yield实现斐波那契序列:
import sys, time def fib(): a,b,c = 0,1,0 while True: yield c a,b = b, c c = a + b if __name__ == '__main__': fib_iter = fib() for i in range(int(sys.argv[1])): print(fib_iter.__next__())
Send:
send函数用来向fib_iter发送数据,这样数据就可以双向流动。我们尝试使用send来模拟一个比较慢的生成器,我们让它一秒钟生成一个数:
import sys, time def fib(): a,b,c = 0,1,0 while True: sleep_sec = yield c time.sleep(sleep_sec) a,b = b, c c = a + b def fib1(): index = 1 a = 0 b = 1 while index: yield b a, b = b, a + b index += 1 if __name__ == '__main__': fib_iter = fib() print(fib_iter.__next__()) #先执行一下,让它停留在yield for i in range(int(sys.argv[1])): result = fib_iter.send(1) # print(result)
yield from 是什么?
在上面的yield中,我们通过for循环使用__next__()方法来获取下一个值,也就是说想要获取下一个值就要重新yield一下;yield from 它可以简化这个过程,看看实例:
import sys, time def fib(n): a,b,c = 0,1,0 while c < n: yield c a,b = b, c c = a + b def gener(n): yield from fib(n) if __name__ == '__main__': print(list(gener(5000)))
执行结果:
D:>python fib.py
[0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4
181]
yield from , send传递信息:
import sys, time def fib(n): a,b,c = 0,1,0 while c < n: sleep_sec = yield c time.sleep(sleep_sec) a,b = b, c c = a + b def gener(n): yield from fib(n) if __name__ == '__main__': gen = gener(10) print(gen.send(None)) print(gen.send(1)) print(gen.send(1))
执行结果:
D:>python fib.py
0
1
1
asyncio 和 yield from:
asyncio是一个基于事件循环的实现异步的I/O模块。通过yield from, 我们可以将协程asynico.sleep的控制权交给事件循环,然后挂起当前协程;之后,由事件循环决定何时唤醒asyncio.sleep,然后接着执行后面的代码; 实例中我们用sleep来模拟阻塞:
import sys, time, asyncio def fib(n): a,b,c = 0,1,0 while c < n: yield from asyncio.sleep(2) print('-->', c) a,b = b, c c = a + b def stupid_fib(n): a,b,c = 0,1,0 while c < n: yield from asyncio.sleep(2) print('==>', c) a,b = b,c c = a + b if __name__ == '__main__': loop = asyncio.get_event_loop() tasks = [asyncio.async(fib(10)), asyncio.async(stupid_fib(10))] loop.run_until_complete(asyncio.wait(tasks)) print('all task finished') loop.close()
执行结果:
D:>python fib.py --> 0 ==> 0 --> 1 ==> 1 --> 1 ==> 1 --> 2 ==> 2 --> 3 ==> 3