Plugin cleanup and tweaks

This commit is contained in:
2023-02-20 19:18:45 -06:00
parent 372e4ff3dc
commit 3ad9e1c7bb
1138 changed files with 48878 additions and 40445 deletions

View File

@@ -1,6 +1,3 @@
# coding: utf-8
from __future__ import unicode_literals
import collections
import itertools
import json
@@ -15,11 +12,14 @@ from ..compat import (
compat_urllib_parse_urlparse,
)
from ..utils import (
ExtractorError,
UserNotLive,
base_url,
clean_html,
dict_get,
ExtractorError,
float_or_none,
int_or_none,
make_archive_id,
parse_duration,
parse_iso8601,
parse_qs,
@@ -55,16 +55,10 @@ class TwitchBaseIE(InfoExtractor):
'VideoPreviewOverlay': '3006e77e51b128d838fa4e835723ca4dc9a05c5efd4466c1085215c6e437e65c',
'VideoMetadata': '226edb3e692509f727fd56821f5653c05740242c82b0388883e0c0e75dcbf687',
'VideoPlayer_ChapterSelectButtonVideo': '8d2793384aac3773beab5e59bd5d6f585aedb923d292800119e03d40cd0f9b41',
'VideoPlayer_VODSeekbarPreviewVideo': '07e99e4d56c5a7c67117a154777b0baf85a5ffefa393b213f4bc712ccaf85dd6',
}
def _real_initialize(self):
self._login()
def _login(self):
username, password = self._get_login_info()
if username is None:
return
def _perform_login(self, username, password):
def fail(message):
raise ExtractorError(
'Unable to login. Twitch said: %s' % message, expected=True)
@@ -139,7 +133,6 @@ class TwitchBaseIE(InfoExtractor):
'quality': 10,
'format_note': 'Source',
})
self._sort_formats(formats)
def _download_base_gql(self, video_id, ops, note, fatal=True):
headers = {
@@ -212,6 +205,14 @@ class TwitchVodIE(TwitchBaseIE):
'uploader_id': 'riotgames',
'view_count': int,
'start_time': 310,
'chapters': [
{
'start_time': 0,
'end_time': 17208,
'title': 'League of Legends'
}
],
'live_status': 'was_live',
},
'params': {
# m3u8 download
@@ -280,10 +281,80 @@ class TwitchVodIE(TwitchBaseIE):
'title': 'Art'
}
],
'live_status': 'was_live',
'thumbnail': r're:^https?://.*\.jpg$',
'view_count': int,
},
'params': {
'skip_download': True
},
}, {
'note': 'Storyboards',
'url': 'https://www.twitch.tv/videos/635475444',
'info_dict': {
'id': 'v635475444',
'format_id': 'sb0',
'ext': 'mhtml',
'title': 'Riot Games',
'duration': 11643,
'uploader': 'Riot Games',
'uploader_id': 'riotgames',
'timestamp': 1590770569,
'upload_date': '20200529',
'chapters': [
{
'start_time': 0,
'end_time': 573,
'title': 'League of Legends'
},
{
'start_time': 573,
'end_time': 3922,
'title': 'Legends of Runeterra'
},
{
'start_time': 3922,
'end_time': 11643,
'title': 'Art'
}
],
'live_status': 'was_live',
'thumbnail': r're:^https?://.*\.jpg$',
'view_count': int,
'columns': int,
'rows': int,
},
'params': {
'format': 'mhtml',
'skip_download': True
}
}, {
'note': 'VOD with single chapter',
'url': 'https://www.twitch.tv/videos/1536751224',
'info_dict': {
'id': 'v1536751224',
'ext': 'mp4',
'title': 'Porter Robinson Star Guardian Stream Tour with LilyPichu',
'duration': 8353,
'uploader': 'Riot Games',
'uploader_id': 'riotgames',
'timestamp': 1658267731,
'upload_date': '20220719',
'chapters': [
{
'start_time': 0,
'end_time': 8353,
'title': 'League of Legends'
}
],
'live_status': 'was_live',
'thumbnail': r're:^https?://.*\.jpg$',
'view_count': int,
},
'params': {
'skip_download': True
},
'expected_warnings': ['Unable to download JSON metadata: HTTP Error 403: Forbidden']
}]
def _download_info(self, item_id):
@@ -300,16 +371,23 @@ class TwitchVodIE(TwitchBaseIE):
'includePrivate': False,
'videoID': item_id,
},
}, {
'operationName': 'VideoPlayer_VODSeekbarPreviewVideo',
'variables': {
'includePrivate': False,
'videoID': item_id,
},
}],
'Downloading stream metadata GraphQL')
video = traverse_obj(data, (0, 'data', 'video'))
video['moments'] = traverse_obj(data, (1, 'data', 'video', 'moments', 'edges', ..., 'node'))
video['storyboard'] = traverse_obj(data, (2, 'data', 'video', 'seekPreviewsURL'), expected_type=url_or_none)
if video is None:
raise ExtractorError(
'Video %s does not exist' % item_id, expected=True)
return self._extract_info_gql(video, item_id)
return video
def _extract_info(self, info):
status = info.get('status')
@@ -348,8 +426,14 @@ class TwitchVodIE(TwitchBaseIE):
'was_live': True,
}
def _extract_moments(self, info, item_id):
for moment in info.get('moments') or []:
def _extract_chapters(self, info, item_id):
if not info.get('moments'):
game = traverse_obj(info, ('game', 'displayName'))
if game:
yield {'title': game}
return
for moment in info['moments']:
start_time = int_or_none(moment.get('positionMilliseconds'), 1000)
duration = int_or_none(moment.get('durationMilliseconds'), 1000)
name = str_or_none(moment.get('description'))
@@ -388,15 +472,49 @@ class TwitchVodIE(TwitchBaseIE):
'uploader_id': try_get(info, lambda x: x['owner']['login'], compat_str),
'timestamp': unified_timestamp(info.get('publishedAt')),
'view_count': int_or_none(info.get('viewCount')),
'chapters': list(self._extract_moments(info, item_id)),
'chapters': list(self._extract_chapters(info, item_id)),
'is_live': is_live,
'was_live': True,
}
def _extract_storyboard(self, item_id, storyboard_json_url, duration):
if not duration or not storyboard_json_url:
return
spec = self._download_json(storyboard_json_url, item_id, 'Downloading storyboard metadata JSON', fatal=False) or []
# sort from highest quality to lowest
# This makes sb0 the highest-quality format, sb1 - lower, etc which is consistent with youtube sb ordering
spec.sort(key=lambda x: int_or_none(x.get('width')) or 0, reverse=True)
base = base_url(storyboard_json_url)
for i, s in enumerate(spec):
count = int_or_none(s.get('count'))
images = s.get('images')
if not (images and count):
continue
fragment_duration = duration / len(images)
yield {
'format_id': f'sb{i}',
'format_note': 'storyboard',
'ext': 'mhtml',
'protocol': 'mhtml',
'acodec': 'none',
'vcodec': 'none',
'url': urljoin(base, images[0]),
'width': int_or_none(s.get('width')),
'height': int_or_none(s.get('height')),
'fps': count / duration,
'rows': int_or_none(s.get('rows')),
'columns': int_or_none(s.get('cols')),
'fragments': [{
'url': urljoin(base, path),
'duration': fragment_duration,
} for path in images],
}
def _real_extract(self, url):
vod_id = self._match_id(url)
info = self._download_info(vod_id)
video = self._download_info(vod_id)
info = self._extract_info_gql(video, vod_id)
access_token = self._download_access_token(vod_id, 'video', 'id')
formats = self._extract_m3u8_formats(
@@ -413,6 +531,8 @@ class TwitchVodIE(TwitchBaseIE):
})),
vod_id, 'mp4', entry_protocol='m3u8_native')
formats.extend(self._extract_storyboard(vod_id, video.get('storyboard'), info.get('duration')))
self._prefer_source(formats)
info['formats'] = formats
@@ -860,7 +980,7 @@ class TwitchStreamIE(TwitchBaseIE):
stream = user['stream']
if not stream:
raise ExtractorError('%s is offline' % channel_name, expected=True)
raise UserNotLive(video_id=channel_name)
access_token = self._download_access_token(
channel_name, 'stream', 'channelName')
@@ -1023,7 +1143,6 @@ class TwitchClipsIE(TwitchBaseIE):
'height': int_or_none(option.get('quality')),
'fps': int_or_none(option.get('frameRate')),
})
self._sort_formats(formats)
thumbnails = []
for thumbnail_id in ('tiny', 'small', 'medium'):
@@ -1042,13 +1161,16 @@ class TwitchClipsIE(TwitchBaseIE):
})
thumbnails.append(thumb)
old_id = self._search_regex(r'%7C(\d+)(?:-\d+)?.mp4', formats[-1]['url'], 'old id', default=None)
return {
'id': clip.get('id') or video_id,
'_old_archive_ids': [make_archive_id(self, old_id)] if old_id else None,
'display_id': video_id,
'title': clip.get('title') or video_id,
'title': clip.get('title'),
'formats': formats,
'duration': int_or_none(clip.get('durationSeconds')),
'views': int_or_none(clip.get('viewCount')),
'view_count': int_or_none(clip.get('viewCount')),
'timestamp': unified_timestamp(clip.get('createdAt')),
'thumbnails': thumbnails,
'creator': try_get(clip, lambda x: x['broadcaster']['displayName'], compat_str),