p4fkfpicj/O.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "executemany() missing 1 required positional argument: 'args'\n",
      "+++ save_country_datas, data len: 214\n",
      "+++ save_country_datas is over.\n"
     ]
    }
   ],
   "source": [
    "import datetime\n",
    "import requests\n",
    "from bs4 import BeautifulSoup\n",
    "import re\n",
    "import json\n",
    "import pymysql\n",
    "\n",
    "# 定义实体类\n",
    "def Outside(ds):\n",
    "    class Country:\n",
    "        def __init__(self):\n",
    "            self.countryName = ''\n",
    "            self.currentConfirmedCount = 0 # 现有确诊病例数\n",
    "            self.confirmedCount = 0 # 累计确诊\n",
    "            self.confirmedCountRank = 0 # 累计确诊排名\n",
    "            self.curedCount = 0  # 累计治愈\n",
    "            self.deadCount = 0 # 累计死亡\n",
    "            self.deadCountRank = 0 # 累计死亡排名\n",
    "            self.deadRate = 0.0 # 死亡率\n",
    "            self.deadRateRank = 0 # 死亡率排名\n",
    "            self.updatedTime = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')\n",
    "        def get_info_tuple(self):\n",
    "            return (self.countryName, self.currentConfirmedCount, self.confirmedCount, self.confirmedCountRank, self.curedCount, self.deadCount, self.deadCountRank, self.deadRate, self.deadRateRank,self.updatedTime)\n",
    "        def __str__(self):\n",
    "            return 'countryName:%s,currentConfirmedCount:%d,confirmedCount:%d,\\\n",
    "    confirmedCountRankt:%d,curedCount:%d,deadCount:%d,deadCountRank:%d,deadRate:%d,deadRateRank:%d,updatedTime:%s' % (self.countryName, self.currentConfirmedCount, self.confirmedCount, self.confirmedCountRank, self.curedCount, self.deadCount, self.deadCountRank, self.deadRate, self.deadRateRank,self.updatedTime)\n",
    "\n",
    "    class MyDB:\n",
    "        def __init__(self, host, user, passwd, db):\n",
    "\n",
    "            self.conn = pymysql.connect(host, user, passwd, db)\n",
    "            self.cursor = self.conn.cursor()\n",
    "\n",
    "\n",
    "\n",
    "        def get_country_list_tuple(self, all_country):\n",
    "            info_tuple = []\n",
    "            for item in all_country:\n",
    "                info_tuple.append(item.get_info_tuple())\n",
    "            return info_tuple\n",
    "\n",
    "        # 保存数据\n",
    "        def save_country_datas(self, all_country):\n",
    "            date=datetime.datetime.now().strftime('%Y-%m-%d')\n",
    "            sql='delete from country_daily_datas where updatedTime like \"%s\"'%(date+\"%\")\n",
    "            try:\n",
    "                self.cursor.executemany(sql)\n",
    "                self.conn.commit()\n",
    "            except Exception as e:\n",
    "                print(e)\n",
    "\n",
    "            sql = 'insert into country_daily_datas(countryName,currentConfirmedCount,\\\n",
    "    confirmedCount,confirmedCountRank,curedCount,deadCount,deadCountRank,deadRate,deadRateRank,pub_time) values(%s,%s,%s,%s,%s,%s,%s,%s,%s,%s)'\n",
    "            res = self.get_country_list_tuple(all_country)\n",
    "\n",
    "            print('+++ save_country_datas, data len: %d' % len(res))\n",
    "            try:\n",
    "                self.cursor.executemany(sql, res)\n",
    "                self.conn.commit()\n",
    "            except Exception as e:\n",
    "                print(e)\n",
    "            print('+++ save_country_datas is over.')\n",
    "        def show_country_datas(self):\n",
    "                self.cursor.execute('select * from country_daily_datas')\n",
    "\n",
    "\n",
    "        def __del__(self):\n",
    "            if self.conn is not None:\n",
    "                self.conn.close()\n",
    "\n",
    "\n",
    "    def forign_data_search(ds):\n",
    "        db = MyDB(host = ds[0],user = ds[1],passwd = ds[2],db = ds[3])\n",
    "        res = requests.get('https://ncov.dxy.cn/ncovh5/view/pneumonia')# 爬取页面\n",
    "        res = res.content.decode('utf-8') # 重新解码\n",
    "        soup = BeautifulSoup(res, 'html.parser')# 构建soup对象\n",
    "        tag = soup.find('script', attrs={'id':'getListByCountryTypeService2true'}) # Tag# 使用soup对象查找实时播报新闻标签\n",
    "        tagStr = tag.string# 获取内容\n",
    "        results = re.findall('\\{\"id\".*?\"showRank\".*?\\}', tagStr) # length: 34, [str, str, ....]# 使用正则表达式匹配\n",
    "        all_country = []\n",
    "        for item in results:\n",
    "            country=Country() \n",
    "            obj = json.loads(item) # obj -> dict\n",
    "            country.countryName = obj['provinceName']\n",
    "            country.currentConfirmedCount = int(obj['currentConfirmedCount'])\n",
    "            country.confirmedCount = int(obj['confirmedCount'])\n",
    "            country.curedCount = int(obj['curedCount'])\n",
    "            country.deadCount = int(obj['deadCount'])\n",
    "            country.deadRate = float(obj['deadRate'])\n",
    "            country.updatedTime = datetime.datetime.now().strftime('%Y-%m-%d %H:%M:%S')\n",
    "            try:\n",
    "                country.deadCountRank = int(obj['deadCountRank'])\n",
    "                country.deadRateRank = int(obj['deadRateRank'])\n",
    "                country.confirmedCountRank = int(obj['confirmedCountRank'])\n",
    "            except KeyError:\n",
    "                country.deadCountRank = 0\n",
    "                country.deadRateRank = 0\n",
    "                country.confirmedCountRank = 0\n",
    "            finally:\n",
    "                all_country.append(country)\n",
    "\n",
    "        db.save_country_datas(all_country)\n",
    "        \n",
    "    forign_data_search(ds)\n",
    "\n",
    "ds=['localhost','root','20Z00t10x28_my','covid19']\n",
    "Outside(ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}