Hugs !! Partial Application Bug
Asked Answered
C

1

6

Hugs seems to have a problem with several non-enbraced !! in a partial application.

While this works fine in GHCi:

([[0]]!!0!!)0

Hugs reports a syntax error for the ).

Is this a bug in Hugs?

Adding an extra brace for the second list index operator works though:

(([[0]]!!)0!!)0

or

(([[0]]!!0)!!)0
Cohl answered 26/3, 2012 at 14:39 Comment(1)
I am not sure. I guess it's a different interpretation of the Haskell standards. This actually happens with every section where more than one operators appear on the lowest precedence.Flexile
C
2

This is a known issue in Hugs. From the Hugs 98 Users Guide section on Expressions:

In Hugs, the expression must be an fexp (or case or do). Legal expressions like (a+b+) and (a*b+) are rejected.

Digression Alert

Maybe this is what FUZxxl was talking about in his comment?

Try defining your own (!!) function in ghc and set it to have right-associative fixity:

import Prelude hiding ((!!))
infixr 5 !! -- infixr will make it right associative
(!!) a b = head . drop b $ a

Now that line won't work in ghci either!

ghci> :t ([[0]] !! 0 !!)

<interactive>:1:1:
    The operator `!!' [infixr 5] of a section
        must have lower precedence than that of the operand,
          namely `!!' [infixr 5]
        in the section: `[[0]] !! 0 !!'

Because (!!) has been set with infixr and is now right-associative. If you use infixl, that line works fine.

This is a completely separate issue from the question you asked though. This is about left vs right associativity, whereas the problem with Hugs is that it just doesn't parse an expression like (a+b+).

Costa answered 29/3, 2012 at 20:38 Comment(0)

© 2022 - 2024 — McMap. All rights reserved.