编写Python脚本批量下载DesktopNexus壁纸的教程

DesktopNexus 是我最喜爱的一个壁纸下载网站，上面有许多高质量的壁纸，几乎每天必上， 每月也必会坚持分享我这个月来收集的壁纸

但是 DesktopNexus 壁纸的下载很麻烦，而且因为壁纸会通过浏览器检测你当前分辨率来展示 合适你当前分辨率的壁纸，再加上是国外的网站，速度上很不乐观。

于是我写了个脚本，检测输入的页面中壁纸页面的链接，然后批量下载到指定文件夹中。

脚本使用 python 写的，所以需要机器上安装有 python 。
用法：

    $ python desktop_nexus.py -p http://www.desktopnexus.com/tag/cat/ -s 1280x800 -o wallpapers

-p 包含 DesktopNexus 壁纸链接的页面，比如我的壁纸分享  
-s 壁纸尺寸，可选，缺省为 1440x900  
-o 壁纸输出的文件夹，可选，缺省为当前目录下的 wallpapers， 如果不存在会自动创建

代码：

    #-*- coding: utf-8 -*-
    from argparse import ArgumentParser

    import os, re, sys
    import urllib2, cookielib, urlparse

    RE_WALLPAPER = r'http\:\/\/[^\/\.]+\.desktopnexus\.com\/wallpaper\/\d+\/'
    CHUNK_SIZE = 1024 * 3

    class DesktopNexus:
      def __init__(self, page=None, size=None, output_dir=None):
        self.page = page
        self.size = size
        self.output_dir = output_dir

      def start(self):
        print 'Making output directory:', self.output_dir
        if not os.path.exists(self.output_dir):
          os.makedirs(self.output_dir)

        # Setup cookie
        cookie = cookielib.CookieJar()
        processer = urllib2.HTTPCookieProcessor(cookie)
        opener = urllib2.build_opener(processer)
        urllib2.install_opener(opener)

        self._read_page()

      def _get_pic_info(self, url):
        pic_id = url.split('/')[-2]
        html = urllib2.urlopen(url).read()
        pattern = r'<a href=\"\/get\/%s\/\?t=(?P<token>.*?)\"' % pic_id
        match = re.search(pattern, html, flags=re.I|re.M|re.S)
        if match:
          return {'id': pic_id,
              'token': match.group('token'),
              'size': self.size}
        else:
          raise Exception('Cound not find wallpaper')

      def _get_pic_file(self, pic_info):
        redirect_url = 'http://www.desktopnexus.com/dl/inline/%(id)s/%(size)s/%(token)s' % pic_info

        request = urllib2.urlopen(redirect_url)
        return request.geturl()

      def _download_pic(self, url):
        pic_info = self._get_pic_info(url)
        pic_file = self._get_pic_file(pic_info)
        filename = os.path.split(urlparse.urlparse(pic_file).path)[-1]
        filename = os.path.join(self.output_dir, filename)
        with open(filename, 'wb') as output:
          resp = urllib2.urlopen(pic_file)
          total_size = int(resp.info().get('Content-Length'))
          saved_size = 0.0
          while saved_size != total_size:
            chunk = resp.read(CHUNK_SIZE)
            saved_size += len(chunk)
            output.write(chunk)
            self._print_progress('Saving file: %s' % filename, \
                saved_size / total_size * 100)

      def _print_progress(self, msg, progress):
        sys.stdout.write('%-71s%3d%%\r' \
            % (len(msg) <= 70 and msg or msg[:67] + '...', progress))
        sys.stdout.flush()
        if progress >= 100:
          sys.stdout.write('\n')

      def _read_page(self):
        try:
          print 'Fetching content:', self.page
          html = urllib2.urlopen(self.page).read()
          links = set(re.findall(RE_WALLPAPER, html, re.M|re.I))
          count = len(links)

          print 'Downloading wallpapers:'
          for i, link in enumerate(links):
            print '[%d/%d]: %s' % (i + 1, count, link)
            try:
              self._download_pic(link)
            except Exception as e:
              print 'Error downloading wallpaper.', e.message
        except Exception as e:
          print 'Error fetching content.', e

    if __name__ == '__main__':
      # Setup argparser
      parser = ArgumentParser('python desktop_nexus.py')
      parser.add_argument('-p', '--page', dest='page', required=True, \
          help='specific a page that includes wallpaper list')
      parser.add_argument('-s', '--size', dest='size', default='1440x900', \
          help='specific the wallpaper size, default to 1440x900')
      parser.add_argument('-o', '--output', dest='output_dir', default='wallpapers', \
          help='specific the output directory, default to "wallpapers"')
      args = parser.parse_args()
      dn = DesktopNexus(**args.__dict__)
      dn.start()