[flickr] Don't use regex for extracting the info from the xml files

This commit is contained in:
Jaime Marquínez Ferrándiz 2015-04-22 19:58:39 +02:00
parent f8e51f60b3
commit c04c3e334c

View file

@ -6,7 +6,7 @@ from .common import InfoExtractor
from ..compat import compat_urllib_request from ..compat import compat_urllib_request
from ..utils import ( from ..utils import (
ExtractorError, ExtractorError,
unescapeHTML, find_xpath_attr,
) )
@ -40,20 +40,21 @@ class FlickrIE(InfoExtractor):
secret = self._search_regex(r'secret"\s*:\s*"(\w+)"', webpage, 'secret') secret = self._search_regex(r'secret"\s*:\s*"(\w+)"', webpage, 'secret')
first_url = 'https://secure.flickr.com/apps/video/video_mtl_xml.gne?v=x&photo_id=' + video_id + '&secret=' + secret + '&bitrate=700&target=_self' first_url = 'https://secure.flickr.com/apps/video/video_mtl_xml.gne?v=x&photo_id=' + video_id + '&secret=' + secret + '&bitrate=700&target=_self'
first_xml = self._download_webpage(first_url, video_id, 'Downloading first data webpage') first_xml = self._download_xml(first_url, video_id, 'Downloading first data webpage')
node_id = self._html_search_regex(r'<Item id="id">(\d+-\d+)</Item>', node_id = find_xpath_attr(
first_xml, 'node_id') first_xml, './/{http://video.yahoo.com/YEP/1.0/}Item', 'id',
'id').text
second_url = 'https://secure.flickr.com/video_playlist.gne?node_id=' + node_id + '&tech=flash&mode=playlist&bitrate=700&secret=' + secret + '&rd=video.yahoo.com&noad=1' second_url = 'https://secure.flickr.com/video_playlist.gne?node_id=' + node_id + '&tech=flash&mode=playlist&bitrate=700&secret=' + secret + '&rd=video.yahoo.com&noad=1'
second_xml = self._download_webpage(second_url, video_id, 'Downloading second data webpage') second_xml = self._download_xml(second_url, video_id, 'Downloading second data webpage')
self.report_extraction(video_id) self.report_extraction(video_id)
mobj = re.search(r'<STREAM APP="(.+?)" FULLPATH="(.+?)"', second_xml) stream = second_xml.find('.//STREAM')
if mobj is None: if stream is None:
raise ExtractorError('Unable to extract video url') raise ExtractorError('Unable to extract video url')
video_url = mobj.group(1) + unescapeHTML(mobj.group(2)) video_url = stream.attrib['APP'] + stream.attrib['FULLPATH']
return { return {
'id': video_id, 'id': video_id,