Three Dollar Quill

Oh yeah, I'm bringing this back.

That's right. The project that I shelved because I kept on going to or past the cutting edge of what Python makes possible. As I recall, when last we left off, I'd removed some code that attempted to emulate the capabilities provided by PEP 570.

And, well, we're so much further along, years down the line. I'm particularly excited to see what PEP 614 will allow for Structured Data. Or, at least, part of it.

See, Structured Data more-or-less broke down into two parts:

• Some Haskell-inspired custom class stuff in the genre of "like attrs, but not attrs"
• A Haskell-inspired matcher system that I'm pretty sure is capable of things that can't be syntactically expressed with the (comparatively) new match statement

I recently took another look at the code, and I was seriously torn between "this is convoluted and absurd" and "but this is really cool actually". So, why not take a fresh start, with my years of experience, and, um, actually understanding the Generator type, wow.

Anyway, one other thing occurred to me to try. Every once in a while, I experiment with a somewhat out-there idea. Sometimes, the result is pain. Just pain. So much pain. But I want to be open to new things. So, I was poking around, and I stumbled across a blog post describing how the authors laid out some JavaScript projects. I figured I'd try adapting the ideas to a Python project, and see what happens if I try slotting the Structured Data matcher rewrite into that framework.

To start with, I'm going to do a bunch of planning, and I've done some of that today, and I'll do some more of it later. For now, bed.

Good night.