A direct port of pikchr.org's pikchr.y to Go, by hand.

Why would anyone hand-port almost 12,000 lines of C to Go? There's no good reason. I just like Go, and I like pikchr, and I wanted to use pikchr cleanly in Go, and on my blog, using Hugo.

Perhaps it's just having to constantly make trade-offs and value expediency at work that leaves us with an irrational, curmudgeony desire to dive all the way down the yak-shaving stack, and do things right, dammit!. Consider gopikchr a gift.

There are probably bugs.

Things work well enough to properly convert files in examples/, and tests/, but I'm sure there are still bugs lurking.

Everything is currently in internal/ and cmd/pikchr; it remains to add a clean Go interface in this root directory.

Pull requests are welcome: I currently intend no support for this project, but if you're obscure enough to want a Go port of pikchr, then we share some strange kind of kinship, and I welcome your contributions.

• Add pikchr support to my blog (which means adding it to Hugo (which means adding it to goldmark)).
• Keep the code structure in pikchr.y as close to the original as possible, to make tracking and applying future changes as straight-forward as possible. This means the code is often very C-like, and not very Goish.
• Convert to Go idioms only where the conversion remains clear.

This code was hand-ported from the C code, on top of a hand-port of the Lemon Parser. It is clearly C code, transliterated to Go. I converted pointers to bytes and pointer math on byte pointers to []byte in Go, but otherwise everything is left alone as much as possible.

• Add a clean Go interface in the root directory
• Create a github action that follows the rss feed for changes to pikchr.y and creates issues.
• (Maybe) Set up fuzzing

• @zellyn