[tvigle] Fix extraction and add support for HLS and DASH formats (closes #21967)

youtube_dl/extractor/tvigle.py

@@ -9,6 +9,8 @@ from ..utils import (
     float_or_none,
     int_or_none,
     parse_age_limit,
+    try_get,
+    url_or_none,
 )
 
 
@@ -23,11 +25,10 @@ class TvigleIE(InfoExtractor):
     _TESTS = [
         {
             'url': 'http://www.tvigle.ru/video/sokrat/',
-            'md5': '36514aed3657d4f70b4b2cef8eb520cd',
             'info_dict': {
                 'id': '1848932',
                 'display_id': 'sokrat',
-                'ext': 'flv',
+                'ext': 'mp4',
                 'title': 'Сократ',
                 'description': 'md5:d6b92ffb7217b4b8ebad2e7665253c17',
                 'duration': 6586,
@@ -37,7 +38,6 @@ class TvigleIE(InfoExtractor):
         },
         {
             'url': 'http://www.tvigle.ru/video/vladimir-vysotskii/vedushchii-teleprogrammy-60-minut-ssha-o-vladimire-vysotskom/',
-            'md5': 'e7efe5350dd5011d0de6550b53c3ba7b',
             'info_dict': {
                 'id': '5142516',
                 'ext': 'flv',
@@ -62,7 +62,7 @@ class TvigleIE(InfoExtractor):
             webpage = self._download_webpage(url, display_id)
             video_id = self._html_search_regex(
                 (r'<div[^>]+class=["\']player["\'][^>]+id=["\'](\d+)',
-                 r'var\s+cloudId\s*=\s*["\'](\d+)',
+                 r'cloudId\s*=\s*["\'](\d+)',
                  r'class="video-preview current_playing" id="(\d+)"'),
                 webpage, 'video id')
 
@@ -90,21 +90,40 @@ class TvigleIE(InfoExtractor):
         age_limit = parse_age_limit(item.get('ageRestrictions'))
 
         formats = []
-        for vcodec, fmts in item['videos'].items():
+        for vcodec, url_or_fmts in item['videos'].items():
             if vcodec == 'hls':
-                continue
+                m3u8_url = url_or_none(url_or_fmts)
-            for format_id, video_url in fmts.items():
+                if not m3u8_url:
-                if format_id == 'm3u8':
                     continue
-                height = self._search_regex(
+                formats.extend(self._extract_m3u8_formats(
-                    r'^(\d+)[pP]$', format_id, 'height', default=None)
+                    m3u8_url, video_id, ext='mp4', entry_protocol='m3u8_native',
-                formats.append({
+                    m3u8_id='hls', fatal=False))
-                    'url': video_url,
+            elif vcodec == 'dash':
-                    'format_id': '%s-%s' % (vcodec, format_id),
+                mpd_url = url_or_none(url_or_fmts)
-                    'vcodec': vcodec,
+                if not mpd_url:
-                    'height': int_or_none(height),
+                    continue
-                    'filesize': int_or_none(item.get('video_files_size', {}).get(vcodec, {}).get(format_id)),
+                formats.extend(self._extract_mpd_formats(
-                })
+                    mpd_url, video_id, mpd_id='dash', fatal=False))
+            else:
+                if not isinstance(url_or_fmts, dict):
+                    continue
+                for format_id, video_url in url_or_fmts.items():
+                    if format_id == 'm3u8':
+                        continue
+                    video_url = url_or_none(video_url)
+                    if not video_url:
+                        continue
+                    height = self._search_regex(
+                        r'^(\d+)[pP]$', format_id, 'height', default=None)
+                    filesize = int_or_none(try_get(
+                        item, lambda x: x['video_files_size'][vcodec][format_id]))
+                    formats.append({
+                        'url': video_url,
+                        'format_id': '%s-%s' % (vcodec, format_id),
+                        'vcodec': vcodec,
+                        'height': int_or_none(height),
+                        'filesize': filesize,
+                    })
         self._sort_formats(formats)
 
         return {