python图片抓取

# -*- coding: cp936 -*-  
import urllib  
import urllib2  
import random  
import os,sys  
from sgmllib import SGMLParser  
  
class URLLister(SGMLParser):  
    '''''获取html中的图片地址\url地址，装入list中'''  
    def reset(self):  
        SGMLParser.reset(self)  
        self.img = []  
        self.urls = []  
    def start_img(self, attrs):  
        img = [v for k, v in attrs if k=='src']  
        if img:  
            self.img.extend(img)  
    def start_a(self, attrs):  
        href = [v for k, v in attrs if k=='href']  
        if href:  
            self.urls.extend(href)  
  
  
def get_docum(url):  
    #url=url+'//'  
    sock=urllib.urlopen(url)  
    file=sock.read()  
    sock.close()  
    return file  
  
def is_img(url):  
    global imglenth  
    reqst=urllib2.Request(url)  
    opener=urllib2.build_opener()  
    try:  
        con=opener.open(reqst)  
        Type=con.headers.dict['content-type'][:5] #判断链接返回的 content-type是不是图片。  
        Length =int(con.headers.dict['content-length'])#判断图片大小  
        if Length>imglenth:  
            return Type  
        else:  
            return 0  
    except:  
        print sys.exc_info()[0],sys.exc_info()[1]  ##一般来说这样就足够了  
        print '该图片无法在服务器找到或者图片地址无法识别！'  
        print url  
     
def get_file_name(ospath,imgname,num):   
    #name = 'P'+str(random.randint(10000000,99999999))  
    #filepath = "%s%s.%s" % (ospath,name,(imgname.split('.'))[-1])  
    #保留原文件名  
    idx=imgname.rfind("/")  
    filename=imgname  
    if idx > -1:  
        filename=imgname[idx+1:]  
    filepath=ospath+'%d_'%num+filename  
    print filepath  
    return filepath                  
  
def get_img(rq):  
    parser = URLLister();    doc=get_docum(rq);    parser.feed(doc);    img = parser.img  
    parser.close()  
    for i in range(0,len(img)):  
        if img[i][0:4]!='http':#处理绝对路径  
            img[i]=rq+img[i]  
    return img  
  
def get_url(rq):  
    parser = URLLister();    doc=get_docum(rq);    parser.feed(doc);    urls = parser.urls  
    parser.close()  
    for i in range(0,len(urls)):  
        if urls[i][0:4] != 'http': #处理绝对路径  
            urls[i] = rq+urls[i]  
    return urls  
  
def depth(url,dep,ospath):  
    '''''三个参数分别是 
    url ： 需要下载的网站地址 
    dep ：需要遍历的深度  
    ospath：图片下载的本地文件夹      
    '''  
    global num  
    if dep<=0:  
        return 0  
    else:  
        img=get_img(url)  
        for j in range(0,len(img)):  
            if is_img(img[j]) == 'image':  
                filepath = get_file_name(ospath,img[j],num+1);  
                if (os.path.exists(filepath)):  
                    pass  
                try:  
                    urllib.urlretrieve(img[j], filepath)  
                    print '已经下载好第%d张图片'%(num+1)  
                    num+=1  
                except:  
                    print '该图片无法下载或者图片地址无法识别！'  
                    print img[j]  
            else:  
                pass  
        urls=get_url(url)  
        if len(urls)>0:  
            for url in urls:  
                depth(url,dep-1,ospath)  
        else:  
            return 0  
        return 1  
  
  
if __name__ == '__main__':     
    num=0
    print '默认存储路径为D:\Picture\GetImageFromWeb（在代码中可自行更改）\n'
    adress=raw_input('请输入网址：')
    imglenth=eval(raw_input('请输入最小图片大小（字节数）：'))
    dept=eval(raw_input('请输入链接遍历深度：'))
    depth(adress,dept,"D:\\Picture\\GetImageFromWeb\\")     
    print '********************************搞定******************************************'