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         /*
  46          * Hooks that are ignored
  47          */
  48         protected static $ignore = array( 'testRunLegacyHooks' );
  49
  50         public function __construct() {
  51                 parent::__construct();
  52                 $this->mDescription = 'Find hooks that are undocumented, missing, or just plain wrong';
  53                 $this->addOption( 'online', 'Check against MediaWiki.org hook documentation' );
  54         }
  55
  56         public function getDbType() {
  57                 return Maintenance::DB_NONE;
  58         }
  59
  60         public function execute() {
  61                 global $IP;
  62
  63                 $documented = $this->getHooksFromDoc( $IP . '/docs/hooks.txt' );
  64                 $potential = array();
  65                 $bad = array();
  66
  67                 // TODO: Don't hardcode the list of directories
  68                 $pathinc = array(
  69                         $IP . '/',
  70                         $IP . '/includes/',
  71                         $IP . '/includes/actions/',
  72                         $IP . '/includes/api/',
  73                         $IP . '/includes/cache/',
  74                         $IP . '/includes/changes/',
  75                         $IP . '/includes/clientpool/',
  76                         $IP . '/includes/content/',
  77                         $IP . '/includes/context/',
  78                         $IP . '/includes/dao/',
  79                         $IP . '/includes/db/',
  80                         $IP . '/includes/debug/',
  81                         $IP . '/includes/deferred/',
  82                         $IP . '/includes/diff/',
  83                         $IP . '/includes/externalstore/',
  84                         $IP . '/includes/filebackend/',
  85                         $IP . '/includes/filerepo/',
  86                         $IP . '/includes/filerepo/file/',
  87                         $IP . '/includes/gallery/',
  88                         $IP . '/includes/htmlform/',
  89                         $IP . '/includes/installer/',
  90                         $IP . '/includes/interwiki/',
  91                         $IP . '/includes/jobqueue/',
  92                         $IP . '/includes/json/',
  93                         $IP . '/includes/logging/',
  94                         $IP . '/includes/media/',
  95                         $IP . '/includes/page/',
  96                         $IP . '/includes/parser/',
  97                         $IP . '/includes/rcfeed/',
  98                         $IP . '/includes/resourceloader/',
  99                         $IP . '/includes/revisiondelete/',
 100                         $IP . '/includes/search/',
 101                         $IP . '/includes/site/',
 102                         $IP . '/includes/specialpage/',
 103                         $IP . '/includes/specials/',
 104                         $IP . '/includes/upload/',
 105                         $IP . '/includes/utils/',
 106                         $IP . '/languages/',
 107                         $IP . '/maintenance/',
 108                         $IP . '/maintenance/language/',
 109                         $IP . '/tests/',
 110                         $IP . '/tests/parser/',
 111                         $IP . '/tests/phpunit/suites/',
 112                         $IP . '/skins/',
 113                         $IP . '/skins/MonoBook/',
 114                         $IP . '/skins/Vector/',
 115                 );
 116
 117                 foreach ( $pathinc as $dir ) {
 118                         $potential = array_merge( $potential, $this->getHooksFromPath( $dir ) );
 119                         $bad = array_merge( $bad, $this->getBadHooksFromPath( $dir ) );
 120                 }
 121
 122                 $potential = array_unique( $potential );
 123                 $bad = array_unique( $bad );
 124                 $todo = array_diff( $potential, $documented );
 125                 $deprecated = array_diff( $documented, $potential );
 126
 127                 // let's show the results:
 128                 $this->printArray( 'Undocumented', $todo );
 129                 $this->printArray( 'Documented and not found', $deprecated );
 130                 $this->printArray( 'Unclear hook calls', $bad );
 131
 132                 if ( count( $todo ) == 0 && count( $deprecated ) == 0 && count( $bad ) == 0 ) {
 133                         $this->output( "Looks good!\n" );
 134                 }
 135         }
 136
 137         /**
 138          * Get the hook documentation, either locally or from MediaWiki.org
 139          * @return array of documented hooks
 140          */
 141         private function getHooksFromDoc( $doc ) {
 142                 if ( $this->hasOption( 'online' ) ) {
 143                         return $this->getHooksFromOnlineDoc();
 144                 } else {
 145                         return $this->getHooksFromLocalDoc( $doc );
 146                 }
 147         }
 148
 149         /**
 150          * Get hooks from a local file (for example docs/hooks.txt)
 151          * @param string $doc filename to look in
 152          * @return array Array of documented hooks
 153          */
 154         private function getHooksFromLocalDoc( $doc ) {
 155                 $m = array();
 156                 $content = file_get_contents( $doc );
 157                 preg_match_all( "/\n'(.*?)':/", $content, $m );
 158
 159                 return array_unique( $m[1] );
 160         }
 161
 162         /**
 163          * Get hooks from www.mediawiki.org using the API
 164          * @return array of documented hooks
 165          */
 166         private function getHooksFromOnlineDoc() {
 167                 // All hooks
 168                 $allhookdata = Http::get(
 169                         'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&'
 170                         . 'cmtitle=Category:MediaWiki_hooks&cmlimit=500&format=php'
 171                 );
 172                 $allhookdata = unserialize( $allhookdata );
 173                 $allhooks = array();
 174                 foreach ( $allhookdata['query']['categorymembers'] as $page ) {
 175                         $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 176                         if ( $found ) {
 177                                 $hook = str_replace( ' ', '_', $matches[1] );
 178                                 $allhooks[] = $hook;
 179                         }
 180                 }
 181                 // Removed hooks
 182                 $oldhookdata = Http::get(
 183                         'http://www.mediawiki.org/w/api.php?action=query&list=categorymembers&'
 184                         . 'cmtitle=Category:Removed_hooks&cmlimit=500&format=php'
 185                 );
 186                 $oldhookdata = unserialize( $oldhookdata );
 187                 $removed = array();
 188                 foreach ( $oldhookdata['query']['categorymembers'] as $page ) {
 189                         $found = preg_match( '/Manual\:Hooks\/([a-zA-Z0-9- :]+)/', $page['title'], $matches );
 190                         if ( $found ) {
 191                                 $hook = str_replace( ' ', '_', $matches[1] );
 192                                 $removed[] = $hook;
 193                         }
 194                 }
 195
 196                 return array_diff( $allhooks, $removed );
 197         }
 198
 199         /**
 200          * Get hooks from a PHP file
 201          * @param string $file Full filename to the PHP file.
 202          * @return array of hooks found.
 203          */
 204         private function getHooksFromFile( $file ) {
 205                 $content = file_get_contents( $file );
 206                 $m = array();
 207                 preg_match_all(
 208                         '/(?:wfRunHooks|Hooks\:\:run|ContentHandler\:\:runLegacyHooks)\(\s*([\'"])(.*?)\1/',
 209                         $content,
 210                         $m
 211                 );
 212
 213                 return $m[2];
 214         }
 215
 216         /**
 217          * Get hooks from the source code.
 218          * @param string $path Directory where the include files can be found
 219          * @return array Array of hooks found.
 220          */
 221         private function getHooksFromPath( $path ) {
 222                 $hooks = array();
 223                 $dh = opendir( $path );
 224                 if ( $dh ) {
 225                         while ( ( $file = readdir( $dh ) ) !== false ) {
 226                                 if ( filetype( $path . $file ) == 'file' ) {
 227                                         $hooks = array_merge( $hooks, $this->getHooksFromFile( $path . $file ) );
 228                                 }
 229                         }
 230                         closedir( $dh );
 231                 }
 232
 233                 return $hooks;
 234         }
 235
 236         /**
 237          * Get bad hooks (where the hook name could not be determined) from a PHP file
 238          * @param string $file Full filename to the PHP file.
 239          * @return array Array of bad wfRunHooks() lines
 240          */
 241         private function getBadHooksFromFile( $file ) {
 242                 $content = file_get_contents( $file );
 243                 $m = array();
 244                 # We want to skip the "function wfRunHooks()" one.  :)
 245                 preg_match_all( '/(?<!function )wfRunHooks\(\s*[^\s\'"].*/', $content, $m );
 246                 $list = array();
 247                 foreach ( $m[0] as $match ) {
 248                         $list[] = $match . "(" . $file . ")";
 249                 }
 250
 251                 return $list;
 252         }
 253
 254         /**
 255          * Get bad hooks from the source code.
 256          * @param string $path Directory where the include files can be found
 257          * @return array Array of bad wfRunHooks() lines
 258          */
 259         private function getBadHooksFromPath( $path ) {
 260                 $hooks = array();
 261                 $dh = opendir( $path );
 262                 if ( $dh ) {
 263                         while ( ( $file = readdir( $dh ) ) !== false ) {
 264                                 # We don't want to read this file as it contains bad calls to wfRunHooks()
 265                                 if ( filetype( $path . $file ) == 'file' && !$path . $file == __FILE__ ) {
 266                                         $hooks = array_merge( $hooks, $this->getBadHooksFromFile( $path . $file ) );
 267                                 }
 268                         }
 269                         closedir( $dh );
 270                 }
 271
 272                 return $hooks;
 273         }
 274
 275         /**
 276          * Nicely output the array
 277          * @param string $msg A message to show before the value
 278          * @param array $arr
 279          * @param bool $sort Whether to sort the array (Default: true)
 280          */
 281         private function printArray( $msg, $arr, $sort = true ) {
 282                 if ( $sort ) {
 283                         asort( $arr );
 284                 }
 285
 286                 foreach ( $arr as $v ) {
 287                         if ( !in_array( $v, self::$ignore ) ) {
 288                                 $this->output( "$msg: $v\n" );
 289                         }
 290                 }
 291         }
 292 }
 293
 294 $maintClass = 'FindHooks';
 295 require_once RUN_MAINTENANCE_IF_MAIN;