# -*- coding:gbk -*-
import re
import requests
import json
from lxml import html
#測試微博4054483400791767
comments=[]

def get_page(weibo_id):
  url='https://m.weibo.cn/status/{}'.format(weibo_id)
  html=requests.get(url).text
  regcount=r'"comments_count": (.*?),'
  comments_count=re.findall(regcount,html)[-1]
  comments_count_number=int(comments_count)
  page=int(comments_count_number/10)
  return page-1

def opt_comment(comment):
  tree=html.fromstring(comment)
  strcom=tree.xpath('string(.)')
  reg1=r'回復(fù)@.*?:'
  reg2=r'回覆@.*?:'
  reg3=r'//@.*'
  newstr=''
  comment1=re.subn(reg1,newstr,strcom)[0]
  comment2=re.subn(reg2,newstr,comment1)[0]
  comment3=re.subn(reg3,newstr,comment2)[0]
  return comment3

def get_responses(id,page):
  url="https://m.weibo.cn/api/comments/show?id={}&page={}".format(id,page)
  response=requests.get(url)
  return response

def get_weibo_comments(response):
  json_response=json.loads(response.text)
  for i in range(0,len(json_response['data'])):
    comment=opt_comment(json_response['data'][i]['text'])
    comments.append(comment)


weibo_id=input("輸入微博id，自動返回前5頁評論：")
weibo_id=int(weibo_id)
print('\n')
page=get_page(weibo_id)
for page in range(1,page+1):
  response=get_responses(weibo_id,page)
  get_weibo_comments(response)

for com in comments:
  print(com)
print(len(comments))