aboutsummaryrefslogtreecommitdiff
path: root/youtube_dl/extractor
diff options
context:
space:
mode:
authorremitamine <remitamine@gmail.com>2016-03-17 16:06:25 +0100
committerremitamine <remitamine@gmail.com>2016-03-17 16:11:29 +0100
commit9f02ff537c6ddfd3f1ea3586f3e44f0ec07a2aea (patch)
treee15f6a4c7005f40bcf35e927c3aee5b01e32aefa /youtube_dl/extractor
parent0436ec0e7a4683539bc7844511ba76fbcab03f7b (diff)
[theplatform] extract brightcove once formats
Diffstat (limited to 'youtube_dl/extractor')
-rw-r--r--youtube_dl/extractor/theplatform.py22
1 files changed, 11 insertions, 11 deletions
diff --git a/youtube_dl/extractor/theplatform.py b/youtube_dl/extractor/theplatform.py
index 9a57b49df..ffe7c57ad 100644
--- a/youtube_dl/extractor/theplatform.py
+++ b/youtube_dl/extractor/theplatform.py
@@ -8,13 +8,12 @@ import binascii
import hashlib
-from .common import InfoExtractor
+from .once import OnceIE
from ..compat import (
compat_parse_qs,
compat_urllib_parse_urlparse,
)
from ..utils import (
- determine_ext,
ExtractorError,
float_or_none,
int_or_none,
@@ -29,7 +28,7 @@ default_ns = 'http://www.w3.org/2005/SMIL21/Language'
_x = lambda p: xpath_with_ns(p, {'smil': default_ns})
-class ThePlatformBaseIE(InfoExtractor):
+class ThePlatformBaseIE(OnceIE):
def _extract_theplatform_smil(self, smil_url, video_id, note='Downloading SMIL data'):
meta = self._download_xml(smil_url, video_id, note=note)
error_element = find_xpath_attr(
@@ -38,17 +37,19 @@ class ThePlatformBaseIE(InfoExtractor):
if error_element is not None:
raise ExtractorError(error_element.attrib['abstract'], expected=True)
- formats = self._parse_smil_formats(
+ smil_formats = self._parse_smil_formats(
meta, smil_url, video_id, namespace=default_ns,
# the parameters are from syfy.com, other sites may use others,
# they also work for nbc.com
f4m_params={'g': 'UXWGVKRWHFSP', 'hdcore': '3.0.3'},
transform_rtmp_url=lambda streamer, src: (streamer, 'mp4:' + src))
- for _format in formats:
- ext = determine_ext(_format['url'])
- if ext == 'once':
- _format['ext'] = 'mp4'
+ formats = []
+ for _format in smil_formats:
+ if OnceIE.suitable(_format['url']):
+ formats.extend(self._extract_once_formats(_format['url']))
+ else:
+ formats.append(_format)
self._sort_formats(formats)
@@ -125,7 +126,7 @@ class ThePlatformIE(ThePlatformBaseIE):
'only_matching': True,
}, {
'url': 'http://player.theplatform.com/p/2E2eJC/nbcNewsOffsite?guid=tdy_or_siri_150701',
- 'md5': '734f3790fb5fc4903da391beeebc4836',
+ 'md5': 'fb96bb3d85118930a5b055783a3bd992',
'info_dict': {
'id': 'tdy_or_siri_150701',
'ext': 'mp4',
@@ -135,7 +136,6 @@ class ThePlatformIE(ThePlatformBaseIE):
'thumbnail': 're:^https?://.*\.jpg$',
'timestamp': 1435752600,
'upload_date': '20150701',
- 'categories': ['Today/Shows/Orange Room', 'Today/Sections/Money', 'Today/Topics/Tech', "Today/Topics/Editor's picks"],
},
}, {
# From http://www.nbc.com/the-blacklist/video/sir-crispin-crandall/2928790?onid=137781#vc137781=1
@@ -250,7 +250,7 @@ class ThePlatformFeedIE(ThePlatformBaseIE):
_TEST = {
# From http://player.theplatform.com/p/7wvmTC/MSNBCEmbeddedOffSite?guid=n_hardball_5biden_140207
'url': 'http://feed.theplatform.com/f/7wvmTC/msnbc_video-p-test?form=json&pretty=true&range=-40&byGuid=n_hardball_5biden_140207',
- 'md5': '22d2b84f058d3586efcd99e57d59d314',
+ 'md5': '6e32495b5073ab414471b615c5ded394',
'info_dict': {
'id': 'n_hardball_5biden_140207',
'ext': 'mp4',