Proposal: add 'state' to the MonadState class

[Antoine Latter]

On Mon, Sep 26, 2011 at 9:50 AM, Edward Kmett <ekmett at gmail.com> wrote:
> On Mon, Sep 26, 2011 at 10:41 AM, Antoine Latter <aslatter at gmail.com> wrote:
>>
>> On Mon, Sep 26, 2011 at 9:25 AM, Twan van Laarhoven <twanvl at gmail.com>
>> wrote:
>> > Hello list,
>> >
>> >
>> > Currently the 'state' function is only defined for the State type (aka
>> > StateT Identity):
>> >
>> >    state :: (s -> (a, s)) -> State s a
>> >    state f = StateT (Identity . f)
>> >
>> > But this function makes sense for all MonadState instances. I therefore
>> > propose to either add it to the MonadState class:
>> >
>> >    class MonadState s m | m -> s where
>> >        ...
>> >        state :: (s -> (a,s)) -> m a
>> >        state f = do s <- get
>> >                     let (a,s') = f s
>> >                     set s'
>> >                     return a
>> >
>
> I am 100% behind this change.
> It has been on the short list of changes that I want to make to mtl.
> The main reason that I like it aside from the fact that it witnesses the
> canonical monad homomorphism from State is that it yields more efficient
> definitions for functions like modify, and the often-needed but needlessly
> expensive operations where you bump a counter _and_ want the result for
> something like a fresh variable supply.
> I would like to encourage further public discussion.

If 'modify' is re-written to use this proposed class method, does that
mean I can finally make "modify" safe to use in "BackwardStateT"?

Not that I actually use "BawkardStateT", but it is a fun demo.

Antoine