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