Coding 2025-12-14

By Max Woerner Chase

Okay, I'm writing late again, so let's see if I can bang something out quickly.

My hope is to get something for this paradigm stuff, and I want to avoid thinking too hard about what features I could possibly have for it, so let's look at the immediate stuff. I think it's reasonable to call case and aspect markers a part of speech, so the combination here boils down to selecting two parts of speech. Like, a noun and a case marking. I'd been thinking of doing this with multiple lexicons, but now that I'm thinking about it, I'm a little unsure how this should look. There's a main lexicon that gets updated, and any number of auxiliary lexicons to add more stuff, and selectors, which right now are any part of speech. If I do add them right back into the output lexicon along with everything else, then I just need to make sure that any post-processing only operates on the word boundaries. Adding them back does make sense, because in some cases, the original root will be removed.

Okay, in talking this out, and thinking some about other things, I've put together a formidable set of command-line options. Let's see how to use them together.

I think I see how to wrangle this, the question is, do I feel like doing that right now? Not... really. Well, I'll see about filtering and validating.

...

Okay, the validating is kind of overly clever, but I can just print the entire lexicon if something goes wrong, so I'm safe to just do the stuff I'm doing.

...

All right, I got carried away and coded the whole thing, really jankily. It's probably embarrassingly broken, but instead of thinking about that or trying it out, I'm going to get ready for bed.

Good night.