This shows up in current HEAD and in 5.1.3.
The two trivial Atom 1.0 feeds attached are the same, except that the order of
the <content> and <summary> elements is swapped.
When <summary> comes first, I get the result I would expect: feedparser returns
<summary> in 'summary' and 'summary_detail', and <content> in 'content'.
When <content> comes first, however, feedparser puts <content> into 'summary',
puts both <content> and <summary> into 'content', and doesn't set
'summary_detail' at all...
Looking at the code, this is because _start_summary looks to see if the current
entry already has a 'summary' field set (which it does, because _end_content
set it), and if so, saves the summary into 'content'. I'm not sure why that's a
useful thing to do, but there are a couple of tests that make sure it does it
(wellformed/rss/item_summary_and_description.xml and
wellformed/rss/item_description_and_summary.xml) -- is there a historical
reason for wanting this behaviour with the nonstandard <summary> element in
RSS? (And if so, can it be disabled for Atom?)
Original issue reported on code.google.com by ats-goog...@offog.org on 31 Jul 2013 at 3:25
Original issue reported on code.google.com by
ats-goog...@offog.orgon 31 Jul 2013 at 3:25Attachments: