sam_handelman's Journal: Cheap and sleazy indexing in apache?

Okay, here's the deal, right now we're taking some simple steps to put some of my research product on an apache web-page. Mostly, this is no big deal.

HOWEVER, the one thing that is something of a stumbling block is the keyword search. The way we handle it now, we start a shell, grep keywords out of the set of all bacterial genes (there are hundreds of thousands) and follow those back to the classes to which those bacterial genes are assigned. This is, heh, inefficient.

Obviously, it would be cleaner and more efficient if the entire endeavor were handled inside of a relational database - but that is not going to happen.

So, I'd like, you know, a perl script that generates an index from a bunch of flat files, and a library routine (to be incorporated into our cgi scripts) that looks up terms and returns filehandles. Anyone know where such a thing might be just lying around?

glympse

[js7a]

I recommend Glimpse [webglimpse.net] for this kind of thing. It's still free for noncommercial use. I can't find an iverted indices package in Perl. There is enough algorithmic complexity that people want to use C for the inner and next-to-inner loops.