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
 702
 703
 704
 705
 706
 707
 708
 709
 710
 711
 712
 713
 714
 715
 716
 717
 718
 719
 720
 721
 722
 723
 724
 725
 726
 727
 728
 729
 730
 731
 732
 733
 734
 735
 736
 737
 738
 739
 740
 741
 742
 743
 744
 745
 746
 747
 748
 749
 750
 751
 752
 753
 754
 755
 756
 757
 758
 759
 760
 761
 762
 763
 764
 765
 766
 767
 768
 769
 770
 771
 772
 773
 774
 775
 776
 777
 778
 779
 780
 781
 782
 783
 784
 785
 786
 787
 788
 789
 790
 791
 792
 793
 794
 795
 796
 797
 798
 799
 800
 801
 802
 803
 804
 805
 806
 807
 808
 809
 810
 811
 812
 813
 814
 815
 816
 817
 818
 819
 820
 821
 822
 823
 824
 825
 826
 827
 828
 829
 830
 831
 832
 833
 834
 835
 836
 837
 838
 839
 840
 841
 842
 843
 844
 845
 846
 847
 848
 849
 850
 851
 852
 853
 854
 855
 856
 857
 858
 859
 860
 861
 862
 863
 864
 865
 866
 867
 868
 869
 870
 871
 872
 873
 874
 875
 876
 877
 878
 879
 880
 881
 882
 883
 884
 885
 886
 887
 888
 889
 890
 891
 892
 893
 894
 895
 896
 897
 898
 899
 900
 901
 902
 903
 904
 905
 906
 907
 908
 909
 910
 911
 912
 913
 914
 915
 916
 917
 918
 919
 920
 921
 922
 923
 924
 925
 926
 927
 928
 929
 930
 931
 932
 933
 934
 935
 936
 937
 938
 939
 940
 941
 942
 943
 944
 945
 946
 947
 948
 949
 950
 951
 952
 953
 954
 955
 956
 957
 958
 959
 960
 961
 962
 963
 964
 965
 966
 967
 968
 969
 970
 971
 972
 973
 974
 975
 976
 977
 978
 979
 980
 981
 982
 983
 984
 985
 986
 987
 988
 989
 990
 991
 992
 993
 994
 995
 996
 997
 998
 999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
// Copyright 2017 bluss and ndarray developers.
//
// Licensed under the Apache License, Version 2.0 <LICENSE-APACHE or
// http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
// <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your
// option. This file may not be copied, modified, or distributed
// except according to those terms.

#[macro_use]
mod zipmacro;

use std::mem::MaybeUninit;

use crate::imp_prelude::*;
use crate::AssignElem;
use crate::IntoDimension;
use crate::Layout;
use crate::NdIndex;

use crate::indexes::{indices, Indices};
use crate::layout::{CORDER, FORDER};

/// Return if the expression is a break value.
macro_rules! fold_while {
    ($e:expr) => {
        match $e {
            FoldWhile::Continue(x) => x,
            x => return x,
        }
    };
}

/// Broadcast an array so that it acts like a larger size and/or shape array.
///
/// See [broadcasting][1] for more information.
///
/// [1]: struct.ArrayBase.html#broadcasting
trait Broadcast<E>
where
    E: IntoDimension,
{
    type Output: NdProducer<Dim = E::Dim>;
    /// Broadcast the array to the new dimensions `shape`.
    ///
    /// ***Panics*** if broadcasting isn’t possible.
    fn broadcast_unwrap(self, shape: E) -> Self::Output;
    private_decl! {}
}

impl<S, D> ArrayBase<S, D>
where
    S: RawData,
    D: Dimension,
{
    pub(crate) fn layout_impl(&self) -> Layout {
        Layout::new(if self.is_standard_layout() {
            if self.ndim() <= 1 {
                FORDER | CORDER
            } else {
                CORDER
            }
        } else if self.ndim() > 1 && self.raw_view().reversed_axes().is_standard_layout() {
            FORDER
        } else {
            0
        })
    }
}

impl<'a, A, D, E> Broadcast<E> for ArrayView<'a, A, D>
where
    E: IntoDimension,
    D: Dimension,
{
    type Output = ArrayView<'a, A, E::Dim>;
    fn broadcast_unwrap(self, shape: E) -> Self::Output {
        let res: ArrayView<'_, A, E::Dim> = (&self).broadcast_unwrap(shape.into_dimension());
        unsafe { ArrayView::new(res.ptr, res.dim, res.strides) }
    }
    private_impl! {}
}

pub trait Splittable: Sized {
    fn split_at(self, axis: Axis, index: Ix) -> (Self, Self);
}

impl<D> Splittable for D
where
    D: Dimension,
{
    fn split_at(self, axis: Axis, index: Ix) -> (Self, Self) {
        let mut d1 = self;
        let mut d2 = d1.clone();
        let i = axis.index();
        let len = d1[i];
        d1[i] = index;
        d2[i] = len - index;
        (d1, d2)
    }
}

/// Argument conversion into a producer.
///
/// Slices and vectors can be used (equivalent to 1-dimensional array views).
///
/// This trait is like `IntoIterator` for `NdProducers` instead of iterators.
pub trait IntoNdProducer {
    /// The element produced per iteration.
    type Item;
    /// Dimension type of the producer
    type Dim: Dimension;
    type Output: NdProducer<Dim = Self::Dim, Item = Self::Item>;
    /// Convert the value into an `NdProducer`.
    fn into_producer(self) -> Self::Output;
}

impl<P> IntoNdProducer for P
where
    P: NdProducer,
{
    type Item = P::Item;
    type Dim = P::Dim;
    type Output = Self;
    fn into_producer(self) -> Self::Output {
        self
    }
}

/// A producer of an n-dimensional set of elements;
/// for example an array view, mutable array view or an iterator
/// that yields chunks.
///
/// Producers are used as a arguments to [`Zip`](struct.Zip.html) and
/// [`azip!()`](macro.azip.html).
///
/// # Comparison to `IntoIterator`
///
/// Most `NdProducers` are *iterable* (implement `IntoIterator`) but not directly
/// iterators. This separation is needed because the producer represents
/// a multidimensional set of items, it can be split along a particular axis for
/// parallelization, and it has no fixed correspondance to a sequence.
///
/// The natural exception is one dimensional producers, like `AxisIter`, which
/// implement `Iterator` directly
/// (`AxisIter` traverses a one dimensional sequence, along an axis, while
/// *producing* multidimensional items).
///
/// See also [`IntoNdProducer`](trait.IntoNdProducer.html)
pub trait NdProducer {
    /// The element produced per iteration.
    type Item;
    // Internal use / Pointee type
    /// Dimension type
    type Dim: Dimension;

    // The pointer Ptr is used by an array view to simply point to the
    // current element. It doesn't have to be a pointer (see Indices).
    // Its main function is that it can be incremented with a particular
    // stride (= along a particular axis)
    #[doc(hidden)]
    /// Pointer or stand-in for pointer
    type Ptr: Offset<Stride = Self::Stride>;
    #[doc(hidden)]
    /// Pointer stride
    type Stride: Copy;

    #[doc(hidden)]
    fn layout(&self) -> Layout;
    #[doc(hidden)]
    fn raw_dim(&self) -> Self::Dim;
    #[doc(hidden)]
    fn equal_dim(&self, dim: &Self::Dim) -> bool {
        self.raw_dim() == *dim
    }
    #[doc(hidden)]
    fn as_ptr(&self) -> Self::Ptr;
    #[doc(hidden)]
    unsafe fn as_ref(&self, ptr: Self::Ptr) -> Self::Item;
    #[doc(hidden)]
    unsafe fn uget_ptr(&self, i: &Self::Dim) -> Self::Ptr;
    #[doc(hidden)]
    fn stride_of(&self, axis: Axis) -> <Self::Ptr as Offset>::Stride;
    #[doc(hidden)]
    fn contiguous_stride(&self) -> Self::Stride;
    #[doc(hidden)]
    fn split_at(self, axis: Axis, index: usize) -> (Self, Self)
    where
        Self: Sized;
    private_decl! {}
}

pub trait Offset: Copy {
    type Stride: Copy;
    unsafe fn stride_offset(self, s: Self::Stride, index: usize) -> Self;
    private_decl! {}
}

impl<T> Offset for *const T {
    type Stride = isize;
    unsafe fn stride_offset(self, s: Self::Stride, index: usize) -> Self {
        self.offset(s * (index as isize))
    }
    private_impl! {}
}

impl<T> Offset for *mut T {
    type Stride = isize;
    unsafe fn stride_offset(self, s: Self::Stride, index: usize) -> Self {
        self.offset(s * (index as isize))
    }
    private_impl! {}
}

trait ZippableTuple: Sized {
    type Item;
    type Ptr: OffsetTuple<Args = Self::Stride> + Copy;
    type Dim: Dimension;
    type Stride: Copy;
    fn as_ptr(&self) -> Self::Ptr;
    unsafe fn as_ref(&self, ptr: Self::Ptr) -> Self::Item;
    unsafe fn uget_ptr(&self, i: &Self::Dim) -> Self::Ptr;
    fn stride_of(&self, index: usize) -> Self::Stride;
    fn contiguous_stride(&self) -> Self::Stride;
    fn split_at(self, axis: Axis, index: usize) -> (Self, Self);
}

/// An array reference is an n-dimensional producer of element references
/// (like ArrayView).
impl<'a, A: 'a, S, D> IntoNdProducer for &'a ArrayBase<S, D>
where
    D: Dimension,
    S: Data<Elem = A>,
{
    type Item = &'a A;
    type Dim = D;
    type Output = ArrayView<'a, A, D>;
    fn into_producer(self) -> Self::Output {
        self.view()
    }
}

/// A mutable array reference is an n-dimensional producer of mutable element
/// references (like ArrayViewMut).
impl<'a, A: 'a, S, D> IntoNdProducer for &'a mut ArrayBase<S, D>
where
    D: Dimension,
    S: DataMut<Elem = A>,
{
    type Item = &'a mut A;
    type Dim = D;
    type Output = ArrayViewMut<'a, A, D>;
    fn into_producer(self) -> Self::Output {
        self.view_mut()
    }
}

/// A slice is a one-dimensional producer
impl<'a, A: 'a> IntoNdProducer for &'a [A] {
    type Item = <Self::Output as NdProducer>::Item;
    type Dim = Ix1;
    type Output = ArrayView1<'a, A>;
    fn into_producer(self) -> Self::Output {
        <_>::from(self)
    }
}

/// A mutable slice is a mutable one-dimensional producer
impl<'a, A: 'a> IntoNdProducer for &'a mut [A] {
    type Item = <Self::Output as NdProducer>::Item;
    type Dim = Ix1;
    type Output = ArrayViewMut1<'a, A>;
    fn into_producer(self) -> Self::Output {
        <_>::from(self)
    }
}

/// A Vec is a one-dimensional producer
impl<'a, A: 'a> IntoNdProducer for &'a Vec<A> {
    type Item = <Self::Output as NdProducer>::Item;
    type Dim = Ix1;
    type Output = ArrayView1<'a, A>;
    fn into_producer(self) -> Self::Output {
        <_>::from(self)
    }
}

/// A mutable Vec is a mutable one-dimensional producer
impl<'a, A: 'a> IntoNdProducer for &'a mut Vec<A> {
    type Item = <Self::Output as NdProducer>::Item;
    type Dim = Ix1;
    type Output = ArrayViewMut1<'a, A>;
    fn into_producer(self) -> Self::Output {
        <_>::from(self)
    }
}

impl<'a, A, D: Dimension> NdProducer for ArrayView<'a, A, D> {
    type Item = &'a A;
    type Dim = D;
    type Ptr = *mut A;
    type Stride = isize;

    private_impl! {}
    #[doc(hidden)]
    fn raw_dim(&self) -> Self::Dim {
        self.raw_dim()
    }

    #[doc(hidden)]
    fn equal_dim(&self, dim: &Self::Dim) -> bool {
        self.dim.equal(dim)
    }

    #[doc(hidden)]
    fn as_ptr(&self) -> *mut A {
        self.as_ptr() as _
    }

    #[doc(hidden)]
    fn layout(&self) -> Layout {
        self.layout_impl()
    }

    #[doc(hidden)]
    unsafe fn as_ref(&self, ptr: *mut A) -> Self::Item {
        &*ptr
    }

    #[doc(hidden)]
    unsafe fn uget_ptr(&self, i: &Self::Dim) -> *mut A {
        self.ptr.as_ptr().offset(i.index_unchecked(&self.strides))
    }

    #[doc(hidden)]
    fn stride_of(&self, axis: Axis) -> isize {
        self.stride_of(axis)
    }

    #[inline(always)]
    fn contiguous_stride(&self) -> Self::Stride {
        1
    }

    #[doc(hidden)]
    fn split_at(self, axis: Axis, index: usize) -> (Self, Self) {
        self.split_at(axis, index)
    }
}

impl<'a, A, D: Dimension> NdProducer for ArrayViewMut<'a, A, D> {
    type Item = &'a mut A;
    type Dim = D;
    type Ptr = *mut A;
    type Stride = isize;

    private_impl! {}
    #[doc(hidden)]
    fn raw_dim(&self) -> Self::Dim {
        self.raw_dim()
    }

    #[doc(hidden)]
    fn equal_dim(&self, dim: &Self::Dim) -> bool {
        self.dim.equal(dim)
    }

    #[doc(hidden)]
    fn as_ptr(&self) -> *mut A {
        self.as_ptr() as _
    }

    #[doc(hidden)]
    fn layout(&self) -> Layout {
        self.layout_impl()
    }

    #[doc(hidden)]
    unsafe fn as_ref(&self, ptr: *mut A) -> Self::Item {
        &mut *ptr
    }

    #[doc(hidden)]
    unsafe fn uget_ptr(&self, i: &Self::Dim) -> *mut A {
        self.ptr.as_ptr().offset(i.index_unchecked(&self.strides))
    }

    #[doc(hidden)]
    fn stride_of(&self, axis: Axis) -> isize {
        self.stride_of(axis)
    }

    #[inline(always)]
    fn contiguous_stride(&self) -> Self::Stride {
        1
    }

    #[doc(hidden)]
    fn split_at(self, axis: Axis, index: usize) -> (Self, Self) {
        self.split_at(axis, index)
    }
}

impl<A, D: Dimension> NdProducer for RawArrayView<A, D> {
    type Item = *const A;
    type Dim = D;
    type Ptr = *const A;
    type Stride = isize;

    private_impl! {}
    #[doc(hidden)]
    fn raw_dim(&self) -> Self::Dim {
        self.raw_dim()
    }

    #[doc(hidden)]
    fn equal_dim(&self, dim: &Self::Dim) -> bool {
        self.dim.equal(dim)
    }

    #[doc(hidden)]
    fn as_ptr(&self) -> *const A {
        self.as_ptr()
    }

    #[doc(hidden)]
    fn layout(&self) -> Layout {
        self.layout_impl()
    }

    #[doc(hidden)]
    unsafe fn as_ref(&self, ptr: *const A) -> *const A {
        ptr
    }

    #[doc(hidden)]
    unsafe fn uget_ptr(&self, i: &Self::Dim) -> *const A {
        self.ptr.as_ptr().offset(i.index_unchecked(&self.strides))
    }

    #[doc(hidden)]
    fn stride_of(&self, axis: Axis) -> isize {
        self.stride_of(axis)
    }

    #[inline(always)]
    fn contiguous_stride(&self) -> Self::Stride {
        1
    }

    #[doc(hidden)]
    fn split_at(self, axis: Axis, index: usize) -> (Self, Self) {
        self.split_at(axis, index)
    }
}

impl<A, D: Dimension> NdProducer for RawArrayViewMut<A, D> {
    type Item = *mut A;
    type Dim = D;
    type Ptr = *mut A;
    type Stride = isize;

    private_impl! {}
    #[doc(hidden)]
    fn raw_dim(&self) -> Self::Dim {
        self.raw_dim()
    }

    #[doc(hidden)]
    fn equal_dim(&self, dim: &Self::Dim) -> bool {
        self.dim.equal(dim)
    }

    #[doc(hidden)]
    fn as_ptr(&self) -> *mut A {
        self.as_ptr() as _
    }

    #[doc(hidden)]
    fn layout(&self) -> Layout {
        self.layout_impl()
    }

    #[doc(hidden)]
    unsafe fn as_ref(&self, ptr: *mut A) -> *mut A {
        ptr
    }

    #[doc(hidden)]
    unsafe fn uget_ptr(&self, i: &Self::Dim) -> *mut A {
        self.ptr.as_ptr().offset(i.index_unchecked(&self.strides))
    }

    #[doc(hidden)]
    fn stride_of(&self, axis: Axis) -> isize {
        self.stride_of(axis)
    }

    #[inline(always)]
    fn contiguous_stride(&self) -> Self::Stride {
        1
    }

    #[doc(hidden)]
    fn split_at(self, axis: Axis, index: usize) -> (Self, Self) {
        self.split_at(axis, index)
    }
}

/// Lock step function application across several arrays or other producers.
///
/// Zip allows matching several producers to each other elementwise and applying
/// a function over all tuples of elements (one item from each input at
/// a time).
///
/// In general, the zip uses a tuple of producers
/// ([`NdProducer`](trait.NdProducer.html) trait) that all have to be of the
/// same shape. The NdProducer implementation defines what its item type is
/// (for example if it's a shared reference, mutable reference or an array
/// view etc).
///
/// If all the input arrays are of the same memory layout the zip performs much
/// better and the compiler can usually vectorize the loop (if applicable).
///
/// The order elements are visited is not specified. The producers don’t have to
/// have the same item type.
///
/// The `Zip` has two methods for function application: `apply` and
/// `fold_while`. The zip object can be split, which allows parallelization.
/// A read-only zip object (no mutable producers) can be cloned.
///
/// See also the [`azip!()` macro][az] which offers a convenient shorthand
/// to common ways to use `Zip`.
///
/// [az]: macro.azip.html
///
/// ```
/// use ndarray::Zip;
/// use ndarray::Array2;
///
/// type M = Array2<f64>;
///
/// // Create four 2d arrays of the same size
/// let mut a = M::zeros((64, 32));
/// let b = M::from_elem(a.dim(), 1.);
/// let c = M::from_elem(a.dim(), 2.);
/// let d = M::from_elem(a.dim(), 3.);
///
/// // Example 1: Perform an elementwise arithmetic operation across
/// // the four arrays a, b, c, d.
///
/// Zip::from(&mut a)
///     .and(&b)
///     .and(&c)
///     .and(&d)
///     .apply(|w, &x, &y, &z| {
///         *w += x + y * z;
///     });
///
/// // Example 2: Create a new array `totals` with one entry per row of `a`.
/// //  Use Zip to traverse the rows of `a` and assign to the corresponding
/// //  entry in `totals` with the sum across each row.
/// //  This is possible because the producer for `totals` and the row producer
/// //  for `a` have the same shape and dimensionality.
/// //  The rows producer yields one array view (`row`) per iteration.
///
/// use ndarray::{Array1, Axis};
///
/// let mut totals = Array1::zeros(a.nrows());
///
/// Zip::from(&mut totals)
///     .and(a.genrows())
///     .apply(|totals, row| *totals = row.sum());
///
/// // Check the result against the built in `.sum_axis()` along axis 1.
/// assert_eq!(totals, a.sum_axis(Axis(1)));
///
///
/// // Example 3: Recreate Example 2 using apply_collect to make a new array
///
/// let mut totals2 = Zip::from(a.genrows()).apply_collect(|row| row.sum());
///
/// // Check the result against the previous example.
/// assert_eq!(totals, totals2);
/// ```
#[derive(Debug, Clone)]
pub struct Zip<Parts, D> {
    parts: Parts,
    dimension: D,
    layout: Layout,
}


impl<P, D> Zip<(P,), D>
where
    D: Dimension,
    P: NdProducer<Dim = D>,
{
    /// Create a new `Zip` from the input array or other producer `p`.
    ///
    /// The Zip will take the exact dimension of `p` and all inputs
    /// must have the same dimensions (or be broadcast to them).
    pub fn from<IP>(p: IP) -> Self
    where
        IP: IntoNdProducer<Dim = D, Output = P, Item = P::Item>,
    {
        let array = p.into_producer();
        let dim = array.raw_dim();
        Zip {
            dimension: dim,
            layout: array.layout(),
            parts: (array,),
        }
    }
}
impl<P, D> Zip<(Indices<D>, P), D>
where
    D: Dimension + Copy,
    P: NdProducer<Dim = D>,
{
    /// Create a new `Zip` with an index producer and the producer `p`.
    ///
    /// The Zip will take the exact dimension of `p` and all inputs
    /// must have the same dimensions (or be broadcast to them).
    ///
    /// *Note:* Indexed zip has overhead.
    pub fn indexed<IP>(p: IP) -> Self
    where
        IP: IntoNdProducer<Dim = D, Output = P, Item = P::Item>,
    {
        let array = p.into_producer();
        let dim = array.raw_dim();
        Zip::from(indices(dim)).and(array)
    }
}

impl<Parts, D> Zip<Parts, D>
where
    D: Dimension,
{
    fn check<P>(&self, part: &P)
    where
        P: NdProducer<Dim = D>,
    {
        ndassert!(
            part.equal_dim(&self.dimension),
            "Zip: Producer dimension mismatch, expected: {:?}, got: {:?}",
            self.dimension,
            part.raw_dim()
        );
    }

    /// Return a the number of element tuples in the Zip
    pub fn size(&self) -> usize {
        self.dimension.size()
    }

    /// Return the length of `axis`
    ///
    /// ***Panics*** if `axis` is out of bounds.
    fn len_of(&self, axis: Axis) -> usize {
        self.dimension[axis.index()]
    }

    /// Return an *approximation* to the max stride axis; if
    /// component arrays disagree, there may be no choice better than the
    /// others.
    fn max_stride_axis(&self) -> Axis {
        let i = match self.layout.flag() {
            FORDER => self
                .dimension
                .slice()
                .iter()
                .rposition(|&len| len > 1)
                .unwrap_or(self.dimension.ndim() - 1),
            /* corder or default */
            _ => self
                .dimension
                .slice()
                .iter()
                .position(|&len| len > 1)
                .unwrap_or(0),
        };
        Axis(i)
    }
}

impl<P, D> Zip<P, D>
where
    D: Dimension,
{
    fn apply_core<F, Acc>(&mut self, acc: Acc, function: F) -> FoldWhile<Acc>
    where
        F: FnMut(Acc, P::Item) -> FoldWhile<Acc>,
        P: ZippableTuple<Dim = D>,
    {
        if self.layout.is(CORDER | FORDER) {
            self.apply_core_contiguous(acc, function)
        } else {
            self.apply_core_strided(acc, function)
        }
    }
    fn apply_core_contiguous<F, Acc>(&mut self, mut acc: Acc, mut function: F) -> FoldWhile<Acc>
    where
        F: FnMut(Acc, P::Item) -> FoldWhile<Acc>,
        P: ZippableTuple<Dim = D>,
    {
        debug_assert!(self.layout.is(CORDER | FORDER));
        let size = self.dimension.size();
        let ptrs = self.parts.as_ptr();
        let inner_strides = self.parts.contiguous_stride();
        for i in 0..size {
            unsafe {
                let ptr_i = ptrs.stride_offset(inner_strides, i);
                acc = fold_while![function(acc, self.parts.as_ref(ptr_i))];
            }
        }
        FoldWhile::Continue(acc)
    }

    fn apply_core_strided<F, Acc>(&mut self, mut acc: Acc, mut function: F) -> FoldWhile<Acc>
    where
        F: FnMut(Acc, P::Item) -> FoldWhile<Acc>,
        P: ZippableTuple<Dim = D>,
    {
        let n = self.dimension.ndim();
        if n == 0 {
            panic!("Unreachable: ndim == 0 is contiguous")
        }
        let unroll_axis = n - 1;
        let inner_len = self.dimension[unroll_axis];
        self.dimension[unroll_axis] = 1;
        let mut index_ = self.dimension.first_index();
        let inner_strides = self.parts.stride_of(unroll_axis);
        while let Some(index) = index_ {
            // Let's “unroll” the loop over the innermost axis
            unsafe {
                let ptr = self.parts.uget_ptr(&index);
                for i in 0..inner_len {
                    let p = ptr.stride_offset(inner_strides, i);
                    acc = fold_while!(function(acc, self.parts.as_ref(p)));
                }
            }

            index_ = self.dimension.next_for(index);
        }
        self.dimension[unroll_axis] = inner_len;
        FoldWhile::Continue(acc)
    }

    pub(crate) fn uninitalized_for_current_layout<T>(&self) -> Array<MaybeUninit<T>, D>
    {
        let is_f = !self.layout.is(CORDER) && self.layout.is(FORDER);
        Array::maybe_uninit(self.dimension.clone().set_f(is_f))
    }
}

/*
trait Offset : Copy {
    unsafe fn offset(self, off: isize) -> Self;
    unsafe fn stride_offset(self, index: usize, stride: isize) -> Self {
        self.offset(index as isize * stride)
    }
}

impl<T> Offset for *mut T {
    unsafe fn offset(self, off: isize) -> Self {
        self.offset(off)
    }
}
*/

trait OffsetTuple {
    type Args;
    unsafe fn stride_offset(self, stride: Self::Args, index: usize) -> Self;
}

impl<T> OffsetTuple for *mut T {
    type Args = isize;
    unsafe fn stride_offset(self, stride: Self::Args, index: usize) -> Self {
        self.offset(index as isize * stride)
    }
}

macro_rules! offset_impl {
    ($([$($param:ident)*][ $($q:ident)*],)+) => {
        $(
        #[allow(non_snake_case)]
        impl<$($param: Offset),*> OffsetTuple for ($($param, )*) {
            type Args = ($($param::Stride,)*);
            unsafe fn stride_offset(self, stride: Self::Args, index: usize) -> Self {
                let ($($param, )*) = self;
                let ($($q, )*) = stride;
                ($(Offset::stride_offset($param, $q, index),)*)
            }
        }
        )+
    }
}

offset_impl! {
    [A ][ a],
    [A B][ a b],
    [A B C][ a b c],
    [A B C D][ a b c d],
    [A B C D E][ a b c d e],
    [A B C D E F][ a b c d e f],
}

macro_rules! zipt_impl {
    ($([$($p:ident)*][ $($q:ident)*],)+) => {
        $(
        #[allow(non_snake_case)]
        impl<Dim: Dimension, $($p: NdProducer<Dim=Dim>),*> ZippableTuple for ($($p, )*) {
            type Item = ($($p::Item, )*);
            type Ptr = ($($p::Ptr, )*);
            type Dim = Dim;
            type Stride = ($($p::Stride,)* );

            fn stride_of(&self, index: usize) -> Self::Stride {
                let ($(ref $p,)*) = *self;
                ($($p.stride_of(Axis(index)), )*)
            }

            fn contiguous_stride(&self) -> Self::Stride {
                let ($(ref $p,)*) = *self;
                ($($p.contiguous_stride(), )*)
            }

            fn as_ptr(&self) -> Self::Ptr {
                let ($(ref $p,)*) = *self;
                ($($p.as_ptr(), )*)
            }
            unsafe fn as_ref(&self, ptr: Self::Ptr) -> Self::Item {
                let ($(ref $q ,)*) = *self;
                let ($($p,)*) = ptr;
                ($($q.as_ref($p),)*)
            }

            unsafe fn uget_ptr(&self, i: &Self::Dim) -> Self::Ptr {
                let ($(ref $p,)*) = *self;
                ($($p.uget_ptr(i), )*)
            }

            fn split_at(self, axis: Axis, index: Ix) -> (Self, Self) {
                let ($($p,)*) = self;
                let ($($p,)*) = (
                    $($p.split_at(axis, index), )*
                );
                (
                    ($($p.0,)*),
                    ($($p.1,)*)
                )
            }
        }
        )+
    }
}

zipt_impl! {
    [A ][ a],
    [A B][ a b],
    [A B C][ a b c],
    [A B C D][ a b c d],
    [A B C D E][ a b c d e],
    [A B C D E F][ a b c d e f],
}

macro_rules! map_impl {
    ($([$notlast:ident $($p:ident)*],)+) => {
        $(
        #[allow(non_snake_case)]
        impl<D, $($p),*> Zip<($($p,)*), D>
            where D: Dimension,
                  $($p: NdProducer<Dim=D> ,)*
        {
            /// Apply a function to all elements of the input arrays,
            /// visiting elements in lock step.
            pub fn apply<F>(mut self, mut function: F)
                where F: FnMut($($p::Item),*)
            {
                self.apply_core((), move |(), args| {
                    let ($($p,)*) = args;
                    FoldWhile::Continue(function($($p),*))
                });
            }

            /// Apply a fold function to all elements of the input arrays,
            /// visiting elements in lock step.
            ///
            /// # Example
            ///
            /// The expression `tr(AᵀB)` can be more efficiently computed as
            /// the equivalent expression `∑ᵢⱼ(A∘B)ᵢⱼ` (i.e. the sum of the
            /// elements of the entry-wise product). It would be possible to
            /// evaluate this expression by first computing the entry-wise
            /// product, `A∘B`, and then computing the elementwise sum of that
            /// product, but it's possible to do this in a single loop (and
            /// avoid an extra heap allocation if `A` and `B` can't be
            /// consumed) by using `Zip`:
            ///
            /// ```
            /// use ndarray::{array, Zip};
            ///
            /// let a = array![[1, 5], [3, 7]];
            /// let b = array![[2, 4], [8, 6]];
            ///
            /// // Without using `Zip`. This involves two loops and an extra
            /// // heap allocation for the result of `&a * &b`.
            /// let sum_prod_nonzip = (&a * &b).sum();
            /// // Using `Zip`. This is a single loop without any heap allocations.
            /// let sum_prod_zip = Zip::from(&a).and(&b).fold(0, |acc, a, b| acc + a * b);
            ///
            /// assert_eq!(sum_prod_nonzip, sum_prod_zip);
            /// ```
            pub fn fold<F, Acc>(mut self, acc: Acc, mut function: F) -> Acc
            where
                F: FnMut(Acc, $($p::Item),*) -> Acc,
            {
                self.apply_core(acc, move |acc, args| {
                    let ($($p,)*) = args;
                    FoldWhile::Continue(function(acc, $($p),*))
                }).into_inner()
            }

            /// Apply a fold function to the input arrays while the return
            /// value is `FoldWhile::Continue`, visiting elements in lock step.
            ///
            pub fn fold_while<F, Acc>(mut self, acc: Acc, mut function: F)
                -> FoldWhile<Acc>
                where F: FnMut(Acc, $($p::Item),*) -> FoldWhile<Acc>
            {
                self.apply_core(acc, move |acc, args| {
                    let ($($p,)*) = args;
                    function(acc, $($p),*)
                })
            }

            /// Tests if every element of the iterator matches a predicate.
            ///
            /// Returns `true` if `predicate` evaluates to `true` for all elements.
            /// Returns `true` if the input arrays are empty.
            ///
            /// Example:
            ///
            /// ```
            /// use ndarray::{array, Zip};
            /// let a = array![1, 2, 3];
            /// let b = array![1, 4, 9];
            /// assert!(Zip::from(&a).and(&b).all(|&a, &b| a * a == b));
            /// ```
            pub fn all<F>(mut self, mut predicate: F) -> bool
                where F: FnMut($($p::Item),*) -> bool
            {
                !self.apply_core((), move |_, args| {
                    let ($($p,)*) = args;
                    if predicate($($p),*) {
                        FoldWhile::Continue(())
                    } else {
                        FoldWhile::Done(())
                    }
                }).is_done()
            }

            expand_if!(@bool [$notlast]

            /// Include the producer `p` in the Zip.
            ///
            /// ***Panics*** if `p`’s shape doesn’t match the Zip’s exactly.
            pub fn and<P>(self, p: P) -> Zip<($($p,)* P::Output, ), D>
                where P: IntoNdProducer<Dim=D>,
            {
                let array = p.into_producer();
                self.check(&array);
                let part_layout = array.layout();
                let ($($p,)*) = self.parts;
                Zip {
                    parts: ($($p,)* array, ),
                    layout: self.layout.and(part_layout),
                    dimension: self.dimension,
                }
            }

            /// Include the producer `p` in the Zip, broadcasting if needed.
            ///
            /// If their shapes disagree, `rhs` is broadcast to the shape of `self`.
            ///
            /// ***Panics*** if broadcasting isn’t possible.
            pub fn and_broadcast<'a, P, D2, Elem>(self, p: P)
                -> Zip<($($p,)* ArrayView<'a, Elem, D>, ), D>
                where P: IntoNdProducer<Dim=D2, Output=ArrayView<'a, Elem, D2>, Item=&'a Elem>,
                      D2: Dimension,
            {
                let array = p.into_producer().broadcast_unwrap(self.dimension.clone());
                let part_layout = array.layout();
                let ($($p,)*) = self.parts;
                Zip {
                    parts: ($($p,)* array, ),
                    layout: self.layout.and(part_layout),
                    dimension: self.dimension,
                }
            }

            /// Apply and collect the results into a new array, which has the same size as the
            /// inputs.
            ///
            /// If all inputs are c- or f-order respectively, that is preserved in the output.
            ///
            /// Restricted to functions that produce copyable results for technical reasons; other
            /// cases are not yet implemented.
            pub fn apply_collect<R>(self, f: impl FnMut($($p::Item,)* ) -> R) -> Array<R, D>
                where R: Copy,
            {
                // To support non-Copy elements, implementation of dropping partial array (on
                // panic) is needed
                let mut output = self.uninitalized_for_current_layout::<R>();
                self.apply_assign_into(&mut output, f);
                unsafe {
                    output.assume_init()
                }
            }

            /// Apply and assign the results into the producer `into`, which should have the same
            /// size as the other inputs.
            ///
            /// The producer should have assignable items as dictated by the `AssignElem` trait,
            /// for example `&mut R`.
            pub fn apply_assign_into<R, Q>(self, into: Q, mut f: impl FnMut($($p::Item,)* ) -> R)
                where Q: IntoNdProducer<Dim=D>,
                      Q::Item: AssignElem<R>
            {
                self.and(into)
                    .apply(move |$($p, )* output_| {
                        output_.assign_elem(f($($p ),*));
                    });
            }


            );

            /// Split the `Zip` evenly in two.
            ///
            /// It will be split in the way that best preserves element locality.
            pub fn split(self) -> (Self, Self) {
                debug_assert_ne!(self.size(), 0, "Attempt to split empty zip");
                debug_assert_ne!(self.size(), 1, "Attempt to split zip with 1 elem");
                // Always split in a way that preserves layout (if any)
                let axis = self.max_stride_axis();
                let index = self.len_of(axis) / 2;
                let (p1, p2) = self.parts.split_at(axis, index);
                let (d1, d2) = self.dimension.split_at(axis, index);
                (Zip {
                    dimension: d1,
                    layout: self.layout,
                    parts: p1,
                },
                Zip {
                    dimension: d2,
                    layout: self.layout,
                    parts: p2,
                })
            }
        }
        )+
    }
}

map_impl! {
    [true P1],
    [true P1 P2],
    [true P1 P2 P3],
    [true P1 P2 P3 P4],
    [true P1 P2 P3 P4 P5],
    [false P1 P2 P3 P4 P5 P6],
}

/// Value controlling the execution of `.fold_while` on `Zip`.
#[derive(Debug, Copy, Clone)]
pub enum FoldWhile<T> {
    /// Continue folding with this value
    Continue(T),
    /// Fold is complete and will return this value
    Done(T),
}

impl<T> FoldWhile<T> {
    /// Return the inner value
    pub fn into_inner(self) -> T {
        match self {
            FoldWhile::Continue(x) | FoldWhile::Done(x) => x,
        }
    }

    /// Return true if it is `Done`, false if `Continue`
    pub fn is_done(&self) -> bool {
        match *self {
            FoldWhile::Continue(_) => false,
            FoldWhile::Done(_) => true,
        }
    }
}