Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
2.34% |
5 / 214 |
|
15.38% |
2 / 13 |
CRAP | |
0.00% |
0 / 1 |
| LiftWingService | |
2.34% |
5 / 214 |
|
15.38% |
2 / 13 |
2473.91 | |
0.00% |
0 / 1 |
| __construct | |
0.00% |
0 / 2 |
|
0.00% |
0 / 1 |
2 | |||
| getBaseUrl | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| getFrontendBaseUrl | |
0.00% |
0 / 3 |
|
0.00% |
0 / 1 |
6 | |||
| getUrl | |
100.00% |
4 / 4 |
|
100.00% |
1 / 1 |
1 | |||
| request | |
0.00% |
0 / 15 |
|
0.00% |
0 / 1 |
42 | |||
| singleLiftWingRequest | |
0.00% |
0 / 37 |
|
0.00% |
0 / 1 |
110 | |||
| revertRiskLiftWingRequest | |
0.00% |
0 / 54 |
|
0.00% |
0 / 1 |
240 | |||
| prepareRevertriskRequest | |
0.00% |
0 / 8 |
|
0.00% |
0 / 1 |
6 | |||
| createHostHeader | |
0.00% |
0 / 13 |
|
0.00% |
0 / 1 |
2 | |||
| parseLiftWingResults | |
0.00% |
0 / 12 |
|
0.00% |
0 / 1 |
72 | |||
| createRevisionNotFoundResponse | |
0.00% |
0 / 21 |
|
0.00% |
0 / 1 |
2 | |||
| createRevisionNotScorableResponse | |
0.00% |
0 / 22 |
|
0.00% |
0 / 1 |
2 | |||
| modifyRevertRiskResponse | |
0.00% |
0 / 22 |
|
0.00% |
0 / 1 |
6 | |||
| 1 | <?php |
| 2 | /** |
| 3 | * This program is free software: you can redistribute it and/or modify |
| 4 | * it under the terms of the GNU General Public License as published by |
| 5 | * the Free Software Foundation, either version 3 of the License, or |
| 6 | * (at your option) any later version. |
| 7 | * |
| 8 | * This program is distributed in the hope that it will be useful, |
| 9 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 10 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 11 | * GNU General Public License for more details. |
| 12 | * |
| 13 | * You should have received a copy of the GNU General Public License |
| 14 | * along with this program. If not, see <http://www.gnu.org/licenses/>. |
| 15 | */ |
| 16 | |
| 17 | namespace ORES; |
| 18 | |
| 19 | use MediaWiki\Config\Config; |
| 20 | use MediaWiki\Http\HttpRequestFactory; |
| 21 | use MediaWiki\Json\FormatJson; |
| 22 | use MediaWiki\MediaWikiServices; |
| 23 | use MediaWiki\Status\Status; |
| 24 | use Psr\Log\LoggerInterface; |
| 25 | use RuntimeException; |
| 26 | |
| 27 | /** |
| 28 | * Common methods for accessing a Lift Wing server. |
| 29 | */ |
| 30 | class LiftWingService extends ORESService { |
| 31 | |
| 32 | public const API_VERSION = 1; |
| 33 | private Config $config; |
| 34 | |
| 35 | public function __construct( |
| 36 | LoggerInterface $logger, HttpRequestFactory $httpRequestFactory, Config $config |
| 37 | ) { |
| 38 | parent::__construct( $logger, $httpRequestFactory ); |
| 39 | $this->config = $config; |
| 40 | } |
| 41 | |
| 42 | /** |
| 43 | * @return string Base URL of ORES service |
| 44 | */ |
| 45 | public static function getBaseUrl() { |
| 46 | global $wgOresLiftWingBaseUrl; |
| 47 | |
| 48 | return $wgOresLiftWingBaseUrl; |
| 49 | } |
| 50 | |
| 51 | /** |
| 52 | * @return string Base URL of ORES service being used externally |
| 53 | */ |
| 54 | public static function getFrontendBaseUrl() { |
| 55 | global $wgOresFrontendBaseUrl, $wgOresLiftWingBaseUrl; |
| 56 | |
| 57 | if ( $wgOresFrontendBaseUrl === null ) { |
| 58 | return $wgOresLiftWingBaseUrl; |
| 59 | } |
| 60 | |
| 61 | return $wgOresFrontendBaseUrl; |
| 62 | } |
| 63 | |
| 64 | /** |
| 65 | * @param string|null $model |
| 66 | * @return string Base URL plus your wiki's `scores` API path. |
| 67 | */ |
| 68 | public function getUrl( $model = null ) { |
| 69 | $wikiId = self::getWikiID(); |
| 70 | $prefix = 'v' . self::API_VERSION; |
| 71 | $baseUrl = self::getBaseUrl(); |
| 72 | return "{$baseUrl}{$prefix}/models/{$wikiId}-{$model}:predict"; |
| 73 | } |
| 74 | |
| 75 | /** |
| 76 | * Make an ORES API request and return the decoded result. |
| 77 | * |
| 78 | * @param array $params |
| 79 | * @param array|null $originalRequest |
| 80 | * |
| 81 | * @return array Decoded response |
| 82 | */ |
| 83 | public function request( array $params, $originalRequest = null ) { |
| 84 | if ( !isset( $params['models'] ) ) { |
| 85 | throw new RuntimeException( 'Missing required parameter: models' ); |
| 86 | } |
| 87 | if ( !isset( $params['revids'] ) ) { |
| 88 | throw new RuntimeException( 'Missing required parameter: revids' ); |
| 89 | } |
| 90 | |
| 91 | $models = explode( '|', $params['models'] ); |
| 92 | $revids = explode( '|', $params['revids'] ); |
| 93 | |
| 94 | $responses = []; |
| 95 | |
| 96 | foreach ( $models as $model ) { |
| 97 | foreach ( $revids as $revid ) { |
| 98 | if ( $model == 'revertrisklanguageagnostic' ) { |
| 99 | // This is a hack to convert the revert-risk response to an ORES response in order to be compatible |
| 100 | // with the rest of the model responses in order for them to be easily merged together in one |
| 101 | // response. |
| 102 | $response = $this->revertRiskLiftWingRequest( $model, $revid ); |
| 103 | } else { |
| 104 | $response = $this->singleLiftWingRequest( $model, $revid ); |
| 105 | } |
| 106 | |
| 107 | $responses[] = $response; |
| 108 | } |
| 109 | } |
| 110 | $wikiId = self::getWikiID(); |
| 111 | return $this->parseLiftWingResults( $wikiId, $responses ); |
| 112 | } |
| 113 | |
| 114 | /** |
| 115 | * Make a single call to LW for one revid and one model and return the decoded result. |
| 116 | * |
| 117 | * @param string $model |
| 118 | * @param string $revid |
| 119 | * |
| 120 | * @return array Decoded response |
| 121 | */ |
| 122 | public function singleLiftWingRequest( $model, $revid ) { |
| 123 | $url = $this->getUrl( $model ); |
| 124 | $this->logger->debug( "Requesting: {$url}" ); |
| 125 | |
| 126 | $req = $this->httpRequestFactory->create( $url, [ |
| 127 | 'method' => 'POST', |
| 128 | 'postData' => json_encode( [ 'rev_id' => (int)$revid ] ), |
| 129 | ], __METHOD__ ); |
| 130 | global $wgOresLiftWingAddHostHeader; |
| 131 | if ( $wgOresLiftWingAddHostHeader ) { |
| 132 | $req->setHeader( 'Content-Type', 'application/json' ); |
| 133 | $req->setHeader( 'Host', self::createHostHeader( $model ) ); |
| 134 | } |
| 135 | $status = $req->execute(); |
| 136 | if ( !$status->isOK() ) { |
| 137 | $message = "Failed to make LiftWing request to [{$url}], " . |
| 138 | Status::wrap( $status )->getMessage()->inLanguage( 'en' )->text(); |
| 139 | |
| 140 | // Server time out, try again once |
| 141 | if ( $req->getStatus() === 504 ) { |
| 142 | $req = $this->httpRequestFactory->create( $url, [ |
| 143 | 'method' => 'POST', |
| 144 | 'postData' => json_encode( [ 'rev_id' => (int)$revid ] ), |
| 145 | ], __METHOD__ ); |
| 146 | if ( $wgOresLiftWingAddHostHeader ) { |
| 147 | $req->setHeader( 'Content-Type', 'application/json' ); |
| 148 | $req->setHeader( 'Host', self::createHostHeader( $model ) ); |
| 149 | } |
| 150 | |
| 151 | $status = $req->execute(); |
| 152 | if ( !$status->isOK() ) { |
| 153 | throw new RuntimeException( $message ); |
| 154 | } |
| 155 | } elseif ( $req->getStatus() === 400 ) { |
| 156 | $this->logger->debug( "400 Bad Request: {$message}" ); |
| 157 | $data = FormatJson::decode( $req->getContent(), true ); |
| 158 | if ( strpos( $data["error"], "The MW API does not have any info related to the rev-id" ) === 0 ) { |
| 159 | return $this->createRevisionNotFoundResponse( $model, $revid ); |
| 160 | } else { |
| 161 | throw new RuntimeException( $message ); |
| 162 | } |
| 163 | } else { |
| 164 | throw new RuntimeException( $message ); |
| 165 | } |
| 166 | } |
| 167 | $json = $req->getContent(); |
| 168 | $this->logger->debug( "Raw response: {$json}" ); |
| 169 | $data = FormatJson::decode( $json, true ); |
| 170 | if ( !$data || !empty( $data['error'] ) ) { |
| 171 | throw new RuntimeException( "Bad response from Lift Wing endpoint [{$url}]: {$json}" ); |
| 172 | } |
| 173 | return $data; |
| 174 | } |
| 175 | |
| 176 | /** |
| 177 | * @param string $model |
| 178 | * @param string|int $revid |
| 179 | * @return array|array[]|mixed |
| 180 | */ |
| 181 | public function revertRiskLiftWingRequest( $model, $revid ) { |
| 182 | $revLookup = MediaWikiServices::getInstance()->getRevisionLookup(); |
| 183 | $revidInt = (int)$revid; |
| 184 | $revidStr = (string)$revid; |
| 185 | // The Language-agnostic revert risk model card states that |
| 186 | // the first or only revision on a page must be excluded. |
| 187 | if ( $revidInt === 0 ) { |
| 188 | return $this->createRevisionNotScorableResponse( |
| 189 | $model, |
| 190 | $revidStr |
| 191 | ); |
| 192 | } |
| 193 | $rev = $revLookup->getRevisionById( $revidInt ); |
| 194 | if ( $rev === null ) { |
| 195 | return $this->createRevisionNotFoundResponse( |
| 196 | $model, |
| 197 | $revidStr |
| 198 | ); |
| 199 | } |
| 200 | $parentId = $rev->getParentId(); |
| 201 | if ( $parentId === null || $parentId === 0 ) { |
| 202 | return $this->createRevisionNotScorableResponse( |
| 203 | $model, |
| 204 | $revidStr |
| 205 | ); |
| 206 | } |
| 207 | $wikiId = self::getWikiID(); |
| 208 | $language = substr( $wikiId, 0, strpos( $wikiId, "wiki" ) ); |
| 209 | $prefix = 'v' . self::API_VERSION; |
| 210 | $baseUrl = self::getBaseUrl(); |
| 211 | $url = "{$baseUrl}{$prefix}/models/revertrisk-language-agnostic:predict"; |
| 212 | |
| 213 | $this->logger->debug( "Requesting: {$url}" ); |
| 214 | |
| 215 | $req = $this->prepareRevertriskRequest( $url, $revid, $language ); |
| 216 | $status = $req->execute(); |
| 217 | if ( !$status->isOK() ) { |
| 218 | $message = "Failed to make LiftWing request to [{$url}], " . |
| 219 | Status::wrap( $status )->getMessage()->inLanguage( 'en' )->text(); |
| 220 | |
| 221 | // Server time out, try again once |
| 222 | if ( $req->getStatus() === 504 ) { |
| 223 | $req = $this->prepareRevertriskRequest( $url, $revid, $language ); |
| 224 | $status = $req->execute(); |
| 225 | if ( !$status->isOK() ) { |
| 226 | throw new RuntimeException( $message ); |
| 227 | } |
| 228 | } elseif ( $req->getStatus() === 400 ) { |
| 229 | $this->logger->debug( "400 Bad Request: {$message}" ); |
| 230 | $data = FormatJson::decode( $req->getContent(), true ); |
| 231 | // This is a way to detect whether an error occurred because the revision was not found. |
| 232 | // Checking for "detail" key in the response is a hack to detect the error message from Lift Wing as |
| 233 | // there is no universal error handling/messaging at the moment. |
| 234 | if ( ( $data && isset( $data['error'] ) && strpos( $data["error"], |
| 235 | "The MW API does not have any info related to the rev-id" ) === 0 ) || |
| 236 | array_key_exists( "detail", $data ) ) { |
| 237 | return $this->createRevisionNotFoundResponse( |
| 238 | $model, |
| 239 | $revidStr |
| 240 | ); |
| 241 | } else { |
| 242 | throw new RuntimeException( $message ); |
| 243 | } |
| 244 | } else { |
| 245 | throw new RuntimeException( $message ); |
| 246 | } |
| 247 | } |
| 248 | $json = $req->getContent(); |
| 249 | $this->logger->debug( "Raw response: {$json}" ); |
| 250 | $data = FormatJson::decode( $json, true ); |
| 251 | if ( !$data || !empty( $data['error'] ) ) { |
| 252 | throw new RuntimeException( "Bad response from Lift Wing endpoint [{$url}]: {$json}" ); |
| 253 | } |
| 254 | return $this->modifyRevertRiskResponse( $data ); |
| 255 | } |
| 256 | |
| 257 | private function prepareRevertriskRequest( string $url, string $revid, string $language ): \MWHttpRequest { |
| 258 | $req = $this->httpRequestFactory->create( $url, [ |
| 259 | 'method' => 'POST', |
| 260 | 'postData' => json_encode( [ 'rev_id' => (int)$revid, 'lang' => $language ] ), |
| 261 | ], __METHOD__ ); |
| 262 | $req->setHeader( 'Content-Type', 'application/json' ); |
| 263 | global $wgOresLiftWingAddHostHeader; |
| 264 | if ( $wgOresLiftWingAddHostHeader ) { |
| 265 | $req->setHeader( 'Host', $this->config->get( 'OresLiftWingRevertRiskHostHeader' ) ); |
| 266 | } |
| 267 | return $req; |
| 268 | } |
| 269 | |
| 270 | /** |
| 271 | * @param string $modelname the model name as requested from ORES |
| 272 | * @return string The hostname required in the header for the Lift Wing call |
| 273 | */ |
| 274 | public static function createHostHeader( string $modelname ) { |
| 275 | $hostnames = [ |
| 276 | 'articlequality' => 'revscoring-articlequality', |
| 277 | 'itemquality' => 'revscoring-articlequality', |
| 278 | 'articletopic' => 'revscoring-articletopic', |
| 279 | 'itemtopic' => 'revscoring-articletopic', |
| 280 | 'draftquality' => 'revscoring-draftquality', |
| 281 | 'drafttopic' => 'revscoring-drafttopic', |
| 282 | 'damaging' => 'revscoring-editquality-damaging', |
| 283 | 'goodfaith' => 'revscoring-editquality-goodfaith', |
| 284 | 'reverted' => 'revscoring-editquality-reverted', |
| 285 | ]; |
| 286 | $wikiID = self::getWikiID(); |
| 287 | return "{$wikiID}-{$modelname}.{$hostnames[$modelname]}.wikimedia.org"; |
| 288 | } |
| 289 | |
| 290 | /** |
| 291 | * This function merges the multiple Lift Wing responses into the appropriate format as returned from ORES |
| 292 | * @param string $context |
| 293 | * @param array $responses |
| 294 | * @return array |
| 295 | */ |
| 296 | private function parseLiftWingResults( string $context, array $responses ): array { |
| 297 | $result = []; |
| 298 | foreach ( $responses as $d ) { |
| 299 | if ( !$d ) { |
| 300 | continue; |
| 301 | } |
| 302 | foreach ( $d[$context] as $k => $v ) { |
| 303 | if ( is_array( $v ) && $k === "scores" ) { |
| 304 | foreach ( $v as $rev_id => $scores ) { |
| 305 | if ( isset( $result[$context][$k][$rev_id] ) ) { |
| 306 | $result[$context][$k][$rev_id] = array_merge( $result[$context][$k][$rev_id], $scores ); |
| 307 | } else { |
| 308 | $result[$context][$k][$rev_id] = $scores; |
| 309 | } |
| 310 | } |
| 311 | } else { |
| 312 | $result[$context][$k] = array_merge( $result[$context][$k] ?? [], $v ); |
| 313 | } |
| 314 | } |
| 315 | } |
| 316 | return $result; |
| 317 | } |
| 318 | |
| 319 | /** |
| 320 | * @param string $model_name |
| 321 | * @param string $rev_id |
| 322 | * @return array |
| 323 | */ |
| 324 | private function createRevisionNotFoundResponse( |
| 325 | string $model_name, |
| 326 | string $rev_id |
| 327 | ) { |
| 328 | global $wgOresModelVersions; |
| 329 | $error_type = "RevisionNotFound"; |
| 330 | $error_message = "{$error_type}: Could not find revision ({revision}:{$rev_id})"; |
| 331 | return [ |
| 332 | self::getWikiID() => [ |
| 333 | "models" => [ |
| 334 | $model_name => [ |
| 335 | "version" => $wgOresModelVersions['models'][$model_name]['version'], |
| 336 | ], |
| 337 | ], |
| 338 | "scores" => [ |
| 339 | $rev_id => [ |
| 340 | $model_name => [ |
| 341 | "error" => [ |
| 342 | "message" => $error_message, |
| 343 | "type" => $error_type, |
| 344 | ], |
| 345 | ], |
| 346 | ], |
| 347 | ], |
| 348 | ], |
| 349 | ]; |
| 350 | } |
| 351 | |
| 352 | /** |
| 353 | * @param string $model_name |
| 354 | * @param string $rev_id |
| 355 | * @return array |
| 356 | */ |
| 357 | private function createRevisionNotScorableResponse( |
| 358 | string $model_name, |
| 359 | string $rev_id |
| 360 | ) { |
| 361 | global $wgOresModelVersions; |
| 362 | $error_type = "RevisionNotScorable"; |
| 363 | $error_message = "{$error_type}: This model may not be used to score ({revision}:{$rev_id})."; |
| 364 | $error_message .= " See Users and uses section of model card."; |
| 365 | return [ |
| 366 | self::getWikiID() => [ |
| 367 | "models" => [ |
| 368 | $model_name => [ |
| 369 | "version" => $wgOresModelVersions['models'][$model_name]['version'], |
| 370 | ], |
| 371 | ], |
| 372 | "scores" => [ |
| 373 | $rev_id => [ |
| 374 | $model_name => [ |
| 375 | "error" => [ |
| 376 | "message" => $error_message, |
| 377 | "type" => $error_type, |
| 378 | ], |
| 379 | ], |
| 380 | ], |
| 381 | ], |
| 382 | ], |
| 383 | ]; |
| 384 | } |
| 385 | |
| 386 | /** |
| 387 | * @param array $response |
| 388 | * @return array[] |
| 389 | */ |
| 390 | private function modifyRevertRiskResponse( array $response ): array { |
| 391 | return [ |
| 392 | $response["wiki_db"] => [ |
| 393 | "models" => [ |
| 394 | "revertrisklanguageagnostic" => [ |
| 395 | "version" => $response["model_version"] |
| 396 | ] |
| 397 | ], |
| 398 | "scores" => [ |
| 399 | (string)$response["revision_id"] => [ |
| 400 | "revertrisklanguageagnostic" => [ |
| 401 | "score" => [ |
| 402 | "prediction" => $response["output"]["prediction"] ? "true" : "false", |
| 403 | "probability" => [ |
| 404 | "true" => $response["output"]["probabilities"]["true"], |
| 405 | "false" => $response["output"]["probabilities"]["false"] |
| 406 | ] |
| 407 | ] |
| 408 | ] |
| 409 | ] |
| 410 | ] |
| 411 | ] |
| 412 | ]; |
| 413 | } |
| 414 | |
| 415 | } |