IO异步

2022-09-14 18:58:55 浏览数 (1)

一、说明

对于耗时的过程,我们将其交给别人(如其另外一个线程)去执行,而我们继续往下处理,当别人执行完耗时操作后再将结果反馈给我们,这就是我们所说的异步

二、回调写法实现原理

​```python import time import threading

def long_io(callback): “”“将耗时的操作交给另一线程来处理”"" def fun(cb): # 回调函数作为参数 “”“耗时操作”"" print(“开始执行IO操作”) time.sleep(5) print(“完成IO操作,并执行回调函数”) cb(“io result”) # 执行回调函数 threading.Thread(target=fun, args=(callback,)).start() # 开启线程执行耗时操作

def on_finish(ret): “”“回调函数”"" print(“开始执行回调函数on_finish”) print(“ret: %s” % ret) print(“完成执行回调函数on_finish”)

def req_a(): print(“开始处理请求req_a”) long_io(on_finish) print(“离开处理请求req_a”)

def req_b(): print(“开始处理请求req_b”) time.sleep(2) # 添加此句来突出显示程序执行的过程 print(“完成处理请求req_b”)

def main(): req_a() req_b()

if name == ‘main’: main()

代码语言:javascript复制
执行过程:

开始处理请求req_a 开始执行IO操作 离开处理请求req_a 开始处理请求req_b 完成处理请求req_b 完成IO操作,并执行回调函数 开始执行回调函数on_finish ret: io result 完成执行回调函数on_finish

代码语言:javascript复制
说明:异步的特点是程序存在多个步调,即本属于同一个过程的代码可能在不同的步调上同时执行

三、协程写法实现原理

说明

在使用回调函数写异步程序时,需将本属于一个执行逻辑(处理请求a)的代码拆分成两个函数req_a和on_finish,这与同步程序的写法相差很大。而同步程序更便于理解业务逻辑,所以我们能否用同步代码的写法来编写异步程序

初始版本

代码语言:javascript复制
<span class="hljs-keyword">import</span> time
<span class="hljs-keyword">import</span> threading

gen = <span class="hljs-keyword">None</span>  <span class="hljs-comment"># 全局生成器,供long_io使用</span>

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">long_io</span><span class="hljs-params">()</span>:</span>
    <span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">fun</span><span class="hljs-params">()</span>:</span>
        print(<span class="hljs-string">"开始执行IO操作"</span>)
        <span class="hljs-keyword">global</span> gen
        time.sleep(<span class="hljs-number">5</span>)
        <span class="hljs-keyword">try</span>:
            print(<span class="hljs-string">"完成IO操作,并send结果唤醒挂起程序继续执行"</span>)
            gen.send(<span class="hljs-string">"io result"</span>)  <span class="hljs-comment"># 使用send返回结果并唤醒程序继续执行</span>
        <span class="hljs-keyword">except</span> StopIteration: <span class="hljs-comment"># 捕获生成器完成迭代,防止程序退出</span>
            <span class="hljs-keyword">pass</span>
    threading.Thread(target=fun).start()

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">req_a</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始处理请求req_a"</span>)
    ret = <span class="hljs-keyword">yield</span> long_io()
    print(<span class="hljs-string">"ret: %s"</span> % ret)
    print(<span class="hljs-string">"完成处理请求req_a"</span>)

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">req_b</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始处理请求req_b"</span>)
    time.sleep(<span class="hljs-number">2</span>)
    print(<span class="hljs-string">"完成处理请求req_b"</span>)

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">main</span><span class="hljs-params">()</span>:</span>
    <span class="hljs-keyword">global</span> gen
    gen = req_a()
    next(gen) <span class="hljs-comment"># 开启生成器req_a的执行</span>
    req_b()


<span class="hljs-keyword">if</span> __name__ == <span class="hljs-string">'__main__'</span>:
    main()

执行过程:

代码语言:javascript复制
  开始处理请求req_a
  开始执行IO操作
  开始处理请求req_b
  完成处理请求req_b
  完成IO操作,并send结果唤醒挂起程序继续执行
  ret: io result
  完成处理请求req_a

升级版本

说明:我们在上面编写出的版本虽然req_a的编写方式很类似与同步代码,但是在main中调用req_a的时候却不能将其简单的视为普通函数,而是需要作为生成器对待

代码语言:javascript复制
<span class="hljs-keyword">import</span> time
<span class="hljs-keyword">import</span> threading

gen = <span class="hljs-keyword">None</span>  <span class="hljs-comment"># 全局生成器,供long_io使用</span>

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">gen_coroutine</span><span class="hljs-params">(f)</span>:</span>
    <span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">wrapper</span><span class="hljs-params">(*args, **kwargs)</span>:</span>
        <span class="hljs-keyword">global</span> gen
        gen = f()
        next(gen)
    <span class="hljs-keyword">return</span> wrapper

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">long_io</span><span class="hljs-params">()</span>:</span>
    <span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">fun</span><span class="hljs-params">()</span>:</span>
        print(<span class="hljs-string">"开始执行IO操作"</span>)
        <span class="hljs-keyword">global</span> gen
        time.sleep(<span class="hljs-number">5</span>)
        <span class="hljs-keyword">try</span>:
            print(<span class="hljs-string">"完成IO操作,并send结果唤醒挂起程序继续执行"</span>)
            gen.send(<span class="hljs-string">"io result"</span>)  <span class="hljs-comment"># 使用send返回结果并唤醒程序继续执行</span>
        <span class="hljs-keyword">except</span> StopIteration: <span class="hljs-comment"># 捕获生成器完成迭代,防止程序退出</span>
            <span class="hljs-keyword">pass</span>
    threading.Thread(target=fun).start()

<span class="hljs-meta">@gen_coroutine</span>
<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">req_a</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始处理请求req_a"</span>)
    ret = <span class="hljs-keyword">yield</span> long_io()
    print(<span class="hljs-string">"ret: %s"</span> % ret)
    print(<span class="hljs-string">"完成处理请求req_a"</span>)

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">req_b</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始处理请求req_b"</span>)
    time.sleep(<span class="hljs-number">2</span>)
    print(<span class="hljs-string">"完成处理请求req_b"</span>)

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">main</span><span class="hljs-params">()</span>:</span>
    req_a()
    req_b()


<span class="hljs-keyword">if</span> __name__ == <span class="hljs-string">'__main__'</span>:
    main()

执行过程:

代码语言:javascript复制
  开始处理请求req_a
  开始执行IO操作
  开始处理请求req_b
  完成处理请求req_b
  完成IO操作,并send结果唤醒挂起程序继续执行
  ret: io result
  完成处理请求req_a

最终版本

代码语言:javascript复制
<span class="hljs-keyword">import</span> time
<span class="hljs-keyword">import</span> threading

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">gen_coroutine</span><span class="hljs-params">(f)</span>:</span>
    <span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">wrapper</span><span class="hljs-params">(*args, **kwargs)</span>:</span>
        gen_f = f()  <span class="hljs-comment"># gen_f为生成器req_a</span>
        r = next(gen_f)  <span class="hljs-comment"># r为生成器long_io</span>
        <span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">fun</span><span class="hljs-params">(g)</span>:</span>
            ret = next(g) <span class="hljs-comment"># 执行生成器long_io</span>
            <span class="hljs-keyword">try</span>:
                gen_f.send(ret) <span class="hljs-comment"># 将结果返回给req_a并使其继续执行</span>
            <span class="hljs-keyword">except</span> StopIteration:
                <span class="hljs-keyword">pass</span>
        threading.Thread(target=fun, args=(r,)).start()
    <span class="hljs-keyword">return</span> wrapper

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">long_io</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始执行IO操作"</span>)
    time.sleep(<span class="hljs-number">5</span>)
    print(<span class="hljs-string">"完成IO操作,yield回操作结果"</span>)
    <span class="hljs-keyword">yield</span> <span class="hljs-string">"io result"</span>

<span class="hljs-meta">@gen_coroutine</span>
<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">req_a</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始处理请求req_a"</span>)
    ret = <span class="hljs-keyword">yield</span> long_io()
    print(<span class="hljs-string">"ret: %s"</span> % ret)
    print(<span class="hljs-string">"完成处理请求req_a"</span>)

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">req_b</span><span class="hljs-params">()</span>:</span>
    print(<span class="hljs-string">"开始处理请求req_b"</span>)
    time.sleep(<span class="hljs-number">2</span>)
    print(<span class="hljs-string">"完成处理请求req_b"</span>)

<span class="hljs-function"><span class="hljs-keyword">def</span> <span class="hljs-title">main</span><span class="hljs-params">()</span>:</span>
    req_a()
    req_b()


<span class="hljs-keyword">if</span> __name__ == <span class="hljs-string">'__main__'</span>:
    main()

执行过程:

代码语言:javascript复制
  开始处理请求req_a
  开始执行IO操作
  开始处理请求req_b
  完成处理请求req_b
  完成IO操作,yield回操作结果
  ret: io result
  完成处理请求req_a

说明:这个最终版本就是理解Tornado异步编程原理的最简易模型,但是,Tornado实现异步的机制不是线程,而是epoll,即将异步过程交给epoll执行并进行监视回调

0 人点赞