爬虫-中国福利彩票开奖信息
import re import requests import json import urllib import os import sys import random import time from datetime import datetime,timedelta import sqlite3 from flask import Flask from flask_sqlalchemy import SQLAlchemy from sqlalchemy import Integer, Float, String, Boolean, DECIMAL, Enum, Text from sqlalchemy import CHAR, SMALLINT, Date, Time, DateTime, TIMESTAMP, Text, JSON from sqlalchemy import Column, ForeignKey, func from sqlalchemy.dialects.mysql import DOUBLE, LONGTEXT from fake_useragent import UserAgent ua = UserAgent(verify_ssl=False) app = Flask(__name__) app.config['SQLALCHEMY_DATABASE_URI'] = f"mysql+pymysql://root:root@localhost:3306/sakila?charset=utf8" app.config['SQLALCHEMY_TRACK_MODIFICATIONS'] = True app.config['SQLALCHEMY_COMMIT_ON_TEARDOWN'] = True db = SQLAlchemy(app) class kjxx(db.Model): __tablename__ = 'cwl_kjxx' # 中国福利彩票开奖信息 name=db.Column(db.String(255), primary_key=True, nullable=False, comment='彩票类型') blue=db.Column(db.String(255), nullable=False, comment='中奖号码(蓝)') code=db.Column(db.String(255), primary_key=True, nullable=False, comment='期号') content=db.Column(db.String(255), nullable=True, comment='一等奖中奖情况') date=db.Column(db.String(255), nullable=False, comment='开奖日期') poolmoney=db.Column(db.String(255), nullable=False, comment='奖池(元)') red=db.Column(db.String(255), nullable=False, comment='中奖号码(红)') sales=db.Column(db.String(255), nullable=False, comment='总销售额(元)') prizegrades=db.Column(db.JSON, nullable=True, comment='中奖情况') def get_kjxx(): '''开奖信息''' day = datetime.now() name_dict = { 'ssq':'双色球', 'kl8':'快乐8', '3d':'福彩3D', 'qlc':'七乐彩', } for name_en,name_ch in name_dict.items(): for i in range(999): headers = { 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/avif,image/webp,image/apng,*/*;q=0.8,application/signed-exchange;v=b3;q=0.9', 'Accept-Encoding': 'gzip, deflate', 'Accept-Language': 'zh-CN,zh;q=0.9', 'Connection': 'keep-alive', 'Upgrade-Insecure-Requests': '1', 'Host': 'www.cwl.gov.cn', #'Referer': 'http://www.cwl.gov.cn/ygkj/wqkjgg/ssq/', 'User-Agent': ua.random, #'X-Requested-With': 'XMLHttpRequest', } end_date = (day - timedelta(days=1)).strftime('%Y-%m-%d') day= day - timedelta(days=70) # 每次统计30期,即每周3期共70天 start_date = day.strftime('%Y-%m-%d') url = f"http://www.cwl.gov.cn/cwl_admin/front/cwlkj/search/kjxx/findDrawNotice?name={name_en}&issueCount=&issueStart=&issueEnd=&dayStart={start_date}&dayEnd={end_date}" res = requests.get(url,headers=headers) #time.sleep(2) results = res.json() if not results['result']: break obj_list = [] for row in results['result']: obj = kjxx() obj.name = row['name'] obj.blue = row['blue'] obj.code = row['code'] obj.content = row['content'] obj.date = row['date'] obj.poolmoney = row['poolmoney'] obj.red = row['red'] obj.sales = row['sales'] obj.prizegrades = row['prizegrades'] obj_list.append(obj) print(name_ch, start_date,'至',end_date,results['message']) db.session.add_all(obj_list) db.session.commit() db.session.close() if __name__=='__main__': db.create_all() get_kjxx()
posted on 2022-08-14 09:37 PythonBaby 阅读(380) 评论(1) 收藏 举报
浙公网安备 33010602011771号