Mikaela/Limnoria
1
0
Fork 0
mirror of https://github.com/Mikaela/Limnoria.git synced 2024-10-01 17:16:31 +02:00
Code Issues Projects Releases Wiki Activity

RSS: If ids are missing, use <link> as id.

Browse Source
This commit is contained in:
Valentin Lorentz 2020-01-31 20:59:42 +01:00
parent 653afdd086
commit 2f49362510
1 changed files with 13 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
plugins/RSS/plugin.py
View File

@ -88,6 +88,17 @@ def only_one_at_once(f):
lock[0] = False lock[0] = False
return newf return newf
def get_entry_id(entry):
# in order, try elements to use as unique identifier.
# http://validator.w3.org/feed/docs/rss2.html#hrelementsOfLtitemgt
id_elements = ('id', 'link', 'title', 'description')
for id_element in id_elements:
try:
return getattr(entry, id_element)
except AttributeError:
pass
raise ValueError('Feed entry is missing both title and description')
class InvalidFeedUrl(ValueError): class InvalidFeedUrl(ValueError):
pass pass
@ -340,17 +351,13 @@ class RSS(callbacks.Plugin):
self.update_feed_if_needed(feed) self.update_feed_if_needed(feed)
def get_new_entries(self, feed): def get_new_entries(self, feed):
# http://validator.w3.org/feed/docs/rss2.html#hrelementsOfLtitemgt
get_id = lambda entry: entry.id if hasattr(entry, 'id') else (
entry.title if hasattr(entry, 'title') else entry.description)
with feed.lock: with feed.lock:
entries = feed.entries entries = feed.entries
new_entries = [entry for entry in entries new_entries = [entry for entry in entries
if get_id(entry) not in feed.announced_entries] if get_entry_id(entry) not in feed.announced_entries]
if not new_entries: if not new_entries:
return [] return []
feed.announced_entries |= set(get_id(entry) for entry in new_entries) feed.announced_entries |= set(get_entry_id(entry) for entry in new_entries)
# We keep a little more because we don't want to re-announce # We keep a little more because we don't want to re-announce
# oldest entries if one of the newest gets removed. # oldest entries if one of the newest gets removed.
feed.announced_entries.truncate(10*len(entries)) feed.announced_entries.truncate(10*len(entries))