- Although this module calls
unsafeInterleaveIOfor you, it cannot take the responsibility from you. Using this module is still as unsafe as calling
unsafeInterleaveIOmanually. Thus we recommend to wrap the lazy I/O monad into a custom
newtypewith a restricted set of operations which is considered safe for interleaving I/O actions.
- Operations like
hCloseare usually not safe within this monad, since they will only get executed, if their result is consumed. Since this result is often
()this is quite unusual. It will also often be the case, that not the complete output is read, and thus the closing action is never reached. It is certainly best to call a closing action after you wrote the complete result of the lazy I/O monad somewhere.
return a :: LazyIO ais very different from
interleave (return a) :: LazyIO a. The first one does not trigger previous IO actions, whereas the second one does. This is the reason why we do not instantiate
liftIO = LazyIO.interleave, despite the matching type signatures. One of the
MonadIOlaws explicitly requires
return a = liftIO (return a).
- We advise to lift strict IO functions into the lazy IO monad.
Lifting a function like
readFilemay lead to unintended interleaving. In the future we may enforce that using the
Use it like
import qualified System.IO.Lazy as LazyIO LazyIO.run $ do LazyIO.interleave $ putStr "enter first line:" x <- LazyIO.interleave getLine LazyIO.interleave $ putStr "enter second line:" y <- LazyIO.interleave getLine return x
Because only the first entered line is needed,
only the first prompt and the first
getLine is executed.