1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
use crate::tokio::{
    self,
    runtime::Runtime,
    sync::mpsc::{channel, Receiver, Sender},
    time::timeout,
};
use crate::{
    error::Error,
    parser,
    relay::{CtrlSignal, ProcessorRelay, RelayMessage, WorkerRelay},
    router::SenderPair,
    Cancel, NodeMessage, ShutdownType,
};
use core::time::Duration;
use ockam_core::compat::{string::String, sync::Arc, vec::Vec};
use ockam_core::{
    Address, AddressSet, LocalMessage, Message, Processor, Result, Route, TransportMessage, Worker,
};

/// A default timeout in seconds
pub const DEFAULT_TIMEOUT: u64 = 30;

enum AddressType {
    Worker,
    Processor,
}

impl AddressType {
    fn str(&self) -> &'static str {
        match self {
            AddressType::Worker => "worker",
            AddressType::Processor => "processor",
        }
    }
}

/// Context contains Node state and references to the runtime.
pub struct Context {
    address: AddressSet,
    sender: Sender<NodeMessage>,
    rt: Arc<Runtime>,
    mailbox: Receiver<RelayMessage>,
}

impl Context {
    /// Return runtime clone
    pub fn runtime(&self) -> Arc<Runtime> {
        self.rt.clone()
    }
    /// Wait for the next message from the mailbox
    pub(crate) async fn mailbox_next(&mut self) -> Option<RelayMessage> {
        self.mailbox.recv().await.map(|msg| {
            trace!("{}: received new message!", self.address());
            msg
        })
    }
}

impl Context {
    /// Create a new context
    ///
    /// This function returns a new instance of Context, the relay
    /// sender pair, and relay control signal receiver.
    pub(crate) fn new(
        rt: Arc<Runtime>,
        sender: Sender<NodeMessage>,
        address: AddressSet,
    ) -> (Self, SenderPair, Receiver<CtrlSignal>) {
        let (mailbox_tx, mailbox) = channel(32);
        let (ctrl_tx, ctrl_rx) = channel(1);
        (
            Self {
                rt,
                sender,
                address,
                mailbox,
            },
            SenderPair {
                msgs: mailbox_tx,
                ctrl: ctrl_tx,
            },
            ctrl_rx,
        )
    }

    /// Return the primary worker address
    pub fn address(&self) -> Address {
        self.address.first()
    }

    /// Return all addresses of this worker
    pub fn aliases(&self) -> AddressSet {
        self.address.clone().into_iter().skip(1).collect()
    }

    /// Utility function to sleep tasks from other crates
    #[doc(hidden)]
    pub async fn sleep(&self, dur: Duration) {
        tokio::time::sleep(dur).await;
    }

    /// Create a new context without spawning a full worker
    pub async fn new_context<S: Into<Address>>(&self, addr: S) -> Result<Context> {
        self.new_context_impl(addr.into()).await
    }

    async fn new_context_impl(&self, addr: Address) -> Result<Context> {
        // Create a new context and get access to the mailbox senders
        let (ctx, sender, _) = Self::new(
            Arc::clone(&self.rt),
            self.sender.clone(),
            addr.clone().into(),
        );

        // Create a "bare relay" and register it with the router
        let (msg, mut rx) = NodeMessage::start_worker(addr.into(), sender, true);
        self.sender
            .send(msg)
            .await
            .map_err(|_| Error::FailedStartWorker)?;

        Ok(rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)?
            .map(|_| ctx)?)
    }

    /// Start a new worker handle at [`Address`](ockam_core::Address)
    pub async fn start_worker<NM, NW, S>(&self, address: S, worker: NW) -> Result<()>
    where
        S: Into<AddressSet>,
        NM: Message + Send + 'static,
        NW: Worker<Context = Context, Message = NM>,
    {
        self.start_worker_impl(address.into(), worker).await
    }

    async fn start_worker_impl<NM, NW>(&self, address: AddressSet, worker: NW) -> Result<()>
    where
        NM: Message + Send + 'static,
        NW: Worker<Context = Context, Message = NM>,
    {
        // Pass it to the context
        let (ctx, sender, ctrl_rx) =
            Context::new(self.rt.clone(), self.sender.clone(), address.clone());

        // Then initialise the worker message relay
        WorkerRelay::<NW, NM>::init(self.rt.as_ref(), worker, ctx, ctrl_rx);

        // Send start request to router
        let (msg, mut rx) = NodeMessage::start_worker(address, sender, false);
        self.sender
            .send(msg)
            .await
            .map_err(|_| Error::FailedStartWorker)?;

        // Wait for the actual return code
        Ok(rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)?
            .map(|_| ())?)
    }

    /// Start a new processor at [`Address`](ockam_core::Address)
    pub async fn start_processor<P>(&self, address: impl Into<Address>, processor: P) -> Result<()>
    where
        P: Processor<Context = Context>,
    {
        self.start_processor_impl(address.into(), processor).await
    }

    async fn start_processor_impl<P>(&self, address: Address, processor: P) -> Result<()>
    where
        P: Processor<Context = Context>,
    {
        let addr = address.clone();

        let (ctx, senders, ctrl_rx) =
            Context::new(self.rt.clone(), self.sender.clone(), addr.into());

        // Initialise the processor relay with the ctrl receiver
        ProcessorRelay::<P>::init(self.rt.as_ref(), processor, ctx, ctrl_rx);

        // Send start request to router
        let (msg, mut rx) = NodeMessage::start_processor(address, senders);
        self.sender
            .send(msg)
            .await
            .map_err(|_| Error::FailedStartProcessor)?;

        // Wait for the actual return code
        Ok(rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)?
            .map(|_| ())?)
    }

    /// Shut down a worker by its primary address
    pub async fn stop_worker<A: Into<Address>>(&self, addr: A) -> Result<()> {
        self.stop_address(addr.into(), AddressType::Worker).await
    }

    /// Shut down a processor by its address
    pub async fn stop_processor<A: Into<Address>>(&self, addr: A) -> Result<()> {
        self.stop_address(addr.into(), AddressType::Processor).await
    }

    async fn stop_address(&self, addr: Address, t: AddressType) -> Result<()> {
        debug!("Shutting down {} {}", t.str(), addr);

        // Send the stop request
        let (req, mut rx) = match t {
            AddressType::Worker => NodeMessage::stop_worker(addr),
            AddressType::Processor => NodeMessage::stop_processor(addr),
        };
        self.sender.send(req).await.map_err(Error::from)?;

        // Then check that address was properly shut down
        Ok(rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)?
            .map(|_| ())?)
    }

    /// Signal to the local runtime to shut down immediately
    ///
    /// **WARNING**: calling this function may result in data loss.
    /// It is recommended to use the much safer
    /// [`Context::stop`](Context::stop) function instead!
    pub async fn stop_now(&mut self) -> Result<()> {
        let tx = self.sender.clone();
        info!("Immediately shutting down all workers");
        let (msg, _) = NodeMessage::stop_node(ShutdownType::Immediate);

        match tx.send(msg).await {
            Ok(()) => Ok(()),
            Err(_e) => Err(Error::FailedStopNode.into()),
        }
    }

    /// Signal to the local runtime to shut down
    ///
    /// This call will hang until a safe shutdown has been completed.
    /// The default timeout for a safe shutdown is 1 second.  You can
    /// change this behaviour by calling
    /// [`Context::stop_timeout`](Context::stop_timeout) directly.
    pub async fn stop(&mut self) -> Result<()> {
        self.stop_timeout(1).await
    }

    /// Signal to the local runtime to shut down
    ///
    /// This call will hang until a safe shutdown has been completed
    /// or the desired timeout has been reached.
    pub async fn stop_timeout(&mut self, seconds: u8) -> Result<()> {
        let (req, mut rx) = NodeMessage::stop_node(ShutdownType::Graceful(seconds));
        self.sender.send(req).await.map_err(Error::from)?;

        // Wait until we get the all-clear
        Ok(rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)?
            .map(|_| ())?)
    }

    /// Send a message via a fully qualified route
    ///
    /// Routes can be constructed from a set of [`Address`]es, or via
    /// the [`RouteBuilder`] type.  Routes can contain middleware
    /// router addresses, which will re-address messages that need to
    /// be handled by specific domain workers.
    ///
    /// [`Address`]: ockam_core::Address
    /// [`RouteBuilder`]: ockam_core::RouteBuilder
    pub async fn send<R, M>(&self, route: R, msg: M) -> Result<()>
    where
        R: Into<Route>,
        M: Message + Send + 'static,
    {
        self.send_from_address(route.into(), msg, self.address())
            .await
    }

    /// Send a message via a fully qualified route using specific Worker address
    ///
    /// Routes can be constructed from a set of [`Address`]es, or via
    /// the [`RouteBuilder`] type.  Routes can contain middleware
    /// router addresses, which will re-address messages that need to
    /// be handled by specific domain workers.
    ///
    /// [`Address`]: ockam_core::Address
    /// [`RouteBuilder`]: ockam_core::RouteBuilder
    pub async fn send_from_address<R, M>(
        &self,
        route: R,
        msg: M,
        sending_address: Address,
    ) -> Result<()>
    where
        R: Into<Route>,
        M: Message + Send + 'static,
    {
        self.send_from_address_impl(route.into(), msg, sending_address)
            .await
    }

    async fn send_from_address_impl<M>(
        &self,
        route: Route,
        msg: M,
        sending_address: Address,
    ) -> Result<()>
    where
        M: Message + Send + 'static,
    {
        if !self.address.as_ref().contains(&sending_address) {
            return Err(Error::SenderAddressDoesNotExist.into());
        }

        let (reply_tx, mut reply_rx) = channel(1);
        let next = route.next().unwrap(); // TODO: communicate bad routes
        let req = NodeMessage::SenderReq(next.clone(), reply_tx);

        // First resolve the next hop in the route
        self.sender.send(req).await.map_err(Error::from)?;
        let (addr, sender, needs_wrapping) = reply_rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)??
            .take_sender()?;

        // Pack the payload into a TransportMessage
        let payload = msg.encode().unwrap();
        let mut transport_msg = TransportMessage::v1(route.clone(), Route::new(), payload);
        transport_msg.return_route.modify().append(sending_address);
        let local_msg = LocalMessage::new(transport_msg, Vec::new());

        // Pack transport message into relay message wrapper
        let msg = if needs_wrapping {
            RelayMessage::pre_router(addr, local_msg, route)
        } else {
            RelayMessage::direct(addr, local_msg, route)
        };

        // Send the packed user message with associated route
        sender.send(msg).await.map_err(Error::from)?;

        Ok(())
    }

    /// Forward a transport message to its next routing destination
    ///
    /// Similar to [`Context::send`], but taking a
    /// [`TransportMessage`], which contains the full destination
    /// route, and calculated return route for this hop.
    ///
    /// **Note:** you most likely want to use
    /// [`Context::send`] instead, unless you are writing an
    /// external router implementation for ockam node.
    ///
    /// [`Context::send`]: crate::Context::send
    /// [`TransportMessage`]: ockam_core::TransportMessage
    pub async fn forward(&self, local_msg: LocalMessage) -> Result<()> {
        // Resolve the sender for the next hop in the messages route
        let (reply_tx, mut reply_rx) = channel(1);
        let next = local_msg.transport().onward_route.next().unwrap(); // TODO: communicate bad routes
        let req = NodeMessage::SenderReq(next.clone(), reply_tx);

        // First resolve the next hop in the route
        self.sender.send(req).await.map_err(Error::from)?;
        let (addr, sender, needs_wrapping) = reply_rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)??
            .take_sender()?;

        // Pack the transport message into a relay message
        let onward = local_msg.transport().onward_route.clone();
        // let msg = RelayMessage::direct(addr, data, onward);
        let msg = if needs_wrapping {
            RelayMessage::pre_router(addr, local_msg, onward)
        } else {
            RelayMessage::direct(addr, local_msg, onward)
        };
        sender.send(msg).await.map_err(Error::from)?;

        Ok(())
    }

    /// Receive a message without a timeout
    pub async fn receive_block<M: Message>(&mut self) -> Result<Cancel<'_, M>> {
        let (msg, data, addr) = self.next_from_mailbox().await?;
        Ok(Cancel::new(msg, data, addr, self))
    }

    /// Block the current worker to wait for a typed message
    ///
    /// This function may return a `Err(FailedLoadData)` if the
    /// underlying worker was shut down, or `Err(Timeout)` if the call
    /// was waiting for longer than the `default timeout`.  Use
    /// [`receive_timeout`](Context::receive_timeout) to adjust the
    /// timeout period.
    ///
    /// Will return `None` if the corresponding worker has been
    /// stopped, or the underlying Node has shut down.
    pub async fn receive<M: Message>(&mut self) -> Result<Cancel<'_, M>> {
        self.receive_timeout(DEFAULT_TIMEOUT).await
    }

    /// Block to wait for a typed message, with explicit timeout
    pub async fn receive_timeout<M: Message>(
        &mut self,
        timeout_secs: u64,
    ) -> Result<Cancel<'_, M>> {
        let (msg, data, addr) = timeout(Duration::from_secs(timeout_secs), async {
            self.next_from_mailbox().await
        })
        .await
        .map_err(Error::from)??;
        Ok(Cancel::new(msg, data, addr, self))
    }

    /// Block the current worker to wait for a message satisfying a conditional
    ///
    /// Will return `Err` if the corresponding worker has been
    /// stopped, or the underlying node has shut down.  This operation
    /// has a [default timeout](DEFAULT_TIMEOUT).
    ///
    /// Internally this function calls `receive` and `.cancel()` in a
    /// loop until a matching message is found.
    pub async fn receive_match<M, F>(&mut self, check: F) -> Result<Cancel<'_, M>>
    where
        M: Message,
        F: Fn(&M) -> bool,
    {
        let (m, data, addr) = timeout(Duration::from_secs(DEFAULT_TIMEOUT), async {
            loop {
                match self.next_from_mailbox().await {
                    Ok((m, data, addr)) if check(&m) => break Ok((m, data, addr)),
                    Ok((_, data, _)) => {
                        // Requeue
                        self.forward(data).await?;
                    }
                    e => break e,
                }
            }
        })
        .await
        .map_err(Error::from)??;

        Ok(Cancel::new(m, data, addr, self))
    }

    /// Assign the current worker to a cluster
    ///
    /// A cluster is a set of workers that should be stopped together
    /// when the node is stopped or parts of the system are reloaded.
    /// **This is not to be confused with supervisors!**
    ///
    /// By adding your worker to a cluster you signal to the runtime
    /// that your worker may be depended on by other workers that
    /// should be stopped first.
    ///
    /// **Your cluster name MUST NOT start with `_internals.` or
    /// `ockam.`!**
    ///
    /// Clusters are de-allocated in reverse order of their
    /// initialisation when the node is stopped.
    pub async fn set_cluster<S: Into<String>>(&self, label: S) -> Result<()> {
        let (msg, mut rx) = NodeMessage::set_cluster(self.address(), label.into());
        self.sender.send(msg).await.map_err(Error::from)?;
        Ok(rx.recv().await.ok_or(Error::InternalIOFailure)??.is_ok()?)
    }

    /// Return a list of all available worker addresses on a node
    pub async fn list_workers(&self) -> Result<Vec<Address>> {
        let (msg, mut reply_rx) = NodeMessage::list_workers();

        self.sender.send(msg).await.map_err(Error::from)?;

        Ok(reply_rx
            .recv()
            .await
            .ok_or(Error::InternalIOFailure)??
            .take_workers()?)
    }

    /// Register a router for a specific address type
    pub async fn register<A: Into<Address>>(&self, type_: u8, addr: A) -> Result<()> {
        self.register_impl(type_, addr.into()).await
    }

    /// Send a shutdown acknowledgement to the router
    pub(crate) async fn send_stop_ack(&self) -> Result<()> {
        self.sender
            .send(NodeMessage::StopAck(self.address()))
            .await
            .map_err(Error::from)?;
        Ok(())
    }

    async fn register_impl(&self, type_: u8, addr: Address) -> Result<()> {
        let (tx, mut rx) = channel(1);
        self.sender
            .send(NodeMessage::Router(type_, addr, tx))
            .await
            .map_err(|_| Error::InternalIOFailure)?;

        Ok(rx.recv().await.ok_or(Error::InternalIOFailure)??.is_ok()?)
    }

    /// A convenience function to get a data 3-tuple from the mailbox
    ///
    /// The reason this function doesn't construct a `Cancel<_, M>` is
    /// to avoid the lifetime collision between the mutation on `self`
    /// and the ref to `Context` passed to `Cancel::new(..)`
    ///
    /// This function will block and re-queue messages into the
    /// mailbox until it can receive the correct message payload.
    ///
    /// WARNING: this will temporarily create a busyloop, this
    /// mechanism should be replaced with a waker system that lets the
    /// mailbox work not yield another message until the relay worker
    /// has woken it.
    async fn next_from_mailbox<M: Message>(&mut self) -> Result<(M, LocalMessage, Address)> {
        loop {
            let msg = self.mailbox_next().await.ok_or(Error::FailedLoadData)?;
            let (addr, data) = msg.local_msg();

            // FIXME: make message parsing idempotent to avoid cloning
            match parser::message(&data.transport().payload).ok() {
                Some(msg) => break Ok((msg, data, addr)),
                None => {
                    // Requeue
                    self.forward(data).await?;
                }
            }
        }
    }
}