Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
Total | |
0.00% |
0 / 72 |
|
0.00% |
0 / 6 |
CRAP | |
0.00% |
0 / 1 |
TableCleanup | |
0.00% |
0 / 71 |
|
0.00% |
0 / 6 |
210 | |
0.00% |
0 / 1 |
__construct | |
0.00% |
0 / 4 |
|
0.00% |
0 / 1 |
2 | |||
execute | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
6 | |||
init | |
0.00% |
0 / 5 |
|
0.00% |
0 / 1 |
2 | |||
progress | |
0.00% |
0 / 24 |
|
0.00% |
0 / 1 |
6 | |||
runTable | |
0.00% |
0 / 31 |
|
0.00% |
0 / 1 |
56 | |||
hexChar | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 |
1 | <?php |
2 | /** |
3 | * Generic class to cleanup a database table. |
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\WikiMap\WikiMap; |
25 | |
26 | require_once __DIR__ . '/Maintenance.php'; |
27 | |
28 | /** |
29 | * Generic class to cleanup a database table. Already subclasses Maintenance. |
30 | * |
31 | * @ingroup Maintenance |
32 | */ |
33 | class TableCleanup extends Maintenance { |
34 | protected $defaultParams = [ |
35 | 'table' => 'page', |
36 | 'conds' => [], |
37 | 'index' => 'page_id', |
38 | 'callback' => 'processRow', |
39 | ]; |
40 | |
41 | protected $dryrun = false; |
42 | protected $reportInterval = 100; |
43 | |
44 | protected int $processed; |
45 | protected int $updated; |
46 | protected int $count; |
47 | protected float $startTime; |
48 | protected string $table; |
49 | |
50 | public function __construct() { |
51 | parent::__construct(); |
52 | $this->addOption( 'dry-run', 'Perform a dry run' ); |
53 | $this->addOption( 'reporting-interval', 'How often to print status line' ); |
54 | $this->setBatchSize( 100 ); |
55 | } |
56 | |
57 | public function execute() { |
58 | $this->reportInterval = $this->getOption( 'reporting-interval', $this->reportInterval ); |
59 | |
60 | $this->dryrun = $this->hasOption( 'dry-run' ); |
61 | |
62 | if ( $this->dryrun ) { |
63 | $this->output( "Checking for bad titles...\n" ); |
64 | } else { |
65 | $this->output( "Checking and fixing bad titles...\n" ); |
66 | } |
67 | |
68 | $this->runTable( $this->defaultParams ); |
69 | } |
70 | |
71 | protected function init( $count, $table ) { |
72 | $this->processed = 0; |
73 | $this->updated = 0; |
74 | $this->count = $count; |
75 | $this->startTime = microtime( true ); |
76 | $this->table = $table; |
77 | } |
78 | |
79 | /** |
80 | * @param int $updated |
81 | */ |
82 | protected function progress( $updated ) { |
83 | $this->updated += $updated; |
84 | $this->processed++; |
85 | if ( $this->processed % $this->reportInterval != 0 ) { |
86 | return; |
87 | } |
88 | $portion = $this->processed / $this->count; |
89 | $updateRate = $this->updated / $this->processed; |
90 | |
91 | $now = microtime( true ); |
92 | $delta = $now - $this->startTime; |
93 | $estimatedTotalTime = $delta / $portion; |
94 | $eta = $this->startTime + $estimatedTotalTime; |
95 | |
96 | $this->output( |
97 | sprintf( "%s %s: %6.2f%% done on %s; ETA %s [%d/%d] %.2f/sec <%.2f%% updated>\n", |
98 | WikiMap::getCurrentWikiDbDomain()->getId(), |
99 | wfTimestamp( TS_DB, intval( $now ) ), |
100 | $portion * 100.0, |
101 | $this->table, |
102 | wfTimestamp( TS_DB, intval( $eta ) ), |
103 | $this->processed, |
104 | $this->count, |
105 | $this->processed / $delta, |
106 | $updateRate * 100.0 |
107 | ) |
108 | ); |
109 | flush(); |
110 | } |
111 | |
112 | /** |
113 | * @param array $params |
114 | */ |
115 | public function runTable( $params ) { |
116 | $dbr = $this->getReplicaDB(); |
117 | |
118 | if ( array_diff( array_keys( $params ), |
119 | [ 'table', 'conds', 'index', 'callback' ] ) |
120 | ) { |
121 | $this->fatalError( __METHOD__ . ': Missing parameter ' . implode( ', ', $params ) ); |
122 | } |
123 | |
124 | $table = $params['table']; |
125 | // count(*) would melt the DB for huge tables, we can estimate here |
126 | $count = $dbr->estimateRowCount( $table, '*', '', __METHOD__ ); |
127 | $this->init( $count, $table ); |
128 | $this->output( "Processing $table...\n" ); |
129 | |
130 | $index = (array)$params['index']; |
131 | $indexConds = []; |
132 | $callback = [ $this, $params['callback'] ]; |
133 | |
134 | while ( true ) { |
135 | $conds = array_merge( $params['conds'], $indexConds ); |
136 | $res = $dbr->newSelectQueryBuilder() |
137 | ->select( '*' ) |
138 | ->from( $table ) |
139 | ->where( $conds ) |
140 | ->orderBy( implode( ',', $index ) ) |
141 | ->limit( $this->getBatchSize() ) |
142 | ->caller( __METHOD__ )->fetchResultSet(); |
143 | if ( !$res->numRows() ) { |
144 | // Done |
145 | break; |
146 | } |
147 | |
148 | foreach ( $res as $row ) { |
149 | call_user_func( $callback, $row ); |
150 | } |
151 | |
152 | if ( $res->numRows() < $this->getBatchSize() ) { |
153 | // Done |
154 | break; |
155 | } |
156 | |
157 | // Update the conditions to select the next batch. |
158 | $conds = []; |
159 | foreach ( $index as $field ) { |
160 | // @phan-suppress-next-line PhanPossiblyUndeclaredVariable $res has at at least one item |
161 | $conds[ $field ] = $row->$field; |
162 | } |
163 | $indexConds = [ $dbr->buildComparison( '>', $conds ) ]; |
164 | } |
165 | |
166 | $this->output( "Finished $table... $this->updated of $this->processed rows updated\n" ); |
167 | } |
168 | |
169 | /** |
170 | * @param string[] $matches |
171 | * @return string |
172 | */ |
173 | protected function hexChar( $matches ) { |
174 | return sprintf( "\\x%02x", ord( $matches[1] ) ); |
175 | } |
176 | } |