Johannes Waldmann | 5 Oct 13:11 2012
Picon

referential transparency? (for fixity of local operators)

I was really surprised at the following:

*Main> 1 + 2 * 3
7

*Main> ( \ (+) (*) -> 1 + 2 * 3 ) (+) (*)
9

because I was somehow assuming that either

a) the Prelude fixities of the operators are kept
b) or they are undefined, so the parser rejects.

but the Haskell standard says "Any operator lacking a fixity declaration 
is assumed to be infixl 9". This really should be "infix 9"?
Roman Cheplyaka | 5 Oct 13:22 2012

Re: referential transparency? (for fixity of local operators)

* Johannes Waldmann <waldmann <at> imn.htwk-leipzig.de> [2012-10-05 11:11:48+0000]
> I was really surprised at the following:
> 
> *Main> 1 + 2 * 3
> 7
> 
> *Main> ( \ (+) (*) -> 1 + 2 * 3 ) (+) (*)
> 9
> 
> because I was somehow assuming that either
> 
> a) the Prelude fixities of the operators are kept
> b) or they are undefined, so the parser rejects.
> 
> but the Haskell standard says "Any operator lacking a fixity declaration 
> is assumed to be infixl 9". This really should be "infix 9"?

This behaviour is really handy when you use functions as operators
(using backticks notation). They typically lack infix annotations, but
having to put parentheses would be very annoying.

Roman
Clark Gaebel | 5 Oct 14:50 2012
Picon
Picon

Re: referential transparency? (for fixity of local operators)

Compile with -Wall and the flaw becomes obvious:

<interactive>:2:5:
    Warning: This binding for `+' shadows the existing binding
               imported from `Prelude' (and originally defined in `GHC.Num')

<interactive>:2:9:
    Warning: This binding for `*' shadows the existing binding
               imported from `Prelude' (and originally defined in `GHC.Num')

<interactive>:2:16:
    Warning: Defaulting the following constraint(s) to type `Integer'
               (Num a0) arising from the literal `1'
    In the first argument of `(+)', namely `1'
    In the first argument of `(*)', namely `1 + 2'
    In the expression: 1 + 2 * 3

<interactive>:2:16:
    Warning: Defaulting the following constraint(s) to type `Integer'
               (Num a0) arising from the literal `1' at <interactive>:2:16
               (Show a0) arising from a use of `print' at <interactive>:2:1-34
    In the first argument of `(+)', namely `1'
    In the first argument of `(*)', namely `1 + 2'
    In the expression: 1 + 2 * 3

Shadowing is bad, and tends (as in this case) to be confusing.

  - Clark

On Fri, Oct 5, 2012 at 7:22 AM, Roman Cheplyaka <roma <at> ro-che.info> wrote:
* Johannes Waldmann <waldmann <at> imn.htwk-leipzig.de> [2012-10-05 11:11:48+0000]
> I was really surprised at the following:
>
> *Main> 1 + 2 * 3
> 7
>
> *Main> ( \ (+) (*) -> 1 + 2 * 3 ) (+) (*)
> 9
>
> because I was somehow assuming that either
>
> a) the Prelude fixities of the operators are kept
> b) or they are undefined, so the parser rejects.
>
> but the Haskell standard says "Any operator lacking a fixity declaration
> is assumed to be infixl 9". This really should be "infix 9"?

This behaviour is really handy when you use functions as operators
(using backticks notation). They typically lack infix annotations, but
having to put parentheses would be very annoying.

Roman

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe

_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe <at> haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
wren ng thornton | 5 Oct 15:36 2012

Re: referential transparency? (for fixity of local operators)

On 10/5/12 7:11 AM, Johannes Waldmann wrote:
> I was really surprised at the following:
>
> *Main> 1 + 2 * 3
> 7
>
> *Main> ( \ (+) (*) -> 1 + 2 * 3 ) (+) (*)
> 9
>
> because I was somehow assuming that either
>
> a) the Prelude fixities of the operators are kept

After dealing with how Coq handles infix operators, I'm profoundly glad 
that Haskell doesn't do this. The fixity of an operator is a property of 
the operator itself, not of the notation used to display the operator. 
There's no reason for my (+) operator to have the same fixity as your 
(+) operator, regardless of whether we happen to use the same notation 
or not.

> b) or they are undefined, so the parser rejects.
>
> but the Haskell standard says "Any operator lacking a fixity declaration
> is assumed to be infixl 9". This really should be "infix 9"?

Yep, default fixity. As Roman Cheplyaka mentioned, this is especially 
helpful for ad hoc uses of backticks. (For non-adhoc uses, it's best to 
give an explicit fixity just like for other infix operators.)

As for whether the default should be "infix 9" instead of "infixl 9", I 
dunno...

--

-- 
Live well,
~wren
Johannes Waldmann | 6 Oct 11:18 2012
Picon

Re: referential transparency? (for fixity of local operators)

wren ng thornton <wren <at> freegeek.org> writes:

> As for whether the default should be "infix 9" instead of "infixl 9" ...

that was exactly the point of my message. - J.
Albert Y. C. Lai | 6 Oct 17:18 2012
Picon

Re: referential transparency? (for fixity of local operators)

On 12-10-06 05:18 AM, Johannes Waldmann wrote:
> wren ng thornton <wren <at> freegeek.org> writes:
>
>> As for whether the default should be "infix 9" instead of "infixl 9" ...
>
> that was exactly the point of my message. - J.

Perhaps, half of the people want infixl, another half of the people want 
infixr, and so at the end the middle ground of neither l nor r is taken.

Perhaps, political decision is about saying no to both the left and the 
right? :)

Gmane