數(shù)據(jù)來源：官方 中華人民共和國環(huán)境保護(hù)部
代碼：
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
# 獲取國內(nèi)城市 30天內(nèi) 每天空氣質(zhì)量指數(shù)（AQI）
# 數(shù)據(jù)來源：中華人民共和國環(huán)境保護(hù)部
import requests
from bs4 import BeautifulSoup
import datetime
ref_url = 'http://datacenter.mep.gov.cn:8099/ths-report/report!list.action'
req_url = 'http://datacenter.mep.gov.cn:8099/ths-report/report!list.action'
def get_html_content(city='西安市', V_DATE='2017-03-01', E_DATE='2017-03-03'):
headers = {
'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,image/webp,*/*;q=0.8',
'Accept-Encoding': 'gzip, deflate',
'Accept-Language': 'zh-CN,zh;q=0.8',
'Content-Type': 'application/x-www-form-urlencoded',
'Referer': ref_url,
'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36'
}
data_dict = {
'page.pageNo': '1',
'xmlname': '1462259560614',
'queryflag': 'close',
'CITY': city,
'isdesignpatterns': 'false',
'V_DATE': V_DATE,
'E_DATE': E_DATE,
}
r = requests.post(req_url, data=data_dict, headers=headers, timeout=60)
return r.text
def date_format():
dd = datetime.datetime.now()
E_DATE = dd.strftime('%Y-%m-%d')
_31DayAgo = (dd - datetime.timedelta(days=31))
V_DATE = _31DayAgo.strftime('%Y-%m-%d')
return (V_DATE, E_DATE)
def get_aqi_info(html_content):
soup = BeautifulSoup(html_content, 'html.parser')
aqi_table = soup.find(id='GridView1')
aqi_trs = aqi_table.find_all('tr')[1:]
result = {}
for aqi_tr in aqi_trs:
aqi_tds = aqi_tr.find_all('td')
aqi = aqi_tds[3].string
dd = aqi_tds[6].string
result[dd] = aqi
return result
if __name__ == '__main__':
city = input('請輸入城市名稱，如"西安市"： ')
(V, E) = date_format()
content = get_html_content(city, V, E)
result = get_aqi_info(content)
print(city, '最近30天空氣質(zhì)量指數(shù)如下：\n')
for key in sorted(result.keys()):
print(key, '\t', result[key])
執(zhí)行結(jié)果如圖：
AQI數(shù)據(jù).png
