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