[generic] Fix support for multiple HTML5 videos on one page (closes #14080)

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

diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index d2fb2627db856cdf4be849a6ceee2c007ac36dcc..7d0edf09c6a2e8428b84b3e7ebf787738db1bfaa 100644 (file)
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -1519,14 +1519,27 @@ class GenericIE(InfoExtractor):
         # LiveLeak embed
         {
             'url': 'http://www.wykop.pl/link/3088787/',
-            'md5': 'ace83b9ed19b21f68e1b50e844fdf95d',
+            'md5': '7619da8c820e835bef21a1efa2a0fc71',
             'info_dict': {
                 'id': '874_1459135191',
                 'ext': 'mp4',
                 'title': 'Man shows poor quality of new apartment building',
                 'description': 'The wall is like a sand pile.',
                 'uploader': 'Lake8737',
-            }
+            },
+            'add_ie': [LiveLeakIE.ie_key()],
+        },
+        # Another LiveLeak embed pattern (#13336)
+        {
+            'url': 'https://milo.yiannopoulos.net/2017/06/concealed-carry-robbery/',
+            'info_dict': {
+                'id': '2eb_1496309988',
+                'ext': 'mp4',
+                'title': 'Thief robs place where everyone was armed',
+                'description': 'md5:694d73ee79e535953cf2488562288eee',
+                'uploader': 'brazilwtf',
+            },
+            'add_ie': [LiveLeakIE.ie_key()],
         },
         # Duplicated embedded video URLs
         {
@@ -1866,6 +1879,15 @@ class GenericIE(InfoExtractor):
                 'title': 'Building A Business Online: Principal Chairs Q & A',
             },
         },
+        {
+            # multiple HTML5 videos on one page
+            'url': 'https://www.paragon-software.com/home/rk-free/keyscenarios.html',
+            'info_dict': {
+                'id': 'keyscenarios',
+                'title': 'Rescue Kit 14 Free Edition - Getting started',
+            },
+            'playlist_count': 4,
+        }
         # {
         #     # TODO: find another test
         #     # http://schema.org/VideoObject
@@ -2230,36 +2252,11 @@ class GenericIE(InfoExtractor):
         if vid_me_embed_url is not None:
             return self.url_result(vid_me_embed_url, 'Vidme')
 
-        # Look for embedded YouTube player
-        matches = re.findall(r'''(?x)
-            (?:
-                <iframe[^>]+?src=|
-                data-video-url=|
-                <embed[^>]+?src=|
-                embedSWF\(?:\s*|
-                <object[^>]+data=|
-                new\s+SWFObject\(
-            )
-            (["\'])
-                (?P<url>(?:https?:)?//(?:www\.)?youtube(?:-nocookie)?\.com/
-                (?:embed|v|p)/.+?)
-            \1''', webpage)
-        if matches:
+        # Look for YouTube embeds
+        youtube_urls = YoutubeIE._extract_urls(webpage)
+        if youtube_urls:
             return self.playlist_from_matches(
-                matches, video_id, video_title, lambda m: unescapeHTML(m[1]))
-
-        # Look for lazyYT YouTube embed
-        matches = re.findall(
-            r'class="lazyYT" data-youtube-id="([^"]+)"', webpage)
-        if matches:
-            return self.playlist_from_matches(matches, video_id, video_title, lambda m: unescapeHTML(m))
-
-        # Look for Wordpress "YouTube Video Importer" plugin
-        matches = re.findall(r'''(?x)<div[^>]+
-            class=(?P<q1>[\'"])[^\'"]*\byvii_single_video_player\b[^\'"]*(?P=q1)[^>]+
-            data-video_id=(?P<q2>[\'"])([^\'"]+)(?P=q2)''', webpage)
-        if matches:
-            return self.playlist_from_matches(matches, video_id, video_title, lambda m: m[-1])
+                youtube_urls, video_id, video_title, ie=YoutubeIE.ie_key())
 
         matches = DailymotionIE._extract_urls(webpage)
         if matches:
@@ -2757,9 +2754,9 @@ class GenericIE(InfoExtractor):
                 self._proto_relative_url(instagram_embed_url), InstagramIE.ie_key())
 
         # Look for LiveLeak embeds
-        liveleak_url = LiveLeakIE._extract_url(webpage)
-        if liveleak_url:
-            return self.url_result(liveleak_url, 'LiveLeak')
+        liveleak_urls = LiveLeakIE._extract_urls(webpage)
+        if liveleak_urls:
+            return self.playlist_from_matches(liveleak_urls, video_id, video_title)
 
         # Look for 3Q SDN embeds
         threeqsdn_url = ThreeQSDNIE._extract_url(webpage)
@@ -2858,22 +2855,23 @@ class GenericIE(InfoExtractor):
                     merged[k] = v
             return merged
 
-        # Looking for http://schema.org/VideoObject
-        json_ld = self._search_json_ld(
-            webpage, video_id, default={}, expected_type='VideoObject')
-        if json_ld.get('url'):
-            return merge_dicts(json_ld, info_dict)
-
         # Look for HTML5 media
         entries = self._parse_html5_media_entries(url, webpage, video_id, m3u8_id='hls')
         if entries:
-            for entry in entries:
-                entry.update({
+            if len(entries) == 1:
+                entries[0].update({
                     'id': video_id,
                     'title': video_title,
                 })
+            else:
+                for num, entry in enumerate(entries, start=1):
+                    entry.update({
+                        'id': '%s-%s' % (video_id, num),
+                        'title': '%s (%d)' % (video_title, num),
+                    })
+            for entry in entries:
                 self._sort_formats(entry['formats'])
-            return self.playlist_result(entries)
+            return self.playlist_result(entries, video_id, video_title)
 
         jwplayer_data = self._find_jwplayer_data(
             webpage, video_id, transform_source=js_to_json)
@@ -2882,6 +2880,12 @@ class GenericIE(InfoExtractor):
                 jwplayer_data, video_id, require_title=False, base_url=url)
             return merge_dicts(info, info_dict)
 
+        # Looking for http://schema.org/VideoObject
+        json_ld = self._search_json_ld(
+            webpage, video_id, default={}, expected_type='VideoObject')
+        if json_ld.get('url'):
+            return merge_dicts(json_ld, info_dict)
+
         def check_video(vurl):
             if YoutubeIE.suitable(vurl):
                 return True