thread-safe-lru-cache

LRU cache and thread safe LRU cache implementation in C++.

Description

A least recently used (LRU) cache is a fixed size cache that behaves just like a regular lookup table. LRU cache remembers the order in which elements are accessed and once its capacity is reached, it uses this information to replace the least recently used element with a newly inserted one.

This repository provides two implementations of an LRU cache: one has only the basic functionality described above, and another containing a thread-safe implementation.

Basic Usage

Two main classes are provided ::wstux::lru::lru_cache and ::wstux::lru::thread_safe_lru_cache. A basic usage example of these may look like so:

::wstux::lru::lru_cache

#include <iostream>
#include "lru/lru_cache.h"

using lru_cache = ::wstux::lru::lru_cache<int, int>;

int fibonacci(int n, lru_cache& cache)
{
    if (n < 2) return 1;

    int value;
    if (cache.find(n, value)) {
        return val;
    }

    value = fibonacci(n - 1, cache) + fibonacci(n - 2, cache);
    cache.emplace(n, value);

    return value;
}

int main()
{
    lru_cache cache(10);
    std::cout << "fibonacci(15)" << fibonacci(15, cache) << std::endl;
    return 0;
}

::wstux::lru::thread_safe_lru_cache

#include <iostream>
#include "lru/thread_safe_lru_cache.h"

using lru_cache = ::wstux::lru::thread_safe_lru_cache<int, int>;

int fibonacci(int n, lru_cache& cache)
{
    if (n < 2) return 1;

    int value;
    if (cache.find(n, value)) {
        return val;
    }

    value = fibonacci(n - 1, cache) + fibonacci(n - 2, cache);
    cache.emplace(n, value);

    return value;
}

int main()
{
    lru_cache cache(10);
    std::cout << "fibonacci(15)" << fibonacci(15, cache) << std::endl;
    return 0;
}

Benchmark

lru_cache

For performance measurements, 100000 search/insert iterations were performed. The table shows the execution time 100000 of the lru_cache lookup/insert.

	std containers, ms	boost::intrusive containers, ms
insert	6.96474	4.02347
find	2.17074	2.09442
insert_overflow	5.01054	2.63364