主要是解决centos中下载文件用wget 速度提不上去
from concurrent.futures import ThreadPoolExecutor, wait from threading import Lock from requests import get, head import sys lock = Lock() class Downloader(): def __init__(self, url, nums, file): self.url = url self.num = nums self.name = file r = head(self.url) # 若资源显示302,则迭代找寻源文件 while r.status_code == 302: self.url = r.headers['Location'] print("该url已重定向至{}".format(self.url)) r = head(self.url) self.size = int(r.headers['Content-Length']) print('该文件大小为：{} bytes'.format(self.size)) def down(self, start, end): headers = {'Range': 'bytes={}-{}'.format(start, end)} # stream = True 下载的数据不会保存在内存中 r = get(self.url, headers=headers, stream=True) # 写入文件对应位置,加入文件锁 lock.acquire() with open(self.name, "rb+") as fp: fp.seek(start) fp.write(r.content) lock.release() # 释放锁 def run(self): # 创建一个和要下载文件一样大小的文件 fp = open(self.name, "wb") fp.truncate(self.size) fp.close() # 启动多线程写文件 part = self.size // self.num pool = ThreadPoolExecutor(max_workers=self.num) futures = [] for i in range(self.num): start = part * i # 最后一块 if i == self.num - 1: end = self.size else: end = start + part - 1 print('{}->{}'.format(start, end)) futures.append(pool.submit(self.down, start, end)) wait(futures) print('%s 下载完成' % self.name) try: url=sys.argv[1] except: url='https://upload.wikimedia.org/wikipedia/commons/thumb/7/79/AL_postcode_area_map.svg/400px-AL_postcode_area_map.svg.png' filename=url.split('/')[-1] ss = Downloader(url,6, filename) ss.run()