#coding=utf-8

 import requests as req

 import re

 import urllib

 from bs4 import BeautifulSoup

 import sys

 import codecs

 import time

 r=req.get('https://www.liaoxuefeng.com/wiki/0014316089557264a6b348958f449949df42a6d3a2e542c000',

          headers={'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/62.0.3202.94 Safari/537.36'})

 content=r.text

 #print(content)

 soup=BeautifulSoup(content,'html.parser')

 #下面2行内容解决UnicodeEncodeError: 'ascii' codec can't encode characters in position 63-64问题,但是加了后print就打印不出来了，需要查原因

 reload(sys)

 sys.setdefaultencoding('utf-8')

 i=0

 for tag in soup.find_all(re.compile(r'^a{1}'),{'class':'x-wiki-index-item'}):

     i=i+1

     if i%3==0:

         time.sleep(30)

     name=tag.get_text()

     href='https://www.liaoxuefeng.com'+tag['href']

     req2=req.get(href,headers={'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/62.0.3202.94 Safari/537.36'})

     time.sleep(3)

     soup2=BeautifulSoup(req2.text,'html.parser')

     count=soup2.find_all('div',{'class':'x-wiki-info'})

     try:

         co=count[0].find('span').get_text()

         co=co[7:]

     except IndexError as e:

         co=''

     with open('E:/sg_articles.xlsx', 'a+') as f:

         f.write(codecs.BOM_UTF8)#解决写入csv后乱码问题

         f.write(name+','+href+','+co+'\n')

 '''

 睡眠是因为网页访问过多就会报503 Service Unavailable for Bot网站超过了iis限制造成的由于2003的操作系统在提示IIS过多时并非像2000系统提示“链接人数过多”

 http://www.51testing.com/html/53/15150753-4455790.html --数据可视化

 http://www.cnblogs.com/xxoome/p/5880693.html --python引入模块时import与from ... import的区别

 https://www.cnblogs.com/amou/p/9184614.html --讲解几种爬取网页的匹配方式

 https://www.cnblogs.com/yinheyi/p/6043571.html --python基本语法

 '''

 #coding=utf-8

 from pyecharts import Bar

 import pandas as pd

 p=pd.read_excel('E:\sg_articles.xls',names=["title","href","count"])

 a=p.sort_values(by='count',ascending=False)[0:3]

 title=a['title']

 count=a['count']

 bar=Bar("点击量TOP3", title_pos='center', title_top='', width=800, height=400)

 bar.add("", title, count, is_convert=True, xaxis_min=10, yaxis_rotate=30, yaxis_label_textsize=10, is_yaxis_boundarygap=True, yaxis_interval=0,

         is_label_show=True, is_legend_show=False, label_pos='right',is_yaxis_inverse=True, is_splitline_show=False)

 bar.render("E:\点击量TOP3.html")