ulci (untyped lambda calculus interpreter)

---

ulci is yet another interpreter for the untyped lambda calculus, written in c. It parses expressions such as (\x.x y) z and (\x. x) (\y.y y) and evaluates them. Includes an assignment statement:

set Name = Expression

For example, after assigning set id = \x.x, id id id x evaluates to x.

It interprets files containing lambda expressions (or assignments) evaluating each line (./ulci filename). Run without parameters it starts an interactive shell.

Dependencies:

The only dependency needed is readline, but not strictly because the interpreter can be built #defineing NO_READLINE or simply building

make build/ulci-no-readline

which will have the same behaviour when passing a file and obviously a different repl.

For (the ridiculously few) unit tests, it uses the following submodule:

https://github.com/sheredom/utest.h

git clone --recursive https://github.com/nsanmartin/untyped-lambda-calculus.git

or git submodule update --init --recursive if already cloned.

This submodue is only used to run build/utests. Neither lexer nor ulci (the interpreter) nor itests need that dependency.

Build:

mkdir build
make build/ulci

Run

You either pass expressions from stdin and use it interactively like: $ ./build/ulci

> set apply = (\x.\y. x y)
\x.\y.x y
> set dup = (\x.x x)
\x.x x
> apply dup A
A A
>

and

$ echo "(\x.A) B" | ./build/ulci
> (\x.A) B
A

Or you can pass the paths of files containing valid ulc expressions

./ulci foo.ulc
$ ./build/ulci samples/0.ulc samples/1.ulc
\x.x
A
\t.y