gtools

GTools

GTools is a golang library that provides some handy ways to handle many of the typical requirements when developing GO applications.

Contains

Goroutine Pool

Supports setting the minimum number of resident concurrent goroutine, the maximum number of concurrent goroutine allowed in the pool, the maximum time a non-resident concurrent goroutine can wait for a new task, and the maximum time a task can afford to wait for a concurrent goroutine to receive it (when no concurrent goroutine is free after that time, another concurrent goroutine is opened to work on it). Also, when submitting tasks to the concurrent pool, setting the priority of the task is supported.

Usage

package main

import (
	"fmt"
	"time"

	"github.com/gtoxlili/gtools/pool"
)

func main() {

	// Creates a concurrent pool with at least two resident concurrent goroutines and up to 1024 worker concurrent goroutines, 
	// while non-resident concurrent goroutines are automatically destroyed after 1 minute of inactivity, 
	// and new worker concurrent goroutines are created when the task waits 5 seconds without being processed by the current worker concurrent goroutine.
	p := pool.New(time.Minute, time.Second*5, 1024, 2)

	for i := 0; i < 100; i++ {
		tmp := i
		p.Execute(func() {
			fmt.Println(tmp)
		}, tmp)
	}
	// Output: like 99,98 ... 0 
	// Because of the priority, in the above code, the tasks that are pushed later are instead given priority

	// if you want to stop the pool, you can call the following method
	p.Done()
	// The current policy is to not execute all cached tasks when the concurrent pool is closed
	// If you want to close the pool after all the tasks of the current push have been executed
	// You can try to post the close command as a task and give him a small enough priority
	p.Execute(func() {
		p.Done()
	}, -1)

	<-make(chan struct{}, 1)
}

package main

import (
	"github.com/gtoxlili/gtools/pool"
)

func main() {

	// Creates a goroutine pool that creates new goroutines as needed, 
	// but will reuse previously constructed goroutines when they are available.
	// These pools will typically improve the performance of programs that execute many short-lived asynchronous tasks
	p := pool.DefaultCachedPool()

	// Creates a goroutine pool that reuses a fixed number of goroutines operating off a shared unbounded queue. 
	// At any point, at most N goroutines will be active processing tasks. 
	p = pool.DefaultFixedPool(10)

	// Creates an Executor that uses a single worker thread operating off an unbounded queue.
	p = pool.DefaultSinglePool()

}

Unbounded Chan

A byproduct of developing Goroutine Pool , a based linked table based Unbounded Channel with support for setting priorities

Usage

package main

import "github.com/gtoxlili/gtools/unboundedChan"

func main() {
	ch := unboundedChan.NewUnboundedChan[string]()
	ch.In <- unboundedChan.ChanItem[string]{
		Value:    "AAA",
		Priority: 1,
	}
	ch.In <- unboundedChan.ChanItem[string]{
		Value:    "BBB",
		Priority: 2,
	}
	ch.In <- unboundedChan.ChanItem[string]{
		Value:    "CCC",
		Priority: 3,
	}
	for i := 0; i < 3; i++ {
		val, ok := <-ch.Out
		if !ok {
			break
		}
		println(val)
	}
	// Output: CCC, BBB, AAA

	// close method
	close(ch.In)
}

2Q Cache

thread-safe memory cache based on two-queue (lru-2) algorithm,

In benchmark tests using bigcache-bench, the concurrent read capability is comparable to FastCache and BigCache, but the write capability is slightly worse, but because of the two-queue algorithm used, the hit rate is perhaps higher than theirs in actual scenarios?

Usage

package main

import "github.com/gtoxlili/gtools/twoQueues"

func main() {
	// Created a memory cache with two shards with a maximum capacity of 1024 
	// (theoretically, the more shards, the lower the possibility of lock conflicts during reading and writing, 
	// but it will also lead to more memory usage, so please decide the number of shards according to the actual maximum capacity to use)
	cache := twoQueues.New[string, int](1024, 2)
	cache.Set("a", 1)
	cache.Set("b", 2)
	println(cache.Get("a")) // 1 true
	println(cache.Get("b")) // 2 true
	println(cache.Get("c")) //   false
}