parser-demo VS re2c

I have a demo project that fixes a lot of the default insanities.

I have an old demo project that demonstrates most of that for flex/bison, but this is C/C++ oriented so might not be super applicable for you.

It's just unfortunate that flex and bison have bad defaults due to historical compatibility with lex and yacc. If you're using them I have a demo project that tweaks the defaults toward sanity, though I haven't updated it for recent warnings.

I made a demo project that ties them together. I'm pretty sure at least one of the fixes was due to header cycle problems, but I haven't touched it for a while. (note also that the demo was fully warning-clean at the time, but there may be new warnings since)

I made a demo project that avoids several historical annoyances with Flex and Bison - in particular, I made it warnings-clean (though I haven't updated it recently), and I avoid global state.

I have a demo project that enables all of the non-default options that you really should be using in all new projects. Note that the demo doesn't demonstrate meaningful grammars, just shows how to arrange the surrounding code and makefile.

I have a demo project that twiddles most of the important knobs.

Bash is a really complicated language. Redirections and such are simple; they are just tokenization, and any lexer/parser tutorial should get you there (I have a demo project using flex/bison).

Bison can be configured to avoid every single one of those problems. I use reentrant versions (both with and without push parsing) in my demo project. I admit I didn't bother with named references or cleaning up after errors.

a lot of the things that people hate about flex/bison is actually just their defaults for compatibility with lex/yacc, which can be changed. I have a demo project that does some of the things: https://github.com/o11c/parser-demo

(1) Zulip Chat - https://zulip.com/ - seems to be reasonably popular, but more people should know about it

I’ve been using it for over 5 years now [1], and it’s as good as ever. It’s way faster than any other chat app I’ve used. It has a good UI and conversation model. It has a simple and functional API that lets me curl threads and write blog posts based on them.

(only problem is that I Ctrl-+ in my browser to make the font bigger – I think it’s too dense for most people)

(2) re2c regex to state machine compiler - https://re2c.org

A gem from the 90’s, which people have done a great job maintaining and improving (getting Go and Rust target support in the last few years). I started using it in 2016, and used for a new program a few months ago. I came to the conclusion that it should have been built into C, because C has shitty string processing – and Ken Thompson both invented C AND brought regular languages to computing !!

In comparison, treesitter lexers are very low level, fiddly, and error prone. I recently saw dozens of ad hoc fixes to the tree-sitter-bash lexer, which is unsurprising if you look at the structure of the code (manually crawling through backslashes and braces in C).

https://github.com/tree-sitter/tree-sitter-bash/blob/master/...

These fixes are definitely appreciated, but I think it indicates a problem with the model itself.

(based on https://lobste.rs/s/endspx/software_you_are_thankful_for#c_y...)

[1] https://www.oilshell.org/blog/2018/04/26.html

The "Papers" section on re2c's web site continues Laurikari's work: http://re2c.org/

... but I haven't found them particularly accessible. And it's not clear it's a viable strategy in a general purpose regex engine. Namely, I'm not sure how much bigger it makes the DFA.

Also, AFAIK, these are DFAs. They are different theoretical structures with explicitly more power.

> and then an NDFA is used to match a third time, to extract the capture groups.

That's the PikeVM. It's an NFA simulation. Although it uses additional storage and is otherwise more computationally powerful than just a plain NFA.

> What do you gain by using it?

Performance, although this possibly depends on your compiler, whether you use PGO, and similar finicky issues.

Example: https://eli.thegreenplace.net/2012/07/12/computed-goto-for-e...

Some prior HN discussion: https://news.ycombinator.com/item?id=18678920

Another example where goto is relevant is implementing finite automata. A (very short) paper from 1988 that discusses three different ways of implementing a finite state machine is "How (Not) to Code a Finite State Machine". The documentation of RE2C may be even more interesting: https://re2c.org

RE2C is a program that compiles finite automata into C, Go, or Rust code. It provides many implementation strategies: it can make use of computed or labelled gotos when the language provides them.

Implementing pushdown automata comes with similar issues.

wget https://github.com/skvadrik/re2c/releases/download/1.0.3/re2c-1.0.3.tar.gz tar -zxvf re2c-1.0.3.tar.gz cd re2c-1.0.3/ ./configure make make install

In Rust I use https://docs.rs/logos/latest/logos/. I think another similar is http://re2c.org

re2c is an open-source lexer generator. It uses regular expressions to recognize tokens.

Those suggested crates are still more or less the popular options. There was also recently added support for Rust in re2c.

First, the implementation of strtotime is a textbook study in why other people's C code is not where you want to spend time. You can see the guts of the implementation logic here. This isn't stock C code -- it's code for a system called re2c. This system allows you to write regular expressions in a custom DSL (domain specific language), and then transform/compile those regular expressions down to C programs (also C++ and Go) that will execute those regular expressions. Something in PHP's make file uses this parse_date.re file to generate parse_date.c. If you don't realize parse_date.c is a generated file, this can be extremely rough going. If you've not familiar with re2c is can be regular rough going. We leave further exploration as an exercise for the reader -- an exercise we haven't taken ourself.