MarcoBuster/youtube-dl
1
1
Fork 0
mirror of https://github.com/ytdl-org/youtube-dl synced 2024-11-27 19:36:50 +01:00
Code Issues Releases Wiki Activity

[spiegel] Test format video URLs for 404 (Closes #4579)

Browse Source
This commit is contained in:
Sergey M․ 2015-01-14 20:27:14 +06:00
parent f2cbc96c3e
commit e92d4a11f5
1 changed files with 33 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

49
youtube_dl/extractor/spiegel.py
View File

@ -4,7 +4,14 @@ from __future__ import unicode_literals
import re
from .common import InfoExtractor
from ..compat import compat_urlparse
from ..compat import (
compat_urlparse,
compat_HTTPError,
)
from ..utils import (
HEADRequest,
ExtractorError,
)
from .spiegeltv import SpiegeltvIE
@ -60,21 +67,31 @@ class SpiegelIE(InfoExtractor):
xml_url = base_url + video_id + '.xml'
idoc = self._download_xml(xml_url, video_id)
formats = [
{
'format_id': n.tag.rpartition('type')[2],
'url': base_url + n.find('./filename').text,
'width': int(n.find('./width').text),
'height': int(n.find('./height').text),
'abr': int(n.find('./audiobitrate').text),
'vbr': int(n.find('./videobitrate').text),
'vcodec': n.find('./codec').text,
'acodec': 'MP4A',
}
for n in list(idoc)
# Blacklist type 6, it's extremely LQ and not available on the same server
if n.tag.startswith('type') and n.tag != 'type6'
]
formats = []
for n in list(idoc):
if n.tag.startswith('type') and n.tag != 'type6':
format_id = n.tag.rpartition('type')[2]
video_url = base_url + n.find('./filename').text
# Test video URLs beforehand as some of them are invalid
try:
self._request_webpage(
HEADRequest(video_url), video_id,
'Checking %s video URL' % format_id)
except ExtractorError as e:
if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404:
self.report_warning(
'%s video URL is invalid, skipping' % format_id, video_id)
continue
formats.append({
'format_id': format_id,
'url': video_url,
'width': int(n.find('./width').text),
'height': int(n.find('./height').text),
'abr': int(n.find('./audiobitrate').text),
'vbr': int(n.find('./videobitrate').text),
'vcodec': n.find('./codec').text,
'acodec': 'MP4A',
})
duration = float(idoc[0].findall('./duration')[0].text)
self._sort_formats(formats)