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 medias from an existing wiki
   8  * - process:
   9  *      - interface with the source wiki, don't use bare files only (see --source-wiki-url).
  10  *      - fetch metadata from source wiki for each file to import.
  11  *      - commit the fetched metadata to the destination wiki while submitting.
  12  *
  13  * This program is free software; you can redistribute it and/or modify
  14  * it under the terms of the GNU General Public License as published by
  15  * the Free Software Foundation; either version 2 of the License, or
  16  * (at your option) any later version.
  17  *
  18  * This program is distributed in the hope that it will be useful,
  19  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  20  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  21  * GNU General Public License for more details.
  22  *
  23  * You should have received a copy of the GNU General Public License along
  24  * with this program; if not, write to the Free Software Foundation, Inc.,
  25  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  26  * http://www.gnu.org/copyleft/gpl.html
  27  *
  28  * @file
  29  * @ingroup Maintenance
  30  * @author Rob Church <robchur@gmail.com>
  31  * @author Mij <mij@bitchx.it>
  32  */
  33
  34 $optionsWithArgs = array(
  35         'extensions', 'comment', 'comment-file', 'comment-ext', 'summary', 'user',
  36         'license', 'sleep', 'limit', 'from', 'source-wiki-url', 'timestamp',
  37 );
  38 require_once( __DIR__ . '/commandLine.inc' );
  39 require_once( __DIR__ . '/importImages.inc' );
  40 $processed = $added = $ignored = $skipped = $overwritten = $failed = 0;
  41
  42 echo( "Import Images\n\n" );
  43
  44 # Need a path
  45 if ( count( $args ) == 0 ) {
  46         showUsage();
  47 }
  48
  49 $dir = $args[0];
  50
  51 # Check Protection
  52 if ( isset( $options['protect'] ) && isset( $options['unprotect'] ) ) {
  53         die( "Cannot specify both protect and unprotect.  Only 1 is allowed.\n" );
  54 }
  55
  56 if ( isset( $options['protect'] ) && $options['protect'] == 1 ) {
  57         die( "You must specify a protection option.\n" );
  58 }
  59
  60 # Prepare the list of allowed extensions
  61 global $wgFileExtensions;
  62 $extensions = isset( $options['extensions'] )
  63         ? explode( ',', strtolower( $options['extensions'] ) )
  64         : $wgFileExtensions;
  65
  66 # Search the path provided for candidates for import
  67 $files = findFiles( $dir, $extensions, isset( $options['search-recursively'] ) );
  68
  69 # Initialise the user for this operation
  70 $user = isset( $options['user'] )
  71         ? User::newFromName( $options['user'] )
  72         : User::newFromName( 'Maintenance script' );
  73 if ( !$user instanceof User ) {
  74         $user = User::newFromName( 'Maintenance script' );
  75 }
  76 $wgUser = $user;
  77
  78 # Get block check. If a value is given, this specified how often the check is performed
  79 if ( isset( $options['check-userblock'] ) ) {
  80         if ( !$options['check-userblock'] ) {
  81                 $checkUserBlock = 1;
  82         } else {
  83                 $checkUserBlock = (int)$options['check-userblock'];
  84         }
  85 } else {
  86         $checkUserBlock = false;
  87 }
  88
  89 # Get --from
  90 $from = @$options['from'];
  91
  92 # Get sleep time.
  93 $sleep = @$options['sleep'];
  94 if ( $sleep ) {
  95         $sleep = (int)$sleep;
  96 }
  97
  98 # Get limit number
  99 $limit = @$options['limit'];
 100 if ( $limit ) {
 101         $limit = (int)$limit;
 102 }
 103
 104 $timestamp = isset( $options['timestamp'] ) ? $options['timestamp'] : false;
 105
 106 # Get the upload comment. Provide a default one in case there's no comment given.
 107 $comment = 'Importing image file';
 108
 109 if ( isset( $options['comment-file'] ) ) {
 110         $comment = file_get_contents( $options['comment-file'] );
 111         if ( $comment === false || $comment === null ) {
 112                 die( "failed to read comment file: {$options['comment-file']}\n" );
 113         }
 114 } elseif ( isset( $options['comment'] ) ) {
 115         $comment = $options['comment'];
 116 }
 117
 118 $commentExt = isset( $options['comment-ext'] ) ? $options['comment-ext'] : false;
 119
 120 $summary = isset( $options['summary'] ) ? $options['summary'] : '';
 121
 122 # Get the license specifier
 123 $license = isset( $options['license'] ) ? $options['license'] : '';
 124
 125 # Batch "upload" operation
 126 $count = count( $files );
 127 if ( $count > 0 ) {
 128
 129         foreach ( $files as $file ) {
 130                 $base = wfBaseName( $file );
 131
 132                 # Validate a title
 133                 $title = Title::makeTitleSafe( NS_FILE, $base );
 134                 if ( !is_object( $title ) ) {
 135                         echo( "{$base} could not be imported; a valid title cannot be produced\n" );
 136                         continue;
 137                 }
 138
 139                 if ( $from ) {
 140                         if ( $from == $title->getDBkey() ) {
 141                                 $from = null;
 142                         } else {
 143                                 $ignored++;
 144                                 continue;
 145                         }
 146                 }
 147
 148                 if ( $checkUserBlock && ( ( $processed % $checkUserBlock ) == 0 ) ) {
 149                         $user->clearInstanceCache( 'name' ); // reload from DB!
 150                         if ( $user->isBlocked() ) {
 151                                 echo( $user->getName() . " was blocked! Aborting.\n" );
 152                                 break;
 153                         }
 154                 }
 155
 156                 # Check existence
 157                 $image = wfLocalFile( $title );
 158                 if ( $image->exists() ) {
 159                         if ( isset( $options['overwrite'] ) ) {
 160                                 echo( "{$base} exists, overwriting..." );
 161                                 $svar = 'overwritten';
 162                         } else {
 163                                 echo( "{$base} exists, skipping\n" );
 164                                 $skipped++;
 165                                 continue;
 166                         }
 167                 } else {
 168                         if ( isset( $options['skip-dupes'] ) ) {
 169                                 $repo = $image->getRepo();
 170                                 $sha1 = File::sha1Base36( $file ); # XXX: we end up calculating this again when actually uploading. that sucks.
 171
 172                                 $dupes = $repo->findBySha1( $sha1 );
 173
 174                                 if ( $dupes ) {
 175                                         echo( "{$base} already exists as " . $dupes[0]->getName() . ", skipping\n" );
 176                                         $skipped++;
 177                                         continue;
 178                                 }
 179                         }
 180
 181                         echo( "Importing {$base}..." );
 182                         $svar = 'added';
 183                 }
 184
 185                 if ( isset( $options['source-wiki-url'] ) ) {
 186                         /* find comment text directly from source wiki, through MW's API */
 187                         $real_comment = getFileCommentFromSourceWiki( $options['source-wiki-url'], $base );
 188                         if ( $real_comment === false )
 189                                 $commentText = $comment;
 190                         else
 191                                 $commentText = $real_comment;
 192
 193                         /* find user directly from source wiki, through MW's API */
 194                         $real_user = getFileUserFromSourceWiki( $options['source-wiki-url'], $base );
 195                         if ( $real_user === false ) {
 196                                 $wgUser = $user;
 197                         } else {
 198                                 $wgUser = User::newFromName( $real_user );
 199                                 if ( $wgUser === false ) {
 200                                         # user does not exist in target wiki
 201                                         echo ( "failed: user '$real_user' does not exist in target wiki." );
 202                                         continue;
 203                                 }
 204                         }
 205                 } else {
 206                         # Find comment text
 207                         $commentText = false;
 208
 209                         if ( $commentExt ) {
 210                                 $f = findAuxFile( $file, $commentExt );
 211                                 if ( !$f ) {
 212                                         echo( " No comment file with extension {$commentExt} found for {$file}, using default comment. " );
 213                                 } else {
 214                                         $commentText = file_get_contents( $f );
 215                                         if ( !$commentText ) {
 216                                                 echo( " Failed to load comment file {$f}, using default comment. " );
 217                                         }
 218                                 }
 219                         }
 220
 221                         if ( !$commentText ) {
 222                                 $commentText = $comment;
 223                         }
 224                 }
 225
 226                 # Import the file
 227                 if ( isset( $options['dry'] ) ) {
 228                         echo( " publishing {$file} by '" . $wgUser->getName() . "', comment '$commentText'... " );
 229                 } else {
 230                         $archive = $image->publish( $file );
 231                         if ( !$archive->isGood() ) {
 232                                 echo( "failed. (" .
 233                                         $archive->getWikiText() .
 234                                         ")\n" );
 235                                 $failed++;
 236                                 continue;
 237                         }
 238                 }
 239
 240                 $commentText = SpecialUpload::getInitialPageText( $commentText, $license );
 241                 if ( !$summary ) {
 242                         $summary = $commentText;
 243                 }
 244
 245                 if ( isset( $options['dry'] ) ) {
 246                         echo( "done.\n" );
 247                 } elseif ( $image->recordUpload2( $archive->value, $summary, $commentText, false, $timestamp ) ) {
 248                         # We're done!
 249                         echo( "done.\n" );
 250
 251                         $doProtect = false;
 252
 253                         global $wgRestrictionLevels;
 254
 255                         $protectLevel = isset( $options['protect'] ) ? $options['protect'] : null;
 256
 257                         if ( $protectLevel && in_array( $protectLevel, $wgRestrictionLevels ) ) {
 258                                 $doProtect = true;
 259                         }
 260                         if ( isset( $options['unprotect'] ) ) {
 261                                 $protectLevel = '';
 262                                 $doProtect = true;
 263                         }
 264
 265                         if ( $doProtect ) {
 266                                         # Protect the file
 267                                         echo "\nWaiting for slaves...\n";
 268                                         // Wait for slaves.
 269                                         sleep( 2.0 ); # Why this sleep?
 270                                         wfWaitForSlaves();
 271
 272                                         echo( "\nSetting image restrictions ... " );
 273
 274                                         $cascade = false;
 275                                         $restrictions = array();
 276                                         foreach( $title->getRestrictionTypes() as $type ) {
 277                                                 $restrictions[$type] = $protectLevel;
 278                                         }
 279
 280                                         $page = WikiPage::factory( $title );
 281                                         $status = $page->doUpdateRestrictions( $restrictions, array(), $cascade, '', $user );
 282                                         echo( ( $status->isOK() ? 'done' : 'failed' ) . "\n" );
 283                         }
 284
 285                 } else {
 286                         echo( "failed. (at recordUpload stage)\n" );
 287                         $svar = 'failed';
 288                 }
 289
 290                 $$svar++;
 291                 $processed++;
 292
 293                 if ( $limit && $processed >= $limit ) {
 294                         break;
 295                 }
 296
 297                 if ( $sleep ) {
 298                         sleep( $sleep );
 299                 }
 300         }
 301
 302         # Print out some statistics
 303         echo( "\n" );
 304         foreach ( array( 'count' => 'Found', 'limit' => 'Limit', 'ignored' => 'Ignored',
 305                 'added' => 'Added', 'skipped' => 'Skipped', 'overwritten' => 'Overwritten',
 306                 'failed' => 'Failed' ) as $var => $desc ) {
 307                 if ( $$var > 0 )
 308                         echo( "{$desc}: {$$var}\n" );
 309         }
 310
 311 } else {
 312         echo( "No suitable files could be found for import.\n" );
 313 }
 314
 315 exit( 0 );
 316
 317 function showUsage( $reason = false ) {
 318         if ( $reason ) {
 319                 echo( $reason . "\n" );
 320         }
 321
 322         echo <<<TEXT
 323 Imports images and other media files into the wiki
 324 USAGE: php importImages.php [options] <dir>
 325
 326 <dir> : Path to the directory containing images to be imported
 327
 328 Options:
 329 --extensions=<exts>     Comma-separated list of allowable extensions, defaults to \$wgFileExtensions
 330 --overwrite             Overwrite existing images with the same name (default is to skip them)
 331 --limit=<num>           Limit the number of images to process. Ignored or skipped images are not counted.
 332 --from=<name>           Ignore all files until the one with the given name. Useful for resuming
 333                         aborted imports. <name> should be the file's canonical database form.
 334 --skip-dupes            Skip images that were already uploaded under a different name (check SHA1)
 335 --search-recursively    Search recursively for files in subdirectories
 336 --sleep=<sec>           Sleep between files. Useful mostly for debugging.
 337 --user=<username>       Set username of uploader, default 'Maintenance script'
 338 --check-userblock       Check if the user got blocked during import.
 339 --comment=<text>        Set file description, default 'Importing image file'.
 340 --comment-file=<file>   Set description to the content of <file>.
 341 --comment-ext=<ext>     Causes the description for each file to be loaded from a file with the same name
 342                         but the extension <ext>. If a global description is also given, it is appended.
 343 --license=<code>        Use an optional license template
 344 --dry                   Dry run, don't import anything
 345 --protect=<protect>     Specify the protect value (autoconfirmed,sysop)
 346 --summary=<summary>     Upload summary, description will be used if not provided
 347 --timestamp=<timestamp> Override upload time/date, all MediaWiki timestamp formats are accepted
 348 --unprotect             Unprotects all uploaded images
 349 --source-wiki-url       If specified, take User and Comment data for each imported file from this URL.
 350                         For example, --source-wiki-url="http://en.wikipedia.org/"
 351
 352 TEXT;
 353         exit( 1 );
 354 }