Are LiquidHaskell annotations being considered

[envp]

I think they would add stronger compile time guarantees as well as serve as documentation about Fortran's AST invariants. This could be useful for source to source transformations as a sort of added compile time sanity check

Link to LiquidHaskell

[madgen]

We never considered LiquidHsakell for fortran-src. What sort of annotations did you have in mind?

What I really like to do one day is to rewrite the AST to use GADTs. There is way too much redundancy.

[envp]

I am still unfamiliar with a lot of the codebase of fortran-src, and hence don't have specific annotations in mind. The general flavor of annotations I had in mind was stuff that could help migrate legacy fortran to more modern standards through code generation. I am yet to flesh out the idea completely.