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
//! A [disjoint-sets/union-find] implementation of a vector partitioned in sets that allows
//! for efficient iteration over elements of a set.
//!
//! The main struct of this crate is [`PartitionVec<T>`] which has the functionality of a `Vec<T>`
//! and in addition devides the elements of this vector in sets.
//! The elements each start in their own set and sets can be joined with the [`union`] method.
//! You can check if elements share a set with the [`same_set`] method and iterate on the elements
//! in a set with the [`set`] method.
//! The [`union`] and [`same_set`] methods are extremely fast and have an amortized complexity of
//! `O(α(n))` where 'α' is the inverse Ackermann function and `n` is the length.
//! This complexity is proven to be optimal and `α(n)` has value below 5 for any `n`
//! that can be written in the observable universe.
//! The next element of the iterator returned by [`set`] is found in `O(1)` time.
//!
//! The Disjoint-Sets algorithm is used in high-performance implementations of unification.
//! It is also a key component in implementing Kruskal's algorithm to find the minimum spanning
//! tree of a graph.
//!
//! This implementation stores three integers as `usize` values for every element in the
//! [`PartitionVec<T>`], two values are needed to get the best complexity of the Disjoint-Sets
//! algorithm and the third is used to allow iteration over sets and other methods like the
//! [`make_singleton`] method that removes the element of its current set and gives it its own set.
//!
//! A more compact implementation is included that has the same functionality but only needs to
//! store an additional two `usize` values instead of three for every element.
//! This is done by using a few bits of these two integers to store the third.
//! Because this third value is always very small we only need three bits on a 32 or 64 bit system.
//! This does mean that the maximum amounts of elements stored on 32 and 64 bit systems are
//! 536 870 912 and 2 305 843 009 213 693 952 respectively.
//! This limit should never be reached under any normal circumstances but if you do the struct
//! will panic.
//! This representation can be enabled by adding the following to your `Cargo.toml` file:
//! ```toml
//! [dependencies.partitions]
//! version = "0.2"
//! features = ["compact"]
//! ```
//!
//! [disjoint-sets/union-find]: https://en.wikipedia.org/wiki/Disjoint-set_data_structure
//! [`PartitionVec<T>`]: partition_vec/struct.PartitionVec.html
//! [`union`]: partition_vec/struct.PartitionVec.html#method.union
//! [`same_set`]: partition_vec/struct.PartitionVec.html#method.same_set
//! [`set`]: partition_vec/struct.PartitionVec.html#method.set
//! [`make_singleton`]: partition_vec/struct.PartitionVec.html#method.make_singleton
extern crate bit_vec;
extern crate rayon;
/// We count the amount of expresions given to this macro.
/// A convenient macro to create a `BitVec` similar to `vec!`.
pub use PartitionVec;
/// This takes an mutable reference and return a mutable reference with a different lifetime.
///
/// This function is highly unsafe and every use of this function will have a
/// comment explaining why it is necessary.
/// The main motivation for making a function for this is that the code is not
/// intuitive and this makes the intend clearer.
unsafe