def scrape_release(source, feed, entry, link):
title_text = entry.get('title')
if not isinstance(title_text, unicode):
title_text = title_text.encode('utf-8', 'ignore')
title = kill_control_characters(title_text)
date_text = (entry.get('published') or
entry.get('updated') or
entry.get('a10:updated'))
date = dateutil.parser.parse(date_text) if date_text else now()
body = get_link_content(link)
if body is None:
return
try:
# Does not use get_or_create because the unique constraint is just the url
# and we don't want the source foreign key field to ever be null.
release = Release.objects.get(url=link)
release.title = title
release.date = date
release.body = body
release.source = source
release.save()
except Release.DoesNotExist:
release = Release.objects.create(url=link,
source=source,
title=title,
date=date,
body=body)
def body(self):
if self._body is None:
response = requests.get(self.url)
response.raise_for_status()
(_junk_title, body) = readability_extract(response.content)
self._body = kill_control_characters(body)
return self._body
def get_link_content(link):
try:
response = requests.get(link)
if response.status_code == 400:
logging.warn(u"404 {}".format(link))
return None
if response.status_code != 200:
raise Exception(u"Unable to fetch release content: {0}".format(link))
except requests.exceptions.InvalidURL as e:
logging.warn(u"Invalid link {0}: {1}".format(link, unicode(e)))
return None
content_type = response.headers.get('content-type')
if not content_type:
logging.warn(u"Response did not contain a Content-Type header: {0}".format(link))
return None
(mime_type, mime_subtype, mt_params) = parse_mime_type(content_type)
if mime_type != 'text' or mime_subtype not in ('html', 'xhtml'):
logging.warn(u"Skipping non-HTML link: {0}".format(link))
return None
if len(response.content) == 0:
logging.warn(u"Server returned an empty body: {0}".format(link))
return None
(title, body) = readability_extract(response.content)
return kill_control_characters(body)
def handle(self, *args, **options):
if not hasattr(settings, "SUPERFASTMATCH"):
raise CommandError("You must configure SUPERFASTMATCH in your project settings.")
self.sfm = from_django_conf()
for url in args:
try:
if url.startswith("http://") or url.startswith("https://"):
release = Release.objects.get(url=url)
body = get_link_content(release.url)
release.title = kill_control_characters(release.title)
release.body = body
release.updated = now()
release.save()
logging.info("Updated release {0}: {1}".format(release.id, release.url))
else:
logging.warning("Skipping non-HTTP link {0}".format(release.url))
except Exception as e:
logging.error("Failed to rescrape {0}: {1}".format(url, str(e)))
