aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorYen Chi Hsuan <yan12125@gmail.com>2016-09-05 13:55:07 +0800
committerYen Chi Hsuan <yan12125@gmail.com>2016-09-05 13:55:07 +0800
commit340e31ca7449197450e0a50c925cf9ebd95a76d4 (patch)
treea0133fce961c5da0c4be2a03f578f49cf543658b
parentabf2c79f954c065d6bf92a2129f04eb07942d022 (diff)
parent973dee491f24382161b1673d4d508dcede1da79e (diff)
downloadyoutube-dl-340e31ca7449197450e0a50c925cf9ebd95a76d4.tar.xz
Merge branch 'PeterDing-bilibili'
-rw-r--r--ChangeLog1
-rw-r--r--youtube_dl/extractor/bilibili.py33
2 files changed, 27 insertions, 7 deletions
diff --git a/ChangeLog b/ChangeLog
index 0be9b0fbb..f6858f1d8 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,6 +1,7 @@
version <unreleased>
Extractors
++ [bilibili] Support episodes (#10190)
+ [tvnoe] New extractor (#10524)
diff --git a/youtube_dl/extractor/bilibili.py b/youtube_dl/extractor/bilibili.py
index a332fbb69..8fa96d3a0 100644
--- a/youtube_dl/extractor/bilibili.py
+++ b/youtube_dl/extractor/bilibili.py
@@ -10,11 +10,12 @@ from ..utils import (
int_or_none,
float_or_none,
unified_timestamp,
+ urlencode_postdata,
)
class BiliBiliIE(InfoExtractor):
- _VALID_URL = r'https?://www\.bilibili\.(?:tv|com)/video/av(?P<id>\d+)'
+ _VALID_URL = r'https?://(?:www\.|bangumi\.|)bilibili\.(?:tv|com)/(?:video/av|anime/v/)(?P<id>\d+)'
_TESTS = [{
'url': 'http://www.bilibili.tv/video/av1074402/',
@@ -77,6 +78,17 @@ class BiliBiliIE(InfoExtractor):
'skip_download': True,
},
'expected_warnings': ['upload time'],
+ }, {
+ 'url': 'http://bangumi.bilibili.com/anime/v/40068',
+ 'md5': '08d539a0884f3deb7b698fb13ba69696',
+ 'info_dict': {
+ 'id': '40068',
+ 'ext': 'mp4',
+ 'duration': 1402.357,
+ 'title': '混沌武士 : 第7集 四面楚歌 A Risky Racket',
+ 'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
+ 'thumbnail': 're:^http?://.+\.jpg',
+ },
}]
_APP_KEY = '6f90a59ac58a4123'
@@ -84,13 +96,19 @@ class BiliBiliIE(InfoExtractor):
def _real_extract(self, url):
video_id = self._match_id(url)
-
webpage = self._download_webpage(url, video_id)
- cid = compat_parse_qs(self._search_regex(
- [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
- r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
- webpage, 'player parameters'))['cid'][0]
+ if 'anime/v' not in url:
+ cid = compat_parse_qs(self._search_regex(
+ [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
+ r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
+ webpage, 'player parameters'))['cid'][0]
+ else:
+ js = self._download_json(
+ 'http://bangumi.bilibili.com/web_api/get_source', video_id,
+ data=urlencode_postdata({'episode_id': video_id}),
+ headers={'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8'})
+ cid = js['result']['cid']
payload = 'appkey=%s&cid=%s&otype=json&quality=2&type=mp4' % (self._APP_KEY, cid)
sign = hashlib.md5((payload + self._BILIBILI_KEY).encode('utf-8')).hexdigest()
@@ -125,6 +143,7 @@ class BiliBiliIE(InfoExtractor):
description = self._html_search_meta('description', webpage)
timestamp = unified_timestamp(self._html_search_regex(
r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time', fatal=False))
+ thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
# TODO 'view_count' requires deobfuscating Javascript
info = {
@@ -132,7 +151,7 @@ class BiliBiliIE(InfoExtractor):
'title': title,
'description': description,
'timestamp': timestamp,
- 'thumbnail': self._html_search_meta('thumbnailUrl', webpage),
+ 'thumbnail': thumbnail,
'duration': float_or_none(video_info.get('timelength'), scale=1000),
}