源码及框架分析
SGI-STL30版本源代码中没有unordered_map和unordered_set,SGI-STL30版本是C++11之前的STL版本,这两个容器是C++11之后才更新的。但是SGI-STL30实现了哈希表,只容器的名字是hash_map和hash_set,他是作为⾮标准的容器出现的,⾮标准是指⾮C++标准规定必须实现的,源代码在hash_map/hash_set/stl_hash_map/stl_hash_set/stl_hashtable.h中hash_map和hash_set的实现结构框架核⼼部分截取出来如下:
c++
// stl_hash_set
template <class Value, class HashFcn = hash<Value>,
class EqualKey = equal_to<Value>,
class Alloc = alloc>
class hash_set
{
private:
typedef hashtable<Value, Value, HashFcn, identity<Value>, EqualKey, Alloc> ht;
ht rep;
public:
typedef typename ht::key_type key_type;
typedef typename ht::value_type value_type;
typedef typename ht::hasher hasher;
typedef typename ht::key_equal key_equal;
typedef typename ht::const_iterator iterator;
typedef typename ht::const_iterator const_iterator;
hasher hash_funct() const { return rep.hash_funct(); }
key_equal key_eq() const { return rep.key_eq(); }
};
// stl_hash_map
template <class Key, class T, class HashFcn = hash<Key>,
class EqualKey = equal_to<Key>,
class Alloc = alloc>
class hash_map
{
private:
typedef hashtable<pair<const Key, T>, Key, HashFcn,
select1st<pair<const Key, T> >, EqualKey, Alloc> ht;
ht rep;
public:
typedef typename ht::key_type key_type;
typedef T data_type;
typedef T mapped_type;
typedef typename ht::value_type value_type;
typedef typename ht::hasher hasher;
typedef typename ht::key_equal key_equal;
typedef typename ht::iterator iterator;
typedef typename ht::const_iterator const_iterator;
};
// stl_hashtable.h
template <class Value, class Key, class HashFcn,
class ExtractKey, class EqualKey,
class Alloc>
class hashtable {
public:
typedef Key key_type;
typedef Value value_type;
typedef HashFcn hasher;
typedef EqualKey key_equal;
private:
hasher hash;
key_equal equals;
ExtractKey get_key;
typedef __hashtable_node<Value> node;
vector<node*,Alloc> buckets;
size_type num_elements;
public:
typedef __hashtable_iterator<Value, Key, HashFcn, ExtractKey, EqualKey, Alloc> iterator;
pair<iterator, bool> insert_unique(const value_type& obj);
const_iterator find(const key_type& key) const;
};
template <class Value>
struct __hashtable_node
{
__hashtable_node* next;
Value val;
};
- 通过源码可以看到,结构上hash_map和hash_set跟map和set的完全类似,复⽤同⼀个hashtable实现key和key/value结构,hash_set传给hash_table的是两个key,hash_map传给hash_table的是pair<const key, value>
- 需要注意的是源码⾥⾯跟map/set源码类似,命名⻛格⽐较乱,这⾥⽐map和set还乱,hash_set模板参数居然⽤的Value命名,hash_map⽤的是Key和T命名
模拟实现unordered_map和unordered_set
实现出复⽤哈希表的框架,并⽀持insert
- 参考源码框架,unordered_map和unordered_set复⽤之前我们实现的哈希表。
- 这⾥key参数就⽤K,value参数就⽤V,哈希表中的数据类型,我们使⽤T。
- 其次跟map和set相⽐⽽⾔unordered_map和unordered_set的模拟实现类结构更复杂⼀点,但是⼤框架和思路是完全类似的。因为HashTable实现了泛型不知道T参数导致是K,还是pair<K, V>,那么insert内部进⾏插⼊时要⽤K对象转换成整形取模和K⽐较相等,因为pair的value不参与计算取模,且默认⽀持的是key和value⼀起⽐较相等,需要任何时候只需要⽐较K对象,所以在unordered_map和unordered_set层分别实现⼀个MapKeyOfT和SetKeyOfT的仿函数传给HashTable的KeyOfT,然后HashTable中通过KeyOfT仿函数取出T类型对象中的K对象,再转换成整形取模和K⽐较相等,具体细节参考如下代码实现。
c++
// MyUnorderedSet.h
namespace bit
{
template<class K, class Hash = HashFunc<K>>
class unordered_set
{
struct SetKeyOfT
{
const K& operator()(const K& key)
{
return key;
}
};
public:
bool insert(const K& key)
{
return _ht.Insert(key);
}
private:
hash_bucket::HashTable<K, K, SetKeyOfT, Hash> _ht;
};
}
// MyUnorderedMap.h
namespace bit
{
template<class K, class V, class Hash = HashFunc<K>>
class unordered_map
{
struct MapKeyOfT
{
const K& operator()(const pair<K, V>& kv)
{
return kv.first;
}
};
public:
bool insert(const pair<K, V>& kv)
{
return _ht.Insert(kv);
}
private:
hash_bucket::HashTable<K, pair<K, V>, MapKeyOfT, Hash> _ht;
};
}
// HashTable.h
template<class K>
struct HashFunc
{
size_t operator()(const K& key)
{
return (size_t)key;
}
};
namespace hash_bucket
{
template<class T>
struct HashNode
{
T _data;
HashNode<T>* _next;
HashNode(const T& data)
:_data(data)
,_next(nullptr)
{}
};
// 实现步骤:
// 1、实现哈希表
// 2、封装unordered_map和unordered_set的框架 解决KeyOfT
// 3、iterator
// 4、const_iterator
// 5、key不⽀持修改的问题
// 6、operator[]
template<class K, class T, class KeyOfT, class Hash>
class HashTable
{
typedef HashNode<T> Node;
inline unsigned long __stl_next_prime(unsigned long n)
{
static const int __stl_num_primes = 28;
static const unsigned long __stl_prime_list[__stl_num_primes]=
{
53, 97, 193, 389, 769,
1543, 3079, 6151, 12289, 24593,
49157, 98317, 196613, 393241, 786433,
1572869, 3145739, 6291469, 12582917, 25165843,
50331653, 100663319, 201326611, 402653189, 805306457,
1610612741, 3221225473, 4294967291
};
const unsigned long* first = __stl_prime_list;
const unsigned long* last = __stl_prime_list + __stl_num_primes;
const unsigned long* pos = lower_bound(first, last, n);
return pos == last ? *(last - 1) : *pos;
}
public:
HashTable()
{
_tables.resize(__stl_next_prime(_tables.size()), nullptr);
}
~HashTable()
{
// 依次把每个桶释放
for (size_t i = 0; i < _tables.size(); i++)
{
Node* cur = _tables[i];
while (cur)
{
Node* next = cur->_next;
delete cur;
cur = next;
}
_tables[i] = nullptr;
}
}
bool Insert(const T& data)
{
KeyOfT kot;
if (Find(kot(data)))
return false;
Hash hs;
size_t hashi = hs(kot(data)) % _tables.size();
// 负载因⼦==1扩容
if (_n == _tables.size())
{
vector<Node*> newtables(__stl_next_prime(_tables.size()),
nullptr);
for (size_t i = 0; i < _tables.size(); i++)
{
Node* cur = _tables[i];
while (cur)
{
Node* next = cur->_next;
// 旧表中结点,挪动新表重新映射的位置
size_t hashi = hs(kot(cur->_data)) % newtables.size();
// 头插到新表
cur->_next = newtables[hashi];
newtables[hashi] = cur;
cur = next;
}
_tables[i] = nullptr;
}
_tables.swap(newtables);
}
// 头插
Node* newnode = new Node(data);
newnode->_next = _tables[hashi];
_tables[hashi] = newnode;
++_n;
return true;
}
private:
vector<Node*> _tables; // 指针数组
size_t _n = 0; // 表中存储数据个数
};
}
⽀持iterator的实现
iterator核⼼源代码
c++
template <class Value, class Key, class HashFcn,
class ExtractKey, class EqualKey, class Alloc>
struct __hashtable_iterator {
typedef hashtable<Value, Key, HashFcn, ExtractKey, EqualKey, Alloc>
hashtable;
typedef __hashtable_iterator<Value, Key, HashFcn,
ExtractKey, EqualKey, Alloc>
iterator;
typedef __hashtable_const_iterator<Value, Key, HashFcn,
ExtractKey, EqualKey, Alloc>
const_iterator;
typedef __hashtable_node<Value> node;
typedef forward_iterator_tag iterator_category;
typedef Value value_type;
node* cur;
hashtable* ht;
__hashtable_iterator(node* n, hashtable* tab) : cur(n), ht(tab) {}
__hashtable_iterator() {}
reference operator*() const { return cur->val; }
#ifndef __SGI_STL_NO_ARROW_OPERATOR
pointer operator->() const { return &(operator*()); }
#endif /* __SGI_STL_NO_ARROW_OPERATOR */
iterator& operator++();
iterator operator++(int);
bool operator==(const iterator& it) const { return cur == it.cur; }
bool operator!=(const iterator& it) const { return cur != it.cur; }
};
template <class V, class K, class HF, class ExK, class EqK, class A>
__hashtable_iterator<V, K, HF, ExK, EqK, A>&
__hashtable_iterator<V, K, HF, ExK, EqK, A>::operator++()
{
const node* old = cur;
cur = cur->next;
if (!cur) {
size_type bucket = ht->bkt_num(old->val);
while (!cur && ++bucket < ht->buckets.size())
cur = ht->buckets[bucket];
}
return *this;
}
iterator实现思路分析
- iterator实现的⼤框架跟list的iterator思路是⼀致的,⽤⼀个类型封装结点的指针,再通过重载运算符实现,迭代器像指针⼀样访问的⾏为,要注意的是哈希表的迭代器是单向迭代器。
- 这⾥的难点是operator++的实现。iterator中有⼀个指向结点的指针,如果当前桶下⾯还有结点,则结点的指针指向下⼀个结点即可。如果当前桶⾛完了,则需要想办法计算找到下⼀个桶。这⾥的难点是反⽽是结构设计的问题,参考上⾯的源码,我们可以看到iterator中除了有结点的指针,还有哈希表对象的指针,这样当前桶⾛完了,要计算下⼀个桶就相对容易多了,⽤key值计算出当前桶位置,依次往后找下⼀个不为空的桶即可。
- begin()返回第⼀个桶中第⼀个节点指针构造的迭代器,这⾥end()返回迭代器可以⽤空表⽰。
- unordered_set的iterator也不⽀持修改,我们把unordered_set的第⼆个模板参数改成const K即可,
HashTable<K, const K, SetKeyOfT, Hash> _ht; - unordered_map的iterator不⽀持修改key但是可以修改value,我们把unordered_map的第⼆个模板参数pair的第⼀个参数改成const K即可,
HashTable<K, pair<const K, V>, MapKeyOfT, Hash> _ht; - ⽀持完整的迭代器还有很多细节需要修改,具体参考下⾯题的代码。

map⽀持[]
- unordered_map要⽀持
[]主要需要修改insert返回值⽀持,修改HashTable中的insert返回值为pair<Iterator, bool> Insert(const T& data) - 有了insert⽀持
[]实现就很简单了,具体参考下⾯代码实现
bit::unordered_map和bit::unordered_set代码实现
c++
// MyUnorderedSet.h
namespace bit
{
template<class K, class Hash = HashFunc<K>>
class unordered_set
{
struct SetKeyOfT
{
const K& operator()(const K& key)
{
return key;
}
};
public:
typedef typename hash_bucket::HashTable<K, const K, SetKeyOfT,
Hash>::Iterator iterator;
typedef typename hash_bucket::HashTable<K, const K, SetKeyOfT,
Hash>::ConstIterator const_iterator;
iterator begin()
{
return _ht.Begin();
}
iterator end()
{
return _ht.End();
}
const_iterator begin() const
{
return _ht.Begin();
}
const_iterator end() const
{
return _ht.End();
}
pair<iterator, bool> insert(const K& key)
{
return _ht.Insert(key);
}
iterator Find(const K& key)
{
return _ht.Find(key);
}
bool Erase(const K& key)
{
return _ht.Erase(key);
}
private:
hash_bucket::HashTable<K, const K, SetKeyOfT, Hash> _ht;
};
void test_set()
{
unordered_set<int> s;
int a[] = { 4, 2, 6, 1, 3, 5, 15, 7, 16, 14, 3,3,15 };
for (auto e : a)
{
s.insert(e);
}
for (auto e : s)
{
cout << e << " ";
}
cout << endl;
unordered_set<int>::iterator it = s.begin();
while (it != s.end())
{
// 不⽀持修改
//*it += 1;
cout << *it << " ";
++it;
}
cout << endl;
}
}
// MyUnorderedMap.h
namespace bit
{
template<class K, class V, class Hash = HashFunc<K>>
class unordered_map
{
struct MapKeyOfT
{
const K& operator()(const pair<K, V>& kv)
{
return kv.first;
}
};
public:
typedef typename hash_bucket::HashTable<K, pair<const K, V>,
MapKeyOfT, Hash>::Iterator iterator;
typedef typename hash_bucket::HashTable<K, pair<const K, V>,
MapKeyOfT, Hash>::ConstIterator const_iterator;
iterator begin()
{
return _ht.Begin();
}
iterator end()
{
return _ht.End();
}
const_iterator begin() const
{
return _ht.Begin();
}
const_iterator end() const
{
return _ht.End();
}
pair<iterator, bool> insert(const pair<K, V>& kv)
{
return _ht.Insert(kv);
}
V& operator[](const K& key)
{
pair<iterator, bool> ret = _ht.Insert(make_pair(key, V()));
return ret.first->second;
}
iterator Find(const K& key)
{
return _ht.Find(key);
}
bool Erase(const K& key)
{
return _ht.Erase(key);
}
private:
hash_bucket::HashTable<K, pair<const K, V>, MapKeyOfT, Hash> _ht;
};
void test_map()
{
unordered_map<string, string> dict;
dict.insert({ "sort", "排序" });
dict.insert({ "left", "左边" });
dict.insert({ "right", "右边" });
dict["left"] = "左边,剩余";
dict["insert"] = "插⼊";
dict["string"];
unordered_map<string, string>::iterator it = dict.begin();
while (it != dict.end())
{
// 不能修改first,可以修改second
//it->first += 'x';
it->second += 'x';
cout << it->first << ":" << it->second << endl;
++it;
}
cout << endl;
}
}
// HashTable.h
template<class K>
struct HashFunc
{
size_t operator()(const K& key)
{
return (size_t)key;
}
};
// 特化
template<>
struct HashFunc<string>
{
size_t operator()(const string& key)
{
size_t hash = 0;
for (auto e : key)
{
hash *= 131;
hash += e;
}
return hash;
}
};
namespace hash_bucket
{
template<class T>
struct HashNode
{
T _data;
HashNode<T>* _next;
HashNode(const T& data)
:_data(data)
,_next(nullptr)
{}
};
// 前置声明
template<class K, class T, class KeyOfT, class Hash>
class HashTable;
template<class K, class T, class Ptr, class Ref, class KeyOfT, class Hash>
struct HTIterator
{
typedef HashNode<T> Node;
typedef HTIterator<K, T, Ptr, Ref, KeyOfT, Hash> Self;
Node* _node;
const HashTable<K, T, KeyOfT, Hash>* _pht;
HTIterator(Node* node, const HashTable<K, T, KeyOfT, Hash>* pht)
:_node(node)
,_pht(pht)
{}
Ref operator*()
{
return _node->_data;
}
Ptr operator->()
{
return &_node->_data;
}
bool operator!=(const Self& s)
{
return _node != s._node;
}
Self& operator++()
{
if (_node->_next)
{
// 当前桶还有节点
_node = _node->_next;
}
else
{
// 当前桶⾛完了,找下⼀个不为空的桶
KeyOfT kot;
Hash hs;
size_t hashi = hs(kot(_node->_data)) % _pht-> _tables.size();
++hashi;
while (hashi < _pht->_tables.size())
{
if (_pht->_tables[hashi])
{
break;
}
++hashi;
}
//后面没有桶了
if (hashi == _pht->_tables.size())
{
_node = nullptr; // end()
}
else
{
_node = _pht->_tables[hashi];
}
}
return *this;
}
};
template<class K, class T, class KeyOfT, class Hash>
class HashTable
{
// 友元声明
template<class K, class T, class Ptr, class Ref, class KeyOfT, class Hash>
friend struct HTIterator;
typedef HashNode<T> Node;
public:
typedef HTIterator<K, T, T*, T&, KeyOfT, Hash> Iterator;
typedef HTIterator<K, T, const T*, const T&, KeyOfT, Hash> ConstIterator;
Iterator Begin()
{
if (_n == 0)
return End();
for (size_t i = 0; i < _tables.size(); i++)
{
Node* cur = _tables[i];
if (cur)
{
return Iterator(cur, this);
}
}
return End();
}
Iterator End()
{
return Iterator(nullptr, this);
}
ConstIterator Begin() const
{
if (_n == 0)
return End();
for (size_t i = 0; i < _tables.size(); i++)
{
Node* cur = _tables[i];
if (cur)
{
return ConstIterator(cur, this);
}
}
return End();
}
ConstIterator End() const
{
return ConstIterator(nullptr, this);
}
inline unsigned long __stl_next_prime(unsigned long n)
{
static const int __stl_num_primes = 28;
static const unsigned long __stl_prime_list[__stl_num_primes]=
{
53,97,193,389,769,
1543,3079,6151,12289,24593,
49157, 98317, 196613, 393241, 786433,
1572869, 3145739, 6291469, 12582917, 25165843,
50331653, 100663319, 201326611, 402653189, 805306457,
1610612741, 3221225473, 4294967291
};
const unsigned long* first = __stl_prime_list;
const unsigned long* last = __stl_prime_list + __stl_num_primes;
const unsigned long* pos = lower_bound(first, last, n);
return pos == last ? *(last - 1) : *pos;
}
HashTable()
{
_tables.resize(__stl_next_prime(_tables.size()), nullptr);
}
~HashTable()
{
// 依次把每个桶释放
for (size_t i = 0; i < _tables.size(); i++)
{
Node* cur = _tables[i];
while (cur)
{
Node* next = cur->_next;
delete cur;
cur = next;
}
_tables[i] = nullptr;
}
}
pair<Iterator, bool> Insert(const T& data)
{
KeyOfT kot;
Iterator it = Find(kot(data));
if (it != End())
return make_pair(it, false);
Hash hs;
size_t hashi = hs(kot(data)) % _tables.size();
// 负载因⼦==1扩容
if (_n == _tables.size())
{
vector<Node*>
newtables(__stl_next_prime(_tables.size()+1), nullptr);
for (size_t i = 0; i < _tables.size(); i++)
{
Node* cur = _tables[i];
while (cur)
{
Node* next = cur->_next;
// 旧表中节点,挪动新表重新映射的位置
size_t hashi = hs(kot(cur->_data)) %
newtables.size();
// 头插到新表
cur->_next = newtables[hashi];
newtables[hashi] = cur;
cur = next;
}
_tables[i] = nullptr;
}
_tables.swap(newtables);
}
// 头插
Node* newnode = new Node(data);
newnode->_next = _tables[hashi];
_tables[hashi] = newnode;
++_n;
return make_pair(Iterator(newnode, this), true);
}
Iterator Find(const K& key)
{
KeyOfT kot;
Hash hs;
size_t hashi = hs(key) % _tables.size();
Node* cur = _tables[hashi];
while (cur)
{
if (kot(cur->_data) == key)
{
return Iterator(cur, this);
}
cur = cur->_next;
}
return End();
}
bool Erase(const K& key)
{
KeyOfT kot;
Hash hs;
size_t hashi = hs(key) % _tables .size();
Node* prev = nullptr;
Node* cur = _tables[hashi];
while (cur)
{
if (kot(cur->_data) == key)
{
if (prev == nullptr)
{
_tables[hashi] = cur->_next;
}
else
{
prev->_next = cur->_next;
}
delete cur;
--_n;
return true;
}
prev = cur;
cur = cur->_next;
}
return false;
}
private:
vector<Node*> _tables; // 指针数组
size_t _n = 0; // 表中存储数据个数
};
}