Added Google sitemap support #371

This patch addes the automatic creation of Google sitemaps. The map
is created in the DokuWiki root dir and named sitemap.xml.gz if gzip
compression is available - if not the gz extion is skipped.
How often the map is recreated is defined through the $conf['sitemap']
option. It accepts a day value.

darcs-hash:20051126234709-7ad00-6ff4b0e79670cdfa39e615ec9dc40146ffcc9dd4.gz

1 files changed, 117 insertions, 51 deletions

diff --git a/lib/exe/indexer.php b/lib/exe/indexer.php
index 80dbb77f5..2eea32f02 100644
--- a/lib/exe/indexer.php
+++ b/lib/exe/indexer.php
@@ -5,77 +5,143 @@
  * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
  * @author     Andreas Gohr <andi@splitbrain.org>
  */
-
-/**
- * Just send a 1x1 pixel blank gif to the browser and exit
-  
- */
-function sendGIF(){
-    $img = base64_decode('R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAEALAAAAAABAAEAAAIBTAA7');
-    header('Content-Type: image/gif');
-    header('Content-Length: '.strlen($img));
-    header('Connection: Close');
-    print $img;
-    // Browser should drop connection after this
-    // Thinks it's got the whole image
-}
-
-// Make sure image is sent to the browser immediately
-ob_implicit_flush(TRUE);
+if(!defined('DOKU_INC')) define('DOKU_INC',realpath(dirname(__FILE__).'/../../').'/');
+require_once(DOKU_INC.'inc/init.php');
+require_once(DOKU_INC.'inc/auth.php');
+session_write_close();  //close session
 
 // keep running after browser closes connection
 @ignore_user_abort(true);
 
+// send gif
 sendGIF();
 
-// Switch off implicit flush again - we don't want to send any more output
-ob_implicit_flush(FALSE);
-
 // Catch any possible output (e.g. errors)
 // - probably not needed but better safe...
 ob_start();
 
-// Called to exit - we don't want any output going anywhere
-function indexer_stop() {
-    ob_end_clean();
-    exit();
-}
-
 // Now start work
-if(!defined('DOKU_INC')) define('DOKU_INC',realpath(dirname(__FILE__).'/../../').'/');
-require_once(DOKU_INC.'inc/init.php');
-//close session
-session_write_close();
+require_once(DOKU_INC.'inc/utf8.php');
+require_once(DOKU_INC.'inc/auth.php');
 
-require_once(DOKU_INC.'inc/indexer.php');
+// run one of the jobs
+runIndexer() or runSitemapper();
 
-$ID = cleanID($_REQUEST['id']);
-if(!$ID) indexer_stop();
+ob_end_clean();
+exit;
 
-// check if indexing needed
-$last = @filemtime(metaFN($ID,'.indexed'));
-if($last > @filemtime(wikiFN($ID))) indexer_stop();
+// --------------------------------------------------------------------
 
-// try to aquire a lock
-$lock = $conf['lockdir'].'/_indexer.lock';
-while(!@mkdir($lock,0777)){
-    if(time()-@filemtime($lock) > 60*5){
-        // looks like a stale lock - remove it
-        @rmdir($lock);
-    }else{
-        indexer_stop();
+/**
+ * Runs the indexer for the current page
+ *
+ * @author Andreas Gohr <andi@splitbrain.org>
+ */
+function runIndexer(){
+    global $conf;
+
+    $ID = cleanID($_REQUEST['id']);
+    if(!$ID) return false;
+
+    // check if indexing needed
+    $last = @filemtime(metaFN($ID,'.indexed'));
+    if($last > @filemtime(wikiFN($ID))) return false;
+
+    // try to aquire a lock
+    $lock = $conf['lockdir'].'/_indexer.lock';
+    while(!@mkdir($lock,0777)){
+        if(time()-@filemtime($lock) > 60*5){
+            // looks like a stale lock - remove it
+            @rmdir($lock);
+        }else{
+            return false;
+        }
     }
+
+    require_once(DOKU_INC.'inc/indexer.php');
+
+    // do the work
+    idx_addPage($ID);
+
+    // we're finished - save and free lock
+    io_saveFile(metaFN($ID,'.indexed'),' ');
+    @rmdir($lock);
+    return true;
 }
 
-// do the work
-idx_addPage($ID);
+/**
+ * Builds a Google Sitemap of all public pages known to the indexer
+ *
+ * The map is placed in the root directory named sitemap.xml.gz - This
+ * file needs to be writable!
+ *
+ * @author Andreas Gohr
+ * @link   https://www.google.com/webmasters/sitemaps/docs/en/about.html
+ */
+function runSitemapper(){
+    global $conf;
+    if(!$conf['sitemap']) return false;
+    if(!defined('NL')) define('NL',"\n");
 
-// we're finished
-io_saveFile(metaFN($ID,'.indexed'),' ');
-@rmdir($lock);
-indexer_stop();
+    if($conf['usegzip']){
+        $sitemap = DOKU_INC.'sitemap.xml.gz';
+    }else{
+        $sitemap = DOKU_INC.'sitemap.xml';
+    }
 
-//Setup VIM: ex: et ts=4 enc=utf-8 :
 
+    if(!is_writable($sitemap)) return false;
+    if(@filesize($sitemap) && 
+       @filemtime($sitemap) > (time()-($conf['sitemap']*60*60*24))){
+       return false;
+    }
+
+    ob_start();
+    $pages = file($conf['cachedir'].'/page.idx');
+
+    print '<?xml version="1.0" encoding="UTF-8"?>'.NL;
+    print '<urlset xmlns="http://www.google.com/schemas/sitemap/0.84">'.NL;
+    foreach($pages as $id){
+        $id = trim($id);
+        $file = wikiFN($id);
+
+        //skip hidden, non existing and restricted files
+        if(isHiddenPage($id)) return false;
+        $date = @filemtime($file);
+        if(!$date) continue;
+        if(auth_aclcheck($id,'','') < AUTH_READ) continue;
+
+        print '  <url>'.NL;
+        print '    <loc>'.wl($id,'',true).'</loc>'.NL;
+        print '    <lastmod>'.date('Y-m-d\TH:i:s',$date).'</lastmod>'.NL;
+        print '  </url>'.NL;
+    }
+    print '</urlset>'.NL;
+
+    $data = ob_get_contents();
+    ob_end_clean();
+
+    io_saveFile($sitemap,$data);
+    return true;
+}
+
+/**
+ * Just send a 1x1 pixel blank gif to the browser
+ * 
+ * @author Andreas Gohr <andi@splitbrain.org>
+ * @author Harry Fuecks <fuecks@gmail.com>
+ */
+function sendGIF(){
+    $img = base64_decode('R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAEALAAAAAABAAEAAAIBTAA7');
+    header('Content-Type: image/gif');
+    header('Content-Length: '.strlen($img));
+    header('Connection: Close');
+    print $img;
+    flush();
+    // Browser should drop connection after this
+    // Thinks it's got the whole image
+}
+
+//Setup VIM: ex: et ts=4 enc=utf-8 :
 // No trailing PHP closing tag - no output please!
 // See Note at http://www.php.net/manual/en/language.basic-syntax.instruction-separation.php