Removing syntactic sugar: List comprehension in Haskell

Question

Can I unsugar list comprehension in this expression:

[(i,j) | i <- [1..4], j <- [i+1..4]]

This is the output:

[(1,2),         


        

Answer 1

There is yet another translation scheme, it is due to Wadler, as far as I know.

This would give:

let
    lc_outer (x:xs) = let lc_inner (y:ys) = (x,y) : lc_inner ys
                          lc_inner []     = lc_outer xs
                      in lc_inner [x+1.. 4]
    lc_outer [] = []
in  lc_outer [1..4]

This translation avoids needless construction of singleton lists in the innermost level that would need to get flattened with concatMap later.