includes/specials/SpecialExport.php

   1 <?php
   2 /**
   3  * Implements Special:Export
   4  *
   5  * Copyright © 2003-2008 Brion Vibber <brion@pobox.com>
   6  *
   7  * This program is free software; you can redistribute it and/or modify
   8  * it under the terms of the GNU General Public License as published by
   9  * the Free Software Foundation; either version 2 of the License, or
  10  * (at your option) any later version.
  11  *
  12  * This program is distributed in the hope that it will be useful,
  13  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  14  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  15  * GNU General Public License for more details.
  16  *
  17  * You should have received a copy of the GNU General Public License along
  18  * with this program; if not, write to the Free Software Foundation, Inc.,
  19  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  20  * http://www.gnu.org/copyleft/gpl.html
  21  *
  22  * @file
  23  * @ingroup SpecialPage
  24  */
  25
  26 /**
  27  * A special page that allows users to export pages in a XML file
  28  *
  29  * @ingroup SpecialPage
  30  */
  31 class SpecialExport extends SpecialPage {
  32
  33         private $curonly, $doExport, $pageLinkDepth, $templates;
  34         private $images;
  35
  36         public function __construct() {
  37                 parent::__construct( 'Export' );
  38         }
  39
  40         public function execute( $par ) {
  41                 global $wgOut, $wgRequest, $wgSitename, $wgExportAllowListContributors;
  42                 global $wgExportAllowHistory, $wgExportMaxHistory, $wgExportMaxLinkDepth;
  43                 global $wgExportFromNamespaces, $wgUser;
  44
  45                 $this->setHeaders();
  46                 $this->outputHeader();
  47
  48                 // Set some variables
  49                 $this->curonly = true;
  50                 $this->doExport = false;
  51                 $this->templates = $wgRequest->getCheck( 'templates' );
  52                 $this->images = $wgRequest->getCheck( 'images' ); // Doesn't do anything yet
  53                 $this->pageLinkDepth = $this->validateLinkDepth(
  54                         $wgRequest->getIntOrNull( 'pagelink-depth' )
  55                 );
  56                 $nsindex = '';
  57
  58                 if ( $wgRequest->getCheck( 'addcat' ) ) {
  59                         $page = $wgRequest->getText( 'pages' );
  60                         $catname = $wgRequest->getText( 'catname' );
  61
  62                         if ( $catname !== '' && $catname !== null && $catname !== false ) {
  63                                 $t = Title::makeTitleSafe( NS_MAIN, $catname );
  64                                 if ( $t ) {
  65                                         /**
  66                                          * @todo Fixme: this can lead to hitting memory limit for very large
  67                                          * categories. Ideally we would do the lookup synchronously
  68                                          * during the export in a single query.
  69                                          */
  70                                         $catpages = $this->getPagesFromCategory( $t );
  71                                         if ( $catpages ) {
  72                                                 $page .= "\n" . implode( "\n", $catpages );
  73                                         }
  74                                 }
  75                         }
  76                 }
  77                 else if( $wgRequest->getCheck( 'addns' ) && $wgExportFromNamespaces ) {
  78                         $page = $wgRequest->getText( 'pages' );
  79                         $nsindex = $wgRequest->getText( 'nsindex', '' );
  80
  81                         if ( strval( $nsindex ) !== ''  ) {
  82                                 /**
  83                                  * Same implementation as above, so same @todo
  84                                  */
  85                                 $nspages = $this->getPagesFromNamespace( $nsindex );
  86                                 if ( $nspages ) {
  87                                         $page .= "\n" . implode( "\n", $nspages );
  88                                 }
  89                         }
  90                 }
  91                 else if( $wgRequest->wasPosted() && $par == '' ) {
  92                         $page = $wgRequest->getText( 'pages' );
  93                         $this->curonly = $wgRequest->getCheck( 'curonly' );
  94                         $rawOffset = $wgRequest->getVal( 'offset' );
  95
  96                         if( $rawOffset ) {
  97                                 $offset = wfTimestamp( TS_MW, $rawOffset );
  98                         } else {
  99                                 $offset = null;
 100                         }
 101
 102                         $limit = $wgRequest->getInt( 'limit' );
 103                         $dir = $wgRequest->getVal( 'dir' );
 104                         $history = array(
 105                                 'dir' => 'asc',
 106                                 'offset' => false,
 107                                 'limit' => $wgExportMaxHistory,
 108                         );
 109                         $historyCheck = $wgRequest->getCheck( 'history' );
 110
 111                         if ( $this->curonly ) {
 112                                 $history = WikiExporter::CURRENT;
 113                         } elseif ( !$historyCheck ) {
 114                                 if ( $limit > 0 && ($wgExportMaxHistory == 0 || $limit < $wgExportMaxHistory ) ) {
 115                                         $history['limit'] = $limit;
 116                                 }
 117                                 if ( !is_null( $offset ) ) {
 118                                         $history['offset'] = $offset;
 119                                 }
 120                                 if ( strtolower( $dir ) == 'desc' ) {
 121                                         $history['dir'] = 'desc';
 122                                 }
 123                         }
 124
 125                         if( $page != '' ) {
 126                                 $this->doExport = true;
 127                         }
 128                 } else {
 129                         // Default to current-only for GET requests.
 130                         $page = $wgRequest->getText( 'pages', $par );
 131                         $historyCheck = $wgRequest->getCheck( 'history' );
 132
 133                         if( $historyCheck ) {
 134                                 $history = WikiExporter::FULL;
 135                         } else {
 136                                 $history = WikiExporter::CURRENT;
 137                         }
 138
 139                         if( $page != '' ) {
 140                                 $this->doExport = true;
 141                         }
 142                 }
 143
 144                 if( !$wgExportAllowHistory ) {
 145                         // Override
 146                         $history = WikiExporter::CURRENT;
 147                 }
 148
 149                 $list_authors = $wgRequest->getCheck( 'listauthors' );
 150                 if ( !$this->curonly || !$wgExportAllowListContributors ) {
 151                         $list_authors = false ;
 152                 }
 153
 154                 if ( $this->doExport ) {
 155                         $wgOut->disable();
 156
 157                         // Cancel output buffering and gzipping if set
 158                         // This should provide safer streaming for pages with history
 159                         wfResetOutputBuffers();
 160                         $wgRequest->response()->header( "Content-type: application/xml; charset=utf-8" );
 161
 162                         if( $wgRequest->getCheck( 'wpDownload' ) ) {
 163                                 // Provide a sane filename suggestion
 164                                 $filename = urlencode( $wgSitename . '-' . wfTimestampNow() . '.xml' );
 165                                 $wgRequest->response()->header( "Content-disposition: attachment;filename={$filename}" );
 166                         }
 167
 168                         $this->doExport( $page, $history, $list_authors );
 169
 170                         return;
 171                 }
 172
 173                 $wgOut->addWikiMsg( 'exporttext' );
 174
 175                 $form = Xml::openElement( 'form', array( 'method' => 'post',
 176                         'action' => $this->getTitle()->getLocalUrl( 'action=submit' ) ) );
 177                 $form .= Xml::inputLabel( wfMsg( 'export-addcattext' )    , 'catname', 'catname', 40 ) . '&#160;';
 178                 $form .= Xml::submitButton( wfMsg( 'export-addcat' ), array( 'name' => 'addcat' ) ) . '<br />';
 179
 180                 if ( $wgExportFromNamespaces ) {
 181                         $form .= Xml::namespaceSelector( $nsindex, null, 'nsindex', wfMsg( 'export-addnstext' ) ) . '&#160;';
 182                         $form .= Xml::submitButton( wfMsg( 'export-addns' ), array( 'name' => 'addns' ) ) . '<br />';
 183                 }
 184
 185                 $form .= Xml::element( 'textarea', array( 'name' => 'pages', 'cols' => 40, 'rows' => 10 ), $page, false );
 186                 $form .= '<br />';
 187
 188                 if( $wgExportAllowHistory ) {
 189                         $form .= Xml::checkLabel(
 190                                 wfMsg( 'exportcuronly' ),
 191                                 'curonly',
 192                                 'curonly',
 193                                 $wgRequest->wasPosted() ? $wgRequest->getCheck( 'curonly' ) : true
 194                         ) . '<br />';
 195                 } else {
 196                         $wgOut->addHTML( wfMsgExt( 'exportnohistory', 'parse' ) );
 197                 }
 198
 199                 $form .= Xml::checkLabel(
 200                         wfMsg( 'export-templates' ),
 201                         'templates',
 202                         'wpExportTemplates',
 203                         $wgRequest->wasPosted() ? $wgRequest->getCheck( 'templates' ) : false
 204                 ) . '<br />';
 205
 206                 if( $wgExportMaxLinkDepth || $this->userCanOverrideExportDepth() ) {
 207                         $form .= Xml::inputLabel( wfMsg( 'export-pagelinks' ), 'pagelink-depth', 'pagelink-depth', 20, 0 ) . '<br />';
 208                 }
 209                 // Enable this when we can do something useful exporting/importing image information. :)
 210                 //$form .= Xml::checkLabel( wfMsg( 'export-images' ), 'images', 'wpExportImages', false ) . '<br />';
 211                 $form .= Xml::checkLabel(
 212                         wfMsg( 'export-download' ),
 213                         'wpDownload',
 214                         'wpDownload',
 215                         $wgRequest->wasPosted() ? $wgRequest->getCheck( 'wpDownload' ) : true
 216                 ) . '<br />';
 217
 218                 $form .= Xml::submitButton( wfMsg( 'export-submit' ), $wgUser->getSkin()->tooltipAndAccessKeyAttribs( 'export' ) );
 219                 $form .= Xml::closeElement( 'form' );
 220
 221                 $wgOut->addHTML( $form );
 222         }
 223
 224         private function userCanOverrideExportDepth() {
 225                 global $wgUser;
 226                 return $wgUser->isAllowed( 'override-export-depth' );
 227         }
 228
 229         /**
 230          * Do the actual page exporting
 231          *
 232          * @param $page String: user input on what page(s) to export
 233          * @param $history Mixed: one of the WikiExporter history export constants
 234          * @param $list_authors Boolean: Whether to add distinct author list (when
 235          *                      not returning full history)
 236          */
 237         private function doExport( $page, $history, $list_authors ) {
 238                 $pageSet = array(); // Inverted index of all pages to look up
 239
 240                 // Split up and normalize input
 241                 foreach( explode( "\n", $page ) as $pageName ) {
 242                         $pageName = trim( $pageName );
 243                         $title = Title::newFromText( $pageName );
 244                         if( $title && $title->getInterwiki() == '' && $title->getText() !== '' ) {
 245                                 // Only record each page once!
 246                                 $pageSet[$title->getPrefixedText()] = true;
 247                         }
 248                 }
 249
 250                 // Set of original pages to pass on to further manipulation...
 251                 $inputPages = array_keys( $pageSet );
 252
 253                 // Look up any linked pages if asked...
 254                 if( $this->templates ) {
 255                         $pageSet = $this->getTemplates( $inputPages, $pageSet );
 256                 }
 257                 $linkDepth = $this->pageLinkDepth;
 258                 if( $linkDepth ) {
 259                         $pageSet = $this->getPageLinks( $inputPages, $pageSet, $linkDepth );
 260                 }
 261
 262                 /*
 263                  // Enable this when we can do something useful exporting/importing image information. :)
 264                  if( $this->images ) ) {
 265                  $pageSet = $this->getImages( $inputPages, $pageSet );
 266                  }
 267                  */
 268
 269                 $pages = array_keys( $pageSet );
 270
 271                 // Normalize titles to the same format and remove dupes, see bug 17374
 272                 foreach( $pages as $k => $v ) {
 273                         $pages[$k] = str_replace( " ", "_", $v );
 274                 }
 275
 276                 $pages = array_unique( $pages );
 277
 278                 /* Ok, let's get to it... */
 279                 if( $history == WikiExporter::CURRENT ) {
 280                         $lb = false;
 281                         $db = wfGetDB( DB_SLAVE );
 282                         $buffer = WikiExporter::BUFFER;
 283                 } else {
 284                         // Use an unbuffered query; histories may be very long!
 285                         $lb = wfGetLBFactory()->newMainLB();
 286                         $db = $lb->getConnection( DB_SLAVE );
 287                         $buffer = WikiExporter::STREAM;
 288
 289                         // This might take a while... :D
 290                         wfSuppressWarnings();
 291                         set_time_limit(0);
 292                         wfRestoreWarnings();
 293                 }
 294
 295                 $exporter = new WikiExporter( $db, $history, $buffer );
 296                 $exporter->list_authors = $list_authors;
 297                 $exporter->openStream();
 298
 299                 foreach( $pages as $page ) {
 300                         /*
 301                          if( $wgExportMaxHistory && !$this->curonly ) {
 302                          $title = Title::newFromText( $page );
 303                          if( $title ) {
 304                          $count = Revision::countByTitle( $db, $title );
 305                          if( $count > $wgExportMaxHistory ) {
 306                          wfDebug( __FUNCTION__ .
 307                          ": Skipped $page, $count revisions too big\n" );
 308                          continue;
 309                          }
 310                          }
 311                          }*/
 312                         #Bug 8824: Only export pages the user can read
 313                         $title = Title::newFromText( $page );
 314                         if( is_null( $title ) ) {
 315                                 continue; #TODO: perhaps output an <error> tag or something.
 316                         }
 317                         if( !$title->userCanRead() ) {
 318                                 continue; #TODO: perhaps output an <error> tag or something.
 319                         }
 320
 321                         $exporter->pageByTitle( $title );
 322                 }
 323
 324                 $exporter->closeStream();
 325
 326                 if( $lb ) {
 327                         $lb->closeAll();
 328                 }
 329         }
 330
 331         private function getPagesFromCategory( $title ) {
 332                 global $wgContLang;
 333
 334                 $name = $title->getDBkey();
 335
 336                 $dbr = wfGetDB( DB_SLAVE );
 337                 $res = $dbr->select(
 338                         array( 'page', 'categorylinks' ),
 339                         array( 'page_namespace', 'page_title' ),
 340                         array( 'cl_from=page_id', 'cl_to' => $name ),
 341                         __METHOD__,
 342                         array( 'LIMIT' => '5000' )
 343                 );
 344
 345                 $pages = array();
 346
 347                 foreach ( $res as $row ) {
 348                         $n = $row->page_title;
 349                         if ($row->page_namespace) {
 350                                 $ns = $wgContLang->getNsText( $row->page_namespace );
 351                                 $n = $ns . ':' . $n;
 352                         }
 353
 354                         $pages[] = $n;
 355                 }
 356                 return $pages;
 357         }
 358
 359         private function getPagesFromNamespace( $nsindex ) {
 360                 global $wgContLang;
 361
 362                 $dbr = wfGetDB( DB_SLAVE );
 363                 $res = $dbr->select(
 364                         'page',
 365                         array( 'page_namespace', 'page_title' ),
 366                         array( 'page_namespace' => $nsindex ),
 367                         __METHOD__,
 368                         array( 'LIMIT' => '5000' )
 369                 );
 370
 371                 $pages = array();
 372
 373                 foreach ( $res as $row ) {
 374                         $n = $row->page_title;
 375
 376                         if ( $row->page_namespace ) {
 377                                 $ns = $wgContLang->getNsText( $row->page_namespace );
 378                                 $n = $ns . ':' . $n;
 379                         }
 380
 381                         $pages[] = $n;
 382                 }
 383                 return $pages;
 384         }
 385
 386         /**
 387          * Expand a list of pages to include templates used in those pages.
 388          * @param $inputPages array, list of titles to look up
 389          * @param $pageSet array, associative array indexed by titles for output
 390          * @return array associative array index by titles
 391          */
 392         private function getTemplates( $inputPages, $pageSet ) {
 393                 return $this->getLinks( $inputPages, $pageSet,
 394                         'templatelinks',
 395                         array( 'tl_namespace AS namespace', 'tl_title AS title' ),
 396                         array( 'page_id=tl_from' )
 397                 );
 398         }
 399
 400         /**
 401          * Validate link depth setting, if available.
 402          */
 403         private function validateLinkDepth( $depth ) {
 404                 global $wgExportMaxLinkDepth;
 405
 406                 if( $depth < 0 ) {
 407                         return 0;
 408                 }
 409
 410                 if ( !$this->userCanOverrideExportDepth() ) {
 411                         if( $depth > $wgExportMaxLinkDepth ) {
 412                                 return $wgExportMaxLinkDepth;
 413                         }
 414                 }
 415
 416                 /*
 417                  * There's a HARD CODED limit of 5 levels of recursion here to prevent a
 418                  * crazy-big export from being done by someone setting the depth
 419                  * number too high. In other words, last resort safety net.
 420                  */
 421                 return intval( min( $depth, 5 ) );
 422         }
 423
 424         /** Expand a list of pages to include pages linked to from that page. */
 425         private function getPageLinks( $inputPages, $pageSet, $depth ) {
 426                 for( ; $depth > 0; --$depth ) {
 427                         $pageSet = $this->getLinks(
 428                                 $inputPages, $pageSet, 'pagelinks',
 429                                 array( 'pl_namespace AS namespace', 'pl_title AS title' ),
 430                                 array( 'page_id=pl_from' )
 431                         );
 432                         $inputPages = array_keys( $pageSet );
 433                 }
 434
 435                 return $pageSet;
 436         }
 437
 438         /**
 439          * Expand a list of pages to include images used in those pages.
 440          *
 441          * @param $inputPages array, list of titles to look up
 442          * @param $pageSet array, associative array indexed by titles for output
 443          *
 444          * @return array associative array index by titles
 445          */
 446         private function getImages( $inputPages, $pageSet ) {
 447                 return $this->getLinks(
 448                         $inputPages,
 449                         $pageSet,
 450                         'imagelinks',
 451                         array( NS_FILE . ' AS namespace', 'il_to AS title' ),
 452                         array( 'page_id=il_from' )
 453                 );
 454         }
 455
 456         /**
 457          * Expand a list of pages to include items used in those pages.
 458          */
 459         private function getLinks( $inputPages, $pageSet, $table, $fields, $join ) {
 460                 $dbr = wfGetDB( DB_SLAVE );
 461
 462                 foreach( $inputPages as $page ) {
 463                         $title = Title::newFromText( $page );
 464
 465                         if( $title ) {
 466                                 $pageSet[$title->getPrefixedText()] = true;
 467                                 /// @todo Fixme: May or may not be more efficient to batch these
 468                                 ///        by namespace when given multiple input pages.
 469                                 $result = $dbr->select(
 470                                         array( 'page', $table ),
 471                                         $fields,
 472                                         array_merge(
 473                                                 $join,
 474                                                 array(
 475                                                         'page_namespace' => $title->getNamespace(),
 476                                                         'page_title' => $title->getDBkey()
 477                                                 )
 478                                         ),
 479                                         __METHOD__
 480                                 );
 481
 482                                 foreach( $result as $row ) {
 483                                         $template = Title::makeTitle( $row->namespace, $row->title );
 484                                         $pageSet[$template->getPrefixedText()] = true;
 485                                 }
 486                         }
 487                 }
 488
 489                 return $pageSet;
 490         }
 491
 492 }