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( dirname( __FILE__ ) . '/Maintenance.php' );
  38
  39 class FindHooks extends Maintenance {
  40         public function __construct() {
  41                 parent::__construct();
  42                 $this->mDescription = 'Find hooks that are undocumented, missing, or just plain wrong';
  43                 $this->addOption( 'online', 'Check against MediaWiki.org hook documentation' );
  44         }
  45
  46         public function getDbType() {
  47                 return Maintenance::DB_NONE;
  48         }
  49
  50         public function execute() {
  51                 global $IP;
  52
  53                 $documented = $this->getHooksFromDoc( $IP . '/docs/hooks.txt' );
  54                 $potential = array();
  55                 $bad = array();
  56                 $pathinc = array(
  57                         $IP . '/',
  58                         $IP . '/includes/',
  59                         $IP . '/includes/actions/',
  60                         $IP . '/includes/api/',
  61                         $IP . '/includes/cache/',
  62                         $IP . '/includes/context/',
  63                         $IP . '/includes/db/',
  64                         $IP . '/includes/diff/',
  65                         $IP . '/includes/filerepo/',
  66                         $IP . '/includes/installer/',
  67                         $IP . '/includes/interwiki/',
  68                         $IP . '/includes/media/',
  69                         $IP . '/includes/parser/',
  70                         $IP . '/includes/resourceloader/',
  71                         $IP . '/includes/revisiondelete/',
  72                         $IP . '/includes/search/',
  73                         $IP . '/includes/specials/',
  74                         $IP . '/includes/upload/',
  75                         $IP . '/languages/',
  76                         $IP . '/maintenance/',
  77                         $IP . '/tests/',
  78                         $IP . '/tests/parser/',
  79                         $IP . '/tests/phpunit/suites/',
  80                         $IP . '/skins/',
  81                 );
  82
  83                 foreach ( $pathinc as $dir ) {
  84                         $potential = array_merge( $potential, $this->getHooksFromPath( $dir ) );
  85                         $bad = array_merge( $bad, $this->getBadHooksFromPath( $dir ) );
  86                 }
  87
  88                 $potential = array_unique( $potential );
  89                 $bad = array_unique( $bad );
  90                 $todo = array_diff( $potential, $documented );
  91                 $deprecated = array_diff( $documented, $potential );
  92
  93                 // let's show the results:
  94                 $this->printArray( 'Undocumented', $todo );
  95                 $this->printArray( 'Documented and not found', $deprecated );
  96                 $this->printArray( 'Unclear hook calls', $bad );
  97
  98                 if ( count( $todo ) == 0 && count( $deprecated ) == 0 && count( $bad ) == 0 )
  99                 {
 100                         $this->output( "Looks good!\n" );
 101                 }
 102         }
 103
 104         /**
 105          * Get the hook documentation, either locally or from MediaWiki.org
 106          * @return array of documented hooks
 107          */
 108         private function getHooksFromDoc( $doc ) {
 109                 if ( $this->hasOption( 'online' ) ) {
 110                         return $this->getHooksFromOnlineDoc( );
 111                 } else {
 112                         return $this->getHooksFromLocalDoc( $doc );
 113                 }
 114         }
 115
 116         /**
 117          * Get hooks from a local file (for example docs/hooks.txt)
 118          * @param $doc string: filename to look in
 119          * @return array of documented hooks
 120          */
 121         private function getHooksFromLocalDoc( $doc ) {
 122                         $m = array();
 123                         $content = file_get_contents( $doc );
 124                         preg_match_all( "/\n'(.*?)'/", $content, $m );
 125                         return array_unique( $m[1] );
 126         }
 127
 128         /**
 129          * Get hooks from www.mediawiki.org using the API
 130          * @return array of documented hooks
 131          */
 132         private function getHooksFromOnlineDoc( ) {
 133                         // All hooks
 134                         $allhookdata = Http::get( 'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:MediaWiki_hooks&cmlimit=500&format=php' );
 135                         $allhookdata = unserialize( $allhookdata );
 136                         $allhooks = array();
 137                         foreach ( $allhookdata['query']['categorymembers'] as $page ) {
 138                                 $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 139                                 if ( $found ) {
 140                                         $hook = str_replace( ' ', '_', $matches[1] );
 141                                         $allhooks[] = $hook;
 142                                 }
 143                         }
 144                         // Removed hooks
 145                         $oldhookdata = Http::get( 'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:Removed_hooks&cmlimit=500&format=php' );
 146                         $oldhookdata = unserialize( $oldhookdata );
 147                         $removed = array();
 148                         foreach ( $oldhookdata['query']['categorymembers'] as $page ) {
 149                                 $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 150                                 if ( $found ) {
 151                                         $hook = str_replace( ' ', '_', $matches[1] );
 152                                         $removed[] = $hook;
 153                                 }
 154                         }
 155                         return array_diff( $allhooks, $removed );
 156         }
 157
 158         /**
 159          * Get hooks from a PHP file
 160          * @param $file Full filename to the PHP file.
 161          * @return array of hooks found.
 162          */
 163         private function getHooksFromFile( $file ) {
 164                 $content = file_get_contents( $file );
 165                 $m = array();
 166                 preg_match_all( '/(?:wfRunHooks|Hooks\:\:run)\(\s*([\'"])(.*?)\1/', $content, $m );
 167                 return $m[2];
 168         }
 169
 170         /**
 171          * Get hooks from the source code.
 172          * @param $path Directory where the include files can be found
 173          * @return array of hooks found.
 174          */
 175         private function getHooksFromPath( $path ) {
 176                 $hooks = array();
 177                 $dh = opendir( $path );
 178                 if ( $dh ) {
 179                         while ( ( $file = readdir( $dh ) ) !== false ) {
 180                                 if ( filetype( $path . $file ) == 'file' ) {
 181                                         $hooks = array_merge( $hooks, $this->getHooksFromFile( $path . $file ) );
 182                                 }
 183                         }
 184                         closedir( $dh );
 185                 }
 186                 return $hooks;
 187         }
 188
 189         /**
 190          * Get bad hooks (where the hook name could not be determined) from a PHP file
 191          * @param $file Full filename to the PHP file.
 192          * @return array of bad wfRunHooks() lines
 193          */
 194         private function getBadHooksFromFile( $file ) {
 195                 $content = file_get_contents( $file );
 196                 $m = array();
 197                 # We want to skip the "function wfRunHooks()" one.  :)
 198                 preg_match_all( '/(?<!function )wfRunHooks\(\s*[^\s\'"].*/', $content, $m );
 199                 $list = array();
 200                 foreach ( $m[0] as $match ) {
 201                         $list[] = $match . "(" . $file . ")";
 202                 }
 203                 return $list;
 204         }
 205
 206         /**
 207          * Get bad hooks from the source code.
 208          * @param $path Directory where the include files can be found
 209          * @return array of bad wfRunHooks() lines
 210          */
 211         private function getBadHooksFromPath( $path ) {
 212                 $hooks = array();
 213                 $dh = opendir( $path );
 214                 if ( $dh ) {
 215                         while ( ( $file = readdir( $dh ) ) !== false ) {
 216                                 # We don't want to read this file as it contains bad calls to wfRunHooks()
 217                                 if ( filetype( $path . $file ) == 'file' && !$path . $file == __FILE__ ) {
 218                                         $hooks = array_merge( $hooks, $this->getBadHooksFromFile( $path . $file ) );
 219                                 }
 220                         }
 221                         closedir( $dh );
 222                 }
 223                 return $hooks;
 224         }
 225
 226         /**
 227          * Nicely output the array
 228          * @param $msg String: a message to show before the value
 229          * @param $arr Array: an array
 230          * @param $sort Boolean: whether to sort the array (Default: true)
 231          */
 232         private function printArray( $msg, $arr, $sort = true ) {
 233                 if ( $sort ) {
 234                         asort( $arr );
 235                 }
 236                 foreach ( $arr as $v ) {
 237                         $this->output( "$msg: $v\n" );
 238                 }
 239         }
 240 }
 241
 242 $maintClass = 'FindHooks';
 243 require_once( RUN_MAINTENANCE_IF_MAIN );