Klib VS HTTP Parser

If you want a generic, extremely high performance implementation. I really recommend khash: https://github.com/attractivechaos/klib/blob/master/khashl.h

Most importantly, sqzlib is fully compatible with klib/kseq.h one of the highest performance fastA/Q parsers. This means that any application that uses kseq.h for fastA/Q parsing, can be easily modified to use sqzlib instead. You can find patched versions of seqstats, minimap2, and bwa-mem2 in my github, or you can patch them yourself with the included patches.

I don;t know if you are aware of khash? It's a very nifty hash table implementation. It is one of the fastest hash table implementations out there. Heavily used in bioinformatics. You said in your post that you are a macro guy, well, you are in for a treat with this library.

This is in fact a pretty common pattern in production code.

Here's an entire suit of type safe generic data structures in C: https://github.com/attractivechaos/klib

You will probably want to invest some time into either finding a good data structures library (preferable) or writing your own data structures (fine to do but not a good use of your time if you're starting out). You can try SGLIB or klib.

It looks not too different from the majority of HTTP parsers out there written in C. Here is an example of NodeJS [0].

[0] https://github.com/nodejs/http-parser/blob/main/http_parser....

NodeJS's HTTP parser used to be a handwritten C lib: http-parser

Whoa, I didn't know that! But yeah, it seems like https://github.com/nodejs/http-parser is based on nginx. It now uses https://github.com/nodejs/llhttp but has some of the same legacy.

On the other hand, deno's HTTP stuff is built on top of Hyper, a Rust library https://github.com/hyperium/hyper

Boost.Beast has its own HTTP parser[0], during the development of which Vinnie Falco (the principle author of Beast) found many bugs/inconsistencies in Node.js's own parser[1]

[0] https://www.boost.org/doc/libs/develop/libs/beast/doc/html/b...

[1] https://github.com/nodejs/http-parser/issues?q=is%3Aissue+au...

For cases where it is necessary to pass local information to/from a callback, the http_parser object's data field can be used.

From nodejs http-parser documentation:

Legacy HTTP/1.1 suffers a few issues, see the current RFC errata:

https://www.rfc-editor.org/errata_search.php?rfc=7230&rec_st...

There are issues particularly around how whitespace and obsolete line folding should be handled

https://github.com/nodejs/http-parser/issues?q=is%3Aissue+wh...

https://github.com/httpwg/http-core/issues/53

It's not as trivial as a few string splits.

There's picohttpparser and the one used in node.js: https://github.com/nodejs/http-parser