作为一个新世纪有思想有文化有道德时刻准备着的屌丝男青年,在现在这样一个社会中,心疼我大慢播抵制大百度的前提下,没事儿上上网逛逛YY看看斗鱼翻翻美女图片那是必不可少的,可是美图虽多翻页费劲!今天我们就搞个爬虫把美图都给扒下来!本次实例有2个:煎蛋上的妹子图,某网站的rosi图。我只是一个学习python的菜鸟,技术不可耻,技术是无罪的!!!
煎蛋:
先说说程序的流程:获取煎蛋妹子图URL,得到网页代码,提取妹子图片地址,访问图片地址并将图片保存到本地。Ready? 先让我们看看煎蛋妹子网页:
我们得到URL为:http://jandan.net/ooxx/page-1764#comments 1764就是页码, 首先我们要得到最新的页码,然后向前寻找,然后得到每页中图片的url。下面我们分析网站代码写出正则表达式!
根据之前文章的方法我们写出如下函数getNewPage:
1 def __getNewPage(self): 2 pageCode = self.Get(self.__Url) 3 type = sys.getfilesystemencoding() 4 pattern = re.compile(r'<div .*?cp-pagenavi">.*?<span .*?current-comment-page">[(.*?)]</span>',re.S) 5 newPage = re.search(pattern,pageCode.decode("UTF-8").encode(type)) 6 print pageCode.decode("UTF-8").encode(type) 7 if newPage != None: 8 return newPage.group(1) 9 return 1500
不要问我为什么如果失败返回1500。。。 因为煎蛋把1500页之前的图片都给吃了。 你也可以返回0。接下来是图片的
1 def __getAllPicUrl(self,pageIndex): 2 realurl = self.__Url + "page-" + str(pageIndex) + "#comments" 3 pageCode = self.Get(realurl) 4 type = sys.getfilesystemencoding() 5 pattern = re.compile('<p>.*?<a .*?view_img_link">.*?</a>.*?<img src="(.*?)".*?</p>',re.S) 6 items = re.findall(pattern,pageCode.decode("UTF-8").encode(type)) 7 for item in items: 8 print item
好了,得到了图片地址,接下来就是访问图片地址然后保存图片了:
1 def __savePics(self,img_addr,folder): 2 for item in img_addr: 3 filename = item.split('/')[-1] 4 print "正在保存图片:" + filename 5 with open(filename,'wb') as file: 6 img = self.Get(item) 7 file.write(img)
当你觉得信心满满的时候,一定会有一盆冷水浇到你的头上,毕竟程序就是这样,考验你的耐性,打磨你的自信。你测试了一会儿,然后你发现你重启程序后再也无法获取最新页码,你觉得我什么也没动啊为什么会这样。别着急,我们将得到的网页代码打印出来看看:
看到了吧,是服务器感觉你不像浏览器访问的结果把你的ip给屏蔽了。 真是给跪了,辛辛苦苦码一年,屏蔽回到解放前!那么这个如何解决呢,答:换ip 找代理。接下来我们要改一下我们的HttpClient.py 将里面的opener设置下代理服务器。具体代理服务器请自行百度之,关键字:http代理 。 想找到一个合适的代理也不容易 自己ie Internet选项挨个试试,测试下网速。
1 # -*- coding: utf-8 -*- 2 import cookielib, urllib, urllib2, socket 3 import zlib,StringIO 4 class HttpClient: 5 __cookie = cookielib.CookieJar() 6 __proxy_handler = urllib2.ProxyHandler({"http" : '42.121.6.80:8080'})#设置代理服务器与端口 7 __req = urllib2.build_opener(urllib2.HTTPCookieProcessor(__cookie),__proxy_handler)#生成opener 8 __req.addheaders = [ 9 ('Accept', 'application/javascript, */*;q=0.8'), 10 ('User-Agent', 'Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; WOW64; Trident/5.0)') 11 ] 12 urllib2.install_opener(__req) 13 14 def Get(self, url, refer=None): 15 try: 16 req = urllib2.Request(url) 17 #req.add_header('Accept-encoding', 'gzip') 18 if not (refer is None): 19 req.add_header('Referer', refer) 20 response = urllib2.urlopen(req, timeout=120) 21 html = response.read() 22 #gzipped = response.headers.get('Content-Encoding') 23 #if gzipped: 24 # html = zlib.decompress(html, 16+zlib.MAX_WBITS) 25 return html 26 except urllib2.HTTPError, e: 27 return e.read() 28 except socket.timeout, e: 29 return '' 30 except socket.error, e: 31 return ''
然后,就可以非常愉快的查看图片了。不过用了代理速度好慢。。。可以设置timeout稍微长一点儿,防止图片下载不下来!
好了,rosi的下篇文章再放!现在是时候上一波代码了:
1 # -*- coding: utf-8 -*- 2 import cookielib, urllib, urllib2, socket 3 import zlib,StringIO 4 class HttpClient: 5 __cookie = cookielib.CookieJar() 6 __proxy_handler = urllib2.ProxyHandler({"http" : '42.121.6.80:8080'}) 7 __req = urllib2.build_opener(urllib2.HTTPCookieProcessor(__cookie),__proxy_handler) 8 __req.addheaders = [ 9 ('Accept', 'application/javascript, */*;q=0.8'), 10 ('User-Agent', 'Mozilla/5.0 (compatible; MSIE 9.0; Windows NT 6.1; WOW64; Trident/5.0)') 11 ] 12 urllib2.install_opener(__req) 13 14 def Get(self, url, refer=None): 15 try: 16 req = urllib2.Request(url) 17 req.add_header('Accept-encoding', 'gzip') 18 if not (refer is None): 19 req.add_header('Referer', refer) 20 response = urllib2.urlopen(req, timeout=120) 21 html = response.read() 22 gzipped = response.headers.get('Content-Encoding') 23 if gzipped: 24 html = zlib.decompress(html, 16+zlib.MAX_WBITS) 25 return html 26 except urllib2.HTTPError, e: 27 return e.read() 28 except socket.timeout, e: 29 return '' 30 except socket.error, e: 31 return '' 32 33 def Post(self, url, data, refer=None): 34 try: 35 #req = urllib2.Request(url, urllib.urlencode(data)) 36 req = urllib2.Request(url,data) 37 if not (refer is None): 38 req.add_header('Referer', refer) 39 return urllib2.urlopen(req, timeout=120).read() 40 except urllib2.HTTPError, e: 41 return e.read() 42 except socket.timeout, e: 43 return '' 44 except socket.error, e: 45 return '' 46 47 def Download(self, url, file): 48 output = open(file, 'wb') 49 output.write(urllib2.urlopen(url).read()) 50 output.close() 51 52 # def urlencode(self, data): 53 # return urllib.quote(data) 54 55 def getCookie(self, key): 56 for c in self.__cookie: 57 if c.name == key: 58 return c.value 59 return '' 60 61 def setCookie(self, key, val, domain): 62 ck = cookielib.Cookie(version=0, name=key, value=val, port=None, port_specified=False, domain=domain, domain_specified=False, domain_initial_dot=False, path='/', path_specified=True, secure=False, expires=None, discard=True, comment=None, comment_url=None, rest={'HttpOnly': None}, rfc2109=False) 63 self.__cookie.set_cookie(ck) 64 #self.__cookie.clear() clean cookie 65 # vim : tabstop=2 shiftwidth=2 softtabstop=2 expandtab
1 # -*- coding: utf-8 -*- 2 from __future__ import unicode_literals 3 from HttpClient import HttpClient 4 import sys,re,os 5 class JianDan(HttpClient): 6 def __init__(self): 7 self.__pageIndex = 1500 #之前的图片被煎蛋吞了 8 self.__Url = "http://jandan.net/ooxx/" 9 self.__floder = "jiandan" 10 def __getAllPicUrl(self,pageIndex): 11 realurl = self.__Url + "page-" + str(pageIndex) + "#comments" 12 pageCode = self.Get(realurl) 13 type = sys.getfilesystemencoding() 14 pattern = re.compile('<p>.*?<a .*?view_img_link">.*?</a>.*?<img src="(.*?)".*?</p>',re.S) 15 items = re.findall(pattern,pageCode.decode("UTF-8").encode(type)) 16 for item in items: 17 print item 18 self.__savePics(items,self.__floder) 19 20 def __savePics(self,img_addr,folder): 21 for item in img_addr: 22 filename = item.split('/')[-1] 23 print "正在保存图片:" + filename 24 with open(filename,'wb') as file: 25 img = self.Get(item) 26 file.write(img) 27 28 def __getNewPage(self): 29 pageCode = self.Get(self.__Url) 30 type = sys.getfilesystemencoding() 31 pattern = re.compile(r'<div .*?cp-pagenavi">.*?<span .*?current-comment-page">[(.*?)]</span>',re.S) 32 newPage = re.search(pattern,pageCode.decode("UTF-8").encode(type)) 33 print pageCode.decode("UTF-8").encode(type) 34 if newPage != None: 35 return newPage.group(1) 36 return 1500 37 38 def start(self): 39 isExists=os.path.exists(self.__floder)#检测是否存在目录 40 print isExists 41 if not isExists: 42 os.mkdir(self.__floder) 43 os.chdir(self.__floder) 44 page = int(self.__getNewPage()) 45 for i in range(self.__pageIndex,page): 46 self.__getAllPicUrl(i) 47 48 if __name__ == '__main__': 49 jd = JianDan() 50 jd.start()