简单介绍tornado是如何实现异步非阻塞的

参考:tornado中的协程是如何工作的

写在之前

基础依旧很弱,欠缺理论知识(实践就更不要说了),很多东西都只知道个大概,哎。

Future

Future对象实际是coroutine函数装饰器和IOLoop的沟通使者,有着非常重要的作用。

tornado.concurrent.Futureconcurrent.futures.Future类似,但感觉比其简易。以tornado.concurrent.Future为例介绍该对象的重要方法:

  • _set_done()done(): 用于设置和获取该future的完成状态,并且在_set_done()时会顺序执行该对象的callback函数。
    def _set_done(self):
        self._done = True
        for cb in self._callbacks:
            try:
                cb(self)
            except Exception:
                app_log.exception('Exception in callback %r for %r',
                                  cb, self)
        self._callbacks = None
  • set_result()result(): 用于设置和获取该future的结果。与concurrent.futures.Future不同,如果该tornado.concurrent.Future没有完成时,调用result()会抛出异常(而concurrent.futures.Future是通过自身的condition条件来控制)。
def _check_done(self): 
    if not self._done: 
        raise Exception("DummyFuture does not support blocking for results")
  • add_done_callback: 为该对象添加回调函数,若该future已完成则直接执行,否则添加到_callbacks列表中在```_set_done()``时顺序执行。
    def add_done_callback(self, fn):
        """Attaches the given callback to the `Future`.

        It will be invoked with the `Future` as its argument when the Future
        has finished running and its result is available.  In Tornado
        consider using `.IOLoop.add_future` instead of calling
        `add_done_callback` directly.
        """
        if self._done:
            fn(self)
        else:
            self._callbacks.append(fn)

为什么别人博客说Future实际是@tornado.gen.coroutine和IOLoop的沟通使者呢?再看一下这两者的相关内容。

IOLoop

简单的理解IOLoop实际上就是一个事件循环,调度处理I/O事件和callback, timeout等事件。主要介绍add_future()add_callback方法(删除了部分注释,可以在源码中查看tornado.ioloop模块)。

    def add_future(self, future, callback):
        """Schedules a callback on the ``IOLoop`` when the given
        `.Future` is finished.

        The callback is invoked with one argument, the
        `.Future`.
        """
        assert is_future(future)
        callback = stack_context.wrap(callback)
        future.add_done_callback(
            lambda future: self.add_callback(callback, future))

    def add_callback(self, callback, *args, **kwargs):
        if thread.get_ident() != self._thread_ident:
            with self._callback_lock:
                if self._closing:
                    return
                list_empty = not self._callbacks
                self._callbacks.append(functools.partial(
                    stack_context.wrap(callback), *args, **kwargs))
                if list_empty:
                    self._waker.wake()
        else:
            if self._closing:
                return
            self._callbacks.append(functools.partial(
                stack_context.wrap(callback), *args, **kwargs))

查看了以上的源码可得,这两个方法实际上都是(在future完成后)通过add_callback()往IOLoop的_callbacks里一系列回调函数,使得IOLoop在下一次循环的时候得以顺序调用这些回调函数,即把控制权重新交还给了主线程。

@tornado.gen.coroutine和yield

一般来说@tornado.gen.coroutine和yield对应,被该装饰器装饰的方法需要存在yield(即是一个生成器)。
第一次看到以下的示例代码时,当时感觉与yield的使用相悖的。

class GenRequestHandler(BaseHandler):
    @tornado.gen.coroutine
    def get(self):
        http = httpclient.AsyncHTTPClient()
        res = yield http.fetch('http://www.baidu.com')
        self.write(res.body)

为什么yield http.fetch('http://www.baidu.com')直接返回了fetch得到的response?对于一般的生成器不应该是先调用next()或者send(None)然后遇到yield后交出控制权然后再通过send(result)来返回结果吗(如下例)?

def gen():
    print('start')
    first_send = yield 1
    print('first_send: %s.' % first_send)
    second_send = yield 2
    print('second_send: %s.' % second_send)
    print('end')


if __name__ == '__main__':
    try:
        g = gen()
        first_yielded = next(g)
        print('first_yielded: %s.' % first_yielded)
        second_yielded = g.send(1)
        print('second_yielded: %s.' % second_yielded)
        g.send(2)
    except StopIteration:
        print('stop')

实际上@tornado.gen.coroutine对原函数装饰了很多额外的操作,使得异步代码写起来和同步代码一样。上代码:

def coroutine(func, replace_callback=True):
    return _make_coroutine_wrapper(func, replace_callback=True)

def _make_coroutine_wrapper(func, replace_callback):
    func = types.coroutine(func)

    @functools.wraps(func)
    def wrapper(*args, **kwargs):
        future = TracebackFuture()

        if replace_callback and 'callback' in kwargs:
            callback = kwargs.pop('callback')
            IOLoop.current().add_future(
                future, lambda future: callback(future.result()))

        try:
            result = func(*args, **kwargs)
        except (Return, StopIteration) as e:
            result = _value_from_stopiteration(e)
        except Exception:
            future.set_exc_info(sys.exc_info())
            return future
        else:
            if isinstance(result, GeneratorType):
                try:
                    orig_stack_contexts = stack_context._state.contexts
                    yielded = next(result)
                    if stack_context._state.contexts is not orig_stack_contexts:
                        yielded = TracebackFuture()
                        yielded.set_exception(
                            stack_context.StackContextInconsistentError(
                                'stack_context inconsistency (probably caused '
                                'by yield within a "with StackContext" block)'))
                except (StopIteration, Return) as e:
                    future.set_result(_value_from_stopiteration(e))
                except Exception:
                    future.set_exc_info(sys.exc_info())
                else:
                    Runner(result, future, yielded)
                try:
                    return future
                finally:
                    future = None
        future.set_result(result)
        return future
    return wrapper

推荐阅读更多精彩内容