X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/b238854ce845f3796daac74edab2e8a373e8ba1a..22bdf6fb907cce994fbfe062879d2a991e597a3f:/youtube_dl/extractor/clipsyndicate.py diff --git a/youtube_dl/extractor/clipsyndicate.py b/youtube_dl/extractor/clipsyndicate.py index d4fc869..02a1667 100644 --- a/youtube_dl/extractor/clipsyndicate.py +++ b/youtube_dl/extractor/clipsyndicate.py @@ -1,9 +1,11 @@ +from __future__ import unicode_literals + import re -import xml.etree.ElementTree from .common import InfoExtractor from ..utils import ( find_xpath_attr, + fix_xml_ampersands ) @@ -11,13 +13,14 @@ class ClipsyndicateIE(InfoExtractor): _VALID_URL = r'http://www\.clipsyndicate\.com/video/play(list/\d+)?/(?P\d+)' _TEST = { - u'url': u'http://www.clipsyndicate.com/video/play/4629301/brick_briscoe', - u'md5': u'4d7d549451bad625e0ff3d7bd56d776c', - u'info_dict': { - u'id': u'4629301', - u'ext': u'mp4', - u'title': u'Brick Briscoe', - u'duration': 612, + 'url': 'http://www.clipsyndicate.com/video/play/4629301/brick_briscoe', + 'md5': '4d7d549451bad625e0ff3d7bd56d776c', + 'info_dict': { + 'id': '4629301', + 'ext': 'mp4', + 'title': 'Brick Briscoe', + 'duration': 612, + 'thumbnail': 're:^https?://.+\.jpg', }, } @@ -26,16 +29,14 @@ class ClipsyndicateIE(InfoExtractor): video_id = mobj.group('id') js_player = self._download_webpage( 'http://eplayer.clipsyndicate.com/embed/player.js?va_id=%s' % video_id, - video_id, u'Downlaoding player') + video_id, 'Downlaoding player') # it includes a required token - flvars = self._search_regex(r'flvars: "(.*?)"', js_player, u'flvars') + flvars = self._search_regex(r'flvars: "(.*?)"', js_player, 'flvars') - playlist_page = self._download_webpage( + pdoc = self._download_xml( 'http://eplayer.clipsyndicate.com/osmf/playlist?%s' % flvars, - video_id, u'Downloading video info') - # Fix broken xml - playlist_page = re.sub('&', '&', playlist_page) - pdoc = xml.etree.ElementTree.fromstring(playlist_page.encode('utf-8')) + video_id, 'Downloading video info', + transform_source=fix_xml_ampersands) track_doc = pdoc.find('trackList/track') def find_param(name):