Merge branch 'develop' into feature/resolver-dlsite-affiliate

app/MetadataResolver/IwaraResolver.php

@@ -3,6 +3,7 @@
 namespace App\MetadataResolver;
 
 use GuzzleHttp\Client;
+use Symfony\Component\DomCrawler\Crawler;
 
 class IwaraResolver implements Resolver
 {
@@ -19,46 +20,40 @@ class IwaraResolver implements Resolver
     public function resolve(string $url): Metadata
     {
         $res = $this->client->get($url);
-
         if ($res->getStatusCode() === 200) {
-            $dom = new \DOMDocument();
-            @$dom->loadHTML(mb_convert_encoding($res->getBody(), 'HTML-ENTITIES', 'UTF-8'));
-            $xpath = new \DOMXPath($dom);
-
             $metadata = new Metadata();
+            $html = (string) $res->getBody();
+            $crawler = new Crawler($html);
 
-            // find title
-            foreach ($xpath->query('//title') as $node) {
-                $content = $node->textContent;
-                if (!empty($content)) {
-                    $metadata->title = $content;
-                    break;
+            $infoElements = $crawler->filter('#video-player + div, .field-name-field-video-url + div, .field-name-field-images + div');
+            $title = $infoElements->filter('h1.title')->text();
+            $author = $infoElements->filter('.username')->text();
+            $description = $infoElements->filter('.field-type-text-with-summary')->text('');
+            $tags =  $infoElements->filter('a[href^="/videos"], a[href^="/images"]')->extract('_text');
+            // 役に立たないタグを削除する
+            $tags = array_values(array_diff($tags, ['Uncategorized', 'Other']));
+            array_push($tags, $author);
+
+            $metadata->title = $title;
+            $metadata->description = '投稿者: ' . $author . PHP_EOL . $description;
+            $metadata->tags = $tags;
+
+            // iwara video
+            if ($crawler->filter('#video-player')->count()) {
+                $metadata->image = 'https:' . $crawler->filter('#video-player')->attr('poster');
+            }
+
+            // youtube
+            if ($crawler->filter('iframe[src^="//www.youtube.com"]')->count()) {
+                if (preg_match('~youtube\.com/embed/(\S+)\?~', $crawler->filter('iframe[src^="//www.youtube.com"]')->attr('src'), $matches) === 1) {
+                    $youtubeId = $matches[1];
+                    $metadata->image = 'https://img.youtube.com/vi/' . $youtubeId . '/maxresdefault.jpg';
                 }
             }
 
-            // find thumbnail
-            foreach ($xpath->query('//*[@id="video-player"]') as $node) {
-                $poster = $node->getAttribute('poster');
-                if (!empty($poster)) {
-                    if (strpos($poster, '//') === 0) {
-                        $poster = 'https:' . $poster;
-                    }
-                    $metadata->image = $poster;
-                    break;
-                }
-            }
-            if (empty($metadata->image)) {
-                // YouTube embedded?
-                foreach ($xpath->query('//div[@class="embedded-video"]//iframe') as $node) {
-                    $src = $node->getAttribute('src');
-                    if (preg_match('~youtube\.com/embed/(\S+)\?~', $src, $matches) !== -1) {
-                        $youtubeId = $matches[1];
-                        $iwaraThumbUrl = 'https://i.iwara.tv/sites/default/files/styles/thumbnail/public/video_embed_field_thumbnails/youtube/' . $youtubeId . '.jpg';
-
-                        $metadata->image = $iwaraThumbUrl;
-                        break;
-                    }
-                }
+            // images
+            if ($crawler->filter('.field-name-field-images')->count()) {
+                $metadata->image = 'https:' . $crawler->filter('.field-name-field-images a')->first()->attr('href');
             }
 
             return $metadata;

app/MetadataResolver/MetadataResolver.php

@@ -15,9 +15,9 @@ class MetadataResolver implements Resolver
         '~www\.melonbooks\.co\.jp/detail/detail\.php~' => MelonbooksResolver::class,
         '~ec\.toranoana\.(jp|shop)/(tora|joshi)(_[rd]+)?/(ec|digi)/item/~' => ToranoanaResolver::class,
         '~iwara\.tv/videos/.*~' => IwaraResolver::class,
-        '~www\.dlsite\.com/.+/(work|announce)/=/product_id/..\d+(\.html)?~' => DLsiteResolver::class,
-        '~www\.dlsite\.com/.+/dlaf/=/link/(work|announce)/aid/.+/..\d+(\.html)?~' => DLsiteResolver::class,
-        '~www\.dlsite\.com/.+/dlaf/=/aid/.+/url/.+~' => DLsiteResolver::class,
+        '~www\.dlsite\.com/.*/(work|announce)/=/product_id/..\d+(\.html)?~' => DLsiteResolver::class,
+        '~www\.dlsite\.com/.*/dlaf/=/link/(work|announce)/aid/.+/..\d+(\.html)?~' => DLsiteResolver::class,
+        '~www\.dlsite\.com/.*/dlaf/=/aid/.+/url/.+~' => DLsiteResolver::class,
         '~dlsite\.jp/...tw/..\d+~' => DLsiteResolver::class,
         '~www\.pixiv\.net/member_illust\.php\?illust_id=\d+~' => PixivResolver::class,
         '~www\.pixiv\.net/user/\d+/series/\d+~' => PixivResolver::class,

app/MetadataResolver/NijieResolver.php

@@ -3,6 +3,7 @@
 namespace App\MetadataResolver;
 
 use GuzzleHttp\Client;
+use Symfony\Component\DomCrawler\Crawler;
 
 class NijieResolver implements Resolver
 {
@@ -30,23 +31,29 @@ class NijieResolver implements Resolver
             $url = preg_replace('~view_popup\.php~', 'view.php', $url);
         }
 
-        $client = $this->client;
-        $res = $client->get($url);
+        $res =  $this->client->get($url);
         if ($res->getStatusCode() === 200) {
-            $metadata = $this->ogpResolver->parse($res->getBody());
+            $html = (string) $res->getBody();
+            $metadata = $this->ogpResolver->parse($html);
+            $crawler = new Crawler($html);
 
-            $dom = new \DOMDocument();
-            @$dom->loadHTML(mb_convert_encoding($res->getBody(), 'HTML-ENTITIES', 'UTF-8'));
-            $xpath = new \DOMXPath($dom);
-            $dataNode = $xpath->query('//script[substring(@type, string-length(@type) - 3, 4) = "json"]');
-            foreach ($dataNode as $node) {
-                // 改行がそのまま入っていることがあるのでデコード前にエスケープが必要
-                $imageData = json_decode(preg_replace('/\r?\n/', '\n', $node->nodeValue), true);
-                if (isset($imageData['thumbnailUrl']) && !ends_with($imageData['thumbnailUrl'], '.gif') && !ends_with($imageData['thumbnailUrl'], '.mp4')) {
-                    $metadata->image = preg_replace('~nijie\\.info/.*/nijie_picture/~', 'nijie.info/nijie_picture/', $imageData['thumbnailUrl']);
-                    break;
-                }
+            // DomCrawler内でjson内の日本語がHTMLエンティティに変換されるのでhtml_entity_decode
+            $json = html_entity_decode($crawler->filter('script[type="application/ld+json"]')->first()->text());
+
+            // 改行がそのまま入っていることがあるのでデコード前にエスケープが必要
+            $data = json_decode(preg_replace('/\r?\n/', '\n', $json), true);
+
+            $metadata->title = $data['name'];
+            $metadata->description = '投稿者: ' . $data['author']['name'] . PHP_EOL . $data['description'];
+            if (
+                isset($data['thumbnailUrl']) &&
+                !ends_with($data['thumbnailUrl'], '.gif') &&
+                !ends_with($data['thumbnailUrl'], '.mp4')
+            ) {
+                // サムネイルからメイン画像に
+                $metadata->image = str_replace('__rs_l160x160/', '', $data['thumbnailUrl']);
             }
+            $metadata->tags = $crawler->filter('#view-tag span.tag_name')->extract('_text');
 
             return $metadata;
         } else {