Copyright | (c) The University of Glasgow 2001-2009 |
---|---|
License | BSD-style (see the file LICENSE) |
Maintainer | libraries@haskell.org |
Stability | stable |
Portability | portable |
Safe Haskell | Safe |
Language | Haskell2010 |
This module provides an overloaded function, deepseq
, for fully
evaluating data structures (that is, evaluating to "Normal Form").
A typical use is to prevent resource leaks in lazy IO programs, by forcing all characters from a file to be read. For example:
import System.IO import Control.DeepSeq main = do h <- openFile "f" ReadMode s <- hGetContents h s `deepseq` hClose h return s
deepseq
differs from seq
as it traverses data structures deeply,
for example, seq
will evaluate only to the first constructor in
the list:
> [1,2,undefined] `seq` 3 3
While deepseq
will force evaluation of all the list elements:
> [1,2,undefined] `deepseq` 3 *** Exception: Prelude.undefined
Another common use is to ensure any exceptions hidden within lazy fields of a data structure do not leak outside the scope of the exception handler, or to force evaluation of a data structure in one thread, before passing to another thread (preventing work moving to the wrong threads).
Since: 1.1.0.0
Documentation
deepseq :: NFData a => a -> b -> b Source
deepseq
: fully evaluates the first argument, before returning the
second.
The name deepseq
is used to illustrate the relationship to seq
:
where seq
is shallow in the sense that it only evaluates the top
level of its argument, deepseq
traverses the entire data structure
evaluating it completely.
deepseq
can be useful for forcing pending exceptions,
eradicating space leaks, or forcing lazy I/O to happen. It is
also useful in conjunction with parallel Strategies (see the
parallel
package).
There is no guarantee about the ordering of evaluation. The
implementation may evaluate the components of the structure in
any order or in parallel. To impose an actual order on
evaluation, use pseq
from Control.Parallel in the
parallel
package.
Since: 1.1.0.0
($!!) :: NFData a => (a -> b) -> a -> b infixr 0 Source
the deep analogue of $!
. In the expression f $!! x
, x
is
fully evaluated before the function f
is applied to it.
Since: 1.2.0.0
force :: NFData a => a -> a Source
a variant of deepseq
that is useful in some circumstances:
force x = x `deepseq` x
force x
fully evaluates x
, and then returns it. Note that
force x
only performs evaluation when the value of force x
itself is demanded, so essentially it turns shallow evaluation into
deep evaluation.
force
can be conveniently used in combination with ViewPatterns
:
{-# LANGUAGE BangPatterns, ViewPatterns #-} import Control.DeepSeq someFun :: ComplexData -> SomeResult someFun (force -> !arg) = {- 'arg' will be fully evaluated -}
Another useful application is to combine force
with
evaluate
in order to force deep evaluation
relative to other IO
operations:
import Control.Exception (evaluate) import Control.DeepSeq main = do result <- evaluate $ force $ pureComputation {- 'result' will be fully evaluated at this point -} return ()
Since: 1.2.0.0
A class of types that can be fully evaluated.
Since: 1.1.0.0
Nothing
rnf
should reduce its argument to normal form (that is, fully
evaluate all sub-components), and then return '()'.
Generic
NFData
deriving
Starting with GHC 7.2, you can automatically derive instances
for types possessing a Generic
instance.
{-# LANGUAGE DeriveGeneric #-} import GHC.Generics (Generic) import Control.DeepSeq data Foo a = Foo a String deriving (Eq, Generic) instance NFData a => NFData (Foo a) data Colour = Red | Green | Blue deriving Generic instance NFData Colour
Starting with GHC 7.10, the example above can be written more
concisely by enabling the new DeriveAnyClass
extension:
{-# LANGUAGE DeriveGeneric, DeriveAnyClass #-} import GHC.Generics (Generic) import Control.DeepSeq data Foo a = Foo a String deriving (Eq, Generic, NFData) data Colour = Red | Green | Blue deriving (Generic, NFData)
Compatibility with previous deepseq
versions
Prior to version 1.4.0.0, the default implementation of the rnf
method was defined as
rnf
a =seq
a ()
However, starting with deepseq-1.4.0.0
, the default
implementation is based on DefaultSignatures
allowing for
more accurate auto-derived NFData
instances. If you need the
previously used exact default rnf
method implementation
semantics, use
instance NFData Colour where rnf x = seq x ()
or alternatively
{-# LANGUAGE BangPatterns #-} instance NFData Colour where rnf !_ = ()
NFData Bool | |
NFData Char | |
NFData Double | |
NFData Float | |
NFData Int | |
NFData Int8 | |
NFData Int16 | |
NFData Int32 | |
NFData Int64 | |
NFData Integer | |
NFData Word | |
NFData Word8 | |
NFData Word16 | |
NFData Word32 | |
NFData Word64 | |
NFData TypeRep | NOTE: Only defined for Since: 1.4.0.0 |
NFData () | |
NFData Void | Since: 1.4.0.0 |
NFData Unique | Since: 1.4.0.0 |
NFData Natural | Since: 1.4.0.0 |
NFData Version | Since: 1.3.0.0 |
NFData ThreadId | Since: 1.4.0.0 |
NFData CChar | Since: 1.4.0.0 |
NFData CSChar | Since: 1.4.0.0 |
NFData CUChar | Since: 1.4.0.0 |
NFData CShort | Since: 1.4.0.0 |
NFData CUShort | Since: 1.4.0.0 |
NFData CInt | Since: 1.4.0.0 |
NFData CUInt | Since: 1.4.0.0 |
NFData CLong | Since: 1.4.0.0 |
NFData CULong | Since: 1.4.0.0 |
NFData CLLong | Since: 1.4.0.0 |
NFData CULLong | Since: 1.4.0.0 |
NFData CFloat | Since: 1.4.0.0 |
NFData CDouble | Since: 1.4.0.0 |
NFData CPtrdiff | Since: 1.4.0.0 |
NFData CSize | Since: 1.4.0.0 |
NFData CWchar | Since: 1.4.0.0 |
NFData CSigAtomic | Since: 1.4.0.0 |
NFData CClock | Since: 1.4.0.0 |
NFData CTime | Since: 1.4.0.0 |
NFData CUSeconds | Since: 1.4.0.0 |
NFData CSUSeconds | Since: 1.4.0.0 |
NFData CFile | Since: 1.4.0.0 |
NFData CFpos | Since: 1.4.0.0 |
NFData CJmpBuf | Since: 1.4.0.0 |
NFData CIntPtr | Since: 1.4.0.0 |
NFData CUIntPtr | Since: 1.4.0.0 |
NFData CIntMax | Since: 1.4.0.0 |
NFData CUIntMax | Since: 1.4.0.0 |
NFData All | Since: 1.4.0.0 |
NFData Any | Since: 1.4.0.0 |
NFData TyCon | NOTE: Only defined for Since: 1.4.0.0 |
NFData Fingerprint | Since: 1.4.0.0 |
NFData a => NFData [a] | |
(Integral a, NFData a) => NFData (Ratio a) | |
NFData (StableName a) | Since: 1.4.0.0 |
NFData a => NFData (Identity a) | Since: 1.4.0.0 |
NFData (Fixed a) | Since: 1.3.0.0 |
NFData a => NFData (Complex a) | |
NFData a => NFData (ZipList a) | Since: 1.4.0.0 |
NFData a => NFData (Dual a) | Since: 1.4.0.0 |
NFData a => NFData (Sum a) | Since: 1.4.0.0 |
NFData a => NFData (Product a) | Since: 1.4.0.0 |
NFData a => NFData (First a) | Since: 1.4.0.0 |
NFData a => NFData (Last a) | Since: 1.4.0.0 |
NFData a => NFData (Down a) | Since: 1.4.0.0 |
NFData a => NFData (Maybe a) | |
NFData (a -> b) | This instance is for convenience and consistency with Since: 1.3.0.0 |
(NFData a, NFData b) => NFData (Either a b) | |
(NFData a, NFData b) => NFData (a, b) | |
(Ix a, NFData a, NFData b) => NFData (Array a b) | |
NFData a => NFData (Const a b) | Since: 1.4.0.0 |
NFData (Proxy * a) | Since: 1.4.0.0 |
(NFData a, NFData b, NFData c) => NFData (a, b, c) | |
(NFData a, NFData b, NFData c, NFData d) => NFData (a, b, c, d) | |
(NFData a1, NFData a2, NFData a3, NFData a4, NFData a5) => NFData (a1, a2, a3, a4, a5) | |
(NFData a1, NFData a2, NFData a3, NFData a4, NFData a5, NFData a6) => NFData (a1, a2, a3, a4, a5, a6) | |
(NFData a1, NFData a2, NFData a3, NFData a4, NFData a5, NFData a6, NFData a7) => NFData (a1, a2, a3, a4, a5, a6, a7) | |
(NFData a1, NFData a2, NFData a3, NFData a4, NFData a5, NFData a6, NFData a7, NFData a8) => NFData (a1, a2, a3, a4, a5, a6, a7, a8) | |
(NFData a1, NFData a2, NFData a3, NFData a4, NFData a5, NFData a6, NFData a7, NFData a8, NFData a9) => NFData (a1, a2, a3, a4, a5, a6, a7, a8, a9) |