xref: /dokuwiki/feed.php (revision d88b5f049b763613b6c2953b8f21105639a53afc)
1<?php
2/**
3 * XML feed export
4 *
5 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
6 * @author     Andreas Gohr <andi@splitbrain.org>
7 */
8
9  if(!defined('DOKU_INC')) define('DOKU_INC',realpath(dirname(__FILE__)).'/');
10  require_once(DOKU_INC.'inc/init.php');
11  require_once(DOKU_INC.'inc/common.php');
12  require_once(DOKU_INC.'inc/events.php');
13  require_once(DOKU_INC.'inc/parserutils.php');
14  require_once(DOKU_INC.'inc/feedcreator.class.php');
15  require_once(DOKU_INC.'inc/auth.php');
16  require_once(DOKU_INC.'inc/pageutils.php');
17
18  //close session
19  session_write_close();
20
21
22  $num   = $_REQUEST['num'];
23  $type  = $_REQUEST['type'];
24  $mode  = $_REQUEST['mode'];
25  $minor = $_REQUEST['minor'];
26  $ns    = $_REQUEST['ns'];
27  $ltype = $_REQUEST['linkto'];
28
29  if($type == '')
30    $type = $conf['rss_type'];
31
32  switch ($type){
33    case 'rss':
34       $type = 'RSS0.91';
35       $mime = 'text/xml';
36       break;
37    case 'rss2':
38       $type = 'RSS2.0';
39       $mime = 'text/xml';
40       break;
41    case 'atom':
42       $type = 'ATOM0.3';
43       $mime = 'application/xml';
44       break;
45    case 'atom1':
46       $type = 'ATOM1.0';
47       $mime = 'application/atom+xml';
48       break;
49    default:
50       $type = 'RSS1.0';
51       $mime = 'application/xml';
52  }
53
54  // the feed is dynamic - we need a cache for each combo
55  // (but most people just use the default feed so it's still effective)
56  $cache = getCacheName($num.$type.$mode.$ns.$ltype.$_SERVER['REMOTE_USER'],'.feed');
57
58  // check cacheage and deliver if nothing has changed since last
59  // time or the update interval has not passed, also handles conditional requests
60  header('Cache-Control: must-revalidate, post-check=0, pre-check=0');
61  header('Pragma: public');
62  header('Content-Type: application/xml; charset=utf-8');
63  $cmod = @filemtime($cache); // 0 if not exists
64  if($cmod && (($cmod+$conf['rss_update']>time()) || ($cmod>@filemtime($conf['changelog'])))){
65    http_conditionalRequest($cmod);
66    print io_readFile($cache);
67    exit;
68  } else {
69    http_conditionalRequest(time());
70  }
71
72  // create new feed
73  $rss = new DokuWikiFeedCreator();
74  $rss->title = $conf['title'].(($ns) ? ' '.$ns : '');
75  $rss->link  = DOKU_URL;
76  $rss->syndicationURL = DOKU_URL.'feed.php';
77  $rss->cssStyleSheet  = DOKU_URL.'lib/styles/feed.css';
78
79  $image = new FeedImage();
80  $image->title = $conf['title'];
81  $image->url = DOKU_URL."lib/images/favicon.ico";
82  $image->link = DOKU_URL;
83  $rss->image = $image;
84
85  if($mode == 'list'){
86    rssListNamespace($rss,$ns);
87  }else{
88    rssRecentChanges($rss,$num,$ltype,$ns,$minor);
89  }
90
91  $feed = $rss->createFeed($type,'utf-8');
92
93  // save cachefile
94  io_saveFile($cache,$feed);
95
96  // finally deliver
97  print $feed;
98
99// ---------------------------------------------------------------- //
100
101/**
102 * Add recent changed to a feed object
103 *
104 * @author Andreas Gohr <andi@splitbrain.org>
105 */
106function rssRecentChanges(&$rss,$num,$ltype,$ns,$minor){
107  global $conf;
108  global $auth;
109
110  if(!$num) $num = $conf['recent'];
111  $guardmail = ($conf['mailguard'] != '' && $conf['mailguard'] != 'none');
112
113
114  $flags = RECENTS_SKIP_DELETED;
115  if(!$minor) $flags += RECENTS_SKIP_MINORS;
116
117  $recents = getRecents(0,$num,$ns,$flags);
118
119  //this can take some time if a lot of recaching has to be done
120  @set_time_limit(90); // set max execution time
121
122  foreach($recents as $recent){
123
124    $item = new FeedItem();
125    $item->title = $recent['id'];
126    $xhtml = p_wiki_xhtml($recent['id'],'',false);
127
128    if($conf['useheading']) {
129        $matches = array();
130        if(preg_match('|<h([1-9])>(.*?)</h\1>|', $xhtml, $matches))
131            $item->title = trim($matches[2]);
132    }
133    if(!empty($recent['sum'])){
134      $item->title .= ' - '.strip_tags($recent['sum']);
135    }
136
137    $desc = cleanDesc($xhtml);
138
139    if(empty($ltype))
140      $ltype = $conf['rss_linkto'];
141
142    switch ($ltype){
143      case 'page':
144        $item->link = wl($recent['id'],'rev='.$recent['date'],true);
145        break;
146      case 'rev':
147        $item->link = wl($recent['id'],'do=revisions&rev='.$recent['date'],true);
148        break;
149      case 'current':
150        $item->link = wl($recent['id'], '', true);
151        break;
152      case 'diff':
153      default:
154        $item->link = wl($recent['id'],'rev='.$recent['date'].'&do=diff'.$recent['date'],true);
155    }
156
157    $item->description = $desc;
158    $item->date        = date('r',$recent['date']);
159    $cat = getNS($recent['id']);
160    if($cat) $item->category = $cat;
161
162    $user = null;
163    $user = @$recent['user']; // the @ spares time repeating lookup
164    $item->author = '';
165
166    if($user){
167      $userInfo = $auth->getUserData($user);
168      $item->author = $userInfo['name'];
169      if($guardmail) {
170        //cannot obfuscate because some RSS readers may check validity
171        $item->authorEmail = $user.'@'.$recent['ip'];
172      }else{
173        $item->authorEmail = $userInfo['mail'];
174      }
175    }else{
176      $item->authorEmail = 'anonymous@'.$recent['ip'];
177    }
178    $rss->addItem($item);
179  }
180}
181
182/**
183 * Add all pages of a namespace to a feedobject
184 *
185 * @author Andreas Gohr <andi@splitbrain.org>
186 */
187function rssListNamespace(&$rss,$ns){
188  require_once(DOKU_INC.'inc/search.php');
189  global $conf;
190
191  $ns=':'.cleanID($ns);
192  $ns=str_replace(':','/',$ns);
193
194  $data = array();
195  sort($data);
196  search($data,$conf['datadir'],'search_list','',$ns);
197  foreach($data as $row){
198    $item = new FeedItem();
199
200    $id    = $row['id'];
201    $date  = filemtime(wikiFN($id));
202    $xhtml = p_wiki_xhtml($id,'',false);
203    $desc  = cleanDesc($xhtml);
204    $item->title       = $id;
205
206    if($conf['useheading']) {
207        $matches = array();
208        if(preg_match('|<h([1-9])>(.*?)</h\1>|', $xhtml, $matches))
209            $item->title = trim($matches[2]);
210    }
211
212    $item->link        = wl($id,'rev='.$date,true);
213    $item->description = $desc;
214    $item->date        = date('r',$date);
215    $rss->addItem($item);
216  }
217}
218
219/**
220 * Clean description for feed inclusion
221 *
222 * Removes HTML tags and line breaks and trims the text to
223 * 250 chars
224 *
225 * @author Andreas Gohr <andi@splitbrain.org>
226 */
227function cleanDesc($desc){
228  //start description at text of first paragraph
229  $matches = array();
230  if(preg_match('/<p>|<p\s.*?>/', $desc, $matches, PREG_OFFSET_CAPTURE))
231      $desc = substr($desc, $matches[0][1]);
232
233  //remove TOC
234  $desc = preg_replace('!<div class="toc">.*?(</div>\n</div>)!s','',$desc);
235  $desc = strip_tags($desc);
236  $desc = preg_replace('/[\n\r\t]/',' ',$desc);
237  $desc = preg_replace('/  /',' ',$desc);
238  $desc = utf8_substr($desc,0,250);
239  $desc = $desc.'...';
240  return $desc;
241}
242
243?>
244