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 require_once __DIR__ . '/Maintenance.php';
  25
  26 /**
  27  * Maintenance script which reads in text files
  28  * and imports their content to a page of the wiki.
  29  *
  30  * @ingroup Maintenance
  31  */
  32 class ImportTextFiles extends Maintenance {
  33         public function __construct() {
  34                 parent::__construct();
  35                 $this->addDescription( 'Reads in text files and imports their content to pages of the wiki' );
  36                 $this->addOption( 'user', 'Username to which edits should be attributed. ' .
  37                         'Default: "Maintenance script"', false, true, 'u' );
  38                 $this->addOption( 'summary', 'Specify edit summary for the edits', false, true, 's' );
  39                 $this->addOption( 'use-timestamp', 'Use the modification date of the text file ' .
  40                         'as the timestamp for the edit' );
  41                 $this->addOption( 'overwrite', 'Overwrite existing pages. If --use-timestamp is passed, this ' .
  42                         'will only overwrite pages if the file has been modified since the page was last modified.' );
  43                 $this->addOption( 'prefix', 'A string to place in front of the file name', false, true, 'p' );
  44                 $this->addOption( 'bot', 'Mark edits as bot edits in the recent changes list.' );
  45                 $this->addOption( 'rc', 'Place revisions in RecentChanges.' );
  46                 $this->addArg( 'files', 'Files to import' );
  47         }
  48
  49         public function execute() {
  50                 $userName = $this->getOption( 'user', false );
  51                 $summary = $this->getOption( 'summary', 'Imported from text file' );
  52                 $useTimestamp = $this->hasOption( 'use-timestamp' );
  53                 $rc = $this->hasOption( 'rc' );
  54                 $bot = $this->hasOption( 'bot' );
  55                 $overwrite = $this->hasOption( 'overwrite' );
  56                 $prefix = $this->getOption( 'prefix', '' );
  57
  58                 // Get all the arguments. A loop is required since Maintenance doesn't
  59                 // support an arbitrary number of arguments.
  60                 $files = [];
  61                 $i = 0;
  62                 while ( $arg = $this->getArg( $i++ ) ) {
  63                         if ( file_exists( $arg ) ) {
  64                                 $files[$arg] = file_get_contents( $arg );
  65                         } else {
  66                                 // use glob to support the Windows shell, which doesn't automatically
  67                                 // expand wildcards
  68                                 $found = false;
  69                                 foreach ( glob( $arg ) as $filename ) {
  70                                         $found = true;
  71                                         $files[$filename] = file_get_contents( $filename );
  72                                 }
  73                                 if ( !$found ) {
  74                                         $this->error( "Fatal error: The file '$arg' does not exist!", 1 );
  75                                 }
  76                         }
  77                 };
  78
  79                 $count = count( $files );
  80                 $this->output( "Importing $count pages...\n" );
  81
  82                 if ( $userName === false ) {
  83                         $user = User::newSystemUser( 'Maintenance script', [ 'steal' => true ] );
  84                 } else {
  85                         $user = User::newFromName( $userName );
  86                 }
  87
  88                 if ( !$user ) {
  89                         $this->error( "Invalid username\n", true );
  90                 }
  91                 if ( $user->isAnon() ) {
  92                         $user->addToDatabase();
  93                 }
  94
  95                 $exit = 0;
  96
  97                 $successCount = 0;
  98                 $failCount = 0;
  99                 $skipCount = 0;
 100
 101                 foreach ( $files as $file => $text ) {
 102                         $pageName = $prefix . pathinfo( $file, PATHINFO_FILENAME );
 103                         $timestamp = $useTimestamp ? wfTimestamp( TS_UNIX, filemtime( $file ) ) : wfTimestampNow();
 104
 105                         $title = Title::newFromText( $pageName );
 106                         // Have to check for # manually, since it gets interpreted as a fragment
 107                         if ( !$title || $title->hasFragment() ) {
 108                                 $this->error( "Invalid title $pageName. Skipping.\n" );
 109                                 $skipCount++;
 110                                 continue;
 111                         }
 112
 113                         $exists = $title->exists();
 114                         $oldRevID = $title->getLatestRevID();
 115                         $oldRev = $oldRevID ? Revision::newFromId( $oldRevID ) : null;
 116                         $actualTitle = $title->getPrefixedText();
 117
 118                         if ( $exists ) {
 119                                 $touched = wfTimestamp( TS_UNIX, $title->getTouched() );
 120                                 if ( !$overwrite ) {
 121                                         $this->output( "Title $actualTitle already exists. Skipping.\n" );
 122                                         $skipCount++;
 123                                         continue;
 124                                 } elseif ( $useTimestamp && intval( $touched ) >= intval( $timestamp ) ) {
 125                                         $this->output( "File for title $actualTitle has not been modified since the " .
 126                                                 "destination page was touched. Skipping.\n" );
 127                                         $skipCount++;
 128                                         continue;
 129                                 }
 130                         }
 131
 132                         $rev = new WikiRevision( ConfigFactory::getDefaultInstance()->makeConfig( 'main' ) );
 133                         $rev->setText( rtrim( $text ) );
 134                         $rev->setTitle( $title );
 135                         $rev->setUserObj( $user );
 136                         $rev->setComment( $summary );
 137                         $rev->setTimestamp( $timestamp );
 138
 139                         if ( $exists && $overwrite && $rev->getContent()->equals( $oldRev->getContent() ) ) {
 140                                 $this->output( "File for title $actualTitle contains no changes from the current " .
 141                                         "revision. Skipping.\n" );
 142                                 $skipCount++;
 143                                 continue;
 144                         }
 145
 146                         $status = $rev->importOldRevision();
 147                         $newId = $title->getLatestRevID();
 148
 149                         if ( $status ) {
 150                                 $action = $exists ? 'updated' : 'created';
 151                                 $this->output( "Successfully $action $actualTitle\n" );
 152                                 $successCount++;
 153                         } else {
 154                                 $action = $exists ? 'update' : 'create';
 155                                 $this->output( "Failed to $action $actualTitle\n" );
 156                                 $failCount++;
 157                                 $exit = 1;
 158                         }
 159
 160                         // Create the RecentChanges entry if necessary
 161                         if ( $rc && $status ) {
 162                                 if ( $exists ) {
 163                                         if ( is_object( $oldRev ) ) {
 164                                                 $oldContent = $oldRev->getContent();
 165                                                 RecentChange::notifyEdit(
 166                                                         $timestamp,
 167                                                         $title,
 168                                                         $rev->getMinor(),
 169                                                         $user,
 170                                                         $summary,
 171                                                         $oldRevID,
 172                                                         $oldRev->getTimestamp(),
 173                                                         $bot,
 174                                                         '',
 175                                                         $oldContent ? $oldContent->getSize() : 0,
 176                                                         $rev->getContent()->getSize(),
 177                                                         $newId,
 178                                                         1 /* the pages don't need to be patrolled */
 179                                                 );
 180                                         }
 181                                 } else {
 182                                         RecentChange::notifyNew(
 183                                                 $timestamp,
 184                                                 $title,
 185                                                 $rev->getMinor(),
 186                                                 $user,
 187                                                 $summary,
 188                                                 $bot,
 189                                                 '',
 190                                                 $rev->getContent()->getSize(),
 191                                                 $newId,
 192                                                 1
 193                                         );
 194                                 }
 195                         }
 196                 }
 197
 198                 $this->output( "Done! $successCount succeeded, $skipCount skipped.\n" );
 199                 if ( $exit ) {
 200                         $this->error( "Import failed with $failCount failed pages.\n", $exit );
 201                 }
 202         }
 203 }
 204
 205 $maintClass = "ImportTextFiles";
 206 require_once RUN_MAINTENANCE_IF_MAIN;