How to get rid of IO

From HaskellWiki
Revision as of 15:17, 8 March 2013 by Dmwit (talk | contribs) (→‎Answer: copy editing: fix whitespace, remove acronyms, remove value judgments, correct grammar)
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Question

I have something of type IO a, but I need something of type a How can I get that?

Answer

You can get rid of it, but you almost certainly don't need to. The special safety belt of Haskell is that you cannot get rid of IO! Nonetheless, the biggest parts of Haskell programs are and should be non-IO functions. Applications using both IO and non-IO functions are written by plugging together these two flavors of functions using functions like (>>=). These functions allow useful IO while having all the safety properties of a pure functional programming language.

You can hide some functions using do notation, which looks like this:

do text <- readFile "foo"
   writeFile "bar" (someComplicatedNonIOOperation text)

Without hiding the functions, this would look like:

writeFile "bar" . someComplicatedNonIOOperation =<< readFile "foo"

What we didn't tell you at the beginning

There is a function which directly answers the initial question, namely, unsafePerformIO. It is however not intended for conveniently getting rid of the IO constructor. It must only be used to wrap IO functions that behave like non-IO functions, Since this property cannot be checked by the compiler, it is your task and thus the unsafe part of the name. (Some library writers have abused that name component for partial functions. Don't get confused!) You will only need this in rare cases and only experienced programmers shall do this.

See also