Follow-up to this.
This is basically a wrap-up post. The nature of the application was quite simple, but I had to learn everything from the ground up, which was extremely time-consuming.
After converting the data to sql (including developing a simple schema) I created a MySQL view which is the sole point of contact for the php code querying it.
As far as features go, the php app presents a single search box from which every useful field in the DB is queried.
All words are assumed to be connected by an OR operator, but +- and "" are supported.
Hits are presented in a thumbnail gallery which is weighted by relevancy - the more hits inside of a given record, the more relevant it is and the closer to the front page it goes. The gallery is paged, with each thumbnail providing a quick view of the complete record (with some info truncated to fit) on hover. On click, a larger version is displayed along with a link that takes the user to a full record view.
The current version lives here: I have a feature-complete version of the GRS media database ready for use here:
http://lettuce.tapor.uvic.ca/~gregster/grs/final/ . I do not know where it will ultimately live.
Impossible to estimate the amount of time I spent on this - it was 50/50 learning and coding. "Minutes Worked" is just a guess.