Bhuvaneswaran Arumugam wrote:
>>> This script is almost inherited from svn2rss.py. So, all the
>>> corrections/changes we do here must be done in that script as well.
>> Ah, I *see*. In that case we must merge the code into a unified feed
>> producing script. To do otherwise is a severe maintenance headache.
>>
>> So, that's what I have done. Please review the latest version of
>> svn2rss.py, in which I think I have incorporated all your changes,
>> though in substantially different form.
>
> Yes, i see you have incorporated all the changes and *more* than that.
> It looks to be fully complete! Thank you very much for spending your
> time on this!
Not _quite_ complete. There is still the issue of removing old items
from atom feeds.
>>>>> + doc.createTextNode("%s" % self.format_date(datetime.datetime.now())))
>>>> Now? Shouldn't the time be taken from the revision's svn:date?
>>> I guess it should be the entry published/updated date instead of
>>> revision date.
>> Is it required to be so for feed readers to work properly? Otherwise, I
>> think it might make more sense to use the revision date.
>
> Not really. Anyhow, the revision date is already part of the feed
> content. So, to avoid redundant data we can retain as it is now.
Many feed readers have a separate display for "Date" apart from the feed
content. I think we should put the revision date into the actual XML
metadata, and stop putting the date into the feed summary at all.
>>>> desc_lines.append('')
>>>> item_desc = "\n".join(desc_lines)
>>> Yes, i have incorporated this change.
>> Actually, I meant, use this to replace the use of individual variables
>> entirely - have a look at my version to see what I meant.
>
> I see you no more use append() function.
Oops. First, refactored the code to use append() to replace the use of
lots of local variables.
Then, later, I came back and refactored my code again to use string
interpolation directly from the pipe results. Sorry for the confusion.
One additional thing I noticed, but forgot to mention in a review until now:
The RSS generator puts new items at the *start* of the XML document.
The new Atom generator however, puts new items at the *end* of the XML
document.
Does this matter?
Max.
Received on Fri Jul 21 11:36:23 2006