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
//! Iterator extension for splitting into batches, each introduced by a batch-starting item
//!
//! See
//! [`IteratorExt::batching_split_before_loose`] and
//! [`IteratorExt::batching_split_before_with_header`].
//!
//! # **UNSTABLE**
//!
//! This whole module is UNSTABLE and not part of the semver guarantees.
//! You'll only see it if you ran rustdoc with `--document-private-items`.
// This is achieved with `#[doc(hidden)]` on the top-level module reexport
// in `lib.rs`, which is the only place all of this isactually exposed.

use std::iter;
use std::marker::PhantomData;

use crate::util::PeekableIterator;

/// Iterator for the header, transformable into a [`Batches`] yielding subsequent batches
///
/// Returned by
/// [`.batching_split_before_with_header()`](IteratorExt::batching_split_before_with_header).
///
/// This type is both:
///  * An [`Iterator`], which returns the items in the header
///    (before the first batch-starting item);
///  * Transformable using [.subsequent()](BatchesWithHeader::subsequent)
///    into a [`Batches`], which yields the remainder of the input,
///    split into batches.
///
/// `II` is the iterator item type.  `I` is the input iterator.
/// `F` is the predicate for testing if an item is batch-starting.
pub struct BatchesWithHeader<II, I, F> {
    /// Input
    input: Input<II, I, F>,
}

/// Input, shared by our public structs
struct Input<II, I, F> {
    /// The input iterator
    unfiltered: I,
    /// Callback to test if this is batch-starting
    batch_starting: F,
    /// We're like a function that yields II
    marker: PhantomData<fn() -> II>,
}

/// An iterator-like object yielding an iterator for each batch.
///
/// Each call to [`.next_batch()`](Batches::next_batch)
/// yields an iterator for one subsequent batch,
/// which in turn will yield the individual items.
///
/// `Batches` is not an [`Iterator`] because
/// its returned item type (the sub-iterator)
/// borrows it mutably.
///
/// `II` is the iterator item type.  `I` is the input iterator.
/// `F` is the predicate for testing if an item is batch-starting.
pub struct Batches<II, I, F> {
    /// Input
    input: Input<II, I, F>,
    /// Should we avoid draining the end of the previous batch
    no_drain: Option<NoDrainToken>,
    /// Should we yield even (one) batch-starting item
    yield_one: Option<EvenYieldOneBatchStarting>,
}

/// Token stored (or not) in the state to indicate not to drain the previous batch
///
/// (We use `Option<NoDrainToken>` rather than `bool` because
/// booleans can be very confusing, and because
/// `Option` has good ergonomics with [`.take()`](Option::take) and `?`.)
struct NoDrainToken;

/// Token stored (or not) in the state to indicate to yield even a batch-starting item
///
/// (We use `Option<NoDrainToken>` rather than `bool` because
/// booleans can be very confusing, and because
/// `Option` has good ergonomics with [`.take()`](Option::take) and `?`.)
struct EvenYieldOneBatchStarting;

/// Iterator to yield the members of a batch.
///
/// This is the iterator returned by
/// [`.next_batch()`](Batches::next_batch).
///
/// `II` is the iterator item type.  `I` is the input iterator.
/// `F` is the predicate for testing if an item is batch-starting.
pub struct Batch<'p, II, I, F> {
    /// The parent, with all the actual state etc.
    ///
    /// It is less confusing to keep all the state in the parent iterator.
    parent: &'p mut Batches<II, I, F>,
}

impl<II, I, F> Input<II, I, F>
where
    I: Iterator<Item = II> + PeekableIterator,
    F: FnMut(&II) -> bool,
{
    /// Yield the next item - unless it is batch-starting.
    fn next_non_starting(&mut self) -> Option<II> {
        let item = self.unfiltered.peek()?;
        if (self.batch_starting)(item) {
            return None;
        };
        self.unfiltered.next()
    }
}

impl<II, I, F> Iterator for BatchesWithHeader<II, I, F>
where
    I: Iterator<Item = II> + PeekableIterator,
    F: FnMut(&II) -> bool,
{
    type Item = II;

    fn next(&mut self) -> Option<II> {
        self.input.next_non_starting()
    }
}

impl<II, I, F> BatchesWithHeader<II, I, F>
where
    I: Iterator<Item = II> + PeekableIterator,
    F: FnMut(&II) -> bool,
{
    /// Proceed from the header to the subsequent batches
    ///
    /// Any un-yielded items remaining in the header will be discarded.
    pub fn subsequent(mut self) -> Batches<II, I, F> {
        // Discard any un-yielded contents of the header
        let _ = self.by_ref().count();

        Batches {
            input: self.input,
            yield_one: None,
            no_drain: None,
        }
    }
}

impl<II, I, F> Iterator for Batch<'_, II, I, F>
where
    I: Iterator<Item = II> + PeekableIterator,
    F: FnMut(&II) -> bool,
{
    type Item = II;

    fn next(&mut self) -> Option<II> {
        if self.parent.yield_one.take().is_some() {
            self.parent.input.unfiltered.next()
        } else {
            self.parent.input.next_non_starting()
        }
    }
}

impl<II, I: Iterator<Item = II> + PeekableIterator, F: FnMut(&II) -> bool> Batches<II, I, F> {
    /// Proceed to the next batch
    ///
    /// If the input is exhausted (ie, there is no next batch), returns `None`.
    ///
    /// Any un-yielded items remaining in the previous batch will be discarded.
    //
    // Batches is a LendingIterator - its returned item type borrows from the
    // iterator itself - so can't impl Iterator.
    // <https://rust-lang.github.io/generic-associated-types-initiative/design_patterns/iterable.html>
    pub fn next_batch(&mut self) -> Option<Batch<'_, II, I, F>> {
        // Drain to the end of the batch
        if self.no_drain.take().is_none() {
            let _ = Batch { parent: self }.count();
        }
        let _: &II = self.input.unfiltered.peek()?;
        self.yield_one = Some(EvenYieldOneBatchStarting);
        Some(Batch { parent: self })
    }

    /// Map each batch
    pub fn map<T>(
        mut self,
        mut f: impl FnMut(Batch<'_, II, I, F>) -> T,
    ) -> impl Iterator<Item = T> {
        iter::from_fn(move || {
            let batch = self.next_batch()?;
            Some(f(batch))
        })
    }
}

/// **Extension trait providing `batching_split_before`**
pub trait IteratorExt: Iterator + Sized {
    /// Splits the input into a header followed by batches started according to a predicate
    ///
    /// The input is divided into:
    ///  * A header, containing no batch-starting items
    ///  * Zero or more subsequent batches, each with precisely one batch-starting item
    ///
    /// The returned value from `batching_split_before_with_header` is an iterator,
    /// which yields the elements in the header - before the first batch-starting item.
    ///
    /// After processing the header, call
    /// [`.subsequent()`](BatchesWithHeader::subsequent)
    /// which will return a [`Batches`],
    /// which is a meta-iterator-like-object which yields the subsequent batches.
    ///
    /// Each subsequent batch is then returned by calling
    /// [`.next_batch()`](Batches::next_batch)
    /// which yields a separate sub-iterator.
    ///
    /// A new batch is recognised for each input item for which `batch_starting` returns true.
    ///
    /// This method is named **with_header** because it separates out the header,
    /// using a typestate pattern, which is convenient for processing the header
    /// separately.
    ///
    /// (You will want to iterate the first batch by reference,
    /// so that the iteration doesn't consume the [`BatchesWithHeader`],
    /// which is what you will need to call `.subsequent()`.
    /// The API insists that you process the batches sequentially:
    /// you can only be processing one batch at a time.)
    ///
    /// # **UNSTABLE**
    ///
    /// This method is UNSTABLE and not part of the semver guarantees.
    /// You'll only see it if you ran rustdoc with `--document-private-items`.
    ///
    /// # Example
    ///
    /// ```
    /// use itertools::Itertools as _;
    /// use tor_netdoc::batching_split_before::IteratorExt as _;
    ///
    /// let mut batches = (1..10).peekable().batching_split_before_with_header(|v| v % 3 == 0);
    /// assert_eq!(batches.by_ref().collect_vec(), [ 1, 2 ]);
    ///
    /// let mut batches = batches.subsequent();
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 3, 4, 5 ]);
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 6, 7, 8 ]);
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 9 ]);
    /// assert!(batches.next_batch().is_none());
    /// ```
    fn batching_split_before_with_header<F>(
        self,
        batch_starting: F,
    ) -> BatchesWithHeader<Self::Item, Self, F>
    where
        F: FnMut(&Self::Item) -> bool,
    {
        let input = Input {
            unfiltered: self,
            batch_starting,
            marker: PhantomData,
        };
        BatchesWithHeader { input }
    }

    /// Splits the input into batches, with new batches started according to a predicate
    ///
    /// The input is divided into batches, just before each batch-starting item.
    /// The batch-starting item is included as the first item of every batch,
    /// except the first batch if the input starts with a non-batch-starting-item.
    ///
    /// If the input iterator is empty, there are no batches.
    ///
    /// This method is named **loose** because it neither
    /// insists that the iterator start with a batch-starting item,
    /// nor returns batches which always start with a batch-starting item.
    /// It is up to the caller to handle a possible first batch with no batch-starting item.
    ///
    /// Each batch is returned by calling
    /// [`.next_batch()`](Batches::next_batch)
    /// which yields a separate sub-iterator.
    ///
    /// A new batch is recognised for each input item for which `batch_start` returns true.
    ///
    /// (The API insists that you process the batches sequentially:
    /// you can only be processing one batch at a time.)
    ///
    /// # **UNSTABLE**
    ///
    /// This method is UNSTABLE and not part of the semver guarantees.
    /// You'll only see it if you ran rustdoc with `--document-private-items`.
    ///
    /// # Example
    ///
    /// ```
    /// use itertools::Itertools as _;
    /// use tor_netdoc::batching_split_before::IteratorExt as _;
    ///
    /// let mut batches = (1..10).peekable().batching_split_before_loose(|v| v % 3 == 0);
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 1, 2 ]);
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 3, 4, 5 ]);
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 6, 7, 8 ]);
    /// assert_eq!(batches.next_batch().unwrap().collect_vec(), [ 9 ]);
    /// assert!(batches.next_batch().is_none());
    /// ```
    fn batching_split_before_loose<F>(self, batch_starting: F) -> Batches<Self::Item, Self, F>
    where
        F: FnMut(&Self::Item) -> bool,
    {
        let input = Input {
            unfiltered: self,
            batch_starting,
            marker: PhantomData,
        };
        Batches {
            input,
            no_drain: Some(NoDrainToken),
            yield_one: None,
        }
    }
}
impl<I: Iterator> IteratorExt for I {}

#[cfg(test)]
mod tests {
    // @@ begin test lint list maintained by maint/add_warning @@
    #![allow(clippy::bool_assert_comparison)]
    #![allow(clippy::clone_on_copy)]
    #![allow(clippy::dbg_macro)]
    #![allow(clippy::mixed_attributes_style)]
    #![allow(clippy::print_stderr)]
    #![allow(clippy::print_stdout)]
    #![allow(clippy::single_char_pattern)]
    #![allow(clippy::unwrap_used)]
    #![allow(clippy::unchecked_duration_subtraction)]
    #![allow(clippy::useless_vec)]
    #![allow(clippy::needless_pass_by_value)]
    //! <!-- @@ end test lint list maintained by maint/add_warning @@ -->
    use super::*;
    use crate::util::*;
    use itertools::chain;
    use std::fmt::Debug;
    use std::iter;

    struct TrackingPeekable<I: Iterator>(Peekable<I>);
    impl<I: Iterator> Iterator for TrackingPeekable<I>
    where
        I::Item: Debug,
    {
        type Item = I::Item;
        fn next(&mut self) -> Option<I::Item> {
            let v = self.0.next();
            eprintln!("        iter yielded {v:?}");
            v
        }
    }
    impl<I: Iterator> PeekableIterator for TrackingPeekable<I>
    where
        I::Item: Debug,
    {
        fn peek(&mut self) -> Option<&I::Item> {
            let v = self.0.peek();
            eprintln!("        iter peeked {v:?}");
            v
        }
    }

    #[test]
    fn test_batching_split_before() {
        fn chk_exp(mut iter: impl Iterator<Item = u32>, exp: &[u32]) {
            eprintln!("    exp {exp:?}");
            for exp in exp.iter().cloned() {
                assert_eq!(iter.next(), Some(exp));
            }
            assert_eq!(iter.next(), None);
            assert_eq!(iter.next(), None);
            assert_eq!(iter.next(), None);
        }

        let chk_breakdown = |input: &[u32], iexp: &[u32], sexp: &[&[u32]]| {
            let chk_batches = |mut subseq: Batches<_, _, _>, sexp: &mut dyn Iterator<Item = _>| {
                loop {
                    match (subseq.next_batch(), sexp.next()) {
                        (Some(batch), Some(sexp)) => chk_exp(batch, sexp),
                        (None, None) => break,
                        (b, e) => panic!("({:?}, {e:?}", b.map(|_| ())),
                    }
                }
                assert!(subseq.next_batch().is_none());
                assert!(subseq.next_batch().is_none());
            };

            eprintln!("input {input:?}");
            let input = || TrackingPeekable(input.iter().cloned().peekable());
            let is_starting = |v: &u32| *v >= 10;

            {
                let mut header = input().batching_split_before_with_header(is_starting);

                chk_exp(&mut header, iexp);
                eprintln!("    subsequent...");
                let subseq = header.subsequent();
                let mut sexp = sexp.iter().cloned();
                chk_batches(subseq, &mut sexp);
            }

            {
                let batches = input().batching_split_before_loose(is_starting);

                let mut sexp =
                    chain!(iter::once(iexp), sexp.iter().cloned(),).filter(|s| !s.is_empty());

                chk_batches(batches, &mut sexp);
            }
        };

        chk_breakdown(&[], &[], &[]);

        chk_breakdown(&[10], &[], &[&[10]]);

        chk_breakdown(
            &[1, 2, 30, 4, 5, 60, 7, 8],
            &[1, 2],
            &[&[30, 4, 5], &[60, 7, 8]],
        );
    }
}