[Haskell-cafe] About the Monad Transformers

[Miguel Mitrofanov]

On 17 Jun 2009, at 19:54, .shawn wrote:

> Why does the type signature of mapTreeM look like this?

Why not? I can't think of any other possibility.

> And what does it mean by "The lift tell us that we're going to be  
> executing a command in an enclosed monad. In this case the enclosed  
> monad is IO"?

Wild guess: I think they mean exactly what they said. Literally.

> Why does the author put lift here?

Because he wanted to work in a more general setting than just IO  
monad. He needed IO, of course, because of "putStrLn", but he wanted  
to give the user of this function an opportunity to do something  
besides just IO.

> How does the lift work?

"lift" is defined in the "MonadTrans" type class, so it works  
differently depending on what transformer you use. You can even define  
your own one and "lift" would work as you write it. The only thing we  
know for sure is that whichever transformer you use, "lift" would have  
the type "m x -> t m x", where "t" is a transformer in question. So,  
it would be OK to apply it to, say, something of type "IO a", and  
you'd have something of type "t IO a" as a result.