diff options
Diffstat (limited to 'src/liballoc')
| -rw-r--r-- | src/liballoc/benches/btree/set.rs | 114 | ||||
| -rw-r--r-- | src/liballoc/collections/btree/set.rs | 298 | ||||
| -rw-r--r-- | src/liballoc/tests/btree/set.rs | 61 |
3 files changed, 351 insertions, 122 deletions
diff --git a/src/liballoc/benches/btree/set.rs b/src/liballoc/benches/btree/set.rs index 08e1db5fbb7..6357ea3ea11 100644 --- a/src/liballoc/benches/btree/set.rs +++ b/src/liballoc/benches/btree/set.rs @@ -3,59 +3,49 @@ use std::collections::BTreeSet; use rand::{thread_rng, Rng}; use test::{black_box, Bencher}; -fn random(n1: u32, n2: u32) -> [BTreeSet<usize>; 2] { +fn random(n: usize) -> BTreeSet<usize> { let mut rng = thread_rng(); - let mut set1 = BTreeSet::new(); - let mut set2 = BTreeSet::new(); - for _ in 0..n1 { - let i = rng.gen::<usize>(); - set1.insert(i); + let mut set = BTreeSet::new(); + while set.len() < n { + set.insert(rng.gen()); } - for _ in 0..n2 { - let i = rng.gen::<usize>(); - set2.insert(i); - } - [set1, set2] + assert_eq!(set.len(), n); + set } -fn staggered(n1: u32, n2: u32) -> [BTreeSet<u32>; 2] { - let mut even = BTreeSet::new(); - let mut odd = BTreeSet::new(); - for i in 0..n1 { - even.insert(i * 2); - } - for i in 0..n2 { - odd.insert(i * 2 + 1); +fn neg(n: usize) -> BTreeSet<i32> { + let mut set = BTreeSet::new(); + for i in -(n as i32)..=-1 { + set.insert(i); } - [even, odd] + assert_eq!(set.len(), n); + set } -fn neg_vs_pos(n1: u32, n2: u32) -> [BTreeSet<i32>; 2] { - let mut neg = BTreeSet::new(); - let mut pos = BTreeSet::new(); - for i in -(n1 as i32)..=-1 { - neg.insert(i); - } - for i in 1..=(n2 as i32) { - pos.insert(i); +fn pos(n: usize) -> BTreeSet<i32> { + let mut set = BTreeSet::new(); + for i in 1..=(n as i32) { + set.insert(i); } - [neg, pos] + assert_eq!(set.len(), n); + set } -fn pos_vs_neg(n1: u32, n2: u32) -> [BTreeSet<i32>; 2] { - let mut neg = BTreeSet::new(); - let mut pos = BTreeSet::new(); - for i in -(n1 as i32)..=-1 { - neg.insert(i); - } - for i in 1..=(n2 as i32) { - pos.insert(i); + +fn stagger(n1: usize, factor: usize) -> [BTreeSet<u32>; 2] { + let n2 = n1 * factor; + let mut sets = [BTreeSet::new(), BTreeSet::new()]; + for i in 0..(n1 + n2) { + let b = i % (factor + 1) != 0; + sets[b as usize].insert(i as u32); } - [pos, neg] + assert_eq!(sets[0].len(), n1); + assert_eq!(sets[1].len(), n2); + sets } -macro_rules! set_intersection_bench { - ($name: ident, $sets: expr) => { +macro_rules! set_bench { + ($name: ident, $set_func: ident, $result_func: ident, $sets: expr) => { #[bench] pub fn $name(b: &mut Bencher) { // setup @@ -63,26 +53,36 @@ macro_rules! set_intersection_bench { // measure b.iter(|| { - let x = sets[0].intersection(&sets[1]).count(); + let x = sets[0].$set_func(&sets[1]).$result_func(); black_box(x); }) } }; } -set_intersection_bench! {intersect_random_100, random(100, 100)} -set_intersection_bench! {intersect_random_10k, random(10_000, 10_000)} -set_intersection_bench! {intersect_random_10_vs_10k, random(10, 10_000)} -set_intersection_bench! {intersect_random_10k_vs_10, random(10_000, 10)} -set_intersection_bench! {intersect_staggered_100, staggered(100, 100)} -set_intersection_bench! {intersect_staggered_10k, staggered(10_000, 10_000)} -set_intersection_bench! {intersect_staggered_10_vs_10k, staggered(10, 10_000)} -set_intersection_bench! {intersect_staggered_10k_vs_10, staggered(10_000, 10)} -set_intersection_bench! {intersect_neg_vs_pos_100, neg_vs_pos(100, 100)} -set_intersection_bench! {intersect_neg_vs_pos_10k, neg_vs_pos(10_000, 10_000)} -set_intersection_bench! {intersect_neg_vs_pos_10_vs_10k,neg_vs_pos(10, 10_000)} -set_intersection_bench! {intersect_neg_vs_pos_10k_vs_10,neg_vs_pos(10_000, 10)} -set_intersection_bench! {intersect_pos_vs_neg_100, pos_vs_neg(100, 100)} -set_intersection_bench! {intersect_pos_vs_neg_10k, pos_vs_neg(10_000, 10_000)} -set_intersection_bench! {intersect_pos_vs_neg_10_vs_10k,pos_vs_neg(10, 10_000)} -set_intersection_bench! {intersect_pos_vs_neg_10k_vs_10,pos_vs_neg(10_000, 10)} +set_bench! {intersection_100_neg_vs_100_pos, intersection, count, [neg(100), pos(100)]} +set_bench! {intersection_100_neg_vs_10k_pos, intersection, count, [neg(100), pos(10_000)]} +set_bench! {intersection_100_pos_vs_100_neg, intersection, count, [pos(100), neg(100)]} +set_bench! {intersection_100_pos_vs_10k_neg, intersection, count, [pos(100), neg(10_000)]} +set_bench! {intersection_10k_neg_vs_100_pos, intersection, count, [neg(10_000), pos(100)]} +set_bench! {intersection_10k_neg_vs_10k_pos, intersection, count, [neg(10_000), pos(10_000)]} +set_bench! {intersection_10k_pos_vs_100_neg, intersection, count, [pos(10_000), neg(100)]} +set_bench! {intersection_10k_pos_vs_10k_neg, intersection, count, [pos(10_000), neg(10_000)]} +set_bench! {intersection_random_100_vs_100, intersection, count, [random(100), random(100)]} +set_bench! {intersection_random_100_vs_10k, intersection, count, [random(100), random(10_000)]} +set_bench! {intersection_random_10k_vs_100, intersection, count, [random(10_000), random(100)]} +set_bench! {intersection_random_10k_vs_10k, intersection, count, [random(10_000), random(10_000)]} +set_bench! {intersection_staggered_100_vs_100, intersection, count, stagger(100, 1)} +set_bench! {intersection_staggered_10k_vs_10k, intersection, count, stagger(10_000, 1)} +set_bench! {intersection_staggered_100_vs_10k, intersection, count, stagger(100, 100)} +set_bench! {difference_random_100_vs_100, difference, count, [random(100), random(100)]} +set_bench! {difference_random_100_vs_10k, difference, count, [random(100), random(10_000)]} +set_bench! {difference_random_10k_vs_100, difference, count, [random(10_000), random(100)]} +set_bench! {difference_random_10k_vs_10k, difference, count, [random(10_000), random(10_000)]} +set_bench! {difference_staggered_100_vs_100, difference, count, stagger(100, 1)} +set_bench! {difference_staggered_10k_vs_10k, difference, count, stagger(10_000, 1)} +set_bench! {difference_staggered_100_vs_10k, difference, count, stagger(100, 100)} +set_bench! {is_subset_100_vs_100, is_subset, clone, [pos(100), pos(100)]} +set_bench! {is_subset_100_vs_10k, is_subset, clone, [pos(100), pos(10_000)]} +set_bench! {is_subset_10k_vs_100, is_subset, clone, [pos(10_000), pos(100)]} +set_bench! {is_subset_10k_vs_10k, is_subset, clone, [pos(10_000), pos(10_000)]} diff --git a/src/liballoc/collections/btree/set.rs b/src/liballoc/collections/btree/set.rs index 2be6455ad59..e71767077ca 100644 --- a/src/liballoc/collections/btree/set.rs +++ b/src/liballoc/collections/btree/set.rs @@ -3,7 +3,7 @@ use core::borrow::Borrow; use core::cmp::Ordering::{self, Less, Greater, Equal}; -use core::cmp::{min, max}; +use core::cmp::max; use core::fmt::{self, Debug}; use core::iter::{Peekable, FromIterator, FusedIterator}; use core::ops::{BitOr, BitAnd, BitXor, Sub, RangeBounds}; @@ -118,17 +118,36 @@ pub struct Range<'a, T: 'a> { /// [`difference`]: struct.BTreeSet.html#method.difference #[stable(feature = "rust1", since = "1.0.0")] pub struct Difference<'a, T: 'a> { - a: Peekable<Iter<'a, T>>, - b: Peekable<Iter<'a, T>>, + inner: DifferenceInner<'a, T>, +} +enum DifferenceInner<'a, T: 'a> { + Stitch { + self_iter: Iter<'a, T>, + other_iter: Peekable<Iter<'a, T>>, + }, + Search { + self_iter: Iter<'a, T>, + other_set: &'a BTreeSet<T>, + }, } #[stable(feature = "collection_debug", since = "1.17.0")] impl<T: fmt::Debug> fmt::Debug for Difference<'_, T> { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - f.debug_tuple("Difference") - .field(&self.a) - .field(&self.b) - .finish() + match &self.inner { + DifferenceInner::Stitch { + self_iter, + other_iter, + } => f + .debug_tuple("Difference") + .field(&self_iter) + .field(&other_iter) + .finish(), + DifferenceInner::Search { + self_iter, + other_set: _, + } => f.debug_tuple("Difference").field(&self_iter).finish(), + } } } @@ -164,17 +183,36 @@ impl<T: fmt::Debug> fmt::Debug for SymmetricDifference<'_, T> { /// [`intersection`]: struct.BTreeSet.html#method.intersection #[stable(feature = "rust1", since = "1.0.0")] pub struct Intersection<'a, T: 'a> { - a: Peekable<Iter<'a, T>>, - b: Peekable<Iter<'a, T>>, + inner: IntersectionInner<'a, T>, +} +enum IntersectionInner<'a, T: 'a> { + Stitch { + small_iter: Iter<'a, T>, // for size_hint, should be the smaller of the sets + other_iter: Iter<'a, T>, + }, + Search { + small_iter: Iter<'a, T>, + large_set: &'a BTreeSet<T>, + }, } #[stable(feature = "collection_debug", since = "1.17.0")] impl<T: fmt::Debug> fmt::Debug for Intersection<'_, T> { fn fmt(&self, f: &mut fmt::Formatter<'_>) -> fmt::Result { - f.debug_tuple("Intersection") - .field(&self.a) - .field(&self.b) - .finish() + match &self.inner { + IntersectionInner::Stitch { + small_iter, + other_iter, + } => f + .debug_tuple("Intersection") + .field(&small_iter) + .field(&other_iter) + .finish(), + IntersectionInner::Search { + small_iter, + large_set: _, + } => f.debug_tuple("Intersection").field(&small_iter).finish(), + } } } @@ -201,6 +239,14 @@ impl<T: fmt::Debug> fmt::Debug for Union<'_, T> { } } +// This constant is used by functions that compare two sets. +// It estimates the relative size at which searching performs better +// than iterating, based on the benchmarks in +// https://github.com/ssomers/rust_bench_btreeset_intersection; +// It's used to divide rather than multiply sizes, to rule out overflow, +// and it's a power of two to make that division cheap. +const ITER_PERFORMANCE_TIPPING_SIZE_DIFF: usize = 16; + impl<T: Ord> BTreeSet<T> { /// Makes a new `BTreeSet` with a reasonable choice of B. /// @@ -268,9 +314,24 @@ impl<T: Ord> BTreeSet<T> { /// ``` #[stable(feature = "rust1", since = "1.0.0")] pub fn difference<'a>(&'a self, other: &'a BTreeSet<T>) -> Difference<'a, T> { - Difference { - a: self.iter().peekable(), - b: other.iter().peekable(), + if self.len() > other.len() / ITER_PERFORMANCE_TIPPING_SIZE_DIFF { + // Self is bigger than or not much smaller than other set. + // Iterate both sets jointly, spotting matches along the way. + Difference { + inner: DifferenceInner::Stitch { + self_iter: self.iter(), + other_iter: other.iter().peekable(), + }, + } + } else { + // Self is much smaller than other set, or both sets are empty. + // Iterate the small set, searching for matches in the large set. + Difference { + inner: DifferenceInner::Search { + self_iter: self.iter(), + other_set: other, + }, + } } } @@ -326,9 +387,29 @@ impl<T: Ord> BTreeSet<T> { /// ``` #[stable(feature = "rust1", since = "1.0.0")] pub fn intersection<'a>(&'a self, other: &'a BTreeSet<T>) -> Intersection<'a, T> { - Intersection { - a: self.iter().peekable(), - b: other.iter().peekable(), + let (small, other) = if self.len() <= other.len() { + (self, other) + } else { + (other, self) + }; + if small.len() > other.len() / ITER_PERFORMANCE_TIPPING_SIZE_DIFF { + // Small set is not much smaller than other set. + // Iterate both sets jointly, spotting matches along the way. + Intersection { + inner: IntersectionInner::Stitch { + small_iter: small.iter(), + other_iter: other.iter(), + }, + } + } else { + // Big difference in number of elements, or both sets are empty. + // Iterate the small set, searching for matches in the large set. + Intersection { + inner: IntersectionInner::Search { + small_iter: small.iter(), + large_set: other, + }, + } } } @@ -462,28 +543,44 @@ impl<T: Ord> BTreeSet<T> { /// ``` #[stable(feature = "rust1", since = "1.0.0")] pub fn is_subset(&self, other: &BTreeSet<T>) -> bool { - // Stolen from TreeMap - let mut x = self.iter(); - let mut y = other.iter(); - let mut a = x.next(); - let mut b = y.next(); - while a.is_some() { - if b.is_none() { - return false; - } + // Same result as self.difference(other).next().is_none() + // but the 3 paths below are faster (in order: hugely, 20%, 5%). + if self.len() > other.len() { + false + } else if self.len() > other.len() / ITER_PERFORMANCE_TIPPING_SIZE_DIFF { + // Self is not much smaller than other set. + // Stolen from TreeMap + let mut x = self.iter(); + let mut y = other.iter(); + let mut a = x.next(); + let mut b = y.next(); + while a.is_some() { + if b.is_none() { + return false; + } - let a1 = a.unwrap(); - let b1 = b.unwrap(); + let a1 = a.unwrap(); + let b1 = b.unwrap(); - match b1.cmp(a1) { - Less => (), - Greater => return false, - Equal => a = x.next(), - } + match b1.cmp(a1) { + Less => (), + Greater => return false, + Equal => a = x.next(), + } - b = y.next(); + b = y.next(); + } + true + } else { + // Big difference in number of elements, or both sets are empty. + // Iterate the small set, searching for matches in the large set. + for next in self { + if !other.contains(next) { + return false; + } + } + true } - true } /// Returns `true` if the set is a superset of another, @@ -1001,8 +1098,22 @@ fn cmp_opt<T: Ord>(x: Option<&T>, y: Option<&T>, short: Ordering, long: Ordering impl<T> Clone for Difference<'_, T> { fn clone(&self) -> Self { Difference { - a: self.a.clone(), - b: self.b.clone(), + inner: match &self.inner { + DifferenceInner::Stitch { + self_iter, + other_iter, + } => DifferenceInner::Stitch { + self_iter: self_iter.clone(), + other_iter: other_iter.clone(), + }, + DifferenceInner::Search { + self_iter, + other_set, + } => DifferenceInner::Search { + self_iter: self_iter.clone(), + other_set, + }, + }, } } } @@ -1011,24 +1122,52 @@ impl<'a, T: Ord> Iterator for Difference<'a, T> { type Item = &'a T; fn next(&mut self) -> Option<&'a T> { - loop { - match cmp_opt(self.a.peek(), self.b.peek(), Less, Less) { - Less => return self.a.next(), - Equal => { - self.a.next(); - self.b.next(); - } - Greater => { - self.b.next(); + match &mut self.inner { + DifferenceInner::Stitch { + self_iter, + other_iter, + } => { + let mut self_next = self_iter.next()?; + loop { + match other_iter + .peek() + .map_or(Less, |other_next| Ord::cmp(self_next, other_next)) + { + Less => return Some(self_next), + Equal => { + self_next = self_iter.next()?; + other_iter.next(); + } + Greater => { + other_iter.next(); + } + } } } + DifferenceInner::Search { + self_iter, + other_set, + } => loop { + let self_next = self_iter.next()?; + if !other_set.contains(&self_next) { + return Some(self_next); + } + }, } } fn size_hint(&self) -> (usize, Option<usize>) { - let a_len = self.a.len(); - let b_len = self.b.len(); - (a_len.saturating_sub(b_len), Some(a_len)) + let (self_len, other_len) = match &self.inner { + DifferenceInner::Stitch { + self_iter, + other_iter + } => (self_iter.len(), other_iter.len()), + DifferenceInner::Search { + self_iter, + other_set + } => (self_iter.len(), other_set.len()), + }; + (self_len.saturating_sub(other_len), Some(self_len)) } } @@ -1073,8 +1212,22 @@ impl<T: Ord> FusedIterator for SymmetricDifference<'_, T> {} impl<T> Clone for Intersection<'_, T> { fn clone(&self) -> Self { Intersection { - a: self.a.clone(), - b: self.b.clone(), + inner: match &self.inner { + IntersectionInner::Stitch { + small_iter, + other_iter, + } => IntersectionInner::Stitch { + small_iter: small_iter.clone(), + other_iter: other_iter.clone(), + }, + IntersectionInner::Search { + small_iter, + large_set, + } => IntersectionInner::Search { + small_iter: small_iter.clone(), + large_set, + }, + }, } } } @@ -1083,24 +1236,39 @@ impl<'a, T: Ord> Iterator for Intersection<'a, T> { type Item = &'a T; fn next(&mut self) -> Option<&'a T> { - loop { - match Ord::cmp(self.a.peek()?, self.b.peek()?) { - Less => { - self.a.next(); - } - Equal => { - self.b.next(); - return self.a.next(); - } - Greater => { - self.b.next(); + match &mut self.inner { + IntersectionInner::Stitch { + small_iter, + other_iter, + } => { + let mut small_next = small_iter.next()?; + let mut other_next = other_iter.next()?; + loop { + match Ord::cmp(small_next, other_next) { + Less => small_next = small_iter.next()?, + Greater => other_next = other_iter.next()?, + Equal => return Some(small_next), + } } } + IntersectionInner::Search { + small_iter, + large_set, + } => loop { + let small_next = small_iter.next()?; + if large_set.contains(&small_next) { + return Some(small_next); + } + }, } } fn size_hint(&self) -> (usize, Option<usize>) { - (0, Some(min(self.a.len(), self.b.len()))) + let min_len = match &self.inner { + IntersectionInner::Stitch { small_iter, .. } => small_iter.len(), + IntersectionInner::Search { small_iter, .. } => small_iter.len(), + }; + (0, Some(min_len)) } } diff --git a/src/liballoc/tests/btree/set.rs b/src/liballoc/tests/btree/set.rs index 4f5168f1ce5..d52814118b3 100644 --- a/src/liballoc/tests/btree/set.rs +++ b/src/liballoc/tests/btree/set.rs @@ -69,6 +69,20 @@ fn test_intersection() { check_intersection(&[11, 1, 3, 77, 103, 5, -5], &[2, 11, 77, -9, -42, 5, 3], &[3, 5, 11, 77]); + let large = (0..1000).collect::<Vec<_>>(); + check_intersection(&[], &large, &[]); + check_intersection(&large, &[], &[]); + check_intersection(&[-1], &large, &[]); + check_intersection(&large, &[-1], &[]); + check_intersection(&[0], &large, &[0]); + check_intersection(&large, &[0], &[0]); + check_intersection(&[999], &large, &[999]); + check_intersection(&large, &[999], &[999]); + check_intersection(&[1000], &large, &[]); + check_intersection(&large, &[1000], &[]); + check_intersection(&[11, 5000, 1, 3, 77, 8924, 103], + &large, + &[1, 3, 11, 77, 103]); } #[test] @@ -84,6 +98,18 @@ fn test_difference() { check_difference(&[-5, 11, 22, 33, 40, 42], &[-12, -5, 14, 23, 34, 38, 39, 50], &[11, 22, 33, 40, 42]); + let large = (0..1000).collect::<Vec<_>>(); + check_difference(&[], &large, &[]); + check_difference(&[-1], &large, &[-1]); + check_difference(&[0], &large, &[]); + check_difference(&[999], &large, &[]); + check_difference(&[1000], &large, &[1000]); + check_difference(&[11, 5000, 1, 3, 77, 8924, 103], + &large, + &[5000, 8924]); + check_difference(&large, &[], &large); + check_difference(&large, &[-1], &large); + check_difference(&large, &[1000], &large); } #[test] @@ -115,6 +141,41 @@ fn test_union() { } #[test] +// Only tests the simple function definition with respect to intersection +fn test_is_disjoint() { + let one = [1].into_iter().collect::<BTreeSet<_>>(); + let two = [2].into_iter().collect::<BTreeSet<_>>(); + assert!(one.is_disjoint(&two)); +} + +#[test] +// Also tests the trivial function definition of is_superset +fn test_is_subset() { + fn is_subset(a: &[i32], b: &[i32]) -> bool { + let set_a = a.iter().collect::<BTreeSet<_>>(); + let set_b = b.iter().collect::<BTreeSet<_>>(); + set_a.is_subset(&set_b) + } + + assert_eq!(is_subset(&[], &[]), true); + assert_eq!(is_subset(&[], &[1, 2]), true); + assert_eq!(is_subset(&[0], &[1, 2]), false); + assert_eq!(is_subset(&[1], &[1, 2]), true); + assert_eq!(is_subset(&[2], &[1, 2]), true); + assert_eq!(is_subset(&[3], &[1, 2]), false); + assert_eq!(is_subset(&[1, 2], &[1]), false); + assert_eq!(is_subset(&[1, 2], &[1, 2]), true); + assert_eq!(is_subset(&[1, 2], &[2, 3]), false); + let large = (0..1000).collect::<Vec<_>>(); + assert_eq!(is_subset(&[], &large), true); + assert_eq!(is_subset(&large, &[]), false); + assert_eq!(is_subset(&[-1], &large), false); + assert_eq!(is_subset(&[0], &large), true); + assert_eq!(is_subset(&[1, 2], &large), true); + assert_eq!(is_subset(&[999, 1000], &large), false); +} + +#[test] fn test_zip() { let mut x = BTreeSet::new(); x.insert(5); |