lmaurits / feedformatter Goto Github PK

hi,
i've started looking at rss formatters and yours stood out amongst the
crowd.  i like that feedformatter is designed to be compatible with
feedparser, but i think the compatibility could be made a bit tighter.

for example, a user should be able to instantiate a Feed class directly
from a parsed feedparser structure. ideally, it might look like this:

feed = feedformatter.Feed( **feedparser.parse(ERROR_FEED) )

it looks like to do this, you would need to implement several keywords on
Feed.__init__:

bozo, encoding, entries (alias for items), feed, namespaces, and version

unfortunately, i have very limited experience with feedparser, so i don't
know exactly what to do with this info on the feedformatter side or i would
do it myself.  however, it does seem that the version keyword would allow
the addition of 'format_file' and 'format_string' functions which would use
the passed version to redirect to the appropriate formatting functions.

lastly, since this module is intended to be used with feedparser, your
example code should show how to do this.

thanks for the module.

-chad

What steps will reproduce the problem?
1. set locale to de-DE
2. generate rss2 feed with 1 item with any pubDate
3. try to display the feed with e.g. firefox

What is the expected output? What do you see instead?
expected: RFC822 date
actual: date string with localised month and weekday names

Proposed patch:

    from datetime import datetime
    from email.util import formatdate

in _format_datetime:

    # Then, convert that to the appropriate string
    if feed_type is "rss2":
        return formatdate(mktime(time))
    elif feed_type is "atom":
        return datetime.utcfromtimestamp(mktime(time)).isoformat()

Adding it would be nice.

I'm trying to read a RSS feed with feedparser and update it with
feedformatter to add a few entries in it.

When creating the output file, feedformatter fails with the following message:
  File "/usr/lib/python2.6/site-packages/feedformatter.py", line 163, in
_convert_datetime
    raise Exception("Unrecongised time format!")

The string that it can't convert is an Unicode string previously created
with feedformatter. It's something like "Tue, 17 Nov 2009 15:21:46 CET",
which is *not* alphanumeric, so "time.isalnum()" fails. (And since
"strptime()" is not imported from the "time" module, it wouldn't work anyway.)

Here is a patch that fixes this bad behaviour. It also uses
types.StringTypes for a (possibly) better Unicode support.

What steps will reproduce the problem?
1. I'd like to add an image to my description.
2. My description like looks like this: item["description"] = "<img src='" + 
img + "'> " + ' '.join(article.text.split("\n")[1:])


What is the expected output? What do you see instead?
I expect:
<description><img src='http://foo.com/nxwquMosquito.png> Safety tips to..

I get: 
<description>&lt;img src='http://foo.com/nxwquMosquito.png'&gt; Safety tips to..


Linux, Python 2.7

What steps will reproduce the problem?
1. Have Mon, 01 Oct 2012 11:24:00 +0000 in your atom feed as published
2. try smth like feedformatter.Feed(feed=feedparser.parse().feed).atom_string
3. see it fail

Exception: Unrecongised time format! local Mon, 01 Oct 2012 11:24:00 +0000

Note that I had to pimp the exception in _convert_datetime. Please consider 
giving more helpful messages.

The strptime format currently is:  "%a, %d %b %Y %H:%M:%S %Z"

time.strptime yields:
ValueError: time data '%a, %d %b %Y %H:%M:%S %Z' does not match format u'Mon, 
24 Sep 2012 12:20:27 +0000'

No idea why it doesn't parse. But it's annoying.

proposed patch:

in _convert_datetime:
(...)
    elif hasattr(time, 'timetuple'):
        # datetime.datetime object
        return time.timetuple()

if the link is a dict, the dict type is returned. It should return the link.

--- feedformatter.py~   2009-10-09 00:38:00.000000000 +0200
+++ feedformatter.py    2009-10-09 01:17:41.000000000 +0200
@@ -182,7 +182,7 @@
 def _atomise_link(link):

     if type(link) is dict:
-        return dict
+        return link
     else:
         return {"href" : link}