maintenance/importImages.php

   1 <?php
   2
   3 /**
   4  * Maintenance script to import one or more images from the local file system into
   5  * the wiki without using the web-based interface.
   6  *
   7  * "Smart import" additions:
   8  * - aim: preserve the essential metadata (user, description) when importing medias 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( 'extensions', 'comment', 'comment-file', 'comment-ext', 'user', 'license', 'sleep', 'limit', 'from', 'source-wiki-url' );
  36 require_once( dirname( __FILE__ ) . '/commandLine.inc' );
  37 require_once( dirname( __FILE__ ) . '/importImages.inc' );
  38 $processed = $added = $ignored = $skipped = $overwritten = $failed = 0;
  39
  40 echo( "Import Images\n\n" );
  41
  42 # Need a path
  43 if ( count( $args ) > 0 ) {
  44
  45         $dir = $args[0];
  46
  47         # Check Protection
  48         if ( isset( $options['protect'] ) && isset( $options['unprotect'] ) )
  49                         die( "Cannot specify both protect and unprotect.  Only 1 is allowed.\n" );
  50
  51         if ( isset( $options['protect'] ) && $options['protect'] == 1 )
  52                         die( "You must specify a protection option.\n" );
  53
  54         # Prepare the list of allowed extensions
  55         global $wgFileExtensions;
  56         $extensions = isset( $options['extensions'] )
  57                 ? explode( ',', strtolower( $options['extensions'] ) )
  58                 : $wgFileExtensions;
  59
  60         # Search the path provided for candidates for import
  61         $files = findFiles( $dir, $extensions );
  62
  63         # Initialise the user for this operation
  64         $user = isset( $options['user'] )
  65                 ? User::newFromName( $options['user'] )
  66                 : User::newFromName( 'Maintenance script' );
  67         if ( !$user instanceof User )
  68                 $user = User::newFromName( 'Maintenance script' );
  69         $wgUser = $user;
  70
  71         # Get block check. If a value is given, this specified how often the check is performed
  72         if ( isset( $options['check-userblock'] ) ) {
  73                 if ( !$options['check-userblock'] ) $checkUserBlock = 1;
  74                 else $checkUserBlock = (int)$options['check-userblock'];
  75         } else {
  76                 $checkUserBlock = false;
  77         }
  78
  79         # Get --from
  80         $from = @$options['from'];
  81
  82         # Get sleep time.
  83         $sleep = @$options['sleep'];
  84         if ( $sleep ) $sleep = (int)$sleep;
  85
  86         # Get limit number
  87         $limit = @$options['limit'];
  88         if ( $limit ) $limit = (int)$limit;
  89
  90         # Get the upload comment. Provide a default one in case there's no comment given.
  91         $comment = 'Importing image file';
  92
  93         if ( isset( $options['comment-file'] ) ) {
  94                 $comment =  file_get_contents( $options['comment-file'] );
  95                 if ( $comment === false || $comment === NULL ) {
  96                         die( "failed to read comment file: {$options['comment-file']}\n" );
  97                 }
  98         }
  99         else if ( isset( $options['comment'] ) ) {
 100                 $comment =  $options['comment'];
 101         }
 102
 103         $commentExt = isset( $options['comment-ext'] ) ? $options['comment-ext'] : false;
 104
 105         # Get the license specifier
 106         $license = isset( $options['license'] ) ? $options['license'] : '';
 107
 108         # Batch "upload" operation
 109         if ( ( $count = count( $files ) ) > 0 ) {
 110
 111                 foreach ( $files as $file ) {
 112                         $base = wfBaseName( $file );
 113
 114                         # Validate a title
 115                         $title = Title::makeTitleSafe( NS_FILE, $base );
 116                         if ( !is_object( $title ) ) {
 117                                 echo( "{$base} could not be imported; a valid title cannot be produced\n" );
 118                                 continue;
 119                         }
 120
 121                         if ( $from ) {
 122                                 if ( $from == $title->getDBkey() ) {
 123                                         $from = NULL;
 124                                 } else {
 125                                         $ignored++;
 126                                         continue;
 127                                 }
 128                         }
 129
 130                         if ( $checkUserBlock && ( ( $processed % $checkUserBlock ) == 0 ) ) {
 131                                 $user->clearInstanceCache( 'name' ); // reload from DB!
 132                                 if ( $user->isBlocked() ) {
 133                                         echo( $user->getName() . " was blocked! Aborting.\n" );
 134                                         break;
 135                                 }
 136                         }
 137
 138                         # Check existence
 139                         $image = wfLocalFile( $title );
 140                         if ( $image->exists() ) {
 141                                 if ( isset( $options['overwrite'] ) ) {
 142                                         echo( "{$base} exists, overwriting..." );
 143                                         $svar = 'overwritten';
 144                                 } else {
 145                                         echo( "{$base} exists, skipping\n" );
 146                                         $skipped++;
 147                                         continue;
 148                                 }
 149                         } else {
 150                                 if ( isset( $options['skip-dupes'] ) ) {
 151                                         $repo = $image->getRepo();
 152                                         $sha1 = File::sha1Base36( $file ); # XXX: we end up calculating this again when actually uploading. that sucks.
 153
 154                                         $dupes = $repo->findBySha1( $sha1 );
 155
 156                                         if ( $dupes ) {
 157                                                 echo( "{$base} already exists as " . $dupes[0]->getName() . ", skipping\n" );
 158                                                 $skipped++;
 159                                                 continue;
 160                                         }
 161                                 }
 162
 163                                 echo( "Importing {$base}..." );
 164                                 $svar = 'added';
 165                         }
 166
 167                         if ( isset( $options['source-wiki-url'] ) ) {
 168                                 /* find comment text directly from source wiki, through MW's API */
 169                                 $real_comment = getFileCommentFromSourceWiki( $options['source-wiki-url'], $base );
 170                                 if ( $real_comment === false )
 171                                         $commentText = $comment;
 172                                 else
 173                                         $commentText = $real_comment;
 174
 175                                 /* find user directly from source wiki, through MW's API */
 176                                 $real_user = getFileUserFromSourceWiki( $options['source-wiki-url'], $base );
 177                                 if ( $real_user === false ) {
 178                                         $wgUser = $user;
 179                                 } else {
 180                                         $wgUser = User::newFromName( $real_user );
 181                                         if ( $wgUser === false ) {
 182                                                 # user does not exist in target wiki
 183                                                 echo ( "failed: user '$real_user' does not exist in target wiki." );
 184                                                 continue;
 185                                         }
 186                                 }
 187                         } else {
 188                                 # Find comment text
 189                                 $commentText = false;
 190
 191                                 if ( $commentExt ) {
 192                                         $f = findAuxFile( $file, $commentExt );
 193                                         if ( !$f ) {
 194                                                 echo( " No comment file with extension {$commentExt} found for {$file}, using default comment. " );
 195                                         } else {
 196                                                 $commentText = file_get_contents( $f );
 197                                                 if ( !$f ) {
 198                                                         echo( " Failed to load comment file {$f}, using default comment. " );
 199                                                 }
 200                                         }
 201                                 }
 202
 203                                 if ( !$commentText ) {
 204                                         $commentText = $comment;
 205                                 }
 206                         }
 207
 208
 209                         # Import the file
 210                         if ( isset( $options['dry'] ) ) {
 211                                 echo( " publishing {$file} by '" . $wgUser->getName() . "', comment '$commentText'... " );
 212                         } else {
 213                                 $archive = $image->publish( $file );
 214                                 if ( !$archive->isGood() ) {
 215                                         echo( "failed. (" .
 216                                                 $archive->getWikiText() .
 217                                                 ")\n" );
 218                                         $failed++;
 219                                         continue;
 220                                 }
 221                         }
 222
 223                         $doProtect = false;
 224                         $restrictions = array();
 225
 226                         global $wgRestrictionLevels;
 227
 228                         $protectLevel = isset( $options['protect'] ) ? $options['protect'] : null;
 229
 230                         if ( $protectLevel && in_array( $protectLevel, $wgRestrictionLevels ) ) {
 231                                         $restrictions['move'] = $protectLevel;
 232                                         $restrictions['edit'] = $protectLevel;
 233                                         $doProtect = true;
 234                         }
 235                         if ( isset( $options['unprotect'] ) ) {
 236                                         $restrictions['move'] = '';
 237                                         $restrictions['edit'] = '';
 238                                         $doProtect = true;
 239                         }
 240
 241
 242                         if ( isset( $options['dry'] ) ) {
 243                                 echo( "done.\n" );
 244                         } else if ( $image->recordUpload( $archive->value, $commentText, $license ) ) {
 245                                 # We're done!
 246                                 echo( "done.\n" );
 247                                 if ( $doProtect ) {
 248                                                 # Protect the file
 249                                                 $article = new Article( $title );
 250                                                 echo "\nWaiting for slaves...\n";
 251                                                 // Wait for slaves.
 252                                                 sleep( 2.0 ); # Why this sleep?
 253                                                 wfWaitForSlaves();
 254
 255                                                 echo( "\nSetting image restrictions ... " );
 256                                                 if ( $article->updateRestrictions( $restrictions ) )
 257                                                         echo( "done.\n" );
 258                                                 else
 259                                                         echo( "failed.\n" );
 260                                 }
 261
 262                         } else {
 263                                 echo( "failed. (at recordUpload stage)\n" );
 264                                 $svar = 'failed';
 265                         }
 266
 267                         $$svar++;
 268                         $processed++;
 269
 270                         if ( $limit && $processed >= $limit )
 271                                 break;
 272
 273                         if ( $sleep )
 274                                 sleep( $sleep );
 275                 }
 276
 277                 # Print out some statistics
 278                 echo( "\n" );
 279                 foreach ( array( 'count' => 'Found', 'limit' => 'Limit', 'ignored' => 'Ignored',
 280                         'added' => 'Added', 'skipped' => 'Skipped', 'overwritten' => 'Overwritten',
 281                         'failed' => 'Failed' ) as $var => $desc ) {
 282                         if ( $$var > 0 )
 283                                 echo( "{$desc}: {$$var}\n" );
 284                 }
 285
 286         } else {
 287                 echo( "No suitable files could be found for import.\n" );
 288         }
 289
 290 } else {
 291         showUsage();
 292 }
 293
 294 exit( 0 );
 295
 296 function showUsage( $reason = false ) {
 297         if ( $reason ) {
 298                 echo( $reason . "\n" );
 299         }
 300
 301         echo <<<TEXT
 302 Imports images and other media files into the wiki
 303 USAGE: php importImages.php [options] <dir>
 304
 305 <dir> : Path to the directory containing images to be imported
 306
 307 Options:
 308 --extensions=<exts>     Comma-separated list of allowable extensions, defaults to \$wgFileExtensions
 309 --overwrite             Overwrite existing images with the same name (default is to skip them)
 310 --limit=<num>           Limit the number of images to process. Ignored or skipped images are not counted.
 311 --from=<name>           Ignore all files until the one with the given name. Useful for resuming
 312                                                 aborted imports. <name> should be the file's canonical database form.
 313 --skip-dupes            Skip images that were already uploaded under a different name (check SHA1)
 314 --sleep=<sec>           Sleep between files. Useful mostly for debugging.
 315 --user=<username>       Set username of uploader, default 'Maintenance script'
 316 --check-userblock       Check if the user got blocked during import.
 317 --comment=<text>        Set upload summary comment, default 'Importing image file'.
 318 --comment-file=<file>   Set upload summary comment the the content of <file>.
 319 --comment-ext=<ext>     Causes the comment for each file to be loaded from a file with the same name
 320                         but the extension <ext>. If a global comment is also given, it is appended.
 321 --license=<code>        Use an optional license template
 322 --dry                   Dry run, don't import anything
 323 --protect=<protect>     Specify the protect value (autoconfirmed,sysop)
 324 --unprotect             Unprotects all uploaded images
 325 --source-wiki-url   if specified, take User and Comment data for each imported file from this URL.
 326                                         For example, --source-wiki-url="http://en.wikipedia.org/"
 327
 328 TEXT;
 329         exit( 1 );
 330 }