Posted 21 January 2006 - 10:26 PM
It depends how the data is being publically presented. If it is only being presented as a web-page (html), then you pretty much have to resort to "screen-scraping", which means creating some kind of script that will parse the raw source code for the web-page to retrieve the data that you want. First, that is sometimes really difficult to do. ANd once you have your script, one little alteration of how they present their information will break your script.
The other option is if the web-site from which you are getting results has some type of web-service with which you can interface using some sort of standard api. Perl tends to be popular for this (shudder), but sometimes they are written in C++ or Java. I rather doubt that any news service is going to have any public interfaces though.
I did a quick search, and I couldn't find any web-services that would fit the bill. I see that the BBC has text only versions.. those would be much easier to use for screen-scraping. They are horrendously slow though...