[willowtv] Add new extractor

willow.tv is a Cricket TV channel and Internet service. On the Internet,
they offer live streaming, full replays, and highlight videos.

This initial commit handles downloading of replays.

Note that replays are only accessible with an active subscription, and
youtube-dl must be passed the cookies for an active login session. That
means that it's not possible to run a test case in general, and I think
not possible at all, as I don't think there's a way to pass cookies when
running the tests.

youtube_dl/extractor/extractors.py

@@ -1411,6 +1411,7 @@ from .weibo import (
     WeiboMobileIE
 )
 from .weiqitv import WeiqiTVIE
+from .willowtv import WillowTvReplayIE
 from .wimp import WimpIE
 from .wistia import WistiaIE
 from .worldstarhiphop import WorldStarHipHopIE

youtube_dl/extractor/willowtv.py

@@ -0,0 +1,105 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    ExtractorError,
+    determine_ext,
+    strip_jsonp,
+    try_get,
+)
+
+
+class WillowTvReplayIE(InfoExtractor):
+    IE_NAME = 'willowtv:replay'
+    _VALID_URL = r'https?://(?:www\.)?willow\.tv/watch-live/(?P<id>.*)(?:/.*)'
+    _TESTS = [{
+        'url': 'https://www.willow.tv/watch-live/iu-vs-kk-streaming-online-eliminator-1-pakistan-super-league-2019/replay',
+        'info_dict': {
+            'id': 'iu-vs-kk-streaming-online-eliminator-1-pakistan-super-league-2019',
+            'ext': 'mp4',
+        },
+        'skip': 'There are no free examples. An account and cookies are required',
+    }]
+
+    def _real_extract(self, url):
+        meta_format = 'https://willowfeedsv2.willow.tv/willowds/mspecific/%s.json'
+        replay_format = 'https://www.willow.tv/match_replay_data_by_id?matchid=%s'
+
+        slug = self._match_id(url)
+
+        meta = self._download_json(
+            meta_format % slug,
+            slug,
+            transform_source=strip_jsonp)
+
+        results = try_get(meta, lambda x: x['result'], list) or []
+        if not results:
+            raise ExtractorError(
+                'No results present for this match. Is it finished yet?',
+                expected=True)
+
+        entries = []
+        for result in results:
+            matchid = result['Id']
+
+            name = try_get(result, lambda x: x['Name'], compat_str)
+            series = try_get(result, lambda x: x['SeriesName'], compat_str)
+            match_type = try_get(result, lambda x: x['Type'], compat_str)
+
+            data = self._download_json(
+                replay_format % matchid,
+                matchid,
+                transform_source=strip_jsonp)
+
+            status = try_get(data, lambda x: x['status'], compat_str)
+            if status != 'success':
+                raise ExtractorError(
+                    'You must pass the cookies for a logged in willow.tv '
+                    'session with --cookies to download replays.',
+                    expected=True)
+
+            replays = try_get(data, lambda x: x['result']['replay'], list) \
+                or []
+            if not replays:
+                raise ExtractorError(
+                    'No replays present for this match. '
+                    'They may not have been uploaded yet.',
+                    expected=True)
+
+            vid_format = try_get(data, lambda x: x['result']['vidFormat'],
+                                 compat_str)
+            if vid_format != 'HLS':
+                raise ExtractorError(
+                    'Unsupported video format "%s".' % vid_format,
+                    expected=True)
+
+            for outer in replays:
+                for inner in outer:
+                    part = inner['priority']
+                    part_id = compat_str(part)
+                    url = inner['secureurl']
+
+                    formats = self._extract_m3u8_formats(
+                        url, part_id, 'mp4',
+                        entry_protocol='m3u8_native',
+                        m3u8_id='hls', fatal=True)
+                    self._sort_formats(formats)
+
+                    title = try_get(inner, lambda x: x['title'], compat_str) \
+                        or name or slug
+
+                    entries.append({
+                        'id': part_id,
+                        'title': title,
+                        'series': series,
+                        'episode_name': name,
+                        'episode_id': matchid,
+                        'chapter_number': part,
+                        'chapter_id': part_id,
+                        'match_type': match_type,
+                        'formats': formats,
+                    })
+
+        return self.playlist_result(entries, slug)