Shuiwu/n20_hunnan.py

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

import requests,json
from setting import proxies
from urllib import parse
from lxml import etree
from mongo_cho import myco20,r_myco15,myco20_b
from rety import retry
r = requests.session()
r.keep_alive = False

@retry(3)
def r1(ny1,ny2,pg):
    url = 'http://hunan.chinatax.gov.cn/hardcasegetdatanew'
    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",
    }
    type_value = str(ny1) + '0' + str(ny2)
    dt = str(ny1)  + '/0' + str(ny2)
    data = {
        "type":"3",
        "type_value": type_value,
        "case_type": "1",
        "page": pg,
        "limit": "10",
        "is_search": "0",
        "taxpayerName": "",
        "taxpayerNumber": "",
        "organizationalCode": "",
        "place": "",
        "legalName": "",
        "legalIdCard": "",
        "financeName": "",
        "financeIdCard": "",
        "personName": "",
        "personIdCard": "",
        "_csrf": "fe7aeeb7-63a9-4770-9f35-84869a82d042",
    }
    response = r.post(url=url,headers=headers,data=data,proxies=proxies)
    # print(response.text)
    rsd = response.json()
    rsl = rsd['data']
    rpg = rsd['hardCasePage']['totalPages']
    # print(rpg)
    list1 = []
    list2 = []
    if rsl:
        for i in rsl:
            i['date'] = dt
            print(i)
            url1 = i['id']
            utf = r_myco15.sismember('n20', url1)  ##更改
            if not utf:
                list1.append(i)
                list2.append(url1)
            else:
                print('已存在，>>>n20')
    if list1:
        myco20.insert_many(list1)
        print('已存入原始库')
    if list2:
        myco20_b.insert_many(list1)
        print('已存入备份原始库')
        for mis in list2:
            r_myco15.sadd('n20', mis)  ##更改
    # if list1:
    #     myco20.insert_many(list1)
    return int(rpg)


def runs(ny1,ny2):
    rpg = r1(ny1,ny2,pg=1)
    if rpg>1:
        for pg in range(2,rpg+1):
            print(pg,'====')
            r1(ny1,ny2,pg)

runs('2022','11')