Atomibulator 1.1

Atomibulator is a script that monitors arbitrary Web sites for changes and reports these updates through Atom and RSS feeds.
RSS/Atom feeds are great, but theyre far from pervasive. How many pages do you check regularly, just to see whether anythings new? For example, how many researchers do you know who provide feeds for their list of publications? Using the Atomibulator, you can monitor any webpage out there for changes and be notified by a custom Atom (or RSS, for the old schoolers out there) feed when the page has been updated.
Better yet, you dont have to install anything. You just edit the Atomibulator watchlist at Wikipedia. Anyone can do this. Heres how it works:
Add the page you would like to have monitored in standard Mediawiki shorthand for external links, i.e. [ ]. Multiple URLs can be tagged with the same description.
Sometimes you will want to ignore parts of the HTML markup of a page. For example, advertisements often result in changing links or images in every download of a page. To do this, the Atomibulator provides filter tags. You add them to the URL description in the Wiki markup. The following filters exist:
nolinks: strips all links from the document.
noimgs: strips all images from the document.
nohtml: strips all HTML tags from the document.
For example, to remove all images from the document located at http://www.foobar.com, tagged with "foobar", you would use [http://www.foobar.com foobar noimgs]. It doesnt matter where in the URL description you put the filter. You can use multiple filters in a single description.
The Atomibulator periodically grabs the listed URLs, applies any tag filtering configured, and runs a cryptographic hash function over the result. If an update is detected, it updates a feed named after the description found on in the watchlist: the feeds address is
http://www.cl.cam.ac.uk/~cpk25/atomibulator/feeds/ .xml.
The description you provide gets converted into lower case, any filter tags are removed, and whitespace becomes "_": "My Page" would be turned into "my_page", "My page nohtml" would still result in "my_page". You can see all currently available tags here. The feed for the page youre looking at, for example, is at
http://www.cl.cam.ac.uk/~cpk25/atomibulator/feeds/atomibulator.xml.
Since a lot of feed readers still struggle with supporting Atom 1.o, the Atomibulator actually goes beyond what its name implies and creates RSS 2.o as well:
http://www.cl.cam.ac.uk/~cpk25/atomibulator/feeds/atomibulator.rss.
NOTE: the Atomibulator is not a feed replacement. Do not use it on pages that provide feeds serving the same purpose anyway. Also please note that the Atomibulator checks the watchlist once per hour. Kindly dont request the feeds more often than that. It wont make any difference.
ALSO NOTE: Some people have said this service is like FeedTier. Its not. According to their page, their code "performs content analysis, picks-up the most prominent cluster of hyperlinks and automatically generates RSS web feeds." In short, it focuses on hyperlinks and tries to extract some RSS-able structure. The Atomibulator notices any kind of page update, subject to the HTML tag filtering you configure for a URL.
Enhancements:
- HTML tag filters and download timeouts were added.