maintenance/findHooks.php

   1 <?php
   2 /**
   3  * Simple script that try to find documented hook and hooks actually
   4  * in the code and show what's missing.
   5  *
   6  * This script assumes that:
   7  * - hooks names in hooks.txt are at the beginning of a line and single quoted.
   8  * - hooks names in code are the first parameter of wfRunHooks.
   9  *
  10  * if --online option is passed, the script will compare the hooks in the code
  11  * with the ones at http://www.mediawiki.org/wiki/Manual:Hooks
  12  *
  13  * Any instance of wfRunHooks that doesn't meet these parameters will be noted.
  14  *
  15  * Copyright © Antoine Musso
  16  *
  17  * This program is free software; you can redistribute it and/or modify
  18  * it under the terms of the GNU General Public License as published by
  19  * the Free Software Foundation; either version 2 of the License, or
  20  * (at your option) any later version.
  21  *
  22  * This program is distributed in the hope that it will be useful,
  23  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  24  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
  25  * GNU General Public License for more details.
  26  *
  27  * You should have received a copy of the GNU General Public License along
  28  * with this program; if not, write to the Free Software Foundation, Inc.,
  29  * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
  30  * http://www.gnu.org/copyleft/gpl.html
  31  *
  32  * @file
  33  * @ingroup Maintenance
  34  * @author Antoine Musso <hashar at free dot fr>
  35  */
  36
  37 require_once __DIR__ . '/Maintenance.php';
  38
  39 /**
  40  * Maintenance script that compares documented and actually present mismatches.
  41  *
  42  * @ingroup Maintenance
  43  */
  44 class FindHooks extends Maintenance {
  45         /*
  46          * Hooks that are ignored
  47          */
  48         protected static $ignore = array( 'testRunLegacyHooks' );
  49
  50         public function __construct() {
  51                 parent::__construct();
  52                 $this->mDescription = 'Find hooks that are undocumented, missing, or just plain wrong';
  53                 $this->addOption( 'online', 'Check against MediaWiki.org hook documentation' );
  54         }
  55
  56         public function getDbType() {
  57                 return Maintenance::DB_NONE;
  58         }
  59
  60         public function execute() {
  61                 global $IP;
  62
  63                 $documented = $this->getHooksFromDoc( $IP . '/docs/hooks.txt' );
  64                 $potential = array();
  65                 $bad = array();
  66
  67                 // TODO: Don't hardcode the list of directories
  68                 $pathinc = array(
  69                         $IP . '/',
  70                         $IP . '/includes/',
  71                         $IP . '/includes/actions/',
  72                         $IP . '/includes/api/',
  73                         $IP . '/includes/cache/',
  74                         $IP . '/includes/changes/',
  75                         $IP . '/includes/clientpool/',
  76                         $IP . '/includes/content/',
  77                         $IP . '/includes/context/',
  78                         $IP . '/includes/dao/',
  79                         $IP . '/includes/db/',
  80                         $IP . '/includes/debug/',
  81                         $IP . '/includes/deferred/',
  82                         $IP . '/includes/diff/',
  83                         $IP . '/includes/externalstore/',
  84                         $IP . '/includes/filebackend/',
  85                         $IP . '/includes/filerepo/',
  86                         $IP . '/includes/filerepo/file/',
  87                         $IP . '/includes/gallery/',
  88                         $IP . '/includes/htmlform/',
  89                         $IP . '/includes/installer/',
  90                         $IP . '/includes/interwiki/',
  91                         $IP . '/includes/jobqueue/',
  92                         $IP . '/includes/json/',
  93                         $IP . '/includes/logging/',
  94                         $IP . '/includes/media/',
  95                         $IP . '/includes/page/',
  96                         $IP . '/includes/parser/',
  97                         $IP . '/includes/rcfeed/',
  98                         $IP . '/includes/resourceloader/',
  99                         $IP . '/includes/revisiondelete/',
 100                         $IP . '/includes/search/',
 101                         $IP . '/includes/site/',
 102                         $IP . '/includes/skins/',
 103                         $IP . '/includes/specialpage/',
 104                         $IP . '/includes/specials/',
 105                         $IP . '/includes/upload/',
 106                         $IP . '/includes/utils/',
 107                         $IP . '/languages/',
 108                         $IP . '/maintenance/',
 109                         $IP . '/maintenance/language/',
 110                         $IP . '/tests/',
 111                         $IP . '/tests/parser/',
 112                         $IP . '/tests/phpunit/suites/',
 113                 );
 114
 115                 foreach ( $pathinc as $dir ) {
 116                         $potential = array_merge( $potential, $this->getHooksFromPath( $dir ) );
 117                         $bad = array_merge( $bad, $this->getBadHooksFromPath( $dir ) );
 118                 }
 119
 120                 $potential = array_unique( $potential );
 121                 $bad = array_unique( $bad );
 122                 $todo = array_diff( $potential, $documented );
 123                 $deprecated = array_diff( $documented, $potential );
 124
 125                 // let's show the results:
 126                 $this->printArray( 'Undocumented', $todo );
 127                 $this->printArray( 'Documented and not found', $deprecated );
 128                 $this->printArray( 'Unclear hook calls', $bad );
 129
 130                 if ( count( $todo ) == 0 && count( $deprecated ) == 0 && count( $bad ) == 0 ) {
 131                         $this->output( "Looks good!\n" );
 132                 }
 133         }
 134
 135         /**
 136          * Get the hook documentation, either locally or from MediaWiki.org
 137          * @param string $doc
 138          * @return array Array of documented hooks
 139          */
 140         private function getHooksFromDoc( $doc ) {
 141                 if ( $this->hasOption( 'online' ) ) {
 142                         return $this->getHooksFromOnlineDoc();
 143                 } else {
 144                         return $this->getHooksFromLocalDoc( $doc );
 145                 }
 146         }
 147
 148         /**
 149          * Get hooks from a local file (for example docs/hooks.txt)
 150          * @param string $doc Filename to look in
 151          * @return array Array of documented hooks
 152          */
 153         private function getHooksFromLocalDoc( $doc ) {
 154                 $m = array();
 155                 $content = file_get_contents( $doc );
 156                 preg_match_all( "/\n'(.*?)':/", $content, $m );
 157
 158                 return array_unique( $m[1] );
 159         }
 160
 161         /**
 162          * Get hooks from www.mediawiki.org using the API
 163          * @return array Array of documented hooks
 164          */
 165         private function getHooksFromOnlineDoc() {
 166                 $allhooks = $this->getHooksFromOnlineDocCategory( 'MediaWiki_hooks' );
 167                 $removed = $this->getHooksFromOnlineDocCategory( 'Removed_hooks' );
 168                 return array_diff( $allhooks, $removed );
 169         }
 170
 171         /**
 172          * @param string $title
 173          * @return array
 174          */
 175         private function getHooksFromOnlineDocCategory( $title ) {
 176                 $params = array(
 177                         'action' => 'query',
 178                         'list' => 'categorymembers',
 179                         'cmtitle' => "Category:$title",
 180                         'cmlimit' => 500,
 181                         'format' => 'json',
 182                         'continue' => '',
 183                 );
 184
 185                 $retval = array();
 186                 while ( true ) {
 187                         $json = Http::get( wfAppendQuery( 'http://www.mediawiki.org/w/api.php', $params ) );
 188                         $data = FormatJson::decode( $json, true );
 189                         foreach ( $data['query']['categorymembers'] as $page ) {
 190                                 if ( preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $m ) ) {
 191                                         $retval[] = str_replace( ' ', '_', $m[1] );
 192                                 }
 193                         }
 194                         if ( !isset( $data['continue'] ) ) {
 195                                 return $retval;
 196                         }
 197                         $params = array_replace( $params, $data['continue'] );
 198                 }
 199         }
 200
 201         /**
 202          * Get hooks from a PHP file
 203          * @param string $file Full filename to the PHP file.
 204          * @return array Array of hooks found
 205          */
 206         private function getHooksFromFile( $file ) {
 207                 $content = file_get_contents( $file );
 208                 $m = array();
 209                 preg_match_all(
 210                         '/(?:wfRunHooks|Hooks\:\:run|ContentHandler\:\:runLegacyHooks)\(\s*([\'"])(.*?)\1/',
 211                         $content,
 212                         $m
 213                 );
 214
 215                 return $m[2];
 216         }
 217
 218         /**
 219          * Get hooks from the source code.
 220          * @param string $path Directory where the include files can be found
 221          * @return array Array of hooks found
 222          */
 223         private function getHooksFromPath( $path ) {
 224                 $hooks = array();
 225                 $dh = opendir( $path );
 226                 if ( $dh ) {
 227                         while ( ( $file = readdir( $dh ) ) !== false ) {
 228                                 if ( filetype( $path . $file ) == 'file' ) {
 229                                         $hooks = array_merge( $hooks, $this->getHooksFromFile( $path . $file ) );
 230                                 }
 231                         }
 232                         closedir( $dh );
 233                 }
 234
 235                 return $hooks;
 236         }
 237
 238         /**
 239          * Get bad hooks (where the hook name could not be determined) from a PHP file
 240          * @param string $file Full filename to the PHP file.
 241          * @return array Array of bad wfRunHooks() lines
 242          */
 243         private function getBadHooksFromFile( $file ) {
 244                 $content = file_get_contents( $file );
 245                 $m = array();
 246                 # We want to skip the "function wfRunHooks()" one.  :)
 247                 preg_match_all( '/(?<!function )wfRunHooks\(\s*[^\s\'"].*/', $content, $m );
 248                 $list = array();
 249                 foreach ( $m[0] as $match ) {
 250                         $list[] = $match . "(" . $file . ")";
 251                 }
 252
 253                 return $list;
 254         }
 255
 256         /**
 257          * Get bad hooks from the source code.
 258          * @param string $path Directory where the include files can be found
 259          * @return array Array of bad wfRunHooks() lines
 260          */
 261         private function getBadHooksFromPath( $path ) {
 262                 $hooks = array();
 263                 $dh = opendir( $path );
 264                 if ( $dh ) {
 265                         while ( ( $file = readdir( $dh ) ) !== false ) {
 266                                 # We don't want to read this file as it contains bad calls to wfRunHooks()
 267                                 if ( filetype( $path . $file ) == 'file' && !$path . $file == __FILE__ ) {
 268                                         $hooks = array_merge( $hooks, $this->getBadHooksFromFile( $path . $file ) );
 269                                 }
 270                         }
 271                         closedir( $dh );
 272                 }
 273
 274                 return $hooks;
 275         }
 276
 277         /**
 278          * Nicely output the array
 279          * @param string $msg A message to show before the value
 280          * @param array $arr
 281          * @param bool $sort Whether to sort the array (Default: true)
 282          */
 283         private function printArray( $msg, $arr, $sort = true ) {
 284                 if ( $sort ) {
 285                         asort( $arr );
 286                 }
 287
 288                 foreach ( $arr as $v ) {
 289                         if ( !in_array( $v, self::$ignore ) ) {
 290                                 $this->output( "$msg: $v\n" );
 291                         }
 292                 }
 293         }
 294 }
 295
 296 $maintClass = 'FindHooks';
 297 require_once RUN_MAINTENANCE_IF_MAIN;