python selenum 爬取淘宝

# -*- coding:utf-8 -*-
# author : yesehngbao
# time:2018/3/29

import re
import pymongo

from lxml import etree
from selenium import webdriver
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.common.by import By

# from selenium.webdriver.common.utils import Keys

MONGO_HOST = 'localhost'
MONGO_PORT = 27017
MONGO_DB = 'test'
MONGO_COLL = 'selenum_tao'

webdir = webdriver.Chrome()

def get_page_num():
    webdir.get('http://www.taobao.com')
    input = WebDriverWait(webdir, 10).until(EC.presence_of_element_located((By.CSS_SELECTOR, '#q')))
    button = WebDriverWait(webdir, 10).until(EC.element_to_be_clickable((By.CSS_SELECTOR, '.btn-search')))
    input.clear()
    input.send_keys('衬衫')
    button.click()
    page_num = WebDriverWait(webdir, 10).until(EC.presence_of_element_located((By.CSS_SELECTOR, '#mainsrp-pager > div > div > div > div.total'))).text
    page_num = re.findall('\d+', page_num)[0]
    return page_num

def gain_page(page):
    try:
        WebDriverWait(webdir, 10).until(EC.presence_of_element_located((By.CSS_SELECTOR, '.items .item .pic a img')))
        input = WebDriverWait(webdir, 10).until(EC.presence_of_element_located((By.CSS_SELECTOR, '.J_Input')))
        button = WebDriverWait(webdir, 10).until(EC.element_to_be_clickable((By.CSS_SELECTOR, '.J_Submit')))
        input.clear()
        input.send_keys(page)
        button.click()
        WebDriverWait(webdir, 10).until(EC.text_to_be_present_in_element((By.CSS_SELECTOR, '#mainsrp-pager > div > div > div > ul > li.item.active > span'), str(page)))
    except Exception:
        gain_page(page)

def get_page_html(page):
    if page:
        html = webdir.page_source
        return html

def analysis_page(html):
    doc = etree.HTML(html)
    div_list = doc.xpath('.//div[@class="items"]//div[contains(@class,"item")]')
    for div in div_list:
        img = div.xpath('.//div[@class="pic"]/a/img/@data-src')[0]
        money = div.xpath('.//div[contains(@class, "price")]/strong/text()')[0]
        yield {
            'img': img,
            'money': money,
        }

def save_mongo(content):
    mongo_client = pymongo.MongoClient(host=MONGO_HOST, port=MONGO_PORT)
    db = mongo_client[MONGO_DB]
    coll = db[MONGO_COLL]
    coll.insert(content)

def main():
    page_num = get_page_num()
    for page in range(1, int(page_num)+1):
        gain_page(page)
        html = get_page_html(page)
        content = analysis_page(html)
        save_mongo(content)

if __name__ == '__main__':
    main()