javascript - Selenium 网络驱动程序 : firefox headless inject javascript to modify browser property

Question

我正在尝试弄清楚如何将 selenium webdriver 与 python 或 java 结合使用来注入(inject) javascript 以修改浏览器属性/属性。我的最终目标是获得类似于 this 的东西使用 selenium 和 firefox，因为它是更开放和灵活的选择。

Puppeter 和 chromium 文件 test.js:

const puppeteer = require("puppeteer");

(async () => {
  const browser = await puppeteer.launch({
    args: ["--no-sandbox"],
    headless: true,
  });
  const page = await browser.newPage();
  const fs = require("fs");

  // In your puppeteer script, assuming the javascriptChromium.js file is in same folder of our script
  const preloadFile = fs.readFileSync("./javascriptChromium.js", "utf8");
  await page.evaluateOnNewDocument(preloadFile);

  const testUrl="https://intoli.com/blog/not-possible-to-block-chrome-headless/chrome-headless-test.html";

  await page.goto(testUrl);

  // save screenshot
  await page.screenshot({path: "puppeteer-chromium-async-script-test.png"});
  await browser.close()
})();

Javascript 文件 javascriptChromium.js

// overwrite the `languages` property to use a custom getter
Object.defineProperty(navigator, "languages", {
  get: function() {
    return ["en-US", "en", "es"];
  }
});

// Overwrite the `plugins` property to use a custom getter.
Object.defineProperty(navigator, 'plugins', {
  get: () => [1, 2, 3, 4, 5],
});

// Pass the Webdriver test
Object.defineProperty(navigator, 'webdriver', {
  get: () => false,
});

此代码运行良好，我检查属性是否已通过此 test Web site 更改.

现在，selenium 和 firefox:

import os
from selenium import webdriver

def readJSFile(scriptFile):
    with open(scriptFile, 'r') as fileHandle:  
        script=fileHandle.read()
    return script
injectedJavascript=readJSFile("./javascriptFirefox.js")

options=webdriver.FirefoxOptions()
options.set_headless(True)
driver=webdriver.Firefox(options=options)
driver.set_script_timeout(3)

# inject JavaScript
try:
    driver.execute_async_script(injectedJavascript)
except:
    print("Timeout")

# solution found here https://stackoverflow.com/questions/17385779/how-do-i-load-a-javascript-file-into-the-dom-using-selenium
driver.execute_script("var s=window.document.createElement('script'); s.src='javascriptFirefox.js';window.document.head.appendChild(s);")
testUrl="https://intoli.com/blog/not-possible-to-block-chrome-headless/chrome-headless-test.html";
driver.get(testUrl)

# example sync script
time=driver.execute_script("return performance.timing.loadEventEnd - performance.timing.navigationStart;")
print(time)
# example async script
time=driver.execute_async_script("var callback = arguments[arguments.length-1]; const time = () => { total=performance.timing.loadEventEnd - performance.timing.navigationStart; callback(total); }; time();")
print(time)

file="selenium-firefox-async-script-test.png"
driver.save_screenshot(file)

driver.quit()

Javascript 文件 javascriptFirefox.js

// overwrite the `languages` property to use a custom getter
const setProperty = () => {
    Object.defineProperty(navigator, "languages", {
        get: function() {
            return ["en-US", "en", "es"];
        }
    });

    // Overwrite the `plugins` property to use a custom getter.
    Object.defineProperty(navigator, 'plugins', {
        get: () => [1, 2, 3, 4, 5],
    });

    // Pass the Webdriver test
    Object.defineProperty(navigator, 'webdriver', {
      get: () => false,
    });
    callback();
};
setProperty();

我是 javascript 的新手，但这两种方法(puppeteer 和 selenium)之间的不同之处在于它们如何管理当前选项卡/页面。前者通过页面类和方法page.evaluateOnNewDocument而对于后者，我没有找到等效的方法。我也尝试过使用 greasemonkey 或 violentlmonkey 来注入(inject) javascript 但没有成功。

你有什么建议吗？

谢谢

Answer 1

我按照这个 post 找到了问题的解决方案.简而言之，通过使用扩展，可以将 javascript 代码注入(inject)到 Firefox 的网页中。为了避免其他用户浪费时间，主要文件有:

Python文件:selenium+firefox

import json
import os
import sys

from selenium import webdriver
from selenium.common.exceptions import NoSuchElementException
from selenium.webdriver.firefox.firefox_profile import AddonFormatError

# Patch in support for WebExtensions in Firefox.
# See: https://intoli.com/blog/firefox-extensions-with-selenium/
class FirefoxProfileWithWebExtensionSupport(webdriver.FirefoxProfile):
    def _addon_details(self, addon_path):
        try:
            return super()._addon_details(addon_path)
        except AddonFormatError:
            try:
                with open(os.path.join(addon_path, "manifest.json"), "r") as f:
                    manifest = json.load(f)
                    return {
                        "id": manifest["applications"]["gecko"]["id"],
                        "version": manifest["version"],
                        "name": manifest["name"],
                        "unpack": False,
                    }
            except (IOError, KeyError) as e:
                raise AddonFormatError(str(e), sys.exc_info()[2])

profile_folder="profile_path"
profile=FirefoxProfileWithWebExtensionSupport(profile_folder)
extension_directory="extension"
profile.add_extension(extension_directory)
# firefox dev it is necessary for custom profile, not for standard one
firefox_binary="/usr/bin/firefox-dev"
options=webdriver.FirefoxOptions()
# firefox 56+ headless mode https://developer.mozilla.org/en-US/Firefox/Headless_mode
options.set_headless(True)
driver=webdriver.Firefox(options=options, firefox_profile=profile, firefox_binary=firefox_binary)

test_url="https://intoli.com/blog/not-possible-to-block-chrome-headless/chrome-headless-test.html";
driver.get(test_url)

file="selenium-firefox-extension-profile-script-second-test.png"
driver.save_screenshot(file)

test_url="https://intoli.com/blog/making-chrome-headless-undetectable/chrome-headless-test.html";
driver.get(test_url)

file="selenium-firefox-extension-profile-script-first-test.png"
driver.save_screenshot(file)

driver.quit()

扩展文件:manifest.js 和 content.js

{
  "manifest_version": 2,
  "name": "Smart Extension",
  "version": "1.0.0",
  "applications": {
    "gecko": {
      "id": "user@protonmail.com"
    }
  },
  "content_scripts": [
    {
      "matches": ["*://*/*"],
      "js": ["content.js"],
      "run_at": "document_start"
    }
  ]
}

var script=document.createElement("script");
script.src=browser.extension.getURL("myscript.js");
script.async=false;
document.documentElement.appendChild(script);

Javascript 文件:myscript.js

// overwrite the `languages` property to use a custom getter
Object.defineProperty(navigator, "languages", {
  get: function() {
    return ["en", "es"];
  }
});

// Overwrite the `plugins` property to use a custom getter.
Object.defineProperty(navigator, "plugins", {
  get: () => new Array(Math.floor(Math.random() * 6) + 1),
});

// Pass the Webdriver test
Object.defineProperty(navigator, "webdriver", {
  get: () => false,
});

// hairline: store the existing descriptor
const elementDescriptor=Object.getOwnPropertyDescriptor(HTMLElement.prototype, "offsetHeight");

// redefine the property with a patched descriptor
Object.defineProperty(HTMLDivElement.prototype, "offsetHeight", {
    ...elementDescriptor,
  get: function() {
    if (this.id === "modernizr") {
      return 1;
    }
    return elementDescriptor.get.apply(this);
  },
});

["height", "width"].forEach(property => {
  // store the existing descriptor
  const imageDescriptor=Object.getOwnPropertyDescriptor(HTMLImageElement.prototype, property);

  // redefine the property with a patched descriptor
  Object.defineProperty(HTMLImageElement.prototype, property, {
    ...imageDescriptor,
    get: function() {
      // return an arbitrary non-zero dimension if the image failed to load
      if (this.complete && this.naturalHeight == 0) {
        return 24;
      }
      // otherwise, return the actual dimension
      return imageDescriptor.get.apply(this);
    },
  });
});

const getParameter=WebGLRenderingContext.getParameter;
WebGLRenderingContext.prototype.getParameter=function(parameter) {
  // UNMASKED_VENDOR_WEBGL WebGLRenderingContext.prototype.VENDOR
  if (parameter === 37445) {
    return "Intel Open Source Technology Center";
  }
  // UNMASKED_RENDERER_WEBGL WebGLRenderingContext.prototype.RENDERER
  if (parameter === 37446) { 
    return "Mesa DRI Intel(R) Ivybridge Mobile";
  }
  return getParameter(parameter);
};

这适用于图形模式下的所有测试，而在 headless 模式下，除了似乎影响 bug 的 WebGL 测试之外的所有测试。 .