Revert r24237, r24239, r24245 for the moment. We expect transwiki system to change...
[lhc/web/wiklou.git] / includes / SpecialImport.php
index 57a0ec7..8dee732 100644 (file)
  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  * http://www.gnu.org/copyleft/gpl.html
  *
- * @package MediaWiki
- * @subpackage SpecialPage
+ * @addtogroup SpecialPage
  */
 
-/** */
-require_once( 'WikiError.php' );
-
 /**
  * Constructor
  */
 function wfSpecialImport( $page = '' ) {
        global $wgUser, $wgOut, $wgRequest, $wgTitle, $wgImportSources;
+       global $wgImportTargetNamespace;
 
-       ###
-#      $wgOut->addWikiText( "Special:Import is not ready for this beta release, sorry." );
-#      return;
-       ###
+       $interwiki = false;
+       $namespace = $wgImportTargetNamespace;
+       $frompage = '';
+       $history = true;
 
        if( $wgRequest->wasPosted() && $wgRequest->getVal( 'action' ) == 'submit') {
+               $isUpload = false;
+               $namespace = $wgRequest->getIntOrNull( 'namespace' );
+
                switch( $wgRequest->getVal( "source" ) ) {
                case "upload":
+                       $isUpload = true;
                        if( $wgUser->isAllowed( 'importupload' ) ) {
                                $source = ImportStreamSource::newFromUpload( "xmlimport" );
                        } else {
@@ -47,19 +48,33 @@ function wfSpecialImport( $page = '' ) {
                        }
                        break;
                case "interwiki":
+                       $interwiki = $wgRequest->getVal( 'interwiki' );
+                       $history = $wgRequest->getCheck( 'interwikiHistory' );
+                       $frompage = $wgRequest->getText( "frompage" );
                        $source = ImportStreamSource::newFromInterwiki(
-                               $wgRequest->getVal( "interwiki" ),
-                               $wgRequest->getText( "frompage" ) );
+                               $interwiki,
+                               $frompage,
+                               $history );
                        break;
                default:
-                       $source = new WikiError( "Unknown import source type" );
+                       $source = new WikiErrorMsg( "importunknownsource" );
                }
 
                if( WikiError::isError( $source ) ) {
                        $wgOut->addWikiText( wfEscapeWikiText( $source->getMessage() ) );
                } else {
+                       $wgOut->addWikiText( wfMsg( "importstart" ) );
+
                        $importer = new WikiImporter( $source );
+                       if( !is_null( $namespace ) ) {
+                               $importer->setTargetNamespace( $namespace );
+                       }
+                       $reporter = new ImportReporter( $importer, $isUpload, $interwiki );
+
+                       $reporter->open();
                        $result = $importer->doImport();
+                       $reporter->close();
+
                        if( WikiError::isError( $result ) ) {
                                $wgOut->addWikiText( wfMsg( "importfailed",
                                        wfEscapeWikiText( $result->getMessage() ) ) );
@@ -70,7 +85,7 @@ function wfSpecialImport( $page = '' ) {
                }
        }
 
-       $action = $wgTitle->escapeLocalUrl( 'action=submit' );
+       $action = $wgTitle->getLocalUrl( 'action=submit' );
 
        if( $wgUser->isAllowed( 'importupload' ) ) {
                $wgOut->addWikiText( wfMsg( "importtext" ) );
@@ -82,7 +97,7 @@ function wfSpecialImport( $page = '' ) {
                <input type='hidden' name='source' value='upload' />
                <input type='hidden' name='MAX_FILE_SIZE' value='2000000' />
                <input type='file' name='xmlimport' value='' size='30' />
-               <input type='submit' value='" . wfMsgHtml( "uploadbtn" ) . "'/>
+               <input type='submit' value=\"" . wfMsgHtml( "uploadbtn" ) . "\" />
        </form>
 </fieldset>
 " );
@@ -96,32 +111,122 @@ function wfSpecialImport( $page = '' ) {
                $wgOut->addHTML( "
 <fieldset>
        <legend>" . wfMsgHtml('importinterwiki') . "</legend>
-       <form method='post' action=\"$action\">
+       <form method='post' action=\"$action\">" .
+               $wgOut->parse( wfMsg( 'import-interwiki-text' ) ) . "
                <input type='hidden' name='action' value='submit' />
                <input type='hidden' name='source' value='interwiki' />
-               <select name='interwiki'>
-" );
-               foreach( $wgImportSources as $interwiki ) {
-                       $iw = htmlspecialchars( $interwiki );
-                       $wgOut->addHTML( "<option value=\"$iw\">$iw</option>\n" );
+               <table>
+                       <tr>
+                               <td>
+                                       <select name='interwiki'>" );
+               foreach( $wgImportSources as $prefix ) {
+                       $iw = htmlspecialchars( $prefix );
+                       $selected = ($interwiki === $prefix) ? ' selected="selected"' : '';
+                       $wgOut->addHTML( "<option value=\"$iw\"$selected>$iw</option>\n" );
                }
                $wgOut->addHTML( "
-               </select>
-               <input name='frompage' />
-               <input type='submit' />
+                                       </select>
+                               </td>
+                               <td>" .
+                                       wfInput( 'frompage', 50, $frompage ) .
+                               "</td>
+                       </tr>
+                       <tr>
+                               <td></td>
+                               <td>" .
+                                       wfCheckLabel( wfMsg( 'import-interwiki-history' ),
+                                               'interwikiHistory', 'interwikiHistory', $history ) .
+                               "</td>
+                       </tr>
+                       <tr>
+                               <td></td>
+                               <td>
+                                       " . wfMsgHtml( 'import-interwiki-namespace' ) . " " .
+                                               HTMLnamespaceselector( $namespace, '' ) . "
+                               </td>
+                       </tr>
+                       <tr>
+                               <td></td>
+                               <td>" .
+                                       wfSubmitButton( wfMsg( 'import-interwiki-submit' ) ) .
+                               "</td>
+                       </tr>
+               </table>
        </form>
 </fieldset>
 " );
        }
 }
 
+/**
+ * Reporting callback
+ * @addtogroup SpecialPage
+ */
+class ImportReporter {
+       function __construct( $importer, $upload, $interwiki ) {
+               $importer->setPageOutCallback( array( $this, 'reportPage' ) );
+               $this->mPageCount = 0;
+               $this->mIsUpload = $upload;
+               $this->mInterwiki = $interwiki;
+       }
+
+       function open() {
+               global $wgOut;
+               $wgOut->addHtml( "<ul>\n" );
+       }
+
+       function reportPage( $title, $origTitle, $revisionCount, $successCount ) {
+               global $wgOut, $wgUser, $wgLang, $wgContLang;
+
+               $skin = $wgUser->getSkin();
+
+               $this->mPageCount++;
+
+               $localCount = $wgLang->formatNum( $successCount );
+               $contentCount = $wgContLang->formatNum( $successCount );
+
+               $wgOut->addHtml( "<li>" . $skin->makeKnownLinkObj( $title ) .
+                       " " .
+                       wfMsgExt( 'import-revision-count', array( 'parsemag', 'escape' ), $localCount ) .
+                       "</li>\n" );
+
+               if( $successCount > 0 ) {
+                       $log = new LogPage( 'import' );
+                       if( $this->mIsUpload ) {
+                               $detail = wfMsgForContent( 'import-logentry-upload-detail',
+                                       $contentCount );
+                               $log->addEntry( 'upload', $title, $detail );
+                       } else {
+                               $interwiki = '[[:' . $this->mInterwiki . ':' .
+                                       $origTitle->getPrefixedText() . ']]';
+                               $detail = wfMsgForContent( 'import-logentry-interwiki-detail',
+                                       $contentCount, $interwiki );
+                               $log->addEntry( 'interwiki', $title, $detail );
+                       }
+
+                       $comment = $detail; // quick
+                       $dbw = wfGetDB( DB_MASTER );
+                       $nullRevision = Revision::newNullRevision(
+                               $dbw, $title->getArticleId(), $comment, true );
+                       $nullRevision->insertOn( $dbw );
+               }
+       }
+
+       function close() {
+               global $wgOut;
+               if( $this->mPageCount == 0 ) {
+                       $wgOut->addHtml( "<li>" . wfMsgHtml( 'importnopages' ) . "</li>\n" );
+               }
+               $wgOut->addHtml( "</ul>\n" );
+       }
+}
+
 /**
  *
- * @package MediaWiki
- * @subpackage SpecialPage
+ * @addtogroup SpecialPage
  */
 class WikiRevision {
-       var $title = NULL;
+       var $title = null;
        var $id = 0;
        var $timestamp = "20010115000000";
        var $user = 0;
@@ -130,8 +235,14 @@ class WikiRevision {
        var $comment = "";
        var $minor = false;
 
-       function setTitle( $text ) {
-               $this->title = Title::newFromText( $text );
+       function setTitle( $title ) {
+               if( is_object( $title ) ) {
+                       $this->title = $title;
+               } elseif( is_null( $title ) ) {
+                       throw new MWException( "WikiRevision given a null title in import. You may need to adjust \$wgLegalTitleChars." );
+               } else {
+                       throw new MWException( "WikiRevision given non-object title in import." );
+               }
        }
 
        function setID( $id ) {
@@ -167,7 +278,7 @@ class WikiRevision {
                return $this->title;
        }
 
-       function getID() { 
+       function getID() {
                return $this->id;
        }
 
@@ -192,8 +303,7 @@ class WikiRevision {
        }
 
        function importOldRevision() {
-               $fname = "WikiImporter::importOldRevision";
-               $dbw =& wfGetDB( DB_MASTER );
+               $dbw = wfGetDB( DB_MASTER );
 
                # Sneak a single revision into place
                $user = User::newFromName( $this->getUser() );
@@ -214,9 +324,20 @@ class WikiRevision {
                if( $pageId == 0 ) {
                        # must create the page...
                        $pageId = $article->insertOn( $dbw );
+                       $created = true;
+               } else {
+                       $created = false;
+
+                       $prior = Revision::loadFromTimestamp( $dbw, $this->title, $this->timestamp );
+                       if( !is_null( $prior ) ) {
+                               // FIXME: this could fail slightly for multiple matches :P
+                               wfDebug( __METHOD__ . ": skipping existing revision for [[" .
+                                       $this->title->getPrefixedText() . "]], timestamp " .
+                                       $this->timestamp . "\n" );
+                               return false;
+                       }
                }
 
-               # FIXME: Check for exact conflicts
                # FIXME: Use original rev_id optionally
                # FIXME: blah blah blah
 
@@ -235,7 +356,27 @@ class WikiRevision {
                        'minor_edit' => $this->minor,
                        ) );
                $revId = $revision->insertOn( $dbw );
-               $article->updateIfNewerOn( $dbw, $revision );
+               $changed = $article->updateIfNewerOn( $dbw, $revision );
+
+               if( $created ) {
+                       wfDebug( __METHOD__ . ": running onArticleCreate\n" );
+                       Article::onArticleCreate( $this->title );
+
+                       wfDebug( __METHOD__ . ": running create updates\n" );
+                       $article->createUpdates( $revision );
+
+               } elseif( $changed ) {
+                       wfDebug( __METHOD__ . ": running onArticleEdit\n" );
+                       Article::onArticleEdit( $this->title );
+
+                       wfDebug( __METHOD__ . ": running edit updates\n" );
+                       $article->editUpdates(
+                               $this->getText(),
+                               $this->getComment(),
+                               $this->minor,
+                               $this->timestamp,
+                               $revId );
+               }
 
                return true;
        }
@@ -243,14 +384,15 @@ class WikiRevision {
 }
 
 /**
- *
- * @package MediaWiki
- * @subpackage SpecialPage
+ * implements Special:Import
+ * @addtogroup SpecialPage
  */
 class WikiImporter {
        var $mSource = null;
        var $mPageCallback = null;
+       var $mPageOutCallback = null;
        var $mRevisionCallback = null;
+       var $mTargetNamespace = null;
        var $lastfield;
 
        function WikiImporter( $source ) {
@@ -302,7 +444,7 @@ class WikiImporter {
                        print "$data\n";
                } else {
                        global $wgOut;
-                       $wgOut->addHTML( "<li>$data</li>\n" );
+                       $wgOut->addHTML( "<li>" . htmlspecialchars( $data ) . "</li>\n" );
                }
        }
 
@@ -317,6 +459,21 @@ class WikiImporter {
                return $previous;
        }
 
+       /**
+        * Sets the action to perform as each page in the stream is completed.
+        * Callback accepts the page title (as a Title object), a second object
+        * with the original title form (in case it's been overridden into a
+        * local namespace), and a count of revisions.
+        *
+        * @param callable $callback
+        * @return callable
+        */
+       function setPageOutCallback( $callback ) {
+               $previous = $this->mPageOutCallback;
+               $this->mPageOutCallback = $callback;
+               return $previous;
+       }
+
        /**
         * Sets the action to perform as each page revision is reached.
         * @param callable $callback
@@ -328,20 +485,35 @@ class WikiImporter {
                return $previous;
        }
 
+       /**
+        * Set a target namespace to override the defaults
+        */
+       function setTargetNamespace( $namespace ) {
+               if( is_null( $namespace ) ) {
+                       // Don't override namespaces
+                       $this->mTargetNamespace = null;
+               } elseif( $namespace >= 0 ) {
+                       // FIXME: Check for validity
+                       $this->mTargetNamespace = intval( $namespace );
+               } else {
+                       return false;
+               }
+       }
+
        /**
         * Default per-revision callback, performs the import.
         * @param WikiRevision $revision
-        * @access private
+        * @private
         */
        function importRevision( &$revision ) {
-               $dbw =& wfGetDB( DB_MASTER );
-               $dbw->deadlockLoop( array( &$revision, 'importOldRevision' ) );
+               $dbw = wfGetDB( DB_MASTER );
+               return $dbw->deadlockLoop( array( &$revision, 'importOldRevision' ) );
        }
 
        /**
         * Alternate per-revision callback, for debugging.
         * @param WikiRevision $revision
-        * @access private
+        * @private
         */
        function debugRevisionHandler( &$revision ) {
                $this->debug( "Got revision:" );
@@ -359,7 +531,7 @@ class WikiImporter {
        /**
         * Notify the callback function when a new <page> is reached.
         * @param Title $title
-        * @access private
+        * @private
         */
        function pageCallback( $title ) {
                if( is_callable( $this->mPageCallback ) ) {
@@ -367,6 +539,21 @@ class WikiImporter {
                }
        }
 
+       /**
+        * Notify the callback function when a </page> is closed.
+        * @param Title $title
+        * @param Title $origTitle
+        * @param int $revisionCount
+        * @param int $successCount number of revisions for which callback returned true
+        * @private
+        */
+       function pageOutCallback( $title, $origTitle, $revisionCount, $successCount ) {
+               if( is_callable( $this->mPageOutCallback ) ) {
+                       call_user_func( $this->mPageOutCallback, $title, $origTitle,
+                               $revisionCount, $successCount );
+               }
+       }
+
 
        # XML parser callbacks from here out -- beware!
        function donothing( $parser, $x, $y="" ) {
@@ -386,6 +573,8 @@ class WikiImporter {
                if( $name == 'siteinfo' ) {
                        xml_set_element_handler( $parser, "in_siteinfo", "out_siteinfo" );
                } elseif( $name == 'page' ) {
+                       $this->workRevisionCount = 0;
+                       $this->workSuccessCount = 0;
                        xml_set_element_handler( $parser, "in_page", "out_page" );
                } else {
                        return $this->throwXMLerror( "Expected <page>, got <$name>" );
@@ -436,8 +625,14 @@ class WikiImporter {
                        xml_set_character_data_handler( $parser, "char_append" );
                        break;
                case "revision":
-                       $this->workRevision = new WikiRevision;
-                       $this->workRevision->setTitle( $this->workTitle );
+                       if( is_object( $this->pageTitle ) ) {
+                               $this->workRevision = new WikiRevision;
+                               $this->workRevision->setTitle( $this->pageTitle );
+                               $this->workRevisionCount++;
+                       } else {
+                               // Skipping items due to invalid page title
+                               $this->workRevision = null;
+                       }
                        xml_set_element_handler( $parser, "in_revision", "out_revision" );
                        break;
                default:
@@ -452,8 +647,15 @@ class WikiImporter {
                }
                xml_set_element_handler( $parser, "in_mediawiki", "out_mediawiki" );
 
-               $this->workTitle = NULL;
-               $this->workRevision = NULL;
+               $this->pageOutCallback( $this->pageTitle, $this->origTitle,
+                       $this->workRevisionCount, $this->workSuccessCount );
+
+               $this->workTitle = null;
+               $this->workRevision = null;
+               $this->workRevisionCount = 0;
+               $this->workSuccessCount = 0;
+               $this->pageTitle = null;
+               $this->origTitle = null;
        }
 
        function in_nothing( $parser, $name, $attribs ) {
@@ -475,30 +677,49 @@ class WikiImporter {
                switch( $this->appendfield ) {
                case "title":
                        $this->workTitle = $this->appenddata;
-                       $this->pageCallback( $this->workTitle );
+                       $this->origTitle = Title::newFromText( $this->workTitle );
+                       if( !is_null( $this->mTargetNamespace ) && !is_null( $this->origTitle ) ) {
+                               $this->pageTitle = Title::makeTitle( $this->mTargetNamespace,
+                                       $this->origTitle->getDbKey() );
+                       } else {
+                               $this->pageTitle = Title::newFromText( $this->workTitle );
+                       }
+                       if( is_null( $this->pageTitle ) ) {
+                               // Invalid page title? Ignore the page
+                               $this->notice( "Skipping invalid page title '$this->workTitle'" );
+                       } else {
+                               $this->pageCallback( $this->workTitle );
+                       }
                        break;
                case "id":
                        if ( $this->parenttag == 'revision' ) {
-                               $this->workRevision->setID( $this->appenddata );
+                               if( $this->workRevision )
+                                       $this->workRevision->setID( $this->appenddata );
                        }
                        break;
                case "text":
-                       $this->workRevision->setText( $this->appenddata );
+                       if( $this->workRevision )
+                               $this->workRevision->setText( $this->appenddata );
                        break;
                case "username":
-                       $this->workRevision->setUsername( $this->appenddata );
+                       if( $this->workRevision )
+                               $this->workRevision->setUsername( $this->appenddata );
                        break;
                case "ip":
-                       $this->workRevision->setUserIP( $this->appenddata );
+                       if( $this->workRevision )
+                               $this->workRevision->setUserIP( $this->appenddata );
                        break;
                case "timestamp":
-                       $this->workRevision->setTimestamp( $this->appenddata );
+                       if( $this->workRevision )
+                               $this->workRevision->setTimestamp( $this->appenddata );
                        break;
                case "comment":
-                       $this->workRevision->setComment( $this->appenddata );
+                       if( $this->workRevision )
+                               $this->workRevision->setComment( $this->appenddata );
                        break;
                case "minor":
-                       $this->workRevision->setMinor( true );
+                       if( $this->workRevision )
+                               $this->workRevision->setMinor( true );
                        break;
                default:
                        $this->debug( "Bad append: {$this->appendfield}" );
@@ -535,11 +756,12 @@ class WikiImporter {
                }
                xml_set_element_handler( $parser, "in_page", "out_page" );
 
-               $out = call_user_func_array( $this->mRevisionCallback,
-                       array( &$this->workRevision, &$this ) );
-               if( !empty( $out ) ) {
-                       global $wgOut;
-                       $wgOut->addHTML( "<li>" . $out . "</li>\n" );
+               if( $this->workRevision ) {
+                       $ok = call_user_func_array( $this->mRevisionCallback,
+                               array( &$this->workRevision, &$this ) );
+                       if( $ok ) {
+                               $this->workSuccessCount++;
+                       }
                }
        }
 
@@ -569,7 +791,10 @@ class WikiImporter {
 
 }
 
-/** @package MediaWiki */
+/**
+ * @todo document (e.g. one-sentence class description).
+ * @addtogroup SpecialPage
+ */
 class ImportStringSource {
        function ImportStringSource( $string ) {
                $this->mString = $string;
@@ -590,7 +815,10 @@ class ImportStringSource {
        }
 }
 
-/** @package MediaWiki */
+/**
+ * @todo document (e.g. one-sentence class description).
+ * @addtogroup SpecialPage
+ */
 class ImportStreamSource {
        function ImportStreamSource( $handle ) {
                $this->mHandle = $handle;
@@ -604,15 +832,15 @@ class ImportStreamSource {
                return fread( $this->mHandle, 32768 );
        }
 
-       function newFromFile( $filename ) {
+       static function newFromFile( $filename ) {
                $file = @fopen( $filename, 'rt' );
                if( !$file ) {
-                       return new WikiError( "Couldn't open import file" );
+                       return new WikiErrorMsg( "importcantopen" );
                }
                return new ImportStreamSource( $file );
        }
 
-       function newFromUpload( $fieldname = "xmlimport" ) {
+       static function newFromUpload( $fieldname = "xmlimport" ) {
                $upload =& $_FILES[$fieldname];
 
                if( !isset( $upload ) || !$upload['name'] ) {
@@ -629,25 +857,36 @@ class ImportStreamSource {
                }
        }
 
-       function newFromURL( $url ) {
-               # fopen-wrappers are normally turned off for security.
-               ini_set( "allow_url_fopen", true );
-               $ret = ImportStreamSource::newFromFile( $url );
-               ini_set( "allow_url_fopen", false );
-               return $ret;
+       function newFromURL( $url, $method = 'GET' ) {
+               wfDebug( __METHOD__ . ": opening $url\n" );
+               # Use the standard HTTP fetch function; it times out
+               # quicker and sorts out user-agent problems which might
+               # otherwise prevent importing from large sites, such
+               # as the Wikimedia cluster, etc.
+               $data = Http::request( $method, $url );
+               if( $data !== false ) {
+                       $file = tmpfile();
+                       fwrite( $file, $data );
+                       fflush( $file );
+                       fseek( $file, 0 );
+                       return new ImportStreamSource( $file );
+               } else {
+                       return new WikiErrorMsg( 'importcantopen' );
+               }
        }
 
-       function newFromInterwiki( $interwiki, $page ) {
-               $base = Title::getInterwikiLink( $interwiki );
-               if( empty( $base ) ) {
-                       return new WikiError( 'Bad interwiki link' );
+       public static function newFromInterwiki( $interwiki, $page, $history=false ) {
+               $link = Title::newFromText( "$interwiki:Special:Export/$page" );
+               if( is_null( $link ) || $link->getInterwiki() == '' ) {
+                       return new WikiErrorMsg( 'importbadinterwiki' );
                } else {
-                       $import = wfUrlencode( "Special:Export/$page" );
-                       $url = str_replace( "$1", $import, $base );
-                       return ImportStreamSource::newFromURL( $url );
+                       $params = $history ? 'history=1' : '';
+                       $url = $link->getFullUrl( $params );
+                       # For interwikis, use POST to avoid redirects.
+                       return ImportStreamSource::newFromURL( $url, "POST" );
                }
        }
 }
 
 
-?>
+