Shuiwu/n31_tj.py

#!/usr/bin/env python
# coding:utf-8

import re
import requests,json
from setting import proxies
from urllib import parse
from lxml import etree
from mongo_cho import myco31,r_myco15,myco31_b
from rety import retry
r = requests.session()
r.keep_alive = False

@retry(3)
def r1(ny1,ny2,pg):
    url = 'http://tianjin.chinatax.gov.cn/wzcx/sjcx_cxqyxx.action'
    headers = {
        "User-Agent": "Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/90.0.4430.85 Safari/537.36",
    }
    data = {
        "szsf":"11200000000",
        "nfjd": str(ny1) + '0' + str(ny2),
        "page":pg,
        "pageCount":"15"
    }
    response = r.post(url=url,headers=headers,data=data,proxies=proxies)
    html = response.text
    # print(html)
    html1 = html.replace(' ','').replace('\r','').replace('\t','').replace('\n','')
    rpg1 = re.findall(r"pageCount='(.*?)'", html1)[0]
    # print(rpg1)
    selector = etree.HTML(html)
    a = selector.xpath('//span[@class="mxxx"]')
    dt = str(ny1) + '/0' + str(ny2)
    list1 = []
    list2 = []
    for i in a:
        dict1 = {}
        nsrm = i.xpath('@data-nsrmc')
        dict1['纳税人名称'] = nsrm[0]
        nsrsbh = i.xpath('@data-nsrsbh')
        dict1['纳税人识别号'] =nsrsbh[0]
        zzjgdm = i.xpath('@data-zzjgdm')
        dict1['组织机构代码'] =zzjgdm[0]
        zcjydz = i.xpath('@data-zcjydz')
        dict1['注册地址'] =zcjydz[0]
        fddbrxm = i.xpath('@data-fddbrxm')
        dict1['姓名'] =fddbrxm[0]
        fddbrxb = i.xpath('@data-fddbrxb')
        dict1['性别'] =fddbrxb[0]
        fddbrzjmc = i.xpath('@data-fddbrzjmc')
        dict1['证件名称'] =fddbrzjmc[0]
        fddbrzjhm = i.xpath('@data-fddbrzjhm')
        dict1['证件号码'] =fddbrzjhm[0]
        ajlxmc = i.xpath('@data-ajlxmc')
        dict1['案件性质'] =ajlxmc[0]
        zywfss = i.xpath('@data-zywfss')
        dict1['主要违法事实'] =zywfss[0]
        clqk = i.xpath('@data-clqk')
        dict1['相关法律依据及税务处理处罚情况'] =clqk[0]
        dict1['date'] = dt
        print(dict1)
        list1.append(dict1)
    # return int(rpg1)
    if list1:
        myco31.insert_many(list1)


# http://tianjin.chinatax.gov.cn/wzcx/cx_zdwfaj.action?szsf=11200000000
# 此数据无法去重，遂单独更新，查看页数，季度
def runs():
    for pg in range(1,2):
        print(pg,'==========')
        r1(2023,5,pg)  ##中间2为季度，每次更新前务必加1季度
r1()