base-4.11.1.0: Basic libraries

Copyright(c) The University of Glasgow 2007
LicenseBSD-style (see the file libraries/base/LICENSE)
Maintainer[email protected]
Stabilityexperimental
Portabilitynon-portable
Safe HaskellSafe
LanguageHaskell2010

System.Timeout

Description

Attach a timeout event to arbitrary IO computations.

Synopsis

Documentation

timeout :: Int -> IO a -> IO (Maybe a) Source #

Wrap an IO computation to time out and return Nothing in case no result is available within n microseconds (1/10^6 seconds). In case a result is available before the timeout expires, Just a is returned. A negative timeout interval means "wait indefinitely". When specifying long timeouts, be careful not to exceed maxBound :: Int.

>>> timeout 1000000 (threadDelay 1000 *> pure "finished on time")
Just "finished on time"
>>> timeout 10000 (threadDelay 100000 *> pure "finished on time")
Nothing

The design of this combinator was guided by the objective that timeout n f should behave exactly the same as f as long as f doesn't time out. This means that f has the same myThreadId it would have without the timeout wrapper. Any exceptions f might throw cancel the timeout and propagate further up. It also possible for f to receive exceptions thrown to it by another thread.

A tricky implementation detail is the question of how to abort an IO computation. This combinator relies on asynchronous exceptions internally. The technique works very well for computations executing inside of the Haskell runtime system, but it doesn't work at all for non-Haskell code. Foreign function calls, for example, cannot be timed out with this combinator simply because an arbitrary C function cannot receive asynchronous exceptions. When timeout is used to wrap an FFI call that blocks, no timeout event can be delivered until the FFI call returns, which pretty much negates the purpose of the combinator. In practice, however, this limitation is less severe than it may sound. Standard I/O functions like hGetBuf, hPutBuf, Network.Socket.accept, or hWaitForInput appear to be blocking, but they really don't because the runtime system uses scheduling mechanisms like select(2) to perform asynchronous I/O, so it is possible to interrupt standard socket I/O or file I/O using this combinator.