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
//! An counting sort implementation for [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html)s. //! //! Provides the trait [`CountingSort`](trait.CountingSort.html) with a blanket implementation for //! [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html)s //! for all types `T` that implement (beyond other `std` or `core` traits) the here defined //! [`TryIntoIndex`](trait.TryIntoIndex.html) trait. //! Types that implement this trait can be tried to be converted to an //! [`usize`](https://doc.rust-lang.org/std/primitive.usize.html), i.e. an index. //! //! This trait is already implemented for the following integer types: //! //! * [`u8`](https://doc.rust-lang.org/std/primitive.u8.html) //! * [`u16`](https://doc.rust-lang.org/std/primitive.u16.html) //! * [`u32`](https://doc.rust-lang.org/std/primitive.u32.html) //! * [`usize`](https://doc.rust-lang.org/std/primitive.usize.html) //! * [`i8`](https://doc.rust-lang.org/std/primitive.i8.html) //! * [`i16`](https://doc.rust-lang.org/std/primitive.i16.html) //! * [`i32`](https://doc.rust-lang.org/std/primitive.i32.html) //! //! This means for all [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html)s, //! [`LinkedList`](https://doc.rust-lang.org/std/collections/struct.LinkedList.html)s, //! [`slice`](https://doc.rust-lang.org/std/primitive.slice.html)s or any other //! of the implementors of the [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html) //! trait holding one of the above integers types, counting sort can be executed. //! //! **Note:** Counting sort is also implemented for [`BTreeSet`](https://doc.rust-lang.org/std/collections/struct.BTreeSet.html), //! however it makes no sense to execute it there, since all elements are already in order and further sorting is completely //! useless. //! //! # Example //! //! ```rust //! /* //! * Add counting sort to your source code. //! * counting sort immediately works "out of the box" //! * for all Iterators and integers like //! * u8, i8, u16, i16. //! */ //! use counting_sort::CountingSort; //! //! let vec = vec![2,4,1,3]; //! // counting sort may fail, therefore a result is returned //! let sorted_vec_result = vec.iter().cnt_sort(); //! //! assert!(sorted_vec_result.is_ok()); //! // if successful sorted elements were copied into a Vec //! assert_eq!(vec![1,2,3,4], sorted_vec_result.unwrap()); //! ``` //! //! # Notes //! //! * The counting sort algorithm has an `O(n+d)` (`d` being the range between the minimum value and the maximum value) asymptotic runtime in comparison to an `O(n*log(n))` //! of the Rust std library implementation of [`slice.sort`](https://doc.rust-lang.org/std/primitive.slice.html#method.sort) //! * However the memory consumption is higher //! * Dependent on the range `d` between the minumum value and the maximum value (`d = max_value - min_value`), //! a [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) of //! [`usize`](https://doc.rust-lang.org/std/primitive.usize.html)'s is allocated //! * This may fast result in GB of memory: the maximum range of [`u32`](https://doc.rust-lang.org/std/primitive.u32.html) is //! 4294967295, if usize is 4 bytes, than the memory consumption is 17179869180 bytes or approximately 16 GB //! (1 GB = 1024*1024*1024 bytes) //! * Additionally the current implementation does not consume the given iterator //! * This means the counting sort algorithm excels whenever there are a lot of elements to be sorted but the range //! range between minumum value and maximum value is small //! * counting sort for e.g. [HashSet](https://doc.rust-lang.org/std/collections/struct.HashSet.html)'s is sub-optimal since every element exists only //! once in a [HashSet](https://doc.rust-lang.org/std/collections/struct.HashSet.html). Counting sort excels when a lot of elements exist in the //! collection but the number of distinct elements is small. //! * **<span style="color:red">Caution:</span>** Be careful using this algorithm when the range between minumum value and maximum value is large //! * An excellent illustration about the counting sort algorithm can be found [here](https://www.cs.usfca.edu/~galles/visualization/CountingSort.html) //! * Wikipedia article on [counting sort](https://en.wikipedia.org/wiki/Counting_sort) #![warn(missing_docs)] #![warn(missing_doc_code_examples)] use core::cmp::{max, min, Ord}; use core::convert::TryInto; use core::fmt; use core::fmt::Display; use std::error::Error; /// This enumeration is a list of all possible errors that can happen during /// [`cnt_sort`](trait.CountingSort.html#method.cnt_sort) or /// [`cnt_sort_min_max`](trait.CountingSort.html#method.cnt_sort_min_max). #[derive(Debug)] pub enum CountingSortError { /// The conversion from a value of the to-be-sorted type `T` into an /// index ([`usize`](https://doc.rust-lang.org/std/primitive.usize.html)) failed. /// Most likely due to an overflow happening. IntoIndexFailed(&'static str), /// The iterator is empty and therefore nothing can be sorted. IteratorEmpty(&'static str), /// The minimum value is equal to the maximum value, this means sorting is unnecessary. SortingUnnecessary(&'static str), /// The minimum value is larger than the maximum value, most likely due to calling /// [`cnt_sort_min_max`](trait.CountingSort.html#method.cnt_sort_min_max) with the switched /// parameters. MinValueLargerMaxValue(&'static str), /// The converted index is still larger than the length of the count value vector. This happens /// when the given maximum value is smaller than the actual maximum value when /// [`cnt_sort_min_max`](trait.CountingSort.html#method.cnt_sort_min_max) is used. IndexOutOfBounds(&'static str), } impl Display for CountingSortError { fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> fmt::Result { match self { CountingSortError::IntoIndexFailed(description) => description.fmt(f), CountingSortError::IteratorEmpty(description) => description.fmt(f), CountingSortError::SortingUnnecessary(description) => description.fmt(f), CountingSortError::MinValueLargerMaxValue(description) => description.fmt(f), CountingSortError::IndexOutOfBounds(description) => description.fmt(f), } } } impl Error for CountingSortError {} impl CountingSortError { /// Create IntoIndexFailed error when conversion to index failed. fn from_try_into_index_failed() -> CountingSortError { CountingSortError::IntoIndexFailed("Conversion into index failed") } /// Create IteratorEmpty error when the iterator is empty. fn from_empty_iterator() -> CountingSortError { CountingSortError::IteratorEmpty("There are no element available in the iterator") } /// Create SortingUnnecessary when minimum value equals maximum value. fn from_sorting_unnecessary() -> CountingSortError { CountingSortError::SortingUnnecessary( "Minimum value is identical to maximum value, therefore no sorting is necessary", ) } /// Create SortingUnnecessary when minimum value equals maximum value. fn from_min_value_larger_max_value() -> CountingSortError { CountingSortError::MinValueLargerMaxValue("Minimum value is larger than maximum value") } /// Create IndexOutOfBounds when minimum value equals maximum value. fn from_index_out_of_bounds() -> CountingSortError { CountingSortError::IndexOutOfBounds( "Index is out of bounds, most likely the given maximum value is too small", ) } } /// The interface for counting sort algorithm. /// /// Interface provides blanket implementation of all collections that implement /// the [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html) /// trait. These collections must also implement /// [`Clone`](https://doc.rust-lang.org/std/clone/trait.Clone.html), since the iterator is iterated several times, /// and [`Sized`](https://doc.rust-lang.org/std/marker/trait.Sized.html). If your collection does provide these, /// you can simply implement this trait "empty": /// /// ```rust,compile_fail /// impl CountingSort for YourType {} /// ``` /// /// However the intention of this trait is to provide an implementation of all collections that /// implement the /// [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html) /// trait like [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html). /// /// The types which are held by the collections must implement /// [`Ord`](https://doc.rust-lang.org/std/cmp/trait.Ord.html) in order to sort the elements, as well /// as [`Copy`](https://doc.rust-lang.org/std/marker/trait.Copy.html), since the elements are copied /// during the count phase as well as the re-order phase. Finally the type must implement the in this /// crate defined [`TryIntoIndex`](trait.TryIntoIndex.html) trait. pub trait CountingSort<'a, T> where T: Ord + Copy + TryIntoIndex + 'a, Self: Clone + Sized + Iterator<Item = &'a T>, { /// Sorts the elements in the /// [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html) /// with the counting sort algorithm. /// /// This sort is stable (i.e., does not reorder equal elements) and `O(n + d)` worst-case, /// where `d` is the distance between the maximum and minimum element in the collection. /// /// Memory usage is `O(n + d)` as well, since all elements of the collection are copied into a new /// [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) and the frequency of all /// elements in the collection are counted in a [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) /// of size `d`. /// /// **<span style="color:red">Caution:</span>** If distance `d` is large, than memory consumption is large /// and you process may run out of memory. /// /// This method iterates [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html) /// in the beginning to identify the maximum and mimumum value in order to identify the distance `d`. This means /// the runtime is longer due to this additional `n` iterations and the checks needed to identify the minimum and /// maximum values. /// /// # Example /// /// ```rust /// use counting_sort::CountingSort; /// /// let slice = [20000,-1000,17,333]; /// let sorted_vec_result = slice.iter().cnt_sort(); /// /// assert_eq!(vec![-1000,17,333,20000], sorted_vec_result.unwrap()); /// ``` /// /// # Errors /// /// * [`CountingSortError::IntoIndexFailed`](enum.CountingSortError.html#variant.IntoIndexFailed) when /// converting into an index fails, this could happen if the distance `d` is larger than /// [`usize::max_value`](https://doc.rust-lang.org/std/primitive.usize.html#method.max_value) /// * [`CountingSortError::IteratorEmpty`](enum.CountingSortError.html#variant.IteratorEmpty) when the iterator /// is empty (and there is nothing to sort) /// * [`CountingSortError::SortingUnnecessary`](enum.CountingSortError.html#variant.SortingUnnecessary)] when /// the minimum value is equal to the maximum value, this means all values are essentially equal and no sorting /// is necessary fn cnt_sort(self) -> Result<Vec<T>, CountingSortError> { counting_sort(self) } /// Sorts the elements in the /// [`Iterator`](https://doc.rust-lang.org/std/iter/trait.Iterator.html) /// with the counting sort algorithm given the minimum and maximum element of the collection. /// /// This sort is stable (i.e., does not reorder equal elements) and `O(n + d)` worst-case, /// where `d` is the distance between the maximum and minimum element in the collection. /// /// Memory usage is `O(n + d)` as well, since all elements of the collection are copied into a new /// [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) and the frequency of all /// elements in the collection are counted in a [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) /// of size `d`. /// /// **<span style="color:red">Caution:</span>** If distance `d` is large, than memory consumption is large /// and you process may run out of memory. /// /// This method uses the given minimum and maximum element and therefore does not need to iterate the iterator /// to identify the minimum and maximum element. /// /// **<span style="color:red">Caution:</span>** If any element is either larger than the given maximum value /// or smaller than the given minimum value, the method will return with an error. Only use this method if /// you know these values. /// /// # Example /// /// ```rust /// use std::collections::LinkedList; /// use counting_sort::CountingSort; /// /// let mut list = LinkedList::new(); /// list.push_back(1000001); /// list.push_back(1000003); /// list.push_back(1000002); /// list.push_back(1000000); /// /// let sorted_vec_result = list.iter().cnt_sort_min_max(&1000000, &1000003); /// /// assert_eq!(vec![1000000, 1000001, 1000002, 1000003], sorted_vec_result.unwrap()); /// /// // minimum value incorrect /// let error = list.iter().cnt_sort_min_max(&1000001, &1000003); /// assert!(error.is_err()); /// ``` /// /// # Errors /// /// * [`CountingSortError::IntoIndexFailed`](enum.CountingSortError.html#variant.IntoIndexFailed) when /// converting into an index fails, this could happen if the distance `d` is larger than /// [`usize::max_value`](https://doc.rust-lang.org/std/primitive.usize.html#method.max_value) /// * [`CountingSortError::SortingUnnecessary`](enum.CountingSortError.html#variant.SortingUnnecessary)] when /// the minimum value is equal to the maximum value, this means all values are essentially equal and no sorting /// is necessary /// * [`CountingSortError::MinValueLargerMaxValue`](enum.CountingSortError.html#variant.MinValueLargerMaxValue)] when /// the given minimum value is larger than the given maximum value /// * [`CountingSortError::IndexOutOfBounds`](enum.CountingSortError.html#variant.IndexOutOfBounds)] when /// the given maximum value is smaller than the actual maximum value of the collection fn cnt_sort_min_max(self, min_value: &T, max_value: &T) -> Result<Vec<T>, CountingSortError> { counting_sort_min_max(self, min_value, max_value) } } // Counting sort implementation for ITER with trait bound Iterator. // This enables that CountingSort is implemented for all implementors of // Iterator, especially for Vec, LinkedList and slice. impl<'a, T, ITER> CountingSort<'a, T> for ITER where T: Ord + Copy + TryIntoIndex + 'a, ITER: Sized + Iterator<Item = &'a T> + Clone, { } /// The interface for converting values into an index. /// /// Index is always [`usize`](https://doc.rust-lang.org/std/primitive.usize.html). Unfortunatelly /// [`TryInto`](https://doc.rust-lang.org/std/convert/trait.TryInto.html) for /// [`usize`](https://doc.rust-lang.org/std/primitive.usize.html) is not sufficient since signed /// integers overflow when calculating `max_value - min_value`. Therefore this trait was added to /// implement an non-overflowing conversion to [`usize`](https://doc.rust-lang.org/std/primitive.usize.html). /// /// You can implement this trait yourself as long as there is a natural conversion from your type to /// [`usize`](https://doc.rust-lang.org/std/primitive.usize.html). However it must hold for your type that if /// `t_1 <= t_2` then `YourType::try_into_index(t_1, min_value)? <= YourType::try_into_index(t_2, min_value)?`. /// Also consider that the size [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) that holds the /// frequency of all elements in the collection is calculated like this /// /// ```rust,compile_fail /// let length = YourType::try_into_index(max_value,min_value)? + 1; /// ``` /// /// It is not highly recommended to do this if your type's order is not simply dependent on one integer value /// of your struct. /// /// # Example /// /// ```rust /// use core::cmp::{Ord, Ordering}; /// use counting_sort::TryIntoIndex; /// /// #[derive(Copy, Clone)] /// struct Person { /// name: &'static str, /// id: usize /// } /// /// impl Ord for Person { /// fn cmp(&self, other: &Self) -> Ordering { /// self.id.cmp(&other.id) /// } /// } /// /// impl PartialOrd for Person { /// fn partial_cmp(&self, other: &Self) -> Option<Ordering> { /// Some(self.cmp(other)) /// } /// } /// /// impl PartialEq for Person { /// fn eq(&self, other: &Self) -> bool { /// self.id == other.id /// } /// } /// /// impl Eq for Person {} /// /// impl TryIntoIndex for Person { /// type Error = &'static str; /// /// fn try_into_index(value: &Self, min_value: &Self) -> Result<usize, Self::Error> { /// Ok(value.id - min_value.id) /// } /// } /// /// let john = Person { name: "John", id: 1234 }; /// let min_value = Person { name: "Jim", id: 234 }; /// let index_result = Person::try_into_index(&john, &min_value); /// assert!(index_result.is_ok()); /// assert_eq!(1000, index_result.unwrap_or(0)); /// ``` pub trait TryIntoIndex { /// The type returned whenever the conversion into an index failed. type Error; /// Tries to convert the value into an index. /// /// The `min_value` parameter is for calculating the offset between the actual value /// and the minimum value. This concept is used in order to only allocate a /// [`Vec`](https://doc.rust-lang.org/std/vec/struct.Vec.html) that only covers the /// distance between the maximum value and the minimum value of the collection. fn try_into_index(value: &Self, min_value: &Self) -> Result<usize, Self::Error>; } // Macro is needed to implement TryIntoIndex for signed integers, which can overflow during // index conversion. macro_rules! try_into_index_impl_for_signed { ($smaller_int:ty,$larger_int:ty) => { impl TryIntoIndex for $smaller_int { type Error = <$larger_int as TryInto<usize>>::Error; fn try_into_index(value: &Self, min_value: &Self) -> Result<usize, Self::Error> { // This conversion can only fail, if the larger integer type has a larger maximum // value then usize. To-be-converted value is always be >= 0 as long as min_value <= // value. <$larger_int>::try_into( // convert smaller signed integer into larger signed integer to // avoid integer overflow for the smaller integer. // Example: 127 - (-128) overflows in i8 since 255 > 127 = i8::max_value(). // It is safe to convert a smaller integer into a larger integer. <$larger_int>::from(*value) - <$larger_int>::from(*min_value), ) } } }; } // Macro used for unsigned integer implementations of TryIntoIndex. macro_rules! try_into_index_impl_for_unsigned { ($unsigned:ty) => { impl TryIntoIndex for $unsigned { type Error = <$unsigned as TryInto<usize>>::Error; #[inline] fn try_into_index(value: &Self, min_value: &Self) -> Result<usize, Self::Error> { // Unsigned integer (e.g. u32) could be larger than usize on some HW. <$unsigned>::try_into(*value - *min_value) } } }; } // Macro used for small unsigned integer implementations of TryIntoIndex. macro_rules! try_into_index_impl_for_small_unsigned { ($unsigned:ty) => { impl TryIntoIndex for $unsigned { type Error = CountingSortError; #[inline] fn try_into_index(value: &Self, min_value: &Self) -> Result<usize, Self::Error> { // u8 and u16 should always fit into an usize. Therefore no TryInto is needed. Ok(usize::from(*value - *min_value)) } } }; } // macro instances for signed integer implementation of TryIntoIndex try_into_index_impl_for_signed!(i8, i16); try_into_index_impl_for_signed!(i16, i32); try_into_index_impl_for_signed!(i32, i64); // i64 was not added, since i64 could be larger than usize and can also // result in huge memory consumption if the distance between max_value and // min_value of the collection is huge. // macro instances for small unsigned integer implementation of TryIntoIndex try_into_index_impl_for_small_unsigned!(u8); try_into_index_impl_for_small_unsigned!(u16); // macro instances for unsigned integer implementation of TryIntoIndex try_into_index_impl_for_unsigned!(u32); try_into_index_impl_for_unsigned!(usize); // u64 was not added, since i64 could be larger than usize and can also // result in huge memory consumption if the distance between max_value and // min_value of the collection is huge. #[inline] fn counting_sort<'a, ITER, T>(iterator: ITER) -> Result<Vec<T>, CountingSortError> where ITER: Iterator<Item = &'a T> + Clone, T: Ord + Copy + TryIntoIndex + 'a, { let optional_tuple = get_min_max(&mut iterator.clone()); if let Some((min_value,max_value)) = optional_tuple { counting_sort_min_max(iterator, min_value, max_value) } else { Err(CountingSortError::from_empty_iterator()) } } #[inline] fn counting_sort_min_max<'a, ITER, T>( iterator: ITER, min_value: &T, max_value: &T, ) -> Result<Vec<T>, CountingSortError> where ITER: Iterator<Item = &'a T> + Clone, T: Ord + Copy + TryIntoIndex + 'a, { if min_value == max_value { return Err(CountingSortError::from_sorting_unnecessary()); } if min_value > max_value { return Err(CountingSortError::from_min_value_larger_max_value()); } let mut count_vector = count_values(&mut iterator.clone(), min_value, max_value)?; calculate_prefix_sum(&mut count_vector); // last element of the count vector depicts the index-1 of the largest element, hence it is its length let length = *count_vector.last().unwrap(); // it's safe to unwrap, since vector has at least one element re_order(iterator, &mut count_vector, length, &min_value) } #[inline] fn re_order<'a, T, ITER>( iterator: ITER, count_vector: &mut Vec<usize>, length: usize, min_value: &T, ) -> Result<Vec<T>, CountingSortError> where T: Ord + Copy + TryIntoIndex + 'a, ITER: Iterator<Item = &'a T>, { let mut sorted_vector: Vec<T> = vec![*min_value; length]; for value in iterator { let index_count_vector_result = T::try_into_index(value, min_value); if index_count_vector_result.is_err() { return Err(CountingSortError::from_try_into_index_failed()); } else { // index_count_vector_result is ok, unwrapping is safe let index_count_vector = index_count_vector_result.unwrap_or(0); if index_count_vector >= count_vector.len() { return Err(CountingSortError::from_index_out_of_bounds()); } // /* Get the cumulative frequency of the value before this. The cumulative frequency of the preceeding value is the index of the first element with this value. In order to avoid checks for the index to be 0 (and therefore not to try to access the -1-th element) we allocated the 0-the element additionally so that we can now safely access it. Additionally it holds the index of the next element which equals the minimum value. */ let mut index = count_vector[index_count_vector]; sorted_vector[index] = *value; /* Increment the index so that successive elements with the same value do not override this one. This additionally ensures that the sort is stable. This actually increments the cumulative frequency of the preceeding value. However at the end of the sorting process this frequency will be the cumulative frequency of this value. */ index += 1; count_vector[index_count_vector] = index; } } Ok(sorted_vector) } #[inline] fn count_values<'a, ITER, T>( iterator: &mut ITER, min_value: &T, max_value: &T, ) -> Result<Vec<usize>, CountingSortError> where ITER: Iterator<Item = &'a T>, T: Ord + Copy + TryIntoIndex + 'a, { let distance_result = T::try_into_index(max_value, min_value); if distance_result.is_ok() { /* Length must hold all possible distinct values of the collection, this means the complete distance + 1. E.g. the distance of 0 and 255 is 255, but there 256 distinct values between 0 and 255. We allocate another value in our vector to represent the value that preceeds the minimum value. This value actually does not exist in the collection but is introduced as an optimization for enabling the stable sort of this algorithm without the need of a DoubleEndedIterator and the reverse iteration of the collection when the given collection is re-ordered. */ let length = distance_result.unwrap_or(0) + 2; // distance_result is okay so unwrapping is safe let mut count_vector: Vec<usize> = vec![0; length]; for value in iterator { let index_result = T::try_into_index(value, min_value); if index_result.is_err() { return Err(CountingSortError::from_try_into_index_failed()); } else { /* Always add + 1 to not use the 0-the element in the vector. This element is just allocated to optimize the re-ordering of the given collection later on. The 0-the element does in a way represent the value that preceeds the minimum value, i.e. this value does not exist in the given collection. */ let index = index_result.unwrap_or(0) + 1; // index_result is ok, unwrapping is safe if index >= count_vector.len() { return Err(CountingSortError::from_index_out_of_bounds()); } let new_count_value = count_vector[index] + 1; count_vector[index] = new_count_value; } } return Ok(count_vector); } Err(CountingSortError::from_try_into_index_failed()) } #[inline] fn calculate_prefix_sum(count_vector: &mut Vec<usize>) { let mut iterator = count_vector.iter_mut(); // skip first element let optional_first_element = iterator.next(); if let Some(first_element) = optional_first_element { let mut total = *first_element; for value in iterator { total += *value; *value = total; } } } #[inline] fn get_min_max<T, ITER>(iterator: &mut ITER) -> Option<(T, T)> where T: Ord + Copy, ITER: Iterator<Item = T>, { // consume first element to initialize as min and max value let min_value_optional = iterator.next(); if let Some(min_value) = min_value_optional { let tuple = iterator.fold((min_value, min_value), |(min_val, max_val), value| { (min(min_val, value), max(max_val, value)) }); return Some(tuple); } None } #[cfg_attr(tarpaulin, skip)] #[cfg(test)] mod unit_tests { use super::*; const TEST_ARRAY_MIN_VALUE: u8 = 1; const TEST_ARRAY_MAX_VALUE: u8 = 30; const TEST_ARRAY_UNSORTED: [u8; 30] = [ 13, 24, 27, 3, 10, 1, 9, 17, 6, 7, 3, 30, 14, 15, 2, 3, 7, 11, 21, 16, 7, 11, 21, 5, 23, 25, 26, 28, 28, 4, ]; const TEST_ARRAY_SORTED: [u8; 30] = [ 1, 2, 3, 3, 3, 4, 5, 6, 7, 7, 7, 9, 10, 11, 11, 13, 14, 15, 16, 17, 21, 21, 23, 24, 25, 26, 27, 28, 28, 30, ]; const TEST_COUNT_VALUES_ARRAY: [usize; 31] = [ 0, 1, 1, 3, 1, 1, 1, 3, 0, 1, 1, 2, 0, 1, 1, 1, 1, 1, 0, 0, 0, 2, 0, 1, 1, 1, 1, 1, 2, 0, 1, ]; const TEST_PREFIX_SUM_ARRAY: [usize; 31] = [ 0, 1, 2, 5, 6, 7, 8, 11, 11, 12, 13, 15, 15, 16, 17, 18, 19, 20, 20, 20, 20, 22, 22, 23, 24, 25, 26, 27, 29, 29, 30, ]; #[test] fn test_cnt_sort_i8_vector() { let test_vector: Vec<i8> = vec![2, -2, 1, -6]; let sorted_vector = counting_sort(test_vector.iter()).unwrap(); assert_eq!(vec![-6, -2, 1, 2], sorted_vector); } #[test] fn test_cnt_sort_i8_vector_with_overflow() { let test_vector: Vec<i8> = vec![2, -100, 50, -6]; let sorted_vector = counting_sort(test_vector.iter()).unwrap(); assert_eq!(vec![-100, -6, 2, 50], sorted_vector); } #[test] fn test_cnt_sort_u8_vector() { let mut test_vector = TEST_ARRAY_UNSORTED.to_vec(); test_vector = test_vector.iter().cnt_sort().unwrap(); let sorted_vector = TEST_ARRAY_SORTED.to_vec(); assert_eq!(sorted_vector, test_vector); } #[test] fn test_cnt_sort_min_max_u8_vector() { let mut test_vector = TEST_ARRAY_UNSORTED.to_vec(); test_vector = test_vector .iter() .cnt_sort_min_max(&TEST_ARRAY_MIN_VALUE, &TEST_ARRAY_MAX_VALUE) .unwrap(); let sorted_vector = TEST_ARRAY_SORTED.to_vec(); assert_eq!(sorted_vector, test_vector); } #[test] fn test_into_index_i8() { assert_eq!(255, i8::try_into_index(&127, &-128).unwrap()); assert_eq!(0, i8::try_into_index(&-128, &-128).unwrap()); assert_eq!(150, i8::try_into_index(&50, &-100).unwrap()); assert_eq!(50, i8::try_into_index(&-50, &-100).unwrap()); assert_eq!(27, i8::try_into_index(&127, &100).unwrap()); } #[test] fn test_into_index_i16() { assert_eq!(0xFFFF, i16::try_into_index(&32767, &-32768).unwrap()); assert_eq!(0, i16::try_into_index(&-32768, &-32768).unwrap()); assert_eq!(0, i16::try_into_index(&32767, &32767).unwrap()); } #[test] fn test_into_index_i32() { assert_eq!( 0xFFFFFFFF, i32::try_into_index(&2147483647, &-2147483648).unwrap() ); assert_eq!(0, i32::try_into_index(&-2147483648, &-2147483648).unwrap()); assert_eq!(1, i32::try_into_index(&-2147483647, &-2147483648).unwrap()); assert_eq!(0, i32::try_into_index(&2147483647, &2147483647).unwrap()); } #[test] fn test_into_index_u8() { assert_eq!(255, u8::try_into_index(&255, &0).unwrap()); assert_eq!(0, u8::try_into_index(&0, &0).unwrap()); assert_eq!(0, u8::try_into_index(&255, &255).unwrap()); assert_eq!(50, u8::try_into_index(&150, &100).unwrap()); assert_eq!(50, u8::try_into_index(&100, &50).unwrap()); assert_eq!(27, i8::try_into_index(&127, &100).unwrap()); } #[test] fn test_into_index_u16() { assert_eq!(0xFFFF, u16::try_into_index(&0xFFFF, &0).unwrap()); assert_eq!(0, u16::try_into_index(&0, &0).unwrap()); assert_eq!(0, u16::try_into_index(&0xFFFF, &0xFFFF).unwrap()); assert_eq!(1, u16::try_into_index(&0xFFFF, &0xFFFE).unwrap()); } #[test] fn test_into_index_u32() { assert_eq!(0xFFFFFFFF, u32::try_into_index(&0xFFFFFFFF, &0).unwrap()); assert_eq!(0, u32::try_into_index(&0, &0).unwrap()); assert_eq!(50, u32::try_into_index(&1000000, &999950).unwrap()); assert_eq!(50, u8::try_into_index(&100, &50).unwrap()); assert_eq!(27, i8::try_into_index(&127, &100).unwrap()); } #[test] fn test_counting_sort() { let test_vector: Vec<u8> = TEST_ARRAY_UNSORTED.to_vec(); let sorted_vector = counting_sort(test_vector.iter()).unwrap(); let expected_vector = TEST_ARRAY_SORTED.to_vec(); assert_eq!(expected_vector, sorted_vector); } #[test] fn test_counting_sort_min_max() { let test_vector: Vec<u8> = TEST_ARRAY_UNSORTED.to_vec(); let sorted_vector = counting_sort_min_max( test_vector.iter(), &TEST_ARRAY_MIN_VALUE, &TEST_ARRAY_MAX_VALUE, ) .unwrap(); let expected_vector = TEST_ARRAY_SORTED.to_vec(); assert_eq!(expected_vector, sorted_vector); } #[test] fn test_count_values() { let test_vector = TEST_ARRAY_UNSORTED.to_vec(); let result_count_value_vector = count_values( &mut test_vector.iter(), &TEST_ARRAY_MIN_VALUE, &TEST_ARRAY_MAX_VALUE, ); assert!(result_count_value_vector.is_ok()); let count_values_vector = result_count_value_vector.unwrap(); let expected_vector = TEST_COUNT_VALUES_ARRAY.to_vec(); assert_eq!(expected_vector, count_values_vector); } #[test] fn test_get_min_max_unsigned() { let test_vector: Vec<u8> = vec![1, 2, 3, 4]; let tuple = get_min_max(&mut test_vector.iter()); assert!(tuple.is_some()); let (min_value, max_value) = tuple.unwrap(); assert_eq!(1, *min_value); assert_eq!(4, *max_value); } #[test] fn test_get_min_max_without_min() { let test_vector: Vec<u8> = Vec::new(); let tuple = get_min_max(&mut test_vector.iter()); assert!(tuple.is_none()); } #[test] fn test_get_min_max_signed() { let test_vector: Vec<i8> = vec![-128, 2, 3, 127]; let tuple = get_min_max(&mut test_vector.iter()); assert!(tuple.is_some()); let (min_value, max_value) = tuple.unwrap(); assert_eq!(-128, *min_value); assert_eq!(127, *max_value); } #[test] fn test_calculate_prefix_sum_1() { let mut test_vector: Vec<usize> = vec![1; 4]; calculate_prefix_sum(&mut test_vector); assert_eq!(vec![1, 2, 3, 4], test_vector); } #[test] fn test_calculate_prefix_sum_2() { let mut test_vector: Vec<usize> = vec![1, 2, 3, 4, 5]; calculate_prefix_sum(&mut test_vector); assert_eq!(vec![1, 3, 6, 10, 15], test_vector); } #[test] fn test_calculate_prefix_sum_3() { let mut test_vector = TEST_COUNT_VALUES_ARRAY.to_vec(); calculate_prefix_sum(&mut test_vector); assert_eq!(TEST_PREFIX_SUM_ARRAY.to_vec(), test_vector); } #[test] fn test_re_order() { let test_vector = TEST_ARRAY_UNSORTED.to_vec(); let mut test_count_vector = TEST_PREFIX_SUM_ARRAY.to_vec(); let result_sorted_vector = re_order( test_vector.iter(), &mut test_count_vector, test_vector.len(), &TEST_ARRAY_MIN_VALUE, ); assert!(result_sorted_vector.is_ok()); let sorted_vector = result_sorted_vector.unwrap(); assert_eq!(TEST_ARRAY_SORTED.to_vec(), sorted_vector); } #[test] fn test_min_value_larger_max_value_error() { let test_vector = vec![1]; let result = counting_sort_min_max(test_vector.iter(), &1, &0); assert!(result.is_err()); assert_eq!( "Minimum value is larger than maximum value", format!("{}", result.unwrap_err()) ); } #[test] fn test_sorting_unnecessary_error() { let test_vector = vec![1]; let result = counting_sort_min_max(test_vector.iter(), &1, &1); assert!(result.is_err()); assert_eq!( "Minimum value is identical to maximum value, therefore no sorting is necessary", format!("{}", result.unwrap_err()) ); } #[test] fn test_empty_iterator_error() { let test_vector: Vec<u8> = vec![]; let result = counting_sort(test_vector.iter()); assert!(result.is_err()); assert_eq!( "There are no element available in the iterator", format!("{}", result.unwrap_err()) ); let test_vector: Vec<u8> = vec![]; let result = counting_sort_min_max(test_vector.iter(), &0, &1); assert!(result.is_ok()); assert_eq!(test_vector, result.unwrap()); } #[test] fn test_incorrect_given_min_max_values() { let vec = vec![4, 3, 2, 1]; let error = vec.iter().cnt_sort_min_max(&2, &4); assert!(error.is_err()); assert_eq!( "Conversion into index failed", format!("{}", error.unwrap_err()) ); let error = vec.iter().cnt_sort_min_max(&1, &3); assert!(error.is_err()); assert_eq!( "Index is out of bounds, most likely the given maximum value is too small", format!("{}", error.unwrap_err()) ); } #[test] fn test_try_into_error() { #[derive(Ord, PartialOrd, PartialEq, Eq, Copy, Clone, Debug)] struct ValueWithTryIntoError { value: u8, }; let min_value = ValueWithTryIntoError { value: 0 }; let max_value = ValueWithTryIntoError { value: u8::max_value(), }; impl TryIntoIndex for ValueWithTryIntoError { type Error = String; fn try_into_index(_value: &Self, _min_value: &Self) -> Result<usize, Self::Error> { Err(String::from("TryInto always fails")) } } let test_vector: Vec<ValueWithTryIntoError> = Vec::new(); let result = counting_sort_min_max(test_vector.iter(), &min_value, &max_value); assert!(result.is_err()); assert_eq!( CountingSortError::from_try_into_index_failed().to_string(), result.unwrap_err().to_string() ); let mut count_vector = vec![0, 0]; let test_vector = vec![max_value, min_value]; let result = re_order(test_vector.iter(), &mut count_vector, 2, &min_value); assert!(result.is_err()); assert_eq!( CountingSortError::from_try_into_index_failed().to_string(), result.unwrap_err().to_string() ); } #[test] fn test_empty_count_values_vector_is_impossible() { #[derive(Ord, PartialOrd, PartialEq, Eq, Copy, Clone, Debug)] struct ValueWithWrongSubstraction { value: usize, }; let min_value = ValueWithWrongSubstraction { value: 0 }; let max_value = ValueWithWrongSubstraction { value: usize::max_value(), }; impl TryIntoIndex for ValueWithWrongSubstraction { type Error = String; fn try_into_index(_value: &Self, _min_value: &Self) -> Result<usize, Self::Error> { Ok(0) } } let test_vector: Vec<ValueWithWrongSubstraction> = Vec::new(); let result = counting_sort_min_max(test_vector.iter(), &min_value, &max_value); assert!(result.is_ok()); assert_eq!(test_vector, result.unwrap()); } #[test] fn test_re_order_index_out_of_bounds_error() { let vec = vec![1, 2]; let mut count_vector = vec![1]; let result = re_order(vec.iter(), &mut count_vector, 2, &1); assert!(result.is_err()); assert_eq!( CountingSortError::from_index_out_of_bounds().to_string(), result.unwrap_err().to_string() ); } } #[cfg_attr(tarpaulin, skip)] #[cfg(doctest)] macro_rules! doc_check { ($x:expr) => { #[doc = $x] extern {} }; } #[cfg_attr(tarpaulin, skip)] #[cfg(doctest)] doc_check!(include_str!("../README.md"));