今天使用python爬虫实现了自动抓取糗事百科的段子,因为糗事百科不需要登录,抓取比较简单。程序每按一次回车输出一条段子,代码参考了 http://cuiqingcai.com/990.html 但该博主的代码似乎有些问题,我自己做了修改,运行成功,下面是代码内容:
# -*- coding:utf-8 -*-
__author__ = 'Jz'
import urllib2
import re
#糗事百科爬虫类
class QSBK:
#初始化
def __init__(self):
self.pageIndex = 1
self.user_agent = 'Mozilla/5.0 (Windows NT 6.1; WOW64)'
self.headers = {'User-Agent': self.user_agent}
#joke的每一个元素是每一页的段子
self.joke = []
#判断是否继续运行
self.enable = False
def getPage(self, pageIndex):
try:
URL = 'http://www.qiushibaike.com/hot/page/' + str(pageIndex)
request = urllib2.Request(url = URL, headers = self.headers)
response = urllib2.urlopen(request)
pageContent = response.read().decode('utf-8')
return pageContent
except urllib2.URLError, e:
if hasattr(e, 'reason'):
print '段子抓取失败,失败原因:', e.reason
return None
def getJokeList(self, pageIndex):
pageContent = self.getPage(pageIndex)
if not pageContent:
print '段子获取失败...'
return None
#第三个组中的内容用于判断段子是否附带图片
pattern = re.compile(r'<div.*?class="author">.*?<a.*?>.*?<img.*?/>\n(.*?)\n</a>.*?</div>.*?<div class="content">\n\n(.*?)\n<!--.*?-->.*?</div>' +
r'(.*?)class="stats">.*?<span.*?class="stats-vote"><i.*?class="number">(.*?)</i>'
, re.S)
jokes = re.findall(pattern, pageContent)
pageJokes = []
#过滤带有图片的段子
for joke in jokes:
hasImg = re.search('img', joke[2])
#joke[0]为发布人,joke[1]为段子内容,joke[3]为点赞数量
if not hasImg:
pageJokes.append([joke[0].strip(), joke[1].strip(), joke[3].strip()])
return pageJokes
def loadPage(self):
if self.enable == True:
#若当前已看的页数少于两页,则加载新的一页
if len(self.joke) < 2:
pageJokes = self.getJokeList(self.pageIndex)
if pageJokes:
self.joke.append(pageJokes)
self.pageIndex += 1
#每输入一次回车,打印一条段子
def getOneJoke(self, pageJokes, page):
jokes = pageJokes
for joke in jokes:
userInput = raw_input('请输入回车键或Q/q: ')
self.loadPage()
if userInput == 'Q' or userInput == 'q':
self.enable = False
print '退出爬虫...'
return
print u'段子内容:%s\n第%d页\t发布人:%s\t赞:%s' % (joke[1], page, joke[0], joke[2])
def start(self):
print '正在从糗事百科抓取段子,按回车键查看新段子,按Q/q退出...'
self.enable = True
self.loadPage()
page = 0
while self.enable:
if len(self.joke) > 0:
pageJokes = self.joke[0]
page += 1
#删除已经读取过的段子页
del self.joke[0]
self.getOneJoke(pageJokes, page)
spider = QSBK()
spider.start()
注释已经附上,其中有几点需要注意的地方:
1.需要加上header验证进行伪装,否则无法抓取网页内容
2.正则表达式的书写,需要将内容提取出来以验证是否有附带图片
3.getOneJoke函数中格式化输出段子时,需要在字符串前加上u,否则会报如下错误:
Traceback (most recent call last):
File "D:\coding_file\python_file\TestPython\src\Test\QSBK.py", line 84, in <module>
spider.start()
File "D:\coding_file\python_file\TestPython\src\Test\QSBK.py", line 81, in start
self.getOneJoke(pageJokes, page)
File "D:\coding_file\python_file\TestPython\src\Test\QSBK.py", line 68, in getOneJoke
print '段子内容:%s\n第%d页\t发布人:%s\t赞:%s' % (joke[1], page, joke[0], joke[2])
UnicodeDecodeError: 'ascii' codec can't decode byte 0xe7 in position 3: ordinal not in range(128)
这是因为Python默认编码方式为Unicode,所以joke[0]等也是Unicode编码,为了格式化输出,前面的字符串也需要转换成Unicode编码