More automated review indexing

Tonight, I finally finished some code that I'd been meaning to finish for several months. The script that I use to generate the indices for my reviews can now recognize a collection (either a short story collection or a magazine) and automatically add the index entries for my short fiction index. That also means that the last two magazines I read have now been added to the index.

The review-post script that does all of this is rather hairy and oddly designed. It's evolved naturally from a few quick automation scripts to automatically post reviews to my journal, and then to automate the changes to the thread files that I was maintaining by hand.

On one hand, it's an interesting case study in doing things that people associate with a database without any database. The output is just regular thread files that I could equally easily write by hand, and it has enough smarts to get this right and do a good job. I like doing that, just to prove a point.

On the other hand, putting all the review metadata in a database would be an obvious database application and would probably make it easier to do some other things (although I've not thought a lot about what I would change). I keep thinking I should do that, although then I have to figure out the best way of generating the pages. I could generate the existing thread pages out of the database data (although then, what am I really gaining over what I'm doing now?), or I could generate them dynamically, but that would require putting the right Apache modules on www.eyrie.org and probably doing something with PHP.

Each time I start thinking about this, I leave well enough alone.