def fetch(self,
url,
etag=None,
modified=None,
autodiscovery=True,
**kwargs):
""" use kwargs to pass extra data to parse_feed in Fetcher subclasses """
# handle local file first
if url.startswith('file://'):
url = url[len('file://'):]
stream = open(url)
return self.parse_feed(url, stream, {}, UPDATED_FEED, **kwargs)
# remote feed
headers = {}
if modified is not None:
headers['If-Modified-Since'] = modified
if etag is not None:
headers['If-None-Match'] = etag
stream = util.urlopen(url, headers)
responses = stream.history + [stream]
for i, resp in enumerate(responses):
if resp.is_permanent_redirect:
# there should always be a next response when a redirect is encountered
# If max redirects is reached, TooManyRedirects is raised
# TODO: since we've got the end contents anyway, modify model.py to accept contents on NEW_LOCATION
return Result(NEW_LOCATION, responses[i + 1].url)
res = self._check_statuscode(stream.status_code, stream.url)
if res == NOT_MODIFIED:
return Result(NOT_MODIFIED, stream.url)
if autodiscovery and stream.headers.get('content-type',
'').startswith('text/html'):
ad = FeedAutodiscovery(url)
# response_text() will assume utf-8 if no charset specified
ad.feed(util.response_text(stream))
if ad._resolved_url and ad._resolved_url != url:
try:
self.fetch(ad._resolved_url,
etag=None,
modified=None,
autodiscovery=False,
**kwargs)
return Result(NEW_LOCATION, ad._resolved_url)
except Exception as e:
logger.warn('Feed autodiscovery failed', exc_info=True)
# Second, try to resolve the URL
new_url = self._resolve_url(url)
if new_url and new_url != url:
return Result(NEW_LOCATION, new_url)
# xml documents specify the encoding inline so better pass encoded body.
# Especially since requests will use ISO-8859-1 for content-type 'text/xml'
# if the server doesn't specify a charset.
return self.parse_feed(url, BytesIO(stream.content), stream.headers,
UPDATED_FEED, **kwargs)
def get_channel_desc(url, feed_data=None):
if 'youtube.com' in url:
class YouTubeHTMLDesc(HTMLParser):
"""This custom html parser searches for the YouTube channel description."""
def __init__(self):
super().__init__()
self.description = ''
def handle_starttag(self, tag, attributes):
attribute_dict = {
attribute[0]: attribute[1]
for attribute in attributes
}
# Get YouTube channel description.
if tag == 'meta' \
and 'name' in attribute_dict \
and attribute_dict['name'] == "description":
self.description = attribute_dict['content']
try:
channel_url = get_channel_id_url(url, feed_data)
r = util.urlopen(channel_url)
if not r.ok:
raise YouTubeError('Youtube "%s": %d %s' %
(url, r.status_code, r.reason))
html_data = util.response_text(r)
parser = YouTubeHTMLDesc()
parser.feed(html_data)
if parser.description:
logger.debug('YouTube description for %s is: %s', url,
parser.description)
return parser.description
else:
logger.debug('YouTube description for %s is not provided.',
url)
return _('No description available')
except Exception:
logger.warning(
'Could not retrieve YouTube channel description for %s.' % url,
exc_info=True)
def get_cover(url, feed_data=None):
if 'youtube.com' in url:
class YouTubeHTMLCoverParser(HTMLParser):
"""This custom html parser searches for the youtube channel thumbnail/avatar"""
def __init__(self):
super().__init__()
self.url = []
def handle_starttag(self, tag, attributes):
attribute_dict = {
attribute[0]: attribute[1]
for attribute in attributes
}
# Look for 900x900px image first.
if tag == 'link' \
and 'rel' in attribute_dict \
and attribute_dict['rel'] == 'image_src':
self.url.append(attribute_dict['href'])
# Fallback to image that may only be 100x100px.
elif tag == 'img' \
and 'class' in attribute_dict \
and attribute_dict['class'] == "channel-header-profile-image":
self.url.append(attribute_dict['src'])
try:
channel_url = get_channel_id_url(url, feed_data)
r = util.urlopen(channel_url)
if not r.ok:
raise YouTubeError('Youtube "%s": %d %s' %
(url, r.status_code, r.reason))
html_data = util.response_text(r)
parser = YouTubeHTMLCoverParser()
parser.feed(html_data)
if parser.url:
logger.debug('Youtube cover art for {} is: {}'.format(
url, parser.url))
return parser.url[0]
except Exception:
logger.warning('Could not retrieve cover art', exc_info=True)
