NNTmux /
newznab-tmux
| Total Complexity | 64 |
| Total Lines | 344 |
| Duplicated Lines | 0 % |
| Changes | 1 | ||
| Bugs | 0 | Features | 0 |
Complex classes like HotmoviesPipe 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 HotmoviesPipe, and based on these observations, apply Extract Interface, too.
| 1 | <?php |
||
| 13 | class HotmoviesPipe extends AbstractAdultProviderPipe |
||
| 14 | { |
||
| 15 | protected int $priority = 50; |
||
| 16 | |||
| 17 | private const BASE_URL = 'https://www.hotmovies.com'; |
||
| 18 | private const SEARCH_URL = '/search.php?words='; |
||
| 19 | private const EXTRA_SEARCH = '&complete=on&search_in=video_title'; |
||
| 20 | |||
| 21 | protected string $directUrl = ''; |
||
| 22 | protected string $title = ''; |
||
| 23 | protected string $response = ''; |
||
| 24 | |||
| 25 | public function getName(): string |
||
| 26 | { |
||
| 27 | return 'hotm'; |
||
| 28 | } |
||
| 29 | |||
| 30 | public function getDisplayName(): string |
||
| 31 | { |
||
| 32 | return 'HotMovies'; |
||
| 33 | } |
||
| 34 | |||
| 35 | protected function getBaseUrl(): string |
||
| 38 | } |
||
| 39 | |||
| 40 | protected function process(AdultProcessingPassable $passable): AdultProcessingResult |
||
| 75 | ); |
||
| 76 | } |
||
| 77 | |||
| 78 | protected function search(string $movie): array|false |
||
| 79 | { |
||
| 80 | if (empty($movie)) { |
||
| 81 | return false; |
||
| 82 | } |
||
| 83 | |||
| 84 | // Initialize session with age verification |
||
| 85 | $this->initializeSession(); |
||
| 86 | |||
| 87 | $searchUrl = self::BASE_URL . self::SEARCH_URL . urlencode($movie) . self::EXTRA_SEARCH; |
||
| 88 | $response = $this->fetchHtml($searchUrl, $this->cookie); |
||
| 89 | |||
| 90 | if ($response === false) { |
||
| 91 | return false; |
||
| 92 | } |
||
| 93 | |||
| 94 | $this->getHtmlParser()->loadHtml($response); |
||
| 95 | |||
| 96 | // Try multiple result selectors |
||
| 97 | $resultSelectors = [ |
||
| 98 | 'h3[class=title] a[title]', |
||
| 99 | 'h3.title a', |
||
| 100 | 'div.movie-title a', |
||
| 101 | ]; |
||
| 102 | |||
| 103 | $bestMatch = null; |
||
| 104 | $highestSimilarity = 0; |
||
| 105 | |||
| 106 | foreach ($resultSelectors as $selector) { |
||
| 107 | $elements = $this->getHtmlParser()->find($selector); |
||
| 108 | if (!empty($elements)) { |
||
| 109 | foreach ($elements as $ret) { |
||
| 110 | $title = $ret->title ?? $ret->plaintext ?? ''; |
||
| 111 | $url = $ret->href ?? ''; |
||
| 112 | |||
| 113 | if (empty($title) || empty($url)) { |
||
| 114 | continue; |
||
| 115 | } |
||
| 116 | |||
| 117 | $similarity = $this->calculateSimilarity($movie, $title); |
||
| 118 | |||
| 119 | if ($similarity > $highestSimilarity) { |
||
| 120 | $highestSimilarity = $similarity; |
||
| 121 | $bestMatch = [ |
||
| 122 | 'title' => trim($title), |
||
| 123 | 'url' => str_starts_with($url, 'http') ? $url : self::BASE_URL . $url, |
||
| 124 | ]; |
||
| 125 | } |
||
| 126 | } |
||
| 127 | |||
| 128 | // If we found results with this selector, don't try others |
||
| 129 | if ($bestMatch !== null) { |
||
| 130 | break; |
||
| 131 | } |
||
| 132 | } |
||
| 133 | } |
||
| 134 | |||
| 135 | if ($bestMatch !== null && $highestSimilarity >= $this->minimumSimilarity) { |
||
| 136 | return $bestMatch; |
||
| 137 | } |
||
| 138 | |||
| 139 | return false; |
||
| 140 | } |
||
| 141 | |||
| 142 | protected function getMovieInfo(): array|false |
||
| 143 | { |
||
| 144 | $results = []; |
||
| 145 | |||
| 146 | if (!empty($this->directUrl)) { |
||
| 147 | if (!empty($this->title)) { |
||
| 148 | $results['title'] = $this->title; |
||
| 149 | } |
||
| 150 | $results['directurl'] = $this->directUrl; |
||
| 151 | } |
||
| 152 | |||
| 153 | // Get all the movie data |
||
| 154 | $synopsis = $this->extractSynopsis(); |
||
| 155 | if (is_array($synopsis)) { |
||
| 156 | $results = array_merge($results, $synopsis); |
||
| 157 | } |
||
| 158 | |||
| 159 | $productInfo = $this->extractProductInfo(true); |
||
| 160 | if (is_array($productInfo)) { |
||
| 161 | $results = array_merge($results, $productInfo); |
||
| 162 | } |
||
| 163 | |||
| 164 | $cast = $this->extractCast(); |
||
| 165 | if (is_array($cast)) { |
||
| 166 | $results = array_merge($results, $cast); |
||
| 167 | } |
||
| 168 | |||
| 169 | $genres = $this->extractGenres(); |
||
| 170 | if (is_array($genres)) { |
||
| 171 | $results = array_merge($results, $genres); |
||
| 172 | } |
||
| 173 | |||
| 174 | $covers = $this->extractCovers(); |
||
| 175 | if (is_array($covers)) { |
||
| 176 | $results = array_merge($results, $covers); |
||
| 177 | } |
||
| 178 | |||
| 179 | if (empty($results)) { |
||
| 180 | return false; |
||
| 181 | } |
||
| 182 | |||
| 183 | return $results; |
||
| 184 | } |
||
| 185 | |||
| 186 | protected function extractSynopsis(): array |
||
| 187 | { |
||
| 188 | $res = []; |
||
| 189 | $res['synopsis'] = 'N/A'; |
||
| 190 | |||
| 191 | // Try multiple selectors |
||
| 192 | $selectors = [ |
||
| 193 | '.video_description', |
||
| 194 | 'div.description', |
||
| 195 | 'div.synopsis', |
||
| 196 | 'meta[name="description"]', |
||
| 197 | ]; |
||
| 198 | |||
| 199 | foreach ($selectors as $selector) { |
||
| 200 | $ret = $this->getHtmlParser()->findOne($selector); |
||
| 201 | if ($ret) { |
||
| 202 | $text = $ret->innerText ?? $ret->plaintext ?? $ret->content ?? ''; |
||
| 203 | if (!empty(trim($text))) { |
||
| 204 | $res['synopsis'] = trim($text); |
||
| 205 | |||
| 206 | return $res; |
||
| 207 | } |
||
| 208 | } |
||
| 209 | } |
||
| 210 | |||
| 211 | return $res; |
||
| 212 | } |
||
| 213 | |||
| 214 | protected function extractProductInfo(bool $extras = false): array |
||
| 215 | { |
||
| 216 | $res = []; |
||
| 217 | $studio = false; |
||
| 218 | $director = false; |
||
| 219 | |||
| 220 | if (($ret = $this->getHtmlParser()->find('div.page_video_info')) && !empty($ret->find('text'))) { |
||
| 221 | $productinfo = []; |
||
| 222 | |||
| 223 | foreach ($ret->find('text') as $e) { |
||
| 224 | $e = trim($e->plaintext); |
||
| 225 | $rArray = [',', '...', ' :']; |
||
| 226 | $e = str_replace($rArray, '', $e); |
||
| 227 | |||
| 228 | if (stripos($e, 'Studio:') !== false) { |
||
| 229 | $studio = true; |
||
| 230 | } |
||
| 231 | |||
| 232 | if (str_contains($e, 'Director:')) { |
||
| 233 | $director = true; |
||
| 234 | $e = null; |
||
| 235 | } |
||
| 236 | |||
| 237 | if ($studio === true) { |
||
| 238 | if ((stripos($e, 'Custodian of Records') === false) && stripos($e, 'Description') === false) { |
||
| 239 | if ($director === true && !empty($e)) { |
||
| 240 | $res['director'] = $e; |
||
| 241 | $e = null; |
||
| 242 | $director = false; |
||
| 243 | } |
||
| 244 | if (!empty($e)) { |
||
| 245 | $productinfo[] = $e; |
||
| 246 | } |
||
| 247 | } else { |
||
| 248 | break; |
||
| 249 | } |
||
| 250 | } |
||
| 251 | } |
||
| 252 | |||
| 253 | if (is_array($productinfo)) { |
||
| 254 | $res['productinfo'] = array_chunk($productinfo, 2, false); |
||
| 255 | } |
||
| 256 | } |
||
| 257 | |||
| 258 | return $res; |
||
| 259 | } |
||
| 260 | |||
| 261 | protected function extractCast(): array |
||
| 262 | { |
||
| 263 | $res = []; |
||
| 264 | $cast = []; |
||
| 265 | |||
| 266 | // Prefer scoped search within stars container to avoid unrelated links |
||
| 267 | if ($container = $this->getHtmlParser()->findOne('.stars')) { |
||
| 268 | foreach ($container->find('a[title]') as $e) { |
||
| 269 | $name = trim($e->title); |
||
| 270 | $name = preg_replace('/\((.*)\)/', '', $name); |
||
| 271 | $name = trim($name); |
||
| 272 | if ($name !== '') { |
||
| 273 | $cast[] = $name; |
||
| 274 | } |
||
| 275 | } |
||
| 276 | } |
||
| 277 | |||
| 278 | // Fallback: anchors that look like performer links |
||
| 279 | if (empty($cast)) { |
||
| 280 | foreach ($this->getHtmlParser()->find('a[href*="/performers/"]') as $e) { |
||
| 281 | $name = trim($e->plaintext); |
||
| 282 | if ($name !== '') { |
||
| 283 | $cast[] = $name; |
||
| 284 | } |
||
| 285 | } |
||
| 286 | } |
||
| 287 | |||
| 288 | if (!empty($cast)) { |
||
| 289 | $res['cast'] = array_values(array_unique($cast)); |
||
| 290 | } |
||
| 291 | |||
| 292 | return $res; |
||
| 293 | } |
||
| 294 | |||
| 295 | protected function extractGenres(): array |
||
| 311 | } |
||
| 312 | |||
| 313 | protected function extractCovers(): array |
||
| 314 | { |
||
| 315 | $res = []; |
||
| 316 | |||
| 317 | // Try multiple selectors |
||
| 318 | $selectors = [ |
||
| 319 | 'img#cover', |
||
| 320 | 'div#large_cover img', |
||
| 321 | 'img.boxcover', |
||
| 322 | 'div.product-image img', |
||
| 323 | ]; |
||
| 324 | |||
| 325 | foreach ($selectors as $selector) { |
||
| 326 | $ret = $this->getHtmlParser()->findOne($selector); |
||
| 327 | if ($ret && isset($ret->src)) { |
||
| 328 | $res['boxcover'] = trim($ret->src); |
||
| 329 | $res['backcover'] = str_ireplace(['.cover', 'front'], ['.back', 'back'], trim($ret->src)); |
||
| 330 | |||
| 331 | return $res; |
||
| 332 | } |
||
| 333 | } |
||
| 334 | |||
| 335 | return $res; |
||
| 336 | } |
||
| 337 | |||
| 338 | /** |
||
| 339 | * Initialize session by visiting the site to establish cookies. |
||
| 340 | * HotMovies uses cookie-based age verification. |
||
| 341 | */ |
||
| 342 | protected function initializeSession(): void |
||
| 357 | } |
||
| 358 | } |
||
| 359 | } |
||
| 360 | |||
| 361 |
Loading history...