includes/Feed.php

   1 <?php
   2 /**
   3  * Basic support for outputting syndication feeds in RSS, other formats.
   4  *
   5  * Contain a feed class as well as classes to build rss / atom ... feeds
   6  * Available feeds are defined in Defines.php
   7  *
   8  * Copyright © 2004 Brion Vibber <brion@pobox.com>
   9  * https://www.mediawiki.org/
  10  *
  11  * This program is free software; you can redistribute it and/or modify
  12  * it under the terms of the GNU General Public License as published by
  13  * the Free Software Foundation; either version 2 of the License, or
  14  * (at your option) any later version.
  15  *
  16  * This program is distributed in the hope that it will be useful,
  17  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  18  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  19  * GNU General Public License for more details.
  20  *
  21  * You should have received a copy of the GNU General Public License along
  22  * with this program; if not, write to the Free Software Foundation, Inc.,
  23  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  24  * http://www.gnu.org/copyleft/gpl.html
  25  *
  26  * @file
  27  */
  28
  29 /**
  30  * @defgroup Feed Feed
  31  */
  32
  33 /**
  34  * A base class for basic support for outputting syndication feeds in RSS and other formats.
  35  *
  36  * @ingroup Feed
  37  */
  38 class FeedItem {
  39         /** @var Title */
  40         public $title;
  41
  42         public $description;
  43
  44         public $url;
  45
  46         public $date;
  47
  48         public $author;
  49
  50         public $uniqueId;
  51
  52         public $comments;
  53
  54         public $rssIsPermalink = false;
  55
  56         /**
  57          * @param string|Title $title Item's title
  58          * @param string $description
  59          * @param string $url URL uniquely designating the item.
  60          * @param string $date Item's date
  61          * @param string $author Author's user name
  62          * @param string $comments
  63          */
  64         function __construct( $title, $description, $url, $date = '', $author = '', $comments = '' ) {
  65                 $this->title = $title;
  66                 $this->description = $description;
  67                 $this->url = $url;
  68                 $this->uniqueId = $url;
  69                 $this->date = $date;
  70                 $this->author = $author;
  71                 $this->comments = $comments;
  72         }
  73
  74         /**
  75          * Encode $string so that it can be safely embedded in a XML document
  76          *
  77          * @param string $string String to encode
  78          * @return string
  79          */
  80         public function xmlEncode( $string ) {
  81                 $string = str_replace( "\r\n", "\n", $string );
  82                 $string = preg_replace( '/[\x00-\x08\x0b\x0c\x0e-\x1f]/', '', $string );
  83                 return htmlspecialchars( $string );
  84         }
  85
  86         /**
  87          * Get the unique id of this item
  88          *
  89          * @return string
  90          */
  91         public function getUniqueId() {
  92                 if ( $this->uniqueId ) {
  93                         return $this->xmlEncode( wfExpandUrl( $this->uniqueId, PROTO_CURRENT ) );
  94                 }
  95         }
  96
  97         /**
  98          * Set the unique id of an item
  99          *
 100          * @param string $uniqueId Unique id for the item
 101          * @param bool $rssIsPermalink Set to true if the guid (unique id) is a permalink (RSS feeds only)
 102          */
 103         public function setUniqueId( $uniqueId, $rssIsPermalink = false ) {
 104                 $this->uniqueId = $uniqueId;
 105                 $this->rssIsPermalink = $rssIsPermalink;
 106         }
 107
 108         /**
 109          * Get the title of this item; already xml-encoded
 110          *
 111          * @return string
 112          */
 113         public function getTitle() {
 114                 return $this->xmlEncode( $this->title );
 115         }
 116
 117         /**
 118          * Get the URL of this item; already xml-encoded
 119          *
 120          * @return string
 121          */
 122         public function getUrl() {
 123                 return $this->xmlEncode( $this->url );
 124         }
 125
 126         /**
 127          * Get the description of this item; already xml-encoded
 128          *
 129          * @return string
 130          */
 131         public function getDescription() {
 132                 return $this->xmlEncode( $this->description );
 133         }
 134
 135         /**
 136          * Get the language of this item
 137          *
 138          * @return string
 139          */
 140         public function getLanguage() {
 141                 global $wgLanguageCode;
 142                 return LanguageCode::bcp47( $wgLanguageCode );
 143         }
 144
 145         /**
 146          * Get the date of this item
 147          *
 148          * @return string
 149          */
 150         public function getDate() {
 151                 return $this->date;
 152         }
 153
 154         /**
 155          * Get the author of this item; already xml-encoded
 156          *
 157          * @return string
 158          */
 159         public function getAuthor() {
 160                 return $this->xmlEncode( $this->author );
 161         }
 162
 163         /**
 164          * Get the comment of this item; already xml-encoded
 165          *
 166          * @return string
 167          */
 168         public function getComments() {
 169                 return $this->xmlEncode( $this->comments );
 170         }
 171
 172         /**
 173          * Quickie hack... strip out wikilinks to more legible form from the comment.
 174          *
 175          * @param string $text Wikitext
 176          * @return string
 177          */
 178         public static function stripComment( $text ) {
 179                 return preg_replace( '/\[\[([^]]*\|)?([^]]+)\]\]/', '\2', $text );
 180         }
 181         /**#@-*/
 182 }
 183
 184 /**
 185  * Class to support the outputting of syndication feeds in Atom and RSS format.
 186  *
 187  * @ingroup Feed
 188  */
 189 abstract class ChannelFeed extends FeedItem {
 190         /**
 191          * Generate Header of the feed
 192          * @par Example:
 193          * @code
 194          * print "<feed>";
 195          * @endcode
 196          */
 197         abstract public function outHeader();
 198
 199         /**
 200          * Generate an item
 201          * @par Example:
 202          * @code
 203          * print "<item>...</item>";
 204          * @endcode
 205          * @param FeedItem $item
 206          */
 207         abstract public function outItem( $item );
 208
 209         /**
 210          * Generate Footer of the feed
 211          * @par Example:
 212          * @code
 213          * print "</feed>";
 214          * @endcode
 215          */
 216         abstract public function outFooter();
 217
 218         /**
 219          * Setup and send HTTP headers. Don't send any content;
 220          * content might end up being cached and re-sent with
 221          * these same headers later.
 222          *
 223          * This should be called from the outHeader() method,
 224          * but can also be called separately.
 225          */
 226         public function httpHeaders() {
 227                 global $wgOut, $wgVaryOnXFP;
 228
 229                 # We take over from $wgOut, excepting its cache header info
 230                 $wgOut->disable();
 231                 $mimetype = $this->contentType();
 232                 header( "Content-type: $mimetype; charset=UTF-8" );
 233
 234                 // Set a sane filename
 235                 $exts = MimeMagic::singleton()->getExtensionsForType( $mimetype );
 236                 $ext = $exts ? strtok( $exts, ' ' ) : 'xml';
 237                 header( "Content-Disposition: inline; filename=\"feed.{$ext}\"" );
 238
 239                 if ( $wgVaryOnXFP ) {
 240                         $wgOut->addVaryHeader( 'X-Forwarded-Proto' );
 241                 }
 242                 $wgOut->sendCacheControl();
 243         }
 244
 245         /**
 246          * Return an internet media type to be sent in the headers.
 247          *
 248          * @return string
 249          */
 250         private function contentType() {
 251                 global $wgRequest;
 252
 253                 $ctype = $wgRequest->getVal( 'ctype', 'application/xml' );
 254                 $allowedctypes = [
 255                         'application/xml',
 256                         'text/xml',
 257                         'application/rss+xml',
 258                         'application/atom+xml'
 259                 ];
 260
 261                 return ( in_array( $ctype, $allowedctypes ) ? $ctype : 'application/xml' );
 262         }
 263
 264         /**
 265          * Output the initial XML headers.
 266          */
 267         protected function outXmlHeader() {
 268                 $this->httpHeaders();
 269                 echo '<?xml version="1.0"?>' . "\n";
 270         }
 271 }
 272
 273 /**
 274  * Generate a RSS feed
 275  *
 276  * @ingroup Feed
 277  */
 278 class RSSFeed extends ChannelFeed {
 279
 280         /**
 281          * Format a date given a timestamp
 282          *
 283          * @param int $ts Timestamp
 284          * @return string Date string
 285          */
 286         function formatTime( $ts ) {
 287                 return gmdate( 'D, d M Y H:i:s \G\M\T', wfTimestamp( TS_UNIX, $ts ) );
 288         }
 289
 290         /**
 291          * Output an RSS 2.0 header
 292          */
 293         function outHeader() {
 294                 global $wgVersion;
 295
 296                 $this->outXmlHeader();
 297                 ?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">
 298         <channel>
 299                 <title><?php print $this->getTitle() ?></title>
 300                 <link><?php print wfExpandUrl( $this->getUrl(), PROTO_CURRENT ) ?></link>
 301                 <description><?php print $this->getDescription() ?></description>
 302                 <language><?php print $this->getLanguage() ?></language>
 303                 <generator>MediaWiki <?php print $wgVersion ?></generator>
 304                 <lastBuildDate><?php print $this->formatTime( wfTimestampNow() ) ?></lastBuildDate>
 305 <?php
 306         }
 307
 308         /**
 309          * Output an RSS 2.0 item
 310          * @param FeedItem $item Item to be output
 311          */
 312         function outItem( $item ) {
 313                 // @codingStandardsIgnoreStart Ignore long lines and formatting issues.
 314         ?>
 315                 <item>
 316                         <title><?php print $item->getTitle(); ?></title>
 317                         <link><?php print wfExpandUrl( $item->getUrl(), PROTO_CURRENT ); ?></link>
 318                         <guid<?php if ( !$item->rssIsPermalink ) { print ' isPermaLink="false"'; } ?>><?php print $item->getUniqueId(); ?></guid>
 319                         <description><?php print $item->getDescription() ?></description>
 320                         <?php if ( $item->getDate() ) { ?><pubDate><?php print $this->formatTime( $item->getDate() ); ?></pubDate><?php } ?>
 321                         <?php if ( $item->getAuthor() ) { ?><dc:creator><?php print $item->getAuthor(); ?></dc:creator><?php }?>
 322                         <?php if ( $item->getComments() ) { ?><comments><?php print wfExpandUrl( $item->getComments(), PROTO_CURRENT ); ?></comments><?php }?>
 323                 </item>
 324 <?php
 325                 // @codingStandardsIgnoreEnd
 326         }
 327
 328         /**
 329          * Output an RSS 2.0 footer
 330          */
 331         function outFooter() {
 332         ?>
 333         </channel>
 334 </rss><?php
 335         }
 336 }
 337
 338 /**
 339  * Generate an Atom feed
 340  *
 341  * @ingroup Feed
 342  */
 343 class AtomFeed extends ChannelFeed {
 344         /**
 345          * Format a date given timestamp.
 346          *
 347          * @param string|int $timestamp
 348          * @return string
 349          */
 350         function formatTime( $timestamp ) {
 351                 // need to use RFC 822 time format at least for rss2.0
 352                 return gmdate( 'Y-m-d\TH:i:s', wfTimestamp( TS_UNIX, $timestamp ) );
 353         }
 354
 355         /**
 356          * Outputs a basic header for Atom 1.0 feeds.
 357          */
 358         function outHeader() {
 359                 global $wgVersion;
 360
 361                 $this->outXmlHeader();
 362                 // @codingStandardsIgnoreStart Ignore long lines and formatting issues.
 363                 ?><feed xmlns="http://www.w3.org/2005/Atom" xml:lang="<?php print $this->getLanguage() ?>">
 364                 <id><?php print $this->getFeedId() ?></id>
 365                 <title><?php print $this->getTitle() ?></title>
 366                 <link rel="self" type="application/atom+xml" href="<?php print wfExpandUrl( $this->getSelfUrl(), PROTO_CURRENT ) ?>"/>
 367                 <link rel="alternate" type="text/html" href="<?php print wfExpandUrl( $this->getUrl(), PROTO_CURRENT ) ?>"/>
 368                 <updated><?php print $this->formatTime( wfTimestampNow() ) ?>Z</updated>
 369                 <subtitle><?php print $this->getDescription() ?></subtitle>
 370                 <generator>MediaWiki <?php print $wgVersion ?></generator>
 371
 372 <?php
 373                 // @codingStandardsIgnoreEnd
 374         }
 375
 376         /**
 377          * Atom 1.0 requires a unique, opaque IRI as a unique identifier
 378          * for every feed we create. For now just use the URL, but who
 379          * can tell if that's right? If we put options on the feed, do we
 380          * have to change the id? Maybe? Maybe not.
 381          *
 382          * @return string
 383          */
 384         private function getFeedId() {
 385                 return $this->getSelfUrl();
 386         }
 387
 388         /**
 389          * Atom 1.0 requests a self-reference to the feed.
 390          * @return string
 391          */
 392         private function getSelfUrl() {
 393                 global $wgRequest;
 394                 return htmlspecialchars( $wgRequest->getFullRequestURL() );
 395         }
 396
 397         /**
 398          * Output a given item.
 399          * @param FeedItem $item
 400          */
 401         function outItem( $item ) {
 402                 global $wgMimeType;
 403                 // @codingStandardsIgnoreStart Ignore long lines and formatting issues.
 404         ?>
 405         <entry>
 406                 <id><?php print $item->getUniqueId(); ?></id>
 407                 <title><?php print $item->getTitle(); ?></title>
 408                 <link rel="alternate" type="<?php print $wgMimeType ?>" href="<?php print wfExpandUrl( $item->getUrl(), PROTO_CURRENT ); ?>"/>
 409                 <?php if ( $item->getDate() ) { ?>
 410                 <updated><?php print $this->formatTime( $item->getDate() ); ?>Z</updated>
 411                 <?php } ?>
 412
 413                 <summary type="html"><?php print $item->getDescription() ?></summary>
 414                 <?php if ( $item->getAuthor() ) { ?><author><name><?php print $item->getAuthor(); ?></name></author><?php }?>
 415         </entry>
 416
 417 <?php /* @todo FIXME: Need to add comments
 418         <?php if( $item->getComments() ) { ?><dc:comment><?php print $item->getComments() ?></dc:comment><?php }?>
 419           */
 420         }
 421
 422         /**
 423          * Outputs the footer for Atom 1.0 feed (basically '\</feed\>').
 424          */
 425         function outFooter() {?>
 426         </feed><?php
 427                 // @codingStandardsIgnoreEnd
 428         }
 429 }