-# coding: utf-8
-from __future__ import unicode_literals
-
+import base64
import hashlib
import itertools
-import json
+import functools
+import math
import re
from .common import InfoExtractor, SearchInfoExtractor
from ..compat import (
- compat_str,
compat_parse_qs,
compat_urlparse,
+ compat_urllib_parse_urlparse
)
from ..utils import (
ExtractorError,
- int_or_none,
+ InAdvancePagedList,
+ OnDemandPagedList,
+ filter_dict,
float_or_none,
+ int_or_none,
+ mimetype2ext,
+ parse_count,
parse_iso8601,
- try_get,
+ qualities,
smuggle_url,
+ srt_subtitles_timecode,
str_or_none,
strip_jsonp,
+ traverse_obj,
unified_timestamp,
unsmuggle_url,
urlencode_postdata,
+ url_or_none,
)
video/[aA][vV]|
anime/(?P<anime_id>\d+)/play\#
)(?P<id>\d+)|
- video/[bB][vV](?P<id_bv>[^/?#&]+)
+ (s/)?video/[bB][vV](?P<id_bv>[^/?#&]+)
)
(?:/?\?p=(?P<page>\d+))?
'''
_TESTS = [{
'url': 'http://www.bilibili.com/video/av1074402/',
- 'md5': '5f7d29e1a2872f3df0cf76b1f87d3788',
+ 'md5': '7ac275ec84a99a6552c5d229659a0fe1',
'info_dict': {
- 'id': '1074402',
- 'ext': 'flv',
+ 'id': '1074402_part1',
+ 'ext': 'mp4',
'title': '【金坷垃】金泡沫',
+ 'uploader_id': '156160',
+ 'uploader': '菊子桑',
+ 'upload_date': '20140420',
'description': 'md5:ce18c2a2d2193f0df2917d270f2e5923',
- 'duration': 308.067,
'timestamp': 1398012678,
- 'upload_date': '20140420',
- 'thumbnail': r're:^https?://.+\.jpg',
- 'uploader': '菊子桑',
- 'uploader_id': '156160',
+ 'tags': ['顶上去报复社会', '该来的总会来的', '金克拉是检验歌曲的唯一标准', '坷垃教主', '金坷垃', '邓紫棋', '治愈系坷垃'],
+ 'bv_id': 'BV11x411K7CN',
+ 'cid': '1554319',
+ 'thumbnail': 'http://i2.hdslb.com/bfs/archive/c79a8cf0347cd7a897c53a2f756e96aead128e8c.jpg',
+ 'duration': 308.36,
},
}, {
# Tested in BiliBiliBangumiIE
'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
'md5': '3f721ad1e75030cc06faf73587cfec57',
'info_dict': {
- 'id': '100643',
+ 'id': '100643_part1',
'ext': 'mp4',
'title': 'CHAOS;CHILD',
'description': '如果你是神明，并且能够让妄想成为现实。那你会进行怎么样的妄想？是淫靡的世界？独裁社会？毁灭性的制裁？还是……2015年，涩谷。从6年前发生的大灾害“涩谷地震”之后复兴了的这个街区里新设立的私立高中...',
},
'skip': 'Geo-restricted to China',
}, {
- # Title with double quotes
'url': 'http://www.bilibili.com/video/av8903802/',
'info_dict': {
- 'id': '8903802',
+ 'id': '8903802_part1',
+ 'ext': 'mp4',
'title': '阿滴英文｜英文歌分享#6 "Closer',
+ 'upload_date': '20170301',
'description': '滴妹今天唱Closer給你聽! 有史以来，被推最多次也是最久的歌曲，其实歌词跟我原本想像差蛮多的，不过还是好听！ 微博@阿滴英文',
+ 'timestamp': 1488382634,
+ 'uploader_id': '65880958',
+ 'uploader': '阿滴英文',
+ 'thumbnail': 'http://i2.hdslb.com/bfs/archive/49267ce20bc246be6304bf369a3ded0256854c23.jpg',
+ 'cid': '14694589',
+ 'duration': 554.117,
+ 'bv_id': 'BV13x41117TL',
+ 'tags': ['人文', '英语', '文化', '公开课', '阿滴英文'],
+ },
+ 'params': {
+ 'skip_download': True,
},
- 'playlist': [{
- 'info_dict': {
- 'id': '8903802_part1',
- 'ext': 'flv',
- 'title': '阿滴英文｜英文歌分享#6 "Closer',
- 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
- 'uploader': '阿滴英文',
- 'uploader_id': '65880958',
- 'timestamp': 1488382634,
- 'upload_date': '20170301',
- },
- 'params': {
- 'skip_download': True, # Test metadata only
- },
- }, {
- 'info_dict': {
- 'id': '8903802_part2',
- 'ext': 'flv',
- 'title': '阿滴英文｜英文歌分享#6 "Closer',
- 'description': 'md5:3b1b9e25b78da4ef87e9b548b88ee76a',
- 'uploader': '阿滴英文',
- 'uploader_id': '65880958',
- 'timestamp': 1488382634,
- 'upload_date': '20170301',
- },
- 'params': {
- 'skip_download': True, # Test metadata only
- },
- }]
}, {
# new BV video id format
'url': 'https://www.bilibili.com/video/BV1JE411F741',
'title': '物语中的人物是如何吐槽自己的OP的'
},
'playlist_count': 17,
+ }, {
+ # Correct matching of single and double quotes in title
+ 'url': 'https://www.bilibili.com/video/BV1NY411E7Rx/',
+ 'info_dict': {
+ 'id': '255513412_part1',
+ 'ext': 'mp4',
+ 'title': 'Vid"eo" Te\'st',
+ 'cid': '570602418',
+ 'thumbnail': 'http://i2.hdslb.com/bfs/archive/0c0de5a90b6d5b991b8dcc6cde0afbf71d564791.jpg',
+ 'upload_date': '20220408',
+ 'timestamp': 1649436552,
+ 'description': 'Vid"eo" Te\'st',
+ 'uploader_id': '1630758804',
+ 'bv_id': 'BV1NY411E7Rx',
+ 'duration': 60.394,
+ 'uploader': 'bili_31244483705',
+ 'tags': ['VLOG'],
+ },
+ 'params': {
+ 'skip_download': True,
+ },
}]
_APP_KEY = 'iVGUTjsxvpLeuDCf'
def _real_extract(self, url):
url, smuggled_data = unsmuggle_url(url, {})
- mobj = re.match(self._VALID_URL, url)
+ mobj = self._match_valid_url(url)
video_id = mobj.group('id_bv') or mobj.group('id')
av_id, bv_id = self._get_video_id_set(video_id, mobj.group('id_bv') is not None)
video_id = av_id
+ info = {}
anime_id = mobj.group('anime_id')
page_id = mobj.group('page')
webpage = self._download_webpage(url, video_id)
if 'anime/' not in url:
cid = self._search_regex(
- r'\bcid(?:["\']:|=)(\d+),["\']page(?:["\']:|=)' + compat_str(page_id), webpage, 'cid',
+ r'\bcid(?:["\']:|=)(\d+),["\']page(?:["\']:|=)' + str(page_id), webpage, 'cid',
default=None
) or self._search_regex(
r'\bcid(?:["\']:|=)(\d+)', webpage, 'cid',
}
headers.update(self.geo_verification_headers())
+ video_info = self._parse_json(
+ self._search_regex(r'window.__playinfo__\s*=\s*({.+?})</script>', webpage, 'video info', default=None) or '{}',
+ video_id, fatal=False)
+ video_info = video_info.get('data') or {}
+
+ durl = traverse_obj(video_info, ('dash', 'video'))
+ audios = traverse_obj(video_info, ('dash', 'audio')) or []
+ flac_audio = traverse_obj(video_info, ('dash', 'flac', 'audio'))
+ if flac_audio:
+ audios.append(flac_audio)
entries = []
RENDITIONS = ('qn=80&quality=80&type=', 'quality=2&type=mp4')
for num, rendition in enumerate(RENDITIONS, start=1):
payload = 'appkey=%s&cid=%s&otype=json&%s' % (self._APP_KEY, cid, rendition)
sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
-
- video_info = self._download_json(
- 'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
- video_id, note='Downloading video info page',
- headers=headers, fatal=num == len(RENDITIONS))
-
if not video_info:
- continue
+ video_info = self._download_json(
+ 'http://interface.bilibili.com/v2/playurl?%s&sign=%s' % (payload, sign),
+ video_id, note='Downloading video info page',
+ headers=headers, fatal=num == len(RENDITIONS))
+ if not video_info:
+ continue
- if 'durl' not in video_info:
+ if not durl and 'durl' not in video_info:
if num < len(RENDITIONS):
continue
self._report_error(video_info)
- for idx, durl in enumerate(video_info['durl']):
- formats = [{
- 'url': durl['url'],
- 'filesize': int_or_none(durl['size']),
- }]
- for backup_url in durl.get('backup_url', []):
+ formats = []
+ for idx, durl in enumerate(durl or video_info['durl']):
+ formats.append({
+ 'url': durl.get('baseUrl') or durl.get('base_url') or durl.get('url'),
+ 'ext': mimetype2ext(durl.get('mimeType') or durl.get('mime_type')),
+ 'fps': int_or_none(durl.get('frameRate') or durl.get('frame_rate')),
+ 'width': int_or_none(durl.get('width')),
+ 'height': int_or_none(durl.get('height')),
+ 'vcodec': durl.get('codecs'),
+ 'acodec': 'none' if audios else None,
+ 'tbr': float_or_none(durl.get('bandwidth'), scale=1000),
+ 'filesize': int_or_none(durl.get('size')),
+ })
+ for backup_url in traverse_obj(durl, 'backup_url', expected_type=list) or []:
formats.append({
'url': backup_url,
- # backup URLs have lower priorities
'quality': -2 if 'hd.mp4' in backup_url else -3,
})
- for a_format in formats:
- a_format.setdefault('http_headers', {}).update({
- 'Referer': url,
+ for audio in audios:
+ formats.append({
+ 'url': audio.get('baseUrl') or audio.get('base_url') or audio.get('url'),
+ 'ext': mimetype2ext(audio.get('mimeType') or audio.get('mime_type')),
+ 'fps': int_or_none(audio.get('frameRate') or audio.get('frame_rate')),
+ 'width': int_or_none(audio.get('width')),
+ 'height': int_or_none(audio.get('height')),
+ 'acodec': audio.get('codecs'),
+ 'vcodec': 'none',
+ 'tbr': float_or_none(audio.get('bandwidth'), scale=1000),
+ 'filesize': int_or_none(audio.get('size'))
+ })
+ for backup_url in traverse_obj(audio, 'backup_url', expected_type=list) or []:
+ formats.append({
+ 'url': backup_url,
+ # backup URLs have lower priorities
+ 'quality': -3,
})
- self._sort_formats(formats)
-
- entries.append({
- 'id': '%s_part%s' % (video_id, idx),
- 'duration': float_or_none(durl.get('length'), 1000),
- 'formats': formats,
- })
+ info.update({
+ 'id': video_id,
+ 'duration': float_or_none(durl.get('length'), 1000),
+ 'formats': formats,
+ 'http_headers': {
+ 'Referer': url,
+ },
+ })
break
- title = self._html_search_regex(
- (r'<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
- r'(?s)<h1[^>]*>(?P<title>.+?)</h1>'), webpage, 'title',
- group='title')
+ self._sort_formats(formats)
+
+ title = self._html_search_regex((
+ r'<h1[^>]+title=(["])(?P<content>[^"]+)',
+ r'<h1[^>]+title=([\'])(?P<content>[^\']+)',
+ r'(?s)<h1[^>]*>(?P<content>.+?)</h1>',
+ self._meta_regex('title')
+ ), webpage, 'title', group='content', fatal=False)
# Get part title for anthologies
if page_id is not None:
- # TODO: The json is already downloaded by _extract_anthology_entries. Don't redownload for each video
- part_title = try_get(
- self._download_json(
- "https://api.bilibili.com/x/player/pagelist?bvid=%s&jsonp=jsonp" % bv_id,
- video_id, note='Extracting videos in anthology'),
- lambda x: x['data'][int(page_id) - 1]['part'])
- title = part_title or title
+ # TODO: The json is already downloaded by _extract_anthology_entries. Don't redownload for each video.
+ part_info = traverse_obj(self._download_json(
+ f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
+ video_id, note='Extracting videos in anthology'), 'data', expected_type=list)
+ title = title if len(part_info) == 1 else traverse_obj(part_info, (int(page_id) - 1, 'part')) or title
description = self._html_search_meta('description', webpage)
timestamp = unified_timestamp(self._html_search_regex(
thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
# TODO 'view_count' requires deobfuscating Javascript
- info = {
- 'id': compat_str(video_id) if page_id is None else '%s_p%s' % (video_id, page_id),
+ info.update({
+ 'id': f'{video_id}_part{page_id or 1}',
'cid': cid,
'title': title,
'description': description,
'timestamp': timestamp,
'thumbnail': thumbnail,
'duration': float_or_none(video_info.get('timelength'), scale=1000),
- }
+ })
uploader_mobj = re.search(
r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]*>\s*(?P<name>[^<]+?)\s*<',
webpage)
if uploader_mobj:
info.update({
- 'uploader': uploader_mobj.group('name'),
+ 'uploader': uploader_mobj.group('name').strip(),
'uploader_id': uploader_mobj.group('id'),
})
info['uploader'] = self._html_search_meta(
'author', webpage, 'uploader', default=None)
- raw_danmaku = self._get_raw_danmaku(video_id, cid)
-
- raw_tags = self._get_tags(video_id)
- tags = list(map(lambda x: x['tag_name'], raw_tags))
-
top_level_info = {
- 'raw_danmaku': raw_danmaku,
- 'tags': tags,
- 'raw_tags': raw_tags,
+ 'tags': traverse_obj(self._download_json(
+ f'https://api.bilibili.com/x/tag/archive/tags?aid={video_id}',
+ video_id, fatal=False, note='Downloading tags'), ('data', ..., 'tag_name')),
}
- if self.get_param('getcomments', False):
- def get_comments():
- comments = self._get_all_comment_pages(video_id)
- return {
- 'comments': comments,
- 'comment_count': len(comments)
- }
- top_level_info['__post_extractor'] = get_comments
+ info['subtitles'] = {
+ 'danmaku': [{
+ 'ext': 'xml',
+ 'url': f'https://comment.bilibili.com/{cid}.xml',
+ }]
+ }
- '''
+ r'''
# Requires https://github.com/m13253/danmaku2ass which is licenced under GPL3
# See https://github.com/animelover1984/youtube-dl
+
+ raw_danmaku = self._download_webpage(
+ f'https://comment.bilibili.com/{cid}.xml', video_id, fatal=False, note='Downloading danmaku comments')
danmaku = NiconicoIE.CreateDanmaku(raw_danmaku, commentType='Bilibili', x=1024, y=576)
entries[0]['subtitles'] = {
'danmaku': [{
}
'''
+ top_level_info['__post_extractor'] = self.extract_comments(video_id)
+
for entry in entries:
entry.update(info)
if len(entries) == 1:
entries[0].update(top_level_info)
return entries[0]
- else:
- for idx, entry in enumerate(entries):
- entry['id'] = '%s_part%d' % (video_id, (idx + 1))
-
- global_info = {
- '_type': 'multi_video',
- 'id': compat_str(video_id),
- 'bv_id': bv_id,
- 'title': title,
- 'description': description,
- 'entries': entries,
- }
- global_info.update(info)
- global_info.update(top_level_info)
+ for idx, entry in enumerate(entries):
+ entry['id'] = '%s_part%d' % (video_id, (idx + 1))
- return global_info
+ return {
+ 'id': str(video_id),
+ 'bv_id': bv_id,
+ 'title': title,
+ 'description': description,
+ **info, **top_level_info
+ }
def _extract_anthology_entries(self, bv_id, video_id, webpage):
title = self._html_search_regex(
(r'<h1[^>]+\btitle=(["\'])(?P<title>(?:(?!\1).)+)\1',
- r'(?s)<h1[^>]*>(?P<title>.+?)</h1>'), webpage, 'title',
+ r'(?s)<h1[^>]*>(?P<title>.+?)</h1>',
+ r'<title>(?P<title>.+?)</title>'), webpage, 'title',
group='title')
json_data = self._download_json(
- "https://api.bilibili.com/x/player/pagelist?bvid=%s&jsonp=jsonp" % bv_id,
+ f'https://api.bilibili.com/x/player/pagelist?bvid={bv_id}&jsonp=jsonp',
video_id, note='Extracting videos in anthology')
- if len(json_data['data']) > 1:
+ if json_data['data']:
return self.playlist_from_matches(
json_data['data'], bv_id, title, ie=BiliBiliIE.ie_key(),
getter=lambda entry: 'https://www.bilibili.com/video/%s?p=%d' % (bv_id, entry['page']))
if response['code'] == -400:
raise ExtractorError('Video ID does not exist', expected=True, video_id=id)
elif response['code'] != 0:
- raise ExtractorError('Unknown error occurred during API check (code %s)' % response['code'], expected=True, video_id=id)
- return (response['data']['aid'], response['data']['bvid'])
-
- # recursive solution to getting every page of comments for the video
- # we can stop when we reach a page without any comments
- def _get_all_comment_pages(self, video_id, commentPageNumber=0):
- comment_url = "https://api.bilibili.com/x/v2/reply?jsonp=jsonp&pn=%s&type=1&oid=%s&sort=2&_=1567227301685" % (commentPageNumber, video_id)
- json_str = self._download_webpage(
- comment_url, video_id,
- note='Extracting comments from page %s' % (commentPageNumber))
- replies = json.loads(json_str)['data']['replies']
- if replies is None:
- return []
- return self._get_all_children(replies) + self._get_all_comment_pages(video_id, commentPageNumber + 1)
-
- # extracts all comments in the tree
- def _get_all_children(self, replies):
- if replies is None:
- return []
-
- ret = []
- for reply in replies:
- author = reply['member']['uname']
- author_id = reply['member']['mid']
- id = reply['rpid']
- text = reply['content']['message']
- timestamp = reply['ctime']
- parent = reply['parent'] if reply['parent'] != 0 else 'root'
-
- comment = {
- "author": author,
- "author_id": author_id,
- "id": id,
- "text": text,
- "timestamp": timestamp,
- "parent": parent,
- }
- ret.append(comment)
-
- # from the JSON, the comment structure seems arbitrarily deep, but I could be wrong.
- # Regardless, this should work.
- ret += self._get_all_children(reply['replies'])
+ raise ExtractorError(f'Unknown error occurred during API check (code {response["code"]})',
+ expected=True, video_id=id)
+ return response['data']['aid'], response['data']['bvid']
- return ret
-
- def _get_raw_danmaku(self, video_id, cid):
- # This will be useful if I decide to scrape all pages instead of doing them individually
- # cid_url = "https://www.bilibili.com/widget/getPageList?aid=%s" % (video_id)
- # cid_str = self._download_webpage(cid_url, video_id, note=False)
- # cid = json.loads(cid_str)[0]['cid']
-
- danmaku_url = "https://comment.bilibili.com/%s.xml" % (cid)
- danmaku = self._download_webpage(danmaku_url, video_id, note='Downloading danmaku comments')
- return danmaku
-
- def _get_tags(self, video_id):
- tags_url = "https://api.bilibili.com/x/tag/archive/tags?aid=%s" % (video_id)
- tags_json = self._download_json(tags_url, video_id, note='Downloading tags')
- return tags_json['data']
+ def _get_comments(self, video_id, commentPageNumber=0):
+ for idx in itertools.count(1):
+ replies = traverse_obj(
+ self._download_json(
+ f'https://api.bilibili.com/x/v2/reply?pn={idx}&oid={video_id}&type=1&jsonp=jsonp&sort=2&_=1567227301685',
+ video_id, note=f'Extracting comments from page {idx}', fatal=False),
+ ('data', 'replies'))
+ if not replies:
+ return
+ for children in map(self._get_all_children, replies):
+ yield from children
+
+ def _get_all_children(self, reply):
+ yield {
+ 'author': traverse_obj(reply, ('member', 'uname')),
+ 'author_id': traverse_obj(reply, ('member', 'mid')),
+ 'id': reply.get('rpid'),
+ 'text': traverse_obj(reply, ('content', 'message')),
+ 'timestamp': reply.get('ctime'),
+ 'parent': reply.get('parent') or 'root',
+ }
+ for children in map(self._get_all_children, reply.get('replies') or []):
+ yield from children
class BiliBiliBangumiIE(InfoExtractor):
season_info.get('bangumi_title'), season_info.get('evaluate'))
-class BilibiliChannelIE(InfoExtractor):
- _VALID_URL = r'https?://space.bilibili\.com/(?P<id>\d+)'
- _API_URL = "https://api.bilibili.com/x/space/arc/search?mid=%s&pn=%d&jsonp=jsonp"
+class BilibiliSpaceBaseIE(InfoExtractor):
+ def _extract_playlist(self, fetch_page, get_metadata, get_entries):
+ first_page = fetch_page(1)
+ metadata = get_metadata(first_page)
+
+ paged_list = InAdvancePagedList(
+ lambda idx: get_entries(fetch_page(idx) if idx > 1 else first_page),
+ metadata['page_count'], metadata['page_size'])
+
+ return metadata, paged_list
+
+
+class BilibiliSpaceVideoIE(BilibiliSpaceBaseIE):
+ _VALID_URL = r'https?://space\.bilibili\.com/(?P<id>\d+)(?P<video>/video)?/?(?:[?#]|$)'
_TESTS = [{
'url': 'https://space.bilibili.com/3985676/video',
- 'info_dict': {},
- 'playlist_mincount': 112,
+ 'info_dict': {
+ 'id': '3985676',
+ },
+ 'playlist_mincount': 178,
}]
- def _entries(self, list_id):
- count, max_count = 0, None
+ def _real_extract(self, url):
+ playlist_id, is_video_url = self._match_valid_url(url).group('id', 'video')
+ if not is_video_url:
+ self.to_screen('A channel URL was given. Only the channel\'s videos will be downloaded. '
+ 'To download audios, add a "/audio" to the URL')
+
+ def fetch_page(page_idx):
+ return self._download_json(
+ 'https://api.bilibili.com/x/space/arc/search', playlist_id,
+ note=f'Downloading page {page_idx}',
+ query={'mid': playlist_id, 'pn': page_idx, 'jsonp': 'jsonp'})['data']
+
+ def get_metadata(page_data):
+ page_size = page_data['page']['ps']
+ entry_count = page_data['page']['count']
+ return {
+ 'page_count': math.ceil(entry_count / page_size),
+ 'page_size': page_size,
+ }
- for page_num in itertools.count(1):
- data = self._parse_json(
- self._download_webpage(
- self._API_URL % (list_id, page_num), list_id,
- note='Downloading page %d' % page_num),
- list_id)['data']
+ def get_entries(page_data):
+ for entry in traverse_obj(page_data, ('list', 'vlist')) or []:
+ yield self.url_result(f'https://www.bilibili.com/video/{entry["bvid"]}', BiliBiliIE, entry['bvid'])
- max_count = max_count or try_get(data, lambda x: x['page']['count'])
+ metadata, paged_list = self._extract_playlist(fetch_page, get_metadata, get_entries)
+ return self.playlist_result(paged_list, playlist_id)
- entries = try_get(data, lambda x: x['list']['vlist'])
- if not entries:
- return
- for entry in entries:
- yield self.url_result(
- 'https://www.bilibili.com/video/%s' % entry['bvid'],
- BiliBiliIE.ie_key(), entry['bvid'])
- count += len(entries)
- if max_count and count >= max_count:
- return
+class BilibiliSpaceAudioIE(BilibiliSpaceBaseIE):
+ _VALID_URL = r'https?://space\.bilibili\.com/(?P<id>\d+)/audio'
+ _TESTS = [{
+ 'url': 'https://space.bilibili.com/3985676/audio',
+ 'info_dict': {
+ 'id': '3985676',
+ },
+ 'playlist_mincount': 1,
+ }]
+
+ def _real_extract(self, url):
+ playlist_id = self._match_id(url)
+
+ def fetch_page(page_idx):
+ return self._download_json(
+ 'https://api.bilibili.com/audio/music-service/web/song/upper', playlist_id,
+ note=f'Downloading page {page_idx}',
+ query={'uid': playlist_id, 'pn': page_idx, 'ps': 30, 'order': 1, 'jsonp': 'jsonp'})['data']
+
+ def get_metadata(page_data):
+ return {
+ 'page_count': page_data['pageCount'],
+ 'page_size': page_data['pageSize'],
+ }
+
+ def get_entries(page_data):
+ for entry in page_data.get('data', []):
+ yield self.url_result(f'https://www.bilibili.com/audio/au{entry["id"]}', BilibiliAudioIE, entry['id'])
+
+ metadata, paged_list = self._extract_playlist(fetch_page, get_metadata, get_entries)
+ return self.playlist_result(paged_list, playlist_id)
+
+
+class BilibiliSpacePlaylistIE(BilibiliSpaceBaseIE):
+ _VALID_URL = r'https?://space.bilibili\.com/(?P<mid>\d+)/channel/collectiondetail\?sid=(?P<sid>\d+)'
+ _TESTS = [{
+ 'url': 'https://space.bilibili.com/2142762/channel/collectiondetail?sid=57445',
+ 'info_dict': {
+ 'id': '2142762_57445',
+ 'title': '《底特律 变人》'
+ },
+ 'playlist_mincount': 31,
+ }]
def _real_extract(self, url):
- list_id = self._match_id(url)
- return self.playlist_result(self._entries(list_id), list_id)
+ mid, sid = self._match_valid_url(url).group('mid', 'sid')
+ playlist_id = f'{mid}_{sid}'
+
+ def fetch_page(page_idx):
+ return self._download_json(
+ 'https://api.bilibili.com/x/polymer/space/seasons_archives_list',
+ playlist_id, note=f'Downloading page {page_idx}',
+ query={'mid': mid, 'season_id': sid, 'page_num': page_idx, 'page_size': 30})['data']
+
+ def get_metadata(page_data):
+ page_size = page_data['page']['page_size']
+ entry_count = page_data['page']['total']
+ return {
+ 'page_count': math.ceil(entry_count / page_size),
+ 'page_size': page_size,
+ 'title': traverse_obj(page_data, ('meta', 'name'))
+ }
+
+ def get_entries(page_data):
+ for entry in page_data.get('archives', []):
+ yield self.url_result(f'https://www.bilibili.com/video/{entry["bvid"]}',
+ BiliBiliIE, entry['bvid'])
+
+ metadata, paged_list = self._extract_playlist(fetch_page, get_metadata, get_entries)
+ return self.playlist_result(paged_list, playlist_id, metadata['title'])
+
+
+class BilibiliCategoryIE(InfoExtractor):
+ IE_NAME = 'Bilibili category extractor'
+ _MAX_RESULTS = 1000000
+ _VALID_URL = r'https?://www\.bilibili\.com/v/[a-zA-Z]+\/[a-zA-Z]+'
+ _TESTS = [{
+ 'url': 'https://www.bilibili.com/v/kichiku/mad',
+ 'info_dict': {
+ 'id': 'kichiku: mad',
+ 'title': 'kichiku: mad'
+ },
+ 'playlist_mincount': 45,
+ 'params': {
+ 'playlistend': 45
+ }
+ }]
+
+ def _fetch_page(self, api_url, num_pages, query, page_num):
+ parsed_json = self._download_json(
+ api_url, query, query={'Search_key': query, 'pn': page_num},
+ note='Extracting results from page %s of %s' % (page_num, num_pages))
+
+ video_list = traverse_obj(parsed_json, ('data', 'archives'), expected_type=list)
+ if not video_list:
+ raise ExtractorError('Failed to retrieve video list for page %d' % page_num)
+
+ for video in video_list:
+ yield self.url_result(
+ 'https://www.bilibili.com/video/%s' % video['bvid'], 'BiliBili', video['bvid'])
+
+ def _entries(self, category, subcategory, query):
+ # map of categories : subcategories : RIDs
+ rid_map = {
+ 'kichiku': {
+ 'mad': 26,
+ 'manual_vocaloid': 126,
+ 'guide': 22,
+ 'theatre': 216,
+ 'course': 127
+ },
+ }
+
+ if category not in rid_map:
+ raise ExtractorError(
+ f'The category {category} isn\'t supported. Supported categories: {list(rid_map.keys())}')
+ if subcategory not in rid_map[category]:
+ raise ExtractorError(
+ f'The subcategory {subcategory} isn\'t supported for this category. Supported subcategories: {list(rid_map[category].keys())}')
+ rid_value = rid_map[category][subcategory]
+
+ api_url = 'https://api.bilibili.com/x/web-interface/newlist?rid=%d&type=1&ps=20&jsonp=jsonp' % rid_value
+ page_json = self._download_json(api_url, query, query={'Search_key': query, 'pn': '1'})
+ page_data = traverse_obj(page_json, ('data', 'page'), expected_type=dict)
+ count, size = int_or_none(page_data.get('count')), int_or_none(page_data.get('size'))
+ if count is None or not size:
+ raise ExtractorError('Failed to calculate either page count or size')
+
+ num_pages = math.ceil(count / size)
+
+ return OnDemandPagedList(functools.partial(
+ self._fetch_page, api_url, num_pages, query), size)
+
+ def _real_extract(self, url):
+ u = compat_urllib_parse_urlparse(url)
+ category, subcategory = u.path.split('/')[2:4]
+ query = '%s: %s' % (category, subcategory)
+
+ return self.playlist_result(self._entries(category, subcategory, query), query, query)
class BiliBiliSearchIE(SearchInfoExtractor):
- IE_DESC = 'Bilibili video search, "bilisearch" keyword'
+ IE_DESC = 'Bilibili video search'
_MAX_RESULTS = 100000
_SEARCH_KEY = 'bilisearch'
- MAX_NUMBER_OF_RESULTS = 1000
- def _get_n_results(self, query, n):
- """Get a specified number of results for a query"""
-
- entries = []
- pageNumber = 0
- while True:
- pageNumber += 1
- # FIXME
- api_url = "https://api.bilibili.com/x/web-interface/search/type?context=&page=%s&order=pubdate&keyword=%s&duration=0&tids_2=&__refresh__=true&search_type=video&tids=0&highlight=1" % (pageNumber, query)
- json_str = self._download_webpage(
- api_url, "None", query={"Search_key": query},
- note='Extracting results from page %s' % pageNumber)
- data = json.loads(json_str)['data']
-
- # FIXME: this is hideous
- if "result" not in data:
- return {
- '_type': 'playlist',
- 'id': query,
- 'entries': entries[:n]
- }
-
- videos = data['result']
+ def _search_results(self, query):
+ for page_num in itertools.count(1):
+ videos = self._download_json(
+ 'https://api.bilibili.com/x/web-interface/search/type', query,
+ note=f'Extracting results from page {page_num}', query={
+ 'Search_key': query,
+ 'keyword': query,
+ 'page': page_num,
+ 'context': '',
+ 'duration': 0,
+ 'tids_2': '',
+ '__refresh__': 'true',
+ 'search_type': 'video',
+ 'tids': 0,
+ 'highlight': 1,
+ })['data'].get('result')
+ if not videos:
+ break
for video in videos:
- e = self.url_result(video['arcurl'], 'BiliBili', compat_str(video['aid']))
- entries.append(e)
-
- if(len(entries) >= n or len(videos) >= BiliBiliSearchIE.MAX_NUMBER_OF_RESULTS):
- return {
- '_type': 'playlist',
- 'id': query,
- 'entries': entries[:n]
- }
+ yield self.url_result(video['arcurl'], 'BiliBili', str(video['aid']))
class BilibiliAudioBaseIE(InfoExtractor):
'vcodec': 'none'
}]
+ for a_format in formats:
+ a_format.setdefault('http_headers', {}).update({
+ 'Referer': url,
+ })
+
song = self._call_api('song/info', au_id)
title = song['title']
statistic = song.get('statistic') or {}
return self.url_result(
'http://www.bilibili.tv/video/av%s/' % video_id,
ie=BiliBiliIE.ie_key(), video_id=video_id)
+
+
+class BiliIntlBaseIE(InfoExtractor):
+ _API_URL = 'https://api.bilibili.tv/intl/gateway'
+ _NETRC_MACHINE = 'biliintl'
+
+ def _call_api(self, endpoint, *args, **kwargs):
+ json = self._download_json(self._API_URL + endpoint, *args, **kwargs)
+ if json.get('code'):
+ if json['code'] in (10004004, 10004005, 10023006):
+ self.raise_login_required()
+ elif json['code'] == 10004001:
+ self.raise_geo_restricted()
+ else:
+ if json.get('message') and str(json['code']) != json['message']:
+ errmsg = f'{kwargs.get("errnote", "Unable to download JSON metadata")}: {self.IE_NAME} said: {json["message"]}'
+ else:
+ errmsg = kwargs.get('errnote', 'Unable to download JSON metadata')
+ if kwargs.get('fatal'):
+ raise ExtractorError(errmsg)
+ else:
+ self.report_warning(errmsg)
+ return json.get('data')
+
+ def json2srt(self, json):
+ data = '\n\n'.join(
+ f'{i + 1}\n{srt_subtitles_timecode(line["from"])} --> {srt_subtitles_timecode(line["to"])}\n{line["content"]}'
+ for i, line in enumerate(traverse_obj(json, (
+ 'body', lambda _, l: l['content'] and l['from'] and l['to']))))
+ return data
+
+ def _get_subtitles(self, *, ep_id=None, aid=None):
+ sub_json = self._call_api(
+ '/web/v2/subtitle', ep_id or aid, fatal=False,
+ note='Downloading subtitles list', errnote='Unable to download subtitles list',
+ query=filter_dict({
+ 'platform': 'web',
+ 's_locale': 'en_US',
+ 'episode_id': ep_id,
+ 'aid': aid,
+ })) or {}
+ subtitles = {}
+ for sub in sub_json.get('subtitles') or []:
+ sub_url = sub.get('url')
+ if not sub_url:
+ continue
+ sub_data = self._download_json(
+ sub_url, ep_id or aid, errnote='Unable to download subtitles', fatal=False,
+ note='Downloading subtitles%s' % f' for {sub["lang"]}' if sub.get('lang') else '')
+ if not sub_data:
+ continue
+ subtitles.setdefault(sub.get('lang_key', 'en'), []).append({
+ 'ext': 'srt',
+ 'data': self.json2srt(sub_data)
+ })
+ return subtitles
+
+ def _get_formats(self, *, ep_id=None, aid=None):
+ video_json = self._call_api(
+ '/web/playurl', ep_id or aid, note='Downloading video formats',
+ errnote='Unable to download video formats', query=filter_dict({
+ 'platform': 'web',
+ 'ep_id': ep_id,
+ 'aid': aid,
+ }))
+ video_json = video_json['playurl']
+ formats = []
+ for vid in video_json.get('video') or []:
+ video_res = vid.get('video_resource') or {}
+ video_info = vid.get('stream_info') or {}
+ if not video_res.get('url'):
+ continue
+ formats.append({
+ 'url': video_res['url'],
+ 'ext': 'mp4',
+ 'format_note': video_info.get('desc_words'),
+ 'width': video_res.get('width'),
+ 'height': video_res.get('height'),
+ 'vbr': video_res.get('bandwidth'),
+ 'acodec': 'none',
+ 'vcodec': video_res.get('codecs'),
+ 'filesize': video_res.get('size'),
+ })
+ for aud in video_json.get('audio_resource') or []:
+ if not aud.get('url'):
+ continue
+ formats.append({
+ 'url': aud['url'],
+ 'ext': 'mp4',
+ 'abr': aud.get('bandwidth'),
+ 'acodec': aud.get('codecs'),
+ 'vcodec': 'none',
+ 'filesize': aud.get('size'),
+ })
+
+ self._sort_formats(formats)
+ return formats
+
+ def _extract_video_info(self, video_data, *, ep_id=None, aid=None):
+ return {
+ 'id': ep_id or aid,
+ 'title': video_data.get('title_display') or video_data.get('title'),
+ 'thumbnail': video_data.get('cover'),
+ 'episode_number': int_or_none(self._search_regex(
+ r'^E(\d+)(?:$| - )', video_data.get('title_display') or '', 'episode number', default=None)),
+ 'formats': self._get_formats(ep_id=ep_id, aid=aid),
+ 'subtitles': self._get_subtitles(ep_id=ep_id, aid=aid),
+ 'extractor_key': BiliIntlIE.ie_key(),
+ }
+
+ def _perform_login(self, username, password):
+ try:
+ from Cryptodome.PublicKey import RSA
+ from Cryptodome.Cipher import PKCS1_v1_5
+ except ImportError:
+ try:
+ from Crypto.PublicKey import RSA
+ from Crypto.Cipher import PKCS1_v1_5
+ except ImportError:
+ raise ExtractorError('pycryptodomex not found. Please install', expected=True)
+
+ key_data = self._download_json(
+ 'https://passport.bilibili.tv/x/intl/passport-login/web/key?lang=en-US', None,
+ note='Downloading login key', errnote='Unable to download login key')['data']
+
+ public_key = RSA.importKey(key_data['key'])
+ password_hash = PKCS1_v1_5.new(public_key).encrypt((key_data['hash'] + password).encode('utf-8'))
+ login_post = self._download_json(
+ 'https://passport.bilibili.tv/x/intl/passport-login/web/login/password?lang=en-US', None, data=urlencode_postdata({
+ 'username': username,
+ 'password': base64.b64encode(password_hash).decode('ascii'),
+ 'keep_me': 'true',
+ 's_locale': 'en_US',
+ 'isTrusted': 'true'
+ }), note='Logging in', errnote='Unable to log in')
+ if login_post.get('code'):
+ if login_post.get('message'):
+ raise ExtractorError(f'Unable to log in: {self.IE_NAME} said: {login_post["message"]}', expected=True)
+ else:
+ raise ExtractorError('Unable to log in')
+
+
+class BiliIntlIE(BiliIntlBaseIE):
+ _VALID_URL = r'https?://(?:www\.)?bili(?:bili\.tv|intl\.com)/(?:[a-zA-Z]{2}/)?(play/(?P<season_id>\d+)/(?P<ep_id>\d+)|video/(?P<aid>\d+))'
+ _TESTS = [{
+ # Bstation page
+ 'url': 'https://www.bilibili.tv/en/play/34613/341736',
+ 'info_dict': {
+ 'id': '341736',
+ 'ext': 'mp4',
+ 'title': 'E2 - The First Night',
+ 'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
+ 'episode_number': 2,
+ }
+ }, {
+ # Non-Bstation page
+ 'url': 'https://www.bilibili.tv/en/play/1033760/11005006',
+ 'info_dict': {
+ 'id': '11005006',
+ 'ext': 'mp4',
+ 'title': 'E3 - Who?',
+ 'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
+ 'episode_number': 3,
+ }
+ }, {
+ # Subtitle with empty content
+ 'url': 'https://www.bilibili.tv/en/play/1005144/10131790',
+ 'info_dict': {
+ 'id': '10131790',
+ 'ext': 'mp4',
+ 'title': 'E140 - Two Heartbeats: Kabuto\'s Trap',
+ 'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
+ 'episode_number': 140,
+ },
+ 'skip': 'According to the copyright owner\'s request, you may only watch the video after you log in.'
+ }, {
+ 'url': 'https://www.biliintl.com/en/play/34613/341736',
+ 'only_matching': True,
+ }, {
+ # User-generated content (as opposed to a series licensed from a studio)
+ 'url': 'https://bilibili.tv/en/video/2019955076',
+ 'only_matching': True,
+ }, {
+ # No language in URL
+ 'url': 'https://www.bilibili.tv/video/2019955076',
+ 'only_matching': True,
+ }, {
+ # Uppercase language in URL
+ 'url': 'https://www.bilibili.tv/EN/video/2019955076',
+ 'only_matching': True,
+ }]
+
+ def _real_extract(self, url):
+ season_id, ep_id, aid = self._match_valid_url(url).group('season_id', 'ep_id', 'aid')
+ video_id = ep_id or aid
+ webpage = self._download_webpage(url, video_id)
+ # Bstation layout
+ initial_data = (
+ self._search_json(r'window\.__INITIAL_(?:DATA|STATE)__\s*=', webpage, 'preload state', video_id, default={})
+ or self._search_nuxt_data(webpage, video_id, '__initialState', fatal=False, traverse=None))
+ video_data = traverse_obj(
+ initial_data, ('OgvVideo', 'epDetail'), ('UgcVideo', 'videoData'), ('ugc', 'archive'), expected_type=dict)
+
+ if season_id and not video_data:
+ # Non-Bstation layout, read through episode list
+ season_json = self._call_api(f'/web/v2/ogv/play/episodes?season_id={season_id}&platform=web', video_id)
+ video_data = traverse_obj(season_json,
+ ('sections', ..., 'episodes', lambda _, v: str(v['episode_id']) == ep_id),
+ expected_type=dict, get_all=False)
+ return self._extract_video_info(video_data or {}, ep_id=ep_id, aid=aid)
+
+
+class BiliIntlSeriesIE(BiliIntlBaseIE):
+ _VALID_URL = r'https?://(?:www\.)?bili(?:bili\.tv|intl\.com)/(?:[a-zA-Z]{2}/)?play/(?P<id>\d+)/?(?:[?#]|$)'
+ _TESTS = [{
+ 'url': 'https://www.bilibili.tv/en/play/34613',
+ 'playlist_mincount': 15,
+ 'info_dict': {
+ 'id': '34613',
+ 'title': 'Fly Me to the Moon',
+ 'description': 'md5:a861ee1c4dc0acfad85f557cc42ac627',
+ 'categories': ['Romance', 'Comedy', 'Slice of life'],
+ 'thumbnail': r're:^https://pic\.bstarstatic\.com/ogv/.+\.png$',
+ 'view_count': int,
+ },
+ 'params': {
+ 'skip_download': True,
+ },
+ }, {
+ 'url': 'https://www.biliintl.com/en/play/34613',
+ 'only_matching': True,
+ }, {
+ 'url': 'https://www.biliintl.com/EN/play/34613',
+ 'only_matching': True,
+ }]
+
+ def _entries(self, series_id):
+ series_json = self._call_api(f'/web/v2/ogv/play/episodes?season_id={series_id}&platform=web', series_id)
+ for episode in traverse_obj(series_json, ('sections', ..., 'episodes', ...), expected_type=dict, default=[]):
+ episode_id = str(episode.get('episode_id'))
+ yield self._extract_video_info(episode, ep_id=episode_id)
+
+ def _real_extract(self, url):
+ series_id = self._match_id(url)
+ series_info = self._call_api(f'/web/v2/ogv/play/season_info?season_id={series_id}&platform=web', series_id).get('season') or {}
+ return self.playlist_result(
+ self._entries(series_id), series_id, series_info.get('title'), series_info.get('description'),
+ categories=traverse_obj(series_info, ('styles', ..., 'title'), expected_type=str_or_none),
+ thumbnail=url_or_none(series_info.get('horizontal_cover')), view_count=parse_count(series_info.get('view')))
+
+
+class BiliLiveIE(InfoExtractor):
+ _VALID_URL = r'https?://live.bilibili.com/(?P<id>\d+)'
+
+ _TESTS = [{
+ 'url': 'https://live.bilibili.com/196',
+ 'info_dict': {
+ 'id': '33989',
+ 'description': "周六杂谈回，其他时候随机游戏。 | \n录播：@下播型泛式录播组。 | \n直播通知群（全员禁言）：666906670，902092584，59971⑧481 （功能一样，别多加）",
+ 'ext': 'flv',
+ 'title': "太空狼人杀联动，不被爆杀就算赢",
+ 'thumbnail': "https://i0.hdslb.com/bfs/live/new_room_cover/e607bc1529057ef4b332e1026e62cf46984c314d.jpg",
+ 'timestamp': 1650802769,
+ },
+ 'skip': 'not live'
+ }, {
+ 'url': 'https://live.bilibili.com/196?broadcast_type=0&is_room_feed=1?spm_id_from=333.999.space_home.strengthen_live_card.click',
+ 'only_matching': True
+ }]
+
+ _FORMATS = {
+ 80: {'format_id': 'low', 'format_note': '流畅'},
+ 150: {'format_id': 'high_res', 'format_note': '高清'},
+ 250: {'format_id': 'ultra_high_res', 'format_note': '超清'},
+ 400: {'format_id': 'blue_ray', 'format_note': '蓝光'},
+ 10000: {'format_id': 'source', 'format_note': '原画'},
+ 20000: {'format_id': '4K', 'format_note': '4K'},
+ 30000: {'format_id': 'dolby', 'format_note': '杜比'},
+ }
+
+ _quality = staticmethod(qualities(list(_FORMATS)))
+
+ def _call_api(self, path, room_id, query):
+ api_result = self._download_json(f'https://api.live.bilibili.com/{path}', room_id, query=query)
+ if api_result.get('code') != 0:
+ raise ExtractorError(api_result.get('message') or 'Unable to download JSON metadata')
+ return api_result.get('data') or {}
+
+ def _parse_formats(self, qn, fmt):
+ for codec in fmt.get('codec') or []:
+ if codec.get('current_qn') != qn:
+ continue
+ for url_info in codec['url_info']:
+ yield {
+ 'url': f'{url_info["host"]}{codec["base_url"]}{url_info["extra"]}',
+ 'ext': fmt.get('format_name'),
+ 'vcodec': codec.get('codec_name'),
+ 'quality': self._quality(qn),
+ **self._FORMATS[qn],
+ }
+
+ def _real_extract(self, url):
+ room_id = self._match_id(url)
+ room_data = self._call_api('room/v1/Room/get_info', room_id, {'id': room_id})
+ if room_data.get('live_status') == 0:
+ raise ExtractorError('Streamer is not live', expected=True)
+
+ formats = []
+ for qn in self._FORMATS.keys():
+ stream_data = self._call_api('xlive/web-room/v2/index/getRoomPlayInfo', room_id, {
+ 'room_id': room_id,
+ 'qn': qn,
+ 'codec': '0,1',
+ 'format': '0,2',
+ 'mask': '0',
+ 'no_playurl': '0',
+ 'platform': 'web',
+ 'protocol': '0,1',
+ })
+ for fmt in traverse_obj(stream_data, ('playurl_info', 'playurl', 'stream', ..., 'format', ...)) or []:
+ formats.extend(self._parse_formats(qn, fmt))
+ self._sort_formats(formats)
+
+ return {
+ 'id': room_id,
+ 'title': room_data.get('title'),
+ 'description': room_data.get('description'),
+ 'thumbnail': room_data.get('user_cover'),
+ 'timestamp': stream_data.get('live_time'),
+ 'formats': formats,
+ 'http_headers': {
+ 'Referer': url,
+ },
+ }
projects / yt-dlp.git / blobdiff