Code Coverage |
||||||||||
Lines |
Functions and Methods |
Classes and Traits |
||||||||
| Total | |
65.71% |
115 / 175 |
|
41.18% |
7 / 17 |
CRAP | |
0.00% |
0 / 1 |
| WikiSEO | |
65.71% |
115 / 175 |
|
41.18% |
7 / 17 |
171.52 | |
0.00% |
0 / 1 |
| __construct | |
100.00% |
2 / 2 |
|
100.00% |
1 / 1 |
1 | |||
| setMetadataFromPageProps | |
100.00% |
7 / 7 |
|
100.00% |
1 / 1 |
2 | |||
| setMetadata | |
75.00% |
9 / 12 |
|
0.00% |
0 / 1 |
6.56 | |||
| getMetadataValue | |
100.00% |
1 / 1 |
|
100.00% |
1 / 1 |
1 | |||
| addMetadataToPage | |
100.00% |
11 / 11 |
|
100.00% |
1 / 1 |
4 | |||
| setMetadataGenerators | |
0.00% |
0 / 22 |
|
0.00% |
0 / 1 |
12 | |||
| loadPagePropsFromDb | |
63.64% |
7 / 11 |
|
0.00% |
0 / 1 |
4.77 | |||
| loadPagePropsFromOutputPage | |
90.91% |
10 / 11 |
|
0.00% |
0 / 1 |
5.02 | |||
| instantiateMetadataPlugins | |
71.43% |
5 / 7 |
|
0.00% |
0 / 1 |
3.21 | |||
| finalize | |
100.00% |
6 / 6 |
|
100.00% |
1 / 1 |
2 | |||
| makeErrorHtml | |
100.00% |
2 / 2 |
|
100.00% |
1 / 1 |
1 | |||
| modifyPageTitle | |
95.24% |
20 / 21 |
|
0.00% |
0 / 1 |
8 | |||
| saveMetadataToProps | |
91.67% |
11 / 12 |
|
0.00% |
0 / 1 |
5.01 | |||
| mergeValidParameterNames | |
0.00% |
0 / 17 |
|
0.00% |
0 / 1 |
2 | |||
| fromTag | |
81.82% |
9 / 11 |
|
0.00% |
0 / 1 |
2.02 | |||
| fromParserFunction | |
63.16% |
12 / 19 |
|
0.00% |
0 / 1 |
4.80 | |||
| protocolizeUrl | |
100.00% |
3 / 3 |
|
100.00% |
1 / 1 |
2 | |||
| 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 2 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 along |
| 14 | * with this program; if not, write to the Free Software Foundation, Inc., |
| 15 | * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. |
| 16 | * |
| 17 | * @file |
| 18 | */ |
| 19 | |
| 20 | declare( strict_types=1 ); |
| 21 | |
| 22 | namespace MediaWiki\Extension\WikiSEO; |
| 23 | |
| 24 | use ConfigException; |
| 25 | use ExtensionRegistry; |
| 26 | use MediaWiki\Extension\WikiSEO\Generator\GeneratorInterface; |
| 27 | use MediaWiki\Extension\WikiSEO\Generator\MetaTag; |
| 28 | use MediaWiki\MediaWikiServices; |
| 29 | use OutputPage; |
| 30 | use PageImages\PageImages; |
| 31 | use Parser; |
| 32 | use ParserOutput; |
| 33 | use PPFrame; |
| 34 | use ReflectionClass; |
| 35 | use ReflectionException; |
| 36 | use Title; |
| 37 | use WebRequest; |
| 38 | use Wikimedia\AtEase\AtEase; |
| 39 | |
| 40 | class WikiSEO { |
| 41 | private const MODE_TAG = 'tag'; |
| 42 | private const MODE_PARSER = 'parser'; |
| 43 | |
| 44 | /** |
| 45 | * @var string 'tag' or 'parser' used to determine the error message |
| 46 | */ |
| 47 | private $mode; |
| 48 | |
| 49 | /** |
| 50 | * prepend, append or replace the new title to the existing title |
| 51 | * |
| 52 | * @var string |
| 53 | */ |
| 54 | private $titleMode = 'replace'; |
| 55 | |
| 56 | /** |
| 57 | * the separator to use when using append or prepend modes |
| 58 | * |
| 59 | * @var string |
| 60 | */ |
| 61 | private $titleSeparator = ' - '; |
| 62 | |
| 63 | /** |
| 64 | * @var string[] Array with generator names |
| 65 | */ |
| 66 | private $generators; |
| 67 | |
| 68 | /** |
| 69 | * @var GeneratorInterface[] |
| 70 | */ |
| 71 | private $generatorInstances = []; |
| 72 | |
| 73 | /** |
| 74 | * @var string[] Possible error messages |
| 75 | */ |
| 76 | private $errors = []; |
| 77 | |
| 78 | /** |
| 79 | * @var array |
| 80 | */ |
| 81 | private $metadata = []; |
| 82 | |
| 83 | /** |
| 84 | * WikiSEO constructor. |
| 85 | * Loads generator names from LocalSettings |
| 86 | * |
| 87 | * @param string $mode the parser mode |
| 88 | */ |
| 89 | public function __construct( $mode = self::MODE_PARSER ) { |
| 90 | $this->setMetadataGenerators(); |
| 91 | |
| 92 | $this->mode = $mode; |
| 93 | } |
| 94 | |
| 95 | /** |
| 96 | * Set the metadata by loading the page props from the db or the OutputPage object |
| 97 | * |
| 98 | * @param OutputPage $outputPage |
| 99 | */ |
| 100 | public function setMetadataFromPageProps( OutputPage $outputPage ): void { |
| 101 | if ( $outputPage->getTitle() === null ) { |
| 102 | $this->errors[] = wfMessage( 'wiki-seo-missing-page-title' ); |
| 103 | |
| 104 | return; |
| 105 | } |
| 106 | |
| 107 | $result = |
| 108 | $this->loadPagePropsFromDb( $outputPage->getTitle() ) ?? |
| 109 | $this->loadPagePropsFromOutputPage( $outputPage ) ?? []; |
| 110 | |
| 111 | $this->setMetadata( $result ); |
| 112 | } |
| 113 | |
| 114 | /** |
| 115 | * Set an array with metadata key value pairs |
| 116 | * Gets validated by Validator |
| 117 | * |
| 118 | * @param array $metadataArray |
| 119 | * @param ParserOutput|null $out ParserOutput is used to set a extension data flag to disable auto description, |
| 120 | * even when the flag is active. |
| 121 | * The reason is, if a description was provided and does not equal 'auto' or 'textextracts' we want to use it. |
| 122 | * @see Validator |
| 123 | */ |
| 124 | public function setMetadata( array $metadataArray, ParserOutput $out = null ): void { |
| 125 | $validMetadata = []; |
| 126 | |
| 127 | // We'll set a flag to don't overwrite manual descriptions |
| 128 | // If the AutoDescription setting is set |
| 129 | if ( $out !== null ) { |
| 130 | if ( isset( $metadataArray['manualDescription'] ) && |
| 131 | !in_array( $metadataArray['manualDescription'], [ 'auto', 'textextracts' ], true ) ) { |
| 132 | |
| 133 | $out->setPageProperty( 'manualDescription', true ); |
| 134 | |
| 135 | $metadataArray['description'] = $metadataArray['manualDescription']; |
| 136 | unset( $metadataArray['manualDescription'] ); |
| 137 | } else { |
| 138 | $out->unsetPageProperty( 'manualDescription' ); |
| 139 | } |
| 140 | } |
| 141 | |
| 142 | foreach ( Validator::validateParams( $metadataArray ) as $k => $v ) { |
| 143 | if ( !empty( $v ) ) { |
| 144 | $validMetadata[$k] = $v; |
| 145 | } |
| 146 | } |
| 147 | |
| 148 | $this->metadata = $validMetadata; |
| 149 | } |
| 150 | |
| 151 | /** |
| 152 | * Get a value from the $metadata array, given a key (or null if |
| 153 | * no value exists). |
| 154 | * |
| 155 | * @param string $key |
| 156 | * @return string|null |
| 157 | */ |
| 158 | public function getMetadataValue( $key ) { |
| 159 | return $this->metadata[$key] ?? null; |
| 160 | } |
| 161 | |
| 162 | /** |
| 163 | * Add the metadata array as meta tags to the page |
| 164 | * |
| 165 | * @param OutputPage $out |
| 166 | */ |
| 167 | public function addMetadataToPage( OutputPage $out ): void { |
| 168 | if ( $out->isArticle() && !isset( $out->getRequest()->getQueryValues()['diff'] ) ) { |
| 169 | $this->modifyPageTitle( $out ); |
| 170 | } |
| 171 | |
| 172 | MediaWikiServices::getInstance()->getHookContainer()->run( |
| 173 | 'WikiSEOPreAddMetadata', |
| 174 | [ |
| 175 | &$this->metadata, |
| 176 | ] |
| 177 | ); |
| 178 | |
| 179 | foreach ( $this->generatorInstances as $generatorInstance ) { |
| 180 | $generatorInstance->init( $this->metadata, $out ); |
| 181 | $generatorInstance->addMetadata(); |
| 182 | } |
| 183 | } |
| 184 | |
| 185 | /** |
| 186 | * Set active metadata generators defined in $wgMetdataGenerators |
| 187 | * And merges all valid parameter names from the generator to the validator |
| 188 | */ |
| 189 | private function setMetadataGenerators(): void { |
| 190 | $defaultGenerators = [ |
| 191 | 'OpenGraph', |
| 192 | 'Twitter', |
| 193 | 'SchemaOrg', |
| 194 | ]; |
| 195 | |
| 196 | try { |
| 197 | $generators = MediaWikiServices::getInstance() |
| 198 | ->getConfigFactory() |
| 199 | ->makeConfig( 'WikiSEO' ) |
| 200 | ->get( 'MetadataGenerators' ); |
| 201 | |
| 202 | if ( empty( $generators ) ) { |
| 203 | $generators = $defaultGenerators; |
| 204 | } |
| 205 | } catch ( ConfigException $e ) { |
| 206 | wfLogWarning( |
| 207 | sprintf( |
| 208 | 'Could not get config for "$wgMetadataGenerators", using default. %s', |
| 209 | $e->getMessage() |
| 210 | ) |
| 211 | ); |
| 212 | |
| 213 | $generators = $defaultGenerators; |
| 214 | } |
| 215 | |
| 216 | $this->generators = $generators; |
| 217 | $this->instantiateMetadataPlugins(); |
| 218 | $this->mergeValidParameterNames(); |
| 219 | } |
| 220 | |
| 221 | /** |
| 222 | * Loads all page props for the given page with pp_propnames in Validator::getValidParams() |
| 223 | * |
| 224 | * @param Title $title |
| 225 | * @return null|array Null if empty |
| 226 | * @see Validator::getValidParams() |
| 227 | */ |
| 228 | private function loadPagePropsFromDb( Title $title ): ?array { |
| 229 | $properties = MediaWikiServices::getInstance()->getPageProps()->getProperties( |
| 230 | $title, |
| 231 | Validator::getValidParams() |
| 232 | ); |
| 233 | |
| 234 | $properties = array_shift( $properties ); |
| 235 | |
| 236 | if ( $properties === null || count( $properties ) === 0 ) { |
| 237 | return null; |
| 238 | } |
| 239 | |
| 240 | $result = []; |
| 241 | |
| 242 | foreach ( $properties as $key => $value ) { |
| 243 | $result[$key] = $value; |
| 244 | } |
| 245 | |
| 246 | return $result; |
| 247 | } |
| 248 | |
| 249 | /** |
| 250 | * Tries to load the page props from OutputPage with keys from Validator::getValidParams() |
| 251 | * |
| 252 | * @see Validator::getValidParams() |
| 253 | * |
| 254 | * @param OutputPage $page |
| 255 | * @return array|null |
| 256 | */ |
| 257 | private function loadPagePropsFromOutputPage( OutputPage $page ): ?array { |
| 258 | $result = []; |
| 259 | |
| 260 | foreach ( Validator::getValidParams() as $param ) { |
| 261 | $prop = $page->getProperty( $param ); |
| 262 | |
| 263 | if ( $prop !== null ) { |
| 264 | AtEase::suppressWarnings(); |
| 265 | $value = unserialize( $prop, [ 'allowed_classes' => false ] ); |
| 266 | AtEase::restoreWarnings(); |
| 267 | |
| 268 | // Value was serialized |
| 269 | if ( $value !== false ) { |
| 270 | $prop = $value; |
| 271 | } |
| 272 | |
| 273 | $result[$param] = $prop; |
| 274 | } |
| 275 | } |
| 276 | |
| 277 | return empty( $result ) ? null : $result; |
| 278 | } |
| 279 | |
| 280 | /** |
| 281 | * Instantiates the metadata generators from $wgMetadataGenerators |
| 282 | */ |
| 283 | private function instantiateMetadataPlugins(): void { |
| 284 | $this->generatorInstances[] = new MetaTag(); |
| 285 | |
| 286 | foreach ( $this->generators as $generator ) { |
| 287 | $classPath = "MediaWiki\\Extension\\WikiSEO\\Generator\\Plugins\\$generator"; |
| 288 | |
| 289 | try { |
| 290 | $class = new ReflectionClass( $classPath ); |
| 291 | $this->generatorInstances[] = $class->newInstance(); |
| 292 | } catch ( ReflectionException $e ) { |
| 293 | $this->errors[] = wfMessage( 'wiki-seo-invalid-generator', $generator )->parse(); |
| 294 | } |
| 295 | } |
| 296 | } |
| 297 | |
| 298 | /** |
| 299 | * Finalize everything. |
| 300 | * Check for errors and save to props if everything is ok. |
| 301 | * |
| 302 | * @param ParserOutput $output |
| 303 | * |
| 304 | * @return string String with errors that happened or empty |
| 305 | */ |
| 306 | public function finalize( ParserOutput $output ): string { |
| 307 | if ( empty( $this->metadata ) ) { |
| 308 | $message = sprintf( 'wiki-seo-empty-attr-%s', $this->mode ); |
| 309 | $this->errors[] = wfMessage( $message ); |
| 310 | |
| 311 | return $this->makeErrorHtml(); |
| 312 | } |
| 313 | |
| 314 | $this->saveMetadataToProps( $output ); |
| 315 | |
| 316 | return ''; |
| 317 | } |
| 318 | |
| 319 | /** |
| 320 | * @return string Concatenated error strings |
| 321 | */ |
| 322 | private function makeErrorHtml(): string { |
| 323 | $text = implode( '<br>', $this->errors ); |
| 324 | |
| 325 | return sprintf( '<div class="errorbox">%s</div>', $text ); |
| 326 | } |
| 327 | |
| 328 | /** |
| 329 | * Modifies the page title based on 'titleMode' |
| 330 | * |
| 331 | * @param OutputPage $out |
| 332 | */ |
| 333 | private function modifyPageTitle( OutputPage $out ): void { |
| 334 | if ( !array_key_exists( 'title', $this->metadata ) ) { |
| 335 | return; |
| 336 | } |
| 337 | |
| 338 | $metaTitle = $this->metadata['title']; |
| 339 | |
| 340 | if ( array_key_exists( 'title_separator', $this->metadata ) ) { |
| 341 | $this->titleSeparator = $this->metadata['title_separator']; |
| 342 | } |
| 343 | |
| 344 | if ( array_key_exists( 'title_mode', $this->metadata ) ) { |
| 345 | $this->titleMode = $this->metadata['title_mode']; |
| 346 | } |
| 347 | |
| 348 | $strippedTitle = strip_tags( $out->getPageTitle() ); |
| 349 | |
| 350 | switch ( $this->titleMode ) { |
| 351 | case 'append': |
| 352 | $pageTitle = sprintf( '%s%s%s', $strippedTitle, $this->titleSeparator, $metaTitle ); |
| 353 | break; |
| 354 | |
| 355 | case 'prepend': |
| 356 | $pageTitle = sprintf( '%s%s%s', $metaTitle, $this->titleSeparator, $strippedTitle ); |
| 357 | break; |
| 358 | |
| 359 | case 'replace': |
| 360 | default: |
| 361 | $pageTitle = $metaTitle; |
| 362 | break; |
| 363 | } |
| 364 | |
| 365 | $pageTitle = preg_replace( "/[\r\n]/", '', $pageTitle ); |
| 366 | $pageTitle = html_entity_decode( $pageTitle, ENT_QUOTES ); |
| 367 | |
| 368 | $out->setHTMLTitle( $pageTitle ); |
| 369 | } |
| 370 | |
| 371 | /** |
| 372 | * Save the metadata array json encoded to the page props table |
| 373 | * |
| 374 | * @param ParserOutput $outputPage |
| 375 | */ |
| 376 | private function saveMetadataToProps( ParserOutput $outputPage ): void { |
| 377 | MediaWikiServices::getInstance()->getHookContainer()->run( |
| 378 | 'WikiSEOPreAddPageProps', |
| 379 | [ |
| 380 | &$this->metadata, |
| 381 | ] |
| 382 | ); |
| 383 | |
| 384 | foreach ( $this->metadata as $key => $value ) { |
| 385 | if ( $outputPage->getPageProperty( $key ) === null ) { |
| 386 | $outputPage->setPageProperty( $key, $value ); |
| 387 | } |
| 388 | |
| 389 | if ( ExtensionRegistry::getInstance()->isLoaded( 'PageImages' ) && |
| 390 | $key === 'image' ) { |
| 391 | $outputPage->setPageProperty( PageImages::PROP_NAME_FREE, $value ); |
| 392 | } |
| 393 | } |
| 394 | } |
| 395 | |
| 396 | /** |
| 397 | * Adds valid tags from all generator instances to the Validator |
| 398 | * Automatically called after instantiating all active generators |
| 399 | */ |
| 400 | private function mergeValidParameterNames(): void { |
| 401 | Validator::$validParameterNames = array_unique( |
| 402 | array_merge( |
| 403 | Validator::$validParameterNames, |
| 404 | array_reduce( |
| 405 | array_map( |
| 406 | static function ( GeneratorInterface $generator ) { |
| 407 | return $generator->getAllowedParameterNames(); |
| 408 | }, |
| 409 | $this->generatorInstances |
| 410 | ), |
| 411 | static function ( array $carry, array $item ) { |
| 412 | return array_merge( $carry, $item ); |
| 413 | }, |
| 414 | [] |
| 415 | ) |
| 416 | ) |
| 417 | ); |
| 418 | } |
| 419 | |
| 420 | /** |
| 421 | * Parse the values input from the <seo> tag extension |
| 422 | * |
| 423 | * @param string|null $input The text content of the tag |
| 424 | * @param array $args The HTML attributes of the tag |
| 425 | * @param Parser $parser The active Parser instance |
| 426 | * @param PPFrame $frame |
| 427 | * |
| 428 | * @return string The HTML comments of cached attributes |
| 429 | */ |
| 430 | public static function fromTag( ?string $input, array $args, Parser $parser, PPFrame $frame ): string { |
| 431 | $seo = new WikiSEO( self::MODE_TAG ); |
| 432 | $tagParser = new TagParser(); |
| 433 | |
| 434 | $parsedInput = $tagParser->parseText( $input, $parser, $frame ); |
| 435 | $tags = array_merge( $parsedInput, $args ); |
| 436 | $tags = $tagParser->expandWikiTextTagArray( $tags, $parser, $frame ); |
| 437 | |
| 438 | if ( isset( $tags['description'] ) ) { |
| 439 | $tags['manualDescription'] = $tags['description']; |
| 440 | unset( $tags['description'] ); |
| 441 | } |
| 442 | $tags = array_merge( $seo->loadPagePropsFromDb( $frame->getTitle() ) ?? [], $tags ); |
| 443 | |
| 444 | $seo->setMetadata( $tags, $parser->getOutput() ); |
| 445 | |
| 446 | return $seo->finalize( $parser->getOutput() ); |
| 447 | } |
| 448 | |
| 449 | /** |
| 450 | * Parse the values input from the {{#seo}} parser function |
| 451 | * |
| 452 | * @param Parser $parser The active Parser instance |
| 453 | * @param PPFrame $frame Frame |
| 454 | * @param array $args Arguments |
| 455 | * |
| 456 | * @return array Parser options and the HTML comments of cached attributes |
| 457 | */ |
| 458 | public static function fromParserFunction( $parser, PPFrame $frame, array $args ): array { |
| 459 | $expandedArgs = []; |
| 460 | |
| 461 | foreach ( $args as $arg ) { |
| 462 | $expandedArgs[] = trim( $frame->expand( $arg ) ); |
| 463 | } |
| 464 | |
| 465 | $seo = new WikiSEO( self::MODE_PARSER ); |
| 466 | $tagParser = new TagParser(); |
| 467 | |
| 468 | $args = $tagParser->parseArgs( $expandedArgs, $parser, $frame ); |
| 469 | if ( isset( $args['description'] ) ) { |
| 470 | $args['manualDescription'] = $args['description']; |
| 471 | unset( $args['description'] ); |
| 472 | } |
| 473 | |
| 474 | $args = array_merge( $seo->loadPagePropsFromDb( $frame->getTitle() ) ?? [], $args ); |
| 475 | |
| 476 | $seo->setMetadata( $args, $parser->getOutput() ); |
| 477 | |
| 478 | $fin = $seo->finalize( $parser->getOutput() ); |
| 479 | if ( !empty( $fin ) ) { |
| 480 | return [ |
| 481 | $fin, |
| 482 | 'noparse' => true, |
| 483 | 'isHTML' => true, |
| 484 | ]; |
| 485 | } |
| 486 | |
| 487 | // See https://github.com/octfx/wiki-seo/issues/30 |
| 488 | return [ '<!-- WikiSEO -->' ]; |
| 489 | } |
| 490 | |
| 491 | /** |
| 492 | * Add the server protocol to the URL if it is missing |
| 493 | * |
| 494 | * @param string $url URL from getFullURL() |
| 495 | * @param WebRequest $request |
| 496 | * |
| 497 | * @return string |
| 498 | */ |
| 499 | public static function protocolizeUrl( string $url, WebRequest $request ): string { |
| 500 | if ( parse_url( $url, PHP_URL_SCHEME ) === null ) { |
| 501 | $url = sprintf( '%s:%s', $request->getProtocol(), $url ); |
| 502 | } |
| 503 | |
| 504 | return $url; |
| 505 | } |
| 506 | } |