爬取瓜子二手车网站项目

import re, time, redis
from lxml import etree
from selenium import webdriver

class Guazi:
    def __call__(self, *args, **kwargs):
        # 初始url
        self.base_url = 'https://www.guazi.com/bj/buy'
        # 创建数据库连接
        self.r = self.get_redis()
        # selsnium方式打开网页
        self.driver = self.get_driver()
        # 所有城市url  缺少/buy
        self.city_url_list = self.get_city_url()
        # 所有品牌后缀
        self.pai_url_list = self.get_city_pai()
        self.run()

    # 主函数
    def run(self):
        # 储存所有url到数据库
        # self.get_cmp_url()
        # 浏览数据库取出所有url组成一个列表
        complete_url_list = self.r.lrange("guazi_url_list",0,-1)
        for complete_url in complete_url_list:
            try:
                complete_url1 = str(complete_url,"utf-8")
                self.driver.get(complete_url1)
                # page_source直接返回网页源码
                html = self.driver.page_source
                # 转换为xpath对象
                html_xml = etree.HTML(html)
                self.get_zuihou(html_xml)
                # print(car_dict)
                # self.r.lpush("cmp_guazi_url",car_dict)
                # print(complete_url2)
            except Exception as e:
                print(e)

    # 取最大页码 与页码组合后48865个url 并存储到redis数据库
    def get_max_yema(self, html_xml, url):
        try:
            # 用xpath取最大页码
            max_ye = html_xml.xpath("//ul[@class='pageLink clearfix']/li[last()-1]/a/span/text()")
            # print(max_ye)
            ba_url = re.findall(r'(.+)#bread',url)
            # print(ba_url)
            if max_ye:
                for eve in range(int(max_ye[0])):
                    # print(eve)
                    complete_url = ba_url[0] + 'o' + str(eve+1)
                    # print(complete_url)
                    # 存储到数据库
                    self.lpush(complete_url)
            else:
                # print(url)
                self.lpush(url)
        except Exception as e:
            print(e)

    """
    把每一个城市的所有品牌的网址打印并爬取，总共43665个网址，
    """
    def get_cmp_url(self):
        for eve in self.city_url_list:
            for cvc in self.pai_url_list:
                # 组成每一个城市的所有牌子的url
                url = eve + cvc
                self.driver.get(url)
                html = self.driver.page_source
                html_xml = etree.HTML(html)
                time.sleep(0.5)
                # 查最大页码 循环组成新的url并储存到数据库
                self.get_max_yema(html_xml, url)

    # 获取城市地区品牌的url  231个牌子
    def get_city_pai(self):
        self.driver.get(self.base_url)
        # page_source直接返回网页源码
        html = self.driver.page_source
        # 转换为xpath对象
        html_xml = etree.HTML(html)
        # 品牌后缀
        str_pai_list = html_xml.xpath("//div[@class='dd-all clearfix js-brand js-option-hid-info']//a/@href")
        pai_url_list = []
        for eve in str_pai_list:
            str_pai_list = re.findall(r'/bj(.+)', eve)[0]
            pai_url_list.append(str_pai_list)
        # print(pai_url_list)
        return pai_url_list

    # 获取城市地区url 缺少/buy 223个城市
    def get_city_url(self):
        self.driver.get(self.base_url)
        # page_source直接返回网页源码
        html = self.driver.page_source
        # print(html)
        # 转换为xpath对象
        html_xml = etree.HTML(html)
        # 取出城市后缀
        str_city_list = html_xml.xpath("//div[@class='city-box all-city']//div/dl/dd/a/@href")
        # print(str_city_list)
        city_url_list = []
        for city_url in str_city_list:
            # 组成城市url
            city_url = "https://www.guazi.com" + city_url
            city_url = re.findall(r'(.+)/buy', city_url)[0]
            if city_url not in city_url_list:
                city_url_list.append(city_url)
        # print(len(city_url_list))
        return city_url_list

    # 取出数据
    def get_zuihou(self, html_xml):
        car_list = html_xml.xpath("//ul[@class='carlist clearfix js-top']//li")
        car_dict = {}
        for car in car_list:
            car_img_list = car.xpath(".//img/@src")
            car_img = car_img_list[0] if car_img_list else ''
            car_name = car.xpath(".//a/@title")[0]
            car_info = car.xpath(".//div[@class='t-i']/text()")
            car_time = car_info[0]
            car_km_num = car_info[1]
            car_price = car.xpath(".//div[@class='t-price']/p/text()")[0] + '万'
            car_detail_url = car.xpath(".//a/@href")[0]
            car_detail_url = 'https://www.guazi.com' + car_detail_url

            car_dict = {
                "car_img": car_img,
                "car_name": car_name,
                "car_time": car_time,
                "car_km_num": car_km_num,
                "car_price": car_price,
                "car_detail_url": car_detail_url,
            }
            # print(car_dict)
            if car_dict:
                # 取出的数据存入数据库
                self.r.lpush("guazi_url", str(car_dict))
                print(car_dict)
            else:
                print("000")
            # return car_dict

    # 存入redis数据
    def lpush(self, guzi_url):
        self.r.lpush("guazi_url_list", guzi_url)
        print(guzi_url)

    # 创建redis连接
    def get_redis(self):
        return redis.Redis(host='127.0.0.1', port=6379)

    # 创建selsnium方式
    def get_driver(self):
        return webdriver.PhantomJS(
            executable_path=r"E:\EVE\phantomjs-2.1.1-windows\bin\phantomjs"
        )

    def __del__(self):
        """
        作用：代码执行完执行此函数
        :return:
        """
        self.driver.close()
        self.driver.quit()


if __name__ == '__main__':
    guazi = Guazi()
    guazi()

项目爬取流程
1：先爬取所有城市爬取瓜子二手车网站
在这里插入图片描述
2：在爬取每一个城市的每一个品牌

品牌数是固定的，所以爬取一次便可以了，然后组合与城市组合成新的URL
3：瓜子网每一个城市的每一个品牌都没超过50页，所以就不在爬取车系了（瓜子网最多显示50页的汽车信息，每一页40辆）
在这里插入图片描述
爬取最大页码，找出页码的规律，循环组成新的url，全部存进数据库。

4：再从数据库取出每一条url在爬取车辆信息，再存入一个新的数据库。