[imgur] improve error check for non-video URLs

This commit is contained in:
John Boehr 2015-02-18 19:47:54 -08:00
parent d236b37ac9
commit 9e2d7dca87

View file

@ -5,6 +5,7 @@
from .common import InfoExtractor from .common import InfoExtractor
from ..utils import ( from ..utils import (
int_or_none, int_or_none,
str_or_none,
js_to_json, js_to_json,
mimetype2ext, mimetype2ext,
ExtractorError, ExtractorError,
@ -35,11 +36,6 @@ def _real_extract(self, url):
video_id = self._match_id(url) video_id = self._match_id(url)
webpage = self._download_webpage(url, video_id) webpage = self._download_webpage(url, video_id)
sources = re.findall(r'<source src="([^"]+)" type="([^"]+)"', webpage)
if not sources:
raise ExtractorError(
'No sources found for video %s' % video_id, expected=True)
width = int_or_none(self._search_regex( width = int_or_none(self._search_regex(
r'<param name="width" value="([0-9]+)"', r'<param name="width" value="([0-9]+)"',
webpage, 'width', fatal=False)) webpage, 'width', fatal=False))
@ -47,10 +43,13 @@ def _real_extract(self, url):
r'<param name="height" value="([0-9]+)"', r'<param name="height" value="([0-9]+)"',
webpage, 'height', fatal=False)) webpage, 'height', fatal=False))
formats = [] video_elements = str_or_none(self._search_regex(
video_elements = self._search_regex(
r'(?s)<div class="video-elements">(.*?)</div>', r'(?s)<div class="video-elements">(.*?)</div>',
webpage, 'video elements') webpage, 'video elements', fatal=False))
if not video_elements:
raise ExtractorError(
'No sources found for video %s' % video_id, expected=True)
formats = [] formats = []
for m in re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements): for m in re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements):
formats.append({ formats.append({