Move this till after ID set
[lhc/web/wiklou.git] / includes / SpecialExport.php
index 15d11d4..38bfc83 100644 (file)
 #
 # You should have received a copy of the GNU General Public License along
 # with this program; if not, write to the Free Software Foundation, Inc.,
-# 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
+# 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
 # http://www.gnu.org/copyleft/gpl.html
 /**
- *
- * @package MediaWiki
- * @subpackage SpecialPage
+ * @file
+ * @ingroup SpecialPage
  */
 
-/** */
-require_once( 'Revision.php' );
-require_once( 'Export.php' );
+function wfExportGetPagesFromCategory( $title ) {
+       global $wgContLang;
+
+       $name = $title->getDBkey();
+
+       $dbr = wfGetDB( DB_SLAVE );
+
+       list( $page, $categorylinks ) = $dbr->tableNamesN( 'page', 'categorylinks' );
+       $sql = "SELECT page_namespace, page_title FROM $page " .
+               "JOIN $categorylinks ON cl_from = page_id " .
+               "WHERE cl_to = " . $dbr->addQuotes( $name );
+
+       $pages = array();
+       $res = $dbr->query( $sql, 'wfExportGetPagesFromCategory' );
+       while ( $row = $dbr->fetchObject( $res ) ) {
+               $n = $row->page_title;
+               if ($row->page_namespace) {
+                       $ns = $wgContLang->getNsText( $row->page_namespace );
+                       $n = $ns . ':' . $n;
+               }
+
+               $pages[] = $n;
+       }
+       $dbr->freeResult($res);
+
+       return $pages;
+}
+
+/**
+ * Expand a list of pages to include templates used in those pages.
+ * @param $inputPages array, list of titles to look up
+ * @param $pageSet array, associative array indexed by titles for output
+ * @return array associative array index by titles
+ */
+function wfExportGetTemplates( $inputPages, $pageSet ) {
+       return wfExportGetLinks( $inputPages, $pageSet,
+               'templatelinks',
+               array( 'tl_namespace AS namespace', 'tl_title AS title' ),
+               array( 'page_id=tl_from' ) );
+}
+
+/**
+ * Expand a list of pages to include images used in those pages.
+ * @param $inputPages array, list of titles to look up
+ * @param $pageSet array, associative array indexed by titles for output
+ * @return array associative array index by titles
+ */
+function wfExportGetImages( $inputPages, $pageSet ) {
+       return wfExportGetLinks( $inputPages, $pageSet,
+               'imagelinks',
+               array( NS_IMAGE . ' AS namespace', 'il_to AS title' ),
+               array( 'page_id=il_from' ) );
+}
+
+/**
+ * Expand a list of pages to include items used in those pages.
+ * @private
+ */
+function wfExportGetLinks( $inputPages, $pageSet, $table, $fields, $join ) {
+       $dbr = wfGetDB( DB_SLAVE );
+       foreach( $inputPages as $page ) {
+               $title = Title::newFromText( $page );
+               if( $title ) {
+                       $pageSet[$title->getPrefixedText()] = true;
+                       /// @fixme May or may not be more efficient to batch these
+                       ///        by namespace when given multiple input pages.
+                       $result = $dbr->select(
+                               array( 'page', $table ),
+                               $fields,
+                               array_merge( $join,
+                                       array(
+                                               'page_namespace' => $title->getNamespace(),
+                                               'page_title' => $title->getDbKey() ) ),
+                               __METHOD__ );
+                       foreach( $result as $row ) {
+                               $template = Title::makeTitle( $row->namespace, $row->title );
+                               $pageSet[$template->getPrefixedText()] = true;
+                       }
+               }
+       }
+       return $pageSet;
+}
+
+/**
+ * Callback function to remove empty strings from the pages array.
+ */
+function wfFilterPage( $page ) {
+       return $page !== '' && $page !== null;
+}
 
 /**
  *
  */
 function wfSpecialExport( $page = '' ) {
-       global $wgOut, $wgRequest, $wgExportAllowListContributors;
-       global $wgExportAllowHistory;
+       global $wgOut, $wgRequest, $wgSitename, $wgExportAllowListContributors;
+       global $wgExportAllowHistory, $wgExportMaxHistory;
 
-       if( $wgRequest->getVal( 'action' ) == 'submit') {
+       $curonly = true;
+       $doexport = false;
+
+       if ( $wgRequest->getCheck( 'addcat' ) ) {
                $page = $wgRequest->getText( 'pages' );
-               if( $wgExportAllowHistory ) {
-                       $curonly = $wgRequest->getCheck( 'curonly' );
+               $catname = $wgRequest->getText( 'catname' );
+
+               if ( $catname !== '' && $catname !== NULL && $catname !== false ) {
+                       $t = Title::makeTitleSafe( NS_CATEGORY, $catname );
+                       if ( $t ) {
+                               /**
+                                * @fixme This can lead to hitting memory limit for very large
+                                * categories. Ideally we would do the lookup synchronously
+                                * during the export in a single query.
+                                */
+                               $catpages = wfExportGetPagesFromCategory( $t );
+                               if ( $catpages ) $page .= "\n" . implode( "\n", $catpages );
+                       }
+               }
+       }
+       else if( $wgRequest->wasPosted() && $page == '' ) {
+               $page = $wgRequest->getText( 'pages' );
+               $curonly = $wgRequest->getCheck( 'curonly' );
+               $rawOffset = $wgRequest->getVal( 'offset' );
+               if( $rawOffset ) {
+                       $offset = wfTimestamp( TS_MW, $rawOffset );
                } else {
-                       $curonly = true;
+                       $offset = null;
                }
+               $limit = $wgRequest->getInt( 'limit' );
+               $dir = $wgRequest->getVal( 'dir' );
+               $history = array(
+                       'dir' => 'asc',
+                       'offset' => false,
+                       'limit' => $wgExportMaxHistory,
+               );
+               $historyCheck = $wgRequest->getCheck( 'history' );
+               if ( $curonly ) {
+                       $history = WikiExporter::CURRENT;
+               } elseif ( !$historyCheck ) {
+                       if ( $limit > 0 && $limit < $wgExportMaxHistory ) {
+                               $history['limit'] = $limit;
+                       }
+                       if ( !is_null( $offset ) ) {
+                               $history['offset'] = $offset;
+                       }
+                       if ( strtolower( $dir ) == 'desc' ) {
+                               $history['dir'] = 'desc';
+                       }
+               }
+
+               if( $page != '' ) $doexport = true;
        } else {
-               # Pre-check the 'current version only' box in the UI
-               $curonly = true;
+               // Default to current-only for GET requests
+               $page = $wgRequest->getText( 'pages', $page );
+               $historyCheck = $wgRequest->getCheck( 'history' );
+               if( $historyCheck ) {
+                       $history = WikiExporter::FULL;
+               } else {
+                       $history = WikiExporter::CURRENT;
+               }
+
+               if( $page != '' ) $doexport = true;
        }
-       
+
+       if( !$wgExportAllowHistory ) {
+               // Override
+               $history = WikiExporter::CURRENT;
+       }
+
        $list_authors = $wgRequest->getCheck( 'listauthors' );
        if ( !$curonly || !$wgExportAllowListContributors ) $list_authors = false ;
 
-       if( $page != '' ) {
+       if ( $doexport ) {
                $wgOut->disable();
+
+               // Cancel output buffering and gzipping if set
+               // This should provide safer streaming for pages with history
+               wfResetOutputBuffers();
                header( "Content-type: application/xml; charset=utf-8" );
-               $pages = explode( "\n", $page );
+               if( $wgRequest->getCheck( 'wpDownload' ) ) {
+                       // Provide a sane filename suggestion
+                       $filename = urlencode( $wgSitename . '-' . wfTimestampNow() . '.xml' );
+                       $wgRequest->response()->header( "Content-disposition: attachment;filename={$filename}" );
+               }
+
+               /* Split up the input and look up linked pages */
+               $inputPages = array_filter( explode( "\n", $page ), 'wfFilterPage' );
+               $pageSet = array_flip( $inputPages );
+
+               if( $wgRequest->getCheck( 'templates' ) ) {
+                       $pageSet = wfExportGetTemplates( $inputPages, $pageSet );
+               }
 
-               $db =& wfGetDB( DB_SLAVE );
-               $history = $curonly ? MW_EXPORT_CURRENT : MW_EXPORT_FULL;
+               /*
+               // Enable this when we can do something useful exporting/importing image information. :)
+               if( $wgRequest->getCheck( 'images' ) ) {
+                       $pageSet = wfExportGetImages( $inputPages, $pageSet );
+               }
+               */
+
+               $pages = array_keys( $pageSet );
+
+               /* Ok, let's get to it... */
+
+               $db = wfGetDB( DB_SLAVE );
                $exporter = new WikiExporter( $db, $history );
                $exporter->list_authors = $list_authors ;
                $exporter->openStream();
-               $exporter->pagesByName( $pages );
+
+               foreach( $pages as $page ) {
+                       /*
+                       if( $wgExportMaxHistory && !$curonly ) {
+                               $title = Title::newFromText( $page );
+                               if( $title ) {
+                                       $count = Revision::countByTitle( $db, $title );
+                                       if( $count > $wgExportMaxHistory ) {
+                                               wfDebug( __FUNCTION__ .
+                                                       ": Skipped $page, $count revisions too big\n" );
+                                               continue;
+                                       }
+                               }
+                       }*/
+
+                       #Bug 8824: Only export pages the user can read
+                       $title = Title::newFromText( $page );
+                       if( is_null( $title ) ) continue; #TODO: perhaps output an <error> tag or something.
+                       if( !$title->userCanRead() ) continue; #TODO: perhaps output an <error> tag or something.
+
+                       $exporter->pageByTitle( $title );
+               }
+
                $exporter->closeStream();
                return;
        }
 
-       $wgOut->addWikiText( wfMsg( "exporttext" ) );
-       $titleObj = Title::makeTitle( NS_SPECIAL, "Export" );
-       $action = $titleObj->escapeLocalURL( 'action=submit' );
+       $self = SpecialPage::getTitleFor( 'Export' );
+       $wgOut->addHtml( wfMsgExt( 'exporttext', 'parse' ) );
+
+       $form = Xml::openElement( 'form', array( 'method' => 'post',
+               'action' => $self->getLocalUrl( 'action=submit' ) ) );
+
+       $form .= Xml::inputLabel( wfMsg( 'export-addcattext' )  , 'catname', 'catname', 40 ) . '&nbsp;';
+       $form .= Xml::submitButton( wfMsg( 'export-addcat' ), array( 'name' => 'addcat' ) ) . '<br />';
+
+       $form .= Xml::openElement( 'textarea', array( 'name' => 'pages', 'cols' => 40, 'rows' => 10 ) );
+       $form .= htmlspecialchars( $page );
+       $form .= Xml::closeElement( 'textarea' );
+       $form .= '<br />';
+
        if( $wgExportAllowHistory ) {
-               $checkbox = "<label><input type='checkbox' name='curonly' value='true' checked='checked' />
-" . wfMsgHtml( 'exportcuronly' ) . "</label><br />";
+               $form .= Xml::checkLabel( wfMsg( 'exportcuronly' ), 'curonly', 'curonly', true ) . '<br />';
        } else {
-               $checkbox = "";
-               $wgOut->addWikiText( wfMsg( "exportnohistory" ) );
+               $wgOut->addHtml( wfMsgExt( 'exportnohistory', 'parse' ) );
        }
-       $wgOut->addHTML( "
-<form method='post' action=\"$action\">
-<input type='hidden' name='action' value='submit' />
-<textarea name='pages' cols='40' rows='10'></textarea><br />
-$checkbox
-<input type='submit' />
-</form>
-" );
-}
+       $form .= Xml::checkLabel( wfMsg( 'export-templates' ), 'templates', 'wpExportTemplates', false ) . '<br />';
+       // Enable this when we can do something useful exporting/importing image information. :)
+       //$form .= Xml::checkLabel( wfMsg( 'export-images' ), 'images', 'wpExportImages', false ) . '<br />';
+       $form .= Xml::checkLabel( wfMsg( 'export-download' ), 'wpDownload', 'wpDownload', true ) . '<br />';
 
-?>
+       $form .= Xml::submitButton( wfMsg( 'export-submit' ) );
+       $form .= Xml::closeElement( 'form' );
+       $wgOut->addHtml( $form );
+}