[crackle] Fix extraction (closes #15969)

youtube_dl/extractor/crackle.py

@@ -1,31 +1,41 @@
 # coding: utf-8
 from __future__ import unicode_literals, division
 
+import re
+
 from .common import InfoExtractor
-from ..utils import int_or_none
+from ..compat import compat_str
+from ..utils import (
+    determine_ext,
+    float_or_none,
+    int_or_none,
+    parse_age_limit,
+    parse_duration,
+)
 
 
 class CrackleIE(InfoExtractor):
     _GEO_COUNTRIES = ['US']
     _VALID_URL = r'(?:crackle:|https?://(?:(?:www|m)\.)?crackle\.com/(?:playlist/\d+/|(?:[^/]+/)+))(?P<id>\d+)'
     _TEST = {
-        'url': 'http://www.crackle.com/comedians-in-cars-getting-coffee/2498934',
+        'url': 'https://www.crackle.com/andromeda/2502343',
         'info_dict': {
-            'id': '2498934',
+            'id': '2502343',
             'ext': 'mp4',
-            'title': 'Everybody Respects A Bloody Nose',
+            'title': 'Under The Night',
-            'description': 'Jerry is kaffeeklatsching in L.A. with funnyman J.B. Smoove (Saturday Night Live, Real Husbands of Hollywood). They’re headed for brew at 10 Speed Coffee in a 1964 Studebaker Avanti.',
+            'description': 'md5:d2b8ca816579ae8a7bf28bfff8cefc8a',
-            'thumbnail': r're:^https?://.*\.jpg',
+            'duration': 2583,
-            'duration': 906,
+            'view_count': int,
-            'series': 'Comedians In Cars Getting Coffee',
+            'average_rating': 0,
-            'season_number': 8,
+            'age_limit': 14,
-            'episode_number': 4,
+            'genre': 'Action, Sci-Fi',
-            'subtitles': {
+            'creator': 'Allan Kroeker',
-                'en-US': [
+            'artist': 'Keith Hamilton Cobb, Kevin Sorbo, Lisa Ryder, Lexa Doig, Robert Hewitt Wolfe',
-                    {'ext': 'vtt'},
+            'release_year': 2000,
-                    {'ext': 'tt'},
+            'series': 'Andromeda',
-                ]
+            'episode': 'Under The Night',
-            },
+            'season_number': 1,
+            'episode_number': 1,
         },
         'params': {
             # m3u8 download
@@ -33,108 +43,95 @@ class CrackleIE(InfoExtractor):
         }
     }
 
-    _THUMBNAIL_RES = [
-        (120, 90),
-        (208, 156),
-        (220, 124),
-        (220, 220),
-        (240, 180),
-        (250, 141),
-        (315, 236),
-        (320, 180),
-        (360, 203),
-        (400, 300),
-        (421, 316),
-        (460, 330),
-        (460, 460),
-        (462, 260),
-        (480, 270),
-        (587, 330),
-        (640, 480),
-        (700, 330),
-        (700, 394),
-        (854, 480),
-        (1024, 1024),
-        (1920, 1080),
-    ]
-
-    # extracted from http://legacyweb-us.crackle.com/flash/ReferrerRedirect.ashx
-    _MEDIA_FILE_SLOTS = {
-        'c544.flv': {
-            'width': 544,
-            'height': 306,
-        },
-        '360p.mp4': {
-            'width': 640,
-            'height': 360,
-        },
-        '480p.mp4': {
-            'width': 852,
-            'height': 478,
-        },
-        '480p_1mbps.mp4': {
-            'width': 852,
-            'height': 478,
-        },
-    }
-
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        config_doc = self._download_xml(
+        media = self._download_json(
-            'http://legacyweb-us.crackle.com/flash/QueryReferrer.ashx?site=16',
+            'https://web-api-us.crackle.com/Service.svc/details/media/%s/%s'
-            video_id, 'Downloading config')
+            % (video_id, self._GEO_COUNTRIES[0]), video_id, query={
+                'disableProtocols': 'true',
+                'format': 'json'
+            })
 
-        item = self._download_xml(
+        title = media['Title']
-            'http://legacyweb-us.crackle.com/app/revamp/vidwallcache.aspx?flags=-1&fm=%s' % video_id,
+
-            video_id, headers=self.geo_verification_headers()).find('i')
+        formats = []
-        title = item.attrib['t']
+        for e in media['MediaURLs']:
+            if e.get('UseDRM') is True:
+                continue
+            format_url = e.get('Path')
+            if not format_url or not isinstance(format_url, compat_str):
+                continue
+            ext = determine_ext(format_url)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id='hls', fatal=False))
+            elif ext == 'mpd':
+                formats.extend(self._extract_mpd_formats(
+                    format_url, video_id, mpd_id='dash', fatal=False))
+        self._sort_formats(formats)
+
+        description = media.get('Description')
+        duration = int_or_none(media.get(
+            'DurationInSeconds')) or parse_duration(media.get('Duration'))
+        view_count = int_or_none(media.get('CountViews'))
+        average_rating = float_or_none(media.get('UserRating'))
+        age_limit = parse_age_limit(media.get('Rating'))
+        genre = media.get('Genre')
+        release_year = int_or_none(media.get('ReleaseYear'))
+        creator = media.get('Directors')
+        artist = media.get('Cast')
+
+        if media.get('MediaTypeDisplayValue') == 'Full Episode':
+            series = media.get('ShowName')
+            episode = title
+            season_number = int_or_none(media.get('Season'))
+            episode_number = int_or_none(media.get('Episode'))
+        else:
+            series = episode = season_number = episode_number = None
 
         subtitles = {}
-        formats = self._extract_m3u8_formats(
+        cc_files = media.get('ClosedCaptionFiles')
-            'http://content.uplynk.com/ext/%s/%s.m3u8' % (config_doc.attrib['strUplynkOwnerId'], video_id),
+        if isinstance(cc_files, list):
-            video_id, 'mp4', m3u8_id='hls', fatal=None)
+            for cc_file in cc_files:
+                if not isinstance(cc_file, dict):
+                    continue
+                cc_url = cc_file.get('Path')
+                if not cc_url or not isinstance(cc_url, compat_str):
+                    continue
+                lang = cc_file.get('Locale') or 'en'
+                subtitles.setdefault(lang, []).append({'url': cc_url})
+
         thumbnails = []
-        path = item.attrib.get('p')
+        images = media.get('Images')
-        if path:
+        if isinstance(images, list):
-            for width, height in self._THUMBNAIL_RES:
+            for image_key, image_url in images.items():
-                res = '%dx%d' % (width, height)
+                mobj = re.search(r'Img_(\d+)[xX](\d+)', image_key)
+                if not mobj:
+                    continue
                 thumbnails.append({
-                    'id': res,
+                    'url': image_url,
-                    'url': 'http://images-us-am.crackle.com/%stnl_%s.jpg' % (path, res),
+                    'width': int(mobj.group(1)),
-                    'width': width,
+                    'height': int(mobj.group(2)),
-                    'height': height,
-                    'resolution': res,
                 })
-            http_base_url = 'http://ahttp.crackle.com/' + path
-            for mfs_path, mfs_info in self._MEDIA_FILE_SLOTS.items():
-                formats.append({
-                    'url': http_base_url + mfs_path,
-                    'format_id': 'http-' + mfs_path.split('.')[0],
-                    'width': mfs_info['width'],
-                    'height': mfs_info['height'],
-                })
-            for cc in item.findall('cc'):
-                locale = cc.attrib.get('l')
-                v = cc.attrib.get('v')
-                if locale and v:
-                    if locale not in subtitles:
-                        subtitles[locale] = []
-                    for url_ext, ext in (('vtt', 'vtt'), ('xml', 'tt')):
-                        subtitles.setdefault(locale, []).append({
-                            'url': '%s/%s%s_%s.%s' % (config_doc.attrib['strSubtitleServer'], path, locale, v, url_ext),
-                            'ext': ext,
-                        })
-        self._sort_formats(formats, ('width', 'height', 'tbr', 'format_id'))
 
         return {
             'id': video_id,
             'title': title,
-            'description': item.attrib.get('d'),
+            'description': description,
-            'duration': int(item.attrib.get('r'), 16) / 1000 if item.attrib.get('r') else None,
+            'duration': duration,
-            'series': item.attrib.get('sn'),
+            'view_count': view_count,
-            'season_number': int_or_none(item.attrib.get('se')),
+            'average_rating': average_rating,
-            'episode_number': int_or_none(item.attrib.get('ep')),
+            'age_limit': age_limit,
+            'genre': genre,
+            'creator': creator,
+            'artist': artist,
+            'release_year': release_year,
+            'series': series,
+            'episode': episode,
+            'season_number': season_number,
+            'episode_number': episode_number,
             'thumbnails': thumbnails,
             'subtitles': subtitles,
             'formats': formats,