京东

# -*- coding: utf-8 -*-
from selenium.webdriver.support.ui import WebDriverWait
from selenium import webdriver
import time
from lxml import etree
from pymongo import MongoClient

browser = webdriver.Chrome(executable_path='C:\My Tool\Anaconda3\envs\learn\Scripts\chromedriver.exe')
wait = WebDriverWait(browser, 10)
browser.maximize_window()

def get_response(page):
        url = 'https://search.jd.com/Search?keyword=华硕电脑&enc=utf-8&page='
        browser.get(url + str(page))
        for i in range(8):
            js ='window.scrollTo(0,%s)'%(i*1500)
            browser.execute_script(js)
            time.sleep(1)
        get_html()

def get_html():
        html = etree.HTML(browser.page_source)
        for i in html.xpath('//*[@id="J_goodsList"]/ul/li'):
            data = {
                "title" : i.xpath('string(//div[3]/a/em)'),
                "place" : i.xpath('string(.//div[2]/strong/i)'),
                "shop" : i.xpath('string(.//div[5]/span/a)'),
                "jpg" : i.xpath('string(./div/div[1]/a/img/@src)'),
                "nums" : i.xpath('string(./div/div[4]/strong/a)'),
                "title" : i.xpath('string(//div[3]/a/em)'),
            }
            save_Mongo_DB(data)
            # print(data)

def save_Mongo_DB(data):
    MongoClient("localhost")['JD']["jd"].insert(dict(data))

if __name__ == "__main__":
    for i in range(0,22,3):
        get_response(i)