diff options
Diffstat (limited to 'includes/content')
| -rw-r--r-- | includes/content/AbstractContent.php | 444 | ||||
| -rw-r--r-- | includes/content/Content.php | 508 | ||||
| -rw-r--r-- | includes/content/ContentHandler.php | 1114 | ||||
| -rw-r--r-- | includes/content/CssContent.php | 65 | ||||
| -rw-r--r-- | includes/content/CssContentHandler.php | 67 | ||||
| -rw-r--r-- | includes/content/JavaScriptContent.php | 66 | ||||
| -rw-r--r-- | includes/content/JavaScriptContentHandler.php | 67 | ||||
| -rw-r--r-- | includes/content/MessageContent.php | 158 | ||||
| -rw-r--r-- | includes/content/TextContent.php | 286 | ||||
| -rw-r--r-- | includes/content/TextContentHandler.php | 115 | ||||
| -rw-r--r-- | includes/content/WikitextContent.php | 322 | ||||
| -rw-r--r-- | includes/content/WikitextContentHandler.php | 98 |
12 files changed, 3310 insertions, 0 deletions
diff --git a/includes/content/AbstractContent.php b/includes/content/AbstractContent.php new file mode 100644 index 00000000..137efb8a --- /dev/null +++ b/includes/content/AbstractContent.php @@ -0,0 +1,444 @@ +<?php +/** + * A content object represents page content, e.g. the text to show on a page. + * Content objects have no knowledge about how they relate to Wiki pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Base implementation for content objects. + * + * @ingroup Content + */ +abstract class AbstractContent implements Content { + + /** + * Name of the content model this Content object represents. + * Use with CONTENT_MODEL_XXX constants + * + * @since 1.21 + * + * @var string $model_id + */ + protected $model_id; + + /** + * @param string|null $modelId + * + * @since 1.21 + */ + public function __construct( $modelId = null ) { + $this->model_id = $modelId; + } + + /** + * @see Content::getModel + * + * @since 1.21 + */ + public function getModel() { + return $this->model_id; + } + + /** + * Throws an MWException if $model_id is not the id of the content model + * supported by this Content object. + * + * @since 1.21 + * + * @param string $modelId The model to check + * + * @throws MWException + */ + protected function checkModelID( $modelId ) { + if ( $modelId !== $this->model_id ) { + throw new MWException( + "Bad content model: " . + "expected {$this->model_id} " . + "but got $modelId." + ); + } + } + + /** + * @see Content::getContentHandler + * + * @since 1.21 + */ + public function getContentHandler() { + return ContentHandler::getForContent( $this ); + } + + /** + * @see Content::getDefaultFormat + * + * @since 1.21 + */ + public function getDefaultFormat() { + return $this->getContentHandler()->getDefaultFormat(); + } + + /** + * @see Content::getSupportedFormats + * + * @since 1.21 + */ + public function getSupportedFormats() { + return $this->getContentHandler()->getSupportedFormats(); + } + + /** + * @see Content::isSupportedFormat + * + * @param string $format + * + * @since 1.21 + * + * @return boolean + */ + public function isSupportedFormat( $format ) { + if ( !$format ) { + return true; // this means "use the default" + } + + return $this->getContentHandler()->isSupportedFormat( $format ); + } + + /** + * Throws an MWException if $this->isSupportedFormat( $format ) does not + * return true. + * + * @since 1.21 + * + * @param string $format + * @throws MWException + */ + protected function checkFormat( $format ) { + if ( !$this->isSupportedFormat( $format ) ) { + throw new MWException( + "Format $format is not supported for content model " . + $this->getModel() + ); + } + } + + /** + * @see Content::serialize + * + * @param string|null $format + * + * @since 1.21 + * + * @return string + */ + public function serialize( $format = null ) { + return $this->getContentHandler()->serializeContent( $this, $format ); + } + + /** + * @see Content::isEmpty + * + * @since 1.21 + * + * @return boolean + */ + public function isEmpty() { + return $this->getSize() === 0; + } + + /** + * @see Content::isValid + * + * @since 1.21 + * + * @return boolean + */ + public function isValid() { + return true; + } + + /** + * @see Content::equals + * + * @since 1.21 + * + * @param Content|null $that + * + * @return boolean + */ + public function equals( Content $that = null ) { + if ( is_null( $that ) ) { + return false; + } + + if ( $that === $this ) { + return true; + } + + if ( $that->getModel() !== $this->getModel() ) { + return false; + } + + return $this->getNativeData() === $that->getNativeData(); + } + + /** + * Returns a list of DataUpdate objects for recording information about this + * Content in some secondary data store. + * + * This default implementation calls + * $this->getParserOutput( $content, $title, null, null, false ), + * and then calls getSecondaryDataUpdates( $title, $recursive ) on the + * resulting ParserOutput object. + * + * Subclasses may override this to determine the secondary data updates more + * efficiently, preferably without the need to generate a parser output object. + * + * @see Content::getSecondaryDataUpdates() + * + * @param $title Title The context for determining the necessary updates + * @param $old Content|null An optional Content object representing the + * previous content, i.e. the content being replaced by this Content + * object. + * @param $recursive boolean Whether to include recursive updates (default: + * false). + * @param $parserOutput ParserOutput|null Optional ParserOutput object. + * Provide if you have one handy, to avoid re-parsing of the content. + * + * @return Array. A list of DataUpdate objects for putting information + * about this content object somewhere. + * + * @since 1.21 + */ + public function getSecondaryDataUpdates( Title $title, + Content $old = null, + $recursive = true, ParserOutput $parserOutput = null + ) { + if ( $parserOutput === null ) { + $parserOutput = $this->getParserOutput( $title, null, null, false ); + } + + return $parserOutput->getSecondaryDataUpdates( $title, $recursive ); + } + + /** + * @see Content::getRedirectChain + * + * @since 1.21 + */ + public function getRedirectChain() { + global $wgMaxRedirects; + $title = $this->getRedirectTarget(); + if ( is_null( $title ) ) { + return null; + } + // recursive check to follow double redirects + $recurse = $wgMaxRedirects; + $titles = array( $title ); + while ( --$recurse > 0 ) { + if ( $title->isRedirect() ) { + $page = WikiPage::factory( $title ); + $newtitle = $page->getRedirectTarget(); + } else { + break; + } + // Redirects to some special pages are not permitted + if ( $newtitle instanceOf Title && $newtitle->isValidRedirectTarget() ) { + // The new title passes the checks, so make that our current + // title so that further recursion can be checked + $title = $newtitle; + $titles[] = $newtitle; + } else { + break; + } + } + return $titles; + } + + /** + * @see Content::getRedirectTarget + * + * @since 1.21 + */ + public function getRedirectTarget() { + return null; + } + + /** + * @see Content::getUltimateRedirectTarget + * @note: migrated here from Title::newFromRedirectRecurse + * + * @since 1.21 + */ + public function getUltimateRedirectTarget() { + $titles = $this->getRedirectChain(); + return $titles ? array_pop( $titles ) : null; + } + + /** + * @see Content::isRedirect + * + * @since 1.21 + * + * @return bool + */ + public function isRedirect() { + return $this->getRedirectTarget() !== null; + } + + /** + * @see Content::updateRedirect + * + * This default implementation always returns $this. + * + * @param Title $target + * + * @since 1.21 + * + * @return Content $this + */ + public function updateRedirect( Title $target ) { + return $this; + } + + /** + * @see Content::getSection + * + * @since 1.21 + */ + public function getSection( $sectionId ) { + return null; + } + + /** + * @see Content::replaceSection + * + * @since 1.21 + */ + public function replaceSection( $section, Content $with, $sectionTitle = '' ) { + return null; + } + + /** + * @see Content::preSaveTransform + * + * @since 1.21 + */ + public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) { + return $this; + } + + /** + * @see Content::addSectionHeader + * + * @since 1.21 + */ + public function addSectionHeader( $header ) { + return $this; + } + + /** + * @see Content::preloadTransform + * + * @since 1.21 + */ + public function preloadTransform( Title $title, ParserOptions $popts ) { + return $this; + } + + /** + * @see Content::prepareSave + * + * @since 1.21 + */ + public function prepareSave( WikiPage $page, $flags, $baseRevId, User $user ) { + if ( $this->isValid() ) { + return Status::newGood(); + } else { + return Status::newFatal( "invalid-content-data" ); + } + } + + /** + * @see Content::getDeletionUpdates + * + * @since 1.21 + * + * @param $page WikiPage the deleted page + * @param $parserOutput null|ParserOutput optional parser output object + * for efficient access to meta-information about the content object. + * Provide if you have one handy. + * + * @return array A list of DataUpdate instances that will clean up the + * database after deletion. + */ + public function getDeletionUpdates( WikiPage $page, + ParserOutput $parserOutput = null ) + { + return array( + new LinksDeletionUpdate( $page ), + ); + } + + /** + * This default implementation always returns false. Subclasses may override this to supply matching logic. + * + * @see Content::matchMagicWord + * + * @since 1.21 + * + * @param MagicWord $word + * + * @return bool + */ + public function matchMagicWord( MagicWord $word ) { + return false; + } + + /** + * @see Content::convert() + * + * This base implementation calls the hook ConvertContent to enable custom conversions. + * Subclasses may override this to implement conversion for "their" content model. + * + * @param string $toModel the desired content model, use the CONTENT_MODEL_XXX flags. + * @param string $lossy flag, set to "lossy" to allow lossy conversion. If lossy conversion is + * not allowed, full round-trip conversion is expected to work without losing information. + * + * @return Content|bool A content object with the content model $toModel, or false if + * that conversion is not supported. + */ + public function convert( $toModel, $lossy = '' ) { + if ( $this->getModel() === $toModel ) { + //nothing to do, shorten out. + return $this; + } + + $lossy = ( $lossy === 'lossy' ); // string flag, convert to boolean for convenience + $result = false; + + wfRunHooks( 'ConvertContent', array( $this, $toModel, $lossy, &$result ) ); + return $result; + } +} diff --git a/includes/content/Content.php b/includes/content/Content.php new file mode 100644 index 00000000..72729b09 --- /dev/null +++ b/includes/content/Content.php @@ -0,0 +1,508 @@ +<?php +/** + * A content object represents page content, e.g. the text to show on a page. + * Content objects have no knowledge about how they relate to wiki pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Base interface for content objects. + * + * @ingroup Content + */ +interface Content { + + /** + * @since 1.21 + * + * @return string A string representing the content in a way useful for + * building a full text search index. If no useful representation exists, + * this method returns an empty string. + * + * @todo: test that this actually works + * @todo: make sure this also works with LuceneSearch / WikiSearch + */ + public function getTextForSearchIndex(); + + /** + * @since 1.21 + * + * @return string|false The wikitext to include when another page includes this + * content, or false if the content is not includable in a wikitext page. + * + * @todo allow native handling, bypassing wikitext representation, like + * for includable special pages. + * @todo allow transclusion into other content models than Wikitext! + * @todo used in WikiPage and MessageCache to get message text. Not so + * nice. What should we use instead?! + */ + public function getWikitextForTransclusion(); + + /** + * Returns a textual representation of the content suitable for use in edit + * summaries and log messages. + * + * @since 1.21 + * + * @param int $maxLength Maximum length of the summary text + * @return string The summary text + */ + public function getTextForSummary( $maxLength = 250 ); + + /** + * Returns native representation of the data. Interpretation depends on + * the data model used, as given by getDataModel(). + * + * @since 1.21 + * + * @return mixed The native representation of the content. Could be a + * string, a nested array structure, an object, a binary blob... + * anything, really. + * + * @note Caller must be aware of content model! + */ + public function getNativeData(); + + /** + * Returns the content's nominal size in bogo-bytes. + * + * @return int + */ + public function getSize(); + + /** + * Returns the ID of the content model used by this Content object. + * Corresponds to the CONTENT_MODEL_XXX constants. + * + * @since 1.21 + * + * @return String The model id + */ + public function getModel(); + + /** + * Convenience method that returns the ContentHandler singleton for handling + * the content model that this Content object uses. + * + * Shorthand for ContentHandler::getForContent( $this ) + * + * @since 1.21 + * + * @return ContentHandler + */ + public function getContentHandler(); + + /** + * Convenience method that returns the default serialization format for the + * content model that this Content object uses. + * + * Shorthand for $this->getContentHandler()->getDefaultFormat() + * + * @since 1.21 + * + * @return String + */ + public function getDefaultFormat(); + + /** + * Convenience method that returns the list of serialization formats + * supported for the content model that this Content object uses. + * + * Shorthand for $this->getContentHandler()->getSupportedFormats() + * + * @since 1.21 + * + * @return Array of supported serialization formats + */ + public function getSupportedFormats(); + + /** + * Returns true if $format is a supported serialization format for this + * Content object, false if it isn't. + * + * Note that this should always return true if $format is null, because null + * stands for the default serialization. + * + * Shorthand for $this->getContentHandler()->isSupportedFormat( $format ) + * + * @since 1.21 + * + * @param string $format The format to check + * @return bool Whether the format is supported + */ + public function isSupportedFormat( $format ); + + /** + * Convenience method for serializing this Content object. + * + * Shorthand for $this->getContentHandler()->serializeContent( $this, $format ) + * + * @since 1.21 + * + * @param $format null|string The desired serialization format (or null for + * the default format). + * @return string Serialized form of this Content object + */ + public function serialize( $format = null ); + + /** + * Returns true if this Content object represents empty content. + * + * @since 1.21 + * + * @return bool Whether this Content object is empty + */ + public function isEmpty(); + + /** + * Returns whether the content is valid. This is intended for local validity + * checks, not considering global consistency. + * + * Content needs to be valid before it can be saved. + * + * This default implementation always returns true. + * + * @since 1.21 + * + * @return boolean + */ + public function isValid(); + + /** + * Returns true if this Content objects is conceptually equivalent to the + * given Content object. + * + * Contract: + * + * - Will return false if $that is null. + * - Will return true if $that === $this. + * - Will return false if $that->getModel() != $this->getModel(). + * - Will return false if $that->getNativeData() is not equal to $this->getNativeData(), + * where the meaning of "equal" depends on the actual data model. + * + * Implementations should be careful to make equals() transitive and reflexive: + * + * - $a->equals( $b ) <=> $b->equals( $a ) + * - $a->equals( $b ) && $b->equals( $c ) ==> $a->equals( $c ) + * + * @since 1.21 + * + * @param $that Content The Content object to compare to + * @return bool True if this Content object is equal to $that, false otherwise. + */ + public function equals( Content $that = null ); + + /** + * Return a copy of this Content object. The following must be true for the + * object returned: + * + * if $copy = $original->copy() + * + * - get_class($original) === get_class($copy) + * - $original->getModel() === $copy->getModel() + * - $original->equals( $copy ) + * + * If and only if the Content object is immutable, the copy() method can and + * should return $this. That is, $copy === $original may be true, but only + * for immutable content objects. + * + * @since 1.21 + * + * @return Content. A copy of this object + */ + public function copy(); + + /** + * Returns true if this content is countable as a "real" wiki page, provided + * that it's also in a countable location (e.g. a current revision in the + * main namespace). + * + * @since 1.21 + * + * @param bool $hasLinks If it is known whether this content contains + * links, provide this information here, to avoid redundant parsing to + * find out. + * @return boolean + */ + public function isCountable( $hasLinks = null ); + + /** + * Parse the Content object and generate a ParserOutput from the result. + * $result->getText() can be used to obtain the generated HTML. If no HTML + * is needed, $generateHtml can be set to false; in that case, + * $result->getText() may return null. + * + * @param $title Title The page title to use as a context for rendering + * @param $revId null|int The revision being rendered (optional) + * @param $options null|ParserOptions Any parser options + * @param $generateHtml Boolean Whether to generate HTML (default: true). If false, + * the result of calling getText() on the ParserOutput object returned by + * this method is undefined. + * + * @since 1.21 + * + * @return ParserOutput + */ + public function getParserOutput( Title $title, + $revId = null, + ParserOptions $options = null, $generateHtml = true ); + // TODO: make RenderOutput and RenderOptions base classes + + /** + * Returns a list of DataUpdate objects for recording information about this + * Content in some secondary data store. If the optional second argument, + * $old, is given, the updates may model only the changes that need to be + * made to replace information about the old content with information about + * the new content. + * + * This default implementation calls + * $this->getParserOutput( $content, $title, null, null, false ), + * and then calls getSecondaryDataUpdates( $title, $recursive ) on the + * resulting ParserOutput object. + * + * Subclasses may implement this to determine the necessary updates more + * efficiently, or make use of information about the old content. + * + * @param $title Title The context for determining the necessary updates + * @param $old Content|null An optional Content object representing the + * previous content, i.e. the content being replaced by this Content + * object. + * @param $recursive boolean Whether to include recursive updates (default: + * false). + * @param $parserOutput ParserOutput|null Optional ParserOutput object. + * Provide if you have one handy, to avoid re-parsing of the content. + * + * @return Array. A list of DataUpdate objects for putting information + * about this content object somewhere. + * + * @since 1.21 + */ + public function getSecondaryDataUpdates( Title $title, + Content $old = null, + $recursive = true, ParserOutput $parserOutput = null + ); + + /** + * Construct the redirect destination from this content and return an + * array of Titles, or null if this content doesn't represent a redirect. + * The last element in the array is the final destination after all redirects + * have been resolved (up to $wgMaxRedirects times). + * + * @since 1.21 + * + * @return Array of Titles, with the destination last + */ + public function getRedirectChain(); + + /** + * Construct the redirect destination from this content and return a Title, + * or null if this content doesn't represent a redirect. + * This will only return the immediate redirect target, useful for + * the redirect table and other checks that don't need full recursion. + * + * @since 1.21 + * + * @return Title: The corresponding Title + */ + public function getRedirectTarget(); + + /** + * Construct the redirect destination from this content and return the + * Title, or null if this content doesn't represent a redirect. + * + * This will recurse down $wgMaxRedirects times or until a non-redirect + * target is hit in order to provide (hopefully) the Title of the final + * destination instead of another redirect. + * + * There is usually no need to override the default behavior, subclasses that + * want to implement redirects should override getRedirectTarget(). + * + * @since 1.21 + * + * @return Title + */ + public function getUltimateRedirectTarget(); + + /** + * Returns whether this Content represents a redirect. + * Shorthand for getRedirectTarget() !== null. + * + * @since 1.21 + * + * @return bool + */ + public function isRedirect(); + + /** + * If this Content object is a redirect, this method updates the redirect target. + * Otherwise, it does nothing. + * + * @since 1.21 + * + * @param Title $target the new redirect target + * + * @return Content a new Content object with the updated redirect (or $this if this Content object isn't a redirect) + */ + public function updateRedirect( Title $target ); + + /** + * Returns the section with the given ID. + * + * @since 1.21 + * + * @param string $sectionId The section's ID, given as a numeric string. + * The ID "0" retrieves the section before the first heading, "1" the + * text between the first heading (included) and the second heading + * (excluded), etc. + * @return Content|Boolean|null The section, or false if no such section + * exist, or null if sections are not supported. + */ + public function getSection( $sectionId ); + + /** + * Replaces a section of the content and returns a Content object with the + * section replaced. + * + * @since 1.21 + * + * @param $section null/false or a section number (0, 1, 2, T1, T2...), or "new" + * @param $with Content: new content of the section + * @param string $sectionTitle new section's subject, only if $section is 'new' + * @return string Complete article text, or null if error + */ + public function replaceSection( $section, Content $with, $sectionTitle = '' ); + + /** + * Returns a Content object with pre-save transformations applied (or this + * object if no transformations apply). + * + * @since 1.21 + * + * @param $title Title + * @param $user User + * @param $parserOptions null|ParserOptions + * @return Content + */ + public function preSaveTransform( Title $title, User $user, ParserOptions $parserOptions ); + + /** + * Returns a new WikitextContent object with the given section heading + * prepended, if supported. The default implementation just returns this + * Content object unmodified, ignoring the section header. + * + * @since 1.21 + * + * @param $header string + * @return Content + */ + public function addSectionHeader( $header ); + + /** + * Returns a Content object with preload transformations applied (or this + * object if no transformations apply). + * + * @since 1.21 + * + * @param $title Title + * @param $parserOptions null|ParserOptions + * @return Content + */ + public function preloadTransform( Title $title, ParserOptions $parserOptions ); + + /** + * Prepare Content for saving. Called before Content is saved by WikiPage::doEditContent() and in + * similar places. + * + * This may be used to check the content's consistency with global state. This function should + * NOT write any information to the database. + * + * Note that this method will usually be called inside the same transaction bracket that will be used + * to save the new revision. + * + * Note that this method is called before any update to the page table is performed. This means that + * $page may not yet know a page ID. + * + * @since 1.21 + * + * @param WikiPage $page The page to be saved. + * @param int $flags bitfield for use with EDIT_XXX constants, see WikiPage::doEditContent() + * @param int $baseRevId the ID of the current revision + * @param User $user + * + * @return Status A status object indicating whether the content was successfully prepared for saving. + * If the returned status indicates an error, a rollback will be performed and the + * transaction aborted. + * + * @see see WikiPage::doEditContent() + */ + public function prepareSave( WikiPage $page, $flags, $baseRevId, User $user ); + + /** + * Returns a list of updates to perform when this content is deleted. + * The necessary updates may be taken from the Content object, or depend on + * the current state of the database. + * + * @since 1.21 + * + * @param $page WikiPage the deleted page + * @param $parserOutput null|ParserOutput optional parser output object + * for efficient access to meta-information about the content object. + * Provide if you have one handy. + * + * @return array A list of DataUpdate instances that will clean up the + * database after deletion. + */ + public function getDeletionUpdates( WikiPage $page, + ParserOutput $parserOutput = null ); + + /** + * Returns true if this Content object matches the given magic word. + * + * @since 1.21 + * + * @param MagicWord $word the magic word to match + * + * @return bool whether this Content object matches the given magic word. + */ + public function matchMagicWord( MagicWord $word ); + + /** + * Converts this content object into another content object with the given content model, + * if that is possible. + * + * @param string $toModel the desired content model, use the CONTENT_MODEL_XXX flags. + * @param string $lossy flag, set to "lossy" to allow lossy conversion. If lossy conversion is + * not allowed, full round-trip conversion is expected to work without losing information. + * + * @return Content|bool A content object with the content model $toModel, or false if + * that conversion is not supported. + */ + public function convert( $toModel, $lossy = '' ); + + // TODO: ImagePage and CategoryPage interfere with per-content action handlers + // TODO: nice&sane integration of GeSHi syntax highlighting + // [11:59] <vvv> Hooks are ugly; make CodeHighlighter interface and a + // config to set the class which handles syntax highlighting + // [12:00] <vvv> And default it to a DummyHighlighter +} diff --git a/includes/content/ContentHandler.php b/includes/content/ContentHandler.php new file mode 100644 index 00000000..9c201955 --- /dev/null +++ b/includes/content/ContentHandler.php @@ -0,0 +1,1114 @@ +<?php +/** + * Base class for content handling. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Exception representing a failure to serialize or unserialize a content object. + * + * @ingroup Content + */ +class MWContentSerializationException extends MWException { + +} + +/** + * A content handler knows how do deal with a specific type of content on a wiki + * page. Content is stored in the database in a serialized form (using a + * serialization format a.k.a. MIME type) and is unserialized into its native + * PHP representation (the content model), which is wrapped in an instance of + * the appropriate subclass of Content. + * + * ContentHandler instances are stateless singletons that serve, among other + * things, as a factory for Content objects. Generally, there is one subclass + * of ContentHandler and one subclass of Content for every type of content model. + * + * Some content types have a flat model, that is, their native representation + * is the same as their serialized form. Examples would be JavaScript and CSS + * code. As of now, this also applies to wikitext (MediaWiki's default content + * type), but wikitext content may be represented by a DOM or AST structure in + * the future. + * + * @ingroup Content + */ +abstract class ContentHandler { + + /** + * Switch for enabling deprecation warnings. Used by ContentHandler::deprecated() + * and ContentHandler::runLegacyHooks(). + * + * Once the ContentHandler code has settled in a bit, this should be set to true to + * make extensions etc. show warnings when using deprecated functions and hooks. + */ + protected static $enableDeprecationWarnings = false; + + /** + * Convenience function for getting flat text from a Content object. This + * should only be used in the context of backwards compatibility with code + * that is not yet able to handle Content objects! + * + * If $content is null, this method returns the empty string. + * + * If $content is an instance of TextContent, this method returns the flat + * text as returned by $content->getNativeData(). + * + * If $content is not a TextContent object, the behavior of this method + * depends on the global $wgContentHandlerTextFallback: + * - If $wgContentHandlerTextFallback is 'fail' and $content is not a + * TextContent object, an MWException is thrown. + * - If $wgContentHandlerTextFallback is 'serialize' and $content is not a + * TextContent object, $content->serialize() is called to get a string + * form of the content. + * - If $wgContentHandlerTextFallback is 'ignore' and $content is not a + * TextContent object, this method returns null. + * - otherwise, the behavior is undefined. + * + * @since 1.21 + * + * @param $content Content|null + * @return null|string the textual form of $content, if available + * @throws MWException if $content is not an instance of TextContent and + * $wgContentHandlerTextFallback was set to 'fail'. + */ + public static function getContentText( Content $content = null ) { + global $wgContentHandlerTextFallback; + + if ( is_null( $content ) ) { + return ''; + } + + if ( $content instanceof TextContent ) { + return $content->getNativeData(); + } + + wfDebugLog( 'ContentHandler', 'Accessing ' . $content->getModel() . ' content as text!' ); + + if ( $wgContentHandlerTextFallback == 'fail' ) { + throw new MWException( + "Attempt to get text from Content with model " . + $content->getModel() + ); + } + + if ( $wgContentHandlerTextFallback == 'serialize' ) { + return $content->serialize(); + } + + return null; + } + + /** + * Convenience function for creating a Content object from a given textual + * representation. + * + * $text will be deserialized into a Content object of the model specified + * by $modelId (or, if that is not given, $title->getContentModel()) using + * the given format. + * + * @since 1.21 + * + * @param string $text the textual representation, will be + * unserialized to create the Content object + * @param $title null|Title the title of the page this text belongs to. + * Required if $modelId is not provided. + * @param $modelId null|string the model to deserialize to. If not provided, + * $title->getContentModel() is used. + * @param $format null|string the format to use for deserialization. If not + * given, the model's default format is used. + * + * @throws MWException + * @return Content a Content object representing $text + * + * @throws MWException if $model or $format is not supported or if $text can + * not be unserialized using $format. + */ + public static function makeContent( $text, Title $title = null, + $modelId = null, $format = null ) + { + if ( is_null( $modelId ) ) { + if ( is_null( $title ) ) { + throw new MWException( "Must provide a Title object or a content model ID." ); + } + + $modelId = $title->getContentModel(); + } + + $handler = ContentHandler::getForModelID( $modelId ); + return $handler->unserializeContent( $text, $format ); + } + + /** + * Returns the name of the default content model to be used for the page + * with the given title. + * + * Note: There should rarely be need to call this method directly. + * To determine the actual content model for a given page, use + * Title::getContentModel(). + * + * Which model is to be used by default for the page is determined based + * on several factors: + * - The global setting $wgNamespaceContentModels specifies a content model + * per namespace. + * - The hook ContentHandlerDefaultModelFor may be used to override the page's default + * model. + * - Pages in NS_MEDIAWIKI and NS_USER default to the CSS or JavaScript + * model if they end in .js or .css, respectively. + * - Pages in NS_MEDIAWIKI default to the wikitext model otherwise. + * - The hook TitleIsCssOrJsPage may be used to force a page to use the CSS + * or JavaScript model. This is a compatibility feature. The ContentHandlerDefaultModelFor + * hook should be used instead if possible. + * - The hook TitleIsWikitextPage may be used to force a page to use the + * wikitext model. This is a compatibility feature. The ContentHandlerDefaultModelFor + * hook should be used instead if possible. + * + * If none of the above applies, the wikitext model is used. + * + * Note: this is used by, and may thus not use, Title::getContentModel() + * + * @since 1.21 + * + * @param $title Title + * @return null|string default model name for the page given by $title + */ + public static function getDefaultModelFor( Title $title ) { + // NOTE: this method must not rely on $title->getContentModel() directly or indirectly, + // because it is used to initialize the mContentModel member. + + $ns = $title->getNamespace(); + + $ext = false; + $m = null; + $model = MWNamespace::getNamespaceContentModel( $ns ); + + // Hook can determine default model + if ( !wfRunHooks( 'ContentHandlerDefaultModelFor', array( $title, &$model ) ) ) { + if ( !is_null( $model ) ) { + return $model; + } + } + + // Could this page contain custom CSS or JavaScript, based on the title? + $isCssOrJsPage = NS_MEDIAWIKI == $ns && preg_match( '!\.(css|js)$!u', $title->getText(), $m ); + if ( $isCssOrJsPage ) { + $ext = $m[1]; + } + + // Hook can force JS/CSS + wfRunHooks( 'TitleIsCssOrJsPage', array( $title, &$isCssOrJsPage ) ); + + // Is this a .css subpage of a user page? + $isJsCssSubpage = NS_USER == $ns + && !$isCssOrJsPage + && preg_match( "/\\/.*\\.(js|css)$/", $title->getText(), $m ); + if ( $isJsCssSubpage ) { + $ext = $m[1]; + } + + // Is this wikitext, according to $wgNamespaceContentModels or the DefaultModelFor hook? + $isWikitext = is_null( $model ) || $model == CONTENT_MODEL_WIKITEXT; + $isWikitext = $isWikitext && !$isCssOrJsPage && !$isJsCssSubpage; + + // Hook can override $isWikitext + wfRunHooks( 'TitleIsWikitextPage', array( $title, &$isWikitext ) ); + + if ( !$isWikitext ) { + switch ( $ext ) { + case 'js': + return CONTENT_MODEL_JAVASCRIPT; + case 'css': + return CONTENT_MODEL_CSS; + default: + return is_null( $model ) ? CONTENT_MODEL_TEXT : $model; + } + } + + // We established that it must be wikitext + + return CONTENT_MODEL_WIKITEXT; + } + + /** + * Returns the appropriate ContentHandler singleton for the given title. + * + * @since 1.21 + * + * @param $title Title + * @return ContentHandler + */ + public static function getForTitle( Title $title ) { + $modelId = $title->getContentModel(); + return ContentHandler::getForModelID( $modelId ); + } + + /** + * Returns the appropriate ContentHandler singleton for the given Content + * object. + * + * @since 1.21 + * + * @param $content Content + * @return ContentHandler + */ + public static function getForContent( Content $content ) { + $modelId = $content->getModel(); + return ContentHandler::getForModelID( $modelId ); + } + + /** + * @var Array A Cache of ContentHandler instances by model id + */ + static $handlers; + + /** + * Returns the ContentHandler singleton for the given model ID. Use the + * CONTENT_MODEL_XXX constants to identify the desired content model. + * + * ContentHandler singletons are taken from the global $wgContentHandlers + * array. Keys in that array are model names, the values are either + * ContentHandler singleton objects, or strings specifying the appropriate + * subclass of ContentHandler. + * + * If a class name is encountered when looking up the singleton for a given + * model name, the class is instantiated and the class name is replaced by + * the resulting singleton in $wgContentHandlers. + * + * If no ContentHandler is defined for the desired $modelId, the + * ContentHandler may be provided by the ContentHandlerForModelID hook. + * If no ContentHandler can be determined, an MWException is raised. + * + * @since 1.21 + * + * @param string $modelId The ID of the content model for which to get a + * handler. Use CONTENT_MODEL_XXX constants. + * @return ContentHandler The ContentHandler singleton for handling the + * model given by $modelId + * @throws MWException if no handler is known for $modelId. + */ + public static function getForModelID( $modelId ) { + global $wgContentHandlers; + + if ( isset( ContentHandler::$handlers[$modelId] ) ) { + return ContentHandler::$handlers[$modelId]; + } + + if ( empty( $wgContentHandlers[$modelId] ) ) { + $handler = null; + + wfRunHooks( 'ContentHandlerForModelID', array( $modelId, &$handler ) ); + + if ( $handler === null ) { + throw new MWException( "No handler for model '$modelId'' registered in \$wgContentHandlers" ); + } + + if ( !( $handler instanceof ContentHandler ) ) { + throw new MWException( "ContentHandlerForModelID must supply a ContentHandler instance" ); + } + } else { + $class = $wgContentHandlers[$modelId]; + $handler = new $class( $modelId ); + + if ( !( $handler instanceof ContentHandler ) ) { + throw new MWException( "$class from \$wgContentHandlers is not compatible with ContentHandler" ); + } + } + + wfDebugLog( 'ContentHandler', 'Created handler for ' . $modelId + . ': ' . get_class( $handler ) ); + + ContentHandler::$handlers[$modelId] = $handler; + return ContentHandler::$handlers[$modelId]; + } + + /** + * Returns the localized name for a given content model. + * + * Model names are localized using system messages. Message keys + * have the form content-model-$name, where $name is getContentModelName( $id ). + * + * @param string $name The content model ID, as given by a CONTENT_MODEL_XXX + * constant or returned by Revision::getContentModel(). + * + * @return string The content format's localized name. + * @throws MWException if the model id isn't known. + */ + public static function getLocalizedName( $name ) { + $key = "content-model-$name"; + + $msg = wfMessage( $key ); + + return $msg->exists() ? $msg->plain() : $name; + } + + public static function getContentModels() { + global $wgContentHandlers; + + return array_keys( $wgContentHandlers ); + } + + public static function getAllContentFormats() { + global $wgContentHandlers; + + $formats = array(); + + foreach ( $wgContentHandlers as $model => $class ) { + $handler = ContentHandler::getForModelID( $model ); + $formats = array_merge( $formats, $handler->getSupportedFormats() ); + } + + $formats = array_unique( $formats ); + return $formats; + } + + // ------------------------------------------------------------------------ + + protected $mModelID; + protected $mSupportedFormats; + + /** + * Constructor, initializing the ContentHandler instance with its model ID + * and a list of supported formats. Values for the parameters are typically + * provided as literals by subclass's constructors. + * + * @param string $modelId (use CONTENT_MODEL_XXX constants). + * @param array $formats List for supported serialization formats + * (typically as MIME types) + */ + public function __construct( $modelId, $formats ) { + $this->mModelID = $modelId; + $this->mSupportedFormats = $formats; + + $this->mModelName = preg_replace( '/(Content)?Handler$/', '', get_class( $this ) ); + $this->mModelName = preg_replace( '/[_\\\\]/', '', $this->mModelName ); + $this->mModelName = strtolower( $this->mModelName ); + } + + /** + * Serializes a Content object of the type supported by this ContentHandler. + * + * @since 1.21 + * + * @param $content Content The Content object to serialize + * @param $format null|String The desired serialization format + * @return string Serialized form of the content + */ + abstract public function serializeContent( Content $content, $format = null ); + + /** + * Unserializes a Content object of the type supported by this ContentHandler. + * + * @since 1.21 + * + * @param string $blob serialized form of the content + * @param $format null|String the format used for serialization + * @return Content the Content object created by deserializing $blob + */ + abstract public function unserializeContent( $blob, $format = null ); + + /** + * Creates an empty Content object of the type supported by this + * ContentHandler. + * + * @since 1.21 + * + * @return Content + */ + abstract public function makeEmptyContent(); + + /** + * Creates a new Content object that acts as a redirect to the given page, + * or null of redirects are not supported by this content model. + * + * This default implementation always returns null. Subclasses supporting redirects + * must override this method. + * + * Note that subclasses that override this method to return a Content object + * should also override supportsRedirects() to return true. + * + * @since 1.21 + * + * @param Title $destination the page to redirect to. + * + * @return Content + */ + public function makeRedirectContent( Title $destination ) { + return null; + } + + /** + * Returns the model id that identifies the content model this + * ContentHandler can handle. Use with the CONTENT_MODEL_XXX constants. + * + * @since 1.21 + * + * @return String The model ID + */ + public function getModelID() { + return $this->mModelID; + } + + /** + * Throws an MWException if $model_id is not the ID of the content model + * supported by this ContentHandler. + * + * @since 1.21 + * + * @param string $model_id The model to check + * + * @throws MWException + */ + protected function checkModelID( $model_id ) { + if ( $model_id !== $this->mModelID ) { + throw new MWException( "Bad content model: " . + "expected {$this->mModelID} " . + "but got $model_id." ); + } + } + + /** + * Returns a list of serialization formats supported by the + * serializeContent() and unserializeContent() methods of this + * ContentHandler. + * + * @since 1.21 + * + * @return array of serialization formats as MIME type like strings + */ + public function getSupportedFormats() { + return $this->mSupportedFormats; + } + + /** + * The format used for serialization/deserialization by default by this + * ContentHandler. + * + * This default implementation will return the first element of the array + * of formats that was passed to the constructor. + * + * @since 1.21 + * + * @return string the name of the default serialization format as a MIME type + */ + public function getDefaultFormat() { + return $this->mSupportedFormats[0]; + } + + /** + * Returns true if $format is a serialization format supported by this + * ContentHandler, and false otherwise. + * + * Note that if $format is null, this method always returns true, because + * null means "use the default format". + * + * @since 1.21 + * + * @param string $format the serialization format to check + * @return bool + */ + public function isSupportedFormat( $format ) { + + if ( !$format ) { + return true; // this means "use the default" + } + + return in_array( $format, $this->mSupportedFormats ); + } + + /** + * Throws an MWException if isSupportedFormat( $format ) is not true. + * Convenient for checking whether a format provided as a parameter is + * actually supported. + * + * @param string $format the serialization format to check + * + * @throws MWException + */ + protected function checkFormat( $format ) { + if ( !$this->isSupportedFormat( $format ) ) { + throw new MWException( + "Format $format is not supported for content model " + . $this->getModelID() + ); + } + } + + /** + * Returns overrides for action handlers. + * Classes listed here will be used instead of the default one when + * (and only when) $wgActions[$action] === true. This allows subclasses + * to override the default action handlers. + * + * @since 1.21 + * + * @return Array + */ + public function getActionOverrides() { + return array(); + } + + /** + * Factory for creating an appropriate DifferenceEngine for this content model. + * + * @since 1.21 + * + * @param $context IContextSource context to use, anything else will be + * ignored + * @param $old Integer Old ID we want to show and diff with. + * @param int|string $new String either 'prev' or 'next'. + * @param $rcid Integer ??? FIXME (default 0) + * @param $refreshCache boolean If set, refreshes the diff cache + * @param $unhide boolean If set, allow viewing deleted revs + * + * @return DifferenceEngine + */ + public function createDifferenceEngine( IContextSource $context, + $old = 0, $new = 0, + $rcid = 0, # FIXME: use everywhere! + $refreshCache = false, $unhide = false + ) { + $diffEngineClass = $this->getDiffEngineClass(); + + return new $diffEngineClass( $context, $old, $new, $rcid, $refreshCache, $unhide ); + } + + /** + * Get the language in which the content of the given page is written. + * + * This default implementation just returns $wgContLang (except for pages in the MediaWiki namespace) + * + * Note that the pages language is not cacheable, since it may in some cases depend on user settings. + * + * Also note that the page language may or may not depend on the actual content of the page, + * that is, this method may load the content in order to determine the language. + * + * @since 1.21 + * + * @param Title $title the page to determine the language for. + * @param Content|null $content the page's content, if you have it handy, to avoid reloading it. + * + * @return Language the page's language + */ + public function getPageLanguage( Title $title, Content $content = null ) { + global $wgContLang, $wgLang; + $pageLang = $wgContLang; + + if ( $title->getNamespace() == NS_MEDIAWIKI ) { + // Parse mediawiki messages with correct target language + list( /* $unused */, $lang ) = MessageCache::singleton()->figureMessage( $title->getText() ); + $pageLang = wfGetLangObj( $lang ); + } + + wfRunHooks( 'PageContentLanguage', array( $title, &$pageLang, $wgLang ) ); + return wfGetLangObj( $pageLang ); + } + + /** + * Get the language in which the content of this page is written when + * viewed by user. Defaults to $this->getPageLanguage(), but if the user + * specified a preferred variant, the variant will be used. + * + * This default implementation just returns $this->getPageLanguage( $title, $content ) unless + * the user specified a preferred variant. + * + * Note that the pages view language is not cacheable, since it depends on user settings. + * + * Also note that the page language may or may not depend on the actual content of the page, + * that is, this method may load the content in order to determine the language. + * + * @since 1.21 + * + * @param Title $title the page to determine the language for. + * @param Content|null $content the page's content, if you have it handy, to avoid reloading it. + * + * @return Language the page's language for viewing + */ + public function getPageViewLanguage( Title $title, Content $content = null ) { + $pageLang = $this->getPageLanguage( $title, $content ); + + if ( $title->getNamespace() !== NS_MEDIAWIKI ) { + // If the user chooses a variant, the content is actually + // in a language whose code is the variant code. + $variant = $pageLang->getPreferredVariant(); + if ( $pageLang->getCode() !== $variant ) { + $pageLang = Language::factory( $variant ); + } + } + + return $pageLang; + } + + /** + * Determines whether the content type handled by this ContentHandler + * can be used on the given page. + * + * This default implementation always returns true. + * Subclasses may override this to restrict the use of this content model to specific locations, + * typically based on the namespace or some other aspect of the title, such as a special suffix + * (e.g. ".svg" for SVG content). + * + * @param Title $title the page's title. + * + * @return bool true if content of this kind can be used on the given page, false otherwise. + */ + public function canBeUsedOn( Title $title ) { + return true; + } + + /** + * Returns the name of the diff engine to use. + * + * @since 1.21 + * + * @return string + */ + protected function getDiffEngineClass() { + return 'DifferenceEngine'; + } + + /** + * Attempts to merge differences between three versions. + * Returns a new Content object for a clean merge and false for failure or + * a conflict. + * + * This default implementation always returns false. + * + * @since 1.21 + * + * @param $oldContent Content|string String + * @param $myContent Content|string String + * @param $yourContent Content|string String + * + * @return Content|Bool + */ + public function merge3( Content $oldContent, Content $myContent, Content $yourContent ) { + return false; + } + + /** + * Return an applicable auto-summary if one exists for the given edit. + * + * @since 1.21 + * + * @param $oldContent Content|null: the previous text of the page. + * @param $newContent Content|null: The submitted text of the page. + * @param int $flags Bit mask: a bit mask of flags submitted for the edit. + * + * @return string An appropriate auto-summary, or an empty string. + */ + public function getAutosummary( Content $oldContent = null, Content $newContent = null, $flags ) { + // Decide what kind of auto-summary is needed. + + // Redirect auto-summaries + + /** + * @var $ot Title + * @var $rt Title + */ + + $ot = !is_null( $oldContent ) ? $oldContent->getRedirectTarget() : null; + $rt = !is_null( $newContent ) ? $newContent->getRedirectTarget() : null; + + if ( is_object( $rt ) ) { + if ( !is_object( $ot ) + || !$rt->equals( $ot ) + || $ot->getFragment() != $rt->getFragment() ) + { + $truncatedtext = $newContent->getTextForSummary( + 250 + - strlen( wfMessage( 'autoredircomment' )->inContentLanguage()->text() ) + - strlen( $rt->getFullText() ) ); + + return wfMessage( 'autoredircomment', $rt->getFullText() ) + ->rawParams( $truncatedtext )->inContentLanguage()->text(); + } + } + + // New page auto-summaries + if ( $flags & EDIT_NEW && $newContent->getSize() > 0 ) { + // If they're making a new article, give its text, truncated, in + // the summary. + + $truncatedtext = $newContent->getTextForSummary( + 200 - strlen( wfMessage( 'autosumm-new' )->inContentLanguage()->text() ) ); + + return wfMessage( 'autosumm-new' )->rawParams( $truncatedtext ) + ->inContentLanguage()->text(); + } + + // Blanking auto-summaries + if ( !empty( $oldContent ) && $oldContent->getSize() > 0 && $newContent->getSize() == 0 ) { + return wfMessage( 'autosumm-blank' )->inContentLanguage()->text(); + } elseif ( !empty( $oldContent ) + && $oldContent->getSize() > 10 * $newContent->getSize() + && $newContent->getSize() < 500 ) + { + // Removing more than 90% of the article + + $truncatedtext = $newContent->getTextForSummary( + 200 - strlen( wfMessage( 'autosumm-replace' )->inContentLanguage()->text() ) ); + + return wfMessage( 'autosumm-replace' )->rawParams( $truncatedtext ) + ->inContentLanguage()->text(); + } + + // If we reach this point, there's no applicable auto-summary for our + // case, so our auto-summary is empty. + return ''; + } + + /** + * Auto-generates a deletion reason + * + * @since 1.21 + * + * @param $title Title: the page's title + * @param &$hasHistory Boolean: whether the page has a history + * @return mixed String containing deletion reason or empty string, or + * boolean false if no revision occurred + * + * @XXX &$hasHistory is extremely ugly, it's here because + * WikiPage::getAutoDeleteReason() and Article::generateReason() + * have it / want it. + */ + public function getAutoDeleteReason( Title $title, &$hasHistory ) { + $dbw = wfGetDB( DB_MASTER ); + + // Get the last revision + $rev = Revision::newFromTitle( $title ); + + if ( is_null( $rev ) ) { + return false; + } + + // Get the article's contents + $content = $rev->getContent(); + $blank = false; + + // If the page is blank, use the text from the previous revision, + // which can only be blank if there's a move/import/protect dummy + // revision involved + if ( !$content || $content->isEmpty() ) { + $prev = $rev->getPrevious(); + + if ( $prev ) { + $rev = $prev; + $content = $rev->getContent(); + $blank = true; + } + } + + $this->checkModelID( $rev->getContentModel() ); + + // Find out if there was only one contributor + // Only scan the last 20 revisions + $res = $dbw->select( 'revision', 'rev_user_text', + array( + 'rev_page' => $title->getArticleID(), + $dbw->bitAnd( 'rev_deleted', Revision::DELETED_USER ) . ' = 0' + ), + __METHOD__, + array( 'LIMIT' => 20 ) + ); + + if ( $res === false ) { + // This page has no revisions, which is very weird + return false; + } + + $hasHistory = ( $res->numRows() > 1 ); + $row = $dbw->fetchObject( $res ); + + if ( $row ) { // $row is false if the only contributor is hidden + $onlyAuthor = $row->rev_user_text; + // Try to find a second contributor + foreach ( $res as $row ) { + if ( $row->rev_user_text != $onlyAuthor ) { // Bug 22999 + $onlyAuthor = false; + break; + } + } + } else { + $onlyAuthor = false; + } + + // Generate the summary with a '$1' placeholder + if ( $blank ) { + // The current revision is blank and the one before is also + // blank. It's just not our lucky day + $reason = wfMessage( 'exbeforeblank', '$1' )->inContentLanguage()->text(); + } else { + if ( $onlyAuthor ) { + $reason = wfMessage( + 'excontentauthor', + '$1', + $onlyAuthor + )->inContentLanguage()->text(); + } else { + $reason = wfMessage( 'excontent', '$1' )->inContentLanguage()->text(); + } + } + + if ( $reason == '-' ) { + // Allow these UI messages to be blanked out cleanly + return ''; + } + + // Max content length = max comment length - length of the comment (excl. $1) + $text = $content ? $content->getTextForSummary( 255 - ( strlen( $reason ) - 2 ) ) : ''; + + // Now replace the '$1' placeholder + $reason = str_replace( '$1', $text, $reason ); + + return $reason; + } + + /** + * Get the Content object that needs to be saved in order to undo all revisions + * between $undo and $undoafter. Revisions must belong to the same page, + * must exist and must not be deleted. + * + * @since 1.21 + * + * @param $current Revision The current text + * @param $undo Revision The revision to undo + * @param $undoafter Revision Must be an earlier revision than $undo + * + * @return mixed String on success, false on failure + */ + public function getUndoContent( Revision $current, Revision $undo, Revision $undoafter ) { + $cur_content = $current->getContent(); + + if ( empty( $cur_content ) ) { + return false; // no page + } + + $undo_content = $undo->getContent(); + $undoafter_content = $undoafter->getContent(); + + $this->checkModelID( $cur_content->getModel() ); + $this->checkModelID( $undo_content->getModel() ); + $this->checkModelID( $undoafter_content->getModel() ); + + if ( $cur_content->equals( $undo_content ) ) { + // No use doing a merge if it's just a straight revert. + return $undoafter_content; + } + + $undone_content = $this->merge3( $undo_content, $undoafter_content, $cur_content ); + + return $undone_content; + } + + /** + * Get parser options suitable for rendering the primary article wikitext + * + * @param IContextSource|User|string $context One of the following: + * - IContextSource: Use the User and the Language of the provided + * context + * - User: Use the provided User object and $wgLang for the language, + * so use an IContextSource object if possible. + * - 'canonical': Canonical options (anonymous user with default + * preferences and content language). + * + * @param IContextSource|User|string $context + * + * @throws MWException + * @return ParserOptions + */ + public function makeParserOptions( $context ) { + global $wgContLang; + + if ( $context instanceof IContextSource ) { + $options = ParserOptions::newFromContext( $context ); + } elseif ( $context instanceof User ) { // settings per user (even anons) + $options = ParserOptions::newFromUser( $context ); + } elseif ( $context === 'canonical' ) { // canonical settings + $options = ParserOptions::newFromUserAndLang( new User, $wgContLang ); + } else { + throw new MWException( "Bad context for parser options: $context" ); + } + + $options->enableLimitReport(); // show inclusion/loop reports + $options->setTidy( true ); // fix bad HTML + + return $options; + } + + /** + * Returns true for content models that support caching using the + * ParserCache mechanism. See WikiPage::isParserCacheUsed(). + * + * @since 1.21 + * + * @return bool + */ + public function isParserCacheSupported() { + return false; + } + + /** + * Returns true if this content model supports sections. + * This default implementation returns false. + * + * Content models that return true here should also implement + * Content::getSection, Content::replaceSection, etc. to handle sections.. + * + * @return boolean whether sections are supported. + */ + public function supportsSections() { + return false; + } + + /** + * Returns true if this content model supports redirects. + * This default implementation returns false. + * + * Content models that return true here should also implement + * ContentHandler::makeRedirectContent to return a Content object. + * + * @return boolean whether redirects are supported. + */ + public function supportsRedirects() { + return false; + } + + /** + * Logs a deprecation warning, visible if $wgDevelopmentWarnings, but only if + * self::$enableDeprecationWarnings is set to true. + * + * @param string $func The name of the deprecated function + * @param string $version The version since the method is deprecated. Usually 1.21 + * for ContentHandler related stuff. + * @param string|bool $component: Component to which the function belongs. + * If false, it is assumed the function is in MediaWiki core. + * + * @see ContentHandler::$enableDeprecationWarnings + * @see wfDeprecated + */ + public static function deprecated( $func, $version, $component = false ) { + if ( self::$enableDeprecationWarnings ) { + wfDeprecated( $func, $version, $component, 3 ); + } + } + + /** + * Call a legacy hook that uses text instead of Content objects. + * Will log a warning when a matching hook function is registered. + * If the textual representation of the content is changed by the + * hook function, a new Content object is constructed from the new + * text. + * + * @param string $event event name + * @param array $args parameters passed to hook functions + * @param bool $warn whether to log a warning. + * Default to self::$enableDeprecationWarnings. + * May be set to false for testing. + * + * @return Boolean True if no handler aborted the hook + * + * @see ContentHandler::$enableDeprecationWarnings + */ + public static function runLegacyHooks( $event, $args = array(), + $warn = null ) { + + if ( $warn === null ) { + $warn = self::$enableDeprecationWarnings; + } + + if ( !Hooks::isRegistered( $event ) ) { + return true; // nothing to do here + } + + if ( $warn ) { + // Log information about which handlers are registered for the legacy hook, + // so we can find and fix them. + + $handlers = Hooks::getHandlers( $event ); + $handlerInfo = array(); + + wfSuppressWarnings(); + + foreach ( $handlers as $handler ) { + if ( is_array( $handler ) ) { + if ( is_object( $handler[0] ) ) { + $info = get_class( $handler[0] ); + } else { + $info = $handler[0]; + } + + if ( isset( $handler[1] ) ) { + $info .= '::' . $handler[1]; + } + } else if ( is_object( $handler ) ) { + $info = get_class( $handler[0] ); + $info .= '::on' . $event; + } else { + $info = $handler; + } + + $handlerInfo[] = $info; + } + + wfRestoreWarnings(); + + wfWarn( "Using obsolete hook $event via ContentHandler::runLegacyHooks()! Handlers: " . implode( ', ', $handlerInfo ), 2 ); + } + + // convert Content objects to text + $contentObjects = array(); + $contentTexts = array(); + + foreach ( $args as $k => $v ) { + if ( $v instanceof Content ) { + /* @var Content $v */ + + $contentObjects[$k] = $v; + + $v = $v->serialize(); + $contentTexts[ $k ] = $v; + $args[ $k ] = $v; + } + } + + // call the hook functions + $ok = wfRunHooks( $event, $args ); + + // see if the hook changed the text + foreach ( $contentTexts as $k => $orig ) { + /* @var Content $content */ + + $modified = $args[ $k ]; + $content = $contentObjects[$k]; + + if ( $modified !== $orig ) { + // text was changed, create updated Content object + $content = $content->getContentHandler()->unserializeContent( $modified ); + } + + $args[ $k ] = $content; + } + + return $ok; + } +} diff --git a/includes/content/CssContent.php b/includes/content/CssContent.php new file mode 100644 index 00000000..569d122d --- /dev/null +++ b/includes/content/CssContent.php @@ -0,0 +1,65 @@ +<?php +/** + * Content object for CSS pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Content object for CSS pages. + * + * @ingroup Content + */ +class CssContent extends TextContent { + public function __construct( $text ) { + parent::__construct( $text, CONTENT_MODEL_CSS ); + } + + /** + * Returns a Content object with pre-save transformations applied using + * Parser::preSaveTransform(). + * + * @param $title Title + * @param $user User + * @param $popts ParserOptions + * @return Content + */ + public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) { + global $wgParser; + // @todo: make pre-save transformation optional for script pages + + $text = $this->getNativeData(); + $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts ); + + return new CssContent( $pst ); + } + + protected function getHtml() { + $html = ""; + $html .= "<pre class=\"mw-code mw-css\" dir=\"ltr\">\n"; + $html .= $this->getHighlightHtml(); + $html .= "\n</pre>\n"; + + return $html; + } +} diff --git a/includes/content/CssContentHandler.php b/includes/content/CssContentHandler.php new file mode 100644 index 00000000..cb5a349d --- /dev/null +++ b/includes/content/CssContentHandler.php @@ -0,0 +1,67 @@ +<?php +/** + * Content handler for CSS pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @file + * @ingroup Content + */ + +/** + * Content handler for CSS pages. + * + * @since 1.21 + * @ingroup Content + */ +class CssContentHandler extends TextContentHandler { + + public function __construct( $modelId = CONTENT_MODEL_CSS ) { + parent::__construct( $modelId, array( CONTENT_FORMAT_CSS ) ); + } + + public function unserializeContent( $text, $format = null ) { + $this->checkFormat( $format ); + + return new CssContent( $text ); + } + + public function makeEmptyContent() { + return new CssContent( '' ); + } + + /** + * Returns the english language, because CSS is english, and should be handled as such. + * + * @return Language wfGetLangObj( 'en' ) + * + * @see ContentHandler::getPageLanguage() + */ + public function getPageLanguage( Title $title, Content $content = null ) { + return wfGetLangObj( 'en' ); + } + + /** + * Returns the english language, because CSS is english, and should be handled as such. + * + * @return Language wfGetLangObj( 'en' ) + * + * @see ContentHandler::getPageViewLanguage() + */ + public function getPageViewLanguage( Title $title, Content $content = null ) { + return wfGetLangObj( 'en' ); + } +} diff --git a/includes/content/JavaScriptContent.php b/includes/content/JavaScriptContent.php new file mode 100644 index 00000000..9cd947f9 --- /dev/null +++ b/includes/content/JavaScriptContent.php @@ -0,0 +1,66 @@ +<?php +/** + * Content for JavaScript pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Content for JavaScript pages. + * + * @ingroup Content + */ +class JavaScriptContent extends TextContent { + public function __construct( $text ) { + parent::__construct( $text, CONTENT_MODEL_JAVASCRIPT ); + } + + /** + * Returns a Content object with pre-save transformations applied using + * Parser::preSaveTransform(). + * + * @param Title $title + * @param User $user + * @param ParserOptions $popts + * @return Content + */ + public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) { + global $wgParser; + // @todo: make pre-save transformation optional for script pages + // See bug #32858 + + $text = $this->getNativeData(); + $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts ); + + return new JavaScriptContent( $pst ); + } + + protected function getHtml() { + $html = ""; + $html .= "<pre class=\"mw-code mw-js\" dir=\"ltr\">\n"; + $html .= $this->getHighlightHtml(); + $html .= "\n</pre>\n"; + + return $html; + } +} diff --git a/includes/content/JavaScriptContentHandler.php b/includes/content/JavaScriptContentHandler.php new file mode 100644 index 00000000..33fa9172 --- /dev/null +++ b/includes/content/JavaScriptContentHandler.php @@ -0,0 +1,67 @@ +<?php +/** + * Content handler for JavaScript pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @file + */ + +/** + * Content handler for JavaScript pages. + * + * @since 1.21 + * @ingroup Content + * @todo make ScriptContentHandler base class, do highlighting stuff there? + */ +class JavaScriptContentHandler extends TextContentHandler { + + public function __construct( $modelId = CONTENT_MODEL_JAVASCRIPT ) { + parent::__construct( $modelId, array( CONTENT_FORMAT_JAVASCRIPT ) ); + } + + public function unserializeContent( $text, $format = null ) { + $this->checkFormat( $format ); + + return new JavaScriptContent( $text ); + } + + public function makeEmptyContent() { + return new JavaScriptContent( '' ); + } + + /** + * Returns the english language, because JS is english, and should be handled as such. + * + * @return Language wfGetLangObj( 'en' ) + * + * @see ContentHandler::getPageLanguage() + */ + public function getPageLanguage( Title $title, Content $content = null ) { + return wfGetLangObj( 'en' ); + } + + /** + * Returns the english language, because JS is english, and should be handled as such. + * + * @return Language wfGetLangObj( 'en' ) + * + * @see ContentHandler::getPageViewLanguage() + */ + public function getPageViewLanguage( Title $title, Content $content = null ) { + return wfGetLangObj( 'en' ); + } +} diff --git a/includes/content/MessageContent.php b/includes/content/MessageContent.php new file mode 100644 index 00000000..b36b670c --- /dev/null +++ b/includes/content/MessageContent.php @@ -0,0 +1,158 @@ +<?php +/** + * Wrapper content object allowing to handle a system message as a Content object. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Wrapper allowing us to handle a system message as a Content object. + * Note that this is generally *not* used to represent content from the + * MediaWiki namespace, and that there is no MessageContentHandler. + * MessageContent is just intended as glue for wrapping a message programatically. + * + * @ingroup Content + */ +class MessageContent extends AbstractContent { + + /** + * @var Message + */ + protected $mMessage; + + /** + * @param Message|String $msg A Message object, or a message key + * @param array|null $params An optional array of message parameters + */ + public function __construct( $msg, $params = null ) { + # XXX: messages may be wikitext, html or plain text! and maybe even something else entirely. + parent::__construct( CONTENT_MODEL_WIKITEXT ); + + if ( is_string( $msg ) ) { + $this->mMessage = wfMessage( $msg ); + } else { + $this->mMessage = clone $msg; + } + + if ( $params ) { + $this->mMessage = $this->mMessage->params( $params ); + } + } + + /** + * Returns the message as rendered HTML + * + * @return string The message text, parsed into html + */ + public function getHtml() { + return $this->mMessage->parse(); + } + + /** + * Returns the message as rendered HTML + * + * @return string The message text, parsed into html + */ + public function getWikitext() { + return $this->mMessage->text(); + } + + /** + * Returns the message object, with any parameters already substituted. + * + * @return Message The message object. + */ + public function getNativeData() { + //NOTE: Message objects are mutable. Cloning here makes MessageContent immutable. + return clone $this->mMessage; + } + + /** + * @see Content::getTextForSearchIndex + */ + public function getTextForSearchIndex() { + return $this->mMessage->plain(); + } + + /** + * @see Content::getWikitextForTransclusion + */ + public function getWikitextForTransclusion() { + return $this->getWikitext(); + } + + /** + * @see Content::getTextForSummary + */ + public function getTextForSummary( $maxlength = 250 ) { + return substr( $this->mMessage->plain(), 0, $maxlength ); + } + + /** + * @see Content::getSize + * + * @return int + */ + public function getSize() { + return strlen( $this->mMessage->plain() ); + } + + /** + * @see Content::copy + * + * @return Content. A copy of this object + */ + public function copy() { + // MessageContent is immutable (because getNativeData() returns a clone of the Message object) + return $this; + } + + /** + * @see Content::isCountable + * + * @return bool false + */ + public function isCountable( $hasLinks = null ) { + return false; + } + + /** + * @see Content::getParserOutput + * + * @return ParserOutput + */ + public function getParserOutput( + Title $title, $revId = null, + ParserOptions $options = null, $generateHtml = true + ) { + + if ( $generateHtml ) { + $html = $this->getHtml(); + } else { + $html = ''; + } + + $po = new ParserOutput( $html ); + return $po; + } +} diff --git a/includes/content/TextContent.php b/includes/content/TextContent.php new file mode 100644 index 00000000..8fafcb63 --- /dev/null +++ b/includes/content/TextContent.php @@ -0,0 +1,286 @@ +<?php +/** + * Content object implementation for representing flat text. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Content object implementation for representing flat text. + * + * TextContent instances are immutable + * + * @ingroup Content + */ +class TextContent extends AbstractContent { + + public function __construct( $text, $model_id = CONTENT_MODEL_TEXT ) { + parent::__construct( $model_id ); + + if ( $text === null || $text === false ) { + wfWarn( "TextContent constructed with \$text = " . var_export( $text, true ) . "! " + . "This may indicate an error in the caller's scope." ); + + $text = ''; + } + + if ( !is_string( $text ) ) { + throw new MWException( "TextContent expects a string in the constructor." ); + } + + $this->mText = $text; + } + + public function copy() { + return $this; # NOTE: this is ok since TextContent are immutable. + } + + public function getTextForSummary( $maxlength = 250 ) { + global $wgContLang; + + $text = $this->getNativeData(); + + $truncatedtext = $wgContLang->truncate( + preg_replace( "/[\n\r]/", ' ', $text ), + max( 0, $maxlength ) ); + + return $truncatedtext; + } + + /** + * returns the text's size in bytes. + * + * @return int The size + */ + public function getSize() { + $text = $this->getNativeData(); + return strlen( $text ); + } + + /** + * Returns true if this content is not a redirect, and $wgArticleCountMethod + * is "any". + * + * @param bool $hasLinks if it is known whether this content contains links, + * provide this information here, to avoid redundant parsing to find out. + * + * @return bool True if the content is countable + */ + public function isCountable( $hasLinks = null ) { + global $wgArticleCountMethod; + + if ( $this->isRedirect() ) { + return false; + } + + if ( $wgArticleCountMethod === 'any' ) { + return true; + } + + return false; + } + + /** + * Returns the text represented by this Content object, as a string. + * + * @return string: the raw text + */ + public function getNativeData() { + $text = $this->mText; + return $text; + } + + /** + * Returns the text represented by this Content object, as a string. + * + * @return string: the raw text + */ + public function getTextForSearchIndex() { + return $this->getNativeData(); + } + + /** + * Returns attempts to convert this content object to wikitext, + * and then returns the text string. The conversion may be lossy. + * + * @note: this allows any text-based content to be transcluded as if it was wikitext. + * + * @return string|false: the raw text, or null if the conversion failed + */ + public function getWikitextForTransclusion() { + $wikitext = $this->convert( CONTENT_MODEL_WIKITEXT, 'lossy' ); + + if ( $wikitext ) { + return $wikitext->getNativeData(); + } else { + return false; + } + } + + /** + * Returns a Content object with pre-save transformations applied. + * This implementation just trims trailing whitespace. + * + * @param $title Title + * @param $user User + * @param $popts ParserOptions + * @return Content + */ + public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) { + $text = $this->getNativeData(); + $pst = rtrim( $text ); + + return ( $text === $pst ) ? $this : new WikitextContent( $pst ); + } + + /** + * Diff this content object with another content object. + * + * @since 1.21diff + * + * @param $that Content: The other content object to compare this content + * object to. + * @param $lang Language: The language object to use for text segmentation. + * If not given, $wgContentLang is used. + * + * @return DiffResult: A diff representing the changes that would have to be + * made to this content object to make it equal to $that. + */ + public function diff( Content $that, Language $lang = null ) { + global $wgContLang; + + $this->checkModelID( $that->getModel() ); + + # @todo: could implement this in DifferenceEngine and just delegate here? + + if ( !$lang ) { + $lang = $wgContLang; + } + + $otext = $this->getNativeData(); + $ntext = $this->getNativeData(); + + # Note: Use native PHP diff, external engines don't give us abstract output + $ota = explode( "\n", $lang->segmentForDiff( $otext ) ); + $nta = explode( "\n", $lang->segmentForDiff( $ntext ) ); + + $diff = new Diff( $ota, $nta ); + return $diff; + } + + /** + * Returns a generic ParserOutput object, wrapping the HTML returned by + * getHtml(). + * + * @param $title Title Context title for parsing + * @param int|null $revId Revision ID (for {{REVISIONID}}) + * @param $options ParserOptions|null Parser options + * @param bool $generateHtml Whether or not to generate HTML + * + * @return ParserOutput representing the HTML form of the text + */ + public function getParserOutput( Title $title, + $revId = null, + ParserOptions $options = null, $generateHtml = true + ) { + global $wgParser, $wgTextModelsToParse; + + if ( !$options ) { + //NOTE: use canonical options per default to produce cacheable output + $options = $this->getContentHandler()->makeParserOptions( 'canonical' ); + } + + if ( in_array( $this->getModel(), $wgTextModelsToParse ) ) { + // parse just to get links etc into the database + $po = $wgParser->parse( $this->getNativeData(), $title, $options, true, true, $revId ); + } else { + $po = new ParserOutput(); + } + + if ( $generateHtml ) { + $html = $this->getHtml(); + } else { + $html = ''; + } + + $po->setText( $html ); + return $po; + } + + /** + * Generates an HTML version of the content, for display. Used by + * getParserOutput() to construct a ParserOutput object. + * + * This default implementation just calls getHighlightHtml(). Content + * models that have another mapping to HTML (as is the case for markup + * languages like wikitext) should override this method to generate the + * appropriate HTML. + * + * @return string An HTML representation of the content + */ + protected function getHtml() { + return $this->getHighlightHtml(); + } + + /** + * Generates a syntax-highlighted version of the content, as HTML. + * Used by the default implementation of getHtml(). + * + * @return string an HTML representation of the content's markup + */ + protected function getHighlightHtml() { + # TODO: make Highlighter interface, use highlighter here, if available + return htmlspecialchars( $this->getNativeData() ); + } + + /** + * @see Content::convert() + * + * This implementation provides lossless conversion between content models based + * on TextContent. + * + * @param string $toModel the desired content model, use the CONTENT_MODEL_XXX flags. + * @param string $lossy flag, set to "lossy" to allow lossy conversion. If lossy conversion is + * not allowed, full round-trip conversion is expected to work without losing information. + * + * @return Content|bool A content object with the content model $toModel, or false if + * that conversion is not supported. + */ + public function convert( $toModel, $lossy = '' ) { + $converted = parent::convert( $toModel, $lossy ); + + if ( $converted !== false ) { + return $converted; + } + + $toHandler = ContentHandler::getForModelID( $toModel ); + + if ( $toHandler instanceof TextContentHandler ) { + //NOTE: ignore content serialization format - it's just text anyway. + $text = $this->getNativeData(); + $converted = $toHandler->unserializeContent( $text ); + } + + return $converted; + } +} diff --git a/includes/content/TextContentHandler.php b/includes/content/TextContentHandler.php new file mode 100644 index 00000000..e7f41e18 --- /dev/null +++ b/includes/content/TextContentHandler.php @@ -0,0 +1,115 @@ +<?php +/** + * Base content handler class for flat text contents. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + */ + +/** + * Base content handler implementation for flat text contents. + * + * @ingroup Content + */ +class TextContentHandler extends ContentHandler { + + public function __construct( $modelId = CONTENT_MODEL_TEXT, $formats = array( CONTENT_FORMAT_TEXT ) ) { + parent::__construct( $modelId, $formats ); + } + + /** + * Returns the content's text as-is. + * + * @param $content Content + * @param $format string|null + * @return mixed + */ + public function serializeContent( Content $content, $format = null ) { + $this->checkFormat( $format ); + return $content->getNativeData(); + } + + /** + * Attempts to merge differences between three versions. Returns a new + * Content object for a clean merge and false for failure or a conflict. + * + * All three Content objects passed as parameters must have the same + * content model. + * + * This text-based implementation uses wfMerge(). + * + * @param $oldContent Content|string String + * @param $myContent Content|string String + * @param $yourContent Content|string String + * + * @return Content|Bool + */ + public function merge3( Content $oldContent, Content $myContent, Content $yourContent ) { + $this->checkModelID( $oldContent->getModel() ); + $this->checkModelID( $myContent->getModel() ); + $this->checkModelID( $yourContent->getModel() ); + + $format = $this->getDefaultFormat(); + + $old = $this->serializeContent( $oldContent, $format ); + $mine = $this->serializeContent( $myContent, $format ); + $yours = $this->serializeContent( $yourContent, $format ); + + $ok = wfMerge( $old, $mine, $yours, $result ); + + if ( !$ok ) { + return false; + } + + if ( !$result ) { + return $this->makeEmptyContent(); + } + + $mergedContent = $this->unserializeContent( $result, $format ); + return $mergedContent; + } + + /** + * Unserializes a Content object of the type supported by this ContentHandler. + * + * @since 1.21 + * + * @param $text string serialized form of the content + * @param $format null|String the format used for serialization + * + * @return Content the TextContent object wrapping $text + */ + public function unserializeContent( $text, $format = null ) { + $this->checkFormat( $format ); + + return new TextContent( $text ); + } + + /** + * Creates an empty TextContent object. + * + * @since 1.21 + * + * @return Content + */ + public function makeEmptyContent() { + return new TextContent( '' ); + } +} diff --git a/includes/content/WikitextContent.php b/includes/content/WikitextContent.php new file mode 100644 index 00000000..8be4ebab --- /dev/null +++ b/includes/content/WikitextContent.php @@ -0,0 +1,322 @@ +<?php +/** + * Content object for wiki text pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + * + * @author Daniel Kinzler + */ + +/** + * Content object for wiki text pages. + * + * @ingroup Content + */ +class WikitextContent extends TextContent { + + public function __construct( $text ) { + parent::__construct( $text, CONTENT_MODEL_WIKITEXT ); + } + + /** + * @see Content::getSection() + */ + public function getSection( $section ) { + global $wgParser; + + $text = $this->getNativeData(); + $sect = $wgParser->getSection( $text, $section, false ); + + if ( $sect === false ) { + return false; + } else { + return new WikitextContent( $sect ); + } + } + + /** + * @see Content::replaceSection() + */ + public function replaceSection( $section, Content $with, $sectionTitle = '' ) { + wfProfileIn( __METHOD__ ); + + $myModelId = $this->getModel(); + $sectionModelId = $with->getModel(); + + if ( $sectionModelId != $myModelId ) { + throw new MWException( "Incompatible content model for section: " . + "document uses $myModelId but " . + "section uses $sectionModelId." ); + } + + $oldtext = $this->getNativeData(); + $text = $with->getNativeData(); + + if ( $section === '' ) { + wfProfileOut( __METHOD__ ); + return $with; # XXX: copy first? + } if ( $section == 'new' ) { + # Inserting a new section + $subject = $sectionTitle ? wfMessage( 'newsectionheaderdefaultlevel' ) + ->rawParams( $sectionTitle )->inContentLanguage()->text() . "\n\n" : ''; + if ( wfRunHooks( 'PlaceNewSection', array( $this, $oldtext, $subject, &$text ) ) ) { + $text = strlen( trim( $oldtext ) ) > 0 + ? "{$oldtext}\n\n{$subject}{$text}" + : "{$subject}{$text}"; + } + } else { + # Replacing an existing section; roll out the big guns + global $wgParser; + + $text = $wgParser->replaceSection( $oldtext, $section, $text ); + } + + $newContent = new WikitextContent( $text ); + + wfProfileOut( __METHOD__ ); + return $newContent; + } + + /** + * Returns a new WikitextContent object with the given section heading + * prepended. + * + * @param $header string + * @return Content + */ + public function addSectionHeader( $header ) { + $text = wfMessage( 'newsectionheaderdefaultlevel' ) + ->rawParams( $header )->inContentLanguage()->text(); + $text .= "\n\n"; + $text .= $this->getNativeData(); + + return new WikitextContent( $text ); + } + + /** + * Returns a Content object with pre-save transformations applied using + * Parser::preSaveTransform(). + * + * @param $title Title + * @param $user User + * @param $popts ParserOptions + * @return Content + */ + public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) { + global $wgParser; + + $text = $this->getNativeData(); + $pst = $wgParser->preSaveTransform( $text, $title, $user, $popts ); + rtrim( $pst ); + + return ( $text === $pst ) ? $this : new WikitextContent( $pst ); + } + + /** + * Returns a Content object with preload transformations applied (or this + * object if no transformations apply). + * + * @param $title Title + * @param $popts ParserOptions + * @return Content + */ + public function preloadTransform( Title $title, ParserOptions $popts ) { + global $wgParser; + + $text = $this->getNativeData(); + $plt = $wgParser->getPreloadText( $text, $title, $popts ); + + return new WikitextContent( $plt ); + } + + /** + * Implement redirect extraction for wikitext. + * + * @return null|Title + * + * @note: migrated here from Title::newFromRedirectInternal() + * + * @see Content::getRedirectTarget + * @see AbstractContent::getRedirectTarget + */ + public function getRedirectTarget() { + global $wgMaxRedirects; + if ( $wgMaxRedirects < 1 ) { + // redirects are disabled, so quit early + return null; + } + $redir = MagicWord::get( 'redirect' ); + $text = trim( $this->getNativeData() ); + if ( $redir->matchStartAndRemove( $text ) ) { + // Extract the first link and see if it's usable + // Ensure that it really does come directly after #REDIRECT + // Some older redirects included a colon, so don't freak about that! + $m = array(); + if ( preg_match( '!^\s*:?\s*\[{2}(.*?)(?:\|.*?)?\]{2}!', $text, $m ) ) { + // Strip preceding colon used to "escape" categories, etc. + // and URL-decode links + if ( strpos( $m[1], '%' ) !== false ) { + // Match behavior of inline link parsing here; + $m[1] = rawurldecode( ltrim( $m[1], ':' ) ); + } + $title = Title::newFromText( $m[1] ); + // If the title is a redirect to bad special pages or is invalid, return null + if ( !$title instanceof Title || !$title->isValidRedirectTarget() ) { + return null; + } + return $title; + } + } + return null; + } + + /** + * @see Content::updateRedirect() + * + * This implementation replaces the first link on the page with the given new target + * if this Content object is a redirect. Otherwise, this method returns $this. + * + * @since 1.21 + * + * @param Title $target + * + * @return Content a new Content object with the updated redirect (or $this if this Content object isn't a redirect) + */ + public function updateRedirect( Title $target ) { + if ( !$this->isRedirect() ) { + return $this; + } + + # Fix the text + # Remember that redirect pages can have categories, templates, etc., + # so the regex has to be fairly general + $newText = preg_replace( '/ \[ \[ [^\]]* \] \] /x', + '[[' . $target->getFullText() . ']]', + $this->getNativeData(), 1 ); + + return new WikitextContent( $newText ); + } + + /** + * Returns true if this content is not a redirect, and this content's text + * is countable according to the criteria defined by $wgArticleCountMethod. + * + * @param bool $hasLinks if it is known whether this content contains + * links, provide this information here, to avoid redundant parsing to + * find out (default: null). + * @param $title Title: (default: null) + * + * @internal param \IContextSource $context context for parsing if necessary + * + * @return bool True if the content is countable + */ + public function isCountable( $hasLinks = null, Title $title = null ) { + global $wgArticleCountMethod; + + if ( $this->isRedirect() ) { + return false; + } + + $text = $this->getNativeData(); + + switch ( $wgArticleCountMethod ) { + case 'any': + return true; + case 'comma': + return strpos( $text, ',' ) !== false; + case 'link': + if ( $hasLinks === null ) { # not known, find out + if ( !$title ) { + $context = RequestContext::getMain(); + $title = $context->getTitle(); + } + + $po = $this->getParserOutput( $title, null, null, false ); + $links = $po->getLinks(); + $hasLinks = !empty( $links ); + } + + return $hasLinks; + } + + return false; + } + + public function getTextForSummary( $maxlength = 250 ) { + $truncatedtext = parent::getTextForSummary( $maxlength ); + + # clean up unfinished links + # XXX: make this optional? wasn't there in autosummary, but required for + # deletion summary. + $truncatedtext = preg_replace( '/\[\[([^\]]*)\]?$/', '$1', $truncatedtext ); + + return $truncatedtext; + } + + /** + * Returns a ParserOutput object resulting from parsing the content's text + * using $wgParser. + * + * @since 1.21 + * + * @param $title Title + * @param int $revId Revision to pass to the parser (default: null) + * @param $options ParserOptions (default: null) + * @param bool $generateHtml (default: false) + * + * @internal param \IContextSource|null $context + * @return ParserOutput representing the HTML form of the text + */ + public function getParserOutput( Title $title, + $revId = null, + ParserOptions $options = null, $generateHtml = true + ) { + global $wgParser; + + if ( !$options ) { + //NOTE: use canonical options per default to produce cacheable output + $options = $this->getContentHandler()->makeParserOptions( 'canonical' ); + } + + $po = $wgParser->parse( $this->getNativeData(), $title, $options, true, true, $revId ); + return $po; + } + + protected function getHtml() { + throw new MWException( + "getHtml() not implemented for wikitext. " + . "Use getParserOutput()->getText()." + ); + } + + /** + * @see Content::matchMagicWord() + * + * This implementation calls $word->match() on the this TextContent object's text. + * + * @param MagicWord $word + * + * @return bool whether this Content object matches the given magic word. + */ + public function matchMagicWord( MagicWord $word ) { + return $word->match( $this->getNativeData() ); + } +} diff --git a/includes/content/WikitextContentHandler.php b/includes/content/WikitextContentHandler.php new file mode 100644 index 00000000..e1dcc668 --- /dev/null +++ b/includes/content/WikitextContentHandler.php @@ -0,0 +1,98 @@ +<?php +/** + * Content handler for wiki text pages. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License along + * with this program; if not, write to the Free Software Foundation, Inc., + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. + * http://www.gnu.org/copyleft/gpl.html + * + * @since 1.21 + * + * @file + * @ingroup Content + */ + +/** + * Content handler for wiki text pages. + * + * @ingroup Content + */ +class WikitextContentHandler extends TextContentHandler { + + public function __construct( $modelId = CONTENT_MODEL_WIKITEXT ) { + parent::__construct( $modelId, array( CONTENT_FORMAT_WIKITEXT ) ); + } + + public function unserializeContent( $text, $format = null ) { + $this->checkFormat( $format ); + + return new WikitextContent( $text ); + } + + /** + * @see ContentHandler::makeEmptyContent + * + * @return Content + */ + public function makeEmptyContent() { + return new WikitextContent( '' ); + } + + /** + * Returns a WikitextContent object representing a redirect to the given destination page. + * + * @see ContentHandler::makeRedirectContent + * + * @param Title $destination the page to redirect to. + * + * @return Content + */ + public function makeRedirectContent( Title $destination ) { + $mwRedir = MagicWord::get( 'redirect' ); + $redirectText = $mwRedir->getSynonym( 0 ) . ' [[' . $destination->getPrefixedText() . ']]'; + + return new WikitextContent( $redirectText ); + } + + /** + * Returns true because wikitext supports redirects. + * + * @see ContentHandler::supportsRedirects + * + * @return boolean whether redirects are supported. + */ + public function supportsRedirects() { + return true; + } + + /** + * Returns true because wikitext supports sections. + * + * @return boolean whether sections are supported. + */ + public function supportsSections() { + return true; + } + + /** + * Returns true, because wikitext supports caching using the + * ParserCache mechanism. + * + * @since 1.21 + * @return bool + */ + public function isParserCacheSupported() { + return true; + } +} |