maintenance/importImages.php

   1 <?php
   2 /**
   3  * Import one or more images from the local file system into the wiki without
   4  * using the web-based interface.
   5  *
   6  * "Smart import" additions:
   7  * - aim: preserve the essential metadata (user, description) when importing media
   8  *   files from an existing wiki.
   9  * - process:
  10  *      - interface with the source wiki, don't use bare files only (see --source-wiki-url).
  11  *      - fetch metadata from source wiki for each file to import.
  12  *      - commit the fetched metadata to the destination wiki while submitting.
  13  *
  14  * This program is free software; you can redistribute it and/or modify
  15  * it under the terms of the GNU General Public License as published by
  16  * the Free Software Foundation; either version 2 of the License, or
  17  * (at your option) any later version.
  18  *
  19  * This program is distributed in the hope that it will be useful,
  20  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  21  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  22  * GNU General Public License for more details.
  23  *
  24  * You should have received a copy of the GNU General Public License along
  25  * with this program; if not, write to the Free Software Foundation, Inc.,
  26  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  27  * http://www.gnu.org/copyleft/gpl.html
  28  *
  29  * @file
  30  * @ingroup Maintenance
  31  * @author Rob Church <robchur@gmail.com>
  32  * @author Mij <mij@bitchx.it>
  33  */
  34
  35 $optionsWithArgs = array(
  36         'extensions', 'comment', 'comment-file', 'comment-ext', 'summary', 'user',
  37         'license', 'sleep', 'limit', 'from', 'source-wiki-url', 'timestamp',
  38 );
  39 require_once __DIR__ . '/commandLine.inc';
  40 require_once __DIR__ . '/importImages.inc';
  41 $processed = $added = $ignored = $skipped = $overwritten = $failed = 0;
  42
  43 echo "Import Images\n\n";
  44
  45 # Need a path
  46 if ( count( $args ) == 0 ) {
  47         showUsage();
  48 }
  49
  50 $dir = $args[0];
  51
  52 # Check Protection
  53 if ( isset( $options['protect'] ) && isset( $options['unprotect'] ) ) {
  54         die( "Cannot specify both protect and unprotect.  Only 1 is allowed.\n" );
  55 }
  56
  57 if ( isset( $options['protect'] ) && $options['protect'] == 1 ) {
  58         die( "You must specify a protection option.\n" );
  59 }
  60
  61 # Prepare the list of allowed extensions
  62 global $wgFileExtensions;
  63 $extensions = isset( $options['extensions'] )
  64         ? explode( ',', strtolower( $options['extensions'] ) )
  65         : $wgFileExtensions;
  66
  67 # Search the path provided for candidates for import
  68 $files = findFiles( $dir, $extensions, isset( $options['search-recursively'] ) );
  69
  70 # Initialise the user for this operation
  71 $user = isset( $options['user'] )
  72         ? User::newFromName( $options['user'] )
  73         : User::newSystemUser( 'Maintenance script', array( 'steal' => true ) );
  74 if ( !$user instanceof User ) {
  75         $user = User::newSystemUser( 'Maintenance script', array( 'steal' => true ) );
  76 }
  77 $wgUser = $user;
  78
  79 # Get block check. If a value is given, this specified how often the check is performed
  80 if ( isset( $options['check-userblock'] ) ) {
  81         if ( !$options['check-userblock'] ) {
  82                 $checkUserBlock = 1;
  83         } else {
  84                 $checkUserBlock = (int)$options['check-userblock'];
  85         }
  86 } else {
  87         $checkUserBlock = false;
  88 }
  89
  90 # Get --from
  91 MediaWiki\suppressWarnings();
  92 $from = $options['from'];
  93 MediaWiki\restoreWarnings();
  94
  95 # Get sleep time.
  96 MediaWiki\suppressWarnings();
  97 $sleep = $options['sleep'];
  98 MediaWiki\restoreWarnings();
  99
 100 if ( $sleep ) {
 101         $sleep = (int)$sleep;
 102 }
 103
 104 # Get limit number
 105 MediaWiki\suppressWarnings();
 106 $limit = $options['limit'];
 107 MediaWiki\restoreWarnings();
 108
 109 if ( $limit ) {
 110         $limit = (int)$limit;
 111 }
 112
 113 $timestamp = isset( $options['timestamp'] ) ? $options['timestamp'] : false;
 114
 115 # Get the upload comment. Provide a default one in case there's no comment given.
 116 $comment = 'Importing file';
 117
 118 if ( isset( $options['comment-file'] ) ) {
 119         $comment = file_get_contents( $options['comment-file'] );
 120         if ( $comment === false || $comment === null ) {
 121                 die( "failed to read comment file: {$options['comment-file']}\n" );
 122         }
 123 } elseif ( isset( $options['comment'] ) ) {
 124         $comment = $options['comment'];
 125 }
 126
 127 $commentExt = isset( $options['comment-ext'] ) ? $options['comment-ext'] : false;
 128
 129 $summary = isset( $options['summary'] ) ? $options['summary'] : '';
 130
 131 # Get the license specifier
 132 $license = isset( $options['license'] ) ? $options['license'] : '';
 133
 134 # Batch "upload" operation
 135 $count = count( $files );
 136 if ( $count > 0 ) {
 137
 138         foreach ( $files as $file ) {
 139                 $base = wfBaseName( $file );
 140
 141                 # Validate a title
 142                 $title = Title::makeTitleSafe( NS_FILE, $base );
 143                 if ( !is_object( $title ) ) {
 144                         echo "{$base} could not be imported; a valid title cannot be produced\n";
 145                         continue;
 146                 }
 147
 148                 if ( $from ) {
 149                         if ( $from == $title->getDBkey() ) {
 150                                 $from = null;
 151                         } else {
 152                                 $ignored++;
 153                                 continue;
 154                         }
 155                 }
 156
 157                 if ( $checkUserBlock && ( ( $processed % $checkUserBlock ) == 0 ) ) {
 158                         $user->clearInstanceCache( 'name' ); // reload from DB!
 159                         if ( $user->isBlocked() ) {
 160                                 echo $user->getName() . " was blocked! Aborting.\n";
 161                                 break;
 162                         }
 163                 }
 164
 165                 # Check existence
 166                 $image = wfLocalFile( $title );
 167                 if ( $image->exists() ) {
 168                         if ( isset( $options['overwrite'] ) ) {
 169                                 echo "{$base} exists, overwriting...";
 170                                 $svar = 'overwritten';
 171                         } else {
 172                                 echo "{$base} exists, skipping\n";
 173                                 $skipped++;
 174                                 continue;
 175                         }
 176                 } else {
 177                         if ( isset( $options['skip-dupes'] ) ) {
 178                                 $repo = $image->getRepo();
 179                                 # XXX: we end up calculating this again when actually uploading. that sucks.
 180                                 $sha1 = FSFile::getSha1Base36FromPath( $file );
 181
 182                                 $dupes = $repo->findBySha1( $sha1 );
 183
 184                                 if ( $dupes ) {
 185                                         echo "{$base} already exists as " . $dupes[0]->getName() . ", skipping\n";
 186                                         $skipped++;
 187                                         continue;
 188                                 }
 189                         }
 190
 191                         echo "Importing {$base}...";
 192                         $svar = 'added';
 193                 }
 194
 195                 if ( isset( $options['source-wiki-url'] ) ) {
 196                         /* find comment text directly from source wiki, through MW's API */
 197                         $real_comment = getFileCommentFromSourceWiki( $options['source-wiki-url'], $base );
 198                         if ( $real_comment === false ) {
 199                                 $commentText = $comment;
 200                         } else {
 201                                 $commentText = $real_comment;
 202                         }
 203
 204                         /* find user directly from source wiki, through MW's API */
 205                         $real_user = getFileUserFromSourceWiki( $options['source-wiki-url'], $base );
 206                         if ( $real_user === false ) {
 207                                 $wgUser = $user;
 208                         } else {
 209                                 $wgUser = User::newFromName( $real_user );
 210                                 if ( $wgUser === false ) {
 211                                         # user does not exist in target wiki
 212                                         echo "failed: user '$real_user' does not exist in target wiki.";
 213                                         continue;
 214                                 }
 215                         }
 216                 } else {
 217                         # Find comment text
 218                         $commentText = false;
 219
 220                         if ( $commentExt ) {
 221                                 $f = findAuxFile( $file, $commentExt );
 222                                 if ( !$f ) {
 223                                         echo " No comment file with extension {$commentExt} found "
 224                                                 . "for {$file}, using default comment. ";
 225                                 } else {
 226                                         $commentText = file_get_contents( $f );
 227                                         if ( !$commentText ) {
 228                                                 echo " Failed to load comment file {$f}, using default comment. ";
 229                                         }
 230                                 }
 231                         }
 232
 233                         if ( !$commentText ) {
 234                                 $commentText = $comment;
 235                         }
 236                 }
 237
 238                 # Import the file
 239                 if ( isset( $options['dry'] ) ) {
 240                         echo " publishing {$file} by '" . $wgUser->getName() . "', comment '$commentText'... ";
 241                 } else {
 242                         $props = FSFile::getPropsFromPath( $file );
 243                         $flags = 0;
 244                         $publishOptions = array();
 245                         $handler = MediaHandler::getHandler( $props['mime'] );
 246                         if ( $handler ) {
 247                                 $publishOptions['headers'] = $handler->getStreamHeaders( $props['metadata'] );
 248                         } else {
 249                                 $publishOptions['headers'] = array();
 250                         }
 251                         $archive = $image->publish( $file, $flags, $publishOptions );
 252                         if ( !$archive->isGood() ) {
 253                                 echo "failed. (" .
 254                                         $archive->getWikiText() .
 255                                         ")\n";
 256                                 $failed++;
 257                                 continue;
 258                         }
 259                 }
 260
 261                 $commentText = SpecialUpload::getInitialPageText( $commentText, $license );
 262                 if ( !isset( $options['summary'] ) ) {
 263                         $summary = $commentText;
 264                 }
 265
 266                 if ( isset( $options['dry'] ) ) {
 267                         echo "done.\n";
 268                 } elseif ( $image->recordUpload2(
 269                         $archive->value,
 270                         $summary,
 271                         $commentText,
 272                         $props,
 273                         $timestamp
 274                 ) ) {
 275                         # We're done!
 276                         echo "done.\n";
 277
 278                         $doProtect = false;
 279
 280                         global $wgRestrictionLevels;
 281
 282                         $protectLevel = isset( $options['protect'] ) ? $options['protect'] : null;
 283
 284                         if ( $protectLevel && in_array( $protectLevel, $wgRestrictionLevels ) ) {
 285                                 $doProtect = true;
 286                         }
 287                         if ( isset( $options['unprotect'] ) ) {
 288                                 $protectLevel = '';
 289                                 $doProtect = true;
 290                         }
 291
 292                         if ( $doProtect ) {
 293                                 # Protect the file
 294                                 echo "\nWaiting for slaves...\n";
 295                                 // Wait for slaves.
 296                                 sleep( 2.0 ); # Why this sleep?
 297                                 wfWaitForSlaves();
 298
 299                                 echo "\nSetting image restrictions ... ";
 300
 301                                 $cascade = false;
 302                                 $restrictions = array();
 303                                 foreach ( $title->getRestrictionTypes() as $type ) {
 304                                         $restrictions[$type] = $protectLevel;
 305                                 }
 306
 307                                 $page = WikiPage::factory( $title );
 308                                 $status = $page->doUpdateRestrictions( $restrictions, array(), $cascade, '', $user );
 309                                 echo ( $status->isOK() ? 'done' : 'failed' ) . "\n";
 310                         }
 311                 } else {
 312                         echo "failed. (at recordUpload stage)\n";
 313                         $svar = 'failed';
 314                 }
 315
 316                 $$svar++;
 317                 $processed++;
 318
 319                 if ( $limit && $processed >= $limit ) {
 320                         break;
 321                 }
 322
 323                 if ( $sleep ) {
 324                         sleep( $sleep );
 325                 }
 326         }
 327
 328         # Print out some statistics
 329         echo "\n";
 330         foreach (
 331                 array(
 332                         'count' => 'Found',
 333                         'limit' => 'Limit',
 334                         'ignored' => 'Ignored',
 335                         'added' => 'Added',
 336                         'skipped' => 'Skipped',
 337                         'overwritten' => 'Overwritten',
 338                         'failed' => 'Failed'
 339                 ) as $var => $desc
 340         ) {
 341                 if ( $$var > 0 ) {
 342                         echo "{$desc}: {$$var}\n";
 343                 }
 344         }
 345 } else {
 346         echo "No suitable files could be found for import.\n";
 347 }
 348
 349 exit( 0 );
 350
 351 function showUsage( $reason = false ) {
 352         if ( $reason ) {
 353                 echo $reason . "\n";
 354         }
 355
 356         echo <<<TEXT
 357 Imports images and other media files into the wiki
 358 USAGE: php importImages.php [options] <dir>
 359
 360 <dir> : Path to the directory containing images to be imported
 361
 362 Options:
 363 --extensions=<exts>     Comma-separated list of allowable extensions, defaults
 364                         to \$wgFileExtensions.
 365 --overwrite             Overwrite existing images with the same name (default
 366                         is to skip them).
 367 --limit=<num>           Limit the number of images to process. Ignored or
 368                         skipped images are not counted.
 369 --from=<name>           Ignore all files until the one with the given name.
 370                         Useful for resuming aborted imports. <name> should be
 371                         the file's canonical database form.
 372 --skip-dupes            Skip images that were already uploaded under a different
 373                         name (check SHA1).
 374 --search-recursively    Search recursively for files in subdirectories.
 375 --sleep=<sec>           Sleep between files. Useful mostly for debugging.
 376 --user=<username>       Set username of uploader, default 'Maintenance script'.
 377 --check-userblock       Check if the user got blocked during import.
 378 --comment=<text>        Set file description, default 'Importing file'.
 379 --comment-file=<file>   Set description to the content of <file>.
 380 --comment-ext=<ext>     Causes the description for each file to be loaded from a
 381                         file with the same name, but the extension <ext>. If a
 382                         global description is also given, it is appended.
 383 --license=<code>        Use an optional license template.
 384 --dry                   Dry run, don't import anything.
 385 --protect=<protect>     Specify the protect value (autoconfirmed,sysop).
 386 --summary=<summary>     Upload summary, description will be used if not
 387                         provided.
 388 --timestamp=<timestamp> Override upload time/date, all MediaWiki timestamp
 389                         formats are accepted.
 390 --unprotect             Unprotects all uploaded images.
 391 --source-wiki-url       If specified, take User and Comment data for each
 392                         imported file from this URL. For example,
 393                         --source-wiki-url="http://en.wikipedia.org/."
 394
 395 TEXT;
 396         exit( 1 );
 397 }