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