playwright反爬虫检测

爬虫检测网站

Antibot: https://bot.sannysoft.com/

• https://bot-detector.rebrowser.net/
• https://fingerprint.com/products/bot-detection/

playwright：https://playwright.net.cn/python/

正常浏览器展示如下：

使用 playwright 打开时，展示如下：

playwright cr https://bot.sannysoft.com/

可见默认情况下使用 playwright 时 WebDriver 那一栏无法通过检查。

绕过方式一：Stealth 插件

文档直达

pip install playwright-stealth

测试代码如下：

import asyncio
from playwright.async_api import async_playwright
from playwright_stealth import Stealth

async def main():
    async with Stealth().use_async(async_playwright()) as p: # 最常见用法
        browser = await p.chromium.launch(  # 启动浏览器
            headless=False, 
            proxy={"server": "socks5://127.0.0.1:7890"},
            ignore_default_args=["--enable-automation"],  # 移除自动化标记
            args=[
                "--disable-blink-features=AutomationControlled",  # 禁用自动化控制特性，隐藏自动化痕迹
                "--ignore-certificate-errors",                  # 忽略SSL证书错误，允许访问自签名证书站点
                "--disable-gpu",                                # 禁用GPU硬件加速，减少资源占用并避免部分兼容问题
                "--disable-dev-shm-usage",                      # 禁用/dev/shm共享内存，防止容器或低内存环境崩溃
                "--disable-webrtc-ip-handling",                 # 禁用WebRTC本地IP泄露，提升匿名性
                "--force-webrtc-ip-handling-policy=disable_non_proxied_udp",  # 强制WebRTC仅使用代理UDP，避免真实IP暴露
            ],
        )
        context = await browser.new_context(
            locale="zh-CN",
            timezone_id="Asia/Shanghai",
            geolocation={"longitude": 116.4074, "latitude": 39.9042},
            permissions=["geolocation"],
            ignore_https_errors=True,
            java_script_enabled=True,
            accept_downloads=False,
            viewport={"width": 1920, "height": 1080},
            user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
        )
        page = await context.new_page()
        await page.goto("https://bot.sannysoft.com/")
        print(await page.title())
        # 在这里继续操作
        await page.wait_for_timeout(10000)

        await browser.close()

asyncio.run(main())

绕过方式二：使用当前浏览器

退出浏览器，再以debug模式启动

"/Applications/Google Chrome.app/Contents/MacOS/Google Chrome" --remote-debugging-port=9222 --user-data-dir=/tmp/test/

编写代码连接：

import asyncio
from playwright.async_api import async_playwright

async def main():
    async with async_playwright() as p:
        browser = await p.chromium.connect_over_cdp("http://localhost:9222") # 连接刚才打开的浏览器
        context = await browser.new_context()
        page = await context.new_page()
        await page.goto("https://bot.sannysoft.com/")
        print(await page.title())
        # 在这里继续操作
        await page.wait_for_timeout(10000)

        await browser.close()

asyncio.run(main())

扩展

既反爬又可以保持网站登录状态

import asyncio
from playwright.async_api import async_playwright
from playwright_stealth import Stealth
import os

storagePath = "/Users/d4m1ts/.playwright/playwright_state.json" # 会话存储路径
if not os.path.exists(storagePath):
    with open(storagePath, "w", encoding="utf-8") as f:
        f.write("{}")

async def main():
    async with Stealth().use_async(async_playwright()) as p: # 最常见用法
        browser = await p.chromium.launch(  # 启动浏览器
            headless=False, 
            proxy={"server": "socks5://127.0.0.1:7890"},
            ignore_default_args=["--enable-automation"],  # 移除自动化标记
            args=[
                "--disable-blink-features=AutomationControlled",  # 禁用自动化控制特性，隐藏自动化痕迹
                "--ignore-certificate-errors",                  # 忽略SSL证书错误，允许访问自签名证书站点
                "--disable-gpu",                                # 禁用GPU硬件加速，减少资源占用并避免部分兼容问题
                "--disable-dev-shm-usage",                      # 禁用/dev/shm共享内存，防止容器或低内存环境崩溃
                "--disable-webrtc-ip-handling",                 # 禁用WebRTC本地IP泄露，提升匿名性
                "--force-webrtc-ip-handling-policy=disable_non_proxied_udp",  # 强制WebRTC仅使用代理UDP，避免真实IP暴露
            ],
        )
        context = await browser.new_context(
            storage_state=storagePath, 
            locale="zh-CN",
            timezone_id="Asia/Shanghai",
            geolocation={"longitude": 116.4074, "latitude": 39.9042},
            permissions=["geolocation"],
            ignore_https_errors=True,
            java_script_enabled=True,
            accept_downloads=False,
            viewport={"width": 1920, "height": 1080},
            user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/122.0.0.0 Safari/537.36",
        )
        context.set_default_timeout(240000) # 设置全局默认超时为 240 秒
        page = await context.new_page()
        
        await page.goto("https://markdown.gm7.org/")
        print(await page.title())

        # 保存状态到文件
        await context.storage_state(path=storagePath)
        await browser.close()

asyncio.run(main())