[dvtv] Improve and fix playlists support (closes #13063)
authorSergey M․ <dstftw@gmail.com>
Sat, 27 May 2017 23:19:21 +0000 (06:19 +0700)
committerSergey M․ <dstftw@gmail.com>
Sat, 27 May 2017 23:19:54 +0000 (06:19 +0700)
youtube_dl/extractor/dvtv.py

index b1ae37d786a13989a5bface40000fe40f64c2992..e85c58bd5669bd14cb069690d43277f77cea49d3 100644 (file)
@@ -5,9 +5,12 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
+    ExtractorError,
+    int_or_none,
     js_to_json,
+    mimetype2ext,
     unescapeHTML,
-    ExtractorError,
 )
 
 
@@ -24,14 +27,7 @@ class DVTVIE(InfoExtractor):
             'id': 'dc0768de855511e49e4b0025900fea04',
             'ext': 'mp4',
             'title': 'Vondra o Českém století: Při pohledu na Havla mi bylo trapně',
-        }
-    }, {
-        'url': 'http://video.aktualne.cz/dvtv/stropnicky-policie-vrbetice-preventivne-nekontrolovala/r~82ed4322849211e4a10c0025900fea04/',
-        'md5': '6388f1941b48537dbd28791f712af8bf',
-        'info_dict': {
-            'id': '72c02230849211e49f60002590604f2e',
-            'ext': 'mp4',
-            'title': 'Stropnický: Policie Vrbětice preventivně nekontrolovala',
+            'duration': 1484,
         }
     }, {
         'url': 'http://video.aktualne.cz/dvtv/dvtv-16-12-2014-utok-talibanu-boj-o-kliniku-uprchlici/r~973eb3bc854e11e498be002590604f2e/',
@@ -44,72 +40,100 @@ class DVTVIE(InfoExtractor):
             'info_dict': {
                 'id': 'b0b40906854d11e4bdad0025900fea04',
                 'ext': 'mp4',
-                'title': 'Drtinová Veselovský TV 16. 12. 2014: Témata dne'
+                'title': 'Drtinová Veselovský TV 16. 12. 2014: Témata dne',
+                'description': 'md5:0916925dea8e30fe84222582280b47a0',
+                'timestamp': 1418760010,
+                'upload_date': '20141216',
             }
         }, {
             'md5': '5f7652a08b05009c1292317b449ffea2',
             'info_dict': {
                 'id': '420ad9ec854a11e4bdad0025900fea04',
                 'ext': 'mp4',
-                'title': 'Školní masakr možná změní boj s Talibanem, říká novinářka'
+                'title': 'Školní masakr možná změní boj s Talibanem, říká novinářka',
+                'description': 'md5:ff2f9f6de73c73d7cef4f756c1c1af42',
+                'timestamp': 1418760010,
+                'upload_date': '20141216',
             }
         }, {
             'md5': '498eb9dfa97169f409126c617e2a3d64',
             'info_dict': {
                 'id': '95d35580846a11e4b6d20025900fea04',
                 'ext': 'mp4',
-                'title': 'Boj o kliniku: Veřejný zájem, nebo právo na majetek?'
+                'title': 'Boj o kliniku: Veřejný zájem, nebo právo na majetek?',
+                'description': 'md5:889fe610a70fee5511dc3326a089188e',
+                'timestamp': 1418760010,
+                'upload_date': '20141216',
             }
         }, {
             'md5': 'b8dc6b744844032dab6ba3781a7274b9',
             'info_dict': {
                 'id': '6fe14d66853511e4833a0025900fea04',
                 'ext': 'mp4',
-                'title': 'Pánek: Odmítání syrských uprchlíků je ostudou české vlády'
+                'title': 'Pánek: Odmítání syrských uprchlíků je ostudou české vlády',
+                'description': 'md5:544f86de6d20c4815bea11bf2ac3004f',
+                'timestamp': 1418760010,
+                'upload_date': '20141216',
             }
         }],
-    }, {
-        'url': 'http://video.aktualne.cz/v-cechach-poprve-zazni-zelenkova-zrestaurovana-mse/r~45b4b00483ec11e4883b002590604f2e/',
-        'only_matching': True,
     }, {
         'url': 'https://video.aktualne.cz/dvtv/zeman-si-jen-leci-mindraky-sobotku-nenavidi-a-babis-se-mu-te/r~960cdb3a365a11e7a83b0025900fea04/',
         'md5': 'f8efe9656017da948369aa099788c8ea',
         'info_dict': {
             'id': '3c496fec365911e7a6500025900fea04',
-            'ext': 'm3u8',
+            'ext': 'mp4',
             'title': 'Zeman si jen léčí mindráky, Sobotku nenávidí a Babiš se mu teď hodí, tvrdí Kmenta',
-        }
+            'duration': 1103,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        'url': 'http://video.aktualne.cz/v-cechach-poprve-zazni-zelenkova-zrestaurovana-mse/r~45b4b00483ec11e4883b002590604f2e/',
+        'only_matching': True,
     }]
 
     def _parse_video_metadata(self, js, video_id):
-        metadata = self._parse_json(js, video_id, transform_source=js_to_json)
+        data = self._parse_json(js, video_id, transform_source=js_to_json)
+
+        title = unescapeHTML(data['title'])
 
         formats = []
-        for video in metadata['sources']:
-            ext = video['type'][6:]
-            if video['label'] != 'adaptive':
+        for video in data['sources']:
+            video_url = video.get('file')
+            if not video_url:
+                continue
+            video_type = video.get('type')
+            ext = determine_ext(video_url, mimetype2ext(video_type))
+            if video_type == 'application/vnd.apple.mpegurl' or ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    video_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id='hls', fatal=False))
+            elif video_type == 'application/dash+xml' or ext == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    video_url, video_id, mpd_id='dash', fatal=False))
+            else:
+                label = video.get('label')
+                height = self._search_regex(
+                    r'^(\d+)[pP]', label or '', 'height', default=None)
+                format_id = ['http']
+                for f in (ext, label):
+                    if f:
+                        format_id.append(f)
                 formats.append({
-                    'url': video['file'],
-                    'ext': ext,
-                    'format_id': '%s-%s' % (ext, video['label']),
-                    'height': int(video['label'].rstrip('p')),
-                    'fps': 25,
+                    'url': video_url,
+                    'format_id': '-'.join(format_id),
+                    'height': int_or_none(height),
                 })
-            elif video['type'] == 'application/dash+xml':
-                formats.extend(self._extract_mpd_formats(video['file'],
-                                                         video_id,
-                                                         fatal=False))
-            elif video['type'] == 'application/vnd.apple.mpegurl':
-                formats.extend(self._extract_m3u8_formats(video['file'],
-                                                          video_id,
-                                                          fatal=False))
-
         self._sort_formats(formats)
 
         return {
-            'id': metadata['mediaid'],
-            'title': unescapeHTML(metadata['title']),
-            'thumbnail': self._proto_relative_url(metadata['image'], 'http:'),
+            'id': data.get('mediaid') or video_id,
+            'title': title,
+            'description': data.get('description'),
+            'thumbnail': data.get('image'),
+            'duration': int_or_none(data.get('duration')),
+            'timestamp': int_or_none(data.get('pubtime')),
             'formats': formats
         }
 
@@ -120,7 +144,7 @@ class DVTVIE(InfoExtractor):
 
         # single video
         item = self._search_regex(
-            r"(?s)embedData[0-9a-f]{32}\['asset'\]\s*=\s*(\{.+?\});",
+            r'(?s)embedData[0-9a-f]{32}\[["\']asset["\']\]\s*=\s*(\{.+?\});',
             webpage, 'video', default=None, fatal=False)
 
         if item:
@@ -130,6 +154,8 @@ class DVTVIE(InfoExtractor):
         items = re.findall(
             r"(?s)BBX\.context\.assets\['[0-9a-f]{32}'\]\.push\(({.+?})\);",
             webpage)
+        if not items:
+            items = re.findall(r'(?s)var\s+asset\s*=\s*({.+?});\n', webpage)
 
         if items:
             return {