2 # Copyright (C) 2003, 2005, 2006 Brion Vibber <brion@pobox.com>
3 # http://www.mediawiki.org/
5 # This program is free software; you can redistribute it and/or modify
6 # it under the terms of the GNU General Public License as published by
7 # the Free Software Foundation; either version 2 of the License, or
8 # (at your option) any later version.
10 # This program is distributed in the hope that it will be useful,
11 # but WITHOUT ANY WARRANTY; without even the implied warranty of
12 # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 # GNU General Public License for more details.
15 # You should have received a copy of the GNU General Public License along
16 # with this program; if not, write to the Free Software Foundation, Inc.,
17 # 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 # http://www.gnu.org/copyleft/gpl.html
25 * @ingroup SpecialPage Dump
28 var $list_authors = false ; # Return distinct author list (when not returning full history)
29 var $author_list = "" ;
31 var $dumpUploads = false;
35 const STABLE
= 4; // extension defined
45 * If using WikiExporter::STREAM to stream a large amount of data,
46 * provide a database connection which is not managed by
47 * LoadBalancer to read from: some history blob types will
48 * make additional queries to pull source data while the
49 * main query is still running.
52 * @param $history Mixed: one of WikiExporter::FULL or WikiExporter::CURRENT,
53 * or an associative array:
54 * offset: non-inclusive offset at which to start the query
55 * limit: maximum number of rows to return
56 * dir: "asc" or "desc" timestamp order
57 * @param $buffer Int: one of WikiExporter::BUFFER or WikiExporter::STREAM
59 function __construct( &$db, $history = WikiExporter
::CURRENT
,
60 $buffer = WikiExporter
::BUFFER
, $text = WikiExporter
::TEXT
) {
62 $this->history
= $history;
63 $this->buffer
= $buffer;
64 $this->writer
= new XmlDumpWriter();
65 $this->sink
= new DumpOutput();
70 * Set the DumpOutput or DumpFilter object which will receive
71 * various row objects and XML output for filtering. Filters
72 * can be chained or used as callbacks.
76 public function setOutputSink( &$sink ) {
80 public function openStream() {
81 $output = $this->writer
->openStream();
82 $this->sink
->writeOpenStream( $output );
85 public function closeStream() {
86 $output = $this->writer
->closeStream();
87 $this->sink
->writeCloseStream( $output );
91 * Dumps a series of page and revision records for all pages
92 * in the database, either including complete history or only
93 * the most recent version.
95 public function allPages() {
96 return $this->dumpFrom( '' );
100 * Dumps a series of page and revision records for those pages
101 * in the database falling within the page_id range given.
102 * @param $start Int: inclusive lower limit (this id is included)
103 * @param $end Int: Exclusive upper limit (this id is not included)
104 * If 0, no upper limit.
106 public function pagesByRange( $start, $end ) {
107 $condition = 'page_id >= ' . intval( $start );
109 $condition .= ' AND page_id < ' . intval( $end );
111 return $this->dumpFrom( $condition );
115 * @param $title Title
117 public function pageByTitle( $title ) {
118 return $this->dumpFrom(
119 'page_namespace=' . $title->getNamespace() .
120 ' AND page_title=' . $this->db
->addQuotes( $title->getDBkey() ) );
123 public function pageByName( $name ) {
124 $title = Title
::newFromText( $name );
125 if( is_null( $title ) ) {
126 return new WikiError( "Can't export invalid title" );
128 return $this->pageByTitle( $title );
132 public function pagesByName( $names ) {
133 foreach( $names as $name ) {
134 $this->pageByName( $name );
138 public function allLogs() {
139 return $this->dumpFrom( '' );
142 public function logsByRange( $start, $end ) {
143 $condition = 'log_id >= ' . intval( $start );
145 $condition .= ' AND log_id < ' . intval( $end );
147 return $this->dumpFrom( $condition );
150 # Generates the distinct list of authors of an article
151 # Not called by default (depends on $this->list_authors)
152 # Can be set by Special:Export when not exporting whole history
153 protected function do_list_authors( $page , $revision , $cond ) {
154 $fname = "do_list_authors" ;
155 wfProfileIn( $fname );
156 $this->author_list
= "<contributors>";
158 $nothidden = '('.$this->db
->bitAnd('rev_deleted', Revision
::DELETED_USER
) . ') = 0';
160 $sql = "SELECT DISTINCT rev_user_text,rev_user FROM {$page},{$revision}
161 WHERE page_id=rev_page AND $nothidden AND " . $cond ;
162 $result = $this->db
->query( $sql, $fname );
163 $resultset = $this->db
->resultObject( $result );
164 while( $row = $resultset->fetchObject() ) {
165 $this->author_list
.= "<contributor>" .
167 htmlentities( $row->rev_user_text
) .
174 wfProfileOut( $fname );
175 $this->author_list
.= "</contributors>";
178 protected function dumpFrom( $cond = '' ) {
179 wfProfileIn( __METHOD__
);
180 # For logging dumps...
181 if( $this->history
& self
::LOGS
) {
182 if( $this->buffer
== WikiExporter
::STREAM
) {
183 $prev = $this->db
->bufferResults( false );
185 $where = array( 'user_id = log_user' );
187 $hideLogs = LogEventsList
::getExcludeClause( $this->db
);
188 if( $hideLogs ) $where[] = $hideLogs;
189 # Add on any caller specified conditions
190 if( $cond ) $where[] = $cond;
191 # Get logging table name for logging.* clause
192 $logging = $this->db
->tableName('logging');
193 $result = $this->db
->select( array('logging','user'),
194 array( "{$logging}.*", 'user_name' ), // grab the user name
197 array( 'ORDER BY' => 'log_id', 'USE INDEX' => array('logging' => 'PRIMARY') )
199 $wrapper = $this->db
->resultObject( $result );
200 $this->outputLogStream( $wrapper );
201 if( $this->buffer
== WikiExporter
::STREAM
) {
202 $this->db
->bufferResults( $prev );
206 $tables = array( 'page', 'revision' );
207 $opts = array( 'ORDER BY' => 'page_id ASC' );
208 $opts['USE INDEX'] = array();
210 # Full history dumps...
211 if( $this->history
& WikiExporter
::FULL
) {
212 $join['revision'] = array('INNER JOIN','page_id=rev_page');
213 # Latest revision dumps...
214 } elseif( $this->history
& WikiExporter
::CURRENT
) {
215 if( $this->list_authors
&& $cond != '' ) { // List authors, if so desired
216 list($page,$revision) = $this->db
->tableNamesN('page','revision');
217 $this->do_list_authors( $page, $revision, $cond );
219 $join['revision'] = array('INNER JOIN','page_id=rev_page AND page_latest=rev_id');
220 # "Stable" revision dumps...
221 } elseif( $this->history
& WikiExporter
::STABLE
) {
222 # Default JOIN, to be overridden...
223 $join['revision'] = array('INNER JOIN','page_id=rev_page AND page_latest=rev_id');
224 # One, and only one hook should set this, and return false
225 if( wfRunHooks( 'WikiExporter::dumpStableQuery', array(&$tables,&$opts,&$join) ) ) {
226 wfProfileOut( __METHOD__
);
227 return new WikiError( __METHOD__
." given invalid history dump type." );
229 # Time offset/limit for all pages/history...
230 } elseif( is_array( $this->history
) ) {
231 $revJoin = 'page_id=rev_page';
233 if( $this->history
['dir'] == 'asc' ) {
235 $opts['ORDER BY'] = 'rev_timestamp ASC';
238 $opts['ORDER BY'] = 'rev_timestamp DESC';
241 if( !empty( $this->history
['offset'] ) ) {
242 $revJoin .= " AND rev_timestamp $op " .
243 $this->db
->addQuotes( $this->db
->timestamp( $this->history
['offset'] ) );
245 $join['revision'] = array('INNER JOIN',$revJoin);
247 if( !empty( $this->history
['limit'] ) ) {
248 $opts['LIMIT'] = intval( $this->history
['limit'] );
250 # Uknown history specification parameter?
252 wfProfileOut( __METHOD__
);
253 return new WikiError( __METHOD__
." given invalid history dump type." );
255 # Query optimization hacks
257 $opts[] = 'STRAIGHT_JOIN';
258 $opts['USE INDEX']['page'] = 'PRIMARY';
260 # Build text join options
261 if( $this->text
!= WikiExporter
::STUB
) { // 1-pass
263 $join['text'] = array('INNER JOIN','rev_text_id=old_id');
266 if( $this->buffer
== WikiExporter
::STREAM
) {
267 $prev = $this->db
->bufferResults( false );
270 wfRunHooks( 'ModifyExportQuery',
271 array( $this->db
, &$tables, &$cond, &$opts, &$join ) );
274 $result = $this->db
->select( $tables, '*', $cond, __METHOD__
, $opts, $join );
275 $wrapper = $this->db
->resultObject( $result );
276 # Output dump results
277 $this->outputPageStream( $wrapper );
278 if( $this->list_authors
) {
279 $this->outputPageStream( $wrapper );
282 if( $this->buffer
== WikiExporter
::STREAM
) {
283 $this->db
->bufferResults( $prev );
286 wfProfileOut( __METHOD__
);
290 * Runs through a query result set dumping page and revision records.
291 * The result set should be sorted/grouped by page to avoid duplicate
292 * page records in the output.
294 * The result set will be freed once complete. Should be safe for
295 * streaming (non-buffered) queries, as long as it was made on a
296 * separate database connection not managed by LoadBalancer; some
297 * blob storage types will make queries to pull source data.
299 * @param $resultset ResultWrapper
301 protected function outputPageStream( $resultset ) {
303 while( $row = $resultset->fetchObject() ) {
304 if( is_null( $last ) ||
305 $last->page_namespace
!= $row->page_namespace ||
306 $last->page_title
!= $row->page_title
) {
307 if( isset( $last ) ) {
309 if( $this->dumpUploads
) {
310 $output .= $this->writer
->writeUploads( $last );
312 $output .= $this->writer
->closePage();
313 $this->sink
->writeClosePage( $output );
315 $output = $this->writer
->openPage( $row );
316 $this->sink
->writeOpenPage( $row, $output );
319 $output = $this->writer
->writeRevision( $row );
320 $this->sink
->writeRevision( $row, $output );
322 if( isset( $last ) ) {
324 if( $this->dumpUploads
) {
325 $output .= $this->writer
->writeUploads( $last );
327 $output .= $this->author_list
;
328 $output .= $this->writer
->closePage();
329 $this->sink
->writeClosePage( $output );
333 protected function outputLogStream( $resultset ) {
334 while( $row = $resultset->fetchObject() ) {
335 $output = $this->writer
->writeLogItem( $row );
336 $this->sink
->writeLogItem( $row, $output );
344 class XmlDumpWriter
{
347 * Returns the export schema version.
350 function schemaVersion() {
351 return "0.3"; // FIXME: upgrade to 0.4 when updated XSD is ready, for the revision deletion bits
355 * Opens the XML output stream's root <mediawiki> element.
356 * This does not include an xml directive, so is safe to include
357 * as a subelement in a larger XML stream. Namespace and XML Schema
358 * references are included.
360 * Output will be encoded in UTF-8.
364 function openStream() {
365 global $wgContLanguageCode;
366 $ver = $this->schemaVersion();
367 return Xml
::element( 'mediawiki', array(
368 'xmlns' => "http://www.mediawiki.org/xml/export-$ver/",
369 'xmlns:xsi' => "http://www.w3.org/2001/XMLSchema-instance",
370 'xsi:schemaLocation' => "http://www.mediawiki.org/xml/export-$ver/ " .
371 "http://www.mediawiki.org/xml/export-$ver.xsd",
373 'xml:lang' => $wgContLanguageCode ),
379 function siteInfo() {
384 $this->caseSetting(),
385 $this->namespaces() );
386 return " <siteinfo>\n " .
387 implode( "\n ", $info ) .
391 function sitename() {
393 return Xml
::element( 'sitename', array(), $wgSitename );
396 function generator() {
398 return Xml
::element( 'generator', array(), "MediaWiki $wgVersion" );
401 function homelink() {
402 return Xml
::element( 'base', array(), Title
::newMainPage()->getFullUrl() );
405 function caseSetting() {
406 global $wgCapitalLinks;
407 // "case-insensitive" option is reserved for future
408 $sensitivity = $wgCapitalLinks ?
'first-letter' : 'case-sensitive';
409 return Xml
::element( 'case', array(), $sensitivity );
412 function namespaces() {
414 $spaces = "<namespaces>\n";
415 foreach( $wgContLang->getFormattedNamespaces() as $ns => $title ) {
416 $spaces .= ' ' . Xml
::element( 'namespace', array( 'key' => $ns ), $title ) . "\n";
418 $spaces .= " </namespaces>";
423 * Closes the output stream with the closing root element.
424 * Call when finished dumping things.
426 function closeStream() {
427 return "</mediawiki>\n";
432 * Opens a <page> section on the output stream, with data
433 * from the given database row.
439 function openPage( $row ) {
441 $title = Title
::makeTitle( $row->page_namespace
, $row->page_title
);
442 $out .= ' ' . Xml
::elementClean( 'title', array(), $title->getPrefixedText() ) . "\n";
443 $out .= ' ' . Xml
::element( 'id', array(), strval( $row->page_id
) ) . "\n";
444 if( $row->page_is_redirect
) {
445 $out .= ' ' . Xml
::element( 'redirect', array() ). "\n";
447 if( '' != $row->page_restrictions
) {
448 $out .= ' ' . Xml
::element( 'restrictions', array(),
449 strval( $row->page_restrictions
) ) . "\n";
452 wfRunHooks( 'XmlDumpWriterOpenPage', array( $this, &$out, $row, $title ) );
458 * Closes a <page> section on the output stream.
462 function closePage() {
467 * Dumps a <revision> section on the output stream, with
468 * data filled in from the given database row.
474 function writeRevision( $row ) {
475 $fname = 'WikiExporter::dumpRev';
476 wfProfileIn( $fname );
478 $out = " <revision>\n";
479 $out .= " " . Xml
::element( 'id', null, strval( $row->rev_id
) ) . "\n";
481 $out .= $this->writeTimestamp( $row->rev_timestamp
);
483 if( $row->rev_deleted
& Revision
::DELETED_USER
) {
484 $out .= " " . Xml
::element( 'contributor', array( 'deleted' => 'deleted' ) ) . "\n";
486 $out .= $this->writeContributor( $row->rev_user
, $row->rev_user_text
);
489 if( $row->rev_minor_edit
) {
490 $out .= " <minor/>\n";
492 if( $row->rev_deleted
& Revision
::DELETED_COMMENT
) {
493 $out .= " " . Xml
::element( 'comment', array( 'deleted' => 'deleted' ) ) . "\n";
494 } elseif( $row->rev_comment
!= '' ) {
495 $out .= " " . Xml
::elementClean( 'comment', null, strval( $row->rev_comment
) ) . "\n";
498 if( $row->rev_deleted
& Revision
::DELETED_TEXT
) {
499 $out .= " " . Xml
::element( 'text', array( 'deleted' => 'deleted' ) ) . "\n";
500 } elseif( isset( $row->old_text
) ) {
501 // Raw text from the database may have invalid chars
502 $text = strval( Revision
::getRevisionText( $row ) );
503 $out .= " " . Xml
::elementClean( 'text',
504 array( 'xml:space' => 'preserve' ),
505 strval( $text ) ) . "\n";
508 $out .= " " . Xml
::element( 'text',
509 array( 'id' => $row->rev_text_id
),
513 wfRunHooks( 'XmlDumpWriterWriteRevision', array( &$this, &$out, $row, $text ) );
515 $out .= " </revision>\n";
517 wfProfileOut( $fname );
522 * Dumps a <logitem> section on the output stream, with
523 * data filled in from the given database row.
529 function writeLogItem( $row ) {
530 $fname = 'WikiExporter::writeLogItem';
531 wfProfileIn( $fname );
533 $out = " <logitem>\n";
534 $out .= " " . Xml
::element( 'id', null, strval( $row->log_id
) ) . "\n";
536 $out .= $this->writeTimestamp( $row->log_timestamp
);
538 if( $row->log_deleted
& LogPage
::DELETED_USER
) {
539 $out .= " " . Xml
::element( 'contributor', array( 'deleted' => 'deleted' ) ) . "\n";
541 $out .= $this->writeContributor( $row->log_user
, $row->user_name
);
544 if( $row->log_deleted
& LogPage
::DELETED_COMMENT
) {
545 $out .= " " . Xml
::element( 'comment', array( 'deleted' => 'deleted' ) ) . "\n";
546 } elseif( $row->log_comment
!= '' ) {
547 $out .= " " . Xml
::elementClean( 'comment', null, strval( $row->log_comment
) ) . "\n";
550 $out .= " " . Xml
::element( 'type', null, strval( $row->log_type
) ) . "\n";
551 $out .= " " . Xml
::element( 'action', null, strval( $row->log_action
) ) . "\n";
553 if( $row->log_deleted
& LogPage
::DELETED_ACTION
) {
554 $out .= " " . Xml
::element( 'text', array( 'deleted' => 'deleted' ) ) . "\n";
556 $title = Title
::makeTitle( $row->log_namespace
, $row->log_title
);
557 $out .= " " . Xml
::elementClean( 'logtitle', null, $title->getPrefixedText() ) . "\n";
558 $out .= " " . Xml
::elementClean( 'params',
559 array( 'xml:space' => 'preserve' ),
560 strval( $row->log_params
) ) . "\n";
563 $out .= " </logitem>\n";
565 wfProfileOut( $fname );
569 function writeTimestamp( $timestamp ) {
570 $ts = wfTimestamp( TS_ISO_8601
, $timestamp );
571 return " " . Xml
::element( 'timestamp', null, $ts ) . "\n";
574 function writeContributor( $id, $text ) {
575 $out = " <contributor>\n";
577 $out .= " " . Xml
::elementClean( 'username', null, strval( $text ) ) . "\n";
578 $out .= " " . Xml
::element( 'id', null, strval( $id ) ) . "\n";
580 $out .= " " . Xml
::elementClean( 'ip', null, strval( $text ) ) . "\n";
582 $out .= " </contributor>\n";
587 * Warning! This data is potentially inconsistent. :(
589 function writeUploads( $row ) {
590 if( $row->page_namespace
== NS_IMAGE
) {
591 $img = wfFindFile( $row->page_title
);
594 foreach( array_reverse( $img->getHistory() ) as $ver ) {
595 $out .= $this->writeUpload( $ver );
597 $out .= $this->writeUpload( $img );
604 function writeUpload( $file ) {
605 return " <upload>\n" .
606 $this->writeTimestamp( $file->getTimestamp() ) .
607 $this->writeContributor( $file->getUser( 'id' ), $file->getUser( 'text' ) ) .
608 " " . Xml
::elementClean( 'comment', null, $file->getDescription() ) . "\n" .
609 " " . Xml
::element( 'filename', null, $file->getName() ) . "\n" .
610 " " . Xml
::element( 'src', null, $file->getFullUrl() ) . "\n" .
611 " " . Xml
::element( 'size', null, $file->getSize() ) . "\n" .
619 * Base class for output stream; prints to stdout or buffer or whereever.
623 function writeOpenStream( $string ) {
624 $this->write( $string );
627 function writeCloseStream( $string ) {
628 $this->write( $string );
631 function writeOpenPage( $page, $string ) {
632 $this->write( $string );
635 function writeClosePage( $string ) {
636 $this->write( $string );
639 function writeRevision( $rev, $string ) {
640 $this->write( $string );
643 function writeLogItem( $rev, $string ) {
644 $this->write( $string );
648 * Override to write to a different stream type.
651 function write( $string ) {
657 * Stream outputter to send data to a file.
660 class DumpFileOutput
extends DumpOutput
{
663 function DumpFileOutput( $file ) {
664 $this->handle
= fopen( $file, "wt" );
667 function write( $string ) {
668 fputs( $this->handle
, $string );
673 * Stream outputter to send data to a file via some filter program.
674 * Even if compression is available in a library, using a separate
675 * program can allow us to make use of a multi-processor system.
678 class DumpPipeOutput
extends DumpFileOutput
{
679 function DumpPipeOutput( $command, $file = null ) {
680 if( !is_null( $file ) ) {
681 $command .= " > " . wfEscapeShellArg( $file );
683 $this->handle
= popen( $command, "w" );
688 * Sends dump output via the gzip compressor.
691 class DumpGZipOutput
extends DumpPipeOutput
{
692 function DumpGZipOutput( $file ) {
693 parent
::DumpPipeOutput( "gzip", $file );
698 * Sends dump output via the bgzip2 compressor.
701 class DumpBZip2Output
extends DumpPipeOutput
{
702 function DumpBZip2Output( $file ) {
703 parent
::DumpPipeOutput( "bzip2", $file );
708 * Sends dump output via the p7zip compressor.
711 class Dump7ZipOutput
extends DumpPipeOutput
{
712 function Dump7ZipOutput( $file ) {
713 $command = "7za a -bd -si " . wfEscapeShellArg( $file );
714 // Suppress annoying useless crap from p7zip
715 // Unfortunately this could suppress real error messages too
716 $command .= ' >' . wfGetNull() . ' 2>&1';
717 parent
::DumpPipeOutput( $command );
724 * Dump output filter class.
725 * This just does output filtering and streaming; XML formatting is done
726 * higher up, so be careful in what you do.
730 function DumpFilter( &$sink ) {
731 $this->sink
=& $sink;
734 function writeOpenStream( $string ) {
735 $this->sink
->writeOpenStream( $string );
738 function writeCloseStream( $string ) {
739 $this->sink
->writeCloseStream( $string );
742 function writeOpenPage( $page, $string ) {
743 $this->sendingThisPage
= $this->pass( $page, $string );
744 if( $this->sendingThisPage
) {
745 $this->sink
->writeOpenPage( $page, $string );
749 function writeClosePage( $string ) {
750 if( $this->sendingThisPage
) {
751 $this->sink
->writeClosePage( $string );
752 $this->sendingThisPage
= false;
756 function writeRevision( $rev, $string ) {
757 if( $this->sendingThisPage
) {
758 $this->sink
->writeRevision( $rev, $string );
762 function writeLogItem( $rev, $string ) {
763 $this->sink
->writeRevision( $rev, $string );
767 * Override for page-based filter types.
770 function pass( $page ) {
776 * Simple dump output filter to exclude all talk pages.
779 class DumpNotalkFilter
extends DumpFilter
{
780 function pass( $page ) {
781 return !MWNamespace
::isTalk( $page->page_namespace
);
786 * Dump output filter to include or exclude pages in a given set of namespaces.
789 class DumpNamespaceFilter
extends DumpFilter
{
791 var $namespaces = array();
793 function DumpNamespaceFilter( &$sink, $param ) {
794 parent
::DumpFilter( $sink );
797 "NS_MAIN" => NS_MAIN
,
798 "NS_TALK" => NS_TALK
,
799 "NS_USER" => NS_USER
,
800 "NS_USER_TALK" => NS_USER_TALK
,
801 "NS_PROJECT" => NS_PROJECT
,
802 "NS_PROJECT_TALK" => NS_PROJECT_TALK
,
803 "NS_FILE" => NS_FILE
,
804 "NS_FILE_TALK" => NS_FILE_TALK
,
805 "NS_IMAGE" => NS_IMAGE
, // NS_IMAGE is an alias for NS_FILE
806 "NS_IMAGE_TALK" => NS_IMAGE_TALK
,
807 "NS_MEDIAWIKI" => NS_MEDIAWIKI
,
808 "NS_MEDIAWIKI_TALK" => NS_MEDIAWIKI_TALK
,
809 "NS_TEMPLATE" => NS_TEMPLATE
,
810 "NS_TEMPLATE_TALK" => NS_TEMPLATE_TALK
,
811 "NS_HELP" => NS_HELP
,
812 "NS_HELP_TALK" => NS_HELP_TALK
,
813 "NS_CATEGORY" => NS_CATEGORY
,
814 "NS_CATEGORY_TALK" => NS_CATEGORY_TALK
);
816 if( $param{0} == '!' ) {
817 $this->invert
= true;
818 $param = substr( $param, 1 );
821 foreach( explode( ',', $param ) as $key ) {
823 if( isset( $constants[$key] ) ) {
824 $ns = $constants[$key];
825 $this->namespaces
[$ns] = true;
826 } elseif( is_numeric( $key ) ) {
827 $ns = intval( $key );
828 $this->namespaces
[$ns] = true;
830 throw new MWException( "Unrecognized namespace key '$key'\n" );
835 function pass( $page ) {
836 $match = isset( $this->namespaces
[$page->page_namespace
] );
837 return $this->invert
xor $match;
843 * Dump output filter to include only the last revision in each page sequence.
846 class DumpLatestFilter
extends DumpFilter
{
847 var $page, $pageString, $rev, $revString;
849 function writeOpenPage( $page, $string ) {
851 $this->pageString
= $string;
854 function writeClosePage( $string ) {
856 $this->sink
->writeOpenPage( $this->page
, $this->pageString
);
857 $this->sink
->writeRevision( $this->rev
, $this->revString
);
858 $this->sink
->writeClosePage( $string );
861 $this->revString
= null;
863 $this->pageString
= null;
866 function writeRevision( $rev, $string ) {
867 if( $rev->rev_id
== $this->page
->page_latest
) {
869 $this->revString
= $string;
875 * Base class for output stream; prints to stdout or buffer or whereever.
878 class DumpMultiWriter
{
879 function DumpMultiWriter( $sinks ) {
880 $this->sinks
= $sinks;
881 $this->count
= count( $sinks );
884 function writeOpenStream( $string ) {
885 for( $i = 0; $i < $this->count
; $i++
) {
886 $this->sinks
[$i]->writeOpenStream( $string );
890 function writeCloseStream( $string ) {
891 for( $i = 0; $i < $this->count
; $i++
) {
892 $this->sinks
[$i]->writeCloseStream( $string );
896 function writeOpenPage( $page, $string ) {
897 for( $i = 0; $i < $this->count
; $i++
) {
898 $this->sinks
[$i]->writeOpenPage( $page, $string );
902 function writeClosePage( $string ) {
903 for( $i = 0; $i < $this->count
; $i++
) {
904 $this->sinks
[$i]->writeClosePage( $string );
908 function writeRevision( $rev, $string ) {
909 for( $i = 0; $i < $this->count
; $i++
) {
910 $this->sinks
[$i]->writeRevision( $rev, $string );
915 function xmlsafe( $string ) {
917 wfProfileIn( $fname );
920 * The page may contain old data which has not been properly normalized.
921 * Invalid UTF-8 sequences or forbidden control characters will make our
922 * XML output invalid, so be sure to strip them out.
924 $string = UtfNormal
::cleanUp( $string );
926 $string = htmlspecialchars( $string );
927 wfProfileOut( $fname );