Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
4.29% |
6 / 140 |
|
5.00% |
1 / 20 |
CRAP | |
0.00% |
0 / 1 |
| BaseBlacklist | |
4.29% |
6 / 140 |
|
5.00% |
1 / 20 |
2242.15 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
6 | |||
| filter | n/a |
0 / 0 |
n/a |
0 / 0 |
0 | |||||
| addBlacklistType | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getBlacklistTypes | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getSpamBlacklist | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getEmailBlacklist | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getInstance | |
0.00% |
0 / 8 |
|
0.00% |
0 / 1 |
20 | |||
| clearInstanceCache | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getBlacklistType | n/a |
0 / 0 |
n/a |
0 / 0 |
0 | |||||
| isLocalSource | |
0.00% |
0 / 22 |
|
0.00% |
0 / 1 |
132 | |||
| getTypeFromTitle | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
2 | |||
| getBlacklists | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
6 | |||
| getLocalBlacklists | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
2 | |||
| getWhitelists | |
0.00% |
0 / 9 |
|
0.00% |
0 / 1 |
2 | |||
| getSharedBlacklists | |
0.00% |
0 / 21 |
|
0.00% |
0 / 1 |
20 | |||
| clearCache | |
0.00% |
0 / 6 |
|
0.00% |
0 / 1 |
2 | |||
| buildSharedBlacklists | |
0.00% |
0 / 17 |
|
0.00% |
0 / 1 |
30 | |||
| getHttpText | |
0.00% |
0 / 16 |
|
0.00% |
0 / 1 |
30 | |||
| getArticleText | |
0.00% |
0 / 10 |
|
0.00% |
0 / 1 |
12 | |||
| getRegexStart | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| getRegexEnd | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
6 | |||
| warmCachesForFilter | |
0.00% |
0 / 1 |
|
0.00% |
0 / 1 |
2 | |||
| 1 | <?php |
| 2 | |
| 3 | namespace MediaWiki\Extension\SpamBlacklist; |
| 4 | |
| 5 | use InvalidArgumentException; |
| 6 | use MediaWiki\MediaWikiServices; |
| 7 | use MediaWiki\Revision\SlotRecord; |
| 8 | use MediaWiki\Title\Title; |
| 9 | use MediaWiki\User\User; |
| 10 | use ObjectCache; |
| 11 | use TextContent; |
| 12 | |
| 13 | /** |
| 14 | * Base class for different kinds of blacklists |
| 15 | */ |
| 16 | abstract class BaseBlacklist { |
| 17 | /** |
| 18 | * Array of blacklist sources |
| 19 | * |
| 20 | * @var string[] |
| 21 | */ |
| 22 | public $files = []; |
| 23 | |
| 24 | /** |
| 25 | * Array containing regexes to test against |
| 26 | * |
| 27 | * @var string[]|false |
| 28 | */ |
| 29 | protected $regexes = false; |
| 30 | |
| 31 | /** |
| 32 | * Chance of receiving a warning when the filter is hit |
| 33 | * |
| 34 | * @var int |
| 35 | */ |
| 36 | public $warningChance = 100; |
| 37 | |
| 38 | /** |
| 39 | * @var int |
| 40 | */ |
| 41 | public $warningTime = 600; |
| 42 | |
| 43 | /** |
| 44 | * @var int |
| 45 | */ |
| 46 | public $expiryTime = 900; |
| 47 | |
| 48 | /** |
| 49 | * Array containing blacklists that extend BaseBlacklist |
| 50 | * |
| 51 | * @var string[] |
| 52 | */ |
| 53 | private static $blacklistTypes = [ |
| 54 | 'spam' => SpamBlacklist::class, |
| 55 | 'email' => EmailBlacklist::class, |
| 56 | ]; |
| 57 | |
| 58 | /** |
| 59 | * Array of blacklist instances |
| 60 | * |
| 61 | * @var self[] |
| 62 | */ |
| 63 | private static $instances = []; |
| 64 | |
| 65 | /** |
| 66 | * @param array $settings |
| 67 | */ |
| 68 | public function __construct( $settings = [] ) { |
| 69 | foreach ( $settings as $name => $value ) { |
| 70 | $this->$name = $value; |
| 71 | } |
| 72 | } |
| 73 | |
| 74 | /** |
| 75 | * @param array $links |
| 76 | * @param ?Title $title |
| 77 | * @param User $user |
| 78 | * @param bool $preventLog |
| 79 | * @return mixed |
| 80 | */ |
| 81 | abstract public function filter( |
| 82 | array $links, |
| 83 | ?Title $title, |
| 84 | User $user, |
| 85 | $preventLog = false |
| 86 | ); |
| 87 | |
| 88 | /** |
| 89 | * Adds a blacklist class to the registry |
| 90 | * |
| 91 | * @param string $type |
| 92 | * @param string $class |
| 93 | */ |
| 94 | public static function addBlacklistType( $type, $class ) { |
| 95 | self::$blacklistTypes[$type] = $class; |
| 96 | } |
| 97 | |
| 98 | /** |
| 99 | * Return the array of blacklist types currently defined |
| 100 | * |
| 101 | * @return string[] |
| 102 | */ |
| 103 | public static function getBlacklistTypes() { |
| 104 | return self::$blacklistTypes; |
| 105 | } |
| 106 | |
| 107 | /** |
| 108 | * @return SpamBlacklist |
| 109 | */ |
| 110 | public static function getSpamBlacklist() { |
| 111 | // @phan-suppress-next-line PhanTypeMismatchReturnSuperType |
| 112 | return self::getInstance( 'spam' ); |
| 113 | } |
| 114 | |
| 115 | /** |
| 116 | * @return EmailBlacklist |
| 117 | */ |
| 118 | public static function getEmailBlacklist() { |
| 119 | // @phan-suppress-next-line PhanTypeMismatchReturnSuperType |
| 120 | return self::getInstance( 'email' ); |
| 121 | } |
| 122 | |
| 123 | /** |
| 124 | * Returns an instance of the given blacklist |
| 125 | * |
| 126 | * @deprecated Use getSpamBlacklist() or getEmailBlacklist() instead |
| 127 | * @param string $type Code for the blacklist |
| 128 | * @return BaseBlacklist |
| 129 | */ |
| 130 | public static function getInstance( $type ) { |
| 131 | if ( !isset( self::$blacklistTypes[$type] ) ) { |
| 132 | throw new InvalidArgumentException( "Invalid blacklist type '$type' passed to " . __METHOD__ ); |
| 133 | } |
| 134 | |
| 135 | if ( !isset( self::$instances[$type] ) ) { |
| 136 | global $wgBlacklistSettings; |
| 137 | |
| 138 | // Prevent notices |
| 139 | if ( !isset( $wgBlacklistSettings[$type] ) ) { |
| 140 | $wgBlacklistSettings[$type] = []; |
| 141 | } |
| 142 | |
| 143 | $class = self::$blacklistTypes[$type]; |
| 144 | self::$instances[$type] = new $class( $wgBlacklistSettings[$type] ); |
| 145 | } |
| 146 | |
| 147 | return self::$instances[$type]; |
| 148 | } |
| 149 | |
| 150 | /** |
| 151 | * Clear instance cache. For use during testing. |
| 152 | */ |
| 153 | public static function clearInstanceCache() { |
| 154 | self::$instances = []; |
| 155 | } |
| 156 | |
| 157 | /** |
| 158 | * Returns the code for the blacklist implementation |
| 159 | * |
| 160 | * @return string |
| 161 | */ |
| 162 | abstract protected function getBlacklistType(); |
| 163 | |
| 164 | /** |
| 165 | * Check if the given local page title is a spam regex source. |
| 166 | * |
| 167 | * @param Title $title |
| 168 | * @return bool |
| 169 | */ |
| 170 | public static function isLocalSource( Title $title ) { |
| 171 | global $wgDBname, $wgBlacklistSettings; |
| 172 | |
| 173 | if ( $title->inNamespace( NS_MEDIAWIKI ) ) { |
| 174 | $sources = []; |
| 175 | foreach ( self::$blacklistTypes as $type => $class ) { |
| 176 | // For the built in types, this results in the use of: |
| 177 | // spam-blacklist, spam-whitelist |
| 178 | // email-blacklist, email-whitelist |
| 179 | $type = ucfirst( $type ); |
| 180 | $sources[] = "$type-blacklist"; |
| 181 | $sources[] = "$type-whitelist"; |
| 182 | } |
| 183 | |
| 184 | if ( in_array( $title->getDBkey(), $sources ) ) { |
| 185 | return true; |
| 186 | } |
| 187 | } |
| 188 | |
| 189 | $thisHttp = wfExpandUrl( $title->getFullUrl( 'action=raw' ), PROTO_HTTP ); |
| 190 | $thisHttpRegex = '/^' . preg_quote( $thisHttp, '/' ) . '(?:&.*)?$/'; |
| 191 | |
| 192 | $files = []; |
| 193 | foreach ( self::$blacklistTypes as $type => $class ) { |
| 194 | if ( isset( $wgBlacklistSettings[$type]['files'] ) ) { |
| 195 | $files += $wgBlacklistSettings[$type]['files']; |
| 196 | } |
| 197 | } |
| 198 | |
| 199 | foreach ( $files as $fileName ) { |
| 200 | $matches = []; |
| 201 | if ( preg_match( '/^DB: (\w*) (.*)$/', $fileName, $matches ) ) { |
| 202 | if ( $wgDBname === $matches[1] && $matches[2] === $title->getPrefixedDbKey() ) { |
| 203 | // Local DB fetch of this page... |
| 204 | return true; |
| 205 | } |
| 206 | } elseif ( preg_match( $thisHttpRegex, $fileName ) ) { |
| 207 | // Raw view of this page |
| 208 | return true; |
| 209 | } |
| 210 | } |
| 211 | |
| 212 | return false; |
| 213 | } |
| 214 | |
| 215 | /** |
| 216 | * Returns the type of blacklist from the given title |
| 217 | * |
| 218 | * @todo building a regex for this is pretty overkill |
| 219 | * @param Title $title |
| 220 | * @return bool|string |
| 221 | */ |
| 222 | public static function getTypeFromTitle( Title $title ) { |
| 223 | $contLang = MediaWikiServices::getInstance()->getContentLanguage(); |
| 224 | |
| 225 | $types = array_map( [ $contLang, 'ucfirst' ], array_keys( self::$blacklistTypes ) ); |
| 226 | $regex = '/(' . implode( '|', $types ) . ')-(?:blacklist|whitelist)/'; |
| 227 | |
| 228 | if ( preg_match( $regex, $title->getDBkey(), $m ) ) { |
| 229 | return strtolower( $m[1] ); |
| 230 | } |
| 231 | |
| 232 | return false; |
| 233 | } |
| 234 | |
| 235 | /** |
| 236 | * Fetch local and (possibly cached) remote blacklists. |
| 237 | * Will be cached locally across multiple invocations. |
| 238 | * @return string[] set of regular expressions, potentially empty. |
| 239 | */ |
| 240 | public function getBlacklists() { |
| 241 | if ( $this->regexes === false ) { |
| 242 | $this->regexes = array_merge( |
| 243 | $this->getLocalBlacklists(), |
| 244 | $this->getSharedBlacklists() |
| 245 | ); |
| 246 | } |
| 247 | return $this->regexes; |
| 248 | } |
| 249 | |
| 250 | /** |
| 251 | * Returns the local blacklist |
| 252 | * |
| 253 | * @return string[] Regular expressions |
| 254 | */ |
| 255 | public function getLocalBlacklists() { |
| 256 | $type = $this->getBlacklistType(); |
| 257 | $cache = MediaWikiServices::getInstance()->getMainWANObjectCache(); |
| 258 | |
| 259 | return $cache->getWithSetCallback( |
| 260 | $cache->makeKey( 'spamblacklist', $type, 'blacklist-regex' ), |
| 261 | $this->expiryTime, |
| 262 | function () use ( $type ) { |
| 263 | return SpamRegexBatch::regexesFromMessage( "{$type}-blacklist", $this ); |
| 264 | } |
| 265 | ); |
| 266 | } |
| 267 | |
| 268 | /** |
| 269 | * Returns the (local) whitelist |
| 270 | * |
| 271 | * @return string[] Regular expressions |
| 272 | */ |
| 273 | public function getWhitelists() { |
| 274 | $type = $this->getBlacklistType(); |
| 275 | $cache = MediaWikiServices::getInstance()->getMainWANObjectCache(); |
| 276 | |
| 277 | return $cache->getWithSetCallback( |
| 278 | $cache->makeKey( 'spamblacklist', $type, 'whitelist-regex' ), |
| 279 | $this->expiryTime, |
| 280 | function () use ( $type ) { |
| 281 | return SpamRegexBatch::regexesFromMessage( "{$type}-whitelist", $this ); |
| 282 | } |
| 283 | ); |
| 284 | } |
| 285 | |
| 286 | /** |
| 287 | * Fetch (possibly cached) remote blacklists. |
| 288 | * @return array |
| 289 | */ |
| 290 | private function getSharedBlacklists() { |
| 291 | $listType = $this->getBlacklistType(); |
| 292 | |
| 293 | wfDebugLog( 'SpamBlacklist', "Loading $listType regex..." ); |
| 294 | |
| 295 | if ( !$this->files ) { |
| 296 | # No lists |
| 297 | wfDebugLog( 'SpamBlacklist', "no files specified\n" ); |
| 298 | return []; |
| 299 | } |
| 300 | |
| 301 | if ( defined( 'MW_PHPUNIT_TEST' ) ) { |
| 302 | wfDebugLog( 'SpamBlacklist', 'remote loading disabled during PHPUnit test' ); |
| 303 | return []; |
| 304 | } |
| 305 | |
| 306 | $miss = false; |
| 307 | $cache = MediaWikiServices::getInstance()->getMainWANObjectCache(); |
| 308 | $regexes = $cache->getWithSetCallback( |
| 309 | // This used to be cached per-site, but that could be bad on a shared |
| 310 | // server where not all wikis have the same configuration. |
| 311 | $cache->makeKey( 'spamblacklist', $listType, 'shared-blacklist-regex' ), |
| 312 | $this->expiryTime, |
| 313 | function () use ( &$miss ) { |
| 314 | $miss = true; |
| 315 | return $this->buildSharedBlacklists(); |
| 316 | } |
| 317 | ); |
| 318 | |
| 319 | if ( !$miss ) { |
| 320 | wfDebugLog( 'SpamBlacklist', "Got shared spam regexes from cache\n" ); |
| 321 | } |
| 322 | |
| 323 | return $regexes; |
| 324 | } |
| 325 | |
| 326 | /** |
| 327 | * Clear all primary blacklist cache keys |
| 328 | */ |
| 329 | public function clearCache() { |
| 330 | $listType = $this->getBlacklistType(); |
| 331 | |
| 332 | $cache = MediaWikiServices::getInstance()->getMainWANObjectCache(); |
| 333 | $cache->delete( $cache->makeKey( 'spamblacklist', $listType, 'shared-blacklist-regex' ) ); |
| 334 | $cache->delete( $cache->makeKey( 'spamblacklist', $listType, 'blacklist-regex' ) ); |
| 335 | $cache->delete( $cache->makeKey( 'spamblacklist', $listType, 'whitelist-regex' ) ); |
| 336 | |
| 337 | wfDebugLog( 'SpamBlacklist', "$listType blacklist local cache cleared.\n" ); |
| 338 | } |
| 339 | |
| 340 | private function buildSharedBlacklists() { |
| 341 | $regexes = []; |
| 342 | $listType = $this->getBlacklistType(); |
| 343 | # Load lists |
| 344 | wfDebugLog( 'SpamBlacklist', "Constructing $listType blacklist\n" ); |
| 345 | foreach ( $this->files as $fileName ) { |
| 346 | $matches = []; |
| 347 | if ( preg_match( '/^DB: ([\w-]*) (.*)$/', $fileName, $matches ) ) { |
| 348 | $text = $this->getArticleText( $matches[1], $matches[2] ); |
| 349 | } elseif ( preg_match( '/^(https?:)?\/\//', $fileName ) ) { |
| 350 | $text = $this->getHttpText( $fileName ); |
| 351 | } else { |
| 352 | $text = file_get_contents( $fileName ); |
| 353 | wfDebugLog( 'SpamBlacklist', "got from file $fileName\n" ); |
| 354 | } |
| 355 | |
| 356 | if ( $text ) { |
| 357 | // Build a separate batch of regexes from each source. |
| 358 | // While in theory we could squeeze a little efficiency |
| 359 | // out of combining multiple sources in one regex, if |
| 360 | // there's a bad line in one of them we'll gain more |
| 361 | // from only having to break that set into smaller pieces. |
| 362 | $regexes = array_merge( |
| 363 | $regexes, |
| 364 | SpamRegexBatch::regexesFromText( $text, $this, $fileName ) |
| 365 | ); |
| 366 | } |
| 367 | } |
| 368 | |
| 369 | return $regexes; |
| 370 | } |
| 371 | |
| 372 | private function getHttpText( $fileName ) { |
| 373 | global $wgMessageCacheType; |
| 374 | // FIXME: This is a hack to use Memcached where possible (incl. WMF), |
| 375 | // but have CACHE_DB as fallback (instead of no cache). |
| 376 | // This might be a good candidate for T248005. |
| 377 | $cache = ObjectCache::getInstance( $wgMessageCacheType ); |
| 378 | |
| 379 | $listType = $this->getBlacklistType(); |
| 380 | // There are two keys, when the warning key expires, a random thread will refresh |
| 381 | // the real key. This reduces the chance of multiple requests under high traffic |
| 382 | // conditions. |
| 383 | $key = $cache->makeGlobalKey( "blacklist_file_{$listType}", $fileName ); |
| 384 | $warningKey = $cache->makeKey( "filewarning_{$listType}", $fileName ); |
| 385 | $httpText = $cache->get( $key ); |
| 386 | $warning = $cache->get( $warningKey ); |
| 387 | |
| 388 | if ( !is_string( $httpText ) || ( !$warning && !mt_rand( 0, $this->warningChance ) ) ) { |
| 389 | wfDebugLog( 'SpamBlacklist', "Loading $listType blacklist from $fileName\n" ); |
| 390 | $httpText = MediaWikiServices::getInstance()->getHttpRequestFactory() |
| 391 | ->get( $fileName, [], __METHOD__ ); |
| 392 | if ( $httpText === false ) { |
| 393 | wfDebugLog( 'SpamBlacklist', "Error loading $listType blacklist from $fileName\n" ); |
| 394 | } |
| 395 | $cache->set( $warningKey, 1, $this->warningTime ); |
| 396 | $cache->set( $key, $httpText, $this->expiryTime ); |
| 397 | } else { |
| 398 | wfDebugLog( 'SpamBlacklist', "Got $listType blacklist from HTTP cache for $fileName\n" ); |
| 399 | } |
| 400 | return $httpText; |
| 401 | } |
| 402 | |
| 403 | /** |
| 404 | * Fetch an article from this or another local MediaWiki database. |
| 405 | * |
| 406 | * @param string $wiki |
| 407 | * @param string $pagename |
| 408 | * @return bool|string|null |
| 409 | */ |
| 410 | private function getArticleText( $wiki, $pagename ) { |
| 411 | wfDebugLog( 'SpamBlacklist', |
| 412 | "Fetching {$this->getBlacklistType()} blacklist from '$pagename' on '$wiki'...\n" ); |
| 413 | |
| 414 | $services = MediaWikiServices::getInstance(); |
| 415 | |
| 416 | // XXX: We do not know about custom namespaces on the target wiki here! |
| 417 | $title = $services->getTitleParser()->parseTitle( $pagename ); |
| 418 | $store = $services->getRevisionStoreFactory()->getRevisionStore( $wiki ); |
| 419 | $rev = $store->getRevisionByTitle( $title ); |
| 420 | |
| 421 | $content = $rev ? $rev->getContent( SlotRecord::MAIN ) : null; |
| 422 | |
| 423 | if ( !( $content instanceof TextContent ) ) { |
| 424 | return false; |
| 425 | } |
| 426 | |
| 427 | return $content->getText(); |
| 428 | } |
| 429 | |
| 430 | /** |
| 431 | * Returns the start of the regex for matches |
| 432 | * |
| 433 | * @return string |
| 434 | */ |
| 435 | public function getRegexStart() { |
| 436 | return '/[a-z0-9_\-.]*'; |
| 437 | } |
| 438 | |
| 439 | /** |
| 440 | * Returns the end of the regex for matches |
| 441 | * |
| 442 | * @param int $batchSize |
| 443 | * @return string |
| 444 | */ |
| 445 | public function getRegexEnd( $batchSize ) { |
| 446 | return ( $batchSize > 0 ) ? '/Sim' : '/im'; |
| 447 | } |
| 448 | |
| 449 | /** |
| 450 | * @param Title $title |
| 451 | * @param string[] $entries |
| 452 | * @param User $user |
| 453 | */ |
| 454 | public function warmCachesForFilter( Title $title, array $entries, User $user ) { |
| 455 | // subclass this |
| 456 | } |
| 457 | } |