爬虫2021广东省普通专升本各院校专业招生计划汇总表

爬虫:2021广东省普通专升本各院校专业招生计划汇总表

代码如下：

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

import requests
import json
import pandas as pd


def Get_yxdm_yxmc_list():
    '''
        1.获取院校信息
        2.数据选择 例如:[｛'yxdm':'12345','yxmc':'华南师范大学'｝] 的字典列表yxdm_yxmc_list
        3.返回 yxdm_yxmc_list  
    '''
    url = "https://www.eeagd.edu.cn/ptzsbks/public/jbxx/getYxxxBm.jsmeb"
	# Cookie注意更新
    headers = {
        'Cookie': 'BIGipServerptzsbks_pool=1053294602.17439.0000',
        'Host': 'www.eeagd.edu.cn',
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/89.0.4389.128 Safari/537.36'
    }
    json = requests.request("POST", url, headers=headers).json()
    datalist = json["result"]["result"]
    yxdm_yxmc_list = []
    for item in datalist:
        a = {}
        a['yxdm'] = item['yxdm']
        a['yxmc'] = item['yxmc']
        yxdm_yxmc_list.append(a)

    return yxdm_yxmc_list


def Get_zyjh_list(yxdm, name, zyjh_lists):
    '''
        参数：院校代码,院校名称,全局列表zyjh_lists
        1. 根据  院校代码  获取  院校的跟专业招生计划信息datalist
        2. 专业招生计划信息 添加 key-value yxmc:华南理工大学
        3. 往 全局列表zyjh_lists 添加  专业招生计划信息
    '''

    url = "https://www.eeagd.edu.cn/ptzsbks/public/jbxx/getZyjhByYxdmBm.jsmeb?"+yxdm
    headers = {
        'Cookie': 'BIGipServerptzsbks_pool=1053294602.17439.0000',
        'Host': 'www.eeagd.edu.cn',
        'User-Agent': 'Mozilla/5.0 (Macintosh; Intel Mac 05 X 10_11_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/52.0.2743.116 Safari/537.36'

    }

    json = requests.request("POST", url, headers=headers).json()
    datalist = json["result"]["result"]
    for item in datalist:
        item['yxmc'] = name
        zyjh_lists.append(item)


def main():
    yxdm_yxmc_list = Get_yxdm_yxmc_list()  # 院校代码院校名称字典集合 yxdm_yxmc_list

    zyjh_lists = []
    for item in yxdm_yxmc_list:
        id = item['yxdm']
        name = item['yxmc']
        Get_zyjh_list(id, name, zyjh_lists)
    # 输出表格
    df = pd.DataFrame(zyjh_lists)
    df.to_excel('所有专业招生计划表.xlsx', 'zyjhsheet')


if __name__ == '__main__':
    main()