1
0
mirror of https://codeberg.org/polarisfm/youtube-dl synced 2024-11-26 02:14:32 +01:00

[tver] Added support for videos distributed by Fuji TV (FOD, Fuji TV On Demand) hosted by its own distribution system

This commit is contained in:
tsukumijima 2020-09-30 02:43:27 +09:00
parent bf6f1048b2
commit bcae9d7a02

View File

@ -12,7 +12,9 @@ from ..utils import (
class TVerIE(InfoExtractor):
_TEST = {
_TESTS = [
{
# Delivery from Brightcove
'url': 'https://tver.jp/feature/f0057485', # In addition to 'feature', there are also categories such as 'corner' and 'episode'.
'md5': '4ae1bc00e6d55af8f7e2b2c17029f1a3', # MD5 hash of a short video downloaded by running youtube-dl with the --test option
'info_dict': {
@ -26,9 +28,28 @@ class TVerIE(InfoExtractor):
'timestamp': 1600308623,
'upload_date': '20200917',
'uploader_id': '4031511847001',
'creator': 'tbs', # Means TBS TV
'uploader': 'TBS FREE', # Content provider
},
'skip': 'Running from test_download.py doesn\'t seem to be able to handle encrypted HLS videos',
}
},
{
# Delivery from FOD (Fuji TV On Demand)
'url': 'https://tver.jp/corner/f0057932', # In addition to 'feature', there are also categories such as 'corner' and 'episode'.
'md5': '6d1970594e532f4b1d6403b5bf9d0d67', # MD5 hash of a short video downloaded by running youtube-dl with the --test option
'info_dict': {
'id': 'f0057932', # TVer ID
'display_id': '5d40810015', # FOD ID
'ext': 'mp4',
'title': 'ちびまる子ちゃん #1258「秋のお楽しみメニューまる子の昔ばなし 『まる子の涼しい大作戦』の巻/『まる子のおむすびころりん』の巻」',
'description': 'md5:328c6ef38bed76588a1f6eb5d69c4a7c',
'thumbnail': r're:https?://.*\.jpg$',
'creator': 'cx', # Means Fuji TV
'uploader': 'FOD見逃し無料', # Content provider
},
'skip': 'Running from test_download.py doesn\'t seem to be able to handle encrypted HLS videos',
},
]
IE_NAME = 'TVer'
IE_DESC = 'TVer'
@ -36,9 +57,6 @@ class TVerIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?tver\.jp/(corner|episode|feature)/(?P<id>f?[0-9]+)'
_GEO_COUNTRIES = ['JP'] # TVer service is limited to Japan only
BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s'
# TODO: FOD対応
def _real_extract(self, url):
video_id = self._match_id(url)
@ -49,17 +67,70 @@ class TVerIE(InfoExtractor):
tver_info_csv = tver_info_csv.replace('\t', '').replace('\n', '').replace('\'', '') # remove \t and \n and '
tver_info = tver_info_csv.split(',')
# extract brightcove information
brightcove_account_id = tver_info[3]
brightcove_video_id = 'ref:' + tver_info[4]
brightcove_url = self.BRIGHTCOVE_URL_TEMPLATE % (brightcove_account_id, brightcove_video_id)
brightcove_info = self._extract_brightcove_info(brightcove_url, 'https://tver.jp/')
# extract tver title
title = tver_info[5] + ' ' + tver_info[6].lstrip() # title + subtitle
# extract tver description
description = \
self._html_search_meta(['og:description', 'twitter:description'], webpage, 'description', default=None) or \
self._html_search_regex(r'<div[^>]+class="description"[^>]*>(?P<description>.*?)</div>', webpage, 'description', default=None, flags=re.DOTALL)
# Note: Of the videos on TVer, only the videos distributed by Fuji TV (FOD, Fuji TV On Demand)
# use our own distribution system instead of Brightcove.
if tver_info[7] == 'cx':
# extract fod information
fod_video_id = tver_info[3]
fod_url = 'https://i.fod.fujitv.co.jp/abr/pc_html5/%s.m3u8' % fod_video_id
fod_thumbnail = 'https://i.fod.fujitv.co.jp/pc/image/wbtn/wbtn_%s.jpg' % fod_video_id
# extract fod formats
fod_formats = self._extract_m3u8_formats(fod_url, fod_video_id, 'mp4', entry_protocol='m3u8_native', m3u8_id='hls')
# Note: All 'RESOLUTION' values in the playlist are 360p,
# but this is a fake value and will be replaced based on what you actually downloaded and measured.
for index, fod_fotmat in enumerate(fod_formats):
# 720p, 2000kbps
if fod_fotmat['format_id'] == 'hls-2000':
fod_formats[index]['width'] = 1280
fod_formats[index]['height'] = 720
# 720p, 1200kbps
elif fod_fotmat['format_id'] == 'hls-1200':
fod_formats[index]['width'] = 1280
fod_formats[index]['height'] = 720
# 360p, 800kbps
elif fod_fotmat['format_id'] == 'hls-800':
fod_formats[index]['width'] = 640
fod_formats[index]['height'] = 360
# 180p, 300kbps
elif fod_fotmat['format_id'] == 'hls-300':
fod_formats[index]['width'] = 320
fod_formats[index]['height'] = 180
# reverse the format order
fod_formats.reverse()
info_dict = {
'id': video_id, # Tver ID
'display_id': fod_video_id, # FOD ID
'formats': fod_formats,
'title': title,
'description': description,
'thumbnail': fod_thumbnail,
'creator': tver_info[7], # Broadcaster name e.g. 'cx'
'uploader': tver_info[8], # Delivery platform name e.g. 'FOD見逃し無料'
'tags': [tver_info[5]],
'is_live': False,
}
else:
# extract brightcove information
brightcove_account_id = tver_info[3]
brightcove_video_id = 'ref:' + tver_info[4]
brightcove_url = 'http://players.brightcove.net/%s/default_default/index.html?videoId=%s' % (brightcove_account_id, brightcove_video_id)
brightcove_info = self._extract_brightcove_info(brightcove_url, 'https://tver.jp/')
# Note: Delegate extraction to BrightcoveNewIE by specifying url_transparent,
# while also making TverIE's own acquired entities such as description available.
info_dict = {
@ -68,7 +139,7 @@ class TVerIE(InfoExtractor):
'ie_key': BrightcoveNewIE.ie_key(),
'id': video_id, # Tver ID
'display_id': brightcove_video_id, # Brightcove ID
'title': brightcove_info.get('name'),
'title': title or brightcove_info.get('name'),
'description': description,
'thumbnail': re.sub(r'/[0-9]+x[0-9]+/', r'/1920x1080/', brightcove_info.get('poster')), # select large thumbnail
'creator': tver_info[7], # Broadcaster name e.g. 'tbs', 'ntv'