基于selenium爬取拉勾网职位信息

1.selenium

　　Selenium 本是一个用于WEB应用程序测试的工具。Selenium测试直接运行在浏览器中，就像真正的用户在操作一样。而这一特性为爬虫开发提供了一个选择及方向，由于其本身依赖于浏览器，所以使用python的selenium库的前提是：需要下载相应的浏览器驱动程序，这里附上Chromedriver的下载地址：chromedriver；（64位的向下兼容，）

 

2.code：

　　说明：

　　　　1.多线程爬取（producer&consumer modal）；

　　　　2.结果集的存取文件类型为csv格式；

　　　　3.大三党一枚，代码噪点多提出来哦，相互学习。

#coding:utf-8

from selenium import webdriver
from lxml import etree
import csv
import time
import re
import threading
import queue
from selenium.webdriver.common.by import By
import random


POSITioNS_INFO = queue.Queue()
event = threading.Event()


def write_to_csv(file):
    with open(file, 'a', encoding='utf-8', newline="") as f:
        writer = csv.DictWriter(f, ['name', 'salary', 'addr', 'experience', 'degree', 'type', 'advantage', 'detail'])
        writer.writeheader()
        while True:
            if POSITIONS_INFO.empty():
                event.wait()
            writer.writerow(POSITIONS_INFO.get())



class LaGouSpider(threading.Thread):
    driver_path = r'D:\Pythonlearning\chromedriver.exe'
    options = webdriver.ChromeOptions()
    options.add_argument('--proxy-server="Http://117.63.87.177:9999"')
    def __init__(self, kw, *args, **kwargs):
        super(LagouSpider, self).__init__(*args, **kwargs)
        self.driver = webdriver.Chrome(executable_path=self.driver_path)
        self.url = r'https://www.lagou.com/jobs/list_%s?labelWords=&fromSearch=true&suginput=' % kw

    def run(self):
        self.driver.get(self.url)
        cur_page_source = self.driver.page_source
        # 总页数
        max_page_num = int(self.driver.find_element(By.XPATH, "//span[contains(@hidefocus, 'hidefocus')][last()-1]").text)
        for page in range(1, max_page_num):
            self.parse_cur_page(cur_page_source)
            # 爬取当前页， 点击下一页进行抓取
            next_page_tag = self.driver.find_element(By.CSS_SELECTOR, "div.pager_container .pager_next")
            next_page_tag.click()

            # 防止被识别， 设置随机等待秒数
            rand_seconds = random.choice([2,3])+random.random()
            time.sleep(rand_seconds)
            cur_page_source = self.driver.page_source


    def parse_cur_page(self, cur_page_source):
        """解析结果页，解析出所有的待爬取职位信息页地址"""
        html = etree.HTML(cur_page_source)
        info_links = html.xpath("//a[contains(@class, 'position_link')]")
        for link in info_links:
            # print(link.get('href'))
            link_addr = link.get('href', None)
            self.request_info_page(link_addr)


    def request_info_page(self, page_url:str):
        """获取职位详情页"""
        js_code = "window.open('%s')" % page_url
        self.driver.execute_script(js_code)
        # 解析当前信息页
        cur_handles = self.driver.window_handles
        self.driver.switch_to_window(cur_handles[-1])
        self.parse_position_page(self.driver.page_source)
        # 解析完成， 关闭当前标签卡并切换回原窗口
        self.driver.close()
        self.driver.switch_to_window(cur_handles[0])

    def parse_position_page(self, html):
        # print(html)
        """解析职位详情页"""
        list_xpath = {
            "job_name_xpath" : "//div[@class='job-name']/span[@class='name']/text()",
            "job_salary_xpath" : "//dd[contains(@class, 'job_request')]//span[@class='salary']/text()",
            "job_addr_xpath" : "//dd[contains(@class, 'job_request')]//span[2]/text()",
            "job_experience_xpath" :"//dd[contains(@class, 'job_request')]//span[3]/text()",
            "job_degree_xpath" : "//dd[contains(@class, 'job_request')]//span[4]/text()",
            "job_type_xpath" : "//dd[contains(@class, 'job_request')]//span[5]/text()",
            "job_advantage_xpath": "//dd[contains(@class, 'job-advantage')]/p/text()",
            "job_detail_xpath": "//div[@class='job-detail']/*/text()",
        }

        key_name = ['name', 'salary', 'addr', 'experience', 'degree', 'type', 'advantage', 'detail']

        cur_position_info = dict()
        cur_html = etree.HTML(html)

        for index, xpath in enumerate(list_xpath.values()):
            cur_val = cur_html.xpath(xpath)
            if len(cur_val) == 1:
                temp_value = re.sub(r'[\s/<.*?>\\]', "", cur_val[0].strip())

            else:
                temp_value = re.sub(r'[\s/<.*?>\\\]\[]', "", "".join(cur_val))
                # 去掉\xa0(&nbsp;)
                temp_value = "".join(temp_value.split())
            temp_key = key_name[index]
            cur_position_info[temp_key] = temp_value

        print(cur_position_info)
        # {'name': '高级java开发工程师', 'salary': '20k-35k', 'addr': '上海', 'experience': '经验5-10年', 'degree': '大专及以上', 'type': '全职', 'advantage': '电商微服务', 'detail': '任职资格：1、计算机或相关专业本科以上学历，具有4年以上JAVA开发经验；2、熟练掌握数据结构、多线程编程，掌握常用的设计模式；熟悉JVM，包括内存模型、类加载机制以及性能优化；3、熟练掌握Java语言，熟悉spring、springMVC、mybatis、SpringBoot等框架；4、掌握数据库设计，熟练使用数据库如Mysql、mongoDB、Redis等；5、熟悉linux常用的shell命令；6、熟悉rpc开发，有过分布式开发经验者优先；7、具有很强的分析复杂问题和解决复杂问题的能力，有强烈的责任心和使命感，良好的沟通表达能力和团队协作能力。8、熟悉常见的中间件、分布式解决方案及其原理：分布式缓存、SOA、消息中间件，负载均衡、连接池等；9、具有一定的项目规划和决策能力，善于捕捉业务需求、架构设计中存在的问题，并给出有效的解决措施和方法；'}
        POSITIONS_INFO.put(cur_position_info)
        # GCondition.notify_all()
        event.set()

def main():
    spider = LagouSpider('java')
    spider.start()
    writer = threading.Thread(target=write_to_csv, args=['lagou.csv',])
    writer.start()

if __name__ == '__main__':
    main()