Well, I was thinking you can always hotlink the EngineCE API I wrote  
which gets data directly from the Oracle DB, with caching so the  
evening downtime doesn't break everything. It has scalable filtering  
mechanisms, and pulling an individual template (rather than a  whole  
page, for instance) is pretty easy. Unfortunately it only lists CE  
courses, and it wouldn't be possible to list all courses for a while  
since I haven't replaced the caching mechanism with a more scalable  
model yet. I had been working on it until around the last week of  
January when I started getting hit with a lot of programming that  
wasn't directly web-related.

I think your best option would be to scrape the registrar's pages, and  
if you're worried about the markup changing, you could always set up  
some graceful error-handling and some caching mechanisms so things  
don't break immediately. Remember, XPath is your friend for scraping  
HTML. Also, if you're not aware, UVM has memcache servers, and you can  
use the PHP Memcache library [ ] to use the  
memcache servers.

There's also a course list mechanism in the UVM templating system,  
though I wouldn't think of using that directly because you don't get  
much control over the content.

Meanwhile, it would be awesome if you suggested for a project someone  
make suggestions and re-style the CE course list templates, though I'd  
probably want to ask in my department first. It would be a good way to  
get some feedback on my code and documentation.

Good luck with pulling the course catalog!


Quoting Robert Michael Erickson <[log in to unmask]>:

> can departments grab text from the course catalog to include on  
> their departmental web site?
> for example:
> i would like to grab the catalog information from the above page  
> without parsing it more like an include("url").
> Be Cool,
>     Bob E.
> Robert M. Erickson
> Senior Lecturer, Computer Science Department, University of Vermont
> 802.656.8137