常用

package mainimport (    "context"    "fmt"    "go.mongodb.org/mongo-driver/bson"    "hugo/mon"    "hugo/publishtask"    "strconv"    "sync"    "sync/atomic"    "time")type Abs struct {    UserName string `bson:"username"`    Email string `bson:"email"`}func b()(i int,err error){    if err := fmt.Errorf("aaaa");err!=nil{        fmt.Println(err,232323)    }    fmt.Println(err,8888)    return}func main(){    publishtask.CreateUser()    return    abs := new(Abs)    fmt.Println(time.Now().Format("2006-01-02 15:04:05"))    fmt.Println(time.Now().Second())    singResult :=mon.Client.Collection("tbl_data").FindOne(context.TODO(),bson.M{"username":"胡工827"})    singResult.Decode(abs)    fmt.Println(time.Now().Format("2006-01-02 15:04:05"))    fmt.Println(time.Now().Second())    return    wg := new(sync.WaitGroup)    for i := 0; i < 100; i++ {        wg.Add(1)        go insertMon(wg)    }    wg.Wait()    fmt.Println(time.Now().Format("2006-01-02 15:04:05"))}func insertMon(wg *sync.WaitGroup){    defer func() {        wg.Done()    }()    data := Abs{"hugo","136586551@163.com"}    for i := 1; i < 100000; i++ {        iStr := strconv.Itoa(i)        data.UserName = fmt.Sprintf("胡工%s",iStr)        insertRes,err:=mon.Client.Collection("tbl_data").InsertOne(context.TODO(),data)        if err != nil {            fmt.Println(err,insertRes)        }    }}func doSth(){    //ctx,cancel :=context.WithTimeout(context.Background(),time.Duration(150)* time.Microsecond)    timer := time.NewTimer(2*time.Second)    //go sth(ctx)    for range timer.C{        fmt.Println(time.Now().Format("2006-01-02 15:04:05"))    }    return    for {        select{        case <-timer.C:            fmt.Println("定时器时间到了，开始要取消cancel()\n")            //cancel()            return        default:            //fmt.Println("还没有开始123")        }    }}func sth(ctx context.Context){    fmt.Println(1111)    for{        time.Sleep(time.Microsecond * 100)        select {        case <-ctx.Done():            fmt.Println("取消结束了")            return        default:            fmt.Println("还没有被结束567")        }    }}//定时器每隔2秒执行一次代码块func timer1(){    timer := time.NewTimer(2*time.Second)    for range timer.C{        fmt.Println(time.Now().Format("2006-01-02 15:04:05"))    }}//定时器隔2秒执行一次代码块,仅仅执行一次func timer2(){    timer := time.NewTimer(2*time.Second)    i :=0    for true {        select{        case <-timer.C:            fmt.Println("定时器执行了")        default:            if i == 0 {                fmt.Println("定时器还没执行了333333")            }        }        i++    }}//context 执行//for 里面的break  会直接终止for循坏//select 里面的case里面的break是终止case里面的代码块执行，不会终止到select外层的for//select 里面不用break也只会执行一个casefunc context1(){    ctx,cancel :=context.WithTimeout(context.Background(),2 * time.Second)    i:=0    go func(ctx context.Context) {        for{            select{                case <-time.After(10 * time.Nanosecond):                    fmt.Println("时间1")                    fmt.Println("78787878")                case <-ctx.Done():                    fmt.Println("取消的命令来了")                    return                default:                    if i == 0 {                        fmt.Println("11122333")                    }                    break            }            i++        }    }(ctx)    time.Sleep(1 * time.Second)    fmt.Println("cancel来取消的")    cancel()}//执行到的switch的case如果有 fallthrough  就会在不判断的情况下直接执行下一个casefunc switch1(){    df :=1    switch df{    case 1:        fmt.Println(11)        fallthrough    case 2,3,4:        fmt.Println(222)        fallthrough    case 22,32,42:        fmt.Println(8989)    case 222,322,422:        fmt.Println(84378434)    }}type Mutex struct {    key  int32    sema uint32}//sync.lock,sync.Rlock,//Lock会让 Lock,Rlock等待，也就是锁住的意思//Rlock会让 lock等待，但是不会让Rlock等待func lock1(){    m := new(sync.RWMutex)    // 多个同时读    go lockRead(1,m)    go lockRead(2,m)    time.Sleep(1*time.Second)    go lockRead3(3,m)    time.Sleep(15*time.Second)}func lockRead(i int,m *sync.RWMutex) {    println(i,"read start")    m.RLock()    println(i,"reading")    time.Sleep(5*time.Second)    m.RUnlock()    println(i,"read over")}func lockRead3(i int,m *sync.RWMutex) {    println(i,"read3 start")    m.Lock()    println(i,"read3ing")    time.Sleep(1*time.Second)    m.Unlock()    println(i,"read3 over")}//后面sync.atomic//加锁是一种悲观策略。原来所有锁的底层又跑到操作系统那里去了，想想也是琐是用来锁住多线程的，线程上游是进程，进程上游是操作系统，最终交汇点进程无疑//操作系统的锁策略是时间换空间的策略，从而保证数据安全//无锁策略，利用到了cps算法，并没有惊动到操作系统那一层，让各线程自身就具备鉴别是否资源已经被抢占，https://blog.csdn.net/yanluandai1985/article/details/82686486func atomic1(){    m := int32(0)    wg := new(sync.WaitGroup)    for i := 1; i < 1000; i++ {        wg.Add(1)        go func() {            if d :=atomic.AddInt32(&m,1);d==1{                //fmt.Println(123)            }else{                //fmt.Println(d,222)            }            wg.Done()        }()    }    wg.Wait()    fmt.Println(m)}

给这篇文章的作者打赏

微信扫一扫打赏

支付宝扫一扫打赏

×

打赏 点赞(1)