Question

我无法掌握这段代码的执行流程：

func middlewareOne(next http.Handler) http.Handler {
    return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
        log.Println("Executing middlewareOne")
        next.ServeHTTP(w, r)
        log.Println("Executing middlewareOne again")
        next.ServeHTTP(w, r)
        log.Println("Executing middlewareOne yupiii")
    })
}

func middlewareTwo(next http.Handler) http.Handler {
    return http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
        log.Println("Executing middlewareTwo")
        next.ServeHTTP(w, r)
        log.Println("Executing middlewareTwo again")
        next.ServeHTTP(w, r)
    })
}

func final(w http.ResponseWriter, r *http.Request) {
    log.Println("Executing finalHandler")
    w.Write([]byte("OK"))
}

func main() {
    finalHandler := http.HandlerFunc(final)
    http.Handle("/", middlewareOne(middlewareTwo(finalHandler)))
    http.ListenAndServe(":3000", nil)
}

输出：

2019/03/25 09:16:32 Executing middlewareOne
2019/03/25 09:16:32 Executing middlewareTwo
2019/03/25 09:16:32 Executing finalHandler
2019/03/25 09:16:32 Executing middlewareTwo again
2019/03/25 09:16:32 Executing finalHandler
2019/03/25 09:16:32 Executing middlewareOne again
2019/03/25 09:16:32 Executing middlewareTwo
2019/03/25 09:16:32 Executing finalHandler
2019/03/25 09:16:32 Executing middlewareTwo again
2019/03/25 09:16:32 Executing finalHandler
2019/03/25 09:16:32 Executing middlewareOne yupiii

为什么在middlewareTwo again之前打印middlewareOne again？

Answer 1

让堆栈成为一棵树，更容易理解：

middlewareOne()
    middlewareTwo()
        finalHandler()
            return finalHandler
        return middlewareTwo

    middlewareTwo again()
        finalHandler()
             return finalHandler()
        return middlewareTwo again

    return middleware one

middlewareOne again()
...

为了完成middleware one并移至middleware one again，它需要middleware two才能返回，因为middleware two包含middleware two again，因此需要运行{在完成之前再次{1}}。

中间件执行流程

1 个答案: