一、全国疫情数据爬取 1.数据表共有两个，分别为details和history，表结构如下： 2.爬取全国疫情数据代码如下： import requestsimport jsonimport timeimport pymysqlimport tracebackdef get_details(): url = '
一、全国疫情数据爬取
1.数据表共有两个，分别为details和history，表结构如下：
2.爬取全国疫情数据代码如下：
import requests
import json
import time
import pymysql
import traceback
def get_details():
url = 'https://view.inews.qq.com/g2/getOnsInfo?name=disease_h5&callback=jQuery34102848205531413024_1584924641755&_=1584924641756'
headers ={
'user-agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.25 Safari/537.36 Core/1.70.3741.400 QQBrowser/10.5.3863.400'
}
res = requests.get(url,headers=headers)
#输出全部信息
# print(res.text)
response_data = json.loads(res.text.replace('jQuery34102848205531413024_1584924641755(','')[:-1])
#输出这个字典的键值 dict_keys(['ret', 'data'])ret是响应值，0代表请求成功，data里是我们需要的数据
# print(response_data.keys())
"""上面已经转化过一次字典，然后获取里面的data，因为data是字符串，所以需要再次转化字典
print(json.loads(reponse_data['data']).keys())
结果：
dict_keys(['lastUpdateTime', 'chinaTotal', 'chinaAdd', 'isShowAdd', 'showAddSwitch',
'areaTree', 'chinaDayList', 'chinaDayAddList', 'dailyNewAddHistory', 'dailyHistory',
'wuhanDayList', 'articleList'])
lastUpdateTime是最新更新时间，chinaTotal是全国疫情总数，chinaAdd是全国新增数据，
isShowAdd代表是否展示新增数据，showAddSwitch是显示哪些数据，areaTree中有全国疫情数据
"""
areaTree_data = json.loads(response_data['data'])['areaTree']
temp=json.loads(response_data['data'])
# print(temp.keys())
# print(areaTree_data[0].keys())
"""
获取上一级字典里的areaTree
然后查看里面中国键值
print(areaTree_data[0].keys())
dict_keys(['name', 'today', 'total', 'children'])
name代表国家名称，today代表今日数据，total代表总数,children里有全国各地数据，我们需要获取全国各地数据，查看children数据
print(areaTree_data[0]['children'])
这里面是
name是地区名称，today是今日数据，total是总数，children是市级数据，
我们通过这个接口可以获取每个地区的总数据。我们遍历这个列表，取出name，这个是省级的数据，还需要获取市级数据，
需要取出name，children（市级数据）下的name、total(历史总数)下的confirm、heal、dead，today(今日数据)下的confirm（增加数），
这些就是我们需要的数据
"""
# print(areaTree_data[0]['children'])
# for province_data in areaTree_data[0]['children']:
# print(province_data)
ds= temp['lastUpdateTime']
details=[]
for pro_infos in areaTree_data[0]['children']:
province_name = pro_infos['name'] # 省名
for city_infos in pro_infos['children']:
city_name = city_infos['name'] # 市名
confirm = city_infos['total']['confirm']#历史总数
confirm_add = city_infos['today']['confirm']#今日增加数
heal = city_infos['total']['heal']#治愈
dead = city_infos['total']['dead']#死亡
# print(ds,province_name,city_name,confirm,confirm_add,heal,dead)
details.append([ds,province_name,city_name,confirm,confirm_add,heal,dead])
return details
def get_history():
url = 'https://view.inews.qq.com/g2/getOnsInfo?name=disease_other&callback=jQuery341026745307075030955_1584946267054&_=1584946267055'
headers ={
'user-agent': 'Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.25 Safari/537.36 Core/1.70.3741.400 QQBrowser/10.5.3863.400'
}
res = requests.get(url,headers=headers)
# print(res.text)
response_data = json.loads(res.text.replace('jQuery341026745307075030955_1584946267054(','')[:-1])
# print(response_data)
data = json.loads(response_data['data'])
# print(data.keys())
chinaDayList = data['chinaDayList']#历史记录
chinaDayAddList = data['chinaDayAddList']#历史新增记录
history = {}
for i in chinaDayList:
ds = '2021.' + i['date']#时间
tup = time.strptime(ds,'%Y.%m.%d')
ds = time.strftime('%Y-%m-%d',tup)#改变时间格式，插入数据库
confirm = i['confirm']
suspect = i['suspect']
heal = i['heal']
dead = i['dead']
history[ds] = {'confirm':confirm,'suspect':suspect,'heal':heal,'dead':dead}
for i in chinaDayAddList:
ds = '2021.' + i['date']#时间
tup = time.strptime(ds,'%Y.%m.%d')
ds = time.strftime('%Y-%m-%d',tup)#改变时间格式，插入数据库
confirm_add = i['confirm']
suspect_add = i['suspect']
heal_add = i['heal']
dead_add = i['dead']
history[ds].update({'confirm_add':confirm_add,'suspect_add':suspect_add,'heal_add':heal_add,'dead_add':dead_add})
return history
def get_conn():
"""
:return: 连接，游标
"""
# 创建连接
conn = pymysql.connect(host="127.0.0.1",
user="root",
password="417020",
db="db",
charset="utf8")
# 创建游标
cursor = conn.cursor() # 执行完毕返回的结果集默认以元组显示
return conn, cursor
def close_conn(conn, cursor):
if cursor:
cursor.close()
if conn:
conn.close()
def update_details():
"""
更新 details 表
:return:
"""
cursor = None
conn = None
try:
li = get_details()
conn, cursor = get_conn()
sql = "insert into details(update_time,province,city,confirm,confirm_add,heal,dead) values(%s,%s,%s,%s,%s,%s,%s)"
sql_query = 'select %s=(select update_time from details order by id desc limit 1)' #对比当前最大时间戳
cursor.execute(sql_query,li[0][0])
if not cursor.fetchone()[0]:
print(f"{time.asctime()}开始更新最新数据")
for item in li:
cursor.execute(sql, item)
conn.commit() # 提交事务 update delete insert操作
print(f"{time.asctime()}更新最新数据完毕")
else:
print(f"{time.asctime()}已是最新数据！")
except:
traceback.print_exc()
finally:
close_conn(conn, cursor)
def insert_history():
"""
插入历史数据
:return:
"""
cursor = None
conn = None
try:
dic = get_history()
print(f"{time.asctime()}开始插入历史数据")
conn, cursor = get_conn()
sql = "insert into history values(%s,%s,%s,%s,%s,%s,%s,%s,%s)"
for k, v in dic.items():
# item 格式 {'2021-01-13': {'confirm': 41, 'suspect': 0, 'heal': 0, 'dead': 1}
cursor.execute(sql, [k, v.get("confirm"), v.get("confirm_add"), v.get("suspect"),
v.get("suspect_add"), v.get("heal"), v.get("heal_add"),
v.get("dead"), v.get("dead_add")])
conn.commit() # 提交事务 update delete insert操作
print(f"{time.asctime()}插入历史数据完毕")
except:
traceback.print_exc()
finally:
close_conn(conn, cursor)
def insert_history():
"""
插入历史数据
:return:
"""
cursor = None
conn = None
try:
dic = get_history()
print(f"{time.asctime()}开始插入历史数据")
conn, cursor = get_conn()
sql = "insert into history values(%s,%s,%s,%s,%s,%s,%s,%s,%s)"
for k, v in dic.items():
# item 格式 {'2021-01-13': {'confirm': 41, 'suspect': 0, 'heal': 0, 'dead': 1}
cursor.execute(sql, [k, v.get("confirm"), v.get("confirm_add"), v.get("suspect"),
v.get("suspect_add"), v.get("heal"), v.get("heal_add"),
v.get("dead"), v.get("dead_add")])
conn.commit() # 提交事务 update delete insert操作
print(f"{time.asctime()}插入历史数据完毕")
except:
traceback.print_exc()
finally:
close_conn(conn, cursor)
def update_history():
"""
更新历史数据
:return:
"""
cursor = None
conn = None
try:
dic = get_history()
print(f"{time.asctime()}开始更新历史数据")
conn, cursor = get_conn()
sql = "insert into history values(%s,%s,%s,%s,%s,%s,%s,%s,%s)"
sql_query = "select confirm from history where ds=%s"
for k, v in dic.items():
# item 格式 {'2020-01-13': {'confirm': 41, 'suspect': 0, 'heal': 0, 'dead': 1}
if not cursor.execute(sql_query, k):
cursor.execute(sql, [k, v.get("confirm"), v.get("confirm_add"), v.get("suspect"),
v.get("suspect_add"), v.get("heal"), v.get("heal_add"),
v.get("dead"), v.get("dead_add")])
conn.commit() # 提交事务 update delete insert操作
print(f"{time.asctime()}历史数据更新完毕")
except:
traceback.print_exc()
finally:
close_conn(conn, cursor)
update_history()
insert_history()
update_details()
二、完整项目代码
1.项目结构：
2..js文件
ec_center.js
var ec_center=echarts.init(document.getElementById('c2'),"dark");
var mydata=[{'name':'上海','value':318},{'name':'云南','value':162}]
var ec_center_option={
title:{
text: '',
subtext: '',
x: 'left'
},
tooltip:{
trigger:'item'
},
//左侧小导航图标
visualMap:{
show:true,
x:'left',
y:'bottom',
textStyle:{
fontSize:8,
},
splitList: [{start:1,end:9},
{start:10,end:99},
{start:100,end:999},
{start:1000,end:9999},
{start:1000000}],
color:['#8A3310','#C64918','#E55B25','#F2AD92','#F9DCD1']
},
series:[{
name:'累计确诊人数',
type:'map',
mapType:'china',
roam:false,
itemStyle:{
normal:{
borderWidth:.5,//区域边框宽度
borderColor:'#009fe8',//区域边框颜色
areaColor:'#ffefd5',//区域颜色
},
emphasis:{//鼠标划过地图高亮
borderWidth:.5,
borderColor:'#4b0082',
areaColor:"#fff",
}
},
label:{
normal:{
show:true,
fontSize:8,
},
emphasis:{
show:true,
fontSize:8,
}
},
data:mydata//数据
}]
};
ec_center.setOption(ec_center_option)
controller.js:
function gettime(){
$.ajax({
url:"/time",
timeout:10000,
success: function(data) {
$("#time").html(data)
},
error:function(xhr,type,errorThrowm){
}
});
}
function get_c1_data(){
$.ajax({
url:"/c1",
success: function(data) {
$(".num h1").eq(0).text(data.confirm)
$(".num h1").eq(1).text(data.suspect)
$(".num h1").eq(2).text(data.heal)
$(".num h1").eq(3).text(data.dead)
},
error:function(xhr,type,errorThrowm){
}
})
}
function get_c2_data(){
$.ajax({
url:"/c2",
success: function(data) {
ec_center_option.series[0].data=data.data
ec_center.setOption(ec_center_option)
},
error:function(xhr,type,errorThrowm){
}
})
}
function get_l1_data(){
$.ajax({
url:"/l1",
success: function(data) {
ec_left1_Option.xAxis[0].data=data.day
ec_left1_Option.series[0].data=data.confirm
ec_left1_Option.series[1].data=data.suspect
ec_left1_Option.series[2].data=data.heal
ec_left1_Option.series[3].data=data.dead
ec_left1.setOption(ec_left1_Option)
},
error:function(xhr,type,errorThrowm){
}
})
}
function get_l2_data(){
$.ajax({
url:"/l2",
success: function(data) {
ec_left2_Option.xAxis[0].data=data.day
ec_left2_Option.series[0].data=data.confirm_add
ec_left2_Option.series[1].data=data.suspect_add
ec_left2.setOption(ec_left2_Option)
},
error:function(xhr,type,errorThrowm){
}
})
}
function get_r1_data(){
$.ajax({
url:"/r1",
success: function(data) {
ec_right1_Option.xAxis.data=data.city;
ec_right1_Option.series[0].data=data.confirm;
ec_right1.setOption(ec_right1_Option);
}
})
}
function get_r2_data(){
$.ajax({
url:"/r2",
success: function(data) {
ec_right2_Option.xAxis.data=data.city;
ec_right2_Option.series[0].data=data.confirm;
ec_right2.setOption(ec_right2_Option);
}
})
}
gettime()
get_c1_data()
get_c2_data()
get_l1_data()
get_l2_data()
get_r1_data()
get_r2_data()
//setInterval(gettime,1000)
//setInterval(get_c1_data,1000)
ec_left1.js:
var ec_left1 = echarts.init(document.getElementById('l1'), "dark");
var ec_left1_Option = {
title: {
text: "全国累计趋势",
textStyle: {
//color:'white',
},
left: 'left',
},
tooltip: {
trigger: 'axis',
axisPointer: {
type: 'line',
lineStyle: {
color: '#7171C6'
}
},
},
legend: {
data: ["累计确诊", "现有疑似", "累积治愈", "累计死亡"],
left: "right"
},
//图形位置
grid: {
left: '4%',
right: '6%',
bottom: '4%',
top: 50,
containLabel: true
},
xAxis: [{
type: 'category',
data: ['01.24', '01.25', '01.26']
}],
yAxis: [{
type: 'value',
axisLabel: {
show: true,
color: 'white',
fontSize: 12,
formatter: function(value) {
if (value >= 1000) {
value = value / 1000 + 'k';
}
return value;
}
},
//y轴线设置显示
axisLine: {
show: true
},
//与x轴平行的线样式
splitLine: {
show: true,
lineStyle: {
color: '#17273B',
width: 1,
type: 'solid',
}
}
}],
series: [{
name: "累计确诊",
type: 'line',
smooth: true,
data: [260, 406, 529]
}, {
name: "现有疑似",
type: 'line',
smooth: true,
data: [54, 37, 3935]},
{
name: "累积治愈",
type: 'line',
smooth: true,
data: [25, 25, 25]
},{
name: "累计死亡",
type: 'line',
smooth: true,
data: [6, 9, 17]
}]
};
ec_left1.setOption(ec_left1_Option)
ec_left2.js:
var ec_left2 = echarts.init(document.getElementById('l2'), "dark");
var ec_left2_Option = {
tooltip: {
trigger: 'axis',
//指示器
axisPointer: {
type: 'line',
lineStyle: {
color: '#7171C6'
}
},
},
legend: {
data: ['新增确诊', '新增疑似'],
left: "right"
},
//标题样式
title: {
text: "全国新增趋势",
textStyle: {
color:'yellow',
fontSize: 16
},
left: 'left'
},
//图形位置
grid: {
left: '4%',
right: '6%',
bottom: '4%',
top: 50,
containLabel: true
},
xAxis: [{
type: 'category',
//x轴坐标点开始与结束点位置都不在最边缘
// boundaryGap : true,
data: []
}],
yAxis: [{
type: 'value',
//y轴字体设置
//y轴线设置显示
axisLine: {
show: true
},
axisLabel: {
show: true,
color: 'white',
fontSize: 12,
formatter: function(value) {
if (value >= 1000) {
value = value / 1000 + 'k';
}
return value;
}
},
//与x轴平行的线样式
splitLine: {
show: true,
lineStyle: {
color: '#17273B',
width: 1,
type: 'solid',
}
}
}],
series: [{
name: "新增确诊",
type: 'line',
smooth: true,
data: []
}, {
name: "新增疑似",
type: 'line',
smooth: true,
data: []
}]
};
ec_left2.setOption(ec_left2_Option)
ec_right1.js:
var ec_right1=echarts.init(document.getElementById('r1'),"dark");
var ec_right1_Option={
title:{
text:"非湖北地区城市确诊TOP5",
textStyle:{
color:'white',
},
left:'left',
},
color:['#3398DB'],
tooltip:{
trigger:'axis',
axisPointer:{
type:'shadow',
}
},
xAxis:{
type:'category',
data:['东莞','珠海','境外输入','邢台','南京']
},
yAxis:{
type:'value',
},
series:[{
data:[99,98,96,94,93],
type:'bar',
barMaxWidth:"50%"
}]
};
ec_right1.setOption(ec_right1_Option)
ec_right2.js:
var ec_right2=echarts.init(document.getElementById('r2'),"dark");
var ec_right2_Option={
title:{
text:"湖北地区城市确诊TOP5",
textStyle:{
color:'white',
},
left:'left',
},
color:['#3398DB'],
tooltip:{
trigger:'axis',
axisPointer:{
type:'shadow',
}
},
xAxis:{
type:'category',
data:['东莞','珠海','境外输入','邢台','南京']
},
yAxis:{
type:'value',
},
series:[{
data:[99,98,96,94,93],
type:'bar',
barMaxWidth:"50%"
}]
};
ec_right2.setOption(ec_right2_Option)
（其余js文件需从Echarts官网下载）
3.main.html:
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>疫情监控</title>
<script src="../static/js/jquery-3.5.1.js"></script>
<script src="../static/js/jquery-3.6.0.min.js"></script>
<script src="../static/js/echarts.min.js"></script>
<script src="../static/js/china.js"></script>
<link href="../static/css/main.css" rel="stylesheet" />
<style>
</style>
</head>
<body>
<div id="title">全国疫情实时追踪</div>
<div id="time">我是时间</div>
<div id="l1">我是左1</div>
<div id="l2">我是左2</div>
<div id="c1">
<div class="num"><h1>123</h1></div>
<div class="num"><h1>123</h1></div>
<div class="num"><h1>123</h1></div>
<div class="num"><h1>123</h1></div>
<div class="txt"><h2>累计确诊</h2></div>
<div class="txt"><h2>剩余疑似</h2></div>
<div class="txt"><h2>累计治愈</h2></div>
<div class="txt"><h2>累计死亡</h2></div>
</div>
<div id="c2">我是中2</div>
<div id="r1">我是右1</div>
<div id="r2"></div>
<script src="../static/js/ec_center.js"></script>
<script src="../static/js/ec_left1.js"></script>
<script src="../static/js/ec_left2.js"></script>
<script src="../static/js/ec_right1.js"></script>
<script src="../static/js/ec_right2.js"></script>
<script src="../static/js/controller.js"></script>
</body>
</html>
4..py文件
main.py:
from flask import Flask
from flask import request
from flask import render_template
from flask import jsonify
import utils
app = Flask(__name__)
@app.route("/")
def hello_world():
return render_template("main.html")
@app.route("/c1")
def get_c1_data():
data=utils.get_c1_data()
return jsonify({"confirm":data[0],"suspect":data[1],"heal":data[2],"dead":data[3]})
@app.route("/c2")
def get_c2_data():
res=[]
for tup in utils.get_c2_data():
print(tup)
res.append({"name":tup[0],"value":int(tup[1])})
return jsonify({"data":res})
@app.route("/l1")
def get_l1_data():
data=utils.get_l1_data()
day,confirm,suspect,heal,dead=[],[],[],[],[]
for a,b,c,d,e in data:
day.append(a.strftime("%m-%d"))
confirm.append(b)
suspect.append(c)
heal.append(d)
dead.append(e)
return jsonify({"day":day,"confirm":confirm,"suspect":suspect,"heal":heal,"dead":dead})
@app.route("/l2")
def get_l2_data():
data=utils.get_l2_data()
day,confirm_add,suspect_add=[],[],[]
for a,b,c in data:
day.append(a.strftime("%m-%d"))
confirm_add.append(b)
suspect_add.append(c)
return jsonify({"day":day,"confirm_add":confirm_add,"suspect_add":suspect_add})
@app.route("/r1")
def get_r1_data():
data=utils.get_r1_data()
city=[]
confirm=[]
for k,v in data:
city.append(k)
confirm.append(int(v))
return jsonify({"city":city,"confirm":confirm})
@app.route("/r2")
def get_r2_data():
data=utils.get_r2_data()
city=[]
confirm=[]
for k,v in data:
city.append(k)
confirm.append(int(v))
return jsonify({"city":city,"confirm":confirm})
@app.route('/ajax',methods=["get","post"])
def index3():
name=request.values.get("name")
score=request.values.get("score")
print(f"name:{name},score:{score}")
return "10000"
@app.route("/time")
def get_time():
return utils.get_time()
if __name__ == '__main__':
app.run(debug=True)
utils.py:
import time
import pymysql
def get_time():
time_str=time.strftime("%Y{}%m{}%d{} %X")
return time_str.format("年","月","日")
def get_conn():
conn=pymysql.connect(host="127.0.0.1",
user="root",
password="417020",
db="db",
charset="utf8")
cursor=conn.cursor()
return conn,cursor
def close_conn(conn,cursor):
cursor.close()
conn.close()
def query(sql,*args):
conn,cursor=get_conn()
cursor.execute(sql,args)
res=cursor.fetchall()
close_conn(conn,cursor)
return res
def get_c1_data():
sql="select sum(confirm)," \
"(select suspect from history order by ds desc limit 1)," \
"sum(heal)," \
"sum(dead) " \
"from details " \
"where update_time=(select update_time from details order by update_time desc limit 1)"
res=query(sql)
return res[0]
def get_c2_data():
sql = "select province,sum(confirm) from details "\
"where update_time=(select update_time from details "\
"order by update_time desc limit 1) "\
"group by province"
res=query(sql)
return res
def get_l1_data():
sql="select ds,confirm,suspect,heal,dead from history"
res = query(sql)
return res
def get_l2_data():
sql="select ds,confirm_add,suspect_add from history"
res = query(sql)
return res
def get_r1_data():
sql='SELECT city,confirm FROM '\
'(select city,confirm from details '\
'where update_time=(select update_time from details order by update_time desc limit 1) '\
'and province not in ("湖北","北京","上海","天津","重庆") '\
'union all '\
'select province as city,sum(confirm) as confirm from details '\
'where update_time=(select update_time from details order by update_time desc limit 1) '\
'and province in ("北京","上海","天津","重庆") group by province) as a '\
'ORDER BY confirm DESC LIMIT 5'
res = query(sql)
return res
def get_r2_data():
sql='SELECT city,confirm FROM '\
'(select city,confirm from details '\
'where update_time=(select update_time from details order by update_time desc limit 1) '\
'and province in ("湖北") '\
'union all '\
'select province as city,sum(confirm) as confirm from details '\
'where update_time=(select update_time from details order by update_time desc limit 1) '\
'and province in ("北京","上海","天津") group by province) as a '\
'ORDER BY confirm DESC LIMIT 5'
res = query(sql)
return res
if __name__=="__main__":
print(get_r1_data())
print(get_l1_data())
print(get_l2_data())
5.最终实现效果展示（更新于2022年3月28日，由于数据爬取时间原因，与官方数据相比可能存在误差）：
官方数据：
