3 * Base classes for dumps and export
5 * Copyright © 2003, 2005, 2006 Brion Vibber <brion@pobox.com>
6 * http://www.mediawiki.org/
8 * This program is free software; you can redistribute it and/or modify
9 * it under the terms of the GNU General Public License as published by
10 * the Free Software Foundation; either version 2 of the License, or
11 * (at your option) any later version.
13 * This program is distributed in the hope that it will be useful,
14 * but WITHOUT ANY WARRANTY; without even the implied warranty of
15 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
16 * GNU General Public License for more details.
18 * You should have received a copy of the GNU General Public License along
19 * with this program; if not, write to the Free Software Foundation, Inc.,
20 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
21 * http://www.gnu.org/copyleft/gpl.html
31 * @ingroup SpecialPage Dump
34 var $list_authors = false ; # Return distinct author list (when not returning full history)
35 var $author_list = "" ;
37 var $dumpUploads = false;
38 var $dumpUploadFileContents = false;
42 const STABLE
= 4; // extension defined
52 * If using WikiExporter::STREAM to stream a large amount of data,
53 * provide a database connection which is not managed by
54 * LoadBalancer to read from: some history blob types will
55 * make additional queries to pull source data while the
56 * main query is still running.
59 * @param $history Mixed: one of WikiExporter::FULL or WikiExporter::CURRENT,
60 * or an associative array:
61 * offset: non-inclusive offset at which to start the query
62 * limit: maximum number of rows to return
63 * dir: "asc" or "desc" timestamp order
64 * @param $buffer Int: one of WikiExporter::BUFFER or WikiExporter::STREAM
65 * @param $text Int: one of WikiExporter::TEXT or WikiExporter::STUB
67 function __construct( &$db, $history = WikiExporter
::CURRENT
,
68 $buffer = WikiExporter
::BUFFER
, $text = WikiExporter
::TEXT
) {
70 $this->history
= $history;
71 $this->buffer
= $buffer;
72 $this->writer
= new XmlDumpWriter();
73 $this->sink
= new DumpOutput();
78 * Set the DumpOutput or DumpFilter object which will receive
79 * various row objects and XML output for filtering. Filters
80 * can be chained or used as callbacks.
84 public function setOutputSink( &$sink ) {
88 public function openStream() {
89 $output = $this->writer
->openStream();
90 $this->sink
->writeOpenStream( $output );
93 public function closeStream() {
94 $output = $this->writer
->closeStream();
95 $this->sink
->writeCloseStream( $output );
99 * Dumps a series of page and revision records for all pages
100 * in the database, either including complete history or only
101 * the most recent version.
103 public function allPages() {
104 return $this->dumpFrom( '' );
108 * Dumps a series of page and revision records for those pages
109 * in the database falling within the page_id range given.
110 * @param $start Int: inclusive lower limit (this id is included)
111 * @param $end Int: Exclusive upper limit (this id is not included)
112 * If 0, no upper limit.
114 public function pagesByRange( $start, $end ) {
115 $condition = 'page_id >= ' . intval( $start );
117 $condition .= ' AND page_id < ' . intval( $end );
119 return $this->dumpFrom( $condition );
123 * @param $title Title
125 public function pageByTitle( $title ) {
126 return $this->dumpFrom(
127 'page_namespace=' . $title->getNamespace() .
128 ' AND page_title=' . $this->db
->addQuotes( $title->getDBkey() ) );
131 public function pageByName( $name ) {
132 $title = Title
::newFromText( $name );
133 if ( is_null( $title ) ) {
134 throw new MWException( "Can't export invalid title" );
136 return $this->pageByTitle( $title );
140 public function pagesByName( $names ) {
141 foreach ( $names as $name ) {
142 $this->pageByName( $name );
146 public function allLogs() {
147 return $this->dumpFrom( '' );
150 public function logsByRange( $start, $end ) {
151 $condition = 'log_id >= ' . intval( $start );
153 $condition .= ' AND log_id < ' . intval( $end );
155 return $this->dumpFrom( $condition );
158 # Generates the distinct list of authors of an article
159 # Not called by default (depends on $this->list_authors)
160 # Can be set by Special:Export when not exporting whole history
161 protected function do_list_authors( $cond ) {
162 wfProfileIn( __METHOD__
);
163 $this->author_list
= "<contributors>";
166 $res = $this->db
->select(
167 array( 'page', 'revision' ),
168 array( 'DISTINCT rev_user_text', 'rev_user' ),
170 $this->db
->bitAnd( 'rev_deleted', Revision
::DELETED_USER
) . ' = 0',
177 foreach ( $res as $row ) {
178 $this->author_list
.= "<contributor>" .
180 htmlentities( $row->rev_user_text
) .
187 $this->author_list
.= "</contributors>";
188 wfProfileOut( __METHOD__
);
191 protected function dumpFrom( $cond = '' ) {
192 wfProfileIn( __METHOD__
);
193 # For logging dumps...
194 if ( $this->history
& self
::LOGS
) {
195 if ( $this->buffer
== WikiExporter
::STREAM
) {
196 $prev = $this->db
->bufferResults( false );
198 $where = array( 'user_id = log_user' );
200 $hideLogs = LogEventsList
::getExcludeClause( $this->db
);
201 if ( $hideLogs ) $where[] = $hideLogs;
202 # Add on any caller specified conditions
203 if ( $cond ) $where[] = $cond;
204 # Get logging table name for logging.* clause
205 $logging = $this->db
->tableName( 'logging' );
206 $result = $this->db
->select( array( 'logging', 'user' ),
207 array( "{$logging}.*", 'user_name' ), // grab the user name
210 array( 'ORDER BY' => 'log_id', 'USE INDEX' => array( 'logging' => 'PRIMARY' ) )
212 $wrapper = $this->db
->resultObject( $result );
213 $this->outputLogStream( $wrapper );
214 if ( $this->buffer
== WikiExporter
::STREAM
) {
215 $this->db
->bufferResults( $prev );
219 $tables = array( 'page', 'revision' );
220 $opts = array( 'ORDER BY' => 'page_id ASC' );
221 $opts['USE INDEX'] = array();
223 if ( is_array( $this->history
) ) {
224 # Time offset/limit for all pages/history...
225 $revJoin = 'page_id=rev_page';
227 if ( $this->history
['dir'] == 'asc' ) {
229 $opts['ORDER BY'] = 'rev_timestamp ASC';
232 $opts['ORDER BY'] = 'rev_timestamp DESC';
235 if ( !empty( $this->history
['offset'] ) ) {
236 $revJoin .= " AND rev_timestamp $op " .
237 $this->db
->addQuotes( $this->db
->timestamp( $this->history
['offset'] ) );
239 $join['revision'] = array( 'INNER JOIN', $revJoin );
241 if ( !empty( $this->history
['limit'] ) ) {
242 $opts['LIMIT'] = intval( $this->history
['limit'] );
244 } elseif ( $this->history
& WikiExporter
::FULL
) {
245 # Full history dumps...
246 $join['revision'] = array( 'INNER JOIN', 'page_id=rev_page' );
247 } elseif ( $this->history
& WikiExporter
::CURRENT
) {
248 # Latest revision dumps...
249 if ( $this->list_authors
&& $cond != '' ) { // List authors, if so desired
250 $this->do_list_authors( $cond );
252 $join['revision'] = array( 'INNER JOIN', 'page_id=rev_page AND page_latest=rev_id' );
253 } elseif ( $this->history
& WikiExporter
::STABLE
) {
254 # "Stable" revision dumps...
255 # Default JOIN, to be overridden...
256 $join['revision'] = array( 'INNER JOIN', 'page_id=rev_page AND page_latest=rev_id' );
257 # One, and only one hook should set this, and return false
258 if ( wfRunHooks( 'WikiExporter::dumpStableQuery', array( &$tables, &$opts, &$join ) ) ) {
259 wfProfileOut( __METHOD__
);
260 throw new MWException( __METHOD__
. " given invalid history dump type." );
263 # Uknown history specification parameter?
264 wfProfileOut( __METHOD__
);
265 throw new MWException( __METHOD__
. " given invalid history dump type." );
267 # Query optimization hacks
269 $opts[] = 'STRAIGHT_JOIN';
270 $opts['USE INDEX']['page'] = 'PRIMARY';
272 # Build text join options
273 if ( $this->text
!= WikiExporter
::STUB
) { // 1-pass
275 $join['text'] = array( 'INNER JOIN', 'rev_text_id=old_id' );
278 if ( $this->buffer
== WikiExporter
::STREAM
) {
279 $prev = $this->db
->bufferResults( false );
282 wfRunHooks( 'ModifyExportQuery',
283 array( $this->db
, &$tables, &$cond, &$opts, &$join ) );
286 $result = $this->db
->select( $tables, '*', $cond, __METHOD__
, $opts, $join );
287 $wrapper = $this->db
->resultObject( $result );
288 # Output dump results
289 $this->outputPageStream( $wrapper );
290 if ( $this->list_authors
) {
291 $this->outputPageStream( $wrapper );
294 if ( $this->buffer
== WikiExporter
::STREAM
) {
295 $this->db
->bufferResults( $prev );
298 wfProfileOut( __METHOD__
);
302 * Runs through a query result set dumping page and revision records.
303 * The result set should be sorted/grouped by page to avoid duplicate
304 * page records in the output.
306 * The result set will be freed once complete. Should be safe for
307 * streaming (non-buffered) queries, as long as it was made on a
308 * separate database connection not managed by LoadBalancer; some
309 * blob storage types will make queries to pull source data.
311 * @param $resultset ResultWrapper
313 protected function outputPageStream( $resultset ) {
315 foreach ( $resultset as $row ) {
316 if ( is_null( $last ) ||
317 $last->page_namespace
!= $row->page_namespace ||
318 $last->page_title
!= $row->page_title
) {
319 if ( isset( $last ) ) {
321 if ( $this->dumpUploads
) {
322 $output .= $this->writer
->writeUploads( $last, $this->dumpUploadFileContents
);
324 $output .= $this->writer
->closePage();
325 $this->sink
->writeClosePage( $output );
327 $output = $this->writer
->openPage( $row );
328 $this->sink
->writeOpenPage( $row, $output );
331 $output = $this->writer
->writeRevision( $row );
332 $this->sink
->writeRevision( $row, $output );
334 if ( isset( $last ) ) {
336 if ( $this->dumpUploads
) {
337 $output .= $this->writer
->writeUploads( $last, $this->dumpUploadFileContents
);
339 $output .= $this->author_list
;
340 $output .= $this->writer
->closePage();
341 $this->sink
->writeClosePage( $output );
345 protected function outputLogStream( $resultset ) {
346 foreach ( $resultset as $row ) {
347 $output = $this->writer
->writeLogItem( $row );
348 $this->sink
->writeLogItem( $row, $output );
356 class XmlDumpWriter
{
357 var $firstPageWritten = 0;
358 var $lastPageWritten = 0;
359 var $pageInProgress = 0;
362 * Returns the export schema version.
365 function schemaVersion() {
370 * Opens the XML output stream's root <mediawiki> element.
371 * This does not include an xml directive, so is safe to include
372 * as a subelement in a larger XML stream. Namespace and XML Schema
373 * references are included.
375 * Output will be encoded in UTF-8.
379 function openStream() {
380 global $wgLanguageCode;
381 $ver = $this->schemaVersion();
382 return Xml
::element( 'mediawiki', array(
383 'xmlns' => "http://www.mediawiki.org/xml/export-$ver/",
384 'xmlns:xsi' => "http://www.w3.org/2001/XMLSchema-instance",
385 'xsi:schemaLocation' => "http://www.mediawiki.org/xml/export-$ver/ " .
386 "http://www.mediawiki.org/xml/export-$ver.xsd",
388 'xml:lang' => $wgLanguageCode ),
394 function siteInfo() {
399 $this->caseSetting(),
400 $this->namespaces() );
401 return " <siteinfo>\n " .
402 implode( "\n ", $info ) .
406 function sitename() {
408 return Xml
::element( 'sitename', array(), $wgSitename );
411 function generator() {
413 return Xml
::element( 'generator', array(), "MediaWiki $wgVersion" );
416 function homelink() {
417 return Xml
::element( 'base', array(), Title
::newMainPage()->getCanonicalUrl() );
420 function caseSetting() {
421 global $wgCapitalLinks;
422 // "case-insensitive" option is reserved for future
423 $sensitivity = $wgCapitalLinks ?
'first-letter' : 'case-sensitive';
424 return Xml
::element( 'case', array(), $sensitivity );
427 function namespaces() {
429 $spaces = "<namespaces>\n";
430 foreach ( $wgContLang->getFormattedNamespaces() as $ns => $title ) {
432 Xml
::element( 'namespace',
434 'case' => MWNamespace
::isCapitalized( $ns ) ?
'first-letter' : 'case-sensitive',
437 $spaces .= " </namespaces>";
442 * Closes the output stream with the closing root element.
443 * Call when finished dumping things.
447 function closeStream() {
448 return "</mediawiki>\n";
452 * Opens a <page> section on the output stream, with data
453 * from the given database row.
459 function openPage( $row ) {
461 $title = Title
::makeTitle( $row->page_namespace
, $row->page_title
);
462 $out .= ' ' . Xml
::elementClean( 'title', array(), $title->getPrefixedText() ) . "\n";
463 $out .= ' ' . Xml
::element( 'id', array(), strval( $row->page_id
) ) . "\n";
464 $this->pageInProgress
= $row->page_id
;
465 if ( $row->page_is_redirect
) {
466 $out .= ' ' . Xml
::element( 'redirect', array() ) . "\n";
468 if ( $row->page_restrictions
!= '' ) {
469 $out .= ' ' . Xml
::element( 'restrictions', array(),
470 strval( $row->page_restrictions
) ) . "\n";
473 wfRunHooks( 'XmlDumpWriterOpenPage', array( $this, &$out, $row, $title ) );
479 * Closes a <page> section on the output stream.
483 function closePage() {
485 if (! $this->firstPageWritten
) {
486 $this->firstPageWritten
= $this->pageInProgress
;
488 $this->lastPageWritten
= $this->pageInProgress
;
492 * Dumps a <revision> section on the output stream, with
493 * data filled in from the given database row.
499 function writeRevision( $row ) {
500 wfProfileIn( __METHOD__
);
502 $out = " <revision>\n";
503 $out .= " " . Xml
::element( 'id', null, strval( $row->rev_id
) ) . "\n";
505 $out .= $this->writeTimestamp( $row->rev_timestamp
);
507 if ( $row->rev_deleted
& Revision
::DELETED_USER
) {
508 $out .= " " . Xml
::element( 'contributor', array( 'deleted' => 'deleted' ) ) . "\n";
510 $out .= $this->writeContributor( $row->rev_user
, $row->rev_user_text
);
513 if ( $row->rev_minor_edit
) {
514 $out .= " <minor/>\n";
516 if ( $row->rev_deleted
& Revision
::DELETED_COMMENT
) {
517 $out .= " " . Xml
::element( 'comment', array( 'deleted' => 'deleted' ) ) . "\n";
518 } elseif ( $row->rev_comment
!= '' ) {
519 $out .= " " . Xml
::elementClean( 'comment', null, strval( $row->rev_comment
) ) . "\n";
523 if ( $row->rev_deleted
& Revision
::DELETED_TEXT
) {
524 $out .= " " . Xml
::element( 'text', array( 'deleted' => 'deleted' ) ) . "\n";
525 } elseif ( isset( $row->old_text
) ) {
526 // Raw text from the database may have invalid chars
527 $text = strval( Revision
::getRevisionText( $row ) );
528 $out .= " " . Xml
::elementClean( 'text',
529 array( 'xml:space' => 'preserve', 'bytes' => $row->rev_len
),
530 strval( $text ) ) . "\n";
533 $out .= " " . Xml
::element( 'text',
534 array( 'id' => $row->rev_text_id
, 'bytes' => $row->rev_len
),
538 wfRunHooks( 'XmlDumpWriterWriteRevision', array( &$this, &$out, $row, $text ) );
540 $out .= " </revision>\n";
542 wfProfileOut( __METHOD__
);
547 * Dumps a <logitem> section on the output stream, with
548 * data filled in from the given database row.
554 function writeLogItem( $row ) {
555 wfProfileIn( __METHOD__
);
557 $out = " <logitem>\n";
558 $out .= " " . Xml
::element( 'id', null, strval( $row->log_id
) ) . "\n";
560 $out .= $this->writeTimestamp( $row->log_timestamp
);
562 if ( $row->log_deleted
& LogPage
::DELETED_USER
) {
563 $out .= " " . Xml
::element( 'contributor', array( 'deleted' => 'deleted' ) ) . "\n";
565 $out .= $this->writeContributor( $row->log_user
, $row->user_name
);
568 if ( $row->log_deleted
& LogPage
::DELETED_COMMENT
) {
569 $out .= " " . Xml
::element( 'comment', array( 'deleted' => 'deleted' ) ) . "\n";
570 } elseif ( $row->log_comment
!= '' ) {
571 $out .= " " . Xml
::elementClean( 'comment', null, strval( $row->log_comment
) ) . "\n";
574 $out .= " " . Xml
::element( 'type', null, strval( $row->log_type
) ) . "\n";
575 $out .= " " . Xml
::element( 'action', null, strval( $row->log_action
) ) . "\n";
577 if ( $row->log_deleted
& LogPage
::DELETED_ACTION
) {
578 $out .= " " . Xml
::element( 'text', array( 'deleted' => 'deleted' ) ) . "\n";
580 $title = Title
::makeTitle( $row->log_namespace
, $row->log_title
);
581 $out .= " " . Xml
::elementClean( 'logtitle', null, $title->getPrefixedText() ) . "\n";
582 $out .= " " . Xml
::elementClean( 'params',
583 array( 'xml:space' => 'preserve' ),
584 strval( $row->log_params
) ) . "\n";
587 $out .= " </logitem>\n";
589 wfProfileOut( __METHOD__
);
593 function writeTimestamp( $timestamp ) {
594 $ts = wfTimestamp( TS_ISO_8601
, $timestamp );
595 return " " . Xml
::element( 'timestamp', null, $ts ) . "\n";
598 function writeContributor( $id, $text ) {
599 $out = " <contributor>\n";
601 $out .= " " . Xml
::elementClean( 'username', null, strval( $text ) ) . "\n";
602 $out .= " " . Xml
::element( 'id', null, strval( $id ) ) . "\n";
604 $out .= " " . Xml
::elementClean( 'ip', null, strval( $text ) ) . "\n";
606 $out .= " </contributor>\n";
611 * Warning! This data is potentially inconsistent. :(
613 function writeUploads( $row, $dumpContents = false ) {
614 if ( $row->page_namespace
== NS_IMAGE
) {
615 $img = wfLocalFile( $row->page_title
);
616 if ( $img && $img->exists() ) {
618 foreach ( array_reverse( $img->getHistory() ) as $ver ) {
619 $out .= $this->writeUpload( $ver, $dumpContents );
621 $out .= $this->writeUpload( $img, $dumpContents );
630 * @param $dumpContents bool
633 function writeUpload( $file, $dumpContents = false ) {
634 if ( $file->isOld() ) {
636 Xml
::element( 'archivename', null, $file->getArchiveName() ) . "\n";
640 if ( $dumpContents ) {
641 # Dump file as base64
642 # Uses only XML-safe characters, so does not need escaping
643 $contents = ' <contents encoding="base64">' .
644 chunk_split( base64_encode( file_get_contents( $file->getPath() ) ) ) .
649 return " <upload>\n" .
650 $this->writeTimestamp( $file->getTimestamp() ) .
651 $this->writeContributor( $file->getUser( 'id' ), $file->getUser( 'text' ) ) .
652 " " . Xml
::elementClean( 'comment', null, $file->getDescription() ) . "\n" .
653 " " . Xml
::element( 'filename', null, $file->getName() ) . "\n" .
655 " " . Xml
::element( 'src', null, $file->getCanonicalUrl() ) . "\n" .
656 " " . Xml
::element( 'size', null, $file->getSize() ) . "\n" .
657 " " . Xml
::element( 'sha1base36', null, $file->getSha1() ) . "\n" .
658 " " . Xml
::element( 'rel', null, $file->getRel() ) . "\n" .
667 * Base class for output stream; prints to stdout or buffer or whereever.
671 function writeOpenStream( $string ) {
672 $this->write( $string );
675 function writeCloseStream( $string ) {
676 $this->write( $string );
679 function writeOpenPage( $page, $string ) {
680 $this->write( $string );
683 function writeClosePage( $string ) {
684 $this->write( $string );
687 function writeRevision( $rev, $string ) {
688 $this->write( $string );
691 function writeLogItem( $rev, $string ) {
692 $this->write( $string );
696 * Override to write to a different stream type.
699 function write( $string ) {
703 function closeRenameAndReopen( $newname ) {
707 function rename( $newname ) {
711 function getFilename() {
717 * Stream outputter to send data to a file.
720 class DumpFileOutput
extends DumpOutput
{
724 function __construct( $file ) {
725 $this->handle
= fopen( $file, "wt" );
726 $this->filename
= $file;
729 function write( $string ) {
730 fputs( $this->handle
, $string );
734 * Close the old file, move it to a specified name,
735 * and reopen new file with the old name. Use this
736 * for writing out a file in multiple pieces
737 * at specified checkpoints (e.g. every n hours).
739 function closeRenameAndReopen( $newname ) {
740 if ( is_array($newname) ) {
741 if (count($newname) > 1) {
742 throw new MWException("Export closeRenameAndReopen: passed multiple argumnts for rename of single file\n");
745 $newname = $newname[0];
749 fclose( $this->handle
);
750 rename( $this->filename
, $newname );
751 $this->handle
= fopen( $this->filename
, "wt" );
755 function rename( $newname ) {
756 if ( is_array($newname) ) {
757 if (count($newname) > 1) {
758 throw new MWException("Export closeRenameAndReopen: passed multiple argumnts for rename of single file\n");
761 $newname = $newname[0];
765 rename( $this->filename
, $newname );
769 function getFilename() {
770 return $this->filename
;
775 * Stream outputter to send data to a file via some filter program.
776 * Even if compression is available in a library, using a separate
777 * program can allow us to make use of a multi-processor system.
780 class DumpPipeOutput
extends DumpFileOutput
{
783 function __construct( $command, $file = null ) {
784 if ( !is_null( $file ) ) {
785 $command .= " > " . wfEscapeShellArg( $file );
787 $this->handle
= popen( $command, "w" );
788 $this->command
= $command;
789 $this->filename
= $file;
793 * Close the old file, move it to a specified name,
794 * and reopen new file with the old name.
796 function closeRenameAndReopen( $newname ) {
797 if ( is_array($newname) ) {
798 if (count($newname) > 1) {
799 throw new MWException("Export closeRenameAndReopen: passed multiple argumnts for rename of single file\n");
802 $newname = $newname[0];
806 pclose( $this->handle
);
807 rename( $this->filename
, $newname );
808 $command = $this->command
;
809 $command .= " > " . wfEscapeShellArg( $this->filename
);
810 $this->handle
= popen( $command, "w" );
814 function rename( $newname ) {
815 if ( is_array($newname) ) {
816 if (count($newname) > 1) {
817 throw new MWException("Export closeRenameAndReopen: passed multiple argumnts for rename of single file\n");
820 $newname = $newname[0];
824 rename( $this->filename
, $newname );
830 * Sends dump output via the gzip compressor.
833 class DumpGZipOutput
extends DumpPipeOutput
{
834 function __construct( $file ) {
835 parent
::__construct( "gzip", $file );
840 * Sends dump output via the bgzip2 compressor.
843 class DumpBZip2Output
extends DumpPipeOutput
{
844 function __construct( $file ) {
845 parent
::__construct( "bzip2", $file );
850 * Sends dump output via the p7zip compressor.
853 class Dump7ZipOutput
extends DumpPipeOutput
{
856 function __construct( $file ) {
857 $command = "7za a -bd -si " . wfEscapeShellArg( $file );
858 // Suppress annoying useless crap from p7zip
859 // Unfortunately this could suppress real error messages too
860 $command .= ' >' . wfGetNull() . ' 2>&1';
861 parent
::__construct( $command );
862 $this->filename
= $file;
865 function closeRenameAndReopen( $newname ) {
866 if ( is_array($newname) ) {
867 if (count($newname) > 1) {
868 throw new MWException("Export closeRenameAndReopen: passed multiple argumnts for rename of single file\n");
871 $newname = $newname[0];
875 pclose( $this->handle
);
876 rename( $this->filename
, $newname );
877 $command = "7za a -bd -si " . wfEscapeShellArg( $file );
878 $command .= ' >' . wfGetNull() . ' 2>&1';
879 $this->handle
= popen( $command, "w" );
883 function rename( $newname ) {
884 if ( is_array($newname) ) {
885 if (count($newname) > 1) {
886 throw new MWException("Export closeRenameAndReopen: passed multiple argumnts for rename of single file\n");
889 $newname = $newname[0];
893 rename( $this->filename
, $newname );
901 * Dump output filter class.
902 * This just does output filtering and streaming; XML formatting is done
903 * higher up, so be careful in what you do.
907 function __construct( &$sink ) {
908 $this->sink
=& $sink;
911 function writeOpenStream( $string ) {
912 $this->sink
->writeOpenStream( $string );
915 function writeCloseStream( $string ) {
916 $this->sink
->writeCloseStream( $string );
919 function writeOpenPage( $page, $string ) {
920 $this->sendingThisPage
= $this->pass( $page, $string );
921 if ( $this->sendingThisPage
) {
922 $this->sink
->writeOpenPage( $page, $string );
926 function writeClosePage( $string ) {
927 if ( $this->sendingThisPage
) {
928 $this->sink
->writeClosePage( $string );
929 $this->sendingThisPage
= false;
933 function writeRevision( $rev, $string ) {
934 if ( $this->sendingThisPage
) {
935 $this->sink
->writeRevision( $rev, $string );
939 function writeLogItem( $rev, $string ) {
940 $this->sink
->writeRevision( $rev, $string );
943 function closeRenameAndReopen( $newname ) {
944 $this->sink
->closeRenameAndReopen( $newname );
947 function rename( $newname ) {
948 $this->sink
->rename( $newname );
951 function getFilename() {
952 return $this->sink
->getFilename();
956 * Override for page-based filter types.
959 function pass( $page ) {
965 * Simple dump output filter to exclude all talk pages.
968 class DumpNotalkFilter
extends DumpFilter
{
969 function pass( $page ) {
970 return !MWNamespace
::isTalk( $page->page_namespace
);
975 * Dump output filter to include or exclude pages in a given set of namespaces.
978 class DumpNamespaceFilter
extends DumpFilter
{
980 var $namespaces = array();
982 function __construct( &$sink, $param ) {
983 parent
::__construct( $sink );
986 "NS_MAIN" => NS_MAIN
,
987 "NS_TALK" => NS_TALK
,
988 "NS_USER" => NS_USER
,
989 "NS_USER_TALK" => NS_USER_TALK
,
990 "NS_PROJECT" => NS_PROJECT
,
991 "NS_PROJECT_TALK" => NS_PROJECT_TALK
,
992 "NS_FILE" => NS_FILE
,
993 "NS_FILE_TALK" => NS_FILE_TALK
,
994 "NS_IMAGE" => NS_IMAGE
, // NS_IMAGE is an alias for NS_FILE
995 "NS_IMAGE_TALK" => NS_IMAGE_TALK
,
996 "NS_MEDIAWIKI" => NS_MEDIAWIKI
,
997 "NS_MEDIAWIKI_TALK" => NS_MEDIAWIKI_TALK
,
998 "NS_TEMPLATE" => NS_TEMPLATE
,
999 "NS_TEMPLATE_TALK" => NS_TEMPLATE_TALK
,
1000 "NS_HELP" => NS_HELP
,
1001 "NS_HELP_TALK" => NS_HELP_TALK
,
1002 "NS_CATEGORY" => NS_CATEGORY
,
1003 "NS_CATEGORY_TALK" => NS_CATEGORY_TALK
);
1005 if ( $param { 0 } == '!' ) {
1006 $this->invert
= true;
1007 $param = substr( $param, 1 );
1010 foreach ( explode( ',', $param ) as $key ) {
1011 $key = trim( $key );
1012 if ( isset( $constants[$key] ) ) {
1013 $ns = $constants[$key];
1014 $this->namespaces
[$ns] = true;
1015 } elseif ( is_numeric( $key ) ) {
1016 $ns = intval( $key );
1017 $this->namespaces
[$ns] = true;
1019 throw new MWException( "Unrecognized namespace key '$key'\n" );
1024 function pass( $page ) {
1025 $match = isset( $this->namespaces
[$page->page_namespace
] );
1026 return $this->invert
xor $match;
1032 * Dump output filter to include only the last revision in each page sequence.
1035 class DumpLatestFilter
extends DumpFilter
{
1036 var $page, $pageString, $rev, $revString;
1038 function writeOpenPage( $page, $string ) {
1039 $this->page
= $page;
1040 $this->pageString
= $string;
1043 function writeClosePage( $string ) {
1045 $this->sink
->writeOpenPage( $this->page
, $this->pageString
);
1046 $this->sink
->writeRevision( $this->rev
, $this->revString
);
1047 $this->sink
->writeClosePage( $string );
1050 $this->revString
= null;
1052 $this->pageString
= null;
1055 function writeRevision( $rev, $string ) {
1056 if ( $rev->rev_id
== $this->page
->page_latest
) {
1058 $this->revString
= $string;
1064 * Base class for output stream; prints to stdout or buffer or whereever.
1067 class DumpMultiWriter
{
1068 function __construct( $sinks ) {
1069 $this->sinks
= $sinks;
1070 $this->count
= count( $sinks );
1073 function writeOpenStream( $string ) {
1074 for ( $i = 0; $i < $this->count
; $i++
) {
1075 $this->sinks
[$i]->writeOpenStream( $string );
1079 function writeCloseStream( $string ) {
1080 for ( $i = 0; $i < $this->count
; $i++
) {
1081 $this->sinks
[$i]->writeCloseStream( $string );
1085 function writeOpenPage( $page, $string ) {
1086 for ( $i = 0; $i < $this->count
; $i++
) {
1087 $this->sinks
[$i]->writeOpenPage( $page, $string );
1091 function writeClosePage( $string ) {
1092 for ( $i = 0; $i < $this->count
; $i++
) {
1093 $this->sinks
[$i]->writeClosePage( $string );
1097 function writeRevision( $rev, $string ) {
1098 for ( $i = 0; $i < $this->count
; $i++
) {
1099 $this->sinks
[$i]->writeRevision( $rev, $string );
1103 function closeRenameAndReopen( $newnames ) {
1104 for( $i = 0; $i < $this->count
; $i++
) {
1105 $this->sinks
[$i]->closeRenameAndReopen( $newnames[$i] );
1109 function rename( $newnames ) {
1110 for( $i = 0; $i < $this->count
; $i++
) {
1111 $this->sinks
[$i]->rename( $newnames[$i] );
1115 function getFilename() {
1116 $filenames = array();
1117 for( $i = 0; $i < $this->count
; $i++
) {
1118 $filenames[] = $this->sinks
[$i]->getFilename();
1125 function xmlsafe( $string ) {
1126 wfProfileIn( __FUNCTION__
);
1129 * The page may contain old data which has not been properly normalized.
1130 * Invalid UTF-8 sequences or forbidden control characters will make our
1131 * XML output invalid, so be sure to strip them out.
1133 $string = UtfNormal
::cleanUp( $string );
1135 $string = htmlspecialchars( $string );
1136 wfProfileOut( __FUNCTION__
);