includes/content/TextContent.php

   1 <?php
   2 /**
   3  * Content object implementation for representing flat text.
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @since 1.21
  21  *
  22  * @file
  23  * @ingroup Content
  24  *
  25  * @author Daniel Kinzler
  26  */
  27
  28 use MediaWiki\MediaWikiServices;
  29
  30 /**
  31  * Content object implementation for representing flat text.
  32  *
  33  * TextContent instances are immutable
  34  *
  35  * @ingroup Content
  36  */
  37 class TextContent extends AbstractContent {
  38
  39         /**
  40          * @var string
  41          */
  42         protected $mText;
  43
  44         /**
  45          * @param string $text
  46          * @param string $model_id
  47          * @throws MWException
  48          */
  49         public function __construct( $text, $model_id = CONTENT_MODEL_TEXT ) {
  50                 parent::__construct( $model_id );
  51
  52                 if ( $text === null || $text === false ) {
  53                         wfWarn( "TextContent constructed with \$text = " . var_export( $text, true ) . "! "
  54                                 . "This may indicate an error in the caller's scope.", 2 );
  55
  56                         $text = '';
  57                 }
  58
  59                 if ( !is_string( $text ) ) {
  60                         throw new MWException( "TextContent expects a string in the constructor." );
  61                 }
  62
  63                 $this->mText = $text;
  64         }
  65
  66         /**
  67          * @note Mutable subclasses MUST override this to return a copy!
  68          *
  69          * @return Content $this
  70          */
  71         public function copy() {
  72                 return $this; # NOTE: this is ok since TextContent are immutable.
  73         }
  74
  75         public function getTextForSummary( $maxlength = 250 ) {
  76                 $text = $this->getNativeData();
  77
  78                 $truncatedtext = MediaWikiServices::getInstance()->getContentLanguage()->
  79                         truncateForDatabase( preg_replace( "/[\n\r]/", ' ', $text ), max( 0, $maxlength ) );
  80
  81                 return $truncatedtext;
  82         }
  83
  84         /**
  85          * Returns the text's size in bytes.
  86          *
  87          * @return int
  88          */
  89         public function getSize() {
  90                 $text = $this->getNativeData();
  91
  92                 return strlen( $text );
  93         }
  94
  95         /**
  96          * Returns true if this content is not a redirect, and $wgArticleCountMethod
  97          * is "any".
  98          *
  99          * @param bool|null $hasLinks If it is known whether this content contains links,
 100          * provide this information here, to avoid redundant parsing to find out.
 101          *
 102          * @return bool
 103          */
 104         public function isCountable( $hasLinks = null ) {
 105                 global $wgArticleCountMethod;
 106
 107                 if ( $this->isRedirect() ) {
 108                         return false;
 109                 }
 110
 111                 if ( $wgArticleCountMethod === 'any' ) {
 112                         return true;
 113                 }
 114
 115                 return false;
 116         }
 117
 118         /**
 119          * Returns the text represented by this Content object, as a string.
 120          *
 121          * @return string The raw text.
 122          */
 123         public function getNativeData() {
 124                 return $this->mText;
 125         }
 126
 127         /**
 128          * Returns the text represented by this Content object, as a string.
 129          *
 130          * @return string The raw text.
 131          */
 132         public function getTextForSearchIndex() {
 133                 return $this->getNativeData();
 134         }
 135
 136         /**
 137          * Returns attempts to convert this content object to wikitext,
 138          * and then returns the text string. The conversion may be lossy.
 139          *
 140          * @note this allows any text-based content to be transcluded as if it was wikitext.
 141          *
 142          * @return string|bool The raw text, or false if the conversion failed.
 143          */
 144         public function getWikitextForTransclusion() {
 145                 $wikitext = $this->convert( CONTENT_MODEL_WIKITEXT, 'lossy' );
 146
 147                 if ( $wikitext ) {
 148                         return $wikitext->getNativeData();
 149                 } else {
 150                         return false;
 151                 }
 152         }
 153
 154         /**
 155          * Do a "\r\n" -> "\n" and "\r" -> "\n" transformation
 156          * as well as trim trailing whitespace
 157          *
 158          * This was formerly part of Parser::preSaveTransform, but
 159          * for non-wikitext content models they probably still want
 160          * to normalize line endings without all of the other PST
 161          * changes.
 162          *
 163          * @since 1.28
 164          * @param string $text
 165          * @return string
 166          */
 167         public static function normalizeLineEndings( $text ) {
 168                 return str_replace( [ "\r\n", "\r" ], "\n", rtrim( $text ) );
 169         }
 170
 171         /**
 172          * Returns a Content object with pre-save transformations applied.
 173          *
 174          * At a minimum, subclasses should make sure to call TextContent::normalizeLineEndings()
 175          * either directly or part of Parser::preSaveTransform().
 176          *
 177          * @param Title $title
 178          * @param User $user
 179          * @param ParserOptions $popts
 180          *
 181          * @return Content
 182          */
 183         public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) {
 184                 $text = $this->getNativeData();
 185                 $pst = self::normalizeLineEndings( $text );
 186
 187                 return ( $text === $pst ) ? $this : new static( $pst, $this->getModel() );
 188         }
 189
 190         /**
 191          * Diff this content object with another content object.
 192          *
 193          * @since 1.21
 194          *
 195          * @param Content $that The other content object to compare this content object to.
 196          * @param Language|null $lang The language object to use for text segmentation.
 197          *    If not given, the content language is used.
 198          *
 199          * @return Diff A diff representing the changes that would have to be
 200          *    made to this content object to make it equal to $that.
 201          */
 202         public function diff( Content $that, Language $lang = null ) {
 203                 $this->checkModelID( $that->getModel() );
 204
 205                 // @todo could implement this in DifferenceEngine and just delegate here?
 206
 207                 if ( !$lang ) {
 208                         $lang = MediaWikiServices::getInstance()->getContentLanguage();
 209                 }
 210
 211                 $otext = $this->getNativeData();
 212                 $ntext = $that->getNativeData();
 213
 214                 # Note: Use native PHP diff, external engines don't give us abstract output
 215                 $ota = explode( "\n", $lang->segmentForDiff( $otext ) );
 216                 $nta = explode( "\n", $lang->segmentForDiff( $ntext ) );
 217
 218                 $diff = new Diff( $ota, $nta );
 219
 220                 return $diff;
 221         }
 222
 223         /**
 224          * Fills the provided ParserOutput object with information derived from the content.
 225          * Unless $generateHtml was false, this includes an HTML representation of the content
 226          * provided by getHtml().
 227          *
 228          * For content models listed in $wgTextModelsToParse, this method will call the MediaWiki
 229          * wikitext parser on the text to extract any (wikitext) links, magic words, etc.
 230          *
 231          * Subclasses may override this to provide custom content processing.
 232          * For custom HTML generation alone, it is sufficient to override getHtml().
 233          *
 234          * @param Title $title Context title for parsing
 235          * @param int $revId Revision ID (for {{REVISIONID}})
 236          * @param ParserOptions $options
 237          * @param bool $generateHtml Whether or not to generate HTML
 238          * @param ParserOutput &$output The output object to fill (reference).
 239          */
 240         protected function fillParserOutput( Title $title, $revId,
 241                 ParserOptions $options, $generateHtml, ParserOutput &$output
 242         ) {
 243                 global $wgParser, $wgTextModelsToParse;
 244
 245                 if ( in_array( $this->getModel(), $wgTextModelsToParse ) ) {
 246                         // parse just to get links etc into the database, HTML is replaced below.
 247                         $output = $wgParser->parse( $this->getNativeData(), $title, $options, true, true, $revId );
 248                 }
 249
 250                 if ( $generateHtml ) {
 251                         $html = $this->getHtml();
 252                 } else {
 253                         $html = '';
 254                 }
 255
 256                 $output->clearWrapperDivClass();
 257                 $output->setText( $html );
 258         }
 259
 260         /**
 261          * Generates an HTML version of the content, for display. Used by
 262          * fillParserOutput() to provide HTML for the ParserOutput object.
 263          *
 264          * Subclasses may override this to provide a custom HTML rendering.
 265          * If further information is to be derived from the content (such as
 266          * categories), the fillParserOutput() method can be overridden instead.
 267          *
 268          * For backwards-compatibility, this default implementation just calls
 269          * getHighlightHtml().
 270          *
 271          * @return string An HTML representation of the content
 272          */
 273         protected function getHtml() {
 274                 return $this->getHighlightHtml();
 275         }
 276
 277         /**
 278          * Generates an HTML version of the content, for display.
 279          *
 280          * This default implementation returns an HTML-escaped version
 281          * of the raw text content.
 282          *
 283          * @note The functionality of this method should really be implemented
 284          * in getHtml(), and subclasses should override getHtml() if needed.
 285          * getHighlightHtml() is kept around for backward compatibility with
 286          * extensions that already override it.
 287          *
 288          * @deprecated since 1.24. Use getHtml() instead. In particular, subclasses overriding
 289          *     getHighlightHtml() should override getHtml() instead.
 290          *
 291          * @return string An HTML representation of the content
 292          */
 293         protected function getHighlightHtml() {
 294                 return htmlspecialchars( $this->getNativeData() );
 295         }
 296
 297         /**
 298          * This implementation provides lossless conversion between content models based
 299          * on TextContent.
 300          *
 301          * @param string $toModel The desired content model, use the CONTENT_MODEL_XXX flags.
 302          * @param string $lossy Flag, set to "lossy" to allow lossy conversion. If lossy conversion is not
 303          *     allowed, full round-trip conversion is expected to work without losing information.
 304          *
 305          * @return Content|bool A content object with the content model $toModel, or false if that
 306          *     conversion is not supported.
 307          *
 308          * @see Content::convert()
 309          */
 310         public function convert( $toModel, $lossy = '' ) {
 311                 $converted = parent::convert( $toModel, $lossy );
 312
 313                 if ( $converted !== false ) {
 314                         return $converted;
 315                 }
 316
 317                 $toHandler = ContentHandler::getForModelID( $toModel );
 318
 319                 if ( $toHandler instanceof TextContentHandler ) {
 320                         // NOTE: ignore content serialization format - it's just text anyway.
 321                         $text = $this->getNativeData();
 322                         $converted = $toHandler->unserializeContent( $text );
 323                 }
 324
 325                 return $converted;
 326         }
 327
 328 }