Add First and Last wrappers around Maybe to Data.Monoid

[jyasskin]

First and Last mappend to the first and last non-Nothing elements of the input. These instances seem to fit with the other simple instances in this file and are useful, for example, to turn foldMap into findFirst or findLast, or to get a really trivial Writer monad.

[jyasskin]

V2 of the First and Last monoids, now with an instance for Maybe itself.

[jyasskin]

Haddock'ed docs from V2

[jyasskin]

Ross Paterson suggested that the "obvious" instance directly on Maybe was to combine two Just's using a Monoid instance on their contents and set mempty==Nothing. David House, Stefan O'Rear, and Jan-Willem Maessen agreed so I added it.

John Meacham objected to the location of all of the types currently in Data.Monoid, but that's a broader discussion than this patch, and nobody expressed agreement anyway.

David Menendez and Spencer Janssen suggested that Maybe's Monoid instance should match its MonadPlus? instance, but neither provided an argument supporting that.

The thread is archived at  http://www.haskell.org/pipermail/libraries/2007-March/006950.html

[igloo]

Patch applied, thanks!