README
¶
go-hashlru
A simple thread-safe, fixed size LRU written in Go. Based on dominictarr's Hashlru Algorithm. 🔃
Uses map[interface{}]interface{} to allow kv pairs of any type. The hlru package contains all the necessary functions.
cache, _ := hlru.NewHLRU(100)
cache.Set("key", "value")
val, _ := cache.Get("key")
fmt.Println(val)
// value
Visit example/example.go in the root directory for a simple example.
Table of Contents :
Installation
$ go get github.com/saurabh0719/go-hashlru
Latest - v0.1.0
API Reference
HashLRU Struct
type HashLRU struct {
maxSize int
size int
oldCache, newCache map[interface{}]interface{}
onEvictedCB func(key, value interface{})
lock sync.RWMutex
}
As explained by dominictarr :
- This algorithm does not give you an ordered list of the N most recently used items, but has the important properties of the LRU (bounded memory use and O(1) time complexity)
The HashLRU algorithm maintains two separate maps
and bulk eviction happens only after both the maps fill up. Hence onEvictedCB is triggered in bulk and is not an accurate measure of timely LRU-ness.
This implementation uses sync.RWMutex to ensure thread-safety.
Go back to the table of contents
func NewHLRU
func NewHLRU(maxSize int) (*HashLRU, error)
Returns a new instance of HashLRU of the given size: maxSize.
func NewWithEvict
func NewWithEvict(maxSize int, onEvict func(key, value interface{})) (*HashLRU, error)
Takes maxSzie and a callback function onEvict as arguments and returns a new instance of HashLRU of the given size.
func (lru *HashLRU) Set
func (lru *HashLRU) Set(key, value interface{})
Adds a new key-value pair to the cache and updates it.
func (lru *HashLRU) Get
func (lru *HashLRU) Get(key interface{}) (interface{}, bool)
Get the value of any key and updates the cache. Returns value, true if the kv pair is found, else returns nil, false.
func (lru *HashLRU) Has
func (lru *HashLRU) Has(key interface{}) bool
Returns true if the key exists, else returns false.
func (lru *HashLRU) Remove
func (lru *HashLRU) Remove(key interface{}) bool
Deletes the key-value pair and returns true if its successful, else returns false.
Go back to the table of contents
func (lru *HashLRU) Peek
func (lru *LRU) Peek(key interface{}) (interface{}, bool)
Get the value of a key without updating the cache. Returns value, true if the kv pair is found, else returns nil, false.
func (lru *HashLRU) Clear
func (lru *HashLRU) Clear()
Empties the Cache.
func (lru *HashLRU) Resize
func (lru *HashLRU) Resize(newSize int) (int, error)
Update the maxSize of the cache. Items will be evicted automatically to adjust. Returns the number of evicted key-value pairs due to the re-size.
func (lru *HashLRU) Len
func (lru *HashLRU) Len() int
Returns the total number of key-value pairs present in the cache.
func (lru *HashLRU) Keys
func (lru* HashLRU) Keys() []interface{}
Returns a slice of all the Keys in the cache.
func (lru *HashLRU) Vals
func (lru *HashLRU) Vals() []interface{}
Returns a slice of all the Values in the cache.
Go back to the table of contents
Tests
$ go test
Use -v for an detailed output.
Benchmark
$ go test -run=XXX -bench=.
HashLRU has a better hit/miss ratio for BenchmarkHLRU_Rand and BenchmarkHLRU_Freq (negligible) as compared to same benchmark tests from golang-lru. However, golang-lru does have a slightly lower ns/op.
Go back to the table of contents
Documentation
¶
Index ¶
- type HashLRU
- func (lru *HashLRU) Clear()
- func (lru *HashLRU) Get(key interface{}) (interface{}, bool)
- func (lru *HashLRU) Has(key interface{}) bool
- func (lru *HashLRU) Keys() []interface{}
- func (lru *HashLRU) Len() int
- func (lru *HashLRU) Peek(key interface{}) (interface{}, bool)
- func (lru *HashLRU) Remove(key interface{}) bool
- func (lru *HashLRU) Resize(newSize int) (int, error)
- func (lru *HashLRU) Set(key, value interface{})
- func (lru *HashLRU) Vals() []interface{}
- type KVPair
Constants ¶
This section is empty.
Variables ¶
This section is empty.
Functions ¶
This section is empty.
Types ¶
type HashLRU ¶
type HashLRU struct {
// contains filtered or unexported fields
}
Source Files
Directories