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          * Constructor
  58          *
  59          * @param string|Title $title Item's title
  60          * @param string $description
  61          * @param string $url URL uniquely designating the item.
  62          * @param string $date Item's date
  63          * @param string $author Author's user name
  64          * @param string $comments
  65          */
  66         function __construct( $title, $description, $url, $date = '', $author = '', $comments = '' ) {
  67                 $this->title = $title;
  68                 $this->description = $description;
  69                 $this->url = $url;
  70                 $this->uniqueId = $url;
  71                 $this->date = $date;
  72                 $this->author = $author;
  73                 $this->comments = $comments;
  74         }
  75
  76         /**
  77          * Encode $string so that it can be safely embedded in a XML document
  78          *
  79          * @param string $string String to encode
  80          * @return string
  81          */
  82         public function xmlEncode( $string ) {
  83                 $string = str_replace( "\r\n", "\n", $string );
  84                 $string = preg_replace( '/[\x00-\x08\x0b\x0c\x0e-\x1f]/', '', $string );
  85                 return htmlspecialchars( $string );
  86         }
  87
  88         /**
  89          * Get the unique id of this item
  90          *
  91          * @return string
  92          */
  93         public function getUniqueId() {
  94                 if ( $this->uniqueId ) {
  95                         return $this->xmlEncode( wfExpandUrl( $this->uniqueId, PROTO_CURRENT ) );
  96                 }
  97         }
  98
  99         /**
 100          * Set the unique id of an item
 101          *
 102          * @param string $uniqueId Unique id for the item
 103          * @param bool $rssIsPermalink Set to true if the guid (unique id) is a permalink (RSS feeds only)
 104          */
 105         public function setUniqueId( $uniqueId, $rssIsPermalink = false ) {
 106                 $this->uniqueId = $uniqueId;
 107                 $this->rssIsPermalink = $rssIsPermalink;
 108         }
 109
 110         /**
 111          * Get the title of this item; already xml-encoded
 112          *
 113          * @return string
 114          */
 115         public function getTitle() {
 116                 return $this->xmlEncode( $this->title );
 117         }
 118
 119         /**
 120          * Get the URL of this item; already xml-encoded
 121          *
 122          * @return string
 123          */
 124         public function getUrl() {
 125                 return $this->xmlEncode( $this->url );
 126         }
 127
 128         /**
 129          * Get the description of this item; already xml-encoded
 130          *
 131          * @return string
 132          */
 133         public function getDescription() {
 134                 return $this->xmlEncode( $this->description );
 135         }
 136
 137         /**
 138          * Get the language of this item
 139          *
 140          * @return string
 141          */
 142         public function getLanguage() {
 143                 global $wgLanguageCode;
 144                 return wfBCP47( $wgLanguageCode );
 145         }
 146
 147         /**
 148          * Get the date of this item
 149          *
 150          * @return string
 151          */
 152         public function getDate() {
 153                 return $this->date;
 154         }
 155
 156         /**
 157          * Get the author of this item; already xml-encoded
 158          *
 159          * @return string
 160          */
 161         public function getAuthor() {
 162                 return $this->xmlEncode( $this->author );
 163         }
 164
 165         /**
 166          * Get the comment of this item; already xml-encoded
 167          *
 168          * @return string
 169          */
 170         public function getComments() {
 171                 return $this->xmlEncode( $this->comments );
 172         }
 173
 174         /**
 175          * Quickie hack... strip out wikilinks to more legible form from the comment.
 176          *
 177          * @param string $text Wikitext
 178          * @return string
 179          */
 180         public static function stripComment( $text ) {
 181                 return preg_replace( '/\[\[([^]]*\|)?([^]]+)\]\]/', '\2', $text );
 182         }
 183         /**#@-*/
 184 }
 185
 186 /**
 187  * Class to support the outputting of syndication feeds in Atom and RSS format.
 188  *
 189  * @ingroup Feed
 190  */
 191 abstract class ChannelFeed extends FeedItem {
 192         /**
 193          * Generate Header of the feed
 194          * @par Example:
 195          * @code
 196          * print "<feed>";
 197          * @endcode
 198          */
 199         abstract public function outHeader();
 200
 201         /**
 202          * Generate an item
 203          * @par Example:
 204          * @code
 205          * print "<item>...</item>";
 206          * @endcode
 207          * @param FeedItem $item
 208          */
 209         abstract public function outItem( $item );
 210
 211         /**
 212          * Generate Footer of the feed
 213          * @par Example:
 214          * @code
 215          * print "</feed>";
 216          * @endcode
 217          */
 218         abstract public function outFooter();
 219
 220         /**
 221          * Setup and send HTTP headers. Don't send any content;
 222          * content might end up being cached and re-sent with
 223          * these same headers later.
 224          *
 225          * This should be called from the outHeader() method,
 226          * but can also be called separately.
 227          */
 228         public function httpHeaders() {
 229                 global $wgOut, $wgVaryOnXFP;
 230
 231                 # We take over from $wgOut, excepting its cache header info
 232                 $wgOut->disable();
 233                 $mimetype = $this->contentType();
 234                 header( "Content-type: $mimetype; charset=UTF-8" );
 235                 if ( $wgVaryOnXFP ) {
 236                         $wgOut->addVaryHeader( 'X-Forwarded-Proto' );
 237                 }
 238                 $wgOut->sendCacheControl();
 239         }
 240
 241         /**
 242          * Return an internet media type to be sent in the headers.
 243          *
 244          * @return string
 245          */
 246         private function contentType() {
 247                 global $wgRequest;
 248
 249                 $ctype = $wgRequest->getVal( 'ctype', 'application/xml' );
 250                 $allowedctypes = [
 251                         'application/xml',
 252                         'text/xml',
 253                         'application/rss+xml',
 254                         'application/atom+xml'
 255                 ];
 256
 257                 return ( in_array( $ctype, $allowedctypes ) ? $ctype : 'application/xml' );
 258         }
 259
 260         /**
 261          * Output the initial XML headers.
 262          */
 263         protected function outXmlHeader() {
 264                 $this->httpHeaders();
 265                 echo '<?xml version="1.0"?>' . "\n";
 266         }
 267 }
 268
 269 /**
 270  * Generate a RSS feed
 271  *
 272  * @ingroup Feed
 273  */
 274 class RSSFeed extends ChannelFeed {
 275
 276         /**
 277          * Format a date given a timestamp
 278          *
 279          * @param int $ts Timestamp
 280          * @return string Date string
 281          */
 282         function formatTime( $ts ) {
 283                 return gmdate( 'D, d M Y H:i:s \G\M\T', wfTimestamp( TS_UNIX, $ts ) );
 284         }
 285
 286         /**
 287          * Output an RSS 2.0 header
 288          */
 289         function outHeader() {
 290                 global $wgVersion;
 291
 292                 $this->outXmlHeader();
 293                 ?><rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/">
 294         <channel>
 295                 <title><?php print $this->getTitle() ?></title>
 296                 <link><?php print wfExpandUrl( $this->getUrl(), PROTO_CURRENT ) ?></link>
 297                 <description><?php print $this->getDescription() ?></description>
 298                 <language><?php print $this->getLanguage() ?></language>
 299                 <generator>MediaWiki <?php print $wgVersion ?></generator>
 300                 <lastBuildDate><?php print $this->formatTime( wfTimestampNow() ) ?></lastBuildDate>
 301 <?php
 302         }
 303
 304         /**
 305          * Output an RSS 2.0 item
 306          * @param FeedItem $item Item to be output
 307          */
 308         function outItem( $item ) {
 309                 // @codingStandardsIgnoreStart Ignore long lines and formatting issues.
 310         ?>
 311                 <item>
 312                         <title><?php print $item->getTitle(); ?></title>
 313                         <link><?php print wfExpandUrl( $item->getUrl(), PROTO_CURRENT ); ?></link>
 314                         <guid<?php if ( !$item->rssIsPermalink ) { print ' isPermaLink="false"'; } ?>><?php print $item->getUniqueId(); ?></guid>
 315                         <description><?php print $item->getDescription() ?></description>
 316                         <?php if ( $item->getDate() ) { ?><pubDate><?php print $this->formatTime( $item->getDate() ); ?></pubDate><?php } ?>
 317                         <?php if ( $item->getAuthor() ) { ?><dc:creator><?php print $item->getAuthor(); ?></dc:creator><?php }?>
 318                         <?php if ( $item->getComments() ) { ?><comments><?php print wfExpandUrl( $item->getComments(), PROTO_CURRENT ); ?></comments><?php }?>
 319                 </item>
 320 <?php
 321                 // @codingStandardsIgnoreEnd
 322         }
 323
 324         /**
 325          * Output an RSS 2.0 footer
 326          */
 327         function outFooter() {
 328         ?>
 329         </channel>
 330 </rss><?php
 331         }
 332 }
 333
 334 /**
 335  * Generate an Atom feed
 336  *
 337  * @ingroup Feed
 338  */
 339 class AtomFeed extends ChannelFeed {
 340         /**
 341          * Format a date given timestamp.
 342          *
 343          * @param string|int $timestamp
 344          * @return string
 345          */
 346         function formatTime( $timestamp ) {
 347                 // need to use RFC 822 time format at least for rss2.0
 348                 return gmdate( 'Y-m-d\TH:i:s', wfTimestamp( TS_UNIX, $timestamp ) );
 349         }
 350
 351         /**
 352          * Outputs a basic header for Atom 1.0 feeds.
 353          */
 354         function outHeader() {
 355                 global $wgVersion;
 356
 357                 $this->outXmlHeader();
 358                 // @codingStandardsIgnoreStart Ignore long lines and formatting issues.
 359                 ?><feed xmlns="http://www.w3.org/2005/Atom" xml:lang="<?php print $this->getLanguage() ?>">
 360                 <id><?php print $this->getFeedId() ?></id>
 361                 <title><?php print $this->getTitle() ?></title>
 362                 <link rel="self" type="application/atom+xml" href="<?php print wfExpandUrl( $this->getSelfUrl(), PROTO_CURRENT ) ?>"/>
 363                 <link rel="alternate" type="text/html" href="<?php print wfExpandUrl( $this->getUrl(), PROTO_CURRENT ) ?>"/>
 364                 <updated><?php print $this->formatTime( wfTimestampNow() ) ?>Z</updated>
 365                 <subtitle><?php print $this->getDescription() ?></subtitle>
 366                 <generator>MediaWiki <?php print $wgVersion ?></generator>
 367
 368 <?php
 369                 // @codingStandardsIgnoreEnd
 370         }
 371
 372         /**
 373          * Atom 1.0 requires a unique, opaque IRI as a unique identifier
 374          * for every feed we create. For now just use the URL, but who
 375          * can tell if that's right? If we put options on the feed, do we
 376          * have to change the id? Maybe? Maybe not.
 377          *
 378          * @return string
 379          */
 380         private function getFeedId() {
 381                 return $this->getSelfUrl();
 382         }
 383
 384         /**
 385          * Atom 1.0 requests a self-reference to the feed.
 386          * @return string
 387          */
 388         private function getSelfUrl() {
 389                 global $wgRequest;
 390                 return htmlspecialchars( $wgRequest->getFullRequestURL() );
 391         }
 392
 393         /**
 394          * Output a given item.
 395          * @param FeedItem $item
 396          */
 397         function outItem( $item ) {
 398                 global $wgMimeType;
 399                 // @codingStandardsIgnoreStart Ignore long lines and formatting issues.
 400         ?>
 401         <entry>
 402                 <id><?php print $item->getUniqueId(); ?></id>
 403                 <title><?php print $item->getTitle(); ?></title>
 404                 <link rel="alternate" type="<?php print $wgMimeType ?>" href="<?php print wfExpandUrl( $item->getUrl(), PROTO_CURRENT ); ?>"/>
 405                 <?php if ( $item->getDate() ) { ?>
 406                 <updated><?php print $this->formatTime( $item->getDate() ); ?>Z</updated>
 407                 <?php } ?>
 408
 409                 <summary type="html"><?php print $item->getDescription() ?></summary>
 410                 <?php if ( $item->getAuthor() ) { ?><author><name><?php print $item->getAuthor(); ?></name></author><?php }?>
 411         </entry>
 412
 413 <?php /* @todo FIXME: Need to add comments
 414         <?php if( $item->getComments() ) { ?><dc:comment><?php print $item->getComments() ?></dc:comment><?php }?>
 415           */
 416         }
 417
 418         /**
 419          * Outputs the footer for Atom 1.0 feed (basically '\</feed\>').
 420          */
 421         function outFooter() {?>
 422         </feed><?php
 423                 // @codingStandardsIgnoreEnd
 424         }
 425 }