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 /**
  29  * Content object implementation for representing flat text.
  30  *
  31  * TextContent instances are immutable
  32  *
  33  * @ingroup Content
  34  */
  35 class TextContent extends AbstractContent {
  36
  37         public function __construct( $text, $model_id = CONTENT_MODEL_TEXT ) {
  38                 parent::__construct( $model_id );
  39
  40                 if ( $text === null || $text === false ) {
  41                         wfWarn( "TextContent constructed with \$text = " . var_export( $text, true ) . "! "
  42                                         . "This may indicate an error in the caller's scope." );
  43
  44                         $text = '';
  45                 }
  46
  47                 if ( !is_string( $text ) ) {
  48                         throw new MWException( "TextContent expects a string in the constructor." );
  49                 }
  50
  51                 $this->mText = $text;
  52         }
  53
  54         public function copy() {
  55                 return $this; # NOTE: this is ok since TextContent are immutable.
  56         }
  57
  58         public function getTextForSummary( $maxlength = 250 ) {
  59                 global $wgContLang;
  60
  61                 $text = $this->getNativeData();
  62
  63                 $truncatedtext = $wgContLang->truncate(
  64                         preg_replace( "/[\n\r]/", ' ', $text ),
  65                         max( 0, $maxlength ) );
  66
  67                 return $truncatedtext;
  68         }
  69
  70         /**
  71          * returns the text's size in bytes.
  72          *
  73          * @return int The size
  74          */
  75         public function getSize() {
  76                 $text = $this->getNativeData();
  77                 return strlen( $text );
  78         }
  79
  80         /**
  81          * Returns true if this content is not a redirect, and $wgArticleCountMethod
  82          * is "any".
  83          *
  84          * @param bool $hasLinks if it is known whether this content contains links,
  85          * provide this information here, to avoid redundant parsing to find out.
  86          *
  87          * @return bool True if the content is countable
  88          */
  89         public function isCountable( $hasLinks = null ) {
  90                 global $wgArticleCountMethod;
  91
  92                 if ( $this->isRedirect() ) {
  93                         return false;
  94                 }
  95
  96                 if ( $wgArticleCountMethod === 'any' ) {
  97                         return true;
  98                 }
  99
 100                 return false;
 101         }
 102
 103         /**
 104          * Returns the text represented by this Content object, as a string.
 105          *
 106          * @return string: the raw text
 107          */
 108         public function getNativeData() {
 109                 $text = $this->mText;
 110                 return $text;
 111         }
 112
 113         /**
 114          * Returns the text represented by this Content object, as a string.
 115          *
 116          * @return string: the raw text
 117          */
 118         public function getTextForSearchIndex() {
 119                 return $this->getNativeData();
 120         }
 121
 122         /**
 123          * Returns attempts to convert this content object to wikitext,
 124          * and then returns the text string. The conversion may be lossy.
 125          *
 126          * @note: this allows any text-based content to be transcluded as if it was wikitext.
 127          *
 128          * @return string|false: the raw text, or null if the conversion failed
 129          */
 130         public function getWikitextForTransclusion() {
 131                 $wikitext = $this->convert( CONTENT_MODEL_WIKITEXT, 'lossy' );
 132
 133                 if ( $wikitext ) {
 134                         return $wikitext->getNativeData();
 135                 } else {
 136                         return false;
 137                 }
 138         }
 139
 140         /**
 141          * Returns a Content object with pre-save transformations applied.
 142          * This implementation just trims trailing whitespace.
 143          *
 144          * @param $title Title
 145          * @param $user User
 146          * @param $popts ParserOptions
 147          * @return Content
 148          */
 149         public function preSaveTransform( Title $title, User $user, ParserOptions $popts ) {
 150                 $text = $this->getNativeData();
 151                 $pst = rtrim( $text );
 152
 153                 return ( $text === $pst ) ? $this : new WikitextContent( $pst );
 154         }
 155
 156         /**
 157          * Diff this content object with another content object.
 158          *
 159          * @since 1.21diff
 160          *
 161          * @param $that Content: The other content object to compare this content
 162          * object to.
 163          * @param $lang Language: The language object to use for text segmentation.
 164          *    If not given, $wgContentLang is used.
 165          *
 166          * @return DiffResult: A diff representing the changes that would have to be
 167          *    made to this content object to make it equal to $that.
 168          */
 169         public function diff( Content $that, Language $lang = null ) {
 170                 global $wgContLang;
 171
 172                 $this->checkModelID( $that->getModel() );
 173
 174                 // @todo could implement this in DifferenceEngine and just delegate here?
 175
 176                 if ( !$lang ) {
 177                         $lang = $wgContLang;
 178                 }
 179
 180                 $otext = $this->getNativeData();
 181                 $ntext = $this->getNativeData();
 182
 183                 # Note: Use native PHP diff, external engines don't give us abstract output
 184                 $ota = explode( "\n", $lang->segmentForDiff( $otext ) );
 185                 $nta = explode( "\n", $lang->segmentForDiff( $ntext ) );
 186
 187                 $diff = new Diff( $ota, $nta );
 188                 return $diff;
 189         }
 190
 191         /**
 192          * Fills the provided ParserOutput object with the HTML returned by getHtml().
 193          *
 194          * Content models in $wgTextModelsToParse will be parsed as wikitext to process links,
 195          * magic words, etc.
 196          *
 197          * Subclasses may override this to provide custom rendering.
 198          *
 199          * @param $title Title Context title for parsing
 200          * @param int|null $revId Revision ID (for {{REVISIONID}})
 201          * @param $options ParserOptions|null Parser options
 202          * @param bool $generateHtml Whether or not to generate HTML
 203          * @param $output ParserOutput The output object to fill (reference).
 204          */
 205         protected function fillParserOutput( Title $title, $revId,
 206                 ParserOptions $options, $generateHtml, ParserOutput &$output
 207         ) {
 208                 global $wgParser, $wgTextModelsToParse;
 209
 210                 if ( in_array( $this->getModel(), $wgTextModelsToParse ) ) {
 211                         // parse just to get links etc into the database, HTML is replaced below.
 212                         $output = $wgParser->parse( $this->getNativeData(), $title, $options, true, true, $revId );
 213                 }
 214
 215                 if ( $generateHtml ) {
 216                         $html = $this->getHtml();
 217                 } else {
 218                         $html = '';
 219                 }
 220
 221                 $output->setText( $html );
 222         }
 223
 224         /**
 225          * Generates an HTML version of the content, for display. Used by
 226          * getParserOutput() to construct a ParserOutput object.
 227          *
 228          * This default implementation runs the text returned by $this->getNativeData()
 229          * through htmlspecialchars and tried to convert line breaks and indentation to HTML..
 230          *
 231          * @return string An HTML representation of the content
 232          */
 233         public static function convertWhiteSpaceToHTML( $msg ) {
 234                 $msg = htmlspecialchars( $msg );
 235                 $msg = preg_replace( '/^ /m', '&#160;', $msg );
 236                 $msg = preg_replace( '/ $/m', '&#160;', $msg );
 237                 $msg = preg_replace( '/  /', '&#160; ', $msg );
 238                 $msg = str_replace( "\n", '<br />', $msg );
 239                 return $msg;
 240         }
 241
 242         /**
 243          * @see Content::convert()
 244          *
 245          * This implementation provides lossless conversion between content models based
 246          * on TextContent.
 247          *
 248          * @param string  $toModel the desired content model, use the CONTENT_MODEL_XXX flags.
 249          * @param string  $lossy flag, set to "lossy" to allow lossy conversion. If lossy conversion is
 250          * not allowed, full round-trip conversion is expected to work without losing information.
 251          *
 252          * @return Content|bool A content object with the content model $toModel, or false if
 253          * that conversion is not supported.
 254          */
 255         public function convert( $toModel, $lossy = '' ) {
 256                 $converted = parent::convert( $toModel, $lossy );
 257
 258                 if ( $converted !== false ) {
 259                         return $converted;
 260                 }
 261
 262                 $toHandler = ContentHandler::getForModelID( $toModel );
 263
 264                 if ( $toHandler instanceof TextContentHandler ) {
 265                         //NOTE: ignore content serialization format - it's just text anyway.
 266                         $text = $this->getNativeData();
 267                         $converted = $toHandler->unserializeContent( $text );
 268                 }
 269
 270                 return $converted;
 271         }
 272 }