def get_feeditem_model(url, timestamp, item):
"""
Returns a new feeditem model with cleaned attributes.
Returns none if item already exists.
Parameters:
url - The feed's url
item - A feedparser rss-entry
"""
for feeditem in FeedItemModel.query(FeedItemModel.key == FeedItemKey(url, item.link)).iter(keys_only=True):
# Existing feed, so ignore
return None
# Some feeds use content, some description.
# Take the longest.
d = item.get("description", "")
try:
c = item.get("content", None)
# Can be many content, with different types
# This might fail for some
c = c[0].value
except:
c = ""
# Use d
if len(c) > len(d):
d = c
clean_description = strip_bloat(d)
return FeedItemModel(key=FeedItemKey(url, item.link),
title=item.title,
description=clean_description,
link=item.link,
title_stripped=get_snippet(item.title),
snippet=get_snippet(clean_description),
timestamp=timestamp,
feed_link=url,
images=get_images(clean_description),
published=convert_timestamp(item.get("published", None)),
author=item.get("author", None),
comments=item.get("comments", None),
enclosures=[e.href for e in item.get("enclosures", [])],
tags=[t.term for t in item.get("tags", [])])
