maintenance/importTextFiles.php

   1 <?php
   2 /**
   3  * Import pages from text files
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 2 of the License, or
   8  * (at your option) any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License along
  16  * with this program; if not, write to the Free Software Foundation, Inc.,
  17  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  18  * http://www.gnu.org/copyleft/gpl.html
  19  *
  20  * @file
  21  * @ingroup Maintenance
  22  */
  23
  24 use MediaWiki\Revision\SlotRecord;
  25 use MediaWiki\Title\Title;
  26 use MediaWiki\User\User;
  27
  28 require_once __DIR__ . '/Maintenance.php';
  29
  30 /**
  31  * Maintenance script which reads in text files
  32  * and imports their content to a page of the wiki.
  33  *
  34  * @ingroup Maintenance
  35  */
  36 class ImportTextFiles extends Maintenance {
  37         public function __construct() {
  38                 parent::__construct();
  39                 $this->addDescription( 'Reads in text files and imports their content to pages of the wiki' );
  40                 $this->addOption( 'user', 'Username to which edits should be attributed. ' .
  41                         'Default: "Maintenance script"', false, true, 'u' );
  42                 $this->addOption( 'summary', 'Specify edit summary for the edits', false, true, 's' );
  43                 $this->addOption( 'use-timestamp', 'Use the modification date of the text file ' .
  44                         'as the timestamp for the edit' );
  45                 $this->addOption( 'overwrite', 'Overwrite existing pages. If --use-timestamp is passed, this ' .
  46                         'will only overwrite pages if the file has been modified since the page was last modified.' );
  47                 $this->addOption( 'prefix', 'A string to place in front of the file name', false, true, 'p' );
  48                 $this->addOption( 'bot', 'Mark edits as bot edits in the recent changes list.' );
  49                 $this->addOption( 'rc', 'Place revisions in RecentChanges.' );
  50                 $this->addArg( 'files', 'Files to import' );
  51         }
  52
  53         public function execute() {
  54                 $userName = $this->getOption( 'user', false );
  55                 $summary = $this->getOption( 'summary', 'Imported from text file' );
  56                 $useTimestamp = $this->hasOption( 'use-timestamp' );
  57                 $rc = $this->hasOption( 'rc' );
  58                 $bot = $this->hasOption( 'bot' );
  59                 $overwrite = $this->hasOption( 'overwrite' );
  60                 $prefix = $this->getOption( 'prefix', '' );
  61
  62                 // Get all the arguments. A loop is required since Maintenance doesn't
  63                 // support an arbitrary number of arguments.
  64                 $files = [];
  65                 $i = 0;
  66                 while ( $arg = $this->getArg( $i++ ) ) {
  67                         if ( file_exists( $arg ) ) {
  68                                 $files[$arg] = file_get_contents( $arg );
  69                         } else {
  70                                 // use glob to support the Windows shell, which doesn't automatically
  71                                 // expand wildcards
  72                                 $found = false;
  73                                 foreach ( glob( $arg ) as $filename ) {
  74                                         $found = true;
  75                                         $files[$filename] = file_get_contents( $filename );
  76                                 }
  77                                 if ( !$found ) {
  78                                         $this->fatalError( "Fatal error: The file '$arg' does not exist!" );
  79                                 }
  80                         }
  81                 }
  82
  83                 $count = count( $files );
  84                 $this->output( "Importing $count pages...\n" );
  85
  86                 if ( $userName === false ) {
  87                         $user = User::newSystemUser( User::MAINTENANCE_SCRIPT_USER, [ 'steal' => true ] );
  88                 } else {
  89                         $user = User::newFromName( $userName );
  90                 }
  91
  92                 if ( !$user ) {
  93                         $this->fatalError( "Invalid username\n" );
  94                 }
  95                 if ( $user->isAnon() ) {
  96                         $user->addToDatabase();
  97                 }
  98
  99                 $exit = 0;
 100
 101                 $successCount = 0;
 102                 $failCount = 0;
 103                 $skipCount = 0;
 104
 105                 $revLookup = $this->getServiceContainer()->getRevisionLookup();
 106                 foreach ( $files as $file => $text ) {
 107                         $pageName = $prefix . pathinfo( $file, PATHINFO_FILENAME );
 108                         $timestamp = $useTimestamp ? wfTimestamp( TS_UNIX, filemtime( $file ) ) : wfTimestampNow();
 109
 110                         $title = Title::newFromText( $pageName );
 111                         // Have to check for # manually, since it gets interpreted as a fragment
 112                         if ( !$title || $title->hasFragment() ) {
 113                                 $this->error( "Invalid title $pageName. Skipping.\n" );
 114                                 $skipCount++;
 115                                 continue;
 116                         }
 117
 118                         $exists = $title->exists();
 119                         $oldRevID = $title->getLatestRevID();
 120                         $oldRevRecord = $oldRevID ? $revLookup->getRevisionById( $oldRevID ) : null;
 121                         $actualTitle = $title->getPrefixedText();
 122
 123                         if ( $exists ) {
 124                                 $touched = wfTimestamp( TS_UNIX, $title->getTouched() );
 125                                 if ( !$overwrite ) {
 126                                         $this->output( "Title $actualTitle already exists. Skipping.\n" );
 127                                         $skipCount++;
 128                                         continue;
 129                                 } elseif ( $useTimestamp && intval( $touched ) >= intval( $timestamp ) ) {
 130                                         $this->output( "File for title $actualTitle has not been modified since the " .
 131                                                 "destination page was touched. Skipping.\n" );
 132                                         $skipCount++;
 133                                         continue;
 134                                 }
 135                         }
 136
 137                         $content = ContentHandler::makeContent( rtrim( $text ), $title );
 138                         $rev = new WikiRevision();
 139                         $rev->setContent( SlotRecord::MAIN, $content );
 140                         $rev->setTitle( $title );
 141                         $rev->setUserObj( $user );
 142                         $rev->setComment( $summary );
 143                         $rev->setTimestamp( $timestamp );
 144
 145                         if ( $exists &&
 146                                 $overwrite &&
 147                                 $rev->getContent()->equals( $oldRevRecord->getContent( SlotRecord::MAIN ) )
 148                         ) {
 149                                 $this->output( "File for title $actualTitle contains no changes from the current " .
 150                                         "revision. Skipping.\n" );
 151                                 $skipCount++;
 152                                 continue;
 153                         }
 154
 155                         $status = $rev->importOldRevision();
 156                         $newId = $title->getLatestRevID();
 157
 158                         if ( $status ) {
 159                                 $action = $exists ? 'updated' : 'created';
 160                                 $this->output( "Successfully $action $actualTitle\n" );
 161                                 $successCount++;
 162                         } else {
 163                                 $action = $exists ? 'update' : 'create';
 164                                 $this->output( "Failed to $action $actualTitle\n" );
 165                                 $failCount++;
 166                                 $exit = 1;
 167                         }
 168
 169                         // Create the RecentChanges entry if necessary
 170                         if ( $rc && $status ) {
 171                                 if ( $exists ) {
 172                                         if ( is_object( $oldRevRecord ) ) {
 173                                                 RecentChange::notifyEdit(
 174                                                         $timestamp,
 175                                                         $title,
 176                                                         $rev->getMinor(),
 177                                                         $user,
 178                                                         $summary,
 179                                                         $oldRevID,
 180                                                         $oldRevRecord->getTimestamp(),
 181                                                         $bot,
 182                                                         '',
 183                                                         $oldRevRecord->getSize(),
 184                                                         $rev->getSize(),
 185                                                         $newId,
 186                                                         // the pages don't need to be patrolled
 187                                                         1
 188                                                 );
 189                                         }
 190                                 } else {
 191                                         RecentChange::notifyNew(
 192                                                 $timestamp,
 193                                                 $title,
 194                                                 $rev->getMinor(),
 195                                                 $user,
 196                                                 $summary,
 197                                                 $bot,
 198                                                 '',
 199                                                 $rev->getSize(),
 200                                                 $newId,
 201                                                 1
 202                                         );
 203                                 }
 204                         }
 205                 }
 206
 207                 $this->output( "Done! $successCount succeeded, $skipCount skipped.\n" );
 208                 if ( $exit ) {
 209                         $this->fatalError( "Import failed with $failCount failed pages.\n", $exit );
 210                 }
 211         }
 212 }
 213
 214 $maintClass = ImportTextFiles::class;
 215 require_once RUN_MAINTENANCE_IF_MAIN;