Incremental untyped lambda calculus.
This is a first attempt to implement an incremental lambda-calculus based on the paper and project by Paolo G. Giarrusso. It's not very useful yet, it's just a toy. This was also based on Stephen Diehl's lambda calculus implementation from "Write You a Haskell".
There are many theoretical problems with this implementation. For example, we're not using any canonical name-capturing scheme (like de Bruijn Indexes), so automatic renaming conflicts are likely.
There are many handy built-in functions that map directly to lambda terms: "true", "false", "and", "or", "not", "ifthenelse", "iszero", "pair", "first", "second", "nil", "succ", "pred", "plus", "mul", "pow", "sub". Significantly, there is also a derive function built-in that will compute the derivative of a given term.
To compile and run:
$ cabal runUntyped> (\x. x) 1
=> \x . x
=> 1
=> x
1
Untyped> derive (\x. x)
(\x dx . dx)Released under MIT license.