1.分析网站

target_url:http://www.dilidili.name/watch3/32061/
目标数据源：

千与千寻.png

2.开个多进程爬取资源

from multiprocessing import Pool
from functools import reduce
import requests
import os

# 安装python, 安装requests,就可以愉快的开车啦！

path = '千与千寻/'

start = 0

length = 1866

speed = 20

if not os.path.exists(path):
    os.mkdir(path)

def get_content(iter):
    for i in iter:
        try:
            if i < 10:
                url = 'http://kbzy.zxziyuan-yun.com/20180404/XVXMBPEV/800kb/hls/2Tb76g334100{}.ts'.format(i)
            elif i < 100:
                url = 'http://kbzy.zxziyuan-yun.com/20180404/XVXMBPEV/800kb/hls/2Tb76g33410{}.ts'.format(i)
            else:
                url = 'http://kbzy.zxziyuan-yun.com/20180404/XVXMBPEV/800kb/hls/2Tb76g3341{}.ts'.format(i)
            r = requests.get(url)
            if b'<h1>\xb7\xfe\xce\xf1\xc6\xf7\xb4\xed\xce\xf3</h1>' not in r.content:
                with open(path + '千与千寻{}.ts'.format(i), 'wb') as f:
                    f.write(r.content)
                    print('加载完成', i)
        except:
            print('下载完成...')


def open_file(x):
    with open(x, 'rb') as f:
        return f.read()


def start_task():
    lst = list(range(start, length))
    result = [lst[x:x + int((len(lst) / speed))] for x in range(len(lst)) if x % (len(lst) / speed) == 0]
    pool = Pool(speed)
    for target in result:
        pool.apply_async(get_content, args=(target,))
    pool.close()
    pool.join()
    print('下载完成...')
    return True


if __name__ == '__main__':
    if start_task():
        # 视频拼接
        lst = [path + '千与千寻{}.ts'.format(x) for x in range(length)]
        source = list(map(open_file, lst))
        z = reduce(lambda x, y: x + y, source)
        with open('千与千寻.ts', 'wb') as f:
            f.write(z)

优化可以把地址做个缓存池避免重复爬取，但我懒。。。

成果：http://59.110.157.193:8000/media/video/%E5%8D%83%E4%B8%8E%E5%8D%83%E5%AF%BB.ts