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         public function __construct() {
  46                 parent::__construct();
  47                 $this->mDescription = 'Find hooks that are undocumented, missing, or just plain wrong';
  48                 $this->addOption( 'online', 'Check against MediaWiki.org hook documentation' );
  49         }
  50
  51         public function getDbType() {
  52                 return Maintenance::DB_NONE;
  53         }
  54
  55         public function execute() {
  56                 global $IP;
  57
  58                 $documented = $this->getHooksFromDoc( $IP . '/docs/hooks.txt' );
  59                 $potential = array();
  60                 $bad = array();
  61                 $pathinc = array(
  62                         $IP . '/',
  63                         $IP . '/includes/',
  64                         $IP . '/includes/actions/',
  65                         $IP . '/includes/api/',
  66                         $IP . '/includes/cache/',
  67                         $IP . '/includes/content/',
  68                         $IP . '/includes/context/',
  69                         $IP . '/includes/db/',
  70                         $IP . '/includes/diff/',
  71                         $IP . '/includes/filerepo/',
  72                         $IP . '/includes/filerepo/file/',
  73                         $IP . '/includes/installer/',
  74                         $IP . '/includes/interwiki/',
  75                         $IP . '/includes/logging/',
  76                         $IP . '/includes/media/',
  77                         $IP . '/includes/parser/',
  78                         $IP . '/includes/resourceloader/',
  79                         $IP . '/includes/revisiondelete/',
  80                         $IP . '/includes/search/',
  81                         $IP . '/includes/specials/',
  82                         $IP . '/includes/upload/',
  83                         $IP . '/languages/',
  84                         $IP . '/maintenance/',
  85                         $IP . '/tests/',
  86                         $IP . '/tests/parser/',
  87                         $IP . '/tests/phpunit/suites/',
  88                         $IP . '/skins/',
  89                 );
  90
  91                 foreach ( $pathinc as $dir ) {
  92                         $potential = array_merge( $potential, $this->getHooksFromPath( $dir ) );
  93                         $bad = array_merge( $bad, $this->getBadHooksFromPath( $dir ) );
  94                 }
  95
  96                 $potential = array_unique( $potential );
  97                 $bad = array_unique( $bad );
  98                 $todo = array_diff( $potential, $documented );
  99                 $deprecated = array_diff( $documented, $potential );
 100
 101                 // let's show the results:
 102                 $this->printArray( 'Undocumented', $todo );
 103                 $this->printArray( 'Documented and not found', $deprecated );
 104                 $this->printArray( 'Unclear hook calls', $bad );
 105
 106                 if ( count( $todo ) == 0 && count( $deprecated ) == 0 && count( $bad ) == 0 ) {
 107                         $this->output( "Looks good!\n" );
 108                 }
 109         }
 110
 111         /**
 112          * Get the hook documentation, either locally or from MediaWiki.org
 113          * @return array of documented hooks
 114          */
 115         private function getHooksFromDoc( $doc ) {
 116                 if ( $this->hasOption( 'online' ) ) {
 117                         return $this->getHooksFromOnlineDoc();
 118                 } else {
 119                         return $this->getHooksFromLocalDoc( $doc );
 120                 }
 121         }
 122
 123         /**
 124          * Get hooks from a local file (for example docs/hooks.txt)
 125          * @param $doc string: filename to look in
 126          * @return array of documented hooks
 127          */
 128         private function getHooksFromLocalDoc( $doc ) {
 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          * Get hooks from www.mediawiki.org using the API
 137          * @return array of documented hooks
 138          */
 139         private function getHooksFromOnlineDoc() {
 140                         // All hooks
 141                         $allhookdata = Http::get( 'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:MediaWiki_hooks&cmlimit=500&format=php' );
 142                         $allhookdata = unserialize( $allhookdata );
 143                         $allhooks = array();
 144                         foreach ( $allhookdata['query']['categorymembers'] as $page ) {
 145                                 $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 146                                 if ( $found ) {
 147                                         $hook = str_replace( ' ', '_', $matches[1] );
 148                                         $allhooks[] = $hook;
 149                                 }
 150                         }
 151                         // Removed hooks
 152                         $oldhookdata = Http::get( 'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&cmtitle=Category:Removed_hooks&cmlimit=500&format=php' );
 153                         $oldhookdata = unserialize( $oldhookdata );
 154                         $removed = array();
 155                         foreach ( $oldhookdata['query']['categorymembers'] as $page ) {
 156                                 $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 157                                 if ( $found ) {
 158                                         $hook = str_replace( ' ', '_', $matches[1] );
 159                                         $removed[] = $hook;
 160                                 }
 161                         }
 162                         return array_diff( $allhooks, $removed );
 163         }
 164
 165         /**
 166          * Get hooks from a PHP file
 167          * @param $file string Full filename to the PHP file.
 168          * @return array of hooks found.
 169          */
 170         private function getHooksFromFile( $file ) {
 171                 $content = file_get_contents( $file );
 172                 $m = array();
 173                 preg_match_all( '/(?:wfRunHooks|Hooks\:\:run|ContentHandler\:\:runLegacyHooks)\(\s*([\'"])(.*?)\1/', $content, $m );
 174                 return $m[2];
 175         }
 176
 177         /**
 178          * Get hooks from the source code.
 179          * @param $path Directory where the include files can be found
 180          * @return array of hooks found.
 181          */
 182         private function getHooksFromPath( $path ) {
 183                 $hooks = array();
 184                 $dh = opendir( $path );
 185                 if ( $dh ) {
 186                         while ( ( $file = readdir( $dh ) ) !== false ) {
 187                                 if ( filetype( $path . $file ) == 'file' ) {
 188                                         $hooks = array_merge( $hooks, $this->getHooksFromFile( $path . $file ) );
 189                                 }
 190                         }
 191                         closedir( $dh );
 192                 }
 193                 return $hooks;
 194         }
 195
 196         /**
 197          * Get bad hooks (where the hook name could not be determined) from a PHP file
 198          * @param $file string Full filename to the PHP file.
 199          * @return array of bad wfRunHooks() lines
 200          */
 201         private function getBadHooksFromFile( $file ) {
 202                 $content = file_get_contents( $file );
 203                 $m = array();
 204                 # We want to skip the "function wfRunHooks()" one.  :)
 205                 preg_match_all( '/(?<!function )wfRunHooks\(\s*[^\s\'"].*/', $content, $m );
 206                 $list = array();
 207                 foreach ( $m[0] as $match ) {
 208                         $list[] = $match . "(" . $file . ")";
 209                 }
 210                 return $list;
 211         }
 212
 213         /**
 214          * Get bad hooks from the source code.
 215          * @param $path Directory where the include files can be found
 216          * @return array of bad wfRunHooks() lines
 217          */
 218         private function getBadHooksFromPath( $path ) {
 219                 $hooks = array();
 220                 $dh = opendir( $path );
 221                 if ( $dh ) {
 222                         while ( ( $file = readdir( $dh ) ) !== false ) {
 223                                 # We don't want to read this file as it contains bad calls to wfRunHooks()
 224                                 if ( filetype( $path . $file ) == 'file' && !$path . $file == __FILE__ ) {
 225                                         $hooks = array_merge( $hooks, $this->getBadHooksFromFile( $path . $file ) );
 226                                 }
 227                         }
 228                         closedir( $dh );
 229                 }
 230                 return $hooks;
 231         }
 232
 233         /**
 234          * Nicely output the array
 235          * @param $msg String: a message to show before the value
 236          * @param $arr Array: an array
 237          * @param $sort Boolean: whether to sort the array (Default: true)
 238          */
 239         private function printArray( $msg, $arr, $sort = true ) {
 240                 if ( $sort ) {
 241                         asort( $arr );
 242                 }
 243                 foreach ( $arr as $v ) {
 244                         $this->output( "$msg: $v\n" );
 245                 }
 246         }
 247 }
 248
 249 $maintClass = 'FindHooks';
 250 require_once RUN_MAINTENANCE_IF_MAIN;