samwilson /
simple-wikidata
| Total Complexity | 64 |
| Total Lines | 433 |
| Duplicated Lines | 8.31 % |
| Coverage | 25.12% |
| Changes | 0 | ||
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 Item 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.
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 Item, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 14 | class Item { |
||
| 15 | |||
| 16 | const PROP_INSTANCE_OF = 'P31'; |
||
| 17 | const PROP_TITLE = 'P1476'; |
||
| 18 | const PROP_IMAGE = 'P18'; |
||
| 19 | const PROP_AUTHOR = 'P50'; |
||
| 20 | |||
| 21 | /** @var string */ |
||
| 22 | protected $id; |
||
| 23 | |||
| 24 | /** @var MediawikiApi */ |
||
| 25 | protected $wdApi; |
||
| 26 | |||
| 27 | /** @var string */ |
||
| 28 | protected $lang; |
||
| 29 | |||
| 30 | /** @var CacheItemPoolInterface */ |
||
| 31 | protected $cache; |
||
| 32 | |||
| 33 | /** @var string The base URL of Wikidata, with trailing slash. */ |
||
| 34 | protected $wikidataUrlBase = 'https://www.wikidata.org/wiki/'; |
||
| 35 | |||
| 36 | 1 | private function __construct( $id, $lang, CacheItemPoolInterface $cache ) { |
|
| 37 | 1 | if ( !is_string( $id ) || preg_match( '/[QP][0-9]*/i', $id ) !== 1 ) { |
|
| 38 | throw new Exception( "Not a valid ID: " . var_export( $id, true ) ); |
||
| 39 | } |
||
| 40 | 1 | $this->id = $id; |
|
| 41 | 1 | $this->wdApi = new MediawikiApi( 'https://www.wikidata.org/w/api.php' ); |
|
| 42 | 1 | $this->entities = []; |
|
|
Bug
Best Practice
introduced
by
|
|||
| 43 | 1 | $this->lang = $lang; |
|
| 44 | 1 | $this->cache = $cache; |
|
| 45 | 1 | } |
|
| 46 | |||
| 47 | /** |
||
| 48 | * Create a new Item object with class based on the item's 'instance of' statement. |
||
| 49 | * |
||
| 50 | * @param string $id |
||
| 51 | * @param string $lang |
||
| 52 | * |
||
| 53 | * @return Item |
||
| 54 | */ |
||
| 55 | 1 | public static function factory( $id, $lang, CacheItemPoolInterface $cache ) { |
|
| 56 | 1 | $item = new Item( $id, $lang, $cache ); |
|
| 57 | 1 | foreach ( $item->getPropertyOfTypeItem( self::PROP_INSTANCE_OF ) as $instanceOf ) { |
|
| 58 | // Try to find a class mating the 'instance of' name. |
||
| 59 | 1 | $possibleClassName = __NAMESPACE__ . '\\Items\\' . Str::toCamelCase( $instanceOf->getItem()->getLabel() ); |
|
| 60 | 1 | if ( class_exists( $possibleClassName ) ) { |
|
| 61 | // This won't re-request the metadata, because that's cached. |
||
| 62 | $specificItem = new $possibleClassName( $id, $lang, $cache ); |
||
| 63 | 1 | return $specificItem; |
|
| 64 | } |
||
| 65 | } |
||
| 66 | |||
| 67 | // If we're here, just leave it as a basic Item. |
||
| 68 | 1 | $item->setCache( $cache ); |
|
| 69 | 1 | return $item; |
|
| 70 | } |
||
| 71 | |||
| 72 | 1 | public function setCache( CacheItemPoolInterface $cache_item_pool ) { |
|
| 73 | 1 | $this->cache = $cache_item_pool; |
|
| 74 | 1 | } |
|
| 75 | |||
| 76 | /** |
||
| 77 | * Get the ID (Q-number) of this item. |
||
| 78 | * @return string|bool The ID or false if it couldn't be determined. |
||
| 79 | */ |
||
| 80 | 1 | public function getId() { |
|
| 81 | 1 | $entity = $this->getEntity( $this->id ); |
|
| 82 | 1 | return isset( $entity['id'] ) ? $entity['id'] : false; |
|
| 83 | } |
||
| 84 | |||
| 85 | /** |
||
| 86 | * Get this item's label. |
||
| 87 | * @return string |
||
| 88 | */ |
||
| 89 | 1 | public function getLabel() { |
|
| 97 | } |
||
| 98 | |||
| 99 | 1 | public function getWikidataUrl() { |
|
| 101 | } |
||
| 102 | |||
| 103 | /** |
||
| 104 | * Wikiprojects list their properties like this: |
||
| 105 | * |
||
| 106 | * {{List of properties/Header}} |
||
| 107 | * {{List of properties/Row|id=31|example-subject=Q923767|example-object=Q3331189}} |
||
| 108 | * </table> |
||
| 109 | * |
||
| 110 | * @param string $wikiProject |
||
| 111 | * @param string $type |
||
| 112 | * @return array |
||
| 113 | */ |
||
| 114 | public function getStandardProperties( $wikiProject = 'WikiProject_Books', $type = 'work' ) { |
||
| 115 | if ( $type !== 'work' ) { |
||
| 116 | $type = 'edition'; |
||
| 117 | } |
||
| 118 | $cacheKey = $type . '_item_property_IDs'; |
||
| 119 | if ( $this->cache->hasItem( $cacheKey ) ) { |
||
| 120 | $propIds = $this->cache->getItem( $cacheKey )->get(); |
||
| 121 | } else { |
||
| 122 | $domCrawler = new Crawler(); |
||
| 123 | $wikiProjectUrl = 'https://www.wikidata.org/wiki/Wikidata:' . $wikiProject; |
||
| 124 | $domCrawler->addHtmlContent( file_get_contents( $wikiProjectUrl ) ); |
||
| 125 | $propAncors = "//h3/span[@id='" . ucfirst( $type ) . "_item_properties']/../following-sibling::table[1]//td[2]/a"; |
||
| 126 | $propCells = $domCrawler->filterXPath( $propAncors ); |
||
| 127 | $propIds = []; |
||
| 128 | $propCells->each( function ( Crawler $node, $i ) use ( &$propIds ) { |
||
| 129 | $propId = $node->text(); |
||
| 130 | $propIds[] = $propId; |
||
| 131 | } ); |
||
| 132 | $cacheItem = $this->cache->getItem( $cacheKey ) |
||
| 133 | ->expiresAfter( new DateInterval( 'PT1H' ) ) |
||
| 134 | ->set( $propIds ); |
||
| 135 | $this->cache->save( $cacheItem ); |
||
| 136 | } |
||
| 137 | $workProperties = []; |
||
| 138 | foreach ( $propIds as $propId ) { |
||
| 139 | $workProperties[] = self::factory( $propId, $this->lang, $this->cache ); |
||
| 140 | } |
||
| 141 | |||
| 142 | return $workProperties; |
||
| 143 | } |
||
| 144 | |||
| 145 | /** |
||
| 146 | * @param string $propertyId |
||
| 147 | * @return bool|Time[] |
||
| 148 | */ |
||
| 149 | View Code Duplication | protected function getPropertyOfTypeTime( $propertyId ) { |
|
| 150 | $times = []; |
||
| 151 | $entity = $this->getEntity(); |
||
| 152 | if ( !isset( $entity['claims'][$propertyId] ) ) { |
||
| 153 | // No statements for this property. |
||
| 154 | return $times; |
||
| 155 | } |
||
| 156 | // print_r($entity['claims'][$propertyId]);exit(); |
||
| 157 | foreach ( $entity['claims'][$propertyId] as $claim ) { |
||
| 158 | // print_r($claim); |
||
| 159 | $times[] = new Time( $claim, $this->lang, $this->cache ); |
||
| 160 | // |
||
| 161 | // $timeValue = $claim['datavalue']['value']['time']; |
||
| 162 | // // Ugly workaround for imprecise dates. :-( |
||
| 163 | // if (preg_match('/([0-9]{1,4})-00-00/', $timeValue, $matches) === 1) { |
||
| 164 | // $timeValue = $matches[1]; |
||
| 165 | // return $timeValue; |
||
| 166 | // } |
||
| 167 | // $time = strtotime($timeValue); |
||
| 168 | // return date($dateFormat, $time); |
||
| 169 | // } |
||
| 170 | } |
||
| 171 | return $times; |
||
| 172 | } |
||
| 173 | |||
| 174 | /** |
||
| 175 | * Get the Item that is referred to by the specified item's property. |
||
| 176 | * |
||
| 177 | * @param string $propertyId |
||
| 178 | * |
||
| 179 | * @return \Samwilson\SimpleWikidata\Properties\Item[] |
||
| 180 | */ |
||
| 181 | 1 | View Code Duplication | protected function getPropertyOfTypeItem( $propertyId ) { |
| 182 | 1 | $entity = $this->getEntity( $this->id ); |
|
| 183 | 1 | if ( !isset( $entity['claims'][$propertyId] ) ) { |
|
| 184 | 1 | return []; |
|
| 185 | } |
||
| 186 | 1 | $items = []; |
|
| 187 | 1 | foreach ( $entity['claims'][$propertyId] as $claim ) { |
|
| 188 | 1 | $items[] = new Properties\Item( $claim, $this->lang, $this->cache ); |
|
| 189 | } |
||
| 190 | |||
| 191 | 1 | return $items; |
|
| 192 | } |
||
| 193 | |||
| 194 | public function setPropertyOfTypeItem( $property, $itemId ) { |
||
| 195 | $itemIdNumeric = substr( $itemId, 1 ); |
||
| 196 | |||
| 197 | // First see if this property already exists, and that it is different from what's being set. |
||
| 198 | $entity = $this->getEntity( $this->id ); |
||
| 199 | if ( !empty( $entity['claims'][$property] ) ) { |
||
| 200 | // Get the first claim, and update it if necessary. |
||
| 201 | $claim = array_shift( $entity['claims'][$property] ); |
||
| 202 | if ( $claim['mainsnak']['datavalue']['value']['id'] == $itemId ) { |
||
| 203 | // Already is the required value, no need to change. |
||
| 204 | return; |
||
| 205 | } |
||
| 206 | $claim['mainsnak']['datavalue']['value']['id'] = $itemId; |
||
| 207 | $claim['mainsnak']['datavalue']['value']['numeric-id'] = $itemIdNumeric; |
||
| 208 | $apiParams = [ |
||
| 209 | 'action' => 'wbsetclaim', |
||
| 210 | 'claim' => json_encode( $claim ), |
||
| 211 | ]; |
||
| 212 | } |
||
| 213 | |||
| 214 | // If no claim was found (and modified) above, create a new claim. |
||
| 215 | if ( !isset( $apiParams ) ) { |
||
| 216 | $apiParams = [ |
||
| 217 | 'action' => 'wbcreateclaim', |
||
| 218 | 'entity' => $this->getId(), |
||
| 219 | 'property' => $property, |
||
| 220 | 'snaktype' => 'value', |
||
| 221 | 'value' => json_encode( [ 'entity-type' => 'item', 'numeric-id' => $itemIdNumeric ] ), |
||
| 222 | ]; |
||
| 223 | } |
||
| 224 | |||
| 225 | // Save the property. |
||
| 226 | $wdWpOauth = new WdOauth(); |
||
| 227 | $wdWpOauth->makeCall( $apiParams, true ); |
||
| 228 | |||
| 229 | // Clear the cache. |
||
| 230 | $this->cache->deleteItem( $this->getEntityCacheKey( $this->id ) ); |
||
| 231 | } |
||
| 232 | |||
| 233 | public function getPropertyOfTypeUrl( $entityId, $propertyId ) { |
||
| 244 | } |
||
| 245 | |||
| 246 | public function getPropertyOfTypeExternalIdentifier( $entityId, $propertyId ) { |
||
| 274 | } |
||
| 275 | |||
| 276 | /** |
||
| 277 | * Get a single-valued text property. |
||
| 278 | * @param string $property One of the PROP_* constants. |
||
| 279 | * @return string|bool The value, or false if it can't be found. |
||
| 280 | */ |
||
| 281 | public function getPropertyOfTypeText( $property ) { |
||
| 282 | $entity = $this->getEntity( $this->id ); |
||
| 283 | if ( isset( $entity['claims'][$property] ) ) { |
||
| 284 | // Use the first title. |
||
| 285 | foreach ( $entity['claims'][$property] as $t ) { |
||
| 286 | if ( !isset( $t['mainsnak']['datavalue']['value']['language'] ) ) { |
||
| 287 | var_dump( $t['mainsnak']['datavalue']['value'] ); |
||
| 288 | exit(); |
||
| 289 | } |
||
| 290 | if ( $t['mainsnak']['datavalue']['value']['language'] == $this->lang |
||
| 291 | && !empty( $t['mainsnak']['datavalue']['value']['text'] ) |
||
| 292 | ) { |
||
| 293 | return $t['mainsnak']['datavalue']['value']['text']; |
||
| 294 | } |
||
| 295 | } |
||
| 296 | } |
||
| 297 | return false; |
||
| 298 | } |
||
| 299 | |||
| 300 | /** |
||
| 301 | * Literal data field for a quantity that relates to some kind of well-defined unit. The actual unit goes in the data values that is entered. |
||
| 302 | * - amount – implicit part of the string (mapping of unit prefix is unclear) |
||
| 303 | * - unit – implicit part of the string that defaults to "1" (mapping to standardizing body is unclear) |
||
| 304 | * - upperbound - quantity's upper bound |
||
| 305 | * - lowerbound - quantity's lower bound |
||
| 306 | * @param $property |
||
| 307 | * @return mixed[]|bool If it's not false it's an array with 'amount', 'unit', etc. |
||
| 308 | */ |
||
| 309 | public function getPropertyOfTypeQuantity( $property ) { |
||
| 310 | $quantities = []; |
||
| 311 | $entity = $this->getEntity( $this->id ); |
||
| 312 | if ( !isset( $entity['claims'][$property] ) ) { |
||
| 313 | return false; |
||
| 314 | } |
||
| 315 | foreach ( $entity['claims'][$property] as $t ) { |
||
| 316 | $quantity = $t['mainsnak']['datavalue']['value']; |
||
| 317 | $unitId = substr( $quantity['unit'], strlen( $this->wikidataUrlBase ) + 1 ); |
||
| 318 | $quantity['unit'] = self::factory( $unitId, $this->lang, $this->cache ); |
||
| 319 | $quantities[] = $quantity; |
||
| 320 | } |
||
| 321 | return $quantities; |
||
| 322 | } |
||
| 323 | |||
| 324 | /** |
||
| 325 | * Set a single-valued text property. |
||
| 326 | * @param string $property One of the PROP_* constants. |
||
| 327 | * @param string $value The value. |
||
| 328 | */ |
||
| 329 | public function setPropertyOfTypeText( $property, $value ) { |
||
| 330 | // First see if this property already exists, and that it is different from what's being set. |
||
| 331 | $entity = $this->getEntity( $this->id ); |
||
| 332 | if ( !empty( $entity['claims'][$property] ) ) { |
||
| 333 | // Find this language's claim (if there is one). |
||
| 334 | foreach ( $entity['claims'][$property] as $claim ) { |
||
| 335 | if ( $claim['mainsnak']['datavalue']['value']['language'] == $this->lang ) { |
||
| 336 | // Modify this claim's text value. |
||
| 337 | $titleClaim = $claim; |
||
| 338 | $titleClaim['mainsnak']['datavalue']['value']['text'] = $value; |
||
| 339 | $setTitleParams = [ |
||
| 340 | 'action' => 'wbsetclaim', |
||
| 341 | 'claim' => \GuzzleHttp\json_encode( $titleClaim ), |
||
| 342 | ]; |
||
| 343 | continue; |
||
| 344 | } |
||
| 345 | } |
||
| 346 | } |
||
| 347 | |||
| 348 | // If no claim was found (and modified) above, create a new claim. |
||
| 349 | if ( !isset( $setTitleParams ) ) { |
||
| 350 | $setTitleParams = [ |
||
| 351 | 'action' => 'wbcreateclaim', |
||
| 352 | 'entity' => $this->getId(), |
||
| 353 | 'property' => $property, |
||
| 354 | 'snaktype' => 'value', |
||
| 355 | 'value' => \GuzzleHttp\json_encode( [ 'text' => $value, 'language' => $this->lang ] ), |
||
| 356 | ]; |
||
| 357 | } |
||
| 358 | |||
| 359 | // Save the property. |
||
| 360 | $wdWpOauth = new WdWpOauth(); |
||
| 361 | $wdWpOauth->makeCall( $setTitleParams, true ); |
||
| 362 | |||
| 363 | // Clear the cache. |
||
| 364 | $this->cache->deleteItem( $this->getEntityCacheKey( $this->id ) ); |
||
| 365 | } |
||
| 366 | |||
| 367 | public function getInstanceOf() { |
||
| 368 | $instancesOf = $this->getPropertyOfTypeItem( $this->getId(), self::PROP_INSTANCE_OF ); |
||
| 369 | return array_shift( $instancesOf ); |
||
| 370 | } |
||
| 371 | |||
| 372 | /** |
||
| 373 | * Does this item exist? |
||
| 374 | * @return bool |
||
| 375 | */ |
||
| 376 | public function exists() { |
||
| 377 | return $this->getId() !== false; |
||
| 378 | } |
||
| 379 | |||
| 380 | public function getWikipediaIntro() { |
||
| 413 | } |
||
| 414 | |||
| 415 | /** |
||
| 416 | * Get the raw entity data from the 'wbgetentities' API call. |
||
| 417 | * @param string $id |
||
| 418 | * @param bool $ignoreCache |
||
| 419 | * @return bool |
||
| 420 | */ |
||
| 421 | 1 | public function getEntity( $id = null, $ignoreCache = false ) { |
|
| 422 | 1 | $idActual = $id ?: $this->id; |
|
| 423 | 1 | $cacheKey = $this->getEntityCacheKey( $idActual ); |
|
| 424 | 1 | if ( !$ignoreCache && $this->cache->hasItem( $cacheKey ) ) { |
|
| 425 | return $this->cache->getItem( $cacheKey )->get(); |
||
| 426 | } |
||
| 427 | 1 | $metadataRequest = new SimpleRequest( 'wbgetentities', [ 'ids' => $idActual ] ); |
|
| 428 | 1 | $itemResult = $this->wdApi->getRequest( $metadataRequest ); |
|
| 429 | 1 | if ( !isset( $itemResult['success'] ) || !isset( $itemResult['entities'][$id] ) ) { |
|
| 430 | return false; |
||
| 431 | } |
||
| 432 | 1 | $metadata = $itemResult['entities'][$idActual]; |
|
| 433 | 1 | $cacheItem = $this->cache->getItem( $cacheKey ) |
|
| 434 | 1 | ->expiresAfter( new DateInterval( 'PT10M' ) ) |
|
| 435 | 1 | ->set( $metadata ); |
|
| 436 | 1 | $this->cache->save( $cacheItem ); |
|
| 437 | 1 | return $metadata; |
|
| 438 | } |
||
| 439 | |||
| 440 | /** |
||
| 441 | * @param $id |
||
| 442 | * |
||
| 443 | * @return string |
||
| 444 | */ |
||
| 445 | 1 | protected function getEntityCacheKey( $id ) { |
|
| 447 | } |
||
| 448 | } |
||
| 449 |
Loading history...