Duplicate code is one of the most pungent code smells. A rule that is often used is to re-structure code once it is duplicated in three or more places.
Common duplication problems, and corresponding solutions are:
Complex classes like MediawikiApi often do a lot of different things. To break such a class down, we need to identify a cohesive component within that class. A common approach to find such a component is to look for fields/methods that share the same prefixes, or suffixes. You can also have a look at the cohesion graph to spot any un-connected, or weakly-connected components.
Once you have determined the fields that belong together, you can apply the Extract Class refactoring. If the component makes sense as a sub-class, Extract Subclass is also a candidate, and is often faster.
While breaking up the class, it is a good idea to analyze how other classes use MediawikiApi, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 27 | class MediawikiApi implements MediawikiApiInterface, LoggerAwareInterface { |
||
| 28 | |||
| 29 | /** |
||
| 30 | * Should be accessed through getClient |
||
| 31 | * @var ClientInterface|null |
||
| 32 | */ |
||
| 33 | private ?ClientInterface $client = null; |
||
|
|
|||
| 34 | |||
| 35 | /** |
||
| 36 | * @var bool|string |
||
| 37 | */ |
||
| 38 | private $isLoggedIn; |
||
| 39 | |||
| 40 | private MediawikiSession $session; |
||
| 41 | |||
| 42 | /** |
||
| 43 | * @var string |
||
| 44 | */ |
||
| 45 | private $version; |
||
| 46 | |||
| 47 | private LoggerInterface $logger; |
||
| 48 | |||
| 49 | private string $apiUrl; |
||
| 50 | |||
| 51 | /** |
||
| 52 | * @since 2.0 |
||
| 53 | * |
||
| 54 | * @param string $apiEndpoint e.g. https://en.wikipedia.org/w/api.php |
||
| 55 | * |
||
| 56 | * @return self returns a MediawikiApi instance using $apiEndpoint |
||
| 57 | */ |
||
| 58 | public static function newFromApiEndpoint( string $apiEndpoint ): MediawikiApi { |
||
| 59 | return new self( $apiEndpoint ); |
||
| 60 | } |
||
| 61 | |||
| 62 | /** |
||
| 63 | * Create a new MediawikiApi object from a URL to any page in a MediaWiki website. |
||
| 64 | * |
||
| 65 | * @since 2.0 |
||
| 66 | * @see https://en.wikipedia.org/wiki/Really_Simple_Discovery |
||
| 67 | * |
||
| 68 | * @param string $url e.g. https://en.wikipedia.org OR https://de.wikipedia.org/wiki/Berlin |
||
| 69 | * @return self returns a MediawikiApi instance using the apiEndpoint provided by the RSD |
||
| 70 | * file accessible on all Mediawiki pages |
||
| 71 | * @throws RsdException If the RSD URL could not be found in the page's HTML. |
||
| 72 | */ |
||
| 73 | public static function newFromPage( string $url ): MediawikiApi { |
||
| 74 | // Set up HTTP client and HTML document. |
||
| 75 | $tempClient = new Client( [ 'headers' => [ 'User-Agent' => 'addwiki-mediawiki-client' ] ] ); |
||
| 76 | $pageHtml = $tempClient->get( $url )->getBody(); |
||
| 77 | $pageDoc = new DOMDocument(); |
||
| 78 | |||
| 79 | // Try to load the HTML (turn off errors temporarily; most don't matter, and if they do get |
||
| 80 | // in the way of finding the API URL, will be reported in the RsdException below). |
||
| 81 | $internalErrors = libxml_use_internal_errors( true ); |
||
| 82 | $pageDoc->loadHTML( $pageHtml ); |
||
| 83 | $libXmlErrors = libxml_get_errors(); |
||
| 84 | libxml_use_internal_errors( $internalErrors ); |
||
| 85 | |||
| 86 | // Extract the RSD link. |
||
| 87 | $xpath = 'head/link[@type="application/rsd+xml"][@href]'; |
||
| 88 | $link = ( new DOMXpath( $pageDoc ) )->query( $xpath ); |
||
| 89 | if ( $link->length === 0 ) { |
||
| 90 | // Format libxml errors for display. |
||
| 91 | $libXmlErrorStr = array_reduce( $libXmlErrors, fn( $prevErr, $err ) => $prevErr . ', ' . $err->message . ' (line ' . $err->line . ')' ); |
||
| 92 | if ( $libXmlErrorStr ) { |
||
| 93 | $libXmlErrorStr = sprintf( 'In addition, libxml had the following errors: %s', $libXmlErrorStr ); |
||
| 94 | } |
||
| 95 | throw new RsdException( sprintf( 'Unable to find RSD URL in page: %s %s', $url, $libXmlErrorStr ) ); |
||
| 96 | } |
||
| 97 | $linkItem = $link->item( 0 ); |
||
| 98 | if ( ( $linkItem->attributes ) === null ) { |
||
| 99 | throw new RsdException( 'Unexpected RSD fetch error' ); |
||
| 100 | } |
||
| 101 | /** @psalm-suppress NullReference */ |
||
| 102 | $rsdUrl = $linkItem->attributes->getNamedItem( 'href' )->nodeValue; |
||
| 103 | |||
| 104 | // Then get the RSD XML, and return the API link. |
||
| 105 | $rsdXml = new SimpleXMLElement( $tempClient->get( $rsdUrl )->getBody() ); |
||
| 106 | return self::newFromApiEndpoint( (string)$rsdXml->service->apis->api->attributes()->apiLink ); |
||
| 107 | } |
||
| 108 | |||
| 109 | /** |
||
| 110 | * @param string $apiUrl The API Url |
||
| 111 | * @param ClientInterface|null $client Guzzle Client |
||
| 112 | * @param MediawikiSession|null $session Inject a custom session here |
||
| 113 | */ |
||
| 114 | public function __construct( string $apiUrl, ClientInterface $client = null, |
||
| 115 | MediawikiSession $session = null ) { |
||
| 116 | if ( !is_string( $apiUrl ) ) { |
||
| 117 | throw new InvalidArgumentException( '$apiUrl must be a string' ); |
||
| 118 | } |
||
| 119 | if ( $session === null ) { |
||
| 120 | $session = new MediawikiSession( $this ); |
||
| 121 | } |
||
| 122 | |||
| 123 | $this->apiUrl = $apiUrl; |
||
| 124 | $this->client = $client; |
||
| 125 | $this->session = $session; |
||
| 126 | |||
| 127 | $this->logger = new NullLogger(); |
||
| 128 | } |
||
| 129 | |||
| 130 | /** |
||
| 131 | * Get the API URL (the URL to which API requests are sent, usually ending in api.php). |
||
| 132 | * This is useful if you've created this object via MediawikiApi::newFromPage(). |
||
| 133 | * |
||
| 134 | * @since 2.3 |
||
| 135 | * |
||
| 136 | * @return string The API URL. |
||
| 137 | */ |
||
| 138 | public function getApiUrl(): string { |
||
| 139 | return $this->apiUrl; |
||
| 140 | } |
||
| 141 | |||
| 142 | private function getClient(): ClientInterface { |
||
| 143 | if ( !$this->client instanceof ClientInterface ) { |
||
| 144 | $clientFactory = new ClientFactory(); |
||
| 145 | $clientFactory->setLogger( $this->logger ); |
||
| 146 | $this->client = $clientFactory->getClient(); |
||
| 147 | } |
||
| 148 | return $this->client; |
||
| 149 | } |
||
| 150 | |||
| 151 | /** |
||
| 152 | * Sets a logger instance on the object |
||
| 153 | * |
||
| 154 | * @since 1.1 |
||
| 155 | * |
||
| 156 | * @param LoggerInterface $logger The new Logger object. |
||
| 157 | * |
||
| 158 | * @return null |
||
| 159 | */ |
||
| 160 | public function setLogger( LoggerInterface $logger ) { |
||
| 161 | $this->logger = $logger; |
||
| 162 | $this->session->setLogger( $logger ); |
||
| 163 | } |
||
| 164 | |||
| 165 | /** |
||
| 166 | * @since 2.0 |
||
| 167 | * |
||
| 168 | * @param Request $request The GET request to send. |
||
| 169 | * |
||
| 170 | * @return PromiseInterface |
||
| 171 | * Normally promising an array, though can be mixed (json_decode result) |
||
| 172 | * Can throw UsageExceptions or RejectionExceptions |
||
| 173 | */ |
||
| 174 | public function getRequestAsync( Request $request ): PromiseInterface { |
||
| 175 | $promise = $this->getClient()->requestAsync( |
||
| 176 | 'GET', |
||
| 177 | $this->apiUrl, |
||
| 178 | $this->getClientRequestOptions( $request, 'query' ) |
||
| 179 | ); |
||
| 180 | |||
| 181 | return $promise->then( fn( ResponseInterface $response ) => call_user_func( fn( ResponseInterface $response ) => $this->decodeResponse( $response ), $response ) ); |
||
| 182 | } |
||
| 183 | |||
| 184 | /** |
||
| 185 | * @since 2.0 |
||
| 186 | * |
||
| 187 | * @param Request $request The POST request to send. |
||
| 188 | * |
||
| 189 | * @return PromiseInterface |
||
| 190 | * Normally promising an array, though can be mixed (json_decode result) |
||
| 191 | * Can throw UsageExceptions or RejectionExceptions |
||
| 192 | */ |
||
| 193 | public function postRequestAsync( Request $request ): PromiseInterface { |
||
| 194 | $promise = $this->getClient()->requestAsync( |
||
| 195 | 'POST', |
||
| 196 | $this->apiUrl, |
||
| 197 | $this->getClientRequestOptions( $request, $this->getPostRequestEncoding( $request ) ) |
||
| 198 | ); |
||
| 199 | |||
| 200 | return $promise->then( fn( ResponseInterface $response ) => call_user_func( fn( ResponseInterface $response ) => $this->decodeResponse( $response ), $response ) ); |
||
| 201 | } |
||
| 202 | |||
| 203 | /** |
||
| 204 | * @since 0.2 |
||
| 205 | * |
||
| 206 | * @param Request $request The GET request to send. |
||
| 207 | * |
||
| 208 | * @return mixed Normally an array |
||
| 209 | */ |
||
| 210 | public function getRequest( Request $request ) { |
||
| 211 | $response = $this->getClient()->request( |
||
| 212 | 'GET', |
||
| 213 | $this->apiUrl, |
||
| 214 | $this->getClientRequestOptions( $request, 'query' ) |
||
| 215 | ); |
||
| 216 | |||
| 217 | return $this->decodeResponse( $response ); |
||
| 218 | } |
||
| 219 | |||
| 220 | /** |
||
| 221 | * @since 0.2 |
||
| 222 | * |
||
| 223 | * @param Request $request The POST request to send. |
||
| 224 | * |
||
| 225 | * @return mixed Normally an array |
||
| 226 | */ |
||
| 227 | public function postRequest( Request $request ) { |
||
| 228 | $response = $this->getClient()->request( |
||
| 229 | 'POST', |
||
| 230 | $this->apiUrl, |
||
| 231 | $this->getClientRequestOptions( $request, $this->getPostRequestEncoding( $request ) ) |
||
| 232 | ); |
||
| 233 | |||
| 234 | return $this->decodeResponse( $response ); |
||
| 235 | } |
||
| 236 | |||
| 237 | /** |
||
| 238 | * @param ResponseInterface $response |
||
| 239 | * |
||
| 240 | * @return mixed |
||
| 241 | * @throws UsageException |
||
| 242 | */ |
||
| 243 | private function decodeResponse( ResponseInterface $response ) { |
||
| 244 | $resultArray = json_decode( $response->getBody(), true ); |
||
| 245 | |||
| 246 | $this->logWarnings( $resultArray ); |
||
| 247 | $this->throwUsageExceptions( $resultArray ); |
||
| 248 | |||
| 249 | return $resultArray; |
||
| 250 | } |
||
| 251 | |||
| 252 | /** |
||
| 253 | * @param Request $request |
||
| 254 | */ |
||
| 255 | private function getPostRequestEncoding( Request $request ): string { |
||
| 256 | if ( $request instanceof MultipartRequest ) { |
||
| 257 | return 'multipart'; |
||
| 258 | } |
||
| 259 | foreach ( $request->getParams() as $value ) { |
||
| 260 | if ( is_resource( $value ) ) { |
||
| 261 | return 'multipart'; |
||
| 262 | } |
||
| 263 | } |
||
| 264 | return 'form_params'; |
||
| 265 | } |
||
| 266 | |||
| 267 | /** |
||
| 268 | * @param string $paramsKey either 'query' or 'multipart' |
||
| 269 | * |
||
| 270 | * @throws RequestException |
||
| 271 | * @return array as needed by ClientInterface::get and ClientInterface::post |
||
| 272 | */ |
||
| 273 | private function getClientRequestOptions( Request $request, string $paramsKey ): array { |
||
| 274 | $params = array_merge( $request->getParams(), [ 'format' => 'json' ] ); |
||
| 275 | if ( $paramsKey === 'multipart' ) { |
||
| 276 | $params = $this->encodeMultipartParams( $request, $params ); |
||
| 277 | } |
||
| 278 | |||
| 279 | return [ |
||
| 280 | $paramsKey => $params, |
||
| 281 | 'headers' => array_merge( $this->getDefaultHeaders(), $request->getHeaders() ), |
||
| 282 | ]; |
||
| 283 | } |
||
| 284 | |||
| 285 | /** |
||
| 286 | * Turn the normal key-value array of request parameters into a multipart array where each |
||
| 287 | * parameter is a new array with a 'name' and 'contents' elements (and optionally more, if the |
||
| 288 | * request is a MultipartRequest). |
||
| 289 | * |
||
| 290 | * @param Request $request The request to which the parameters belong. |
||
| 291 | * @param string[] $params The existing parameters. Not the same as $request->getParams(). |
||
| 292 | * |
||
| 293 | * @return array <int mixed[]> |
||
| 294 | */ |
||
| 295 | private function encodeMultipartParams( Request $request, array $params ): array { |
||
| 296 | // See if there are any multipart parameters in this request. |
||
| 297 | $multipartParams = ( $request instanceof MultipartRequest ) |
||
| 298 | ? $request->getMultipartParams() |
||
| 299 | : []; |
||
| 300 | return array_map( |
||
| 301 | function ( $name, $value ) use ( $multipartParams ): array { |
||
| 302 | $partParams = [ |
||
| 303 | 'name' => $name, |
||
| 304 | 'contents' => $value, |
||
| 305 | ]; |
||
| 306 | if ( isset( $multipartParams[ $name ] ) ) { |
||
| 307 | // If extra parameters have been set for this part, use them. |
||
| 308 | $partParams = array_merge( $multipartParams[ $name ], $partParams ); |
||
| 309 | } |
||
| 310 | return $partParams; |
||
| 311 | }, |
||
| 312 | array_keys( $params ), |
||
| 313 | $params |
||
| 314 | ); |
||
| 315 | } |
||
| 316 | |||
| 317 | /** |
||
| 318 | * @return array <string mixed> |
||
| 319 | */ |
||
| 320 | private function getDefaultHeaders(): array { |
||
| 321 | return [ |
||
| 322 | 'User-Agent' => $this->getUserAgent(), |
||
| 323 | ]; |
||
| 324 | } |
||
| 325 | |||
| 326 | private function getUserAgent(): string { |
||
| 327 | $loggedIn = $this->isLoggedin(); |
||
| 328 | if ( $loggedIn ) { |
||
| 329 | return 'addwiki-mediawiki-client/' . $loggedIn; |
||
| 330 | } |
||
| 331 | return 'addwiki-mediawiki-client'; |
||
| 332 | } |
||
| 333 | |||
| 334 | private function logWarnings( $result ): void { |
||
| 335 | if ( is_array( $result ) ) { |
||
| 336 | // Let's see if there is 'warnings' key on the first level of the array... |
||
| 337 | if ( $this->logWarning( $result ) ) { |
||
| 338 | return; |
||
| 339 | } |
||
| 340 | |||
| 341 | // ...if no then go one level deeper and check there for it. |
||
| 342 | foreach ( $result as $value ) { |
||
| 343 | if ( !is_array( $value ) ) { |
||
| 344 | continue; |
||
| 345 | } |
||
| 346 | |||
| 347 | $this->logWarning( $value ); |
||
| 348 | } |
||
| 349 | } |
||
| 350 | } |
||
| 351 | |||
| 352 | /** |
||
| 353 | * @param array $array Array response to look for warning in. |
||
| 354 | * |
||
| 355 | * @return bool Whether any warning has been logged or not. |
||
| 356 | */ |
||
| 357 | protected function logWarning( array $array ): bool { |
||
| 358 | $found = false; |
||
| 359 | |||
| 360 | if ( !array_key_exists( 'warnings', $array ) ) { |
||
| 361 | return false; |
||
| 362 | } |
||
| 363 | |||
| 364 | foreach ( $array['warnings'] as $module => $warningData ) { |
||
| 365 | // Accommodate both formatversion=2 and old-style API results |
||
| 366 | $logPrefix = $module . ': '; |
||
| 367 | if ( isset( $warningData['*'] ) ) { |
||
| 368 | $this->logger->warning( $logPrefix . $warningData['*'], [ 'data' => $warningData ] ); |
||
| 369 | } elseif ( isset( $warningData['warnings'] ) ) { |
||
| 370 | $this->logger->warning( $logPrefix . $warningData['warnings'], [ 'data' => $warningData ] ); |
||
| 371 | } else { |
||
| 372 | $this->logger->warning( $logPrefix, [ 'data' => $warningData ] ); |
||
| 373 | } |
||
| 374 | |||
| 375 | $found = true; |
||
| 376 | } |
||
| 377 | |||
| 378 | return $found; |
||
| 379 | } |
||
| 380 | |||
| 381 | /** |
||
| 382 | * @throws UsageException |
||
| 383 | */ |
||
| 384 | private function throwUsageExceptions( $result ): void { |
||
| 385 | if ( is_array( $result ) && array_key_exists( 'error', $result ) ) { |
||
| 386 | throw new UsageException( |
||
| 387 | $result['error']['code'], |
||
| 388 | $result['error']['info'], |
||
| 389 | $result |
||
| 390 | ); |
||
| 391 | } |
||
| 392 | } |
||
| 393 | |||
| 394 | /** |
||
| 395 | * @since 0.1 |
||
| 396 | * |
||
| 397 | * @return bool|string false or the name of the current user |
||
| 398 | */ |
||
| 399 | public function isLoggedin() { |
||
| 400 | return $this->isLoggedIn; |
||
| 401 | } |
||
| 402 | |||
| 403 | /** |
||
| 404 | * @since 0.1 |
||
| 405 | * |
||
| 406 | * @param ApiUser $apiUser The ApiUser to log in as. |
||
| 407 | * |
||
| 408 | * @throws UsageException |
||
| 409 | * @return bool success |
||
| 410 | */ |
||
| 411 | public function login( ApiUser $apiUser ): bool { |
||
| 412 | $this->logger->log( LogLevel::DEBUG, 'Logging in' ); |
||
| 413 | $credentials = $this->getLoginParams( $apiUser ); |
||
| 414 | $result = $this->postRequest( new SimpleRequest( 'login', $credentials ) ); |
||
| 415 | if ( $result['login']['result'] == "NeedToken" ) { |
||
| 416 | $params = array_merge( [ 'lgtoken' => $result['login']['token'] ], $credentials ); |
||
| 417 | $result = $this->postRequest( new SimpleRequest( 'login', $params ) ); |
||
| 418 | } |
||
| 419 | if ( $result['login']['result'] == "Success" ) { |
||
| 420 | $this->isLoggedIn = $apiUser->getUsername(); |
||
| 421 | return true; |
||
| 422 | } |
||
| 423 | |||
| 424 | $this->isLoggedIn = false; |
||
| 425 | $this->logger->log( LogLevel::DEBUG, 'Login failed.', $result ); |
||
| 426 | $this->throwLoginUsageException( $result ); |
||
| 427 | return false; |
||
| 428 | } |
||
| 429 | |||
| 430 | /** |
||
| 431 | * @param ApiUser $apiUser |
||
| 432 | * |
||
| 433 | * @return string[] |
||
| 434 | */ |
||
| 435 | private function getLoginParams( ApiUser $apiUser ): array { |
||
| 436 | $params = [ |
||
| 437 | 'lgname' => $apiUser->getUsername(), |
||
| 438 | 'lgpassword' => $apiUser->getPassword(), |
||
| 439 | ]; |
||
| 440 | |||
| 441 | if ( $apiUser->getDomain() !== null ) { |
||
| 442 | $params['lgdomain'] = $apiUser->getDomain(); |
||
| 443 | } |
||
| 444 | return $params; |
||
| 445 | } |
||
| 446 | |||
| 447 | /** |
||
| 448 | * |
||
| 449 | * @throws UsageException |
||
| 450 | */ |
||
| 451 | private function throwLoginUsageException( array $result ): void { |
||
| 452 | $loginResult = $result['login']['result']; |
||
| 453 | |||
| 454 | throw new UsageException( |
||
| 455 | 'login-' . $loginResult, |
||
| 456 | array_key_exists( 'reason', $result['login'] ) |
||
| 457 | ? $result['login']['reason'] |
||
| 458 | : 'No Reason given', |
||
| 459 | $result |
||
| 460 | ); |
||
| 461 | } |
||
| 462 | |||
| 463 | /** |
||
| 464 | * @since 0.1 |
||
| 465 | * |
||
| 466 | * @return bool success |
||
| 467 | */ |
||
| 468 | public function logout(): bool { |
||
| 469 | $this->logger->log( LogLevel::DEBUG, 'Logging out' ); |
||
| 470 | $result = $this->postRequest( new SimpleRequest( 'logout', [ |
||
| 471 | 'token' => $this->getToken() |
||
| 472 | ] ) ); |
||
| 473 | if ( $result === [] ) { |
||
| 474 | $this->isLoggedIn = false; |
||
| 475 | $this->clearTokens(); |
||
| 476 | return true; |
||
| 477 | } |
||
| 478 | return false; |
||
| 479 | } |
||
| 480 | |||
| 481 | /** |
||
| 482 | * @since 0.1 |
||
| 483 | * |
||
| 484 | * @param string $type The token type to get. |
||
| 485 | * |
||
| 486 | * @return string |
||
| 487 | */ |
||
| 488 | public function getToken( $type = 'csrf' ): string { |
||
| 489 | return $this->session->getToken( $type ); |
||
| 490 | } |
||
| 491 | |||
| 492 | /** |
||
| 493 | * Clear all tokens stored by the API. |
||
| 494 | * |
||
| 495 | * @since 0.1 |
||
| 496 | */ |
||
| 497 | public function clearTokens() { |
||
| 498 | $this->session->clearTokens(); |
||
| 499 | } |
||
| 500 | |||
| 501 | /** |
||
| 502 | * @return string |
||
| 503 | */ |
||
| 504 | public function getVersion(): string { |
||
| 505 | if ( $this->version === null ) { |
||
| 506 | $result = $this->getRequest( new SimpleRequest( 'query', [ |
||
| 507 | 'meta' => 'siteinfo', |
||
| 508 | 'continue' => '', |
||
| 509 | ] ) ); |
||
| 510 | preg_match( |
||
| 511 | '#\d+(?:\.\d+)+#', |
||
| 512 | $result['query']['general']['generator'], |
||
| 513 | $versionParts |
||
| 514 | ); |
||
| 515 | $this->version = $versionParts[0]; |
||
| 516 | } |
||
| 517 | return $this->version; |
||
| 518 | } |
||
| 519 | |||
| 520 | } |
||
| 521 |