| Index: trunk/phase3/RELEASE-NOTES-1.18 |
| — | — | @@ -205,6 +205,7 @@ |
| 206 | 206 | targets and $wgServer. |
| 207 | 207 | * Introduced $wgVaryOnXFPForAPI which will cause the API to send |
| 208 | 208 | Vary: X-Forwarded-Proto headers. |
| | 209 | +* New maintenance script to refresh image metadata (maintenance/refreshImageMetadata.php) |
| 209 | 210 | |
| 210 | 211 | === Bug fixes in 1.18 === |
| 211 | 212 | * mw.util.getScript has been implemented (like wfScript in GlobalFunctions.php) |
| Index: trunk/phase3/maintenance/refreshImageMetadata.php |
| — | — | @@ -0,0 +1,199 @@ |
| | 2 | +<?php |
| | 3 | +/** |
| | 4 | + * Script to refresh image metadata fields. See also rebuildImages.php |
| | 5 | + * |
| | 6 | + * Usage: php refreshImageMetadata.php |
| | 7 | + * |
| | 8 | + * Copyright © 2011 Brian Wolff |
| | 9 | + * http://www.mediawiki.org/ |
| | 10 | + * |
| | 11 | + * This program is free software; you can redistribute it and/or modify |
| | 12 | + * it under the terms of the GNU General Public License as published by |
| | 13 | + * the Free Software Foundation; either version 2 of the License, or |
| | 14 | + * (at your option) any later version. |
| | 15 | + * |
| | 16 | + * This program is distributed in the hope that it will be useful, |
| | 17 | + * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| | 18 | + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| | 19 | + * GNU General Public License for more details. |
| | 20 | + * |
| | 21 | + * You should have received a copy of the GNU General Public License along |
| | 22 | + * with this program; if not, write to the Free Software Foundation, Inc., |
| | 23 | + * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
| | 24 | + * http://www.gnu.org/copyleft/gpl.html |
| | 25 | + * |
| | 26 | + * @file |
| | 27 | + * @author Brian Wolff |
| | 28 | + * @ingroup maintenance |
| | 29 | + */ |
| | 30 | + |
| | 31 | +require_once( dirname( __FILE__ ) . '/Maintenance.php' ); |
| | 32 | + |
| | 33 | +class RefreshImageMetadata extends Maintenance { |
| | 34 | + |
| | 35 | + /** |
| | 36 | + * @var DatabaseBase |
| | 37 | + */ |
| | 38 | + protected $dbw; |
| | 39 | + |
| | 40 | + function __construct() { |
| | 41 | + parent::__construct(); |
| | 42 | + |
| | 43 | + $this->mDescription = 'Script to update image metadata records'; |
| | 44 | + $this->setBatchSize( 200 ); |
| | 45 | + |
| | 46 | + $this->addOption( 'force', 'Reload metadata from file even if the metadata looks ok', false, false, 'f' ); |
| | 47 | + $this->addOption( 'broken-only', 'Only fix really broken records, leave old but still compatible records alone.' ); |
| | 48 | + $this->addOption( 'verbose', 'Output extra information about each upgraded/non-upgraded file.', false, false, 'v' ); |
| | 49 | + $this->addOption( 'start', 'Name of file to start with', false, true ); |
| | 50 | + $this->addOption( 'end', 'Name of file to end with', false, true ); |
| | 51 | + |
| | 52 | + $this->addOption( 'mime', '(Inefficient!) Only refresh files with this mime type. Can accept wild-card image/*' , false, true ); |
| | 53 | + $this->addOption( 'metadata-contains', '(Inefficient!) Only refresh files where the img_metadata field contains this string. Can be used if its known a specific property was being extracted incorrectly.', false, true ); |
| | 54 | + |
| | 55 | + } |
| | 56 | + |
| | 57 | + public function execute() { |
| | 58 | + $force = $this->hasOption( 'force' ); |
| | 59 | + $brokenOnly = $this->hasOption( 'broken-only' ); |
| | 60 | + $verbose = $this->hasOption( 'verbose' ); |
| | 61 | + $start = $this->getOption( 'start', false ); |
| | 62 | + $this->setupParameters( $force, $brokenOnly ); |
| | 63 | + |
| | 64 | + $upgraded = 0; |
| | 65 | + $leftAlone = 0; |
| | 66 | + $error = 0; |
| | 67 | + |
| | 68 | + $dbw = wfGetDB( DB_MASTER ); |
| | 69 | + if ( $this->mBatchSize <= 0 ) { |
| | 70 | + $this->error( "Batch size is too low...", 12 ); |
| | 71 | + } |
| | 72 | + |
| | 73 | + $repo = RepoGroup::singleton()->getLocalRepo(); |
| | 74 | + $conds = $this->getConditions( $dbw ); |
| | 75 | + |
| | 76 | + // For the WHERE img_name > 'foo' condition that comes after doing a batch |
| | 77 | + $conds2 = array(); |
| | 78 | + if ( $start !== false ) { |
| | 79 | + $conds2[] = 'img_name >= ' . $dbw->addQuotes( $start ); |
| | 80 | + } |
| | 81 | + |
| | 82 | + $options = array( |
| | 83 | + 'LIMIT' => $this->mBatchSize, |
| | 84 | + 'ORDER BY' => 'img_name ASC', |
| | 85 | + ); |
| | 86 | + |
| | 87 | + do { |
| | 88 | + $res = $dbw->select( |
| | 89 | + 'image', |
| | 90 | + '*', |
| | 91 | + array_merge( $conds, $conds2 ), |
| | 92 | + __METHOD__, |
| | 93 | + $options |
| | 94 | + ); |
| | 95 | + |
| | 96 | + if ( $res->numRows() > 0 ) { |
| | 97 | + $row1 = $res->current(); |
| | 98 | + $this->output( "Processing next {$this->mBatchSize} rows starting with {$row1->img_name}.\n"); |
| | 99 | + $res->rewind(); |
| | 100 | + } else { |
| | 101 | + $this->error( "No images to process.", 4 ); |
| | 102 | + } |
| | 103 | + |
| | 104 | + foreach ( $res as $row ) { |
| | 105 | + $file = $repo->newFileFromRow( $row ); |
| | 106 | + if ( $file->getUpgraded() ) { |
| | 107 | + // File was upgraded. |
| | 108 | + $upgraded++; |
| | 109 | + $newLength = strlen( $file->getMetadata() ); |
| | 110 | + $oldLength = strlen( $row->img_metadata ); |
| | 111 | + if ( $newLength < $oldLength - 5 ) { |
| | 112 | + // If after updating, the metadata is smaller then |
| | 113 | + // what it was before, that's probably not a good thing |
| | 114 | + // because we extract more data with time, not less. |
| | 115 | + // Thus this probably indicates an error of some sort, |
| | 116 | + // or at the very least is suspicious. Have the - 5 just |
| | 117 | + // to weed out any inconsequential changes. |
| | 118 | + $error++; |
| | 119 | + $this->output( "Warning: File:{$row->img_name} used to have " . |
| | 120 | + "$oldLength bytes of metadata but now has $newLength bytes.\n" ); |
| | 121 | + } elseif ( $verbose ) { |
| | 122 | + $this->output("Refreshed File:{$row->img_name}.\n" ); |
| | 123 | + } |
| | 124 | + } else { |
| | 125 | + $leftAlone++; |
| | 126 | + if ( $force ) { |
| | 127 | + $file->upgradeRow(); |
| | 128 | + $newLength = strlen( $file->getMetadata() ); |
| | 129 | + $oldLength = strlen( $row->img_metadata ); |
| | 130 | + if ( $newLength < $oldLength - 5 ) { |
| | 131 | + $error++; |
| | 132 | + $this->output( "Warning: File:{$row->img_name} used to have " . |
| | 133 | + "$oldLength bytes of metadata but now has $newLength bytes. (forced)\n" ); |
| | 134 | + |
| | 135 | + } |
| | 136 | + if ( $verbose ) { |
| | 137 | + $this->output("Forcibly refreshed File:{$row->img_name}.\n" ); |
| | 138 | + } |
| | 139 | + } |
| | 140 | + else { |
| | 141 | + if ( $verbose ) { |
| | 142 | + $this->output( "Skipping File:{$row->img_name}.\n" ); |
| | 143 | + } |
| | 144 | + } |
| | 145 | + } |
| | 146 | + |
| | 147 | + } |
| | 148 | + $conds2 = array( 'img_name > ' . $dbw->addQuotes( $row->img_name ) ); |
| | 149 | + wfWaitForSlaves(); |
| | 150 | + } while( $res->numRows() === $this->mBatchSize ); |
| | 151 | + |
| | 152 | + $total = $upgraded + $leftAlone; |
| | 153 | + if ( $force ) { |
| | 154 | + $this->output( "\nFinished refreshing file metadata for $total files. $upgraded needed to be refreshed, $leftAlone did not need to be but were refreshed anyways, and $error refreshes were suspicious.\n" ); |
| | 155 | + } else { |
| | 156 | + $this->output( "\nFinished refreshing file metadata for $total files. $upgraded were refreshed, $leftAlone were already up to date, and $error refreshes were suspicious.\n" ); |
| | 157 | + } |
| | 158 | + } |
| | 159 | + |
| | 160 | + function getConditions( $dbw ) { |
| | 161 | + $conds = array(); |
| | 162 | + |
| | 163 | + $end = $this->getOption( 'end', false ); |
| | 164 | + $mime = $this->getOption( 'mime', false ); |
| | 165 | + $like = $this->getOption( 'metadata-contains', false ); |
| | 166 | + |
| | 167 | + if ( $end !== false ) { |
| | 168 | + $conds[] = 'img_name <= ' . $dbw->addQuotes( $end ) ; |
| | 169 | + } |
| | 170 | + if ( $mime !== false ) { |
| | 171 | + list( $major, $minor ) = File::splitMime( $mime ); |
| | 172 | + $conds['img_major_mime'] = $major; |
| | 173 | + if ( $minor !== '*' ) { |
| | 174 | + $conds['img_minor_mime'] = $minor; |
| | 175 | + } |
| | 176 | + } |
| | 177 | + if ( $like ) { |
| | 178 | + $conds[] = 'img_metadata ' . $dbw->buildLike( $dbw->anyString(), $like, $dbw->anyString() ); |
| | 179 | + } |
| | 180 | + return $conds; |
| | 181 | + } |
| | 182 | + |
| | 183 | + function setupParameters( $force, $brokenOnly ) { |
| | 184 | + global $wgUpdateCompatibleMetadata, $wgReadOnly; |
| | 185 | + |
| | 186 | + if ( $brokenOnly ) { |
| | 187 | + $wgUpdateCompatibleMetadata = false; |
| | 188 | + } else { |
| | 189 | + $wgUpdateCompatibleMetadata = true; |
| | 190 | + } |
| | 191 | + |
| | 192 | + if ( $brokenOnly && $force ) { |
| | 193 | + $this->error( 'Cannot use --broken-only and --force together. ', 2 ); |
| | 194 | + } |
| | 195 | + } |
| | 196 | +} |
| | 197 | + |
| | 198 | + |
| | 199 | +$maintClass = 'RefreshImageMetadata'; |
| | 200 | +require_once( RUN_MAINTENANCE_IF_MAIN ); |
| Property changes on: trunk/phase3/maintenance/refreshImageMetadata.php |
| ___________________________________________________________________ |
| Added: svn:eol-style |
| 1 | 201 | + native |
| Index: trunk/phase3/maintenance/Maintenance.php |
| — | — | @@ -699,7 +699,7 @@ |
| 700 | 700 | $this->mQuiet = true; |
| 701 | 701 | } |
| 702 | 702 | if ( $this->hasOption( 'batch-size' ) ) { |
| 703 | | - $this->mBatchSize = $this->getOption( 'batch-size' ); |
| | 703 | + $this->mBatchSize = intval( $this->getOption( 'batch-size' ) ); |
| 704 | 704 | } |
| 705 | 705 | } |
| 706 | 706 | |