goawk VS tectonic

TIL: GoAWK [1] - A POSIX-compliant AWK interpreter written in Go, with CSV support.

[1]: https://github.com/benhoyt/goawk

Another nice project that I have used from time to time (and a very good source for insight) is the awk interpreter written in go https://github.com/benhoyt/goawk

No, I want exactly the opposite - it should be a , b,c as a single string field containing a literal comma, and c. For example, https://github.com/benhoyt/goawk has csv support. https://github.com/benhoyt/goawk/blob/master/docs/csv.md - more info.

I've had a go at solving this in this PR -- feedback welcome. I don't love it, but oh well, it solves the problem at hand. Your comment pointed me in the right direction, thanks again.

There are quite a few re-implementations of scripting languages like Lua in Go. I've written an AWK interpreter in Go.

github.com/benhoyt/goawk -> this library lets you embed an AWK runtime in your applications, very easy to use and useful for enabling some powerful scripting in things you build

Yes, that's right. With my simplistic UTF-8-based implementation it turned length() -- for example -- from O(1) to O(N), turning O(N) algorithms which use length() into O(N^2). See this issue: https://github.com/benhoyt/goawk/issues/93

Similar with substr() and other string functions, which when operating as bytes are O(1), but become O(N) when trying to count the number of codepoints as UTF-8.

GNU Gawk has a fancier approach, which stores strings as UTF-8 as long as it can, but converts to UTF-32 if it needs to (eg: the string is non-ASCII and you call substr).

It looks like Brian Kernighan's code has the same issue with length() and substr(). I'm going to try to email him about this, as I think it's kind of a performance blocker.

I'm not sure if it was via my personal website or just my GitHub profile, but I got my current job at Canonical due to the CTO there reaching out about my GoAWK project (https://github.com/benhoyt/goawk). I get regular recruitment emails because I have my CV/resume online: most of them are very low-effort, but 1 in 20 or something are interesting emails where the recruiter has actually looked at my website and will tailor it personally. I also just enjoy technical writing, and get joy out of sharing it on HN. So it's "worth it" for me.

You may want to try https://github.com/tectonic-typesetting/tectonic, which downloads files from TeXLive on-demand.

v2 has improved TeX engine lookup, improved PDF template look&feel, proper support for MS Windows (where it comes integrated with the Tectonic engine) and a few more new features.

There's https://github.com/tectonic-typesetting/tectonic but I think the issue with that idea is that sure, you can re-implement TeX (it's sufficiently simple) in Rust and then run LaTeX packages on top of it, but then you're back to LaTeX and all its weirdness so you haven't really gained anything compared to LaTeX itself.

In terms of TeX, I would recommend taking a look at tectonic, a self-contained TeX distribution that auto-installs packages you need when you need them, and “just works” when you call the binary to compile… Because screw messing around with package managers, CTAN and XeTeX. I’ve been using it for around a year and it’s so much easier than any other TeX distribution.

For example, you can have your main application in something like Deno/Node/python that acts as a server, and then delegate the actual pdf generation to tectonic (https://github.com/tectonic-typesetting/tectonic) or Typst https://typst.app/blog/2023/beta-oss-launch/

tectonic: https://github.com/tectonic-typesetting/tectonic

How does it compare with Tectonic?

It's sad that Tectonic conversion to Rust[1] was never finished. For now it's just a wrapper around C and C++ code. By far, it was the most promising thing in this distribution.

[1] https://github.com/tectonic-typesetting/tectonic/issues/459

The only thing I've seen is https://github.com/tectonic-typesetting/tectonic but that's an actual re-implementation of TeX Rust.