百度图片爬虫-python版

时间:2023-03-08 15:45:55
百度图片爬虫-python版
  1 #coding:utf-8
  2 
  3 """
  4 
  5 Created on 2015-9-17
  6 
  7  
  8 
  9 @author: huangxie
 10 
 11 """
 12 
 13 import time,math,os,re,urllib,urllib2,cookielib 
 14 
 15 from bs4 import BeautifulSoup
 16 
 17 import time  
 18 
 19 import re
 20 
 21 import uuid
 22 
 23 import json
 24 
 25 from threading import Thread
 26 
 27 from Queue import Queue 
 28 
 29 import MySQLdb as mdb
 30 
 31 import sys
 32 
 33 import threading
 34 
 35 import utils
 36 
 37 import imitate_browser
 38 
 39 from MySQLdb.constants.REFRESH import STATUS
 40 
 41 reload(sys)
 42 
 43 sys.setdefaultencoding('utf-8')
 44 
 45  
 46 
 47 DB_HOST = '127.0.0.1'
 48 
 49 DB_USER = 'root'
 50 
 51 DB_PASS = 'root'
 52 
 53 proxy = {u'http':u'222.39.64.13:8118'}
 54 
 55 TOP_URL="http://image.baidu.com/i?tn=resultjsonavatarnew&ie=utf-8&word={word}&pn={pn}&rn={rn}"
 56 
 57 KEYWORD_URL="https://www.baidu.com/s?ie=utf-8&f=8&tn=baidu&wd={wd}"
 58 
 59  
 60 
 61 """
 62 
 63 i_headers = {'User-Agent':'Mozilla/5.0 (Windows NT 6.1) AppleWebKit/537.11 (KHTML, like Gecko) Chrome/23.0.1271.64 Safari/537.11',
 64 
 65               'Accept':'json;q=0.9,*/*;q=0.8',
 66 
 67               'Accept-Charset':'utf-8;q=0.7,*;q=0.3',
 68 
 69               'Accept-Encoding':'gzip',
 70 
 71               'Connection':'close',
 72 
 73               'Referer':None #注意如果依然不能抓取的话,这里可以设置抓取网站的host
 74 
 75             }
 76 
 77 """
 78 
 79 i_headers = {'User-Agent':'Mozilla/5.0 (Windows NT 6.1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/31.0.1650.48'}
 80 
 81  
 82 
 83 def GetDateString():
 84 
 85     x = time.localtime(time.time())
 86 
 87     foldername = str(x.__getattribute__("tm_year"))+"-"+str(x.__getattribute__("tm_mon"))+"-"+str(x.__getattribute__("tm_mday"))
 88 
 89     return foldername 
 90 
 91  
 92 
 93 class BaiduImage(threading.Thread):     
 94 
 95  
 96 
 97     def __init__(self):
 98 
 99         Thread.__init__(self)
 
         self.browser=imitate_browser.BrowserBase()
 
         self.chance=0
 
         self.chance1=0
 
         self.request_queue=Queue()
 
         self.wait_ana_queue=Queue()
 
         #self.key_word_queue.put((("动态图", 0, 24)))
 
         self.count=0
 
         self.mutex = threading.RLock() #可重入锁,使单线程可以再次获得已经获得的锁
 
         self.commit_count=0
 
         self.ID=500
 
         self.next_proxy_set = set()
 
         self.dbconn = mdb.connect(DB_HOST, DB_USER, DB_PASS, 'sosogif', charset='utf8')
 
         self.dbconn.autocommit(False)
 
         self.dbcurr = self.dbconn.cursor()
 
         self.dbcurr.execute('SET NAMES utf8')
 
         
 
     """
 
     def run(self):
 
         while True:
 
             self.get_pic()
 
     """
 
     
 
     def work(self,item):
 
         print "start thread",item
 
         while True: #MAX_REQUEST条以上则等待
 
             self.get_pic()
 
             self.prepare_request()
 
     
 
     def format_keyword_url(self,keyword):
 
   
 
         return KEYWORD_URL.format(wd=keyword).encode('utf-8') 
 
            
 
     def generateSeed(self,url):
 
         
 
         html = self.browser.openurl(url).read()
 
         if html:
 
             try:
 
                 soup = BeautifulSoup(html)
 
                 trs = soup.find('div', id='rs').find('table').find_all('tr') #获得所有行
 
                 for tr in trs:
 
                     ths=tr.find_all('th')
 
                     for th in ths:
 
                         a=th.find_all('a')[0]
 
                         keyword=a.text.strip()
 
                         if "动态图" in keyword or "gif" in keyword:
 
                             print "keyword",keyword
 
                             self.dbcurr.execute('select id from info where word=%s',(keyword))
 
                             y = self.dbcurr.fetchone()
 
                             if not y:
 
                                 self.dbcurr.execute('INSERT INTO info(word,status,page_num,left_num,how_many) VALUES(%s,0,0,0,0)',(keyword))
 
                     self.dbconn.commit()
 
             except:
 
                 pass
 
                 
 
                
 
     def prepare_request(self):
 
         self.lock()
 
         self.dbcurr.execute('select * from info where status=0')
 
         result = self.dbcurr.fetchone()
 
         if result:
 
             id,word,status,page_num,left_num,how_many=result
 
             self.request_queue.put((id,word,page_num)) 
 
             if page_num==0 and left_num==0 and how_many==0:
 
                 url=self.format_keyword_url(word)
 
                 self.generateSeed(url)
 
                 html=""
 
                 try:
 
                     url=self.format_top_url(word, page_num, 24)
 
                     html = self.browser.openurl(url).read()
 
                 except Exception as err:
 
                     print "err",err
 
                     #pass
 
                 if html!="":
 
                     how_many=self.how_many(html)
 
                     print "how_many",how_many
 
                     if how_many==None:
 
                         how_many=0
 
                     t=math.ceil(how_many/24*100) #只要前1/100即可
 
                     num = int(t)
 
                     for i  in xrange(0,num-1):
 
                         self.dbcurr.execute('INSERT INTO info(word,status,page_num,left_num,how_many) VALUES(%s,%s,%s,%s,%s)',(word,0,i*24,num-i,how_many))
 
                     self.dbcurr.execute('update info SET status=1 WHERE id=%s',(id)) #置为已经访问
 
                     self.dbconn.commit()
 
         self.unlock()
 
                 
 
             
 
     def start_work(self,req_max):
 
         for item in xrange(req_max):
 
             t = threading.Thread(target=self.work, args=(item,))
 
             t.setDaemon(True)
 
             t.start()
 
             
 
     def lock(self): #加锁
 
         self.mutex.acquire()
 
  
 
     def unlock(self): #解锁
 
         self.mutex.release()
 
  
 
     def get_para(self,url,key):
 
         values = url.split('?')[-1]
 
         for key_value in values.split('&'):
 
             value=key_value.split('=')
 
             if value[0]==key:
 
                 return value[1]
 
         return None  
 
     
 
     def makeDateFolder( self,par,child):
 
         #self.lock()
 
         if os.path.isdir( par ):
 
             path=par + '//' + GetDateString()
 
             newFolderName = path+'//'+child
 
             if not os.path.isdir(path):
 
                 os.mkdir(path)
 
             if not os.path.isdir( newFolderName ):
 
                 os.mkdir( newFolderName )
 
             return newFolderName
 
         else:
 
             return par 
 
         #self.unlock()
 
         
 
     def parse_json(self,data):
 
         
 
         ipdata = json.loads(data)
 
         try:
 
             if ipdata['imgs']:  
 
                 for n in ipdata['imgs']: #data子项 
 
                     if n['objURL']:  
 
                         try:
 
                             proxy_support = urllib2.ProxyHandler(proxy)
 
                             opener = urllib2.build_opener(proxy_support)
 
                             urllib2.install_opener(opener)
 
                             #print "proxy",proxy
 
                             self.lock()
 
                             self.dbcurr.execute('select ID from pic_info where objURL=%s', (n['objURL']))
 
                             y = self.dbcurr.fetchone()
 
                             #print "y=",y
 
                             if y:
 
                                 print "database exist"
 
                                 self.unlock() #continue 前解锁
 
                                 continue
 
                             else:
 
                                 real_extension=utils.get_extension(n['objURL'])
 
                                 req = urllib2.Request(n['objURL'],headers=i_headers)
 
                                 resp = urllib2.urlopen(req,None,5)
 
                                 dataimg=resp.read()
 
                                 name=str(uuid.uuid1())
 
                                 filename=""
 
                                 if len(real_extension)>4:
 
                                     real_extension=".gif"
 
                                 real_extension=real_extension.lower()
 
                                 if real_extension==".gif":
 
                                     filename  =self.makeDateFolder("E://sosogif", "d"+str(self.count % 60))+"//"+name+"-www.sosogif.com-搜搜gif贡献"+real_extension
 
                                     self.count+=1
 
                                 else:
 
                                     filename  =self.makeDateFolder("E://sosogif", "o"+str(self.count % 20))+"//"+name+"-www.sosogif.com-搜搜gif贡献"+real_extension
 
                                     self.count+=1
 
                                 """
 
                                 name=str(uuid.uuid1())
 
                                 filename=""
 
                                 if len(real_extension)>4:
 
                                     real_extension=".gif"
 
                                 filename  =self.makeDateFolder("E://sosogif", "d"+str(self.count % 60))+"//"+name+"-www.sosogif.com-搜搜gif贡献"+real_extension
 
                                 self.count+=1 
 
                                 """
 
                                 try: 
 
                                     if not os.path.exists(filename): 
 
                                         file_object = open(filename,'w+b')  
 
                                         file_object.write(dataimg)  
 
                                         file_object.close()
 
                                         self.anaylis_info(n,filename,real_extension) #入库操作
 
                                     else:
 
                                         print "file exist" 
 
                                 except IOError,e1:  
 
                                     print "e1=",e1
 
                                     pass
 
                             self.unlock()
 
                         except IOError,e2:  
 
                             #print "e2=",e2 
 
                             pass  
 
                             self.chance1+=1
 
         except Exception as parse_error:
 
             print "parse_error",parse_error
 
             pass
 
     
 
     def title_dealwith(self,title):
 
         
 
         #print "title",title
 
         a=title.find("<strong>")
 
         temp1=title[0:a]
 
         b=title.find("</strong>")
 
         temp2=title[a+8:b]
 
         temp3=title[b+9:len(title)]
 
         return (temp1+temp2+temp3).strip()
 
         
 
     def anaylis_info(self,n,filename,real_extension):
 
         print "success."
 
         
 
         #if self.wait_ana_queue.qsize()!=0:
 
             #n,filename,real_extension=self.wait.ana_queue.get()
 
         #self.lock()
 
         objURL=n['objURL'] #图片地址
 
         fromURLHost=n['fromURLHost'] #来源网站
 
         width=n['width']  #宽度
 
         height=n['height'] #高度
 
         di=n['di'] #用来唯一标识
 
         type=n['type'] #格式
 
         fromPageTitle=n['fromPageTitle'] #来自网站
 
         keyword=self.title_dealwith(fromPageTitle)
 
         cs=n['cs'] #未知
 
         os=n['os'] #未知
 
         temp = time.time()
 
         x = time.localtime(float(temp))
 
         acTime = time.strftime("%Y-%m-%d %H:%M:%S",x) #爬取时间
 
         self.dbcurr.execute('select ID from pic_info where cs=%s', (cs))
 
         y = self.dbcurr.fetchone()
 
         if not y:
 
             print 'add pic',filename
 
             self.commit_count+=1
 
             self.dbcurr.execute('INSERT INTO pic_info(objURL,fromURLHost,width,height,di,type,keyword,cs,os,acTime,filename,real_extension) VALUES(%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s)',(objURL,fromURLHost,width,height,di,type,keyword,cs,os,acTime,filename,real_extension))
 
             if self.commit_count==10:
 
                 self.dbconn.commit()
 
                 self.commit_count=0
 
         #self.unlock()
 
            
 
  
 
     def format_top_url(self,word,pn,rn):
 
  
 
         url = TOP_URL.format(word=word, pn=pn,rn=rn).encode('utf-8') 
 
         return url
 
  
 
     def how_many(self,data):
 
         try:
 
             ipdata = json.loads(data)
 
             if ipdata['displayNum']>0:
 
                 how_many=ipdata['displayNum']
 
                 return int(how_many)
 
             else:
 
                 return 0
 
         except Exception as e:
 
             pass
 
         
 
     def get_pic(self):
 
         """
 
         word="gif"
 
         pn=0
 
         rn=24
 
         if self.key_word_queue.qsize()!=0:
 
             word,pn,rn=self.key_word_queue.get()
 
         url=self.format_top_url(word,pn,rn)
 
         global proxy
 
         if url:
 
             try:
 
                 html=""
 
                 try:
 
                     req = urllib2.Request(url,headers=i_headers)
 
                     response = urllib2.urlopen(req, None,5)
 
                     #print "url",url
 
                     html = self.browser.openurl(url).read()
 
                 except Exception as err:
 
                     print "err",err
 
                     #pass
 
                 if html:
 
                     how_many=self.how_many(html)
 
                     #how_many=10000
 
                     print "how_many",how_many
 
                     word=self.get_para(url,"word")
 
                     rn=int(self.get_para(url,"rn"))
 
                     t=math.ceil(how_many/rn)
 
                     num = int(t)
 
                     for item  in xrange(0,num-1):
 
         """
 
         try:
 
             global proxy
 
             print "size of queue",self.request_queue.qsize()
 
             if self.request_queue.qsize()!=0:
 
                 id,word,page_num = self.request_queue.get()            
 
                 u=self.format_top_url(word,page_num,24)
 
                 self.lock()
 
                 self.dbcurr.execute('update info SET status=1 WHERE id=%s',(id))
 
                 self.dbconn.commit()
 
                 if self.chance >0 or self.chance1>1: #任何一个出问题都给换代理
 
                     if self.ID % 100==0:
 
                         self.dbcurr.execute("select count(*) from proxy")
 
                         for r in self.dbcurr:
 
                             count=r[0]
 
                         if self.ID>count:
 
                             self.ID=50
 
                     self.dbcurr.execute("select * from proxy where ID=%s",(self.ID))
 
                     results = self.dbcurr.fetchall()
 
                     for r in results:
 
                         protocol=r[1]
 
                         ip=r[2]
 
                         port=r[3]
 
                         pro=(protocol,ip+":"+port)
 
                         if pro not in self.next_proxy_set:
 
                             self.next_proxy_set.add(pro)
 
                     self.chance=0
 
                     self.chance1=0
 
                     self.ID+=1
 
                 self.unlock() 
 
                 proxy_support = urllib2.ProxyHandler(proxy)
 
                 opener = urllib2.build_opener(proxy_support)
 
                 urllib2.install_opener(opener)
 
                 html=""
 
                 try:
 
                     req = urllib2.Request(u,headers=i_headers)
 
                     #print "u=",u
 
                     response = urllib2.urlopen(req, None,5)
 
                     html = response.read()
 
                     if html:
 
                         #print "html",type(html)
 
                         self.parse_json(html)
 
                 except Exception as ex1:
 
                     #print "error=",ex1
 
                     pass
 
                     self.chance+=1
 
                     if self.chance>0 or self.chance1>1:
 
                         if len(self.next_proxy_set)>0:
 
                             protocol,socket=self.next_proxy_set.pop()
 
                             proxy= {protocol:socket}
 
                             print "change proxy finished<<",proxy,self.ID
 
         except Exception as e:
 
             print "error1",e
 
             pass
 
             
 
 if __name__ == '__main__':
 
  
 
     app = BaiduImage() 
 
     app.start_work(80)
 
     #app.generateSeed()
 
     while 1:
 
         pass