2 from __future__
import unicode_literals
6 from .common
import InfoExtractor
12 from ..compat
import compat_urlparse
15 class UDNEmbedIE(InfoExtractor
):
17 _PROTOCOL_RELATIVE_VALID_URL
= r
'//video\.udn\.com/(?:embed|play)/news/(?P<id>\d+)'
18 _VALID_URL
= r
'https?:' + _PROTOCOL_RELATIVE_VALID_URL
20 'url': 'http://video.udn.com/embed/news/300040',
24 'title': '生物老師男變女 全校挺"做自己"',
25 'thumbnail': r
're:^https?://.*\.jpg$',
29 'skip_download': True,
31 'expected_warnings': ['Failed to parse JSON Expecting value'],
33 'url': 'https://video.udn.com/embed/news/300040',
34 'only_matching': True,
36 # From https://video.udn.com/news/303776
37 'url': 'https://video.udn.com/play/news/303776',
38 'only_matching': True,
41 def _real_extract(self
, url
):
42 video_id
= self
._match
_id
(url
)
44 page
= self
._download
_webpage
(url
, video_id
)
46 options_str
= self
._html
_search
_regex
(
47 r
'var\s+options\s*=\s*([^;]+);', page
, 'options')
48 trans_options_str
= js_to_json(options_str
)
49 options
= self
._parse
_json
(trans_options_str
, 'options', fatal
=False) or {}
51 video_urls
= options
['video']
52 title
= options
['title']
53 poster
= options
.get('poster')
55 video_urls
= self
._parse
_json
(self
._html
_search
_regex
(
56 r
'"video"\s*:\s*({.+?})\s*,', trans_options_str
, 'video urls'), 'video urls')
57 title
= self
._html
_search
_regex
(
58 r
"title\s*:\s*'(.+?)'\s*,", options_str
, 'title')
59 poster
= self
._html
_search
_regex
(
60 r
"poster\s*:\s*'(.+?)'\s*,", options_str
, 'poster', default
=None)
62 if video_urls
.get('youtube'):
63 return self
.url_result(video_urls
.get('youtube'), 'Youtube')
66 for video_type
, api_url
in video_urls
.items():
70 video_url
= self
._download
_webpage
(
71 compat_urlparse
.urljoin(url
, api_url
), video_id
,
72 note
='retrieve url for %s video' % video_type
)
74 ext
= determine_ext(video_url
)
76 formats
.extend(self
._extract
_m
3u8_formats
(
77 video_url
, video_id
, ext
='mp4', m3u8_id
='hls'))
79 formats
.extend(self
._extract
_f
4m
_formats
(
80 video_url
, video_id
, f4m_id
='hds'))
82 mobj
= re
.search(r
'_(?P<height>\d+)p_(?P<tbr>\d+)\.mp4', video_url
)
85 # video_type may be 'mp4', which confuses YoutubeDL
86 'format_id': 'http-' + video_type
,
90 'height': int_or_none(mobj
.group('height')),
91 'tbr': int_or_none(mobj
.group('tbr')),
93 formats
.append(a_format
)
95 self
._sort
_formats
(formats
)