前言

最近在自學(xué)Python，正好2014 CSDN博客之星投票搞得如火如荼，拿來(lái)練練手。

環(huán)境：Win7 64位 Python 2.7；

用到了正則表達(dá)式、函數(shù)、寫文件、urllib2；

沒(méi)有用到線程；

程序也不怎么規(guī)范，但終歸是能夠達(dá)到目的了，哈哈。

源碼

# -*- coding: utf-8 -*-

import urllib2;

import re;

import os;

import thread;

def loadBlogSort(url):

pageCount = getPageCount(url);

print 'pageCount == ',pageCount;

baseUrl = 'http://vote.blog.csdn.net/Blogstar2014/Selection?PageIndex=';

urlSuffix = '#content';

filepath = 'csdn_blog_star_vote.txt';

if os.path.exists(filepath):

os.remove(filepath);

f = open(filepath,'w+');

for pageIndex in range(1,int(pageCount)+1):

contentUrl = baseUrl + str(pageIndex) + urlSuffix;

print 'pageIndex == ',pageIndex, ' contentUrl == ',contentUrl;

user_agent = 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36'

headers = { 'User-Agent' : user_agent }

request = urllib2.Request(contentUrl, headers = headers)

response = urllib2.urlopen(request);

result = response.read();

# unicodeResult = result.decode("utf-8");

# 名稱

#

(.+?)

names = re.findall('

(.+?)',result,re.S);

nameList = [];

for name in names:

# print '昵稱：',name[2];

nameList.append(name[ 2 ]);

# 博客地址

#

blogUrlList = [];

detailUrls = re.findall('

',result,re.S);

for detailUrl in detailUrls:

blogUrlList.append(getBlogUrl(detailUrl[0]));

# 得票

#

得票：(.+?)

votes = re.findall('

(.+?)(.+?)

',result,re.S);

voteList = [];

for vote in votes:

# print ' 得票：',str(vote[2]);

voteList.append(vote[ 2 ]);

# 博文瀏覽量、博文數(shù)、評(píng)論數(shù)

#

(.+?)(.+?)(.+?)

infos = re.findall('

(.+?)(.+?)(.+?)',result,re.S);

infoIndex = 0;

blankSize = 20;

for info in infos:

user = '昵稱：'+nameList[infoIndex] + ( blankSize - len(nameList[infoIndex]) )*' '+'得票：'+voteList[infoIndex] + ( blankSize - len(voteList[infoIndex]) )*' '+'博文瀏覽量: '+str(info[0]) + ( blankSize - len(str(info[0])) )*' '+'博文數(shù)：'+str(info[1]) + ( blankSize - len(str(info[1])) )*' '+'評(píng)論數(shù)：'+str(info[2])+ + ( blankSize - len(str(info[2])) )*' '+'博客地址：' + blogUrlList[infoIndex]+ '\n'

# print user;

f.write(user);

infoIndex += 1;

f.close();

print '寫文件完畢！';

# 得到博客鏈接

def getBlogUrl(detailUrl):

url = 'http://vote.blog.csdn.net/' + detailUrl;

user_agent = 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36'

headers = { 'User-Agent' : user_agent }

request = urllib2.Request(url, headers = headers)

response = urllib2.urlopen(request);

result = response.read();

blogUrls = re.findall('

(.+?)(.+?)

',result,re.S);

print 'blogUrl == ',url + '\n' + str(blogUrls[0][1]);

return str(blogUrls[0][1]);

# 得到總頁(yè)碼數(shù)

def getPageCount(url):

user_agent = 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36'

headers = { 'User-Agent' : user_agent }

request = urllib2.Request(url, headers = headers)

response = urllib2.urlopen(request);

result = response.read();

pageCount = re.findall('

(.+?)',result,re.S);

return pageCount[0];

url = 'http://vote.blog.csdn.net/Blogstar2014/Selection?PageIndex=1#content';

loadBlogSort(url);

效果

2014_csdn_blog_star_vote

總結(jié)

以上是生活随笔為你收集整理的python 投票脚本_使用Python脚本拉取2014 CSDN博客之星投票情况的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺(jué)得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。