[vidzi] Fix extraction

[youtube-dl] / youtube_dl / extractor / twitter.py

diff --git a/youtube_dl/extractor/twitter.py b/youtube_dl/extractor/twitter.py
index 64c429f02c61b1a1fea49c7900e93757740bc28d..5d2b5ec3515277980f2da45a91a0c73fae17923d 100644 (file)
--- a/youtube_dl/extractor/twitter.py
+++ b/youtube_dl/extractor/twitter.py
@@ -165,6 +165,7 @@ class TwitterIE(InfoExtractor):
             'uploader': 'Gifs',
             'uploader_id': 'giphz',
         },
+        'expected_warnings': ['height', 'width'],
     }, {
         'url': 'https://twitter.com/starwars/status/665052190608723968',
         'md5': '39b7199856dee6cd4432e72c74bc69d4',
@@ -212,24 +213,28 @@ class TwitterIE(InfoExtractor):
             return info
 
         mobj = re.search(r'''(?x)
-            <video[^>]+class="animated-gif"[^>]+
-                (?:data-height="(?P<height>\d+)")?[^>]+
-                (?:data-width="(?P<width>\d+)")?[^>]+
-                (?:poster="(?P<poster>[^"]+)")?[^>]*>\s*
+            <video[^>]+class="animated-gif"(?P<more_info>[^>]+)>\s*
                 <source[^>]+video-src="(?P<url>[^"]+)"
         ''', webpage)
 
         if mobj:
+            more_info = mobj.group('more_info')
+            height = int_or_none(self._search_regex(
+                r'data-height="(\d+)"', more_info, 'height', fatal=False))
+            width = int_or_none(self._search_regex(
+                r'data-width="(\d+)"', more_info, 'width', fatal=False))
+            thumbnail = self._search_regex(
+                r'poster="([^"]+)"', more_info, 'poster', fatal=False)
             info.update({
                 'id': twid,
                 'url': mobj.group('url'),
-                'height': int_or_none(mobj.group('height')),
-                'width': int_or_none(mobj.group('width')),
-                'thumbnail': mobj.group('poster'),
+                'height': height,
+                'width': width,
+                'thumbnail': thumbnail,
             })
             return info
 
-        raise ExtractorError('There\'s not video in this tweet.')
+        raise ExtractorError('There\'s no video in this tweet.')
 
 
 class TwitterAmplifyIE(TwitterBaseIE):
@@ -243,6 +248,7 @@ class TwitterAmplifyIE(TwitterBaseIE):
             'id': '0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
             'ext': 'mp4',
             'title': 'Twitter Video',
+            'thumbnail': 're:^https?://.*',
         },
     }
 
@@ -254,8 +260,35 @@ class TwitterAmplifyIE(TwitterBaseIE):
             'twitter:amplify:vmap', webpage, 'vmap url')
         video_url = self._get_vmap_video_url(vmap_url, video_id)
 
+        thumbnails = []
+        thumbnail = self._html_search_meta(
+            'twitter:image:src', webpage, 'thumbnail', fatal=False)
+
+        def _find_dimension(target):
+            w = int_or_none(self._html_search_meta(
+                'twitter:%s:width' % target, webpage, fatal=False))
+            h = int_or_none(self._html_search_meta(
+                'twitter:%s:height' % target, webpage, fatal=False))
+            return w, h
+
+        if thumbnail:
+            thumbnail_w, thumbnail_h = _find_dimension('image')
+            thumbnails.append({
+                'url': thumbnail,
+                'width': thumbnail_w,
+                'height': thumbnail_h,
+            })
+
+        video_w, video_h = _find_dimension('player')
+        formats = [{
+            'url': video_url,
+            'width': video_w,
+            'height': video_h,
+        }]
+
         return {
             'id': video_id,
             'title': 'Twitter Video',
-            'url': video_url,
+            'formats': formats,
+            'thumbnails': thumbnails,
         }