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