[youtube] Fix extraction.

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

diff --git a/youtube_dl/extractor/firsttv.py b/youtube_dl/extractor/firsttv.py
index 4463d3d207bb18375a215c0b0900cda2957eca42..28617d83c7d2afe398184b758b60380446314fc2 100644 (file)
--- a/youtube_dl/extractor/firsttv.py
+++ b/youtube_dl/extractor/firsttv.py
@@ -10,6 +10,7 @@ from ..utils import (
     int_or_none,
     qualities,
     unified_strdate,
+    url_or_none,
 )
 
 
@@ -26,7 +27,7 @@ class FirstTVIE(InfoExtractor):
             'id': '40049',
             'ext': 'mp4',
             'title': 'Гость Людмила Сенчина. Наедине со всеми. Выпуск от 12.02.2015',
-            'thumbnail': 're:^https?://.*\.(?:jpg|JPG)$',
+            'thumbnail': r're:^https?://.*\.(?:jpg|JPG)$',
             'upload_date': '20150212',
             'duration': 2694,
         },
@@ -37,7 +38,7 @@ class FirstTVIE(InfoExtractor):
             'id': '364746',
             'ext': 'mp4',
             'title': 'Весенняя аллергия. Доброе утро. Фрагмент выпуска от 07.04.2016',
-            'thumbnail': 're:^https?://.*\.(?:jpg|JPG)$',
+            'thumbnail': r're:^https?://.*\.(?:jpg|JPG)$',
             'upload_date': '20160407',
             'duration': 179,
             'formats': 'mincount:3',
@@ -86,18 +87,45 @@ class FirstTVIE(InfoExtractor):
             title = item['title']
             quality = qualities(QUALITIES)
             formats = []
+            path = None
             for f in item.get('mbr', []):
-                src = f.get('src')
-                if not src or not isinstance(src, compat_str):
+                src = url_or_none(f.get('src'))
+                if not src:
                     continue
                 tbr = int_or_none(self._search_regex(
                     r'_(\d{3,})\.mp4', src, 'tbr', default=None))
+                if not path:
+                    path = self._search_regex(
+                        r'//[^/]+/(.+?)_\d+\.mp4', src,
+                        'm3u8 path', default=None)
                 formats.append({
                     'url': src,
                     'format_id': f.get('name'),
                     'tbr': tbr,
-                    'quality': quality(f.get('name')),
+                    'source_preference': quality(f.get('name')),
+                    # quality metadata of http formats may be incorrect
+                    'preference': -1,
                 })
+            # m3u8 URL format is reverse engineered from [1] (search for
+            # master.m3u8). dashEdges (that is currently balancer-vod.1tv.ru)
+            # is taken from [2].
+            # 1. http://static.1tv.ru/player/eump1tv-current/eump-1tv.all.min.js?rnd=9097422834:formatted
+            # 2. http://static.1tv.ru/player/eump1tv-config/config-main.js?rnd=9097422834
+            if not path and len(formats) == 1:
+                path = self._search_regex(
+                    r'//[^/]+/(.+?$)', formats[0]['url'],
+                    'm3u8 path', default=None)
+            if path:
+                if len(formats) == 1:
+                    m3u8_path = ','
+                else:
+                    tbrs = [compat_str(t) for t in sorted(f['tbr'] for f in formats)]
+                    m3u8_path = '_,%s,%s' % (','.join(tbrs), '.mp4')
+                formats.extend(self._extract_m3u8_formats(
+                    'http://balancer-vod.1tv.ru/%s%s.urlset/master.m3u8'
+                    % (path, m3u8_path),
+                    display_id, 'mp4',
+                    entry_protocol='m3u8_native', m3u8_id='hls', fatal=False))
             self._sort_formats(formats)
 
             thumbnail = item.get('poster') or self._og_search_thumbnail(webpage)
@@ -107,7 +135,7 @@ class FirstTVIE(InfoExtractor):
                 'ya:ovs:upload_date', webpage, 'upload date', default=None))
 
             entries.append({
-                'id': item.get('id') or uid,
+                'id': compat_str(item.get('id') or item['uid']),
                 'thumbnail': thumbnail,
                 'title': title,
                 'upload_date': upload_date,