| Copyright | (C) 2014-2017 Merijn Verstraaten |
|---|---|
| License | BSD-style (see the file LICENSE) |
| Maintainer | Merijn Verstraaten <merijn@inconsistent.nl> |
| Stability | experimental |
| Portability | haha |
| Safe Haskell | Trustworthy |
| Language | Haskell2010 |
BroadcastChan
Description
A closable, fair, single-wakeup channel that avoids the 0 reader space leak
that Control.Concurrent.Chan from base suffers from.
The Chan type from Control.Concurrent.Chan consists of both a read
and write end combined into a single value. This means there is always at
least 1 read end for a Chan, which keeps any values written to it alive.
This is a problem for applications/libraries that want to have a channel
that can have zero listeners.
Suppose we have an library that produces events and we want to let users
register to receive events. If we use a channel and write all events to it,
we would like to drop and garbage collect any events that take place when
there are 0 listeners. The always present read end of Chan from base
makes this impossible. We end up with a Chan that forever accumulates
more and more events that will never get removed, resulting in a memory
leak.
BroadcastChan splits channels into separate read and write ends. Any
message written to a a channel with no existing read end is immediately
dropped so it can be garbage collected. Once a read end is created, all
messages written to the channel will be accessible to that read end.
Once all read ends for a channel have disappeared and been garbage collected, the channel will return to dropping messages as soon as they are written.
Why should I use BroadcastChan over Control.Concurrent.Chan?
BroadcastChanis closable,BroadcastChanhas no 0 reader space leak,BroadcastChanhas comparable or better performance.
Why should I use BroadcastChan over various (closable) STM channels?
BroadcastChanis single-wakeup,BroadcastChanis fair,BroadcastChanperforms better under contention.
- data BroadcastChan d a
- data Direction
- type In = In
- type Out = Out
- newBroadcastChan :: IO (BroadcastChan In a)
- newBChanListener :: BroadcastChan In a -> IO (BroadcastChan Out a)
- readBChan :: BroadcastChan Out a -> IO (Maybe a)
- writeBChan :: BroadcastChan In a -> a -> IO Bool
- closeBChan :: BroadcastChan In a -> IO Bool
- isClosedBChan :: BroadcastChan In a -> IO Bool
- foldBChan :: MonadIO m => (x -> a -> m x) -> m x -> (x -> m b) -> BroadcastChan In a -> m (m b)
- foldBChan' :: MonadIO m => (x -> a -> m x) -> m x -> (x -> m b) -> BroadcastChan In a -> m (m b)
- getBChanContents :: BroadcastChan In a -> IO [a]
Datatypes
data BroadcastChan d a Source #
The abstract type representing the read or write end of a BroadcastChan.
Instances
| Eq (BroadcastChan d a) Source # | |
Used with DataKinds as phantom type indicating whether a BroadcastChan
value is a read or write end.
Constructors
| In | Indicates a write |
| Out | Indicates a read |
Alias for the In type from the Direction kind, allows users to write
the type without enabling BroadcastChan In aDataKinds.
Alias for the Out type from the Direction kind, allows users to write
the type without enabling BroadcastChan Out aDataKinds.
Construction
newBroadcastChan :: IO (BroadcastChan In a) Source #
Creates a new BroadcastChan write end.
newBChanListener :: BroadcastChan In a -> IO (BroadcastChan Out a) Source #
Create a new read end for a BroadcastChan. Will receive all messages
written to the channel after this read end is created.
Basic Operations
readBChan :: BroadcastChan Out a -> IO (Maybe a) Source #
Read the next value from the read end of a BroadcastChan. Returns
Nothing if the BroadcastChan is closed and empty.
See BroadcastChan.Throw.readBChan for an exception
throwing variant.
writeBChan :: BroadcastChan In a -> a -> IO Bool Source #
Write a value to write end of a BroadcastChan. Any messages written
while there are no live read ends are dropped on the floor and can be
immediately garbage collected, thus avoiding space leaks.
The return value indicates whether the write succeeded, i.e., True if the
message was written, False is the channel is closed.
See BroadcastChan.Throw.writeBChan for an
exception throwing variant.
closeBChan :: BroadcastChan In a -> IO Bool Source #
Close a BroadcastChan, disallowing further writes. Returns True if the
BroadcastChan was closed. Returns False if the BroadcastChan was
already closed.
isClosedBChan :: BroadcastChan In a -> IO Bool Source #
Check whether a BroadcastChan is closed. True means it's closed,
False means it's writable. However:
Beware of TOC-TOU races: It is possible for a BroadcastChan to be
closed by another thread. If multiple threads use the same BroadcastChan a
closeBChan from another thread might result in the channel being closed
right after isClosedBChan returns.
Utility functions
foldBChan :: MonadIO m => (x -> a -> m x) -> m x -> (x -> m b) -> BroadcastChan In a -> m (m b) Source #
Monadic fold of the BroadcastChan's elements. Can be used with Tekmo's
foldl package:
Control.Foldl.impurely foldBChan :: MonadIO m => FoldM m a b -> BroadcastChan In a -> m (m b)
foldBChan' :: MonadIO m => (x -> a -> m x) -> m x -> (x -> m b) -> BroadcastChan In a -> m (m b) Source #
- Strict* version of
foldBChan.
getBChanContents :: BroadcastChan In a -> IO [a] Source #
Return a lazy list representing everything written to the supplied
BroadcastChan after this IO action returns. Similar to
getChanContents.
Uses unsafeInterleaveIO to defer the IO operations.