[nationalgeographic] Add extractor (closes #4960)

This commit is contained in:
Jaime Marquínez Ferrándiz 2015-02-19 18:17:31 +01:00
parent 8fc642eb5b
commit 6140baf4e1
3 changed files with 42 additions and 1 deletions

View File

@ -285,6 +285,7 @@
from .myspass import MySpassIE from .myspass import MySpassIE
from .myvideo import MyVideoIE from .myvideo import MyVideoIE
from .myvidster import MyVidsterIE from .myvidster import MyVidsterIE
from .nationalgeographic import NationalGeographicIE
from .naver import NaverIE from .naver import NaverIE
from .nba import NBAIE from .nba import NBAIE
from .nbc import ( from .nbc import (

View File

@ -0,0 +1,38 @@
from __future__ import unicode_literals
from .common import InfoExtractor
from ..utils import (
smuggle_url,
url_basename,
)
class NationalGeographicIE(InfoExtractor):
_VALID_URL = r'http://video\.nationalgeographic\.com/video/.*?'
_TEST = {
'url': 'http://video.nationalgeographic.com/video/news/150210-news-crab-mating-vin?source=featuredvideo',
'info_dict': {
'id': '4DmDACA6Qtk_',
'ext': 'flv',
'title': 'Mating Crabs Busted by Sharks',
'description': 'md5:16f25aeffdeba55aaa8ec37e093ad8b3',
},
'add_ie': ['ThePlatform'],
}
def _real_extract(self, url):
name = url_basename(url)
webpage = self._download_webpage(url, name)
feed_url = self._search_regex(r'data-feed-url="([^"]+)"', webpage, 'feed url')
guid = self._search_regex(r'data-video-guid="([^"]+)"', webpage, 'guid')
feed = self._download_xml('%s?byGuid=%s' % (feed_url, guid), name)
content = feed.find('.//{http://search.yahoo.com/mrss/}content')
theplatform_id = url_basename(content.attrib.get('url'))
return self.url_result(smuggle_url(
'http://link.theplatform.com/s/ngs/%s?format=SMIL&formats=MPEG4&manifest=f4m' % theplatform_id,
# For some reason, the normal links don't work and we must force the use of f4m
{'force_smil_url': True}))

View File

@ -71,7 +71,9 @@ def _real_extract(self, url):
if not provider_id: if not provider_id:
provider_id = 'dJ5BDC' provider_id = 'dJ5BDC'
if mobj.group('config'): if smuggled_data.get('force_smil_url', False):
smil_url = url
elif mobj.group('config'):
config_url = url + '&form=json' config_url = url + '&form=json'
config_url = config_url.replace('swf/', 'config/') config_url = config_url.replace('swf/', 'config/')
config_url = config_url.replace('onsite/', 'onsite/config/') config_url = config_url.replace('onsite/', 'onsite/config/')