youtube_dl/extractor/streamcz.py

@@ -1,126 +1,105 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import json
+import hashlib
-import re
+import time
 
 from .common import InfoExtractor
 from ..utils import (
-    float_or_none,
     int_or_none,
-    merge_dicts,
+    sanitized_Request,
-    parse_codecs,
-    urljoin,
 )
 
 
+def _get_api_key(api_path):
+    if api_path.endswith('?'):
+        api_path = api_path[:-1]
+
+    api_key = 'fb5f58a820353bd7095de526253c14fd'
+    a = '{0:}{1:}{2:}'.format(api_key, api_path, int(round(time.time() / 24 / 3600)))
+    return hashlib.md5(a.encode('ascii')).hexdigest()
+
+
 class StreamCZIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?(?:stream|televizeseznam)\.cz/[^?#]+/(?P<display_id>[^?#]+)-(?P<id>[0-9]+)'
+    _VALID_URL = r'https?://(?:www\.)?stream\.cz/.+/(?P<id>[0-9]+)'
+    _API_URL = 'http://www.stream.cz/API'
+
     _TESTS = [{
-        'url': 'https://www.televizeseznam.cz/video/lajna/buh-57953890',
+        'url': 'http://www.stream.cz/peklonataliri/765767-ecka-pro-deti',
-        'md5': '40c41ade1464a390a0b447e333df4239',
+        'md5': '934bb6a6d220d99c010783c9719960d5',
         'info_dict': {
-            'id': '57953890',
+            'id': '765767',
             'ext': 'mp4',
-            'title': 'Bůh',
+            'title': 'Peklo na talíři: Éčka pro děti',
-            'display_id': 'buh',
+            'description': 'Taška s grónskou pomazánkou a další pekelnosti ZDE',
-            'description': 'md5:8f5f09b9b7bc67df910486cdd88f7165',
+            'thumbnail': 're:^http://im.stream.cz/episode/52961d7e19d423f8f06f0100',
-            'duration': 1369.6,
+            'duration': 256,
-            'view_count': int,
+        },
-        }
     }, {
-        'url': 'https://www.stream.cz/kdo-to-mluvi/kdo-to-mluvi-velke-odhaleni-prinasi-novy-porad-uz-od-25-srpna-64087937',
+        'url': 'http://www.stream.cz/blanik/10002447-tri-roky-pro-mazanka',
-        'md5': '41fd358000086a1ccdb068c77809b158',
+        'md5': '849a88c1e1ca47d41403c2ba5e59e261',
         'info_dict': {
-            'id': '64087937',
+            'id': '10002447',
             'ext': 'mp4',
-            'title': 'Kdo to mluví? Velké odhalení přináší nový pořad už od 25. srpna',
+            'title': 'Kancelář Blaník: Tři roky pro Mazánka',
-            'display_id': 'kdo-to-mluvi-velke-odhaleni-prinasi-novy-porad-uz-od-25-srpna',
+            'description': 'md5:3862a00ba7bf0b3e44806b544032c859',
-            'description': 'md5:97a811000a6460266029d6c1c2ebcd59',
+            'thumbnail': 're:^http://im.stream.cz/episode/537f838c50c11f8d21320000',
-            'duration': 50.2,
+            'duration': 368,
-            'view_count': int,
+        },
-        }
-    }, {
-        'url': 'https://www.stream.cz/tajemno/znicehonic-jim-skrz-strechu-prolitnul-zahadny-predmet-badatele-vse-objasnili-64147267',
-        'md5': '3ee4d0be040e8f4a543e67e509d55e3f',
-        'info_dict': {
-            'id': '64147267',
-            'ext': 'mp4',
-            'title': 'Zničehonic jim skrz střechu prolítnul záhadný předmět. Badatelé vše objasnili',
-            'display_id': 'znicehonic-jim-skrz-strechu-prolitnul-zahadny-predmet-badatele-vse-objasnili',
-            'description': 'md5:4b8ada6718d34bb011c4e04ca4bc19bf',
-            'duration': 442.84,
-            'view_count': int,
-        }
     }]
 
-    def _extract_formats(self, spl_url, video):
-        for ext, pref, streams in (
-                ('ts', -1, video.get('http_stream', {}).get('qualities', {})),
-                ('mp4', 1, video.get('mp4'))):
-            for format_id, stream in streams.items():
-                if not stream.get('url'):
-                    continue
-                yield merge_dicts({
-                    'format_id': '{}-{}'.format(format_id, ext),
-                    'ext': ext,
-                    'source_preference': pref,
-                    'url': urljoin(spl_url, stream['url']),
-                    'tbr': float_or_none(stream.get('bandwidth'), scale=1000),
-                    'duration': float_or_none(stream.get('duration'), scale=1000),
-                    'width': stream.get('resolution', 2 * [0])[0] or None,
-                    'height': stream.get('resolution', 2 * [0])[1] or int_or_none(format_id.replace('p', '')),
-                }, parse_codecs(stream.get('codec')))
-
     def _real_extract(self, url):
-        display_id, video_id = re.match(self._VALID_URL, url).groups()
+        video_id = self._match_id(url)
+        api_path = '/episode/%s' % video_id
 
-        data = self._download_json(
+        req = sanitized_Request(self._API_URL + api_path)
-            'https://www.televizeseznam.cz/api/graphql', video_id, 'Downloading GraphQL result',
+        req.add_header('Api-Password', _get_api_key(api_path))
-            data=json.dumps({
+        data = self._download_json(req, video_id)
-                'variables': {'urlName': video_id},
-                'query': '''
-                    query LoadEpisode($urlName : String){ episode(urlName: $urlName){ ...VideoDetailFragmentOnEpisode } }
-                    fragment VideoDetailFragmentOnEpisode on Episode {
-                        id
-                        spl
-                        urlName
-                        name
-                        perex
-                        duration
-                        views
-                    }'''
-            }).encode('utf-8'),
-            headers={'Content-Type': 'application/json;charset=UTF-8'}
-        )['data']['episode']
 
-        spl_url = data['spl'] + 'spl2,3'
+        formats = []
-        metadata = self._download_json(spl_url, video_id, 'Downloading playlist')
+        for quality, video in enumerate(data['video_qualities']):
-        if 'Location' in metadata and 'data' not in metadata:
+            for f in video['formats']:
-            spl_url = metadata['Location']
+                typ = f['type'].partition('/')[2]
-            metadata = self._download_json(spl_url, video_id, 'Downloading redirected playlist')
+                qlabel = video.get('quality_label')
-        video = metadata['data']
+                formats.append({
+                    'format_note': '%s-%s' % (qlabel, typ) if qlabel else typ,
+                    'format_id': '%s-%s' % (typ, f['quality']),
+                    'url': f['source'],
+                    'height': int_or_none(f['quality'].rstrip('p')),
+                    'quality': quality,
+                })
+        self._sort_formats(formats)
+
+        image = data.get('image')
+        if image:
+            thumbnail = self._proto_relative_url(
+                image.replace('{width}', '1240').replace('{height}', '697'),
+                scheme='http:',
+            )
+        else:
+            thumbnail = None
+
+        stream = data.get('_embedded', {}).get('stream:show', {}).get('name')
+        if stream:
+            title = '%s: %s' % (stream, data['name'])
+        else:
+            title = data['name']
 
         subtitles = {}
-        for subs in video.get('subtitles', {}).values():
+        srt_url = data.get('subtitles_srt')
-            if not subs.get('language'):
+        if srt_url:
-                continue
+            subtitles['cs'] = [{
-            for ext, sub_url in subs.get('urls').items():
+                'ext': 'srt',
-                subtitles.setdefault(subs['language'], []).append({
+                'url': srt_url,
-                    'ext': ext,
+            }]
-                    'url': urljoin(spl_url, sub_url)
-                })
-
-        formats = list(self._extract_formats(spl_url, video))
-        self._sort_formats(formats)
 
         return {
             'id': video_id,
-            'display_id': display_id,
+            'title': title,
-            'title': data.get('name'),
+            'thumbnail': thumbnail,
-            'description': data.get('perex'),
-            'duration': float_or_none(data.get('duration')),
-            'view_count': int_or_none(data.get('views')),
             'formats': formats,
+            'description': data.get('web_site_text'),
+            'duration': int_or_none(data.get('duration')),
+            'view_count': int_or_none(data.get('views')),
             'subtitles': subtitles,
         }