python3源码剖析之concurrent.futures.ThreadPoolExecutor

来源:互联网 发布:陕西师范网络教育首页 编辑:程序博客网 时间:2024/06/05 11:08

0x00 文档路径:

  • Python » 3.6.2 Documentation » The Python Standard Library » 17. Concurrent Execution »

0x01 模块简述:

添加于python3.2

提供更加高效的接口来实现异步执行

通过具体实现来剖析

0x02 具体实现

参考官方文档给出的例子

from concurrent.futures import ThreadPoolExecutor
with ThreadPoolExecutor(max_workers=1) as executor:    future = executor.submit(pow, 323, 1235)    print(future.result())
  1. 通过ThreadPoolExecutor来生成一个Executor对象

    源码位置

    Lib\concurrent\futures\thread.py 83行

    一共两个参数max_workers=None、thread_name_profix=”

    max_workers

    用来指定最大线程数

    if max_workers is None:    # Use this number because ThreadPoolExecutor is often    # used to overlap I/O instead of CPU work.    max_workers = (os.cpu_count() or 1) * 5

    先判断是否是None,即未指定时,默认是cpu数量*5,比如你是四核的cpu,那么默认最大线程20

    再判断是否小于0,若小于则抛出ValueError

  2. 调用ThreadPoolExecutor对象的submit方法

    ThreadPoolExecutor继承了Executor对象,并实现了部分重写
    源码位置

    Lib\concurrent\futures\thread.py 106行

    一共三个参数 fn , *rags , **kwargs
    submit方法执行 fn(*args,**kwargs) , 然后返回一个Future对象

  3. 调用Future对象的result方法

    返回被执行函数的结果
    源码位置

    Lib\concurrent\futures\_base.py 378行

    一个参数timeout = None

    不指定时,将不会限制等待时间,即一直等到函数完成
    如果在指定时间内函数未完成,则抛出异常TimtoutError

  4. 因为with上下文管理器的原因,自动调用Executor对象的shutdown方法来释放资源

    直接使用with 语句即可

另一个例子

from telnetlib import Telnetdef detect_port(port):    try:        print('{} testing .. '.format(port))        Telnet('127.0.0.1',port,timeout=5)        print('{} opened '.format(port))    except:        passwith ThreadPoolExecutor(100)as executor:  executor.map(detect_port,range(500))

map方法

完全继承于Executor

源码位置

Lib\concurrent\futures\_base.py

一共四个参数 fn , *iterables , timeout = None , chunksize = 1

chunsize用于ProcessPoolExecutor,其他的参数同上面的submit

def map(self, fn, *iterables, timeout=None, chunksize=1):    if timeout is not None:        end_time = timeout + time.time()    fs = [self.submit(fn, *args) for args in zip(*iterables)]    def result_iterator():        try:            for future in fs:                if timeout is None:                    yield future.result()                else:                    yield future.result(end_time - time.time())        finally:            for future in fs:                future.cancel()    return result_iterator()

map里面调用了内置函数zip,zip函数的特点是

In [1]: print(*zip('abc',[1,2,3,4]))('a', 1) ('b', 2) ('c', 3)

举例说

def detect_port(ip,port):    检测端口with ThreadPoolExecutor()as ecexutor:    executor.map(detect_port,['127.0.0.1','192.168.1.10'],['21','22','23'])

原本是打算逐个ip扫描这三个端口是否开放

但是此时只会扫描127.0.0.1的21端口,以及192.168.1.10的23端口

def detect_port(ip,port):    检测端口def detect_all_ip(ip):    ports = [xx,xx,..]    with ThreadPoolExecutor()as executor:        [executor.submit(detect_port,ip,ports) for port in ports]def main():    ips = ['xxx','xxx',..]    with ThreadPoolExecutor()as executor:        executor.map(detect_all_ip,ips)

0x03 myself

如下是之前写的通过telnet扫描端口的脚本,便是利用了ThreadPoolExecutor,可供参考
detectPortByTelnet.py下载

新搭了博客,有兴趣的朋友们可以去踩
楼兰’s Blog