827 public function __construct( $msg_key, $params = null, $options = null ) {
828 parent
::__construct(null, CONTENT_MODEL_WIKITEXT
); #XXX: messages may be wikitext, html or plain text! and maybe even something else entirely.
830 $this->mMessageKey
= $msg_key;
832 $this->mParameters
= $params;
834 if ( is_null( $options ) ) {
837 elseif ( is_string( $options ) ) {
838 $options = array( $options );
841 $this->mOptions
= $options;
843 $this->mHtmlOptions
= null;
847 * Returns the message as rendered HTML, using the options supplied to the constructor plus "parse".
849 protected function getHtml( ) {
850 $opt = array_merge( $this->mOptions
, array('parse') );
852 return wfMsgExt( $this->mMessageKey
, $this->mParameters
, $opt );
857 * Returns the message as raw text, using the options supplied to the constructor minus "parse" and "parseinline".
859 public function getNativeData( ) {
860 $opt = array_diff( $this->mOptions
, array('parse', 'parseinline') );
862 return wfMsgExt( $this->mMessageKey
, $this->mParameters
, $opt );
870 class JavaScriptContent
extends TextContent
{
871 public function __construct( $text ) {
872 parent
::__construct($text, CONTENT_MODEL_JAVASCRIPT
);
875 protected function getHtml( ) {
877 $html .= "<pre class=\"mw-code mw-js\" dir=\"ltr\">\n";
878 $html .= htmlspecialchars( $this->getNativeData() );
879 $html .= "\n</pre>\n";
889 class CssContent
extends TextContent
{
890 public function __construct( $text ) {
891 parent
::__construct($text, CONTENT_MODEL_CSS
);
894 protected function getHtml( ) {
896 $html .= "<pre class=\"mw-code mw-css\" dir=\"ltr\">\n";
897 $html .= htmlspecialchars( $this->getNativeData() );
898 $html .= "\n</pre>\n";