Question

When I program in Javascript, I find it extremely convenient to be able to use the debugger to halt program execution at any point and to be able to runs commands and inspect variables from there.

Now, back in Haskell, is there a way to run arbitrary functions in the interactive GHCI REPL or am I limited to things declared at the toplevel?

What is the “standard” approach to working and debugging inner functions and values?

Answer 1

When you are stopped at a breakpoint in GHCi, you can access anything that’s in scope. Let’s say you have a function like this:

foo :: Int -> Int
foo x = g (x + 2)
  where g y = x^y 

You can set a breakpoint on foo and try calling it:

> :break foo
Breakpoint 1 activated at /tmp/Foo.hs:(2,1)-(3,17)
> foo 42
Stopped at /tmp/Foo.hs:(2,1)-(3,17)
_result :: Int = _

g is not in scope yet at this point, so we have to step once:

[/tmp/Foo.hs:(2,1)-(3,17)] > :step
Stopped at /tmp/Foo.hs:2:9-17
_result :: Int = _
g :: Integral b => b -> Int = _
x :: Int = 42

Now that g is in scope, we can use it like any top-level function:

[/tmp/Foo.hs:2:9-17] > g 2
1764
[/tmp/Foo.hs:2:9-17] > g 3
74088