Getting Stack Overflow error when using chumsky

[FussballAndy]

So I tried implementing chumsky as a parser into a lang I'm working on, however when executing the test I wrote, I get a stack overflow error.

Posting the entire code here would be a huge thing so, I put it into a repo. It can be found in this branch: https://github.com/FussballAndy/unreal-lang/tree/chumsky

The files with chumsky content are the following:

• ulc_parser/chumsky_parser.rs
• ulc_parser/chumsky_parser/
• ulc_parser/chumsky (overflowing test)

And files with AST are located in ulc_ast

I mainly used the nano rust example for reference, if this information is helpful to anybody.

In order to only build and run the necessary test you can run the following cargo test -p ulc_parser --test chumsky.

I hope for anyone trying to help me, it is not too bad that the code is embedded in an entire project instead of some code snippets. Additionally, I hope the entire structure of me having expressions as well as statements being completely separated is not too big of a deal too.

It would be very nice if someone could help me here hence so far I really like chumsky as a parser.

If there happen to be any questions about my code don't hesitate to ask them!

Note: I put this up as a discussion as I am not sure if it is my code being not well written or a problem with the api.

[zesterer]

I've not fully looked into the code, but this sounds very typical of left-recursion. Chumsky is, underneath, a PEG parser and left recursion will cause infinite recursion.

You can resolve this by restating your grammar in a way that avoids left recursion.

For example, the left-recursive pattern

expr = (expr + "+" + atom) | atom

can be restated as

expr = atom + ("+" + atom)*

This section in the tutorial about binary operators demonstrates using this technique to properly parse binary operators with precedence in this way.

[FussballAndy]

Thank you very much, After a little bit of debugging I got it working! 🎉