爬蟲微信支付服務商

#!/usr/bin/python
# -*- coding: UTF-8 -*-

from selenium import webdriver
from selenium.webdriver.common.keys import Keys
import time
import sys
import pymysql
reload(sys)
sys.setdefaultencoding('utf-8')

driver = webdriver.Firefox(executable_path="/Users/chanming/Desktop/geckodriver")
driver.get("https://pay.weixin.qq.com/index.php/partner/public/search")
time.sleep(2)
st=driver.find_element_by_id('searchPortalText')
st.send_keys(u"付")
driver.find_element_by_id("searchPortalSubmit").click()
time.sleep(3)
alist=list()
# map={'拓展地域：':'dis','聯 系 人 ：':'userName','電　　話：':'phone','郵　　箱：':'email','地　　址：':'address','所屬行業：':'industry'}
map={}
flag=True
for x in xrange(1,23):
	nextInput=driver.find_element_by_css_selector('.goto-area input')
	nextInput.clear()
	nextInput.send_keys(x)
	nextButton=driver.find_element_by_css_selector('.goto-area a')
	nextButton.click()
	time.sleep(1)
	searchList=driver.find_elements_by_css_selector('#searchResultList dl')
	for comp in searchList:
		obj={}
		obj['title']=comp.find_element_by_tag_name('dt').text
		propNameList=comp.find_elements_by_class_name('lbl')
		if flag:
			map[propNameList[0].text]='dis'
			map[propNameList[1].text]='userName'
			map[propNameList[2].text]='phone'
			map[propNameList[3].text]='email'
			map[propNameList[4].text]='address'
			map[propNameList[5].text]='industry'
			flag=False
		propList=comp.find_elements_by_class_name('ele')
		k=0
		for propName in propNameList:
			obj[map[propName.text]]=propList[k].text
			k=k+1
		alist.append(obj)
# print len(alist)
driver.close()
db = pymysql.connect(host = '127.0.0.1', port = 3306, user = 'json', passwd = '123456', db = 'youxia', charset="utf8")
cursor = db.cursor()
sql = 'insert into wx_pay_company(dis,userName,phone,email,address,industry,title) values(%s,%s,%s,%s,%s,%s,%s)'
for company in alist:
	if not company.has_key('dis'):
		company['dis']=''
	if not company.has_key('userName'):
		company['userName']=''
	if not company.has_key('phone'):
		company['phone']=''
	if not company.has_key('email'):
		company['email']=''
	if not company.has_key('address'):
		company['address']=''
	if not company.has_key('industry'):
		company['industry']=''
	cursor.execute(sql,(company['dis'],company['userName'],company['phone'],company['email'],company['address'],company['industry'],company['title']))
	db.commit()
cursor.close()
db.close()


print '運行結束'