Disclaimer: I've never read through YAHT, so I don't know if I'm
missing any context...
Your intuition was correct: you do want to lift the MonadPlus property
through the StateT transformer. This is the key to the transformer
libraries, each transformer both contributes a computational feature
(MonadState, MonadReader, MonadPlus, MonadError, ...) as well as
maintains all features of the monad it's transforming.
You've noticed that we need an instance in order to have the StateT
monad transformer maintain the MonadPlus property of its inner monad.
Beyond the encouragement that your intuition was dead on, I would
suggest tryng to use mzero and mplus sooner than you are doing so. In
other words, your method defintions should be simpler.
Good luck!
Nick
On 10/30/06, Magnus Therning <[EMAIL PROTECTED]> wrote:
I can't get searchAll5[1] in Yet Another Haskell Tutorial to run. Ghci
complains that it can't find a MonadPlus that satisfies the required
type; it needs a MonadPlus.
I suspect this is due to the use of 'mzero' and 'mplus', without making
StateT a MonadPlus. My thought for this was to "push" mzero into the
inner monad and have 'mplus' pass on to the inner monad:
instance Monad m => MonadPlus (StateT state m) where
mzero = StateT (\s -> return (s, mzero))
(StateT m1) `mplus` (StateT m2) = StateT (\s -> do
(s1, a) <- m1 s
(s2, b) <- m2 s1
return (s2, a `mplus` b))
What's above isn't accepted by ghci, and I don't even know for sure that
my thought makes sense.
/M
[1]: http://en.wikibooks.org/wiki/Haskell/YAHT/Monads#Monad_Transformers
--
Magnus Therning (OpenPGP: 0xAB4DFBA4)
[EMAIL PROTECTED] Jabber: [EMAIL PROTECTED]
http://therning.org/magnus
Software is not manufactured, it is something you write and publish.
Keep Europe free from software patents, we do not want censorship
by patent law on written works.
Finagle's Fifth Law:
Always draw your curves, then plot your readings.
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
_______________________________________________
Haskell-Cafe mailing list
Haskell-Cafe@haskell.org
http://www.haskell.org/mailman/listinfo/haskell-cafe
