python_获取历史天气网信息，并保存在Excel中

获取历史天气网信息，并保存在Excel中

# -*- coding: utf-8 -*-
#获取天气信息并存入excel

import requests
from bs4 import BeautifulSoup as bs
import pandas as pd
import datetime
import time
from dateutil.relativedelta import relativedelta
import xlsxwriter as xlw

#pd.date_range(start_date, end_date) 产生时间序列
#datetime.strftime(x, '%Y%m') 有了datetime对象，要把它格式化为字符串显示给用户，就需要转换为str,格式化为201801
#sort 是应用在 list 上的方法，sorted 可以对所有可迭代的对象进行排序操作。
#list(set())函数：对原列表去重并按从小到大排序
def date_range(start_date, end_date): #计算日期范围，返回日期的list
    date_list1 = [datetime.datetime.strftime(x, '%Y%m') for x in list(pd.date_range(start_date, end_date))] #获得时间序列
    date_list = sorted(list(set(date_list1)))#对list去重并排序
    #print(date_list1[0])
    return date_list

#计算间隔的天数
def date_day_range(start_date, end_date):
    strptime, strftime = datetime.datetime.strptime, datetime.datetime.strftime
    end = strptime(end_date, "%Y-%m") + relativedelta(months=+1)#月份加一
    days = (end - strptime(start_date, "%Y-%m")).days  #计算天数差
    print(days) #返回天数
    return days

def get_data_city(city, start_date, end_date):
    temp = []
    k = 0
    days = date_day_range(start_date, end_date)
    weather_data = [[] for i in range(days)] #创建一个二维list, 一个一维list对应一天的信息
    date_list = date_range(start_date, end_date)
    
    for i in date_list:
        url = 'https://lishi.tianqi.com/' + city +  '/' + i + '.html'
        headers = {'User-Agent':'Mozilla/5.0'}
        res = requests.get(url, headers=headers)#请求数据
        soup = bs(res.text, 'html.parser')#解析数据
        data1 = soup.find_all('ul', class_='thrui')#寻找数据
        data = data1[0].find_all('li')#寻找数据
        #print(data)
        for item in data:
            data_list = item.get_text().splitlines()#按行分到一个list
            #print(data_list)
            #print(len(data_list))
            
            for j in range(len(data_list)):#清空list里的空字符
                if len(data_list[j]) != 0:
                    temp.append(data_list[j])
            #print(temp)
            weather_data[k].extend(temp)#添加在二维list的一维list
            temp.clear()
            k += 1
            #time.sleep(0.5)#延时，防止被认为是爬虫
            
    #print(weather_data)
    print('geted weather_data')
    list_to_excel(city, start_date, end_date, weather_data)

def list_to_excel(city, start_date, end_date, data):
    filename = city 
    print(filename)
    workbook = xlw.Workbook('%s.xlsx' %  filename)
    sheet_name = start_date + '_' + end_date
    sheet = workbook.add_worksheet(sheet_name)
    title = ['日期', '最高气温', '最低气温', '天气', '风向', '风力'] #写第一排
    for i in range(len(title)):
        sheet.write_string(0, i, title[i], workbook.add_format({'bold': True}))  # 写入表头，字体加粗
    
    row, col = 1, 0
    for item in data:   #写内容
        for i in item:
            sheet.write_string(row, col, i)
            col += 1
        row += 1
        col = 0
        
    workbook.close()


if __name__ == '__main__':
    print('start')
    get_data_city('chongqing', '2011-01', '2021-02')
    print('end')