2006-10-23 17:48:43 +02:00
|
|
|
//
|
|
|
|
// SimpleHashTable.h
|
|
|
|
//
|
2007-04-18 18:22:57 +02:00
|
|
|
// $Id: //poco/Main/Foundation/include/Poco/SimpleHashTable.h#11 $
|
2006-10-23 17:48:43 +02:00
|
|
|
//
|
|
|
|
// Library: Foundation
|
2006-11-17 17:21:51 +01:00
|
|
|
// Package: Hashing
|
2006-10-23 17:48:43 +02:00
|
|
|
// Module: SimpleHashTable
|
|
|
|
//
|
|
|
|
// Definition of the SimpleHashTable class.
|
|
|
|
//
|
|
|
|
// Copyright (c) 2006, Applied Informatics Software Engineering GmbH.
|
|
|
|
// and Contributors.
|
|
|
|
//
|
|
|
|
// Permission is hereby granted, free of charge, to any person or organization
|
|
|
|
// obtaining a copy of the software and accompanying documentation covered by
|
|
|
|
// this license (the "Software") to use, reproduce, display, distribute,
|
|
|
|
// execute, and transmit the Software, and to prepare derivative works of the
|
|
|
|
// Software, and to permit third-parties to whom the Software is furnished to
|
|
|
|
// do so, all subject to the following:
|
|
|
|
//
|
|
|
|
// The copyright notices in the Software and this entire statement, including
|
|
|
|
// the above license grant, this restriction and the following disclaimer,
|
|
|
|
// must be included in all copies of the Software, in whole or in part, and
|
|
|
|
// all derivative works of the Software, unless such copies or derivative
|
|
|
|
// works are solely in the form of machine-executable object code generated by
|
|
|
|
// a source language processor.
|
|
|
|
//
|
|
|
|
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
// IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
// FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT
|
|
|
|
// SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
|
|
|
|
// FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
|
|
|
|
// ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
|
|
|
|
// DEALINGS IN THE SOFTWARE.
|
|
|
|
//
|
|
|
|
|
|
|
|
|
|
|
|
#ifndef Foundation_SimpleHashTable_INCLUDED
|
|
|
|
#define Foundation_SimpleHashTable_INCLUDED
|
|
|
|
|
|
|
|
|
|
|
|
#include "Poco/Foundation.h"
|
|
|
|
#include "Poco/Exception.h"
|
|
|
|
#include "Poco/HashFunction.h"
|
|
|
|
#include "Poco/HashStatistic.h"
|
|
|
|
#include <vector>
|
|
|
|
#include <map>
|
2006-11-17 17:21:51 +01:00
|
|
|
#include <cstddef>
|
|
|
|
#include <algorithm>
|
2006-10-23 17:48:43 +02:00
|
|
|
|
|
|
|
|
|
|
|
namespace Poco {
|
|
|
|
|
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
//@ deprecated
|
|
|
|
template <class Key, class Value, class KeyHashFunction = HashFunction<Key> >
|
2006-10-23 17:48:43 +02:00
|
|
|
class SimpleHashTable
|
|
|
|
/// A SimpleHashTable stores a key value pair that can be looked up via a hashed key.
|
|
|
|
///
|
|
|
|
/// In comparision to a HashTable, this class handles collisions by sequentially searching the next
|
|
|
|
/// free location. This also means that the maximum size of this table is limited, i.e. if the hash table
|
|
|
|
/// is full, it will throw an exception and that this class does not support remove operations.
|
|
|
|
/// On the plus side it is faster than the HashTable.
|
|
|
|
///
|
|
|
|
/// This class is NOT thread safe.
|
|
|
|
{
|
|
|
|
public:
|
|
|
|
class HashEntry
|
|
|
|
{
|
|
|
|
public:
|
2006-11-17 17:21:51 +01:00
|
|
|
Key key;
|
2006-10-23 17:48:43 +02:00
|
|
|
Value value;
|
2006-11-17 17:21:51 +01:00
|
|
|
HashEntry(const Key k, const Value v): key(k), value(v)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
|
|
|
}
|
|
|
|
};
|
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
typedef std::vector<HashEntry*> HashTableVector;
|
2006-10-23 17:48:43 +02:00
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
SimpleHashTable(UInt32 capacity = 251): _entries(capacity, 0), _size(0), _capacity(capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
/// Creates the SimpleHashTable.
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
SimpleHashTable(const SimpleHashTable& ht):
|
|
|
|
_size(ht._size),
|
2006-11-17 17:21:51 +01:00
|
|
|
_capacity(ht._capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
_entries.reserve(ht._capacity);
|
|
|
|
for (typename HashTableVector::iterator it = ht._entries.begin(); it != ht._entries.end(); ++it)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (*it)
|
|
|
|
_entries.push_back(new HashEntry(*it));
|
2006-10-23 17:48:43 +02:00
|
|
|
else
|
2006-11-17 17:21:51 +01:00
|
|
|
_entries.push_back(0);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
~SimpleHashTable()
|
|
|
|
/// Destroys the SimpleHashTable.
|
|
|
|
{
|
|
|
|
clear();
|
|
|
|
}
|
|
|
|
|
|
|
|
SimpleHashTable& operator = (const SimpleHashTable& ht)
|
|
|
|
{
|
|
|
|
if (this != &ht)
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
SimpleHashTable tmp(ht);
|
|
|
|
swap(tmp);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
return *this;
|
|
|
|
}
|
2006-11-17 17:21:51 +01:00
|
|
|
|
|
|
|
void swap(SimpleHashTable& ht)
|
|
|
|
{
|
|
|
|
using std::swap;
|
|
|
|
swap(_entries, ht._entries);
|
|
|
|
swap(_size, ht._size);
|
|
|
|
swap(_capacity, ht._capacity);
|
|
|
|
}
|
2006-10-23 17:48:43 +02:00
|
|
|
|
|
|
|
void clear()
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
for (typename HashTableVector::iterator it = _entries.begin(); it != _entries.end(); ++it)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
delete *it;
|
|
|
|
*it = 0;
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
2006-11-17 17:21:51 +01:00
|
|
|
_size = 0;
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
UInt32 insert(const Key& key, const Value& value)
|
|
|
|
/// Returns the hash value of the inserted item.
|
|
|
|
/// Throws an exception if the entry was already inserted
|
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
insertRaw(key, hsh, value);
|
|
|
|
return hsh;
|
|
|
|
}
|
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
Value& insertRaw(const Key& key, UInt32 hsh, const Value& value)
|
2006-10-23 17:48:43 +02:00
|
|
|
/// Returns the hash value of the inserted item.
|
|
|
|
/// Throws an exception if the entry was already inserted
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
UInt32 pos = hsh;
|
|
|
|
if (!_entries[pos])
|
|
|
|
_entries[pos] = new HashEntry(key, value);
|
2006-10-23 17:48:43 +02:00
|
|
|
else
|
|
|
|
{
|
|
|
|
UInt32 origHash = hsh;
|
2006-11-17 17:21:51 +01:00
|
|
|
while (_entries[hsh % _capacity])
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
|
|
|
throw ExistsException();
|
|
|
|
if (hsh - origHash > _capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
throw PoolOverflowException("SimpleHashTable full");
|
|
|
|
hsh++;
|
|
|
|
}
|
2006-11-17 17:21:51 +01:00
|
|
|
pos = hsh % _capacity;
|
|
|
|
_entries[pos] = new HashEntry(key, value);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
_size++;
|
2006-11-17 17:21:51 +01:00
|
|
|
return _entries[pos]->value;
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
UInt32 update(const Key& key, const Value& value)
|
|
|
|
/// Returns the hash value of the inserted item.
|
|
|
|
/// Replaces an existing entry if it finds one
|
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
updateRaw(key, hsh, value);
|
|
|
|
return hsh;
|
|
|
|
}
|
|
|
|
|
|
|
|
void updateRaw(const Key& key, UInt32 hsh, const Value& value)
|
|
|
|
/// Returns the hash value of the inserted item.
|
|
|
|
/// Replaces an existing entry if it finds one
|
|
|
|
{
|
|
|
|
if (!_entries[hsh])
|
|
|
|
_entries[hsh] = new HashEntry(key, value);
|
|
|
|
else
|
|
|
|
{
|
|
|
|
UInt32 origHash = hsh;
|
2006-11-17 17:21:51 +01:00
|
|
|
while (_entries[hsh % _capacity])
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
_entries[hsh % _capacity]->value = value;
|
2006-10-23 17:48:43 +02:00
|
|
|
return;
|
|
|
|
}
|
2006-11-17 17:21:51 +01:00
|
|
|
if (hsh - origHash > _capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
throw PoolOverflowException("SimpleHashTable full");
|
|
|
|
hsh++;
|
|
|
|
}
|
2006-11-17 17:21:51 +01:00
|
|
|
_entries[hsh % _capacity] = new HashEntry(key, value);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
_size++;
|
|
|
|
}
|
|
|
|
|
|
|
|
UInt32 hash(const Key& key) const
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
return _hash(key, _capacity);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
const Value& get(const Key& key) const
|
|
|
|
/// Throws an exception if the value does not exist
|
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
return getRaw(key, hsh);
|
|
|
|
}
|
|
|
|
|
|
|
|
const Value& getRaw(const Key& key, UInt32 hsh) const
|
|
|
|
/// Throws an exception if the value does not exist
|
|
|
|
{
|
|
|
|
UInt32 origHash = hsh;
|
|
|
|
while (true)
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity])
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
return _entries[hsh % _capacity]->value;
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
throw InvalidArgumentException("value not found");
|
2006-11-17 17:21:51 +01:00
|
|
|
if (hsh - origHash > _capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
throw InvalidArgumentException("value not found");
|
|
|
|
hsh++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
Value& get(const Key& key)
|
|
|
|
/// Throws an exception if the value does not exist
|
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
return const_cast<Value&>(getRaw(key, hsh));
|
|
|
|
}
|
|
|
|
|
|
|
|
const Value& operator [] (const Key& key) const
|
|
|
|
{
|
|
|
|
return get(key);
|
|
|
|
}
|
|
|
|
|
|
|
|
Value& operator [] (const Key& key)
|
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
UInt32 origHash = hsh;
|
|
|
|
while (true)
|
|
|
|
{
|
|
|
|
if (_entries[hsh % _capacity])
|
|
|
|
{
|
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
|
|
|
{
|
|
|
|
return _entries[hsh % _capacity]->value;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else return insertRaw(key, hsh, Value());
|
|
|
|
if (hsh - origHash > _capacity)
|
|
|
|
return insertRaw(key, hsh, Value());
|
|
|
|
hsh++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2006-10-23 17:48:43 +02:00
|
|
|
const Key& getKeyRaw(const Key& key, UInt32 hsh)
|
|
|
|
/// Throws an exception if the key does not exist. returns a reference to the internally
|
|
|
|
/// stored key. Useful when someone does an insert and wants for performance reason only to store
|
|
|
|
/// a pointer to the key in another collection
|
|
|
|
{
|
|
|
|
UInt32 origHash = hsh;
|
|
|
|
while (true)
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity])
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
return _entries[hsh % _capacity]->key;
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
throw InvalidArgumentException("key not found");
|
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
if (hsh - origHash > _capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
throw InvalidArgumentException("key not found");
|
|
|
|
hsh++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
bool get(const Key& key, Value& v) const
|
|
|
|
/// Sets v to the found value, returns false if no value was found
|
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
return getRaw(key, hsh, v);
|
|
|
|
}
|
|
|
|
|
|
|
|
bool getRaw(const Key& key, UInt32 hsh, Value& v) const
|
|
|
|
/// Sets v to the found value, returns false if no value was found
|
|
|
|
{
|
|
|
|
UInt32 origHash = hsh;
|
|
|
|
while (true)
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity])
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
v = _entries[hsh % _capacity]->value;
|
2006-10-23 17:48:43 +02:00
|
|
|
return true;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
return false;
|
2006-11-17 17:21:51 +01:00
|
|
|
if (hsh - origHash > _capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
return false;
|
|
|
|
hsh++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2007-04-18 18:22:57 +02:00
|
|
|
bool exists(const Key& key) const
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
|
|
|
UInt32 hsh = hash(key);
|
|
|
|
return existsRaw(key, hsh);
|
|
|
|
}
|
|
|
|
|
2007-04-18 18:22:57 +02:00
|
|
|
bool existsRaw(const Key& key, UInt32 hsh) const
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
|
|
|
UInt32 origHash = hsh;
|
|
|
|
while (true)
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity])
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_entries[hsh % _capacity]->key == key)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
|
|
|
return true;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
else
|
|
|
|
return false;
|
2006-11-17 17:21:51 +01:00
|
|
|
if (hsh - origHash > _capacity)
|
2006-10-23 17:48:43 +02:00
|
|
|
return false;
|
|
|
|
hsh++;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
size_t size() const
|
|
|
|
/// Returns the number of elements already inserted into the SimpleHashTable
|
|
|
|
{
|
|
|
|
return _size;
|
|
|
|
}
|
|
|
|
|
2006-11-17 17:21:51 +01:00
|
|
|
UInt32 capacity() const
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
return _capacity;
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
void resize(UInt32 newSize)
|
|
|
|
/// Resizes the hashtable, rehashes all existing entries. Expensive!
|
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (_capacity != newSize)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
SimpleHashTable tmp(newSize);
|
|
|
|
swap(tmp);
|
|
|
|
for (typename HashTableVector::const_iterator it = tmp._entries.begin(); it != tmp._entries.end(); ++it)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
if (*it)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
2006-11-17 17:21:51 +01:00
|
|
|
insertRaw((*it)->key, hash((*it)->key), (*it)->value);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
HashStatistic currentState(bool details = false) const
|
|
|
|
/// Returns the current internal state
|
|
|
|
{
|
|
|
|
UInt32 numberOfEntries = (UInt32)_size;
|
|
|
|
UInt32 numZeroEntries = 0;
|
|
|
|
UInt32 maxEntriesPerHash = 0;
|
|
|
|
std::vector<UInt32> detailedEntriesPerHash;
|
|
|
|
#ifdef _DEBUG
|
|
|
|
UInt32 totalSize = 0;
|
|
|
|
#endif
|
2006-11-17 17:21:51 +01:00
|
|
|
for (int i=0; i < _capacity; ++i)
|
2006-10-23 17:48:43 +02:00
|
|
|
{
|
|
|
|
if (_entries[i])
|
|
|
|
{
|
|
|
|
maxEntriesPerHash = 1;
|
|
|
|
UInt32 size = 1;
|
|
|
|
if (details)
|
|
|
|
detailedEntriesPerHash.push_back(size);
|
|
|
|
#ifdef DEBUG
|
|
|
|
totalSize += size;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
else
|
|
|
|
{
|
|
|
|
numZeroEntries++;
|
|
|
|
if (details)
|
|
|
|
detailedEntriesPerHash.push_back(0);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
poco_assert_dbg(totalSize == numberOfEntries);
|
2006-11-17 17:21:51 +01:00
|
|
|
|
|
|
|
return HashStatistic(_capacity, numberOfEntries, numZeroEntries, maxEntriesPerHash, detailedEntriesPerHash);
|
2006-10-23 17:48:43 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
private:
|
|
|
|
HashTableVector _entries;
|
2006-11-17 17:21:51 +01:00
|
|
|
size_t _size;
|
|
|
|
UInt32 _capacity;
|
|
|
|
KeyHashFunction _hash;
|
2006-10-23 17:48:43 +02:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
} // namespace Poco
|
|
|
|
|
|
|
|
|
|
|
|
#endif // Foundation_HashTable_INCLUDED
|