Python之selenium爬取考题代码并写入excel实践

忘是亡心i 2022-11-14 14:52 175阅读 0赞

import time
    
    import xlwt
    from selenium import webdriver
    from selenium.webdriver.common.by import By
    from selenium.webdriver.support import expected_conditions as EC
    from selenium.webdriver.support.ui import WebDriverWait as wwait
    # chrome和driver版本要匹配
    # http://npm.taobao.org/mirrors/chromedriver/
    browser = webdriver.Chrome(executable_path=r'F:\downloads\chromedriver_win32\chromedriver.exe')
    url = "https://www.xiaoqiqiao.com/unlogin/#/login"
    
    browser.get(url)
    time.sleep(3)
    
    # 3.刷新浏览器
    browser.refresh()
    # 4.设置浏览器的大小
    browser.maximize_window()
    
    
    username = "xxxxxx"
    passwd = "xxxxxx"
    # 输入用户名
    browser.find_element_by_xpath('//*[@id="username"]').send_keys(username)
    browser.find_element_by_xpath('//*[@id="password"]').send_keys(passwd)
    time.sleep(1)
    browser.find_element_by_xpath('//*[@id="root"]/div/div/div[2]/div[4]/div/div/form/div[3]/div/div/div/button').click()
    
    time.sleep(1)
    
    # 找到考试测评
    wwait(browser, 10).until(
        EC.presence_of_element_located((By.XPATH, '//*[@id="rootWrapper"]/div/div/div[2]/div[1]/div/ul/li[3]'))).click()
    # 找到题库练习
    wwait(browser, 10).until(EC.presence_of_element_located(
        (By.XPATH, '//*[@id="rootWrapper"]/div/div/div[2]/div[2]/div/div/div[2]/ul/li[3]'))).click()
    time.sleep(1)
    # 进入题库
    wwait(browser, 10).until(EC.presence_of_element_located(
        (By.XPATH, '//*[@id="rootWrapper"]/div/div/div[2]/div[2]/div/div/div[2]/table/tbody/tr/td[6]/span/a'))).click()
    
    time.sleep(1)
    
    handles = browser.window_handles
    browser.switch_to.window(handles[1])
    
    total_list = []
    for i in range(600):
        total_data = []
        # 点击查看答案
        for_answer = wwait(browser, 10).until(
            EC.presence_of_element_located((By.XPATH, '//div[@class="look_answer downskip"]')))
            # 报错解决https://blog.csdn.net/qq_41424519/article/details/87906207
        webdriver.ActionChains(browser).move_to_element(for_answer).click(for_answer).perform()
        # 题目
        element_topic = wwait(browser, 10).until(EC.presence_of_element_located(
            (By.XPATH, '//*[@id="rootWrapper"]/div/div/div[2]/div/div/div/div/div/div[1]/div[3]/div[1]')))
        total_data.append(element_topic.text)
    
        elements = wwait(browser, 10).until(EC.presence_of_element_located(
            (By.XPATH, '//*[@id="rootWrapper"]/div/div/div[2]/div/div/div/div/div/div[1]/div[3]')))
    
        topic_infor = elements.find_elements_by_xpath('//div[@class="topic_infor"]')
        for option in topic_infor:
            total_data.append(option.text)
        true_answer = elements.find_elements_by_xpath('//span[@class="true_answer"]')
        answer_list = []
        for answer in true_answer:
            answer_list.append(answer.text)
        total_data.append(answer_list)
        # 格式化数据，为写入excel做准备
        if len(total_data) < 8:
            count = 8 - len(total_data)
            for k in range(count):
                total_data.insert(-1, "")
        total_list.append(total_data)
        # 下一题
        # 最后一题没有下一题按钮，故做此判断
        if i < 599:
            next_question = wwait(browser, 10).until(EC.presence_of_element_located((By.XPATH, '//div[@class="downskip"]')))
            webdriver.ActionChains(browser).move_to_element(next_question).click(next_question).perform()
        else:
            break
    
    f = xlwt.Workbook()
    sheet = f.add_sheet('QA', cell_overwrite_ok=True)
    
    for hi in enumerate(total_list):
        for num in range(len(hi[1])):
            sheet.write(hi[0], num, hi[1][num])
    
    f.save("QA.xls")