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
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
//! Arranges a collection into a re-usable trace structure.
//!
//! The `arrange` operator applies to a differential dataflow `Collection` and returns an `Arranged`
//! structure, provides access to both an indexed form of accepted updates as well as a stream of
//! batches of newly arranged updates.
//!
//! Several operators (`join`, `group`, and `cogroup`, among others) are implemented against `Arranged`,
//! and can be applied directly to arranged data instead of the collection. Internally, the operators
//! will borrow the shared state, and listen on the timely stream for shared batches of data. The
//! resources to index the collection---communication, computation, and memory---are spent only once,
//! and only one copy of the index needs to be maintained as the collection changes.
//!
//! The arranged collection is stored in a trace, whose append-only operation means that it is safe to
//! share between the single `arrange` writer and multiple readers. Each reader is expected to interrogate
//! the trace only at times for which it knows the trace is complete, as indicated by the frontiers on its
//! incoming channels. Failing to do this is "safe" in the Rust sense of memory safety, but the reader may
//! see ill-defined data at times for which the trace is not complete. (All current implementations
//! commit only completed data to the trace).

use std::rc::{Rc, Weak};
use std::cell::RefCell;
use std::default::Default;
use std::ops::DerefMut;
use std::collections::VecDeque;

use timely::dataflow::operators::{Enter, Map};
use timely::order::PartialOrder;
use timely::dataflow::{Scope, Stream};
use timely::dataflow::operators::generic::{Unary, Operator, source};
use timely::dataflow::channels::pact::{Pipeline, Exchange};
use timely::progress::Timestamp;
use timely::progress::frontier::Antichain;
use timely::dataflow::operators::Capability;
use timely::dataflow::scopes::Child;

use timely_sort::Unsigned;

use ::{Data, Diff, Collection, AsCollection, Hashable};
use lattice::Lattice;
use trace::{Trace, TraceReader, Batch, BatchReader, Batcher, Cursor};
// use trace::implementations::hash::HashValSpine as DefaultValTrace;
// use trace::implementations::hash::HashKeySpine as DefaultKeyTrace;
use trace::implementations::ord::OrdValSpine as DefaultValTrace;
use trace::implementations::ord::OrdKeySpine as DefaultKeyTrace;

use trace::wrappers::enter::{TraceEnter, BatchEnter};
use trace::wrappers::rc::TraceBox;

/// Wrapper type to permit transfer of `Rc` types, as in batch.
///
/// The `BatchWrapper`s sole purpose in life is to implement `Abomonation` with methods that panic
/// when called. This allows the wrapped data to be transited along timely's `Pipeline` channels.
/// The wrapper cannot fake out `Send`, and so cannot be used on timely's `Exchange` channels, which
/// is good.
#[derive(Clone,Eq,PartialEq,Debug)]
pub struct BatchWrapper<T> {
    /// The wrapped item.
    pub item: T,
}

// NOTE: This is all horrible. Don't look too hard.
impl<T> ::abomonation::Abomonation for BatchWrapper<T> {
   unsafe fn entomb<W: ::std::io::Write>(&self, _write: &mut W) -> ::std::io::Result<()> { panic!("BatchWrapper Abomonation impl") }
   unsafe fn exhume<'a,'b>(&'a mut self, _bytes: &'b mut [u8]) -> Option<&'b mut [u8]> { panic!("BatchWrapper Abomonation impl")  }
}

/// A trace writer capability.
pub struct TraceWriter<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: Trace<K,V,T,R>, Tr::Batch: Batch<K,V,T,R> {
    phantom: ::std::marker::PhantomData<(K, V, R)>,
    trace: Weak<RefCell<TraceBox<K, V, T, R, Tr>>>,
    queues: Rc<RefCell<Vec<Weak<RefCell<VecDeque<(Vec<T>, Option<(T, Tr::Batch)>)>>>>>>,
}

impl<K, V, T, R, Tr> TraceWriter<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: Trace<K,V,T,R>, Tr::Batch: Batch<K,V,T,R> {

    /// Advances the trace to `frontier`, providing batch data if it exists.
    pub fn seal(&mut self, frontier: &[T], data: Option<(T, Tr::Batch)>) {

        // push information to each listener that still exists.
        let mut borrow = self.queues.borrow_mut();
        for queue in borrow.iter_mut() {
            queue.upgrade().map(|queue| {
                queue.borrow_mut().push_back((frontier.to_vec(), data.clone()));
            });
        }
        borrow.retain(|w| w.upgrade().is_some());

        // push data to the trace, if it still exists.
        if let Some((_time, batch)) = data {
            if let Some(trace) = self.trace.upgrade() {
                trace.borrow_mut().trace.insert(batch);
            }
        }
    }
}

impl<K, V, T, R, Tr> Drop for TraceWriter<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: Trace<K,V,T,R>, Tr::Batch: Batch<K,V,T,R> {
    fn drop(&mut self) {
        let mut borrow = self.queues.borrow_mut();
        for queue in borrow.iter_mut() {
            queue.upgrade().map(|queue| {
                queue.borrow_mut().push_back((Vec::new(), None));
            });
        }
        borrow.retain(|w| w.upgrade().is_some());
    }
}


/// A `TraceReader` wrapper which can be imported into other dataflows.
///
/// The `TraceAgent` is the default trace type produced by `arranged`, and it can be extracted
/// from the dataflow in which it was defined, and imported into other dataflows.
pub struct TraceAgent<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: TraceReader<K,V,T,R> {
    phantom: ::std::marker::PhantomData<(K, V, R)>,
    trace: Rc<RefCell<TraceBox<K, V, T, R, Tr>>>,
    queues: Weak<RefCell<Vec<Weak<RefCell<VecDeque<(Vec<T>, Option<(T, Tr::Batch)>)>>>>>>,
    advance: Vec<T>,
    through: Vec<T>,
}

impl<K, V, T, R, Tr> TraceReader<K, V, T, R> for TraceAgent<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: TraceReader<K,V,T,R> {
    type Batch = Tr::Batch;
    type Cursor = Tr::Cursor;
    fn advance_by(&mut self, frontier: &[T]) {
        self.trace.borrow_mut().adjust_advance_frontier(&self.advance[..], frontier);
        self.advance.clear();
        self.advance.extend(frontier.iter().cloned());
    }
    fn advance_frontier(&mut self) -> &[T] {
        &self.advance[..]
    }
    fn distinguish_since(&mut self, frontier: &[T]) {
        self.trace.borrow_mut().adjust_through_frontier(&self.through[..], frontier);
        self.through.clear();
        self.through.extend(frontier.iter().cloned());
    }
    fn distinguish_frontier(&mut self) -> &[T] {
        &self.through[..]
    }
    fn cursor_through(&mut self, frontier: &[T]) -> Option<(Tr::Cursor, <Tr::Cursor as Cursor<K, V, T, R>>::Storage)> { self.trace.borrow_mut().trace.cursor_through(frontier) }
    fn map_batches<F: FnMut(&Self::Batch)>(&mut self, f: F) { self.trace.borrow_mut().trace.map_batches(f) }
}

impl<K, V, T, R, Tr> TraceAgent<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: TraceReader<K,V,T,R> {

    /// Creates a new agent from a trace reader.
    pub fn new(trace: Tr) -> (Self, TraceWriter<K,V,T,R,Tr>) where Tr: Trace<K,V,T,R>, Tr::Batch: Batch<K,V,T,R> {

        let trace = Rc::new(RefCell::new(TraceBox::new(trace)));
        let queues = Rc::new(RefCell::new(Vec::new()));

        let reader = TraceAgent {
            phantom: ::std::marker::PhantomData,
            trace: trace.clone(),
            queues: Rc::downgrade(&queues),
            advance: trace.borrow().advance_frontiers.frontier().to_vec(),
            through: trace.borrow().through_frontiers.frontier().to_vec(),
        };

        let writer = TraceWriter {
            phantom: ::std::marker::PhantomData,
            trace: Rc::downgrade(&trace),
            queues: queues,
        };

        (reader, writer)
    }

    /// Attaches a new shared queue to the trace.
    ///
    /// The queue will be immediately populated with existing historical batches from the trace, and until the reference
    /// is dropped the queue will receive new batches as produced by the source `arrange` operator.
    pub fn new_listener(&mut self) -> Rc<RefCell<VecDeque<(Vec<T>, Option<(T, <Tr as TraceReader<K,V,T,R>>::Batch)>)>>> where T: Default {

        // create a new queue for progress and batch information.
        let mut new_queue = VecDeque::new();

        // add the existing batches from the trace
        self.trace.borrow_mut().trace.map_batches(|batch| new_queue.push_back((vec![T::default()], Some((T::default(), batch.clone())))));

        let reference = Rc::new(RefCell::new(new_queue));

        // wraps the queue in a ref-counted ref cell and enqueue/return it.
        if let Some(queue) = self.queues.upgrade() {
            let mut borrow = queue.borrow_mut();
            borrow.push(Rc::downgrade(&reference));
        }
        else {
            // if the trace is closed, send a final signal.
            reference.borrow_mut().push_back((Vec::new(), None));
        }

        reference
    }
}

impl<K, V, T, R, Tr> TraceAgent<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: TraceReader<K,V,T,R> {

    /// Copies an existing collection into the supplied scope.
    ///
    /// This method creates an `Arranged` collection that should appear indistinguishable from applying `arrange`
    /// directly to the source collection brought into the local scope. The only caveat is that the initial state
    /// of the collection is its current state, and updates occur from this point forward. The historical changes
    /// the collection experienced in the past are accumulated, and the distinctions from the initial collection
    /// are no longer evident.
    ///
    /// The current behavior is that the introduced collection accumulates updates to some times less or equal
    /// to `self.advance_frontier()`. There is *not* currently a guarantee that the updates are accumulated *to*
    /// the frontier, and the resulting collection history may be weirdly partial until this point. In particular,
    /// the historical collection may move through configurations that did not actually occur, even if eventually
    /// arriving at the correct collection. This is probably a bug; although we get to the right place in the end,
    /// the intermediate computation could do something that the original computation did not, like diverge.
    ///
    /// I would expect the semantics to improve to "updates are advanced to `self.advance_frontier()`", which
    /// means the computation will run as if starting from exactly this frontier. It is not currently clear whose
    /// responsibility this should be (the trace/batch should only reveal these times, or an operator should know
    /// to advance times before using them).
    ///
    /// # Examples
    ///
    /// ```
    /// extern crate timely;
    /// extern crate timely_communication;
    /// extern crate differential_dataflow;
    ///
    /// use timely_communication::Configuration;
    /// use differential_dataflow::input::Input;
    /// use differential_dataflow::operators::arrange::Arrange;
    /// use differential_dataflow::operators::group::GroupArranged;
    /// use differential_dataflow::trace::Trace;
    /// use differential_dataflow::trace::implementations::ord::OrdValSpine;
    /// use differential_dataflow::hashable::OrdWrapper;
    ///
    /// fn main() {
    ///     ::timely::execute(Configuration::Thread, |worker| {
    ///
    ///         // create a first dataflow
    ///         let mut trace = worker.dataflow::<u32,_,_>(|scope| {
    ///             // create input handle and collection.
    ///             scope.new_collection_from(0 .. 10).1
    ///                  .map(|x| (OrdWrapper { item: x }, x))
    ///                  .arrange(OrdValSpine::new())
    ///                  .trace
    ///         });
    ///
    ///         // do some work.
    ///         worker.step();
    ///         worker.step();
    ///
    ///         // create a second dataflow
    ///         worker.dataflow(move |scope| {
    ///             trace.import(scope)
    ///                  .group_arranged(
    ///                      move |_key, src, dst| dst.push((*src[0].0, 1)),
    ///                      OrdValSpine::new()
    ///                  );
    ///         });
    ///
    ///     }).unwrap();
    /// }
    /// ```
    pub fn import<G: Scope<Timestamp=T>>(&mut self, scope: &G) -> Arranged<G, K, V, R, TraceAgent<K, V, T, R, Tr>> where T: Timestamp {

        let queue = self.new_listener();

        let collection = source(scope, "ArrangedSource", move |capability| {

            // capabilities the source maintains.
            let mut capabilities = vec![capability];

            move |output| {

                let mut borrow = queue.borrow_mut();
                while let Some((frontier, sent)) = borrow.pop_front() {
                    // if data are associated, send em!
                    if let Some((time, batch)) = sent {
                        if let Some(cap) = capabilities.iter().find(|c| c.time().less_equal(&time)) {
                            let delayed = cap.delayed(&time);
                            output.session(&delayed).give(BatchWrapper { item: batch });
                        }
                        else {
                            panic!("failed to find capability for {:?} in {:?}", time, capabilities);
                        }
                    }

                    // advance capabilities to look like `frontier`.
                    let mut new_capabilities = Vec::new();
                    for time in frontier.iter() {
                        if let Some(cap) = capabilities.iter().find(|c| c.time().less_equal(&time)) {
                            new_capabilities.push(cap.delayed(&time));
                        }
                        else {
                            panic!("failed to find capability for {:?} in {:?}", time, capabilities);
                        }
                    }
                    // println!("downgrading {:?} -> {:?}", capabilities, new_capabilities);
                    capabilities = new_capabilities;
                }
            }
        });

        Arranged {
            stream: collection,
            trace: self.clone(),
        }
    }
}

impl<K, V, T, R, Tr> Clone for TraceAgent<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: TraceReader<K,V,T,R> {
    fn clone(&self) -> Self {

        // increase counts for wrapped `TraceBox`.
        self.trace.borrow_mut().adjust_advance_frontier(&[], &self.advance[..]);
        self.trace.borrow_mut().adjust_through_frontier(&[], &self.through[..]);

        TraceAgent {
            phantom: ::std::marker::PhantomData,
            trace: self.trace.clone(),
            queues: self.queues.clone(),
            advance: self.advance.clone(),
            through: self.through.clone(),
        }
    }
}


impl<K, V, T, R, Tr> Drop for TraceAgent<K, V, T, R, Tr>
where T: Lattice+Ord+Clone+'static, Tr: TraceReader<K,V,T,R> {
    fn drop(&mut self) {
        // decrement borrow counts to remove all holds
        self.trace.borrow_mut().adjust_advance_frontier(&self.advance[..], &[]);
        self.trace.borrow_mut().adjust_through_frontier(&self.through[..], &[]);
    }
}

/// An arranged collection of `(K,V)` values.
///
/// An `Arranged` allows multiple differential operators to share the resources (communication,
/// computation, memory) required to produce and maintain an indexed representation of a collection.
pub struct Arranged<G: Scope, K, V, R, T> where G::Timestamp: Lattice+Ord, T: TraceReader<K, V, G::Timestamp, R>+Clone {
    /// A stream containing arranged updates.
    ///
    /// This stream contains the same batches of updates the trace itself accepts, so there should
    /// be no additional overhead to receiving these records. The batches can be navigated just as
    /// the batches in the trace, by key and by value.
    pub stream: Stream<G, BatchWrapper<T::Batch>>,
    /// A shared trace, updated by the `Arrange` operator and readable by others.
    pub trace: T,
    // TODO : We might have an `Option<Collection<G, (K, V)>>` here, which `as_collection` sets and
    // returns when invoked, so as to not duplicate work with multiple calls to `as_collection`.
}

impl<G: Scope, K, V, R, T> Clone for Arranged<G, K, V, R, T> 
where G::Timestamp: Lattice+Ord, T: TraceReader<K, V, G::Timestamp, R>+Clone {
    fn clone(&self) -> Self {
        Arranged {
            stream: self.stream.clone(),
            trace: self.trace.clone(),
        }
    }
}

impl<G: Scope, K, V, R, T> Arranged<G, K, V, R, T> where G::Timestamp: Lattice+Ord, T: TraceReader<K, V, G::Timestamp, R>+Clone {

    /// Brings an arranged collection into a nested scope.
    ///
    /// This method produces a proxy trace handle that uses the same backing data, but acts as if the timestamps
    /// have all been extended with an additional coordinate with the default value. The resulting collection does
    /// not vary with the new timestamp coordinate.
    pub fn enter<'a, TInner>(&self, child: &Child<'a, G, TInner>)
        -> Arranged<Child<'a, G, TInner>, K, V, R, TraceEnter<K, V, G::Timestamp, R, T, TInner>>
        where
            T::Batch: Clone,
            K: 'static,
            V: 'static,
            G::Timestamp: Clone+Default+'static,
            TInner: Lattice+Timestamp+Clone+Default+'static,
            R: 'static {

        Arranged {
            stream: self.stream.enter(child).map(|bw| BatchWrapper { item: BatchEnter::make_from(bw.item) }),
            trace: TraceEnter::make_from(self.trace.clone()),
        }
    }

    /// Flattens the stream into a `Collection`.
    ///
    /// The underlying `Stream<G, BatchWrapper<T::Batch>>` is a much more efficient way to access the data,
    /// and this method should only be used when the data need to be transformed or exchanged, rather than
    /// supplied as arguments to an operator using the same key-value structure.
    pub fn as_collection<D: Data, L>(&self, logic: L) -> Collection<G, D, R>
        where
            R: Diff,
            T::Batch: Clone+'static,
            K: Clone, V: Clone,
            L: Fn(&K, &V) -> D+'static,
    {
        self.stream.unary_stream(Pipeline, "AsCollection", move |input, output| {

            input.for_each(|time, data| {
                let mut session = output.session(&time);
                for wrapper in data.drain(..) {
                    let batch = wrapper.item;
                    let mut cursor = batch.cursor();
                    while cursor.key_valid(&batch) {
                        let key: &K = cursor.key(&batch);
                        while cursor.val_valid(&batch) {
                            let val: &V = cursor.val(&batch);
                            cursor.map_times(&batch, |time, diff| {
                                session.give((logic(key, val), time.clone(), diff.clone()));
                            });
                            cursor.step_val(&batch);
                        }
                        cursor.step_key(&batch);
                    }
                }
            });
        })
        .as_collection()
    }
}

/// Report values associated with keys at certain times.
///
/// This method consumes a stream of (key, time) queries and reports the corresponding stream of
/// (key, value, time, diff) accumulations in the `self` trace.
pub fn query<G: Scope, K, V, R, T>(queries: &Stream<G, (K, G::Timestamp)>, mut trace: T) -> Stream<G, (K, V, G::Timestamp, R)> 
where
    K: Data+Hashable,
    V: Data,
    G::Timestamp: Data+Lattice+Ord,
    R: Diff,
    T: TraceReader<K, V, G::Timestamp, R>+Clone+'static,
{
    // release `distinguish_since` capability.
    trace.distinguish_since(&[]);

    let mut trace = Some(trace);
    let mut stash = ::std::collections::HashMap::new();
    let mut frontier = Antichain::new();

    // while the arrangement is already correctly distributed, the query stream may not be.
    let exchange = Exchange::new(move |update: &(K,G::Timestamp)| update.0.hashed().as_u64());

    queries.unary_frontier(exchange, "TraceQuery", move |_capability|
        move |input, output| {

            // drain the query input, stashing requests.
            input.for_each(|capability, data|
                stash.entry(capability)
                     .or_insert(Vec::new())
                     .extend(data.drain(..).map(|(k,t)| (k,t,1)))
            );

            if let Some(ref mut trace) = trace {

                frontier.clear();
                for time in trace.advance_frontier().iter() {
                    frontier.insert(time.clone());
                }

                for (capability, prefixes) in stash.iter_mut() {

                    // defer requests at incomplete times.
                    // NOTE: not all updates may be at complete times, but if this test fails then none of them are.
                    if !frontier.less_equal(capability.time()) {

                        let mut session = output.session(capability);

                        prefixes.sort_by(|x,y| x.0.cmp(&y.0));

                        let (mut cursor, storage) = trace.cursor();

                        for &mut (ref key, ref time, ref mut cnt) in prefixes.iter_mut() {

                            if !frontier.less_equal(time) {
                                cursor.seek_key(&storage, key);
                                if cursor.get_key(&storage) == Some(key) {

                                    while let Some(val) = cursor.get_val(&storage) {
                                        let mut count = R::zero();
                                        cursor.map_times(&storage, |t, d| if t.less_equal(time) { 
                                            count = count + d;
                                        });
                                        if !count.is_zero() {
                                            session.give((key.clone(), val.clone(), time.clone(), count));
                                        }
                                        cursor.step_val(&storage);
                                    }

                                }
                                *cnt = 0;
                            }
                        }

                        prefixes.retain(|ptd| ptd.2 != 0);
                    }
                }
            }

            // drop fully processed capabilities.
            stash.retain(|_,prefixes| !prefixes.is_empty());
            trace.as_mut().map(|trace| trace.advance_by(input.frontier().frontier()));
            if input.frontier().is_empty() && stash.is_empty() {
                trace = None;
            }
        }
    )
}

/// A type that can be arranged into a trace of type `T`.
///
/// This trait is implemented for appropriately typed collections and all traces that might accommodate them, 
/// as well as by arranged data for their corresponding trace type.
pub trait Arrange<G: Scope, K, V, R: Diff, T> 
where
    G::Timestamp: Lattice,
    T: Trace<K, V, G::Timestamp, R>+'static,
    T::Batch: Batch<K, V, G::Timestamp, R> 
    {
    /// Arranges a stream of `(Key, Val)` updates by `Key`. Accepts an empty instance of the trace type.
    ///
    /// This operator arranges a stream of values into a shared trace, whose contents it maintains.
    /// This trace is current for all times marked completed in the output stream, and probing this stream
    /// is the correct way to determine that times in the shared trace are committed.
    fn arrange(&self, empty_trace: T) -> Arranged<G, K, V, R, TraceAgent<K, V, G::Timestamp, R, T>>;
}

impl<G: Scope, K: Data+Hashable, V: Data, R: Diff, T> Arrange<G, K, V, R, T> for Collection<G, (K, V), R> 
where
    G::Timestamp: Lattice+Ord,
    T: Trace<K, V, G::Timestamp, R>+'static,
    T::Batch: Batch<K, V, G::Timestamp, R> {

    fn arrange(&self, empty_trace: T) -> Arranged<G, K, V, R, TraceAgent<K, V, G::Timestamp, R, T>> {

        let (reader, mut writer) = TraceAgent::new(empty_trace);

        // Where we will deposit received updates, and from which we extract batches.
        let mut batcher = <T::Batch as Batch<K,V,G::Timestamp,R>>::Batcher::new();

        // Capabilities for the lower envelope of updates in `batcher`.
        let mut capabilities = Antichain::<Capability<G::Timestamp>>::new();

        // fabricate a data-parallel operator using the `unary_notify` pattern.
        let exchange = Exchange::new(move |update: &((K,V),G::Timestamp,R)| (update.0).0.hashed().as_u64());
        let stream = self.inner.unary_frontier(exchange, "Arrange", move |_capability|
            move |input, output| {

            // As we receive data, we need to (i) stash the data and (ii) keep *enough* capabilities.
            // We don't have to keep all capabilities, but we need to be able to form output messages
            // when we realize that time intervals are complete.

            input.for_each(|cap, data| {
                capabilities.insert(cap);
                batcher.push_batch(data.deref_mut());
            });

            // The frontier may have advanced by multiple elements, which is an issue because
            // timely dataflow currently only allows one capability per message. This means we
            // must pretend to process the frontier advances one element at a time, batching
            // and sending smaller bites than we might have otherwise done.

            // If there is at least one capability no longer in advance of the input frontier ...
            if capabilities.elements().iter().any(|c| !input.frontier().less_equal(c.time())) {

                let mut upper = Antichain::new();   // re-used allocation for sealing batches.

                // For each capability not in advance of the input frontier ...
                for (index, capability) in capabilities.elements().iter().enumerate() {

                    if !input.frontier().less_equal(capability.time()) {

                        // Assemble the upper bound on times we can commit with this capabilities.
                        // We must respect the input frontier, and *subsequent* capabilities, as
                        // we are pretending to retire the capability changes one by one.
                        upper.clear();
                        for time in input.frontier().frontier().iter() {
                            upper.insert(time.clone());
                        }
                        for other_capability in &capabilities.elements()[(index + 1) .. ] {
                            upper.insert(other_capability.time().clone());
                        }

                        // Extract updates not in advance of `upper`.
                        let batch = batcher.seal(upper.elements());

                        writer.seal(upper.elements(), Some((capability.time().clone(), batch.clone())));

                        // send the batch to downstream consumers, empty or not.
                        output.session(&capabilities.elements()[index]).give(BatchWrapper { item: batch });
                    }
                }

                // Having extracted and sent batches between each capability and the input frontier,
                // we should downgrade all capabilities to match the batcher's lower update frontier.
                // This may involve discarding capabilities, which is fine as any new updates arrive
                // in messages with new capabilities.

                let mut new_capabilities = Antichain::new();
                for time in batcher.frontier() {
                    if let Some(capability) = capabilities.elements().iter().find(|c| c.time().less_equal(time)) {
                        new_capabilities.insert(capability.delayed(time));
                    }
                    else {
                        panic!("failed to find capability");
                    }
                }

                capabilities = new_capabilities;
            }

            // Announce progress updates.
            // TODO: This is very noisy; consider tracking the previous frontier, and issuing an update
            //       if and when it changes.
            writer.seal(input.frontier().frontier(), None);
        });

        Arranged { stream: stream, trace: reader }
    }
}

impl<G: Scope, K: Data+Hashable, R: Diff, T> Arrange<G, K, (), R, T> for Collection<G, K, R>
where
    G::Timestamp: Lattice+Ord,
    T: Trace<K, (), G::Timestamp, R>+'static,
    T::Batch: Batch<K, (), G::Timestamp, R> {

    fn arrange(&self, empty_trace: T) -> Arranged<G, K, (), R, TraceAgent<K, (), G::Timestamp, R, T>> {
        self.map(|k| (k, ()))
            .arrange(empty_trace)
    }
}

impl<G, K, V, R, T> Arrange<G, K, V, R, T> for Arranged<G, K, V, R, TraceAgent<K, V, G::Timestamp, R, T>>
where 
    G: Scope,
    G::Timestamp: Lattice,
    R: Diff,
    T: Trace<K, V, G::Timestamp, R>+Clone+'static,
    T::Batch: Batch<K, V, G::Timestamp, R> 
{
    fn arrange(&self, _: T) -> Arranged<G, K, V, R, TraceAgent<K, V, G::Timestamp, R, T>> {
        (*self).clone()
    }
}

/// Arranges something as `(Key,Val)` pairs according to a type `T` of trace.
///
/// This arrangement requires `Key: Hashable`, and uses the `hashed()` method to place keys in a hashed
/// map. This can result in many hash calls, and in some cases it may help to first transform `K` to the
/// pair `(u64, K)` of hash value and key.
pub trait ArrangeByKey<G: Scope, K: Data+Hashable, V: Data, R: Diff>
where G::Timestamp: Lattice+Ord {
    /// Arranges a collection of `(Key, Val)` records by `Key`.
    ///
    /// This operator arranges a stream of values into a shared trace, whose contents it maintains.
    /// This trace is current for all times completed by the output stream, which can be used to
    /// safely identify the stable times and values in the trace.
    fn arrange_by_key(&self) -> Arranged<G, K, V, R, TraceAgent<K, V, G::Timestamp, R, DefaultValTrace<K, V, G::Timestamp, R>>>;
}

impl<G: Scope, K: Data+Hashable, V: Data, R: Diff> ArrangeByKey<G, K, V, R> for Collection<G, (K,V), R>
where G::Timestamp: Lattice+Ord {
    fn arrange_by_key(&self) -> Arranged<G, K, V, R, TraceAgent<K, V, G::Timestamp, R, DefaultValTrace<K, V, G::Timestamp, R>>> {
        self.arrange(DefaultValTrace::new())
    }
}

/// Arranges something as `(Key, ())` pairs according to a type `T` of trace.
///
/// This arrangement requires `Key: Hashable`, and uses the `hashed()` method to place keys in a hashed
/// map. This can result in many hash calls, and in some cases it may help to first transform `K` to the
/// pair `(u64, K)` of hash value and key.
pub trait ArrangeBySelf<G: Scope, K: Data+Hashable, R: Diff>
where G::Timestamp: Lattice+Ord {
    /// Arranges a collection of `Key` records by `Key`.
    ///
    /// This operator arranges a collection of records into a shared trace, whose contents it maintains.
    /// This trace is current for all times complete in the output stream, which can be used to safely
    /// identify the stable times and values in the trace.
    fn arrange_by_self(&self) -> Arranged<G, K, (), R, TraceAgent<K, (), G::Timestamp, R, DefaultKeyTrace<K, G::Timestamp, R>>>;
}


impl<G: Scope, K: Data+Hashable, R: Diff> ArrangeBySelf<G, K, R> for Collection<G, K, R>
where G::Timestamp: Lattice+Ord {
    fn arrange_by_self(&self) -> Arranged<G, K, (), R, TraceAgent<K, (), G::Timestamp, R, DefaultKeyTrace<K, G::Timestamp, R>>> {
        self.map(|k| (k, ()))
            .arrange(DefaultKeyTrace::new())
    }
}