rust/src/libstd/hashmap.rs

1027 lines
31 KiB
Rust
Raw Normal View History

// Copyright 2013 The Rust Project Developers. See the COPYRIGHT
// file at the top-level directory of this distribution and at
// http://rust-lang.org/COPYRIGHT.
//
// 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.
2013-03-27 02:03:10 -05:00
//! An unordered map and set type implemented as hash tables
//!
//! The tables use a keyed hash with new random keys generated for each container, so the ordering
//! of a set of keys in a hash table is randomized.
#[mutable_doc];
2013-04-03 07:45:14 -05:00
use container::{Container, Mutable, Map, Set};
use cmp::{Eq, Equiv};
use hash::Hash;
2013-04-24 19:35:49 -05:00
use old_iter::BaseIter;
use old_iter;
2013-04-03 07:45:14 -05:00
use option::{None, Option, Some};
use rand::RngUtil;
use rand;
use uint;
use vec;
use kinds::Copy;
2013-05-04 08:54:58 -05:00
use util::{replace, unreachable};
2013-04-03 07:45:14 -05:00
static INITIAL_CAPACITY: uint = 32u; // 2^5
struct Bucket<K,V> {
hash: uint,
key: K,
value: V,
}
/// A hash map implementation which uses linear probing along with the SipHash
/// hash function for internal state. This means that the order of all hash maps
/// is randomized by keying each hash map randomly on creation.
///
/// It is required that the keys implement the `Eq` and `Hash` traits, although
/// this can frequently be achieved by just implementing the `Eq` and
/// `IterBytes` traits as `Hash` is automatically implemented for types that
/// implement `IterBytes`.
pub struct HashMap<K,V> {
2013-04-03 07:45:14 -05:00
priv k0: u64,
priv k1: u64,
priv resize_at: uint,
priv size: uint,
priv buckets: ~[Option<Bucket<K, V>>],
}
2013-04-03 07:45:14 -05:00
// We could rewrite FoundEntry to have type Option<&Bucket<K, V>>
// which would be nifty
enum SearchResult {
FoundEntry(uint), FoundHole(uint), TableFull
}
2013-01-24 21:00:58 -06:00
2013-04-03 07:45:14 -05:00
#[inline(always)]
fn resize_at(capacity: uint) -> uint {
((capacity as float) * 3. / 4.) as uint
}
/// Creates a new hash map with the specified capacity.
2013-04-03 07:45:14 -05:00
pub fn linear_map_with_capacity<K:Eq + Hash,V>(
initial_capacity: uint) -> HashMap<K, V> {
let mut r = rand::task_rng();
linear_map_with_capacity_and_keys(r.gen(), r.gen(),
2013-04-03 07:45:14 -05:00
initial_capacity)
}
fn linear_map_with_capacity_and_keys<K:Eq + Hash,V>(
k0: u64, k1: u64,
initial_capacity: uint) -> HashMap<K, V> {
2013-05-30 12:03:11 -05:00
let cap = uint::max(INITIAL_CAPACITY, initial_capacity);
HashMap {
2013-04-03 07:45:14 -05:00
k0: k0, k1: k1,
2013-05-30 12:03:11 -05:00
resize_at: resize_at(cap),
2013-04-03 07:45:14 -05:00
size: 0,
2013-05-30 12:03:11 -05:00
buckets: vec::from_fn(cap, |_| None)
}
2013-04-03 07:45:14 -05:00
}
impl<K:Hash + Eq,V> HashMap<K, V> {
#[inline(always)]
2013-04-03 07:45:14 -05:00
fn to_bucket(&self, h: uint) -> uint {
// A good hash function with entropy spread over all of the
// bits is assumed. SipHash is more than good enough.
h % self.buckets.len()
}
2013-04-03 07:45:14 -05:00
#[inline(always)]
fn next_bucket(&self, idx: uint, len_buckets: uint) -> uint {
let n = (idx + 1) % len_buckets;
debug!("next_bucket(%?, %?) = %?", idx, len_buckets, n);
n
}
2013-04-03 07:45:14 -05:00
#[inline(always)]
2013-05-02 17:33:18 -05:00
fn bucket_sequence(&self, hash: uint,
op: &fn(uint) -> bool) -> bool {
let start_idx = self.to_bucket(hash);
let len_buckets = self.buckets.len();
let mut idx = start_idx;
loop {
if !op(idx) { return false; }
2013-04-03 07:45:14 -05:00
idx = self.next_bucket(idx, len_buckets);
if idx == start_idx {
2013-05-02 17:33:18 -05:00
return true;
2013-04-03 07:45:14 -05:00
}
}
}
2013-04-03 07:45:14 -05:00
#[inline(always)]
fn bucket_for_key(&self, k: &K) -> SearchResult {
let hash = k.hash_keyed(self.k0, self.k1) as uint;
self.bucket_for_key_with_hash(hash, k)
}
2013-04-03 07:45:14 -05:00
#[inline(always)]
fn bucket_for_key_equiv<Q:Hash + Equiv<K>>(&self, k: &Q)
-> SearchResult {
2013-04-03 07:45:14 -05:00
let hash = k.hash_keyed(self.k0, self.k1) as uint;
self.bucket_for_key_with_hash_equiv(hash, k)
}
2013-04-03 07:45:14 -05:00
#[inline(always)]
fn bucket_for_key_with_hash(&self,
hash: uint,
k: &K)
-> SearchResult {
2013-05-02 17:33:18 -05:00
for self.bucket_sequence(hash) |i| {
2013-04-03 07:45:14 -05:00
match self.buckets[i] {
Some(ref bkt) => if bkt.hash == hash && *k == bkt.key {
return FoundEntry(i);
},
None => return FoundHole(i)
}
2013-05-02 17:33:18 -05:00
}
2013-04-03 07:45:14 -05:00
TableFull
}
2013-04-03 07:45:14 -05:00
#[inline(always)]
fn bucket_for_key_with_hash_equiv<Q:Equiv<K>>(&self,
hash: uint,
k: &Q)
-> SearchResult {
2013-05-02 17:33:18 -05:00
for self.bucket_sequence(hash) |i| {
2013-04-03 07:45:14 -05:00
match self.buckets[i] {
Some(ref bkt) => {
if bkt.hash == hash && k.equiv(&bkt.key) {
return FoundEntry(i);
2013-04-03 07:45:14 -05:00
}
},
None => return FoundHole(i)
}
2013-05-02 17:33:18 -05:00
}
2013-04-03 07:45:14 -05:00
TableFull
}
2013-04-03 07:45:14 -05:00
/// Expand the capacity of the array to the next power of two
/// and re-insert each of the existing buckets.
#[inline(always)]
fn expand(&mut self) {
let new_capacity = self.buckets.len() * 2;
self.resize(new_capacity);
}
2013-04-03 07:45:14 -05:00
/// Expands the capacity of the array and re-insert each of the
/// existing buckets.
fn resize(&mut self, new_capacity: uint) {
self.resize_at = resize_at(new_capacity);
2013-05-05 23:42:54 -05:00
let old_buckets = replace(&mut self.buckets,
vec::from_fn(new_capacity, |_| None));
2013-04-03 07:45:14 -05:00
self.size = 0;
2013-05-05 23:42:54 -05:00
do vec::consume(old_buckets) |_, bucket| {
2013-04-03 07:45:14 -05:00
self.insert_opt_bucket(bucket);
}
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
fn insert_opt_bucket(&mut self, bucket: Option<Bucket<K, V>>) {
match bucket {
Some(Bucket{hash: hash, key: key, value: value}) => {
self.insert_internal(hash, key, value);
}
2013-04-03 07:45:14 -05:00
None => {}
}
2013-04-03 07:45:14 -05:00
}
2013-04-10 15:11:35 -05:00
#[inline(always)]
fn value_for_bucket<'a>(&'a self, idx: uint) -> &'a V {
match self.buckets[idx] {
Some(ref bkt) => &bkt.value,
None => fail!("HashMap::find: internal logic error"),
2013-04-10 15:11:35 -05:00
}
}
#[inline(always)]
fn mut_value_for_bucket<'a>(&'a mut self, idx: uint) -> &'a mut V {
match self.buckets[idx] {
Some(ref mut bkt) => &mut bkt.value,
None => unreachable()
}
}
2013-04-03 07:45:14 -05:00
/// Inserts the key value pair into the buckets.
/// Assumes that there will be a bucket.
/// True if there was no previous entry with that key
2013-05-04 08:54:58 -05:00
fn insert_internal(&mut self, hash: uint, k: K, v: V) -> Option<V> {
2013-04-03 07:45:14 -05:00
match self.bucket_for_key_with_hash(hash, &k) {
TableFull => { fail!("Internal logic error"); }
2013-04-03 07:45:14 -05:00
FoundHole(idx) => {
debug!("insert fresh (%?->%?) at idx %?, hash %?",
k, v, idx, hash);
self.buckets[idx] = Some(Bucket{hash: hash, key: k,
value: v});
self.size += 1;
2013-05-04 08:54:58 -05:00
None
2013-04-03 07:45:14 -05:00
}
FoundEntry(idx) => {
debug!("insert overwrite (%?->%?) at idx %?, hash %?",
k, v, idx, hash);
2013-05-04 08:54:58 -05:00
match self.buckets[idx] {
None => { fail!("insert_internal: Internal logic error") }
2013-05-04 08:54:58 -05:00
Some(ref mut b) => {
b.hash = hash;
b.key = k;
Some(replace(&mut b.value, v))
}
}
}
}
2013-04-03 07:45:14 -05:00
}
2012-08-22 23:01:30 -05:00
2013-04-03 07:45:14 -05:00
fn pop_internal(&mut self, hash: uint, k: &K) -> Option<V> {
// Removing from an open-addressed hashtable
// is, well, painful. The problem is that
// the entry may lie on the probe path for other
// entries, so removing it would make you think that
// those probe paths are empty.
//
// To address this we basically have to keep walking,
// re-inserting entries we find until we reach an empty
// bucket. We know we will eventually reach one because
// we insert one ourselves at the beginning (the removed
// entry).
//
// I found this explanation elucidating:
// http://www.maths.lse.ac.uk/Courses/MA407/del-hash.pdf
let mut idx = match self.bucket_for_key_with_hash(hash, k) {
TableFull | FoundHole(_) => return None,
FoundEntry(idx) => idx
};
let len_buckets = self.buckets.len();
2013-05-05 23:42:54 -05:00
let bucket = replace(&mut self.buckets[idx], None);
2013-04-03 07:45:14 -05:00
let value = match bucket {
None => None,
2013-05-05 23:42:54 -05:00
Some(Bucket{value, _}) => {
2013-04-03 07:45:14 -05:00
Some(value)
},
};
/* re-inserting buckets may cause changes in size, so remember
what our new size is ahead of time before we start insertions */
let size = self.size - 1;
idx = self.next_bucket(idx, len_buckets);
while self.buckets[idx].is_some() {
2013-05-05 23:42:54 -05:00
let bucket = replace(&mut self.buckets[idx], None);
2013-04-03 07:45:14 -05:00
self.insert_opt_bucket(bucket);
idx = self.next_bucket(idx, len_buckets);
}
2013-04-03 07:45:14 -05:00
self.size = size;
2013-04-03 07:45:14 -05:00
value
}
2013-04-03 07:45:14 -05:00
fn search(&self, hash: uint,
op: &fn(x: &Option<Bucket<K, V>>) -> bool) {
let _ = self.bucket_sequence(hash, |i| op(&self.buckets[i]));
}
}
impl<K:Hash + Eq,V> Container for HashMap<K, V> {
2013-04-03 07:45:14 -05:00
/// Return the number of elements in the map
fn len(&const self) -> uint { self.size }
2013-01-23 15:47:27 -06:00
2013-04-03 07:45:14 -05:00
/// Return true if the map contains no elements
fn is_empty(&const self) -> bool { self.len() == 0 }
}
2013-01-21 20:59:19 -06:00
impl<K:Hash + Eq,V> Mutable for HashMap<K, V> {
2013-04-03 07:45:14 -05:00
/// Clear the map, removing all key-value pairs.
fn clear(&mut self) {
for uint::range(0, self.buckets.len()) |idx| {
self.buckets[idx] = None;
}
2013-04-03 07:45:14 -05:00
self.size = 0;
}
2013-04-03 07:45:14 -05:00
}
impl<K:Hash + Eq,V> Map<K, V> for HashMap<K, V> {
2013-04-03 07:45:14 -05:00
/// Return true if the map contains a value for the specified key
fn contains_key(&self, k: &K) -> bool {
match self.bucket_for_key(k) {
FoundEntry(_) => {true}
TableFull | FoundHole(_) => {false}
2013-01-21 17:22:03 -06:00
}
2013-04-03 07:45:14 -05:00
}
2013-01-21 17:22:03 -06:00
/// Visit all key-value pairs
2013-05-02 17:33:18 -05:00
fn each<'a>(&'a self, blk: &fn(&K, &'a V) -> bool) -> bool {
2013-05-20 11:40:29 -05:00
for self.buckets.each |bucket| {
for bucket.each |pair| {
if !blk(&pair.key, &pair.value) {
2013-05-02 17:33:18 -05:00
return false;
}
}
}
return true;
}
2013-04-03 07:45:14 -05:00
/// Visit all keys
2013-05-02 17:33:18 -05:00
fn each_key(&self, blk: &fn(k: &K) -> bool) -> bool {
self.each(|k, _| blk(k))
}
2013-04-03 07:45:14 -05:00
/// Visit all values
2013-05-02 17:33:18 -05:00
fn each_value<'a>(&'a self, blk: &fn(v: &'a V) -> bool) -> bool {
self.each(|_, v| blk(v))
}
2013-04-03 07:45:14 -05:00
/// Iterate over the map and mutate the contained values
2013-05-02 17:33:18 -05:00
fn mutate_values(&mut self, blk: &fn(&K, &mut V) -> bool) -> bool {
for uint::range(0, self.buckets.len()) |i| {
match self.buckets[i] {
Some(Bucket{key: ref key, value: ref mut value, _}) => {
if !blk(key, value) { return false; }
}
None => ()
}
}
return true;
}
2013-04-03 07:45:14 -05:00
/// Return a reference to the value corresponding to the key
2013-04-10 15:11:35 -05:00
fn find<'a>(&'a self, k: &K) -> Option<&'a V> {
match self.bucket_for_key(k) {
FoundEntry(idx) => Some(self.value_for_bucket(idx)),
TableFull | FoundHole(_) => None,
}
}
2013-04-03 07:45:14 -05:00
/// Return a mutable reference to the value corresponding to the key
fn find_mut<'a>(&'a mut self, k: &K) -> Option<&'a mut V> {
let idx = match self.bucket_for_key(k) {
FoundEntry(idx) => idx,
TableFull | FoundHole(_) => return None
};
Some(self.mut_value_for_bucket(idx))
}
2013-04-03 07:45:14 -05:00
/// Insert a key-value pair into the map. An existing value for a
/// key is replaced by the new value. Return true if the key did
/// not already exist in the map.
fn insert(&mut self, k: K, v: V) -> bool {
2013-05-04 08:54:58 -05:00
self.swap(k, v).is_none()
}
/// Remove a key-value pair from the map. Return true if the key
/// was present in the map, otherwise false.
fn remove(&mut self, k: &K) -> bool {
self.pop(k).is_some()
}
/// Insert a key-value pair from the map. If the key already had a value
/// present in the map, that value is returned. Otherwise None is returned.
fn swap(&mut self, k: K, v: V) -> Option<V> {
// this could be faster.
2013-04-03 07:45:14 -05:00
if self.size >= self.resize_at {
// n.b.: We could also do this after searching, so
// that we do not resize if this call to insert is
// simply going to update a key in place. My sense
// though is that it's worse to have to search through
// buckets to find the right spot twice than to just
// resize in this corner case.
self.expand();
}
let hash = k.hash_keyed(self.k0, self.k1) as uint;
self.insert_internal(hash, k, v)
2013-01-21 17:22:03 -06:00
}
2013-05-04 08:54:58 -05:00
/// Removes a key from the map, returning the value at the key if the key
/// was previously in the map.
fn pop(&mut self, k: &K) -> Option<V> {
let hash = k.hash_keyed(self.k0, self.k1) as uint;
self.pop_internal(hash, k)
2013-04-03 07:45:14 -05:00
}
}
2013-03-28 22:30:50 -05:00
impl<K: Hash + Eq, V> HashMap<K, V> {
/// Create an empty HashMap
pub fn new() -> HashMap<K, V> {
HashMap::with_capacity(INITIAL_CAPACITY)
2013-04-03 07:45:14 -05:00
}
/// Create an empty HashMap with space for at least `n` elements in
2013-04-03 07:45:14 -05:00
/// the hash table.
pub fn with_capacity(capacity: uint) -> HashMap<K, V> {
2013-04-03 07:45:14 -05:00
linear_map_with_capacity(capacity)
}
2013-04-03 07:45:14 -05:00
/// Reserve space for at least `n` elements in the hash table.
pub fn reserve_at_least(&mut self, n: uint) {
2013-04-03 07:45:14 -05:00
if n > self.buckets.len() {
let buckets = n * 4 / 3 + 1;
self.resize(uint::next_power_of_two(buckets));
}
2013-04-03 07:45:14 -05:00
}
/// Modify and return the value corresponding to the key in the map, or
/// insert and return a new value if it doesn't exist.
pub fn mangle<'a,A>(&'a mut self, k: K, a: A, not_found: &fn(&K, A) -> V,
found: &fn(&K, &mut V, A)) -> &'a mut V {
if self.size >= self.resize_at {
// n.b.: We could also do this after searching, so
// that we do not resize if this call to insert is
// simply going to update a key in place. My sense
// though is that it's worse to have to search through
// buckets to find the right spot twice than to just
// resize in this corner case.
self.expand();
}
let hash = k.hash_keyed(self.k0, self.k1) as uint;
let idx = match self.bucket_for_key_with_hash(hash, &k) {
TableFull => fail!("Internal logic error"),
FoundEntry(idx) => { found(&k, self.mut_value_for_bucket(idx), a); idx }
FoundHole(idx) => {
let v = not_found(&k, a);
self.buckets[idx] = Some(Bucket{hash: hash, key: k, value: v});
self.size += 1;
idx
}
};
self.mut_value_for_bucket(idx)
}
/// Return the value corresponding to the key in the map, or insert
/// and return the value if it doesn't exist.
pub fn find_or_insert<'a>(&'a mut self, k: K, v: V) -> &'a mut V {
self.mangle(k, v, |_k, a| a, |_k,_v,_a| ())
}
2013-04-03 07:45:14 -05:00
/// Return the value corresponding to the key in the map, or create,
/// insert, and return a new value if it doesn't exist.
pub fn find_or_insert_with<'a>(&'a mut self, k: K, f: &fn(&K) -> V)
-> &'a mut V {
self.mangle(k, (), |k,_a| f(k), |_k,_v,_a| ())
}
/// Insert a key-value pair into the map if the key is not already present.
/// Otherwise, modify the existing value for the key.
/// Returns the new or modified value for the key.
pub fn insert_or_update_with<'a>(&'a mut self, k: K, v: V,
f: &fn(&K, &mut V)) -> &'a mut V {
self.mangle(k, v, |_k,a| a, |k,v,_a| f(k,v))
}
/// Calls a function on each element of a hash map, destroying the hash
/// map in the process.
pub fn consume(&mut self, f: &fn(K, V)) {
2013-05-30 12:03:11 -05:00
let buckets = replace(&mut self.buckets,
vec::from_fn(INITIAL_CAPACITY, |_| None));
2013-04-03 07:45:14 -05:00
self.size = 0;
do vec::consume(buckets) |_, bucket| {
match bucket {
None => {},
2013-05-05 23:42:54 -05:00
Some(Bucket{key, value, _}) => {
2013-04-03 07:45:14 -05:00
f(key, value)
}
}
}
2013-04-03 07:45:14 -05:00
}
/// Retrieves a value for the given key, failing if the key is not
/// present.
pub fn get<'a>(&'a self, k: &K) -> &'a V {
2013-04-10 15:11:35 -05:00
match self.find(k) {
Some(v) => v,
None => fail!("No entry found for key: %?", k),
2013-04-10 15:11:35 -05:00
}
}
2013-04-03 07:45:14 -05:00
/// Return true if the map contains a value for the specified key,
/// using equivalence
pub fn contains_key_equiv<Q:Hash + Equiv<K>>(&self, key: &Q) -> bool {
2013-04-03 07:45:14 -05:00
match self.bucket_for_key_equiv(key) {
FoundEntry(_) => {true}
TableFull | FoundHole(_) => {false}
}
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
/// Return the value corresponding to the key in the map, using
/// equivalence
pub fn find_equiv<'a, Q:Hash + Equiv<K>>(&'a self, k: &Q)
-> Option<&'a V> {
2013-04-10 15:11:35 -05:00
match self.bucket_for_key_equiv(k) {
FoundEntry(idx) => Some(self.value_for_bucket(idx)),
TableFull | FoundHole(_) => None,
}
}
2013-04-03 07:45:14 -05:00
}
impl<K: Hash + Eq, V: Copy> HashMap<K, V> {
/// Like `find`, but returns a copy of the value.
pub fn find_copy(&self, k: &K) -> Option<V> {
self.find(k).map_consume(|v| copy *v)
}
/// Like `get`, but returns a copy of the value.
pub fn get_copy(&self, k: &K) -> V {
copy *self.get(k)
}
}
impl<K:Hash + Eq,V:Eq> Eq for HashMap<K, V> {
fn eq(&self, other: &HashMap<K, V>) -> bool {
2013-04-03 07:45:14 -05:00
if self.len() != other.len() { return false; }
2012-12-10 11:00:52 -06:00
for self.each |key, value| {
2013-04-03 07:45:14 -05:00
match other.find(key) {
None => return false,
Some(v) => if value != v { return false },
2012-12-10 11:00:52 -06:00
}
}
2013-04-03 07:45:14 -05:00
true
2012-12-10 11:00:52 -06:00
}
fn ne(&self, other: &HashMap<K, V>) -> bool { !self.eq(other) }
2013-04-03 07:45:14 -05:00
}
/// An implementation of a hash set using the underlying representation of a
/// HashMap where the value is (). As with the `HashMap` type, a `HashSet`
/// requires that the elements implement the `Eq` and `Hash` traits.
pub struct HashSet<T> {
priv map: HashMap<T, ()>
2013-04-03 07:45:14 -05:00
}
impl<T:Hash + Eq> BaseIter<T> for HashSet<T> {
2013-04-03 07:45:14 -05:00
/// Visit all values in order
2013-05-02 17:33:18 -05:00
fn each(&self, f: &fn(&T) -> bool) -> bool { self.map.each_key(f) }
2013-04-03 07:45:14 -05:00
fn size_hint(&self) -> Option<uint> { Some(self.len()) }
}
impl<T:Hash + Eq> Eq for HashSet<T> {
fn eq(&self, other: &HashSet<T>) -> bool { self.map == other.map }
fn ne(&self, other: &HashSet<T>) -> bool { self.map != other.map }
2013-04-03 07:45:14 -05:00
}
2013-01-23 15:47:27 -06:00
impl<T:Hash + Eq> Container for HashSet<T> {
2013-04-03 07:45:14 -05:00
/// Return the number of elements in the set
fn len(&const self) -> uint { self.map.len() }
2013-01-21 20:59:19 -06:00
2013-04-03 07:45:14 -05:00
/// Return true if the set contains no elements
fn is_empty(&const self) -> bool { self.map.is_empty() }
}
impl<T:Hash + Eq> Mutable for HashSet<T> {
2013-04-03 07:45:14 -05:00
/// Clear the set, removing all values.
fn clear(&mut self) { self.map.clear() }
}
impl<T:Hash + Eq> Set<T> for HashSet<T> {
2013-04-03 07:45:14 -05:00
/// Return true if the set contains a value
fn contains(&self, value: &T) -> bool { self.map.contains_key(value) }
2013-04-03 07:45:14 -05:00
/// Add a value to the set. Return true if the value was not already
/// present in the set.
fn insert(&mut self, value: T) -> bool { self.map.insert(value, ()) }
2013-04-03 07:45:14 -05:00
/// Remove a value from the set. Return true if the value was
/// present in the set.
fn remove(&mut self, value: &T) -> bool { self.map.remove(value) }
2013-01-29 16:04:25 -06:00
2013-04-03 07:45:14 -05:00
/// Return true if the set has no elements in common with `other`.
/// This is equivalent to checking for an empty intersection.
fn is_disjoint(&self, other: &HashSet<T>) -> bool {
2013-04-24 19:35:49 -05:00
old_iter::all(self, |v| !other.contains(v))
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
/// Return true if the set is a subset of another
fn is_subset(&self, other: &HashSet<T>) -> bool {
2013-04-24 19:35:49 -05:00
old_iter::all(self, |v| other.contains(v))
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
/// Return true if the set is a superset of another
fn is_superset(&self, other: &HashSet<T>) -> bool {
2013-04-03 07:45:14 -05:00
other.is_subset(self)
}
/// Visit the values representing the difference
2013-05-02 17:33:18 -05:00
fn difference(&self, other: &HashSet<T>, f: &fn(&T) -> bool) -> bool {
self.each(|v| other.contains(v) || f(v))
}
2013-04-03 07:45:14 -05:00
/// Visit the values representing the symmetric difference
2013-05-02 17:33:18 -05:00
fn symmetric_difference(&self,
other: &HashSet<T>,
f: &fn(&T) -> bool) -> bool {
self.difference(other, f) && other.difference(self, f)
}
2013-04-03 07:45:14 -05:00
/// Visit the values representing the intersection
2013-05-02 17:33:18 -05:00
fn intersection(&self, other: &HashSet<T>, f: &fn(&T) -> bool) -> bool {
self.each(|v| !other.contains(v) || f(v))
}
2013-04-03 07:45:14 -05:00
/// Visit the values representing the union
2013-05-02 17:33:18 -05:00
fn union(&self, other: &HashSet<T>, f: &fn(&T) -> bool) -> bool {
self.each(f) && other.each(|v| self.contains(v) || f(v))
}
2013-04-03 07:45:14 -05:00
}
impl<T:Hash + Eq> HashSet<T> {
/// Create an empty HashSet
pub fn new() -> HashSet<T> {
HashSet::with_capacity(INITIAL_CAPACITY)
2013-04-03 07:45:14 -05:00
}
2013-03-28 22:30:50 -05:00
/// Create an empty HashSet with space for at least `n` elements in
2013-04-03 07:45:14 -05:00
/// the hash table.
pub fn with_capacity(capacity: uint) -> HashSet<T> {
HashSet { map: HashMap::with_capacity(capacity) }
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
/// Reserve space for at least `n` elements in the hash table.
pub fn reserve_at_least(&mut self, n: uint) {
2013-04-03 07:45:14 -05:00
self.map.reserve_at_least(n)
}
2013-03-21 14:41:37 -05:00
2013-04-03 07:45:14 -05:00
/// Consumes all of the elements in the set, emptying it out
pub fn consume(&mut self, f: &fn(T)) {
2013-04-03 07:45:14 -05:00
self.map.consume(|k, _| f(k))
}
/// Returns true if the hash set contains a value equivalent to the
/// given query value.
pub fn contains_equiv<Q:Hash + Equiv<T>>(&self, value: &Q) -> bool {
self.map.contains_key_equiv(value)
}
2013-04-03 07:45:14 -05:00
}
#[cfg(test)]
2013-04-03 07:45:14 -05:00
mod test_map {
use container::{Container, Map, Set};
use option::{None, Some};
use super::*;
use uint;
2013-05-30 12:03:11 -05:00
#[test]
fn test_create_capacity_zero() {
let mut m = HashMap::with_capacity(0);
assert!(m.insert(1, 1));
}
#[test]
fn test_insert() {
let mut m = HashMap::new();
2013-04-03 07:45:14 -05:00
assert!(m.insert(1, 2));
assert!(m.insert(2, 4));
assert_eq!(*m.get(&1), 2);
assert_eq!(*m.get(&2), 4);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_find_mut() {
let mut m = HashMap::new();
2013-04-03 07:45:14 -05:00
assert!(m.insert(1, 12));
assert!(m.insert(2, 8));
assert!(m.insert(5, 14));
let new = 100;
match m.find_mut(&5) {
None => fail!(), Some(x) => *x = new
}
assert_eq!(m.find(&5), Some(&new));
}
2013-03-24 18:07:36 -05:00
2013-04-03 07:45:14 -05:00
#[test]
fn test_insert_overwrite() {
let mut m = HashMap::new();
2013-04-03 07:45:14 -05:00
assert!(m.insert(1, 2));
assert_eq!(*m.get(&1), 2);
2013-04-03 07:45:14 -05:00
assert!(!m.insert(1, 3));
assert_eq!(*m.get(&1), 3);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_insert_conflicts() {
2013-04-03 07:45:14 -05:00
let mut m = linear_map_with_capacity(4);
assert!(m.insert(1, 2));
assert!(m.insert(5, 3));
assert!(m.insert(9, 4));
assert_eq!(*m.get(&9), 4);
assert_eq!(*m.get(&5), 3);
assert_eq!(*m.get(&1), 2);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_conflict_remove() {
2013-04-03 07:45:14 -05:00
let mut m = linear_map_with_capacity(4);
assert!(m.insert(1, 2));
assert!(m.insert(5, 3));
assert!(m.insert(9, 4));
assert!(m.remove(&1));
assert_eq!(*m.get(&9), 4);
assert_eq!(*m.get(&5), 3);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_is_empty() {
2013-04-03 07:45:14 -05:00
let mut m = linear_map_with_capacity(4);
assert!(m.insert(1, 2));
assert!(!m.is_empty());
assert!(m.remove(&1));
assert!(m.is_empty());
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_pop() {
let mut m = HashMap::new();
2013-04-03 07:45:14 -05:00
m.insert(1, 2);
assert_eq!(m.pop(&1), Some(2));
assert_eq!(m.pop(&1), None);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_swap() {
let mut m = HashMap::new();
assert_eq!(m.swap(1, 2), None);
assert_eq!(m.swap(1, 3), Some(2));
assert_eq!(m.swap(1, 4), Some(3));
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_find_or_insert() {
let mut m = HashMap::new::<int, int>();
assert_eq!(*m.find_or_insert(1, 2), 2);
assert_eq!(*m.find_or_insert(1, 3), 2);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_find_or_insert_with() {
let mut m = HashMap::new::<int, int>();
assert_eq!(*m.find_or_insert_with(1, |_| 2), 2);
assert_eq!(*m.find_or_insert_with(1, |_| 3), 2);
2013-04-03 07:45:14 -05:00
}
#[test]
fn test_insert_or_update_with() {
let mut m = HashMap::new::<int, int>();
assert_eq!(*m.insert_or_update_with(1, 2, |_,x| *x+=1), 2);
assert_eq!(*m.insert_or_update_with(1, 2, |_,x| *x+=1), 3);
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_consume() {
let mut m = HashMap::new();
2013-04-03 07:45:14 -05:00
assert!(m.insert(1, 2));
assert!(m.insert(2, 3));
let mut m2 = HashMap::new();
2013-04-03 07:45:14 -05:00
do m.consume |k, v| {
m2.insert(k, v);
}
assert_eq!(m.len(), 0);
assert_eq!(m2.len(), 2);
assert_eq!(m2.get(&1), &2);
assert_eq!(m2.get(&2), &3);
2013-04-03 07:45:14 -05:00
}
2013-05-30 12:03:11 -05:00
#[test]
fn test_consume_still_usable() {
let mut m = HashMap::new();
assert!(m.insert(1, 2));
do m.consume |_, _| {}
assert!(m.insert(1, 2));
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_iterate() {
2013-04-03 07:45:14 -05:00
let mut m = linear_map_with_capacity(4);
for uint::range(0, 32) |i| {
assert!(m.insert(i, i*2));
}
2013-04-03 07:45:14 -05:00
let mut observed = 0;
for m.each |k, v| {
assert_eq!(*v, *k * 2);
2013-04-03 07:45:14 -05:00
observed |= (1 << *k);
}
assert_eq!(observed, 0xFFFF_FFFF);
2013-04-03 07:45:14 -05:00
}
2012-12-10 11:00:52 -06:00
2013-04-03 07:45:14 -05:00
#[test]
fn test_find() {
let mut m = HashMap::new();
2013-04-03 07:45:14 -05:00
assert!(m.find(&1).is_none());
m.insert(1, 2);
match m.find(&1) {
None => fail!(),
Some(v) => assert!(*v == 2)
}
}
2012-12-10 11:00:52 -06:00
2013-04-03 07:45:14 -05:00
#[test]
fn test_eq() {
let mut m1 = HashMap::new();
2013-04-03 07:45:14 -05:00
m1.insert(1, 2);
m1.insert(2, 3);
m1.insert(3, 4);
2012-12-10 11:00:52 -06:00
let mut m2 = HashMap::new();
2013-04-03 07:45:14 -05:00
m2.insert(1, 2);
m2.insert(2, 3);
2012-12-10 11:00:52 -06:00
2013-04-03 07:45:14 -05:00
assert!(m1 != m2);
2012-12-10 11:00:52 -06:00
2013-04-03 07:45:14 -05:00
m2.insert(3, 4);
assert_eq!(m1, m2);
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_expand() {
let mut m = HashMap::new();
assert_eq!(m.len(), 0);
2013-04-03 07:45:14 -05:00
assert!(m.is_empty());
2013-04-03 07:45:14 -05:00
let mut i = 0u;
let old_resize_at = m.resize_at;
while old_resize_at == m.resize_at {
m.insert(i, i);
i += 1;
}
2013-04-03 07:45:14 -05:00
assert_eq!(m.len(), i);
2013-04-03 07:45:14 -05:00
assert!(!m.is_empty());
}
2013-04-03 07:45:14 -05:00
}
#[cfg(test)]
2013-04-03 07:45:14 -05:00
mod test_set {
use super::*;
use container::{Container, Map, Set};
use vec;
2013-04-03 07:45:14 -05:00
#[test]
fn test_disjoint() {
let mut xs = HashSet::new();
let mut ys = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(xs.is_disjoint(&ys));
assert!(ys.is_disjoint(&xs));
assert!(xs.insert(5));
assert!(ys.insert(11));
assert!(xs.is_disjoint(&ys));
assert!(ys.is_disjoint(&xs));
assert!(xs.insert(7));
assert!(xs.insert(19));
assert!(xs.insert(4));
assert!(ys.insert(2));
assert!(ys.insert(-11));
assert!(xs.is_disjoint(&ys));
assert!(ys.is_disjoint(&xs));
assert!(ys.insert(7));
assert!(!xs.is_disjoint(&ys));
assert!(!ys.is_disjoint(&xs));
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_subset_and_superset() {
let mut a = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(a.insert(0));
assert!(a.insert(5));
assert!(a.insert(11));
assert!(a.insert(7));
let mut b = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(b.insert(0));
assert!(b.insert(7));
assert!(b.insert(19));
assert!(b.insert(250));
assert!(b.insert(11));
assert!(b.insert(200));
assert!(!a.is_subset(&b));
assert!(!a.is_superset(&b));
assert!(!b.is_subset(&a));
assert!(!b.is_superset(&a));
assert!(b.insert(5));
assert!(a.is_subset(&b));
assert!(!a.is_superset(&b));
assert!(!b.is_subset(&a));
assert!(b.is_superset(&a));
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_intersection() {
let mut a = HashSet::new();
let mut b = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(a.insert(11));
assert!(a.insert(1));
assert!(a.insert(3));
assert!(a.insert(77));
assert!(a.insert(103));
assert!(a.insert(5));
assert!(a.insert(-5));
assert!(b.insert(2));
assert!(b.insert(11));
assert!(b.insert(77));
assert!(b.insert(-9));
assert!(b.insert(-42));
assert!(b.insert(5));
assert!(b.insert(3));
let mut i = 0;
let expected = [3, 5, 11, 77];
for a.intersection(&b) |x| {
assert!(vec::contains(expected, x));
i += 1
}
assert_eq!(i, expected.len());
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_difference() {
let mut a = HashSet::new();
let mut b = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(a.insert(1));
assert!(a.insert(3));
assert!(a.insert(5));
assert!(a.insert(9));
assert!(a.insert(11));
assert!(b.insert(3));
assert!(b.insert(9));
let mut i = 0;
let expected = [1, 5, 11];
for a.difference(&b) |x| {
assert!(vec::contains(expected, x));
i += 1
}
assert_eq!(i, expected.len());
2013-04-03 07:45:14 -05:00
}
2013-04-03 07:45:14 -05:00
#[test]
fn test_symmetric_difference() {
let mut a = HashSet::new();
let mut b = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(a.insert(1));
assert!(a.insert(3));
assert!(a.insert(5));
assert!(a.insert(9));
assert!(a.insert(11));
assert!(b.insert(-2));
assert!(b.insert(3));
assert!(b.insert(9));
assert!(b.insert(14));
assert!(b.insert(22));
let mut i = 0;
let expected = [-2, 1, 5, 11, 14, 22];
for a.symmetric_difference(&b) |x| {
assert!(vec::contains(expected, x));
i += 1
}
assert_eq!(i, expected.len());
2013-04-03 07:45:14 -05:00
}
#[test]
fn test_union() {
let mut a = HashSet::new();
let mut b = HashSet::new();
2013-04-03 07:45:14 -05:00
assert!(a.insert(1));
assert!(a.insert(3));
assert!(a.insert(5));
assert!(a.insert(9));
assert!(a.insert(11));
assert!(a.insert(16));
assert!(a.insert(19));
assert!(a.insert(24));
assert!(b.insert(-2));
assert!(b.insert(1));
assert!(b.insert(5));
assert!(b.insert(9));
assert!(b.insert(13));
assert!(b.insert(19));
let mut i = 0;
let expected = [-2, 1, 3, 5, 9, 11, 13, 16, 19, 24];
for a.union(&b) |x| {
assert!(vec::contains(expected, x));
i += 1
}
assert_eq!(i, expected.len());
}
}