[crackle] Improve extraction (See desc)

Closes #282

* Refactor authorization as an extension to `_download_json`
* Better error messages and warnings
* Respect `--ignore-no-formats-error`
* Extract subtitles from manifests
* Try with crackle's geo-location service if all hard-coded countries fail

yt_dlp/extractor/crackle.py

@@ -12,6 +12,7 @@
     determine_ext,
     float_or_none,
     int_or_none,
+    orderedSet,
     parse_age_limit,
     parse_duration,
     url_or_none,
@@ -66,59 +67,104 @@ class CrackleIE(InfoExtractor):
         },
     }
 
+    def _download_json(self, url, *args, **kwargs):
+        # Authorization generation algorithm is reverse engineered from:
+        # https://www.sonycrackle.com/static/js/main.ea93451f.chunk.js
+        timestamp = time.strftime('%Y%m%d%H%M', time.gmtime())
+        h = hmac.new(b'IGSLUQCBDFHEOIFM', '|'.join([url, timestamp]).encode(), hashlib.sha1).hexdigest().upper()
+        headers = {
+            'Accept': 'application/json',
+            'Authorization': '|'.join([h, timestamp, '117', '1']),
+        }
+        return InfoExtractor._download_json(self, url, *args, headers=headers, **kwargs)
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        country_code = self._downloader.params.get('geo_bypass_country', None)
+        geo_bypass_country = self._downloader.params.get('geo_bypass_country', None)
-        countries = [country_code] if country_code else (
+        countries = orderedSet((geo_bypass_country, 'US', 'AU', 'CA', 'AS', 'FM', 'GU', 'MP', 'PR', 'PW', 'MH', 'VI', ''))
-            'US', 'AU', 'CA', 'AS', 'FM', 'GU', 'MP', 'PR', 'PW', 'MH', 'VI')
+        num_countries, num = len(countries) - 1, 0
 
-        last_e = None
+        media = {}
+        for num, country in enumerate(countries):
+            if num == 1:  # start hard-coded list
+                self.report_warning('%s. Trying with a list of known countries' % (
+                    'Unable to obtain video formats from %s API' % geo_bypass_country if geo_bypass_country
+                    else 'No country code was given using --geo-bypass-country'))
+            elif num == num_countries:  # end of list
+                geo_info = self._download_json(
+                    'https://web-api-us.crackle.com/Service.svc/geo/country',
+                    video_id, fatal=False, note='Downloading geo-location information from crackle API',
+                    errnote='Unable to fetch geo-location information from crackle') or {}
+                country = geo_info.get('CountryCode')
+                if country is None:
+                    continue
+                self.to_screen('%s identified country as %s' % (self.IE_NAME, country))
+                if country in countries:
+                    self.to_screen('Downloading from %s API was already attempted. Skipping...' % country)
+                    continue
 
-        for country in countries:
+            if country is None:
+                continue
             try:
-                # Authorization generation algorithm is reverse engineered from:
-                # https://www.sonycrackle.com/static/js/main.ea93451f.chunk.js
-                media_detail_url = 'https://web-api-us.crackle.com/Service.svc/details/media/%s/%s?disableProtocols=true' % (video_id, country)
-                timestamp = time.strftime('%Y%m%d%H%M', time.gmtime())
-                h = hmac.new(b'IGSLUQCBDFHEOIFM', '|'.join([media_detail_url, timestamp]).encode(), hashlib.sha1).hexdigest().upper()
                 media = self._download_json(
-                    media_detail_url, video_id, 'Downloading media JSON as %s' % country,
+                    'https://web-api-us.crackle.com/Service.svc/details/media/%s/%s?disableProtocols=true' % (video_id, country),
-                    'Unable to download media JSON', headers={
+                    video_id, note='Downloading media JSON from %s API' % country,
-                        'Accept': 'application/json',
+                    errnote='Unable to download media JSON')
-                        'Authorization': '|'.join([h, timestamp, '117', '1']),
-                    })
             except ExtractorError as e:
                 # 401 means geo restriction, trying next country
                 if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
-                    last_e = e
                     continue
                 raise
 
-            media_urls = media.get('MediaURLs')
+            status = media.get('status')
-            if not media_urls or not isinstance(media_urls, list):
+            if status.get('messageCode') != '0':
-                continue
+                raise ExtractorError(
+                    '%s said: %s %s - %s' % (
+                        self.IE_NAME, status.get('messageCodeDescription'), status.get('messageCode'), status.get('message')),
+                    expected=True)
 
+            # Found video formats
+            if isinstance(media.get('MediaURLs'), list):
+                break
+
+        ignore_no_formats = self._downloader.params.get('ignore_no_formats_error')
+        allow_unplayable_formats = self._downloader.params.get('allow_unplayable_formats')
+
+        if not media or (not media.get('MediaURLs') and not ignore_no_formats):
+            raise ExtractorError(
+                'Unable to access the crackle API. Try passing your country code '
+                'to --geo-bypass-country. If it still does not work and the '
+                'video is available in your country')
         title = media['Title']
 
-            formats = []
+        formats, subtitles = [], {}
-            for e in media['MediaURLs']:
+        has_drm = False
-                if not self._downloader.params.get('allow_unplayable_formats') and e.get('UseDRM') is True:
+        for e in media.get('MediaURLs') or []:
+            if e.get('UseDRM'):
+                has_drm = True
+                if not allow_unplayable_formats:
                     continue
             format_url = url_or_none(e.get('Path'))
             if not format_url:
                 continue
             ext = determine_ext(format_url)
             if ext == 'm3u8':
-                    formats.extend(self._extract_m3u8_formats(
+                fmts, subs = self._extract_m3u8_formats_and_subtitles(
                     format_url, video_id, 'mp4', entry_protocol='m3u8_native',
-                        m3u8_id='hls', fatal=False))
+                    m3u8_id='hls', fatal=False)
+                formats.extend(fmts)
+                subtitles = self._merge_subtitles(subtitles, subs)
             elif ext == 'mpd':
-                    formats.extend(self._extract_mpd_formats(
+                fmts, subs = self._extract_mpd_formats_and_subtitles(
-                        format_url, video_id, mpd_id='dash', fatal=False))
+                    format_url, video_id, mpd_id='dash', fatal=False)
+                formats.extend(fmts)
+                subtitles = self._merge_subtitles(subtitles, subs)
             elif format_url.endswith('.ism/Manifest'):
-                    formats.extend(self._extract_ism_formats(
+                fmts, subs = self._extract_ism_formats_and_subtitles(
-                        format_url, video_id, ism_id='mss', fatal=False))
+                    format_url, video_id, ism_id='mss', fatal=False)
+                formats.extend(fmts)
+                subtitles = self._merge_subtitles(subtitles, subs)
             else:
                 mfs_path = e.get('Type')
                 mfs_info = self._MEDIA_FILE_SLOTS.get(mfs_path)
@@ -130,6 +176,8 @@ def _real_extract(self, url):
                     'width': mfs_info['width'],
                     'height': mfs_info['height'],
                 })
+        if not formats and has_drm and not ignore_no_formats:
+            raise ExtractorError('The video is DRM protected', expected=True)
         self._sort_formats(formats)
 
         description = media.get('Description')
@@ -151,7 +199,6 @@ def _real_extract(self, url):
         else:
             series = episode = season_number = episode_number = None
 
-            subtitles = {}
         cc_files = media.get('ClosedCaptionFiles')
         if isinstance(cc_files, list):
             for cc_file in cc_files:
@@ -196,5 +243,3 @@ def _real_extract(self, url):
             'subtitles': subtitles,
             'formats': formats,
         }
-
-        raise last_e