Mirrors
/
yt-dlp
mirror of https://github.com/yt-dlp/yt-dlp.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

[niconico] Partial fix 

TODO:
Smile formats don't work
Extraction code that no longer works needs to be removed

Adapted from:
b5eff52dd9
by animelover1984

Related:
https://github.com/yt-dlp/yt-dlp/issues/171
https://github.com/animelover1984/youtube-dl/issues/29
https://github.com/ytdl-org/youtube-dl/issues/28470
This commit is contained in:
pukkandan 2021-03-17 15:30:24 +05:30
parent 605d299f83
commit 7978e172f3
No known key found for this signature in database
GPG Key ID: 0F00D95A001F4698
1 changed files with 97 additions and 86 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

183
yt_dlp/extractor/niconico.py
View File

@ -8,6 +8,7 @@ import datetime
from .common import InfoExtractor from .common import InfoExtractor
from ..postprocessor.ffmpeg import FFmpegPostProcessor from ..postprocessor.ffmpeg import FFmpegPostProcessor
from ..compat import ( from ..compat import (
compat_str,
compat_parse_qs, compat_parse_qs,
compat_urllib_parse_urlparse, compat_urllib_parse_urlparse,
) )
@ -20,6 +21,7 @@ from ..utils import (
parse_duration, parse_duration,
parse_iso8601, parse_iso8601,
PostProcessingError, PostProcessingError,
str_or_none,
remove_start, remove_start,
try_get, try_get,
unified_timestamp, unified_timestamp,
@ -34,7 +36,7 @@ class NiconicoIE(InfoExtractor):
_TESTS = [{ _TESTS = [{
'url': 'http://www.nicovideo.jp/watch/sm22312215', 'url': 'http://www.nicovideo.jp/watch/sm22312215',
'md5': 'd1a75c0823e2f629128c43e1212760f9', 'md5': 'a5bad06f1347452102953f323c69da34s',
'info_dict': { 'info_dict': {
'id': 'sm22312215', 'id': 'sm22312215',
'ext': 'mp4', 'ext': 'mp4',
@ -203,7 +205,7 @@ class NiconicoIE(InfoExtractor):
'data-api-data="([^"]+)"', webpage, 'data-api-data="([^"]+)"', webpage,
'API data', default='{}'), video_id) 'API data', default='{}'), video_id)
session_api_data = try_get(api_data, lambda x: x['video']['dmcInfo']['session_api']) session_api_data = try_get(api_data, lambda x: x['media']['delivery']['movie']['session'])
session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0]) session_api_endpoint = try_get(session_api_data, lambda x: x['urls'][0])
# ping # ping
@ -220,7 +222,7 @@ class NiconicoIE(InfoExtractor):
yesno = lambda x: 'yes' if x else 'no' yesno = lambda x: 'yes' if x else 'no'
# m3u8 (encryption) # m3u8 (encryption)
if 'encryption' in try_get(api_data, lambda x: x['video']['dmcInfo']) or {}: if 'encryption' in (try_get(api_data, lambda x: x['media']['delivery']['movie']) or {}):
protocol = 'm3u8' protocol = 'm3u8'
session_api_http_parameters = { session_api_http_parameters = {
'parameters': { 'parameters': {
@ -244,8 +246,8 @@ class NiconicoIE(InfoExtractor):
session_api_http_parameters = { session_api_http_parameters = {
'parameters': { 'parameters': {
'http_output_download_parameters': { 'http_output_download_parameters': {
'use_ssl': yesno(session_api_endpoint['is_ssl']), 'use_ssl': yesno(session_api_endpoint['isSsl']),
'use_well_known_port': yesno(session_api_endpoint['is_well_known_port']), 'use_well_known_port': yesno(session_api_endpoint['isWellKnownPort']),
} }
} }
} }
@ -258,15 +260,15 @@ class NiconicoIE(InfoExtractor):
data=json.dumps({ data=json.dumps({
'session': { 'session': {
'client_info': { 'client_info': {
'player_id': session_api_data.get('player_id'), 'player_id': session_api_data.get('playerId'),
}, },
'content_auth': { 'content_auth': {
'auth_type': try_get(session_api_data, lambda x: x['auth_types'][session_api_data['protocols'][0]]), 'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
'content_key_timeout': session_api_data.get('content_key_timeout'), 'content_key_timeout': session_api_data.get('contentKeyTimeout'),
'service_id': 'nicovideo', 'service_id': 'nicovideo',
'service_user_id': session_api_data.get('service_user_id') 'service_user_id': session_api_data.get('serviceUserId')
}, },
'content_id': session_api_data.get('content_id'), 'content_id': session_api_data.get('contentId'),
'content_src_id_sets': [{ 'content_src_id_sets': [{
'content_src_ids': [{ 'content_src_ids': [{
'src_id_to_mux': { 'src_id_to_mux': {
@ -279,7 +281,7 @@ class NiconicoIE(InfoExtractor):
'content_uri': '', 'content_uri': '',
'keep_method': { 'keep_method': {
'heartbeat': { 'heartbeat': {
'lifetime': session_api_data.get('heartbeat_lifetime') 'lifetime': session_api_data.get('heartbeatLifetime')
} }
}, },
'priority': session_api_data.get('priority'), 'priority': session_api_data.get('priority'),
@ -289,7 +291,7 @@ class NiconicoIE(InfoExtractor):
'http_parameters': session_api_http_parameters 'http_parameters': session_api_http_parameters
} }
}, },
'recipe_id': session_api_data.get('recipe_id'), 'recipe_id': session_api_data.get('recipeId'),
'session_operation_auth': { 'session_operation_auth': {
'session_operation_auth_by_signature': { 'session_operation_auth_by_signature': {
'signature': session_api_data.get('signature'), 'signature': session_api_data.get('signature'),
@ -308,7 +310,7 @@ class NiconicoIE(InfoExtractor):
'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT', 'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
'data': json.dumps(session_response['data']), 'data': json.dumps(session_response['data']),
# interval, convert milliseconds to seconds, then halve to make a buffer. # interval, convert milliseconds to seconds, then halve to make a buffer.
'interval': float_or_none(session_api_data.get('heartbeat_lifetime'), scale=2000), 'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=2000),
} }
return info_dict, heartbeat_info_dict return info_dict, heartbeat_info_dict
@ -327,15 +329,17 @@ class NiconicoIE(InfoExtractor):
format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality])) format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality]))
vdict = parse_format_id(video_quality['id']) vdict = parse_format_id(video_quality['id'])
adict = parse_format_id(audio_quality['id']) adict = parse_format_id(audio_quality['id'])
resolution = video_quality.get('resolution', {'height': vdict.get('res')}) resolution = try_get(video_quality, lambda x: x['metadata']['resolution'], dict) or {'height': vdict.get('res')}
vbr = try_get(video_quality, lambda x: x['metadata']['bitrate'], float)
return { return {
'url': '%s:%s/%s/%s' % (protocol, video_id, video_quality['id'], audio_quality['id']), 'url': '%s:%s/%s/%s' % (protocol, video_id, video_quality['id'], audio_quality['id']),
'format_id': format_id, 'format_id': format_id,
'format_note': 'DMC %s' % try_get(video_quality, lambda x: x['metadata']['label'], compat_str),
'ext': 'mp4', # Session API are used in HTML5, which always serves mp4 'ext': 'mp4', # Session API are used in HTML5, which always serves mp4
'vcodec': vdict.get('codec'), 'vcodec': vdict.get('codec'),
'acodec': adict.get('codec'), 'acodec': adict.get('codec'),
'vbr': float_or_none(video_quality.get('bitrate'), 1000) or float_or_none(vdict.get('br')), 'vbr': float_or_none(vbr, 1000) or float_or_none(vdict.get('br')),
'abr': float_or_none(audio_quality.get('bitrate'), 1000) or float_or_none(adict.get('br')), 'abr': float_or_none(audio_quality.get('bitrate'), 1000) or float_or_none(adict.get('br')),
'height': int_or_none(resolution.get('height', vdict.get('res'))), 'height': int_or_none(resolution.get('height', vdict.get('res'))),
'width': int_or_none(resolution.get('width')), 'width': int_or_none(resolution.get('width')),
@ -394,92 +398,93 @@ class NiconicoIE(InfoExtractor):
formats = [] formats = []
# Get HTML5 videos info # Get HTML5 videos info
try: quality_info = try_get(api_data, lambda x: x['media']['delivery']['movie'])
dmc_info = api_data['video']['dmcInfo'] if not quality_info:
except KeyError: raise ExtractorError('The video can\'t downloaded.', expected=True)
raise ExtractorError('The video can\'t downloaded.',
expected=True)
quality_info = dmc_info.get('quality')
for audio_quality in quality_info.get('audios') or {}: for audio_quality in quality_info.get('audios') or {}:
for video_quality in quality_info.get('videos') or {}: for video_quality in quality_info.get('videos') or {}:
if not audio_quality.get('available') or not video_quality.get('available'): if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
continue continue
formats.append(self._extract_format_for_quality( formats.append(self._extract_format_for_quality(
api_data, video_id, audio_quality, video_quality)) api_data, video_id, audio_quality, video_quality))
# Get flv/swf info # Get flv/swf info
timestamp = None
video_real_url = try_get(api_data, lambda x: x['video']['smileInfo']['url']) video_real_url = try_get(api_data, lambda x: x['video']['smileInfo']['url'])
is_economy = video_real_url.endswith('low') if not video_real_url:
self.report_warning('Unable to obtain smile video information')
else:
is_economy = video_real_url.endswith('low')
if is_economy: if is_economy:
self.report_warning('Site is currently in economy mode! You will only have access to lower quality streams') self.report_warning('Site is currently in economy mode! You will only have access to lower quality streams')
# Invoking ffprobe to determine resolution # Invoking ffprobe to determine resolution
pp = FFmpegPostProcessor(self._downloader) pp = FFmpegPostProcessor(self._downloader)
cookies = self._get_cookies('https://nicovideo.jp').output(header='', sep='; path=/; domain=nicovideo.jp;\n') cookies = self._get_cookies('https://nicovideo.jp').output(header='', sep='; path=/; domain=nicovideo.jp;\n')
self.to_screen('%s: %s' % (video_id, 'Checking smile format with ffprobe')) self.to_screen('%s: %s' % (video_id, 'Checking smile format with ffprobe'))
try: try:
metadata = pp.get_metadata_object(video_real_url, ['-cookies', cookies]) metadata = pp.get_metadata_object(video_real_url, ['-cookies', cookies])
except PostProcessingError as err: except PostProcessingError as err:
raise ExtractorError(err.msg, expected=True) raise ExtractorError(err.msg, expected=True)
v_stream = a_stream = {} v_stream = a_stream = {}
# Some complex swf files doesn't have video stream (e.g. nm4809023) # Some complex swf files doesn't have video stream (e.g. nm4809023)
for stream in metadata['streams']: for stream in metadata['streams']:
if stream['codec_type'] == 'video': if stream['codec_type'] == 'video':
v_stream = stream v_stream = stream
elif stream['codec_type'] == 'audio': elif stream['codec_type'] == 'audio':
a_stream = stream a_stream = stream
# Community restricted videos seem to have issues with the thumb API not returning anything at all # Community restricted videos seem to have issues with the thumb API not returning anything at all
filesize = int( filesize = int(
(get_video_info_xml('size_high') if not is_economy else get_video_info_xml('size_low')) (get_video_info_xml('size_high') if not is_economy else get_video_info_xml('size_low'))
or metadata['format']['size'] or metadata['format']['size']
) )
extension = ( extension = (
get_video_info_xml('movie_type') get_video_info_xml('movie_type')
or 'mp4' if 'mp4' in metadata['format']['format_name'] else metadata['format']['format_name'] or 'mp4' if 'mp4' in metadata['format']['format_name'] else metadata['format']['format_name']
) )
# 'creation_time' tag on video stream of re-encoded SMILEVIDEO mp4 files are '1970-01-01T00:00:00.000000Z'. # 'creation_time' tag on video stream of re-encoded SMILEVIDEO mp4 files are '1970-01-01T00:00:00.000000Z'.
timestamp = ( timestamp = (
parse_iso8601(get_video_info_web('first_retrieve')) parse_iso8601(get_video_info_web('first_retrieve'))
or unified_timestamp(get_video_info_web('postedDateTime')) or unified_timestamp(get_video_info_web('postedDateTime'))
) )
metadata_timestamp = ( metadata_timestamp = (
parse_iso8601(try_get(v_stream, lambda x: x['tags']['creation_time'])) parse_iso8601(try_get(v_stream, lambda x: x['tags']['creation_time']))
or timestamp if extension != 'mp4' else 0 or timestamp if extension != 'mp4' else 0
) )
# According to compconf, smile videos from pre-2017 are always better quality than their DMC counterparts # According to compconf, smile videos from pre-2017 are always better quality than their DMC counterparts
smile_threshold_timestamp = parse_iso8601('2016-12-08T00:00:00+09:00') smile_threshold_timestamp = parse_iso8601('2016-12-08T00:00:00+09:00')
is_source = timestamp < smile_threshold_timestamp or metadata_timestamp > 0 is_source = timestamp < smile_threshold_timestamp or metadata_timestamp > 0
# If movie file size is unstable, old server movie is not source movie. # If movie file size is unstable, old server movie is not source movie.
if filesize > 1: if filesize > 1:
formats.append({ formats.append({
'url': video_real_url, 'url': video_real_url,
'format_id': 'smile' if not is_economy else 'smile_low', 'format_id': 'smile' if not is_economy else 'smile_low',
'format_note': 'SMILEVIDEO source' if not is_economy else 'SMILEVIDEO low quality', 'format_note': 'SMILEVIDEO source' if not is_economy else 'SMILEVIDEO low quality',
'ext': extension, 'ext': extension,
'container': extension, 'container': extension,
'vcodec': v_stream.get('codec_name'), 'vcodec': v_stream.get('codec_name'),
'acodec': a_stream.get('codec_name'), 'acodec': a_stream.get('codec_name'),
# Some complex swf files doesn't have total bit rate metadata (e.g. nm6049209) # Some complex swf files doesn't have total bit rate metadata (e.g. nm6049209)
'tbr': int_or_none(metadata['format'].get('bit_rate'), scale=1000), 'tbr': int_or_none(metadata['format'].get('bit_rate'), scale=1000),
'vbr': int_or_none(v_stream.get('bit_rate'), scale=1000), 'vbr': int_or_none(v_stream.get('bit_rate'), scale=1000),
'abr': int_or_none(a_stream.get('bit_rate'), scale=1000), 'abr': int_or_none(a_stream.get('bit_rate'), scale=1000),
'height': int_or_none(v_stream.get('height')), 'height': int_or_none(v_stream.get('height')),
'width': int_or_none(v_stream.get('width')), 'width': int_or_none(v_stream.get('width')),
'source_preference': 5 if not is_economy else -2, 'source_preference': 5 if not is_economy else -2,
'quality': 5 if is_source and not is_economy else None, 'quality': 5 if is_source and not is_economy else None,
'filesize': filesize 'filesize': filesize
}) })
if len(formats) == 0: if len(formats) == 0:
raise ExtractorError('Unable to find video info.') raise ExtractorError('Unable to find video info.')
@ -517,6 +522,7 @@ class NiconicoIE(InfoExtractor):
timestamp = parse_iso8601( timestamp = parse_iso8601(
video_detail['postedAt'].replace('/', '-'), video_detail['postedAt'].replace('/', '-'),
delimiter=' ', timezone=datetime.timedelta(hours=9)) delimiter=' ', timezone=datetime.timedelta(hours=9))
timestamp = timestamp or try_get(api_data, lambda x: parse_iso8601(x['video']['registeredAt']))
view_count = int_or_none(get_video_info_web(['view_counter', 'viewCount'])) view_count = int_or_none(get_video_info_web(['view_counter', 'viewCount']))
if not view_count: if not view_count:
@ -525,11 +531,16 @@ class NiconicoIE(InfoExtractor):
webpage, 'view count', default=None) webpage, 'view count', default=None)
if match: if match:
view_count = int_or_none(match.replace(',', '')) view_count = int_or_none(match.replace(',', ''))
view_count = view_count or video_detail.get('viewCount') view_count = (
view_count
or video_detail.get('viewCount')
or try_get(api_data, lambda x: x['video']['count']['view']))
comment_count = (
int_or_none(get_video_info_web('comment_num'))
or video_detail.get('commentCount')
or try_get(api_data, lambda x: x['video']['count']['comment']))
comment_count = (int_or_none(get_video_info_web('comment_num'))
or video_detail.get('commentCount')
or try_get(api_data, lambda x: x['thread']['commentCount']))
if not comment_count: if not comment_count:
match = self._html_search_regex( match = self._html_search_regex(
r'>Comments: <strong[^>]*>([^<]+)</strong>', r'>Comments: <strong[^>]*>([^<]+)</strong>',
@ -559,7 +570,7 @@ class NiconicoIE(InfoExtractor):
# Note: cannot use api_data.get('owner', {}) because owner may be set to "null" # Note: cannot use api_data.get('owner', {}) because owner may be set to "null"
# in the JSON, which will cause None to be returned instead of {}. # in the JSON, which will cause None to be returned instead of {}.
owner = try_get(api_data, lambda x: x.get('owner'), dict) or {} owner = try_get(api_data, lambda x: x.get('owner'), dict) or {}
uploader_id = ( uploader_id = str_or_none(
get_video_info_web(['ch_id', 'user_id']) get_video_info_web(['ch_id', 'user_id'])
or owner.get('id') or owner.get('id')
or channel_id or channel_id