闲来无聊,用python抓取天气信息,简单就是美啊
/*** 作者：张荣华
* 日期：2009-02-26
**/
晚上闲来无聊,于是用python写了一个抓取天气信息的脚本,您要是不想去网站上看天气,每天开机的时候运行一下脚本就完事了
同学们请看
#encoding:UTF-8'''@author: ahuaxuan (张荣华)@date: 2009-02-06'''import refrom httplib import HTTPConnectionfrom datetime import datetimeimport urllibimport urllib2import sysheader = {}header['User-Agent'] = 'Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.0.3) Gecko/2008092510 Ubuntu/8.04 (hardy) Firefox/3.0.3'header['Accept-Language'] = 'en-us,en;q=0.7,zh-cn;q=0.3'header['Accept-Encoding'] = 'UTF-8'header['Accept-Charset'] = 'ISO-8859-1,utf-8;q=0.7,*;q=0.7'header['Keep-Alive'] = '300'header['Connection'] = 'keep-alive'header['Referer'] = 'http://www.weather.com.cn/'def postData(url, body): opener = urllib2.build_opener() urllib2.install_opener(opener) req = urllib2.Request(url = url, data = urllib.urlencode(body), headers=header) u = urllib2.urlopen(req) htmlSource = u.read() return htmlSourcedef getData(url): req = urllib2.Request(url,headers = header) res = urllib2.urlopen(req) html = res.read() res.close() return htmltodayPattern = '<div class="box_contenttodayinwea" id="c_1_1">[\s]*?'+\ '<p>[\s]*?<span>[\s\S]*?</span>[\s]*?' + \ '<em><strong>(?P<wea>[\s\S]*?)</strong></em>[\s]*?' + \ '<em class="no_today">(?P<temp>[\s\S]*?)</em>[\s]*?' + \ '<em>(?P<wind>[\s\S]*?)</em><br/>[\s]*?</p>' futurePattern = '<div class="fut_weatherbox7">[\s]*?' + \ '<h3>(?P<date>[\s\S]*?)</h3>[\s]*?'+ \ '<p>(?P<aa>[\s\S]*?)</p>[\s]*?'+ \ '<h4 class="temp00_dn">(?P<wea>[\s\S]*?)</h4>[\s]*?'+ \ '<h4 class="temp01_dn">(?P<tempH>[\s\S]*?)</h4>[\s]*?'+ \ '<h4 class="temp02_dn">(?P<tempL>[\s\S]*?)</h4>[\s]*?'+ \ '<h4 class="temp03_dn"><a name="sk">(?P<wind>[\s\S]*?)</a></h4>[\s]*?'+ \ '</div>''''return the json format data of weatherhttp://search.weather.com.cn/static/url.php'''def getWeather(cityName, url): body = {} body['cityinfo'] = cityName text = postData(url, body) pattern = re.compile("<meta http-equiv=\"refresh\" content=\"0;URL=([\\s\\S]*?)\">") rlst = pattern.findall(text) text = getData(rlst) fieldList = ['wind', 'wea', 'temp'] lst = parserWeaPattern(fieldList, text, todayPattern) fieldList2 = ['wind', 'wea', 'date', 'tempH', 'tempL'] lst2 = parserWeaPattern(fieldList2, text, futurePattern) print u"--------------今日天气:-------------".encode("GBK") for aa in lst: for key,value in aa.items(): print value print u"\r\n--------------未来几天:--------------".encode("GBK") for aa in lst2: print '%s,%s,%s,%s,%s'%(aa['date'],aa['wind'],aa['wea'],aa['tempH'],aa['tempL']) return lstdef parserWeaPattern(fieldList, text, pattern): list = [] p = re.compile(pattern) iterator = p.finditer(text) for matcher in iterator: data = {} for field in fieldList: data = matcher.group(field).encode("GBK") list.append(data) return list if __name__ == '__main__': #也可以用区号 city = u'上海'.encode("GBK") reload(sys) sys.setdefaultencoding('UTF-8') print city getWeather(city, 'http://search.weather.com.cn/static/url.php')
运行后的结果为:
上海--------------今日天气:-------------多云14℃/ 6℃东南风3-4级--------------未来几天:--------------7日星期六,东南风3-4级,多云,高温：15℃,低温：7℃8日星期日,东南风4-5级,多云转阴,高温：14℃,低温：9℃9日星期一,微风,晴,高温：10℃,低温：4℃10日星期二,微风,晴,高温：16℃,低温：9℃11日星期三,微风转北风3-4级,晴,高温：21℃,低温：14℃12日星期四,北风3-4级转微风,晴转中雨,高温：21℃,低温：16℃
页:
[1]