[niconico] Back-port extractor from yt-dlp

Add Nico search extractors, fix extraction
This commit is contained in:
dirkf 2021-12-06 20:52:21 +00:00
parent 73e1ab6125
commit 91278f4b6b
2 changed files with 477 additions and 178 deletions

View File

@ -789,7 +789,14 @@ from .nick import (
NickNightIE,
NickRuIE,
)
from .niconico import NiconicoIE, NiconicoPlaylistIE
from .niconico import (
NiconicoIE,
NiconicoPlaylistIE,
NiconicoUserIE,
NicovideoSearchIE,
NicovideoSearchDateIE,
NicovideoSearchURLIE,
)
from .ninecninemedia import NineCNineMediaIE
from .ninegag import NineGagIE
from .ninenow import NineNowIE

View File

@ -2,25 +2,28 @@
from __future__ import unicode_literals
import datetime
import functools
import itertools
import json
import math
import re
from .common import InfoExtractor
from .common import InfoExtractor, SearchInfoExtractor
from ..postprocessor.ffmpeg import FFmpegPostProcessor
from ..compat import (
compat_parse_qs,
compat_str,
compat_urllib_parse_urlparse,
)
from ..utils import (
determine_ext,
dict_get,
ExtractorError,
dict_get,
float_or_none,
InAdvancePagedList,
int_or_none,
OnDemandPagedList,
parse_duration,
parse_iso8601,
PostProcessingError,
remove_start,
str_or_none,
try_get,
unified_timestamp,
urlencode_postdata,
@ -34,7 +37,7 @@ class NiconicoIE(InfoExtractor):
_TESTS = [{
'url': 'http://www.nicovideo.jp/watch/sm22312215',
'md5': 'd1a75c0823e2f629128c43e1212760f9',
'md5': 'a5bad06f1347452102953f323c69da34s',
'info_dict': {
'id': 'sm22312215',
'ext': 'mp4',
@ -162,6 +165,11 @@ class NiconicoIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'
_NETRC_MACHINE = 'niconico'
_API_HEADERS = {
'X-Frontend-ID': '6',
'X-Frontend-Version': '0'
}
def _real_initialize(self):
self._login()
@ -191,37 +199,89 @@ class NiconicoIE(InfoExtractor):
self._downloader.report_warning('unable to log in: bad username or password')
return login_ok
def _extract_format_for_quality(self, api_data, video_id, audio_quality, video_quality):
def yesno(boolean):
return 'yes' if boolean else 'no'
def _get_heartbeat_info(self, info_dict):
session_api_data = api_data['video']['dmcInfo']['session_api']
session_api_endpoint = session_api_data['urls'][0]
video_id, video_src_id, audio_src_id = info_dict['url'].split(':')[1].split('/')
format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality]))
api_data = (
info_dict.get('_api_data')
or self._parse_json(
self._html_search_regex(
'data-api-data="([^"]+)"',
self._download_webpage('http://www.nicovideo.jp/watch/' + video_id, video_id),
'API data', default='{}'),
video_id))
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])
def ping():
status = try_get(
self._download_json(
'https://nvapi.nicovideo.jp/v1/2ab0cbaa/watch', video_id,
query={'t': try_get(api_data, lambda x: x['media']['delivery']['trackingId'])},
note='Acquiring permission for downloading video',
headers=self._API_HEADERS),
lambda x: x['meta']['status'])
if status != 200:
self.report_warning('Failed to acquire permission for playing video. The video may not download.')
yesno = lambda x: 'yes' if x else 'no'
# m3u8 (encryption)
if try_get(api_data, lambda x: x['media']['delivery']['encryption']) is not None:
protocol = 'm3u8'
encryption = self._parse_json(session_api_data['token'], video_id)['hls_encryption']
session_api_http_parameters = {
'parameters': {
'hls_parameters': {
'encryption': {
encryption: {
'encrypted_key': try_get(api_data, lambda x: x['media']['delivery']['encryption']['encryptedKey']),
'key_uri': try_get(api_data, lambda x: x['media']['delivery']['encryption']['keyUri'])
}
},
'transfer_preset': '',
'use_ssl': yesno(session_api_endpoint['isSsl']),
'use_well_known_port': yesno(session_api_endpoint['isWellKnownPort']),
'segment_duration': 6000,
}
}
}
# http
else:
protocol = 'http'
session_api_http_parameters = {
'parameters': {
'http_output_download_parameters': {
'use_ssl': yesno(session_api_endpoint['isSsl']),
'use_well_known_port': yesno(session_api_endpoint['isWellKnownPort']),
}
}
}
session_response = self._download_json(
session_api_endpoint['url'], video_id,
query={'_format': 'json'},
headers={'Content-Type': 'application/json'},
note='Downloading JSON metadata for %s' % format_id,
note='Downloading JSON metadata for %s' % info_dict['format_id'],
data=json.dumps({
'session': {
'client_info': {
'player_id': session_api_data['player_id'],
'player_id': session_api_data.get('playerId'),
},
'content_auth': {
'auth_type': session_api_data['auth_types'][session_api_data['protocols'][0]],
'content_key_timeout': session_api_data['content_key_timeout'],
'auth_type': try_get(session_api_data, lambda x: x['authTypes'][session_api_data['protocols'][0]]),
'content_key_timeout': session_api_data.get('contentKeyTimeout'),
'service_id': 'nicovideo',
'service_user_id': session_api_data['service_user_id']
'service_user_id': session_api_data.get('serviceUserId')
},
'content_id': session_api_data['content_id'],
'content_id': session_api_data.get('contentId'),
'content_src_id_sets': [{
'content_src_ids': [{
'src_id_to_mux': {
'audio_src_ids': [audio_quality['id']],
'video_src_ids': [video_quality['id']],
'audio_src_ids': [audio_src_id],
'video_src_ids': [video_src_id],
}
}]
}],
@ -229,52 +289,81 @@ class NiconicoIE(InfoExtractor):
'content_uri': '',
'keep_method': {
'heartbeat': {
'lifetime': session_api_data['heartbeat_lifetime']
'lifetime': session_api_data.get('heartbeatLifetime')
}
},
'priority': session_api_data['priority'],
'priority': session_api_data.get('priority'),
'protocol': {
'name': 'http',
'parameters': {
'http_parameters': {
'parameters': {
'http_output_download_parameters': {
'use_ssl': yesno(session_api_endpoint['is_ssl']),
'use_well_known_port': yesno(session_api_endpoint['is_well_known_port']),
}
}
}
'http_parameters': session_api_http_parameters
}
},
'recipe_id': session_api_data['recipe_id'],
'recipe_id': session_api_data.get('recipeId'),
'session_operation_auth': {
'session_operation_auth_by_signature': {
'signature': session_api_data['signature'],
'token': session_api_data['token'],
'signature': session_api_data.get('signature'),
'token': session_api_data.get('token'),
}
},
'timing_constraint': 'unlimited'
}
}).encode())
resolution = video_quality.get('resolution', {})
info_dict['url'] = session_response['data']['session']['content_uri']
info_dict['protocol'] = protocol
# get heartbeat info
heartbeat_info_dict = {
'url': session_api_endpoint['url'] + '/' + session_response['data']['session']['id'] + '?_format=json&_method=PUT',
'data': json.dumps(session_response['data']),
# interval, convert milliseconds to seconds, then halve to make a buffer.
'interval': float_or_none(session_api_data.get('heartbeatLifetime'), scale=3000),
'ping': ping
}
return info_dict, heartbeat_info_dict
def _extract_format_for_quality(self, api_data, video_id, audio_quality, video_quality):
def parse_format_id(id_code):
mobj = re.match(r'''(?x)
(?:archive_)?
(?:(?P<codec>[^_]+)_)?
(?:(?P<br>[\d]+)kbps_)?
(?:(?P<res>[\d+]+)p_)?
''', '%s_' % id_code)
return mobj.groupdict() if mobj else {}
protocol = 'niconico_dmc'
format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality]))
vdict = parse_format_id(video_quality['id'])
adict = parse_format_id(audio_quality['id'])
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 {
'url': session_response['data']['session']['content_uri'],
'url': '%s:%s/%s/%s' % (protocol, video_id, video_quality['id'], audio_quality['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
'abr': float_or_none(audio_quality.get('bitrate'), 1000),
'vbr': float_or_none(video_quality.get('bitrate'), 1000),
'height': resolution.get('height'),
'width': resolution.get('width'),
'vcodec': vdict.get('codec'),
'acodec': adict.get('codec'),
'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')),
'height': int_or_none(resolution.get('height', vdict.get('res'))),
'width': int_or_none(resolution.get('width')),
'quality': -2 if 'low' in format_id else -1, # Default quality value is -1
'protocol': protocol,
'http_headers': {
'Origin': 'https://www.nicovideo.jp',
'Referer': 'https://www.nicovideo.jp/watch/' + video_id,
}
}
def _real_extract(self, url):
video_id = self._match_id(url)
# Get video webpage. We are not actually interested in it for normal
# cases, but need the cookies in order to be able to download the
# info webpage
# Get video webpage for API data.
webpage, handle = self._download_webpage_handle(
'http://www.nicovideo.jp/watch/' + video_id, video_id)
if video_id.startswith('so'):
@ -284,86 +373,136 @@ class NiconicoIE(InfoExtractor):
'data-api-data="([^"]+)"', webpage,
'API data', default='{}'), video_id)
def _format_id_from_url(video_url):
return 'economy' if video_real_url.endswith('low') else 'normal'
def get_video_info_web(items):
return dict_get(api_data['video'], items)
try:
video_real_url = api_data['video']['smileInfo']['url']
except KeyError: # Flash videos
# Get flv info
flv_info_webpage = self._download_webpage(
'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
video_id, 'Downloading flv info')
# Get video info
video_info_xml = self._download_xml(
'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id,
video_id, note='Downloading video info page')
flv_info = compat_parse_qs(flv_info_webpage)
if 'url' not in flv_info:
if 'deleted' in flv_info:
raise ExtractorError('The video has been deleted.',
expected=True)
elif 'closed' in flv_info:
raise ExtractorError('Niconico videos now require logging in',
expected=True)
elif 'error' in flv_info:
raise ExtractorError('%s reports error: %s' % (
self.IE_NAME, flv_info['error'][0]), expected=True)
else:
raise ExtractorError('Unable to find video URL')
def get_video_info_xml(items):
if not isinstance(items, list):
items = [items]
for item in items:
ret = xpath_text(video_info_xml, './/' + item)
if ret:
return ret
video_info_xml = self._download_xml(
'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id,
video_id, note='Downloading video info page')
if get_video_info_xml('error'):
error_code = get_video_info_xml('code')
def get_video_info(items):
if not isinstance(items, list):
items = [items]
for item in items:
ret = xpath_text(video_info_xml, './/' + item)
if ret:
return ret
if error_code == 'DELETED':
raise ExtractorError('The video has been deleted.',
expected=True)
elif error_code == 'NOT_FOUND':
raise ExtractorError('The video is not found.',
expected=True)
elif error_code == 'COMMUNITY':
self.to_screen('%s: The video is community members only.' % video_id)
else:
raise ExtractorError('%s reports error: %s' % (self.IE_NAME, error_code))
video_real_url = flv_info['url'][0]
# Start extracting video formats
formats = []
extension = get_video_info('movie_type')
if not extension:
extension = determine_ext(video_real_url)
# Get HTML5 videos info
quality_info = try_get(api_data, lambda x: x['media']['delivery']['movie'])
if not quality_info:
raise ExtractorError('The video can\'t be downloaded', expected=True)
formats = [{
'url': video_real_url,
'ext': extension,
'format_id': _format_id_from_url(video_real_url),
}]
else:
formats = []
for audio_quality in quality_info.get('audios') or {}:
for video_quality in quality_info.get('videos') or {}:
if not audio_quality.get('isAvailable') or not video_quality.get('isAvailable'):
continue
formats.append(self._extract_format_for_quality(
api_data, video_id, audio_quality, video_quality))
dmc_info = api_data['video'].get('dmcInfo')
if dmc_info: # "New" HTML5 videos
quality_info = dmc_info['quality']
for audio_quality in quality_info['audios']:
for video_quality in quality_info['videos']:
if not audio_quality['available'] or not video_quality['available']:
continue
formats.append(self._extract_format_for_quality(
api_data, video_id, audio_quality, video_quality))
# Get flv/swf info
timestamp = None
video_real_url = try_get(api_data, lambda x: x['video']['smileInfo']['url'])
if video_real_url:
is_economy = video_real_url.endswith('low')
self._sort_formats(formats)
else: # "Old" HTML5 videos
formats = [{
if is_economy:
self.report_warning('Site is currently in economy mode! You will only have access to lower quality streams')
# Invoking ffprobe to determine resolution
pp = FFmpegPostProcessor(self._downloader)
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'))
try:
metadata = pp.get_metadata_object(video_real_url, ['-cookies', cookies])
except PostProcessingError as err:
raise ExtractorError(err.msg, expected=True)
v_stream = a_stream = {}
# Some complex swf files doesn't have video stream (e.g. nm4809023)
for stream in metadata['streams']:
if stream['codec_type'] == 'video':
v_stream = stream
elif stream['codec_type'] == 'audio':
a_stream = stream
# Community restricted videos seem to have issues with the thumb API not returning anything at all
filesize = int(
(get_video_info_xml('size_high') if not is_economy else get_video_info_xml('size_low'))
or metadata['format']['size']
)
extension = (
get_video_info_xml('movie_type')
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'.
timestamp = (
parse_iso8601(get_video_info_web('first_retrieve'))
or unified_timestamp(get_video_info_web('postedDateTime'))
)
metadata_timestamp = (
parse_iso8601(try_get(v_stream, lambda x: x['tags']['creation_time']))
or timestamp if extension != 'mp4' else 0
)
# 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')
is_source = timestamp < smile_threshold_timestamp or metadata_timestamp > 0
# If movie file size is unstable, old server movie is not source movie.
if filesize > 1:
formats.append({
'url': video_real_url,
'ext': 'mp4',
'format_id': _format_id_from_url(video_real_url),
}]
'format_id': 'smile' if not is_economy else 'smile_low',
'format_note': 'SMILEVIDEO source' if not is_economy else 'SMILEVIDEO low quality',
'ext': extension,
'container': extension,
'vcodec': v_stream.get('codec_name'),
'acodec': a_stream.get('codec_name'),
# 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),
'vbr': int_or_none(v_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')),
'width': int_or_none(v_stream.get('width')),
'source_preference': 5 if not is_economy else -2,
'quality': 5 if is_source and not is_economy else None,
'filesize': filesize
})
def get_video_info(items):
return dict_get(api_data['video'], items)
self._sort_formats(formats)
# Start extracting information
title = get_video_info('title')
if not title:
title = self._og_search_title(webpage, default=None)
if not title:
title = self._html_search_regex(
title = (
get_video_info_xml('title') # prefer to get the untranslated original title
or get_video_info_web(['originalTitle', 'title'])
or self._og_search_title(webpage, default=None)
or self._html_search_regex(
r'<span[^>]+class="videoHeaderTitle"[^>]*>([^<]+)</span>',
webpage, 'video title')
webpage, 'video title'))
watch_api_data_string = self._html_search_regex(
r'<div[^>]+id="watchAPIDataContainer"[^>]+>([^<]+)</div>',
@ -372,14 +511,15 @@ class NiconicoIE(InfoExtractor):
video_detail = watch_api_data.get('videoDetail', {})
thumbnail = (
get_video_info(['thumbnail_url', 'thumbnailURL'])
self._html_search_regex(r'<meta property="og:image" content="([^"]+)">', webpage, 'thumbnail data', default=None)
or dict_get( # choose highest from 720p to 240p
get_video_info_web('thumbnail'),
['ogp', 'player', 'largeUrl', 'middleUrl', 'url'])
or self._html_search_meta('image', webpage, 'thumbnail', default=None)
or video_detail.get('thumbnail'))
description = get_video_info('description')
description = get_video_info_web('description')
timestamp = (parse_iso8601(get_video_info('first_retrieve'))
or unified_timestamp(get_video_info('postedDateTime')))
if not timestamp:
match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
if match:
@ -388,19 +528,25 @@ class NiconicoIE(InfoExtractor):
timestamp = parse_iso8601(
video_detail['postedAt'].replace('/', '-'),
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(['view_counter', 'viewCount']))
view_count = int_or_none(get_video_info_web(['view_counter', 'viewCount']))
if not view_count:
match = self._html_search_regex(
r'>Views: <strong[^>]*>([^<]+)</strong>',
webpage, 'view count', default=None)
if match:
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('comment_num'))
or video_detail.get('commentCount')
or try_get(api_data, lambda x: x['thread']['commentCount']))
if not comment_count:
match = self._html_search_regex(
r'>Comments: <strong[^>]*>([^<]+)</strong>',
@ -409,22 +555,41 @@ class NiconicoIE(InfoExtractor):
comment_count = int_or_none(match.replace(',', ''))
duration = (parse_duration(
get_video_info('length')
get_video_info_web('length')
or self._html_search_meta(
'video:duration', webpage, 'video duration', default=None))
or video_detail.get('length')
or get_video_info('duration'))
or get_video_info_web('duration'))
webpage_url = get_video_info('watch_url') or url
webpage_url = get_video_info_web('watch_url') or url
# for channel movie and community movie
channel_id = try_get(
api_data,
(lambda x: x['channel']['globalId'],
lambda x: x['community']['globalId']))
channel = try_get(
api_data,
(lambda x: x['channel']['name'],
lambda x: x['community']['name']))
# 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 {}.
owner = try_get(api_data, lambda x: x.get('owner'), dict) or {}
uploader_id = get_video_info(['ch_id', 'user_id']) or owner.get('id')
uploader = get_video_info(['ch_name', 'user_nickname']) or owner.get('nickname')
uploader_id = str_or_none(
get_video_info_web(['ch_id', 'user_id'])
or owner.get('id')
or channel_id
)
uploader = (
get_video_info_web(['ch_name', 'user_nickname'])
or owner.get('nickname')
or channel
)
return {
'id': video_id,
'_api_data': api_data,
'title': title,
'formats': formats,
'thumbnail': thumbnail,
@ -432,6 +597,8 @@ class NiconicoIE(InfoExtractor):
'uploader': uploader,
'timestamp': timestamp,
'uploader_id': uploader_id,
'channel': channel,
'channel_id': channel_id,
'view_count': view_count,
'comment_count': comment_count,
'duration': duration,
@ -440,7 +607,7 @@ class NiconicoIE(InfoExtractor):
class NiconicoPlaylistIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/(?:user/\d+/)?mylist/(?P<id>\d+)'
_VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/(?:user/\d+/|my/)?mylist/(?P<id>\d+)'
_TESTS = [{
'url': 'http://www.nicovideo.jp/mylist/27411728',
@ -456,60 +623,185 @@ class NiconicoPlaylistIE(InfoExtractor):
'url': 'https://www.nicovideo.jp/user/805442/mylist/27411728',
'only_matching': True,
}]
_PAGE_SIZE = 100
def _call_api(self, list_id, resource, query):
return self._download_json(
'https://nvapi.nicovideo.jp/v2/mylists/' + list_id, list_id,
'Downloading %s JSON metatdata' % resource, query=query,
headers={'X-Frontend-Id': 6})['data']['mylist']
def _parse_owner(self, item):
owner = item.get('owner') or {}
if owner:
return {
'uploader': owner.get('name'),
'uploader_id': owner.get('id'),
}
return {}
def _fetch_page(self, list_id, page):
page += 1
items = self._call_api(list_id, 'page %d' % page, {
'page': page,
'pageSize': self._PAGE_SIZE,
})['items']
for item in items:
video = item.get('video') or {}
video_id = video.get('id')
if not video_id:
continue
count = video.get('count') or {}
get_count = lambda x: int_or_none(count.get(x))
info = {
'_type': 'url',
'id': video_id,
'title': video.get('title'),
'url': 'https://www.nicovideo.jp/watch/' + video_id,
'description': video.get('shortDescription'),
'duration': int_or_none(video.get('duration')),
'view_count': get_count('view'),
'comment_count': get_count('comment'),
'ie_key': NiconicoIE.ie_key(),
}
info.update(self._parse_owner(video))
yield info
_API_HEADERS = {
'X-Frontend-ID': '6',
'X-Frontend-Version': '0'
}
def _real_extract(self, url):
list_id = self._match_id(url)
mylist = self._call_api(list_id, 'list', {
'pageSize': 1,
})
entries = InAdvancePagedList(
functools.partial(self._fetch_page, list_id),
math.ceil(mylist['totalItemCount'] / self._PAGE_SIZE),
self._PAGE_SIZE)
result = self.playlist_result(
entries, list_id, mylist.get('name'), mylist.get('description'))
result.update(self._parse_owner(mylist))
return result
def get_page_data(pagenum, pagesize):
return self._download_json(
'http://nvapi.nicovideo.jp/v2/mylists/' + list_id, list_id,
query={'page': 1 + pagenum, 'pageSize': pagesize},
headers=self._API_HEADERS).get('data').get('mylist')
data = get_page_data(0, 1)
title = data.get('name')
description = data.get('description')
uploader = data.get('owner').get('name')
uploader_id = data.get('owner').get('id')
def pagefunc(pagenum):
data = get_page_data(pagenum, 25)
return ({
'_type': 'url',
'url': 'http://www.nicovideo.jp/watch/' + item.get('watchId'),
} for item in data.get('items'))
return {
'_type': 'playlist',
'id': list_id,
'title': title,
'description': description,
'uploader': uploader,
'uploader_id': uploader_id,
'entries': OnDemandPagedList(pagefunc, 25),
}
class NicovideoSearchBaseIE(InfoExtractor):
_MAX_RESULTS = float('inf')
def _entries(self, url, item_id, query=None, note='Downloading page %(page)s'):
query = query or {}
pages = [query['page']] if 'page' in query else itertools.count(1)
for page_num in pages:
query['page'] = str(page_num)
webpage = self._download_webpage(url, item_id, query=query, note=note % {'page': page_num})
results = re.findall(r'(?<=data-video-id=)["\']?(?P<videoid>.+?)(?=["\'])', webpage)
for item in results:
yield self.url_result('http://www.nicovideo.jp/watch/%s' % item, 'Niconico', item)
if not results:
break
def _get_n_results(self, query, n):
entries = self._entries(self._proto_relative_url('//www.nicovideo.jp/search/%s' % query), query)
if n < self._MAX_RESULTS:
entries = itertools.islice(entries, 0, n)
return self.playlist_result(entries, query, query)
class NicovideoSearchIE(NicovideoSearchBaseIE, SearchInfoExtractor):
IE_DESC = 'Nico video search'
IE_NAME = 'nicovideo:search'
_SEARCH_KEY = 'nicosearch'
def _search_results(self, query):
return self._entries(
self._proto_relative_url('//www.nicovideo.jp/search/%s' % query), query)
class NicovideoSearchURLIE(NicovideoSearchBaseIE):
IE_NAME = '%s_url' % NicovideoSearchIE.IE_NAME
IE_DESC = 'Nico video search URLs'
_VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/search/(?P<id>[^?#&]+)?'
_TESTS = [{
'url': 'http://www.nicovideo.jp/search/sm9',
'info_dict': {
'id': 'sm9',
'title': 'sm9'
},
'playlist_mincount': 40,
}, {
'url': 'https://www.nicovideo.jp/search/sm9?sort=h&order=d&end=2020-12-31&start=2020-01-01',
'info_dict': {
'id': 'sm9',
'title': 'sm9'
},
'playlist_count': 31,
}]
def _real_extract(self, url):
query = self._match_id(url)
return self.playlist_result(self._entries(url, query), query, query)
class NicovideoSearchDateIE(NicovideoSearchBaseIE, SearchInfoExtractor):
IE_DESC = 'Nico video search, newest first'
IE_NAME = '%s:date' % NicovideoSearchIE.IE_NAME
_SEARCH_KEY = 'nicosearchdate'
_TESTS = [{
'url': 'nicosearchdateall:a',
'info_dict': {
'id': 'a',
'title': 'a'
},
'playlist_mincount': 1610,
}]
_START_DATE = datetime.date(2007, 1, 1)
_RESULTS_PER_PAGE = 32
_MAX_PAGES = 50
def _entries(self, url, item_id, start_date=None, end_date=None):
start_date, end_date = start_date or self._START_DATE, end_date or datetime.datetime.now().date()
# If the last page has a full page of videos, we need to break down the query interval further
last_page_len = len(list(self._get_entries_for_date(
url, item_id, start_date, end_date, self._MAX_PAGES,
note='Checking number of videos from {0} to {1}'.format(start_date, end_date))))
if (last_page_len == self._RESULTS_PER_PAGE and start_date != end_date):
midpoint = start_date + ((end_date - start_date) // 2)
for entry in itertools.chain(
iter(self._entries(url, item_id, midpoint, end_date)),
iter(self._entries(url, item_id, start_date, midpoint))):
yield entry
else:
self.to_screen('{0}: Downloading results from {1} to {2}'.format(item_id, start_date, end_date))
for entry in iter(self._get_entries_for_date(
url, item_id, start_date, end_date, note=' Downloading page %(page)s')):
yield entry
def _get_entries_for_date(self, url, item_id, start_date, end_date=None, page_num=None, note=None):
query = {
'start': compat_str(start_date),
'end': compat_str(end_date or start_date),
'sort': 'f',
'order': 'd',
}
if page_num:
query['page'] = compat_str(page_num)
for entry in iter(super(NicovideoSearchDateIE, self)._entries(url, item_id, query=query, note=note)):
yield entry
class NiconicoUserIE(InfoExtractor):
_VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/user/(?P<id>\d+)/?(?:$|[#?])'
_TEST = {
'url': 'https://www.nicovideo.jp/user/419948',
'info_dict': {
'id': '419948',
},
'playlist_mincount': 101,
}
_API_URL = "https://nvapi.nicovideo.jp/v1/users/%s/videos?sortKey=registeredAt&sortOrder=desc&pageSize=%s&page=%s"
_PAGE_SIZE = 100
_API_HEADERS = {
'X-Frontend-ID': '6',
'X-Frontend-Version': '0'
}
def _entries(self, list_id):
total_count = 1
count = page_num = 0
while count < total_count:
json_parsed = self._download_json(
self._API_URL % (list_id, self._PAGE_SIZE, page_num + 1), list_id,
headers=self._API_HEADERS,
note='Downloading JSON metadata%s' % (' page %d' % page_num if page_num else ''))
if not page_num:
total_count = int_or_none(json_parsed['data'].get('totalCount'))
for entry in json_parsed["data"]["items"]:
count += 1
yield self.url_result('https://www.nicovideo.jp/watch/%s' % entry['id'])
page_num += 1
def _real_extract(self, url):
list_id = self._match_id(url)
return self.playlist_result(self._entries(list_id), list_id)