[编程经验] CVPR2017论文全集下载代码脚本分享

import urllib
import re
import os
import urlparse


def auto_down(url, filename):
    try:
        image = urllib.urlretrieve(url, filename)
    except urllib.ContentTooShortError:
        image = auto_down(url, filename)
    return image


def down_load_cvpr():
    urls = "http://openaccess.thecvf.com/CVPR2017.py"
    try:
        page = urllib.urlopen(urls)
        html = page.read()
        reg = r'href="(.+?\.pdf)"'
        imgre = re.compile(reg)
        pdf_list = re.findall(imgre, html)
        for pdf in pdf_list:
            target_path = "./CVPR_2017_paper"
            pdf_url = "http://openaccess.thecvf.com/" + str(pdf)
            url_parse = urlparse.urlparse(pdf_url)
            file_name = url_parse.path.split("/")[-1]
            if not os.path.exists(target_path):
                os.makedirs(target_path)
            target = os.path.join(target_path, '{}'.
                                    format(file_name))
            print "[*] Downloading paper :{}".format(pdf_url)
            auto_down(pdf_url, target)
    except:
        pass


if __name__ == "__main__":
    down_load_cvpr()