Update to r42587 -- use insert() DB wrapper instead of constructing raw SQL with...
[lhc/web/wiklou.git] / maintenance / generateSitemap.php
index 09dc148..cc3f523 100644 (file)
@@ -2,17 +2,16 @@
 define( 'GS_MAIN', -2 );
 define( 'GS_TALK', -1 );
 /**
- * Creates a Google sitemap for the site
+ * Creates a sitemap for the site
  *
- * @package MediaWiki
- * @subpackage Maintenance
+ * @ingroup Maintenance
  *
  * @copyright Copyright © 2005, Ævar Arnfjörð Bjarmason
  * @copyright Copyright © 2005, Jens Frank <jeluf@gmx.de>
  * @copyright Copyright © 2005, Brion Vibber <brion@pobox.com>
  *
- * @link http://www.google.com/webmasters/sitemaps/docs/en/about.html
- * @link http://www.google.com/schemas/sitemap/0.84/sitemap.xsd
+ * @see http://www.sitemaps.org/
+ * @see http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd
  *
  * @license http://www.gnu.org/copyleft/gpl.html GNU General Public License 2.0 or later
  */
@@ -21,7 +20,7 @@ class GenerateSitemap {
        /**
         * The maximum amount of urls in a sitemap file
         *
-        * @link http://www.google.com/schemas/sitemap/0.84/sitemap.xsd
+        * @link http://www.sitemaps.org/schemas/sitemap/0.9/sitemap.xsd
         *
         * @var int
         */
@@ -30,7 +29,7 @@ class GenerateSitemap {
        /**
         * The maximum size of a sitemap file
         *
-        * @link http://www.google.com/webmasters/sitemaps/docs/en/protocol.html#faq_sitemap_size
+        * @link http://www.sitemaps.org/faq.php#faq_sitemap_size
         *
         * @var int
         */
@@ -144,20 +143,37 @@ class GenerateSitemap {
         * @param string $path The path to append to the domain name
         * @param bool $compress Whether to compress the sitemap files
         */
-       function GenerateSitemap( $fspath, $path, $compress ) {
-               global $wgDBname, $wgScriptPath;
+       function GenerateSitemap( $fspath, $compress ) {
+               global $wgScriptPath;
 
                $this->url_limit = 50000;
                $this->size_limit = pow( 2, 20 ) * 10;
-               $this->fspath = isset( $fspath ) ? $fspath : '';
-               $this->path = isset( $path ) ? $path : $wgScriptPath;
+               $this->fspath = self::init_path( $fspath );
+
                $this->compress = $compress;
 
                $this->stderr = fopen( 'php://stderr', 'wt' );
-               $this->dbr =& wfGetDB( DB_SLAVE );
+               $this->dbr = wfGetDB( DB_SLAVE );
                $this->generateNamespaces();
                $this->timestamp = wfTimestamp( TS_ISO_8601, wfTimestampNow() );
-               $this->findex = fopen( "{$this->fspath}sitemap-index-$wgDBname.xml", 'wb' );
+
+
+               $this->findex = fopen( "{$this->fspath}sitemap-index-" . wfWikiID() . ".xml", 'wb' );
+       }
+
+       /**
+        * Create directory if it does not exist and return pathname with a trailing slash
+        */
+       private static function init_path( $fspath ) {
+               if( !isset( $fspath ) ) {
+                       return null;
+               }
+               # Create directory if needed
+               if( $fspath && !is_dir( $fspath ) ) {
+                       mkdir( $fspath, 0755 ) or die("Can not create directory $fspath.\n");
+               }
+
+               return realpath( $fspath ). DIRECTORY_SEPARATOR ;
        }
 
        /**
@@ -166,6 +182,13 @@ class GenerateSitemap {
        function generateNamespaces() {
                $fname = 'GenerateSitemap::generateNamespaces';
 
+               // Only generate for specific namespaces if $wgSitemapNamespaces is an array.
+               global $wgSitemapNamespaces;
+               if( is_array( $wgSitemapNamespaces ) ) {
+                       $this->namespaces = $wgSitemapNamespaces;
+                       return;
+               }
+
                $res = $this->dbr->select( 'page',
                        array( 'page_namespace' ),
                        array(),
@@ -202,7 +225,7 @@ class GenerateSitemap {
         * @return string
         */
        function guessPriority( $namespace ) {
-               return Namespace::isMain( $namespace ) ? $this->priorities[GS_MAIN] : $this->priorities[GS_TALK];
+               return MWNamespace::isMain( $namespace ) ? $this->priorities[GS_MAIN] : $this->priorities[GS_TALK];
        }
 
        /**
@@ -232,7 +255,7 @@ class GenerateSitemap {
         * @access public
         */
        function main() {
-               global $wgDBname, $wgContLang;
+               global $wgContLang;
 
                fwrite( $this->findex, $this->openIndex() );
 
@@ -255,7 +278,7 @@ class GenerateSitemap {
                                        $this->file = $this->open( $this->fspath . $filename, 'wb' );
                                        $this->write( $this->file, $this->openFile() );
                                        fwrite( $this->findex, $this->indexEntry( $filename ) );
-                                       $this->debug( "\t$filename" );
+                                       $this->debug( "\t$this->fspath$filename" );
                                        $length = $this->limit[0];
                                        $i = 1;
                                }
@@ -264,6 +287,16 @@ class GenerateSitemap {
                                $entry = $this->fileEntry( $title->getFullURL(), $date, $this->priority( $namespace ) );
                                $length += strlen( $entry );
                                $this->write( $this->file, $entry );
+                               // generate pages for language variants
+                               if($wgContLang->hasVariants()){
+                                       $variants = $wgContLang->getVariants();
+                                       foreach($variants as $vCode){
+                                               if($vCode==$wgContLang->getCode()) continue; // we don't want default variant
+                                               $entry = $this->fileEntry( $title->getFullURL('',$vCode), $date, $this->priority( $namespace ) );
+                                               $length += strlen( $entry );
+                                               $this->write( $this->file, $entry );
+                                       }
+                               }
                        }
                        if ( $this->file ) {
                                $this->write( $this->file, $this->closeFile() );
@@ -314,11 +347,8 @@ class GenerateSitemap {
         * @return string
         */
        function sitemapFilename( $namespace, $count ) {
-               global $wgDBname;
-
                $ext = $this->compress ? '.gz' : '';
-
-               return "sitemap-$wgDBname-NS_$namespace-$count.xml$ext";
+               return "sitemap-".wfWikiID()."-NS_$namespace-$count.xml$ext";
        }
 
        /**
@@ -340,7 +370,7 @@ class GenerateSitemap {
         * @returns string
         */
        function xmlSchema() {
-               return 'http://www.google.com/schemas/sitemap/0.84';
+               return 'http://www.sitemaps.org/schemas/sitemap/0.9';
        }
 
        /**
@@ -440,23 +470,30 @@ class GenerateSitemap {
        }
 }
 
-if ( in_array( '--help', $argv ) )
-       die(
-               "Usage: php generateSitemap.php [host] [options]\n" .
-               "\thost = hostname\n" .
-               "\toptions:\n" .
-               "\t\t--help\tshow this message\n" .
-               "\t\t--fspath\tThe file system path to save to, e.g /tmp/sitemap/\n" .
-               "\t\t--path\tThe http path to use, e.g. /wiki\n" .
-               "\t\t--compress=[yes|no]\tcompress the sitemap files, default yes\n"
-       );
+if ( in_array( '--help', $argv ) ) {
+       echo <<<EOT
+Usage: php generateSitemap.php [options]
+       --help                  show this message
+
+       --fspath=<path>         The file system path to save to, e.g /tmp/sitemap
 
-if ( isset( $argv[1] ) && strpos( $argv[1], '--' ) !== 0 )
-       $_SERVER['SERVER_NAME'] = $argv[1];
+       --server=<server>       The protocol and server name to use in URLs, e.g.
+               http://en.wikipedia.org. This is sometimes necessary because
+               server name detection may fail in command line scripts.
 
-$optionsWithArgs = array( 'fspath', 'path', 'compress' );
-require_once 'commandLine.inc';
+       --compress=[yes|no]     compress the sitemap files, default yes
 
-$gs = new GenerateSitemap( @$options['fspath'], @$options['path'], @$options['compress'] !== 'no' );
+EOT;
+       die( -1 );
+}
+
+$optionsWithArgs = array( 'fspath', 'server', 'compress' );
+require_once( dirname( __FILE__ ) . '/commandLine.inc' );
+
+if ( isset( $options['server'] ) ) {
+       $wgServer = $options['server'];
+}
+
+$gs = new GenerateSitemap( @$options['fspath'], @$options['compress'] !== 'no' );
 $gs->main();
-?>
+