黄聪：Python网站采集功能（多线程的采集、WDPYSPIDER类、pycurl）

Python

1
import urllib

2
urlItem = urllib.urlopen("http://www.baidu.com")

3
htmSource = urlItem.read()

4
urlItem.close()

5
print htmSource

pycurl

http://pycurl.sourceforge.net/download/

http://pycurl.sourceforge.net/doc/curlobject.html

Python

01
import pycurl

02
c = pycurl.Curl()

03
c.setopt(pycurl.URL, "http://www.whiledo.com/")

04
c.setopt(pycurl.HTTPHEADER, ["Accept:"])

05
import StringIO

06
b = StringIO.StringIO()

07
c.setopt(pycurl.WRITEFUNCTION, b.write)

08
c.setopt(pycurl.FOLLOWLOCATION, 1)

09
c.setopt(pycurl.MAXREDIRS, 5)

10
c.perform()

11
print b.getvalue()

12
print c.getinfo(pycurl.INFO_FILETIME)

curl_easy_setopt


告诉 libcurl 的如何做事
CURLOPT_WRITEFUNCTION: 写(下载)回传函数,传递一个写指针供外部操作, 一次回调内容大小在 CURL_MAX_WRITE_SIZE (curl.h头文件)中设置
CURLOPT_WRITEDATA: 直接写文件,指定一个文件名如c.setopt(pycurl.WRITEDATA, 'E:\WebSite\py\1.txt') 注win下不能用
CURLOPT_READFUNCTION: 读(上传)回传函数
CURLOPT_SEEKFUNCTION: 数据指针移动,int function(void *instream, curl_off_t offset, int origin);SEEK_SET, SEEK_CUR and SEEK_END,返回CURL_SEEKFUNC_OK或CURL_SEEKFUNC_FAIL或CURL_SEEKFUNC_CANTSEEK (0,1,2)
CURLOPT_OPENSOCKETFUNCTION:
CURLOPT_HEADERFUNCTION:只接收头数据 size_t function( void *ptr, size_t size, size_t nmemb, void *userdata);
CURLOPT_DEBUGFUNCTION: int curl_debug_callback (CURL *, curl_infotype, char *, size_t, void *);
CURLOPT_VERBOSE: 参数设置为1 能显示更多详细信息
CURLOPT_HEADER: 设为 1 将在返回的文本中包含头信息
CURLOPT_NOSIGNAL: 不超时
CURLOPT_FOLLOWLOCATION:设置为1告诉libcurl遵循任何访问
CURLOPT_MAXREDIRS: 设定重定向的数目限制,设置为-1表示无限的重定向（默认）
CURLOPT_PUT:数据上载相关
CURLOPT_POST:
CURLOPT_POSTREDIR:
CURLOPT_POSTFIELDS:
CURLOPT_POSTFIELDSIZE:
CURLOPT_POSTFIELDSIZE_LARGE:
CURLOPT_COPYPOSTFIELDS:
CURLOPT_HTTPPOST:
CURLOPT_UPLOAD:
CURLOPT_AUTOREFERER:libcurl自动设置Referer
CURLOPT_REFERER: 伪造来源路径
CURLOPT_USERAGENT:自定义USERAGENT
CURLOPT_HTTPHEADER:自定义头
CURLOPT_COOKIE: "name1=content1; name2=content2;"
CURLOPT_COOKIEFILE:
CURLOPT_COOKIEJAR:
CURLOPT_COOKIESESSION: 默认情况下，libcurl始终加载和存储所有Cookie
CURLOPT_COOKIELIST
CURLOPT_HTTPGET
CURLOPT_HTTP_VERSION: CURL_HTTP_VERSION_NONE,CURL_HTTP_VERSION_1_0,CURL_HTTP_VERSION_1_1
CURLOPT_IGNORE_CONTENT_LENGTH:忽略内容长度头,针对类似Apache 1.x的服务器
CURLOPT_HTTP_TRANSFER_DECODING:告诉libcurl如何对传输解码,(0,=1)
CURLOPT_HTTP200ALIASES:自定义HTTP 200响应别名,有些服务器对200返回不是标准的
CURLOPT_ENCODING:设置接收的内容编码,同 Accept-Encoding, ('','gzip',....)
CURLOPT_UNRESTRICTED_AUTH:数设置为1,继续发送认证（用户+密码）
NETWORK OPTIONS
CURLOPT_URL: http://xxxx,ftp://xxxx
CURLOPT_PROXY:HTTP代理,主机名或IP地址
CURLOPT_PROXYPORT:代理端口,也可在PROXY的地址后加":端口",如 :8080
CURLOPT_PROXYTYPE:代理类型,CURLPROXY_HTTP(默认), CURLPROXY_HTTP_1_0,CURLPROXY_SOCKS4,CURLPROXY_SOCKS5,CURLPROXY_SOCKS4A,CURLPROXY_SOCKS5_HOSTNAME,
CURLOPT_NOPROXY:不使用代理的域
CURLOPT_HTTPPROXYTUNNEL:
CURLOPT_BUFFERSIZE: libcurl的缓冲区大小（以字节为单位）
（认证）
CURLOPT_NETRC: 此参数控制你的密码,CURL_NETRC_OPTIONAL使用 ~/.netrc 文件, CURL_NETRC_IGNORED(默认):忽略文件,CURL_NETRC_REQUIRED:告诉该文件的使用所需的库，要忽略的URL信息
CURLOPT_NETRC_FILE: 指定 ~/.netrc 文件
CURLOPT_USERNAME:
CURLOPT_USERPWD:
CURLOPT_PASSWORD:
CURLOPT_PROXYUSERNAME:
CURLOPT_PROXYUSERPWD:
CURLOPT_HTTPAUTH:
CURLOPT_PROXYAUTH:

CURLAUTH_BASIC: HTTP基本验证
CURLAUTH_DIGEST: HTTP摘要身份验证
CURLAUTH_DIGEST_IE:
CURLAUTH_GSSNEGOTIATE: Kerberos5认证要建立GSS - API
CURLAUTH_NTLM: NTLM身份验证
CURLAUTH_ANY: 设置所有选项,ibcurl自动选择一个它认为合适的,安全的验证<
CURLAUTH_ANYSAFE: 设置基本选项....
CURLAUTH_ONLY: 强制所有请求使用验证

getinfo


CURLINFO_RESPONSE_CODE: 获得最后收到的HTTP或FTP的代码,如200,404,403,505 代理的CONNECT响应要参考 CURLINFO_HTTP_CONNECTCODE
CURLINFO_EFFECTIVE_URL: 最后一次使用有效的URL
CURLINFO_HTTP_CONNECTCODE : 长期接受最后收到的代理响应代码
CURLINFO_FILETIME:
CURLINFO_TOTAL_TIME:
CURLINFO_CONNECT_TIME:
CURLINFO_NUM_CONNECTS: 多少个连接
CURLINFO_CONTENT_TYPE: 例：text/html
CURLINFO_REQUEST_SIZE:
CURLINFO_HEADER_SIZE:
CURLINFO_SIZE_DOWNLOAD: 下载总字节量
CURLINFO_SIZE_UPLOAD:
CURLINFO_HTTPAUTH_AVAIL: 接收掩码表明身份验证
CURLINFO_PROXYAUTH_AVAIL: 接收掩码表明代理身份验证
CURLINFO_COOKIELIST:
部份使用 INFO_ 如：INFO_COOKIELIST

一个粗糙的共用对象的采集示例

Python

01
import pycurl

02
import StringIO

03
import string

04
import random

05
class spider:

06
    def __init__(self,addHeader = []):

07
        self.httpheader = [

08
             'Accept:application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5'

09
            #,'USER_AGENT:Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64; Trident/4.0)'

10
        ] + addHeader

11
        self.curl = pycurl.Curl()

12
        self.curl.setopt(pycurl.HTTPHEADER, self.httpheader)

13
        self.curl.setopt(pycurl.REFERER, 'http://www.google.com/search?sourceid=chrome&ie=UTF-8&q='+self.rand_str())

14
        #self.curl.setopt(pycurl.AUTOREFERER, 1)

15
        self.curl.setopt(pycurl.FOLLOWLOCATION, 1)

16
        self.curl.setopt(pycurl.MAXREDIRS, 5)

17
 
18
    def __del__(self):

19
        pass

20
 
21
    def rand_str(self):

22
        return ''.join(random.sample(['a','b','c','d','e','f','g','h','i','j','k','l','m','n'], 6))

23
 
24
    def tofile(self,url,filename):

25
        fp = open(filename, 'w');

26
        self.curl.setopt(pycurl.URL, url)

27
        self.curl.setopt(pycurl.WRITEFUNCTION, fp.write)

28
        self.curl.perform()

29
        fp.close()

30
        return True

31
 
32
    def html(self, url):

33
        sio = StringIO.StringIO()

34
        self.curl.setopt(pycurl.URL, url)

35
        self.curl.setopt(pycurl.WRITEFUNCTION, sio.write)

36
        self.curl.perform()

37
        reval = sio.getvalue()

38
        sio.close()

39
        return reval

40
 
41
if __name__ == "__main__":

42
    get = spider(['USER_AGENT:Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64; Trident/4.0)'])

43
    print get.html("http://localhost/spider_for_test.php")

44
    print get.tofile("http://localhost/spider_for_test.php",r'E:\WebSite\wwwroot\test.txt')

一个多线程的采集示例

Python

01
import pycurl

02
import threading

03
import StringIO

04
import string

05
import random

06
class spider:

07
    def __init__(self,referer='',httpheader = []):

08
        self.httpheader = [

09
             'Accept:application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5'

10
            ,'USER_AGENT:Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64; Trident/4.0)'

11
        ] + httpheader

12
        self.referer = referer

13
    def __del__(self):

14
        pass

15
 
16
    def fetch(self,url,stream):

17
        curl = pycurl.Curl()

18
        curl.setopt(pycurl.HTTPHEADER, self.httpheader)

19
        if self.referer == '':

20
            curl.setopt(pycurl.AUTOREFERER, 1)

21
        else:

22
            curl.setopt(pycurl.REFERER, self.referer)

23
        curl.setopt(pycurl.FOLLOWLOCATION, 1)

24
        curl.setopt(pycurl.MAXREDIRS, 5)

25
        curl.setopt(pycurl.URL, url)

26
        curl.setopt(pycurl.WRITEFUNCTION, stream.write)

27
        curl.perform()

28
        curl.close()

29
 
30
    def rand_str(self):

31
        return ''.join(random.sample(['a','b','c','d','e','f','g','h','i','j','k','l','m','n'], 6))

32
 
33
    def tofile(self,url,filename):

34
        fp = open(filename, 'w');

35
        self.fetch(url,fp)

36
        fp.close()

37
        return True

38
 
39
    def html(self, url):

40
        sio = StringIO.StringIO()

41
        self.fetch(url,sio)

42
        reval = sio.getvalue()

43
        sio.close()

44
        return reval

45
 
46
def gethtml(url,get):

47
    print get.html(url)

48
 
49
if __name__ == "__main__":

50
    import time,datetime

51
    dstart = datetime.datetime.now()

52
    get = spider()

53
    get.referer = 'http://www.google.com/search?sourceid=chrome&ie=UTF-8&q='+get.rand_str()

54
    thread_pool = []

55
    acc = Account(100)

56
    for i in range(10):

57
        url = "http://localhost/test.php?n="+str(i)

58
        th = threading.Thread(target=gethtml,args=(url,get))

59
        thread_pool.append(th)

60
    for i in range(10):

61
        thread_pool[i].start()

62
    for i in range(10):

63
        threading.Thread.join(thread_pool[i])

64
    dend = datetime.datetime.now()

65
    print "Time span:" , dend-dstart;

WDPYSPIDER类（支持，多线程，代理，登陆验证,POST）

Python

001
#coding:utf-8

002
import pycurl

003
import urllib

004
import threading

005
import StringIO

006
import string

007
import random

008
class spider:

009
    '''WDPYSPIDER(Whiledo Python Spider Class) 采集类

010
 
011
    @author HzqGhost admin@whiledo.com QQ:313143468

012
    get = spider()

013
    get.referer = 'http://www.google.com/search?sourceid=chrome&ie=UTF-8&q='+get.rand_str()

014
    get.proxyuse = True

015
    get.proxyip = ['059148233056.ctinets.com:80']

016
    url = "http://www.whiledo.com"

017
    print get.html(url=url)'''

018
    def __init__(self):

019
        self.httpheader = [

020
             'Accept:application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5'

021
            ,'USER_AGENT:Mozilla/4.0 (compatible; MSIE 8.0; Windows NT 6.1; WOW64; Trident/4.0)'

022
        ] #http头信息

023
        self.referer = '' #伪造来源路径

024
        self.connnecttimeout = 60 #获取联接超时(秒)

025
        self.timeout = 300 #读定超时(秒)

026
        self.backheader = 0 #是否返回服务器http头信息(一般用于测试)

027
        self.cookesfile = "./cookes.dat" #cookesfile 自动读写处理文件

028
        self.proxyuse = False  #是否使用代理服务器

029
        self.proxyip = []  #代理服务器[IP:PORT]列表，随机使用列表中的IP

030
        self.proxynodomain = ['localhost','127.0.0.1']  #不使用代理服务器的域

031
        self.http200alias = [] #200返回信息别名列表

032
        self.error = 'WDPYERROR' #非200状态时返回的错误标识

033
    def __del__(self):

034
        pass

035
 
036
    def fetch(self,url,stream, post={}):

037
        '''

038
        --url

039
        --stream [stream] StringIO or fp

040
        --post [dict] {'username':'hzq','password':'blog'}'''

041
        curl = pycurl.Curl()

042
        curl.setopt(pycurl.CONNECTTIMEOUT, self.connnecttimeout)

043
        curl.setopt(pycurl.TIMEOUT, self.timeout)

044
        curl.setopt(pycurl.HTTPHEADER, self.httpheader)

045
        curl.setopt(pycurl.HTTP200ALIASES, self.http200alias)

046
        curl.setopt(pycurl.HEADER, self.backheader)

047
        curl.setopt(pycurl.FOLLOWLOCATION, 1)

048
        curl.setopt(pycurl.MAXREDIRS, 5)

049
        if self.referer == '':

050
            curl.setopt(pycurl.AUTOREFERER, 1)

051
        else:

052
            curl.setopt(pycurl.REFERER, self.referer)

053
        curl.setopt(pycurl.COOKIEJAR, self.cookesfile)

054
        curl.setopt(pycurl.COOKIEFILE, self.cookesfile)

055
        curl.setopt(pycurl.WRITEFUNCTION, stream.write)

056
        curl.setopt(pycurl.URL, url)

057
        if self.proxyuse:

058
            proxyip = self.proxyip[random.randint(0,len(self.proxyip)-1)];

059
            curl.setopt(pycurl.PROXY, proxyip)

060
            #curl.setopt(pycurl.PROXYNO, self.proxynodomain) #需要7.19.4 的pycurl版本

061
        if len(post)>0 :

062
            curl.setopt(pycurl.POSTFIELDS, post)

063
        status = ''

064
        try:

065
            curl.perform()

066
            status = curl.getinfo(pycurl.RESPONSE_CODE)

067
        except:

068
            status = curl.errstr()

069
        finally:

070
            curl.close()

071
            status = str(status);

072
            if status != '200':

073
                status = self.error

074
            return status;

075
 
076
    def rand_str(self):

077
        return ''.join(random.sample(['a','b','c','d','e','f','g','h','i','j','k','l','m','n'], 6))

078
 
079
    def tofile(self,url,filename, post={}):

080
        fp = open(filename, 'wb');

081
        self.fetch(url,fp,post)

082
        fp.close()

083
        return True

084
 
085
    def html(self, url, post={}):

086
        sio = StringIO.StringIO()

087
        reval = self.fetch(url,sio, post)

088
        if reval == '200':

089
            reval = sio.getvalue()

090
        sio.close()

091
        return reval

092
 
093
def gethtml(url,get):

094
    print get.html(url)

095
 
096
if __name__ == "__main__":

097
    get = spider()

098
    get.referer = 'http://www.google.com/search?sourceid=chrome&ie=UTF-8&q='+get.rand_str()

099
    get.proxyuse = True

100
    get.proxyip = ['059148233056.ctinets.com:80']

101
    url = "http://www.whiledo.com"

102
    print get.html(url=url)