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                 $pathinc = array(
  67                         $IP . '/',
  68                         $IP . '/includes/',
  69                         $IP . '/includes/actions/',
  70                         $IP . '/includes/api/',
  71                         $IP . '/includes/cache/',
  72                         $IP . '/includes/changes/',
  73                         $IP . '/includes/clientpool/',
  74                         $IP . '/includes/content/',
  75                         $IP . '/includes/context/',
  76                         $IP . '/includes/dao/',
  77                         $IP . '/includes/db/',
  78                         $IP . '/includes/debug/',
  79                         $IP . '/includes/deferred/',
  80                         $IP . '/includes/diff/',
  81                         $IP . '/includes/externalstore/',
  82                         $IP . '/includes/filebackend/',
  83                         $IP . '/includes/filerepo/',
  84                         $IP . '/includes/filerepo/file/',
  85                         $IP . '/includes/gallery/',
  86                         $IP . '/includes/htmlform/',
  87                         $IP . '/includes/installer/',
  88                         $IP . '/includes/interwiki/',
  89                         $IP . '/includes/job/',
  90                         $IP . '/includes/json/',
  91                         $IP . '/includes/logging/',
  92                         $IP . '/includes/media/',
  93                         $IP . '/includes/parser/',
  94                         $IP . '/includes/rcfeed/',
  95                         $IP . '/includes/resourceloader/',
  96                         $IP . '/includes/revisiondelete/',
  97                         $IP . '/includes/search/',
  98                         $IP . '/includes/site/',
  99                         $IP . '/includes/specialpage/',
 100                         $IP . '/includes/specials/',
 101                         $IP . '/includes/upload/',
 102                         $IP . '/languages/',
 103                         $IP . '/maintenance/',
 104                         $IP . '/maintenance/language/',
 105                         $IP . '/tests/',
 106                         $IP . '/tests/parser/',
 107                         $IP . '/tests/phpunit/suites/',
 108                         $IP . '/skins/',
 109                 );
 110
 111                 foreach ( $pathinc as $dir ) {
 112                         $potential = array_merge( $potential, $this->getHooksFromPath( $dir ) );
 113                         $bad = array_merge( $bad, $this->getBadHooksFromPath( $dir ) );
 114                 }
 115
 116                 $potential = array_unique( $potential );
 117                 $bad = array_unique( $bad );
 118                 $todo = array_diff( $potential, $documented );
 119                 $deprecated = array_diff( $documented, $potential );
 120
 121                 // let's show the results:
 122                 $this->printArray( 'Undocumented', $todo );
 123                 $this->printArray( 'Documented and not found', $deprecated );
 124                 $this->printArray( 'Unclear hook calls', $bad );
 125
 126                 if ( count( $todo ) == 0 && count( $deprecated ) == 0 && count( $bad ) == 0 ) {
 127                         $this->output( "Looks good!\n" );
 128                 }
 129         }
 130
 131         /**
 132          * Get the hook documentation, either locally or from MediaWiki.org
 133          * @return array of documented hooks
 134          */
 135         private function getHooksFromDoc( $doc ) {
 136                 if ( $this->hasOption( 'online' ) ) {
 137                         return $this->getHooksFromOnlineDoc();
 138                 } else {
 139                         return $this->getHooksFromLocalDoc( $doc );
 140                 }
 141         }
 142
 143         /**
 144          * Get hooks from a local file (for example docs/hooks.txt)
 145          * @param $doc string: filename to look in
 146          * @return array of documented hooks
 147          */
 148         private function getHooksFromLocalDoc( $doc ) {
 149                         $m = array();
 150                         $content = file_get_contents( $doc );
 151                         preg_match_all( "/\n'(.*?)':/", $content, $m );
 152                         return array_unique( $m[1] );
 153         }
 154
 155         /**
 156          * Get hooks from www.mediawiki.org using the API
 157          * @return array of documented hooks
 158          */
 159         private function getHooksFromOnlineDoc() {
 160                         // All hooks
 161                         $allhookdata = Http::get( 'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:MediaWiki_hooks&cmlimit=500&format=php' );
 162                         $allhookdata = unserialize( $allhookdata );
 163                         $allhooks = array();
 164                         foreach ( $allhookdata['query']['categorymembers'] as $page ) {
 165                                 $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 166                                 if ( $found ) {
 167                                         $hook = str_replace( ' ', '_', $matches[1] );
 168                                         $allhooks[] = $hook;
 169                                 }
 170                         }
 171                         // Removed hooks
 172                         $oldhookdata = Http::get( 'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:Removed_hooks&cmlimit=500&format=php' );
 173                         $oldhookdata = unserialize( $oldhookdata );
 174                         $removed = array();
 175                         foreach ( $oldhookdata['query']['categorymembers'] as $page ) {
 176                                 $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 177                                 if ( $found ) {
 178                                         $hook = str_replace( ' ', '_', $matches[1] );
 179                                         $removed[] = $hook;
 180                                 }
 181                         }
 182                         return array_diff( $allhooks, $removed );
 183         }
 184
 185         /**
 186          * Get hooks from a PHP file
 187          * @param $file string Full filename to the PHP file.
 188          * @return array of hooks found.
 189          */
 190         private function getHooksFromFile( $file ) {
 191                 $content = file_get_contents( $file );
 192                 $m = array();
 193                 preg_match_all( '/(?:wfRunHooks|Hooks\:\:run|ContentHandler\:\:runLegacyHooks)\(\s*([\'"])(.*?)\1/', $content, $m );
 194                 return $m[2];
 195         }
 196
 197         /**
 198          * Get hooks from the source code.
 199          * @param $path Directory where the include files can be found
 200          * @return array of hooks found.
 201          */
 202         private function getHooksFromPath( $path ) {
 203                 $hooks = array();
 204                 $dh = opendir( $path );
 205                 if ( $dh ) {
 206                         while ( ( $file = readdir( $dh ) ) !== false ) {
 207                                 if ( filetype( $path . $file ) == 'file' ) {
 208                                         $hooks = array_merge( $hooks, $this->getHooksFromFile( $path . $file ) );
 209                                 }
 210                         }
 211                         closedir( $dh );
 212                 }
 213                 return $hooks;
 214         }
 215
 216         /**
 217          * Get bad hooks (where the hook name could not be determined) from a PHP file
 218          * @param $file string Full filename to the PHP file.
 219          * @return array of bad wfRunHooks() lines
 220          */
 221         private function getBadHooksFromFile( $file ) {
 222                 $content = file_get_contents( $file );
 223                 $m = array();
 224                 # We want to skip the "function wfRunHooks()" one.  :)
 225                 preg_match_all( '/(?<!function )wfRunHooks\(\s*[^\s\'"].*/', $content, $m );
 226                 $list = array();
 227                 foreach ( $m[0] as $match ) {
 228                         $list[] = $match . "(" . $file . ")";
 229                 }
 230                 return $list;
 231         }
 232
 233         /**
 234          * Get bad hooks from the source code.
 235          * @param $path Directory where the include files can be found
 236          * @return array of bad wfRunHooks() lines
 237          */
 238         private function getBadHooksFromPath( $path ) {
 239                 $hooks = array();
 240                 $dh = opendir( $path );
 241                 if ( $dh ) {
 242                         while ( ( $file = readdir( $dh ) ) !== false ) {
 243                                 # We don't want to read this file as it contains bad calls to wfRunHooks()
 244                                 if ( filetype( $path . $file ) == 'file' && !$path . $file == __FILE__ ) {
 245                                         $hooks = array_merge( $hooks, $this->getBadHooksFromFile( $path . $file ) );
 246                                 }
 247                         }
 248                         closedir( $dh );
 249                 }
 250                 return $hooks;
 251         }
 252
 253         /**
 254          * Nicely output the array
 255          * @param $msg String: a message to show before the value
 256          * @param $arr Array: an array
 257          * @param $sort Boolean: whether to sort the array (Default: true)
 258          */
 259         private function printArray( $msg, $arr, $sort = true ) {
 260                 if ( $sort ) {
 261                         asort( $arr );
 262                 }
 263
 264                 foreach ( $arr as $v ) {
 265                         if ( !in_array( $v, self::$ignore ) ) {
 266                                 $this->output( "$msg: $v\n" );
 267                         }
 268                 }
 269         }
 270 }
 271
 272 $maintClass = 'FindHooks';
 273 require_once RUN_MAINTENANCE_IF_MAIN;