def importopml(request):
theFile = request.FILES["opml"].read()
count = 0
dom = minidom.parseString(theFile)
imported = []
sources = dom.getElementsByTagName("outline")
for s in sources:
url = s.getAttribute("xmlUrl")
if url.strip() != "":
ns = Source.objects.filter(feed_url=url)
if ns.count() > 0:
# feed already exists - so there may already be a user subscription for it
ns = ns[0]
us = Subscription.objects.filter(source=ns).filter(
user=request.user)
if us.count() == 0:
us = Subscription(source=ns,
user=request.user,
name=ns.display_name)
if ns.max_index > 10: # don't flood people with all these old things
us.last_read = ns.max_index - 10
us.save()
count += 1
ns.num_subs = ns.subscriptions.count()
ns.save()
else:
# Feed does not already exist it must also be a new sub
ns = Source()
ns.due_poll = datetime.datetime.utcnow().replace(tzinfo=utc)
ns.site_url = s.getAttribute("htmlUrl")
ns.feed_url = url # probably best to see that there isn't a match here :)
ns.name = s.getAttribute("title")
ns.save()
us = Subscription(source=ns,
user=request.user,
name=ns.display_name)
us.save()
count += 1
imported.append(ns)
vals = {}
vals["imported"] = imported
vals["count"] = count
return render(request, 'importopml.html', vals)
def addfeed(request):
try:
feed = ""
if request.method == 'GET':
if request.GET.__contains__("feed"):
feed = request.GET["feed"]
groups = Subscription.objects.filter(
Q(user=request.user) & Q(source=None))
return render(request, "addfeed.html", {
"feed": feed,
"groups": groups
})
else:
if request.POST.__contains__("feed"):
feed = request.POST["feed"]
headers = {
"User-Agent":
"FeedThing/3.3 (+http://%s; Initial Feed Crawler)" %
request.META["HTTP_HOST"],
"Cache-Control":
"no-cache,max-age=0",
"Pragma":
"no-cache"
} # identify ourselves and also stop our requests getting picked up by google's cache
ret = requests.get(feed, headers=headers, verify=False, timeout=15)
# can I be bothered to check return codes here? I think not on balance
isFeed = False
content_type = "Not Set"
if "Content-Type" in ret.headers:
content_type = ret.headers["Content-Type"]
feed_title = feed
body = ret.text.strip()
if "xml" in content_type or body[0:1] == "<":
ff = feedparser.parse(
body) # are we a feed? # imported by django-feed-reader
isFeed = (len(ff.entries) > 0)
if isFeed:
feed_title = ff.feed.title
if "json" in content_type or body[0:1] == "{":
data = json.loads(body)
isFeed = "items" in data and len(data["items"]) > 0
if isFeed:
feed_title = data["title"]
if not isFeed:
soup = BeautifulSoup(body)
feedcount = 0
rethtml = ""
for l in soup.findAll(name='link'):
if l.has_attr("rel") and l.has_attr("type"):
if l['rel'][0] == "alternate" and (
l['type'] == 'application/atom+xml'
or l['type'] == 'application/rss+xml'
or l['type'] == 'application/json'):
feedcount += 1
try:
name = l['title']
except Exception as ex:
name = "Feed %d" % feedcount
rethtml += '<li><form method="post" onsubmit="return false;"> <input type="hidden" name="feed" id="feed-%d" value="%s"><a href="#" onclick="addFeed(%d)" class="btn btn-xs btn-default">Subscribe</a> - %s</form></li>' % (
feedcount, urljoin(feed,
l['href']), feedcount, name)
feed = urljoin(
feed, l['href']
) # store this in case there is only one feed and we wind up importing it
# TODO: need to accout for relative URLs here
# if feedcount == 1:
# just 1 feed found, let's import it now
# ret = fetch(f)
# isFeed = True
if feedcount == 0:
return HttpResponse("No feeds found")
else:
return HttpResponse(rethtml)
if isFeed:
parent = None
if request.POST["group"] != "0":
parent = get_object_or_404(Subscription,
id=int(request.POST["group"]))
if parent.user != request.user:
return HttpResponse(
"<div>Internal error.<!--bad group --></div>")
s = Source.objects.filter(feed_url=feed)
if s.count() > 0:
# feed already exists
s = s[0]
us = Subscription.objects.filter(
Q(user=request.user) & Q(source=s))
if us.count() > 0:
return HttpResponse(
"<div>Already subscribed to this feed </div>")
else:
us = Subscription(source=s,
user=request.user,
name=s.display_name,
parent=parent)
if s.max_index > 10: # don't flood people with all these old things
us.last_read = s.max_index - 10
us.save()
s.num_subs = s.subscriptions.count()
s.save()
return HttpResponse("<div>Imported feed %s</div>" %
us.name)
# need to start checking feed parser errors here
ns = Source()
# ns.due_poll = datetime.datetime.utcnow().replace(tzinfo=utc)
# print request.POST["river"]
# ns.inRiver = (request.POST["river"] == "yes")
ns.name = feed_title
ns.feed_url = feed
ns.save()
us = Subscription(source=ns,
user=request.user,
name=ns.display_name,
parent=parent)
us.save()
# you see really, I could parse out the items here and insert them rather than wait for them to come back round in the refresh cycle
return HttpResponse("<div>Imported feed %s</div>" % ns.name)
except Exception as xx:
return HttpResponse("<div>Error %s: %s</div>" %
(xx.__class__.__name__, str(xx)))
