diff options
author | Valentin Popov <valentin@popov.link> | 2024-01-08 00:21:28 +0300 |
---|---|---|
committer | Valentin Popov <valentin@popov.link> | 2024-01-08 00:21:28 +0300 |
commit | 1b6a04ca5504955c571d1c97504fb45ea0befee4 (patch) | |
tree | 7579f518b23313e8a9748a88ab6173d5e030b227 /vendor/flume/src/lib.rs | |
parent | 5ecd8cf2cba827454317368b68571df0d13d7842 (diff) | |
download | fparkan-1b6a04ca5504955c571d1c97504fb45ea0befee4.tar.xz fparkan-1b6a04ca5504955c571d1c97504fb45ea0befee4.zip |
Initial vendor packages
Signed-off-by: Valentin Popov <valentin@popov.link>
Diffstat (limited to 'vendor/flume/src/lib.rs')
-rw-r--r-- | vendor/flume/src/lib.rs | 1142 |
1 files changed, 1142 insertions, 0 deletions
diff --git a/vendor/flume/src/lib.rs b/vendor/flume/src/lib.rs new file mode 100644 index 0000000..c9bb3ee --- /dev/null +++ b/vendor/flume/src/lib.rs @@ -0,0 +1,1142 @@ +//! # Flume +//! +//! A blazingly fast multi-producer, multi-consumer channel. +//! +//! *"Do not communicate by sharing memory; instead, share memory by communicating."* +//! +//! ## Why Flume? +//! +//! - **Featureful**: Unbounded, bounded and rendezvous queues +//! - **Fast**: Always faster than `std::sync::mpsc` and sometimes `crossbeam-channel` +//! - **Safe**: No `unsafe` code anywhere in the codebase! +//! - **Flexible**: `Sender` and `Receiver` both implement `Send + Sync + Clone` +//! - **Familiar**: Drop-in replacement for `std::sync::mpsc` +//! - **Capable**: Additional features like MPMC support and send timeouts/deadlines +//! - **Simple**: Few dependencies, minimal codebase, fast to compile +//! - **Asynchronous**: `async` support, including mix 'n match with sync code +//! - **Ergonomic**: Powerful `select`-like interface +//! +//! ## Example +//! +//! ``` +//! let (tx, rx) = flume::unbounded(); +//! +//! tx.send(42).unwrap(); +//! assert_eq!(rx.recv().unwrap(), 42); +//! ``` + +#![deny(missing_docs)] + +#[cfg(feature = "select")] +pub mod select; +#[cfg(feature = "async")] +pub mod r#async; + +mod signal; + +// Reexports +#[cfg(feature = "select")] +pub use select::Selector; + +use std::{ + collections::VecDeque, + sync::{Arc, atomic::{AtomicUsize, AtomicBool, Ordering}, Weak}, + time::{Duration, Instant}, + marker::PhantomData, + thread, + fmt, +}; + +#[cfg(feature = "spin")] +use spin1::{Mutex as Spinlock, MutexGuard as SpinlockGuard}; +use crate::signal::{Signal, SyncSignal}; + +/// An error that may be emitted when attempting to send a value into a channel on a sender when +/// all receivers are dropped. +#[derive(Copy, Clone, PartialEq, Eq)] +pub struct SendError<T>(pub T); + +impl<T> SendError<T> { + /// Consume the error, yielding the message that failed to send. + pub fn into_inner(self) -> T { self.0 } +} + +impl<T> fmt::Debug for SendError<T> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + "SendError(..)".fmt(f) + } +} + +impl<T> fmt::Display for SendError<T> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + "sending on a closed channel".fmt(f) + } +} + +impl<T> std::error::Error for SendError<T> {} + +/// An error that may be emitted when attempting to send a value into a channel on a sender when +/// the channel is full or all receivers are dropped. +#[derive(Copy, Clone, PartialEq, Eq)] +pub enum TrySendError<T> { + /// The channel the message is sent on has a finite capacity and was full when the send was attempted. + Full(T), + /// All channel receivers were dropped and so the message has nobody to receive it. + Disconnected(T), +} + +impl<T> TrySendError<T> { + /// Consume the error, yielding the message that failed to send. + pub fn into_inner(self) -> T { + match self { + Self::Full(msg) | Self::Disconnected(msg) => msg, + } + } +} + +impl<T> fmt::Debug for TrySendError<T> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match *self { + TrySendError::Full(..) => "Full(..)".fmt(f), + TrySendError::Disconnected(..) => "Disconnected(..)".fmt(f), + } + } +} + +impl<T> fmt::Display for TrySendError<T> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + TrySendError::Full(..) => "sending on a full channel".fmt(f), + TrySendError::Disconnected(..) => "sending on a closed channel".fmt(f), + } + } +} + +impl<T> std::error::Error for TrySendError<T> {} + +impl<T> From<SendError<T>> for TrySendError<T> { + fn from(err: SendError<T>) -> Self { + match err { + SendError(item) => Self::Disconnected(item), + } + } +} + +/// An error that may be emitted when sending a value into a channel on a sender with a timeout when +/// the send operation times out or all receivers are dropped. +#[derive(Copy, Clone, PartialEq, Eq)] +pub enum SendTimeoutError<T> { + /// A timeout occurred when attempting to send the message. + Timeout(T), + /// All channel receivers were dropped and so the message has nobody to receive it. + Disconnected(T), +} + +impl<T> SendTimeoutError<T> { + /// Consume the error, yielding the message that failed to send. + pub fn into_inner(self) -> T { + match self { + Self::Timeout(msg) | Self::Disconnected(msg) => msg, + } + } +} + +impl<T> fmt::Debug for SendTimeoutError<T> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + "SendTimeoutError(..)".fmt(f) + } +} + +impl<T> fmt::Display for SendTimeoutError<T> { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + SendTimeoutError::Timeout(..) => "timed out sending on a full channel".fmt(f), + SendTimeoutError::Disconnected(..) => "sending on a closed channel".fmt(f), + } + } +} + +impl<T> std::error::Error for SendTimeoutError<T> {} + +impl<T> From<SendError<T>> for SendTimeoutError<T> { + fn from(err: SendError<T>) -> Self { + match err { + SendError(item) => Self::Disconnected(item), + } + } +} + +enum TrySendTimeoutError<T> { + Full(T), + Disconnected(T), + Timeout(T), +} + +/// An error that may be emitted when attempting to wait for a value on a receiver when all senders +/// are dropped and there are no more messages in the channel. +#[derive(Copy, Clone, Debug, PartialEq, Eq)] +pub enum RecvError { + /// All senders were dropped and no messages are waiting in the channel, so no further messages can be received. + Disconnected, +} + +impl fmt::Display for RecvError { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + RecvError::Disconnected => "receiving on a closed channel".fmt(f), + } + } +} + +impl std::error::Error for RecvError {} + +/// An error that may be emitted when attempting to fetch a value on a receiver when there are no +/// messages in the channel. If there are no messages in the channel and all senders are dropped, +/// then `TryRecvError::Disconnected` will be returned. +#[derive(Copy, Clone, Debug, PartialEq, Eq)] +pub enum TryRecvError { + /// The channel was empty when the receive was attempted. + Empty, + /// All senders were dropped and no messages are waiting in the channel, so no further messages can be received. + Disconnected, +} + +impl fmt::Display for TryRecvError { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + TryRecvError::Empty => "receiving on an empty channel".fmt(f), + TryRecvError::Disconnected => "channel is empty and closed".fmt(f), + } + } +} + +impl std::error::Error for TryRecvError {} + +impl From<RecvError> for TryRecvError { + fn from(err: RecvError) -> Self { + match err { + RecvError::Disconnected => Self::Disconnected, + } + } +} + +/// An error that may be emitted when attempting to wait for a value on a receiver with a timeout +/// when the receive operation times out or all senders are dropped and there are no values left +/// in the channel. +#[derive(Copy, Clone, Debug, PartialEq, Eq)] +pub enum RecvTimeoutError { + /// A timeout occurred when attempting to receive a message. + Timeout, + /// All senders were dropped and no messages are waiting in the channel, so no further messages can be received. + Disconnected, +} + +impl fmt::Display for RecvTimeoutError { + fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { + match self { + RecvTimeoutError::Timeout => "timed out waiting on a channel".fmt(f), + RecvTimeoutError::Disconnected => "channel is empty and closed".fmt(f), + } + } +} + +impl std::error::Error for RecvTimeoutError {} + +impl From<RecvError> for RecvTimeoutError { + fn from(err: RecvError) -> Self { + match err { + RecvError::Disconnected => Self::Disconnected, + } + } +} + +enum TryRecvTimeoutError { + Empty, + Timeout, + Disconnected, +} + +// TODO: Investigate some sort of invalidation flag for timeouts +#[cfg(feature = "spin")] +struct Hook<T, S: ?Sized>(Option<Spinlock<Option<T>>>, S); + +#[cfg(not(feature = "spin"))] +struct Hook<T, S: ?Sized>(Option<Mutex<Option<T>>>, S); + +#[cfg(feature = "spin")] +impl<T, S: ?Sized + Signal> Hook<T, S> { + pub fn slot(msg: Option<T>, signal: S) -> Arc<Self> + where + S: Sized, + { + Arc::new(Self(Some(Spinlock::new(msg)), signal)) + } + + fn lock(&self) -> Option<SpinlockGuard<'_, Option<T>>> { + self.0.as_ref().map(|s| s.lock()) + } +} + +#[cfg(not(feature = "spin"))] +impl<T, S: ?Sized + Signal> Hook<T, S> { + pub fn slot(msg: Option<T>, signal: S) -> Arc<Self> + where + S: Sized, + { + Arc::new(Self(Some(Mutex::new(msg)), signal)) + } + + fn lock(&self) -> Option<MutexGuard<'_, Option<T>>> { + self.0.as_ref().map(|s| s.lock().unwrap()) + } +} + +impl<T, S: ?Sized + Signal> Hook<T, S> { + pub fn fire_recv(&self) -> (T, &S) { + let msg = self.lock().unwrap().take().unwrap(); + (msg, self.signal()) + } + + pub fn fire_send(&self, msg: T) -> (Option<T>, &S) { + let ret = match self.lock() { + Some(mut lock) => { + *lock = Some(msg); + None + } + None => Some(msg), + }; + (ret, self.signal()) + } + + pub fn is_empty(&self) -> bool { + self.lock().map(|s| s.is_none()).unwrap_or(true) + } + + pub fn try_take(&self) -> Option<T> { + self.lock().unwrap().take() + } + + pub fn trigger(signal: S) -> Arc<Self> + where + S: Sized, + { + Arc::new(Self(None, signal)) + } + + pub fn signal(&self) -> &S { + &self.1 + } + + pub fn fire_nothing(&self) -> bool { + self.signal().fire() + } +} + +impl<T> Hook<T, SyncSignal> { + pub fn wait_recv(&self, abort: &AtomicBool) -> Option<T> { + loop { + let disconnected = abort.load(Ordering::SeqCst); // Check disconnect *before* msg + let msg = self.lock().unwrap().take(); + if let Some(msg) = msg { + break Some(msg); + } else if disconnected { + break None; + } else { + self.signal().wait() + } + } + } + + // Err(true) if timeout + pub fn wait_deadline_recv(&self, abort: &AtomicBool, deadline: Instant) -> Result<T, bool> { + loop { + let disconnected = abort.load(Ordering::SeqCst); // Check disconnect *before* msg + let msg = self.lock().unwrap().take(); + if let Some(msg) = msg { + break Ok(msg); + } else if disconnected { + break Err(false); + } else if let Some(dur) = deadline.checked_duration_since(Instant::now()) { + self.signal().wait_timeout(dur); + } else { + break Err(true); + } + } + } + + pub fn wait_send(&self, abort: &AtomicBool) { + loop { + let disconnected = abort.load(Ordering::SeqCst); // Check disconnect *before* msg + if disconnected || self.lock().unwrap().is_none() { + break; + } + + self.signal().wait(); + } + } + + // Err(true) if timeout + pub fn wait_deadline_send(&self, abort: &AtomicBool, deadline: Instant) -> Result<(), bool> { + loop { + let disconnected = abort.load(Ordering::SeqCst); // Check disconnect *before* msg + if self.lock().unwrap().is_none() { + break Ok(()); + } else if disconnected { + break Err(false); + } else if let Some(dur) = deadline.checked_duration_since(Instant::now()) { + self.signal().wait_timeout(dur); + } else { + break Err(true); + } + } + } +} + +#[cfg(feature = "spin")] +#[inline] +fn wait_lock<T>(lock: &Spinlock<T>) -> SpinlockGuard<T> { + let mut i = 4; + loop { + for _ in 0..10 { + if let Some(guard) = lock.try_lock() { + return guard; + } + thread::yield_now(); + } + // Sleep for at most ~1 ms + thread::sleep(Duration::from_nanos(1 << i.min(20))); + i += 1; + } +} + +#[cfg(not(feature = "spin"))] +#[inline] +fn wait_lock<'a, T>(lock: &'a Mutex<T>) -> MutexGuard<'a, T> { + lock.lock().unwrap() +} + +#[cfg(not(feature = "spin"))] +use std::sync::{Mutex, MutexGuard}; + +#[cfg(feature = "spin")] +type ChanLock<T> = Spinlock<T>; +#[cfg(not(feature = "spin"))] +type ChanLock<T> = Mutex<T>; + + +type SignalVec<T> = VecDeque<Arc<Hook<T, dyn signal::Signal>>>; +struct Chan<T> { + sending: Option<(usize, SignalVec<T>)>, + queue: VecDeque<T>, + waiting: SignalVec<T>, +} + +impl<T> Chan<T> { + fn pull_pending(&mut self, pull_extra: bool) { + if let Some((cap, sending)) = &mut self.sending { + let effective_cap = *cap + pull_extra as usize; + + while self.queue.len() < effective_cap { + if let Some(s) = sending.pop_front() { + let (msg, signal) = s.fire_recv(); + signal.fire(); + self.queue.push_back(msg); + } else { + break; + } + } + } + } + + fn try_wake_receiver_if_pending(&mut self) { + if !self.queue.is_empty() { + while Some(false) == self.waiting.pop_front().map(|s| s.fire_nothing()) {} + } + } +} + +struct Shared<T> { + chan: ChanLock<Chan<T>>, + disconnected: AtomicBool, + sender_count: AtomicUsize, + receiver_count: AtomicUsize, +} + +impl<T> Shared<T> { + fn new(cap: Option<usize>) -> Self { + Self { + chan: ChanLock::new(Chan { + sending: cap.map(|cap| (cap, VecDeque::new())), + queue: VecDeque::new(), + waiting: VecDeque::new(), + }), + disconnected: AtomicBool::new(false), + sender_count: AtomicUsize::new(1), + receiver_count: AtomicUsize::new(1), + } + } + + fn send<S: Signal, R: From<Result<(), TrySendTimeoutError<T>>>>( + &self, + msg: T, + should_block: bool, + make_signal: impl FnOnce(T) -> Arc<Hook<T, S>>, + do_block: impl FnOnce(Arc<Hook<T, S>>) -> R, + ) -> R { + let mut chan = wait_lock(&self.chan); + + if self.is_disconnected() { + Err(TrySendTimeoutError::Disconnected(msg)).into() + } else if !chan.waiting.is_empty() { + let mut msg = Some(msg); + + loop { + let slot = chan.waiting.pop_front(); + match slot.as_ref().map(|r| r.fire_send(msg.take().unwrap())) { + // No more waiting receivers and msg in queue, so break out of the loop + None if msg.is_none() => break, + // No more waiting receivers, so add msg to queue and break out of the loop + None => { + chan.queue.push_back(msg.unwrap()); + break; + } + Some((Some(m), signal)) => { + if signal.fire() { + // Was async and a stream, so didn't acquire the message. Wake another + // receiver, and do not yet push the message. + msg.replace(m); + continue; + } else { + // Was async and not a stream, so it did acquire the message. Push the + // message to the queue for it to be received. + chan.queue.push_back(m); + drop(chan); + break; + } + }, + Some((None, signal)) => { + drop(chan); + signal.fire(); + break; // Was sync, so it has acquired the message + }, + } + } + + Ok(()).into() + } else if chan.sending.as_ref().map(|(cap, _)| chan.queue.len() < *cap).unwrap_or(true) { + chan.queue.push_back(msg); + Ok(()).into() + } else if should_block { // Only bounded from here on + let hook = make_signal(msg); + chan.sending.as_mut().unwrap().1.push_back(hook.clone()); + drop(chan); + + do_block(hook) + } else { + Err(TrySendTimeoutError::Full(msg)).into() + } + } + + fn send_sync( + &self, + msg: T, + block: Option<Option<Instant>>, + ) -> Result<(), TrySendTimeoutError<T>> { + self.send( + // msg + msg, + // should_block + block.is_some(), + // make_signal + |msg| Hook::slot(Some(msg), SyncSignal::default()), + // do_block + |hook| if let Some(deadline) = block.unwrap() { + hook.wait_deadline_send(&self.disconnected, deadline) + .or_else(|timed_out| { + if timed_out { // Remove our signal + let hook: Arc<Hook<T, dyn signal::Signal>> = hook.clone(); + wait_lock(&self.chan).sending + .as_mut() + .unwrap().1 + .retain(|s| s.signal().as_ptr() != hook.signal().as_ptr()); + } + hook.try_take().map(|msg| if self.is_disconnected() { + Err(TrySendTimeoutError::Disconnected(msg)) + } else { + Err(TrySendTimeoutError::Timeout(msg)) + }) + .unwrap_or(Ok(())) + }) + } else { + hook.wait_send(&self.disconnected); + + match hook.try_take() { + Some(msg) => Err(TrySendTimeoutError::Disconnected(msg)), + None => Ok(()), + } + }, + ) + } + + fn recv<S: Signal, R: From<Result<T, TryRecvTimeoutError>>>( + &self, + should_block: bool, + make_signal: impl FnOnce() -> Arc<Hook<T, S>>, + do_block: impl FnOnce(Arc<Hook<T, S>>) -> R, + ) -> R { + let mut chan = wait_lock(&self.chan); + chan.pull_pending(true); + + if let Some(msg) = chan.queue.pop_front() { + drop(chan); + Ok(msg).into() + } else if self.is_disconnected() { + drop(chan); + Err(TryRecvTimeoutError::Disconnected).into() + } else if should_block { + let hook = make_signal(); + chan.waiting.push_back(hook.clone()); + drop(chan); + + do_block(hook) + } else { + drop(chan); + Err(TryRecvTimeoutError::Empty).into() + } + } + + fn recv_sync(&self, block: Option<Option<Instant>>) -> Result<T, TryRecvTimeoutError> { + self.recv( + // should_block + block.is_some(), + // make_signal + || Hook::slot(None, SyncSignal::default()), + // do_block + |hook| if let Some(deadline) = block.unwrap() { + hook.wait_deadline_recv(&self.disconnected, deadline) + .or_else(|timed_out| { + if timed_out { // Remove our signal + let hook: Arc<Hook<T, dyn Signal>> = hook.clone(); + wait_lock(&self.chan).waiting + .retain(|s| s.signal().as_ptr() != hook.signal().as_ptr()); + } + match hook.try_take() { + Some(msg) => Ok(msg), + None => { + let disconnected = self.is_disconnected(); // Check disconnect *before* msg + if let Some(msg) = wait_lock(&self.chan).queue.pop_front() { + Ok(msg) + } else if disconnected { + Err(TryRecvTimeoutError::Disconnected) + } else { + Err(TryRecvTimeoutError::Timeout) + } + }, + } + }) + } else { + hook.wait_recv(&self.disconnected) + .or_else(|| wait_lock(&self.chan).queue.pop_front()) + .ok_or(TryRecvTimeoutError::Disconnected) + }, + ) + } + + /// Disconnect anything listening on this channel (this will not prevent receivers receiving + /// msgs that have already been sent) + fn disconnect_all(&self) { + self.disconnected.store(true, Ordering::Relaxed); + + let mut chan = wait_lock(&self.chan); + chan.pull_pending(false); + if let Some((_, sending)) = chan.sending.as_ref() { + sending.iter().for_each(|hook| { + hook.signal().fire(); + }) + } + chan.waiting.iter().for_each(|hook| { + hook.signal().fire(); + }); + } + + fn is_disconnected(&self) -> bool { + self.disconnected.load(Ordering::SeqCst) + } + + fn is_empty(&self) -> bool { + self.len() == 0 + } + + fn is_full(&self) -> bool { + self.capacity().map(|cap| cap == self.len()).unwrap_or(false) + } + + fn len(&self) -> usize { + let mut chan = wait_lock(&self.chan); + chan.pull_pending(false); + chan.queue.len() + } + + fn capacity(&self) -> Option<usize> { + wait_lock(&self.chan).sending.as_ref().map(|(cap, _)| *cap) + } + + fn sender_count(&self) -> usize { + self.sender_count.load(Ordering::Relaxed) + } + + fn receiver_count(&self) -> usize { + self.receiver_count.load(Ordering::Relaxed) + } +} + +/// A transmitting end of a channel. +pub struct Sender<T> { + shared: Arc<Shared<T>>, +} + +impl<T> Sender<T> { + /// Attempt to send a value into the channel. If the channel is bounded and full, or all + /// receivers have been dropped, an error is returned. If the channel associated with this + /// sender is unbounded, this method has the same behaviour as [`Sender::send`]. + pub fn try_send(&self, msg: T) -> Result<(), TrySendError<T>> { + self.shared.send_sync(msg, None).map_err(|err| match err { + TrySendTimeoutError::Full(msg) => TrySendError::Full(msg), + TrySendTimeoutError::Disconnected(msg) => TrySendError::Disconnected(msg), + _ => unreachable!(), + }) + } + + /// Send a value into the channel, returning an error if all receivers have been dropped. + /// If the channel is bounded and is full, this method will block until space is available + /// or all receivers have been dropped. If the channel is unbounded, this method will not + /// block. + pub fn send(&self, msg: T) -> Result<(), SendError<T>> { + self.shared.send_sync(msg, Some(None)).map_err(|err| match err { + TrySendTimeoutError::Disconnected(msg) => SendError(msg), + _ => unreachable!(), + }) + } + + /// Send a value into the channel, returning an error if all receivers have been dropped + /// or the deadline has passed. If the channel is bounded and is full, this method will + /// block until space is available, the deadline is reached, or all receivers have been + /// dropped. + pub fn send_deadline(&self, msg: T, deadline: Instant) -> Result<(), SendTimeoutError<T>> { + self.shared.send_sync(msg, Some(Some(deadline))).map_err(|err| match err { + TrySendTimeoutError::Disconnected(msg) => SendTimeoutError::Disconnected(msg), + TrySendTimeoutError::Timeout(msg) => SendTimeoutError::Timeout(msg), + _ => unreachable!(), + }) + } + + /// Send a value into the channel, returning an error if all receivers have been dropped + /// or the timeout has expired. If the channel is bounded and is full, this method will + /// block until space is available, the timeout has expired, or all receivers have been + /// dropped. + pub fn send_timeout(&self, msg: T, dur: Duration) -> Result<(), SendTimeoutError<T>> { + self.send_deadline(msg, Instant::now().checked_add(dur).unwrap()) + } + + /// Returns true if all receivers for this channel have been dropped. + pub fn is_disconnected(&self) -> bool { + self.shared.is_disconnected() + } + + /// Returns true if the channel is empty. + /// Note: Zero-capacity channels are always empty. + pub fn is_empty(&self) -> bool { + self.shared.is_empty() + } + + /// Returns true if the channel is full. + /// Note: Zero-capacity channels are always full. + pub fn is_full(&self) -> bool { + self.shared.is_full() + } + + /// Returns the number of messages in the channel + pub fn len(&self) -> usize { + self.shared.len() + } + + /// If the channel is bounded, returns its capacity. + pub fn capacity(&self) -> Option<usize> { + self.shared.capacity() + } + + /// Get the number of senders that currently exist, including this one. + pub fn sender_count(&self) -> usize { + self.shared.sender_count() + } + + /// Get the number of receivers that currently exist. + /// + /// Note that this method makes no guarantees that a subsequent send will succeed; it's + /// possible that between `receiver_count()` being called and a `send()`, all open receivers + /// could drop. + pub fn receiver_count(&self) -> usize { + self.shared.receiver_count() + } + + /// Creates a [`WeakSender`] that does not keep the channel open. + /// + /// The channel is closed once all `Sender`s are dropped, even if there + /// are still active `WeakSender`s. + pub fn downgrade(&self) -> WeakSender<T> { + WeakSender { + shared: Arc::downgrade(&self.shared), + } + } + + /// Returns whether the senders are belong to the same channel. + pub fn same_channel(&self, other: &Sender<T>) -> bool { + Arc::ptr_eq(&self.shared, &other.shared) + } +} + +impl<T> Clone for Sender<T> { + /// Clone this sender. [`Sender`] acts as a handle to the ending a channel. Remaining channel + /// contents will only be cleaned up when all senders and the receiver have been dropped. + fn clone(&self) -> Self { + self.shared.sender_count.fetch_add(1, Ordering::Relaxed); + Self { shared: self.shared.clone() } + } +} + +impl<T> fmt::Debug for Sender<T> { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.debug_struct("Sender").finish() + } +} + +impl<T> Drop for Sender<T> { + fn drop(&mut self) { + // Notify receivers that all senders have been dropped if the number of senders drops to 0. + if self.shared.sender_count.fetch_sub(1, Ordering::Relaxed) == 1 { + self.shared.disconnect_all(); + } + } +} + +/// A sender that does not prevent the channel from being closed. +/// +/// Weak senders do not count towards the number of active senders on the channel. As soon as +/// all normal [`Sender`]s are dropped, the channel is closed, even if there is still a +/// `WeakSender`. +/// +/// To send messages, a `WeakSender` must first be upgraded to a `Sender` using the [`upgrade`] +/// method. +pub struct WeakSender<T> { + shared: Weak<Shared<T>>, +} + +impl<T> WeakSender<T> { + /// Tries to upgrade the `WeakSender` to a [`Sender`], in order to send messages. + /// + /// Returns `None` if the channel was closed already. Note that a `Some` return value + /// does not guarantee that the channel is still open. + pub fn upgrade(&self) -> Option<Sender<T>> { + self.shared + .upgrade() + // check that there are still live senders + .filter(|shared| { + shared + .sender_count + .fetch_update(Ordering::Relaxed, Ordering::Relaxed, |count| { + if count == 0 { + // all senders are closed already -> don't increase the sender count + None + } else { + // there is still at least one active sender + Some(count + 1) + } + }) + .is_ok() + }) + .map(|shared| Sender { shared }) + } +} + +/// The receiving end of a channel. +/// +/// Note: Cloning the receiver *does not* turn this channel into a broadcast channel. +/// Each message will only be received by a single receiver. This is useful for +/// implementing work stealing for concurrent programs. +pub struct Receiver<T> { + shared: Arc<Shared<T>>, +} + +impl<T> Receiver<T> { + /// Attempt to fetch an incoming value from the channel associated with this receiver, + /// returning an error if the channel is empty or if all senders have been dropped. + pub fn try_recv(&self) -> Result<T, TryRecvError> { + self.shared.recv_sync(None).map_err(|err| match err { + TryRecvTimeoutError::Disconnected => TryRecvError::Disconnected, + TryRecvTimeoutError::Empty => TryRecvError::Empty, + _ => unreachable!(), + }) + } + + /// Wait for an incoming value from the channel associated with this receiver, returning an + /// error if all senders have been dropped. + pub fn recv(&self) -> Result<T, RecvError> { + self.shared.recv_sync(Some(None)).map_err(|err| match err { + TryRecvTimeoutError::Disconnected => RecvError::Disconnected, + _ => unreachable!(), + }) + } + + /// Wait for an incoming value from the channel associated with this receiver, returning an + /// error if all senders have been dropped or the deadline has passed. + pub fn recv_deadline(&self, deadline: Instant) -> Result<T, RecvTimeoutError> { + self.shared.recv_sync(Some(Some(deadline))).map_err(|err| match err { + TryRecvTimeoutError::Disconnected => RecvTimeoutError::Disconnected, + TryRecvTimeoutError::Timeout => RecvTimeoutError::Timeout, + _ => unreachable!(), + }) + } + + /// Wait for an incoming value from the channel associated with this receiver, returning an + /// error if all senders have been dropped or the timeout has expired. + pub fn recv_timeout(&self, dur: Duration) -> Result<T, RecvTimeoutError> { + self.recv_deadline(Instant::now().checked_add(dur).unwrap()) + } + + /// Create a blocking iterator over the values received on the channel that finishes iteration + /// when all senders have been dropped. + /// + /// You can also create a self-owned iterator with [`Receiver::into_iter`]. + pub fn iter(&self) -> Iter<T> { + Iter { receiver: &self } + } + + /// A non-blocking iterator over the values received on the channel that finishes iteration + /// when all senders have been dropped or the channel is empty. + pub fn try_iter(&self) -> TryIter<T> { + TryIter { receiver: &self } + } + + /// Take all msgs currently sitting in the channel and produce an iterator over them. Unlike + /// `try_iter`, the iterator will not attempt to fetch any more values from the channel once + /// the function has been called. + pub fn drain(&self) -> Drain<T> { + let mut chan = wait_lock(&self.shared.chan); + chan.pull_pending(false); + let queue = std::mem::take(&mut chan.queue); + + Drain { queue, _phantom: PhantomData } + } + + /// Returns true if all senders for this channel have been dropped. + pub fn is_disconnected(&self) -> bool { + self.shared.is_disconnected() + } + + /// Returns true if the channel is empty. + /// Note: Zero-capacity channels are always empty. + pub fn is_empty(&self) -> bool { + self.shared.is_empty() + } + + /// Returns true if the channel is full. + /// Note: Zero-capacity channels are always full. + pub fn is_full(&self) -> bool { + self.shared.is_full() + } + + /// Returns the number of messages in the channel. + pub fn len(&self) -> usize { + self.shared.len() + } + + /// If the channel is bounded, returns its capacity. + pub fn capacity(&self) -> Option<usize> { + self.shared.capacity() + } + + /// Get the number of senders that currently exist. + pub fn sender_count(&self) -> usize { + self.shared.sender_count() + } + + /// Get the number of receivers that currently exist, including this one. + pub fn receiver_count(&self) -> usize { + self.shared.receiver_count() + } + + /// Returns whether the receivers are belong to the same channel. + pub fn same_channel(&self, other: &Receiver<T>) -> bool { + Arc::ptr_eq(&self.shared, &other.shared) + } +} + +impl<T> Clone for Receiver<T> { + /// Clone this receiver. [`Receiver`] acts as a handle to the ending a channel. Remaining + /// channel contents will only be cleaned up when all senders and the receiver have been + /// dropped. + /// + /// Note: Cloning the receiver *does not* turn this channel into a broadcast channel. + /// Each message will only be received by a single receiver. This is useful for + /// implementing work stealing for concurrent programs. + fn clone(&self) -> Self { + self.shared.receiver_count.fetch_add(1, Ordering::Relaxed); + Self { shared: self.shared.clone() } + } +} + +impl<T> fmt::Debug for Receiver<T> { + fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result { + f.debug_struct("Receiver").finish() + } +} + +impl<T> Drop for Receiver<T> { + fn drop(&mut self) { + // Notify senders that all receivers have been dropped if the number of receivers drops + // to 0. + if self.shared.receiver_count.fetch_sub(1, Ordering::Relaxed) == 1 { + self.shared.disconnect_all(); + } + } +} + +/// This exists as a shorthand for [`Receiver::iter`]. +impl<'a, T> IntoIterator for &'a Receiver<T> { + type Item = T; + type IntoIter = Iter<'a, T>; + + fn into_iter(self) -> Self::IntoIter { + Iter { receiver: self } + } +} + +impl<T> IntoIterator for Receiver<T> { + type Item = T; + type IntoIter = IntoIter<T>; + + /// Creates a self-owned but semantically equivalent alternative to [`Receiver::iter`]. + fn into_iter(self) -> Self::IntoIter { + IntoIter { receiver: self } + } +} + +/// An iterator over the msgs received from a channel. +pub struct Iter<'a, T> { + receiver: &'a Receiver<T>, +} + +impl<'a, T> Iterator for Iter<'a, T> { + type Item = T; + + fn next(&mut self) -> Option<Self::Item> { + self.receiver.recv().ok() + } +} + +/// An non-blocking iterator over the msgs received from a channel. +pub struct TryIter<'a, T> { + receiver: &'a Receiver<T>, +} + +impl<'a, T> Iterator for TryIter<'a, T> { + type Item = T; + + fn next(&mut self) -> Option<Self::Item> { + self.receiver.try_recv().ok() + } +} + +/// An fixed-sized iterator over the msgs drained from a channel. +#[derive(Debug)] +pub struct Drain<'a, T> { + queue: VecDeque<T>, + /// A phantom field used to constrain the lifetime of this iterator. We do this because the + /// implementation may change and we don't want to unintentionally constrain it. Removing this + /// lifetime later is a possibility. + _phantom: PhantomData<&'a ()>, +} + +impl<'a, T> Iterator for Drain<'a, T> { + type Item = T; + + fn next(&mut self) -> Option<Self::Item> { + self.queue.pop_front() + } +} + +impl<'a, T> ExactSizeIterator for Drain<'a, T> { + fn len(&self) -> usize { + self.queue.len() + } +} + +/// An owned iterator over the msgs received from a channel. +pub struct IntoIter<T> { + receiver: Receiver<T>, +} + +impl<T> Iterator for IntoIter<T> { + type Item = T; + + fn next(&mut self) -> Option<Self::Item> { + self.receiver.recv().ok() + } +} + +/// Create a channel with no maximum capacity. +/// +/// Create an unbounded channel with a [`Sender`] and [`Receiver`] connected to each end respectively. Values sent in +/// one end of the channel will be received on the other end. The channel is thread-safe, and both [`Sender`] and +/// [`Receiver`] may be sent to or shared between threads as necessary. In addition, both [`Sender`] and [`Receiver`] +/// may be cloned. +/// +/// # Examples +/// ``` +/// let (tx, rx) = flume::unbounded(); +/// +/// tx.send(42).unwrap(); +/// assert_eq!(rx.recv().unwrap(), 42); +/// ``` +pub fn unbounded<T>() -> (Sender<T>, Receiver<T>) { + let shared = Arc::new(Shared::new(None)); + ( + Sender { shared: shared.clone() }, + Receiver { shared }, + ) +} + +/// Create a channel with a maximum capacity. +/// +/// Create a bounded channel with a [`Sender`] and [`Receiver`] connected to each end respectively. Values sent in one +/// end of the channel will be received on the other end. The channel is thread-safe, and both [`Sender`] and +/// [`Receiver`] may be sent to or shared between threads as necessary. In addition, both [`Sender`] and [`Receiver`] +/// may be cloned. +/// +/// Unlike an [`unbounded`] channel, if there is no space left for new messages, calls to +/// [`Sender::send`] will block (unblocking once a receiver has made space). If blocking behaviour +/// is not desired, [`Sender::try_send`] may be used. +/// +/// Like `std::sync::mpsc`, `flume` supports 'rendezvous' channels. A bounded queue with a maximum capacity of zero +/// will block senders until a receiver is available to take the value. You can imagine a rendezvous channel as a +/// ['Glienicke Bridge'](https://en.wikipedia.org/wiki/Glienicke_Bridge)-style location at which senders and receivers +/// perform a handshake and transfer ownership of a value. +/// +/// # Examples +/// ``` +/// let (tx, rx) = flume::bounded(32); +/// +/// for i in 1..33 { +/// tx.send(i).unwrap(); +/// } +/// assert!(tx.try_send(33).is_err()); +/// +/// assert_eq!(rx.try_iter().sum::<u32>(), (1..33).sum()); +/// ``` +pub fn bounded<T>(cap: usize) -> (Sender<T>, Receiver<T>) { + let shared = Arc::new(Shared::new(Some(cap))); + ( + Sender { shared: shared.clone() }, + Receiver { shared }, + ) +} |