自动化测试 selenium还是puppeteer？

这两种自动化控制浏览器的模块我都用过，因为我用的是python操作，所以用的是pyppeteer库。
首先，对于基础操作来说，比如输密码登录，点击各种元素之类，selenium和puppeteer都能满足
环境上来说，selenium能支持更多的浏览器，puppeteer貌似只支持chrome？
部署上来说，selenium需要自己下载目前chrome浏览器对应版本的chromedriver，puppteer可以自己下载chrome开发版本，这个是puppteer稍微方便一点。
如果你是想处理某些坑爹的case，比如淘宝滑块验证码这种，貌似只能用puppteer。
语法上来看，selenium主要是同步语法，puppteer则是一堆异步，如果你是python或者java开发，那用selenium好太多了，如果是js开发可以考虑puppteer。

最后，selenium真香。

附两种代码例子

1 我封装的功能比较完整的selenium操作类

import os
from selenium import webdriver
from selenium.webdriver.support import expected_conditions as EC
from selenium.webdriver.common.by import By
from selenium.webdriver.support.ui import WebDriverWait
from selenium.webdriver.common.action_chains import ActionChains
from selenium.webdriver.common.keys import Keys


# 文档https://www.cnblogs.com/taceywong/p/6602927.html?utm_source=tuicool&utm_medium=referral

# 中文apihttp://selenium-python-zh.readthedocs.io/en/latest/
class ChromeOperate():
    def __init__(self, url='', executable_path='', User_data_dir='', arguments=[], headless=False, no_pic=False):
        option = webdriver.ChromeOptions()
        if User_data_dir:
            option.add_argument('--user-data-dir=%s' % User_data_dir)  # 设置成用户自己的数据目录
        else:
            import getpass
            username = getpass.getuser()
            default_path = r'C:\Users\%s\AppData\Local\Google\Chrome\User Data' % username  # echo %LOCALAPPDATA%\Google\Chrome\User Data
            print(default_path)
            if os.path.exists(default_path):
                # option.add_argument('--user-data-dir=%s' % default_path)
                pass
        if no_pic:
            prefs = {"profile.managed_default_content_settings.images": 2}  # 如果出现图片显示不了需要在配置文件里改回1
            option.add_experimental_option("prefs", prefs)
        option.add_argument('--start-maximized')
        if headless:
            option.add_argument('headless')
        option.add_argument('google-base-url=%s' % 'https://www.baidu.com/')
        for argument in arguments:
            option.add_argument(argument)
        if not executable_path: executable_path = r'C:\Users\Administrator\Desktop\chromedriver.exe'
        self.driver = webdriver.Chrome(executable_path=executable_path, chrome_options=option)

        if url: self.open(url)

    def open(self, url):
        self.driver.get(url)  # self.driver.get(url).page_source

    def open_source(self):
        return self.driver.page_source

    def title(self):
        self.title = self.driver.title
        return self.title

    def quit(self):
        self.driver.quit()

    def find_element_by_name(self, name):
        return self.driver.find_element_by_name(name)

    def find_elements_by_xpath(self, xpath):  # 貌似不能用/text
        return self.driver.find_elements_by_xpath(xpath)

    def find_element_by_id(self, id):
        try:
            return self.driver.find_element_by_id(id)
        except:
            return None

    def input_words(self, element, words):
        element.clear()
        element.send_keys(words)

    def click_by_id(self, id):
        self.driver.find_element_by_id(id).click()

    def send_file(self, element, path):
        element.sendKeys(path);

    def wait_element(self, element_id):
        WebDriverWait(self.driver, 10).until(
            EC.presence_of_element_located((By.ID, element_id))
        )

    def get_title(self):
        print(self.driver.title)
        return self.driver.title

    def refresh(self):
        self.driver.refresh()  #

    def down_page(self):
        from selenium.webdriver.common.keys import Keys
        self.driver.execute_script("window.scrollBy(0,3000)")

    def get_cookie(self):
        cookies_list = self.driver.get_cookies()
        cookie_dict = {i["name"]: i["value"] for i in cookies_list}
        return cookie_dict

    def close_tab(self):
        ActionChains(self.driver).key_down(Keys.CONTROL).send_keys("w").key_up(Keys.CONTROL).perform()

    def close_other_tab(self):
        handle = self.driver.current_window_handle
        # 获取当前所有窗口句柄（窗口A、B）
        handles = self.driver.window_handles
        # 对窗口进行遍历
        for newhandle in handles:
            # 筛选新打开的窗口B
            if newhandle != handle:
                # 切换到新打开的窗口B
                # browser.switch_to_window(newhandle) 旧版本
                self.driver.switch_to.window(newhandle)

                # # 在新打开的窗口B中操作
                # self.driver.find_element_by_id('xx').click()

                # 关闭当前窗口B
                self.driver.close()

                # 切换回窗口A
                self.driver.switch_to_window(handles[0])


'''
验证码截取
链式操作ActionChains(driver).move_to_element(menu).click(hidden_submenu).perform()

// Copy the element screenshot to disk
File screenshotLocation = new File("C:\\images\\GoogleLogo_screenshot.png");
FileUtils.copyFile(screenshot, screenshotLocation);

'''

if __name__ == '__main__':
    cop = ChromeOperate(executable_path=r'chromedriver.exe')

    cop.open("https://videoquickedit.com/")
    print(cop.driver.current_url)

2 pyppeteer

import asyncio
import time
from pyppeteer.launcher import launch

class test(object):
    def get_test(self):
        loop = asyncio.new_event_loop()
        self.loop = loop
        asyncio.set_event_loop(loop)
        loop.run_until_complete(self.main())
       # self.loop.close()

    async def main(self):
        # 使用前需要确保关闭
        url = 'https://www.baidu.com/s?wd=coroutine%20%27Launcher.killChrome%27%20was%20never%20awaited&rsv_spt=1&rsv_iqid=0xaa093ece0008eafe&issp=1&f=8&rsv_bp=1&rsv_idx=2&ie=utf-8&rqlang=cn&tn=baiduhome_pg&rsv_enter=1&rsv_t=2122Edd7Utwtaq3%2BC1k8RAJXlzih14qru1Nksz%2BhJn%2F1QqRgEI9GjSCqkLCfGY2YpbPq&oq=puppeteer%2520%25E6%25B8%2585%25E7%25A9%25BA%25E8%25BE%2593%25E5%2585%25A5&inputT=312&rsv_pq=f45dc490000b0918&rsv_sug3=31&rsv_sug1=18&rsv_sug7=100&rsv_n=2&rsv_sug2=0&rsv_sug4=745&rsv_sug=1'
        browser = await launch({'headless': True, 'args': ['--no-sandbox'],'dumpio':True}) #executablePath 服务器需要设置 否则会下载个默认的headless': True,
        self.browser = browser
        page = await browser.newPage()
        self.page = page
        await page.setUserAgent(
            'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/58.0.3029.110 Safari/537.36 Edge/16.16299')
        await page.goto(url)
        await page.type('#kw','啦啦')
        await page.waitFor(2000)
        for i in range(20):
            await page.keyboard.press('ArrowRight');
            await page.keyboard.press('Backspace');

        pass
        print(await page.content())



if __name__ == '__main__':
    test_obj = test()
    test_obj.get_test()

实际上对于更复杂的拦截我还有一种方案，就是自己写一个chrome插件，和后端脚本(我用的是python)用socket通信。我称之为js版本的webdriver。。这个插件我已经写好了，不过因为没有实际应用到所以就放着了，如果有人有兴趣的可以一起讨论下。