Python实现的爬取网易动态评论操作示例

#coding=utf-8
__author__ = 'kongmengfan123'
import urllib
import re
import json
import time
def gethothtml(url):#最热评论
  page=urllib.urlopen(url)
  html=page.read()
  get_json(html)
def gethnewtml():#最新评论有5页
  for i in range(1,6):
    url = 'http://comment.news.163.com/api/v1/products/a2869674571f77b5a0867c3d71db5856/threads/C4QFIJNS0001875O/comments/newList?offset=%d&limit=30&showLevelThreshold=72&headLimit=1&tailLimit=2&callback=getData&ibc=newspc&_=1478010624978'%i*30
    page = urllib.urlopen(url)
    html=page.read()
    time.sleep(1)
    get_json(html)
def get_json(json_):
  end_=re.compile(r'\);')#将json网页转化成python数据
  begain=re.compile(r'getData\(')
  json_=begain.sub('',json_)
  json_=end_.sub('',json_)
  ajson=json.loads(json_)
  lis=ajson["commentIds"]#获得每条评论的键
  n=0
  for i in range(1,len(lis)):
    try:
      xulie=re.compile('\d{10,}')#取得准确评论的键（去掉回复）
      bia=re.findall(xulie,lis[n])
      w.write(ajson['comments'][bia[len(bia)-1]]['user']['nickname'].encode('utf-8')+'|')
    except KeyError:
      w.write(ajson['comments'][bia[len(bia)-1]]['user']['location'].encode('utf-8')+'|')
    if (len(lis[n])>13):
      xulie=re.compile('\d{10,}')
      bia=re.findall(xulie,lis[n])
      w.write(ajson['comments'][bia[len(bia)-1]]['content'].encode('utf-8')+'\n')
    else:
       w.write(ajson['comments'][lis[n]]['content'].encode('utf-8')+'\n')
    n=n+1
  return lis
w=open('wangyi.txt','w')
w.write('用户名'+'|'+'热门评论'+'\n')
hot_=gethothtml('http://comment.news.163.com/api/v1/products/a2869674571f77b5a0867c3d71db5856/threads/C4QFIJNS0001875O/comments/hotList?offset=0&limit=40&showLevelThreshold=72&headLimit=1&tailLimit=2&callback=getData&ibc=newspc')
w.write('用户名'+'|'+'最新评论'+'\n')
gethnewtml()
w.close()