As well as working on what I’ve started calling REx I’ve been messing around with syndication, both in Atom and iCalendar forms.
Not quite satisfied with Dave Jacoby’s RSS feed for Roger Ebert’s movie reviews, I decided to roll my own that also includes the summaries from the website. It’s the first thing I’ve written from scratch in Python—that is, with the help of Steve Jenson’s atomfeed.py and the built-in SGML, URL and regex libraries—and while it’s not pretty, it suits my purposes. If I thought it would be anything less than extremely fragile (mainly due to the wonky markup on Ebert’s site than anything inherent in the code) I’d probably post it… still might so people can point and laugh.
Also took some time on the weekend to write an iCalendar translator for Allison Brown’s performance calendar. It’s a simple little bit of Perl but took a lot longer to write than it probably should have; the RFC isn’t the easiest thing to read and there doesn’t appear to be a validator anywhere. The converted file appears to work in Mozilla Calendar, though, and that’s good enough for me.
Finally, I’m considering converting the Sirens’ tour calendar to hCalendar format. I’d probably still generate the RSS and iCal feeds, since the XSLT is there already, but I’d love to deprecate them in favour of a single human- and machine-readable page.
Oh yeah, and there was the third aborted attempt to set up an LDAP server so I could share my address books between the four primary machines and three OSes I use. Nothing about LDAP seems to have gotten any easier in the last seven years—when it’s impossible to enter or modify entries from an otherwise-capable address book, there’s something seriously wrong somewhere. Guess I’ll try again in another two years.
Good to hear that you are considering hCalendar. If you’re working on reviews, look at hReview as well.
I’m glad atomfeed worked out for you. Did you have any problems with it? I plan to release an update to track the atomfeed RFC as it comes out of the IETF.
Anyway, please feel free to shoot me an email, I will probably forget to check back here.
-steve
Despite not having docs, it was actually pretty straightforward and worked exactly as I expected. Most of the trouble I had was Python-related, such as requiring
.toprettyxml()after.toxml()to get a printable document out of theatomfeedobject.It might be nice to be able to set the
modeandtypeattributes on various elements—for example, mytitleis plain text, and I’m providing valid XHTML in mycontentthat’s being escaped and reported astext/html—but that’s a quibble.