[MOBY-dev] screen-scraping MOBY

[Mark Wilkinson]

Hi MOBYers!

In an effort to make your lives easier (and especially those at PlaNet
who are currently displaying the raw XML of the MOBY Objects in their
portal :-P ) I have just made it easier to ask the MOBY CGI client
program at mobycentral to do what you want it to do, and then
screen-scrape the result.

The client program accepts the following CGI GET parameters:

namespace - a namespace
id - an id
servicename - the name of the service you wish to execute
authority - the authority providing that service
object - this is a URL-encoded MOBY object (object, no message struct)

What comes out of it is HTML, but I have made the HTML scrape-easy! 
Each object is surrounded by a pair of HTML comments:

<!-- SCRAPE_ME_START -->
<!-- SCRAPE_ME_END -->

between those comment tags is a <tr>....</td> containing the rendering
of a single MOBY response object into HTML.  Thus you can simply do a
wget or an LWP GET from inside of your own web pages, do a regexp on the
result, and stick it into a <TABLE> on your own web page in whatever
format you like.

Ta Daaaaaa!

Who loves you more than I do ;-)

These small changes have not yet been migrated into the CVS copy of the
browser (@Gbrowse) because they are aiming for a release soon and I
don't want to bugger that up.  If you want a copy of this code, let me
know and I'll mail it to you.

Mark

-- 
Mark Wilkinson (mwilkinson at mrl.ubc.ca)
University of British Columbia iCAPTURE Centre