# selenium是如何启动浏览器的 前几天有同学问到selenium是怎么样启动浏览器的(selenium启动浏览器的原理)，当时稍微讲解了一下，不过自我感觉不够具体，现在特地把启动原理通过代码和一系列操作给串联起来，希望可以帮助大家更好的理解。 以chrome浏览器为例，selenium启动chrome浏览器的代码如下： [源码](https://github.com/SeleniumHQ/selenium/blob/1a99e2161f8b777672f15fda9b329ee8e89af94a/py/selenium/webdriver/chrome/webdriver.py#L33) ```python def __init__(self, executable_path="chromedriver", port=0, options=None, service_args=None, desired_capabilities=None, service_log_path=None, chrome_options=None): """ Creates a new instance of the chrome driver. Starts the service and then creates new instance of chrome driver. :Args: - executable_path - path to the executable. If the default is used it assumes the executable is in the $PATH - port - port you would like the service to run, if left as 0, a free port will be found. - desired_capabilities: Dictionary object with non-browser specific capabilities only, such as "proxy" or "loggingPref". - options: this takes an instance of ChromeOptions """ if chrome_options: warnings.warn('use options instead of chrome_options', DeprecationWarning) options = chrome_options if options is None: # desired_capabilities stays as passed in if desired_capabilities is None: desired_capabilities = self.create_options().to_capabilities() else: if desired_capabilities is None: desired_capabilities = options.to_capabilities() else: desired_capabilities.update(options.to_capabilities()) self.service = Service( executable_path, port=port, service_args=service_args, log_path=service_log_path) self.service.start() try: RemoteWebDriver.__init__( self, command_executor=ChromeRemoteConnection( remote_server_addr=self.service.service_url), desired_capabilities=desired_capabilities) except Exception: self.quit() raise self._is_remote = False ``` 其中跟浏览器启动密切相关的是这几句 ```python self.service = Service( executable_path, port=port, service_args=service_args, log_path=service_log_path) self.service.start() ``` 通过查看跟[Service](https://github.com/SeleniumHQ/selenium/blob/1a99e2161f8b777672f15fda9b329ee8e89af94a/py/selenium/webdriver/common/service.py#L35)相关的代码复盘得到启动逻辑: 调用chromedriver可执行文件运行chromedirver。这也是为什么我们需要把chromedriver放到系统PATH里的原因。 所以selenium先启动了chromedriver。当然，我们可以手工启动chromedriver来模拟这个启动过程。 在命令行中运行下面的命令```chromedirver``` 你应该可以看来类似的结果 ``` Starting ChromeDriver 2.38.552518 (183d19265345f54ce39cbb94cf81ba5f15905011) on port 9515 Only local connections are allowed. ``` 这样我们就手工启动了chromedriver。driver监听的端口是9515. 启动了driver之后，我们需要告诉driver打开浏览器。selenium的[源码](https://github.com/SeleniumHQ/selenium/blob/1a99e2161f8b777672f15fda9b329ee8e89af94a/py/selenium/webdriver/remote/webdriver.py#L224)里这一过程如下 ```python def start_session(self, capabilities, browser_profile=None): """ Creates a new session with the desired capabilities. :Args: - browser_name - The name of the browser to request. - version - Which browser version to request. - platform - Which platform to request the browser on. - javascript_enabled - Whether the new session should support JavaScript. - browser_profile - A selenium.webdriver.firefox.firefox_profile.FirefoxProfile object. Only used if Firefox is requested. """ if not isinstance(capabilities, dict): raise InvalidArgumentException("Capabilities must be a dictionary") if browser_profile: if "moz:firefoxOptions" in capabilities: capabilities["moz:firefoxOptions"]["profile"] = browser_profile.encoded else: capabilities.update({'firefox_profile': browser_profile.encoded}) w3c_caps = _make_w3c_caps(capabilities) parameters = {"capabilities": w3c_caps, "desiredCapabilities": capabilities} response = self.execute(Command.NEW_SESSION, parameters) if 'sessionId' not in response: response = response['value'] self.session_id = response['sessionId'] self.capabilities = response.get('value') # if capabilities is none we are probably speaking to # a W3C endpoint if self.capabilities is None: self.capabilities = response.get('capabilities') # Double check to see if we have a W3C Compliant browser self.w3c = response.get('status') is None self.command_executor.w3c = self.w3c ``` 这一过程的核心就是就是向```localhost:9515/session```发送1个POST请求，并发送1个json对象，默认情况下，这个对象应该是下面这个样子。 ``` { "capabilities": { "alwaysMatch": { "browserName": "chrome", "goog:chromeOptions": { "args": [], "extensions": [] }, "platformName": "any" }, "firstMatch": [ {} ] }, "desiredCapabilities": { "browserName": "chrome", "goog:chromeOptions": { "args": [], "extensions": [] }, "platform": "ANY", "version": "" } } ``` 简单理解就是告诉remote driver打开什么浏览器，上面的例子里我们打开的是chrome浏览器。 我们可以手工还原这个过程。 确保chromedriver是在运行中的，然后打开postman，构造1个POST请求，路径是localhost:9515/session。在Body里选择raw和JSON(application/json), 贴入上面的json字符串，如下图所示 ![postman](http://img.testclass.net/%E5%B1%8F%E5%B9%95%E5%BF%AB%E7%85%A7%202018-05-17%20%E4%B8%8B%E5%8D%882.02.44.png) 点击send，几秒之后chrome浏览器应该可以正常启动，并且postman的response里会有大致如下的返回值 ``` { "sessionId": "ad4407e133cfd5f3f49bff4c2f1f087a", "status": 0, "value": { "acceptInsecureCerts": false, "acceptSslCerts": false, "applicationCacheEnabled": false, "browserConnectionEnabled": false, "browserName": "chrome", "chrome": { "chromedriverVersion": "2.38.552518 (183d19265345f54ce39cbb94cf81ba5f15905011)", "userDataDir": "/var/folders/s6/f2_brc114wv2g8w0qggk_m2c0000gn/T/.org.chromium.Chromium.NMsAKJ" }, "cssSelectorsEnabled": true, "databaseEnabled": false, "handlesAlerts": true, "hasTouchScreen": false, "javascriptEnabled": true, "locationContextEnabled": true, "mobileEmulationEnabled": false, "nativeEvents": true, "networkConnectionEnabled": false, "pageLoadStrategy": "normal", "platform": "Mac OS X", "rotatable": false, "setWindowRect": true, "takesHeapSnapshot": true, "takesScreenshot": true, "unexpectedAlertBehaviour": "", "version": "66.0.3359.181", "webStorageEnabled": true } } ``` 上面的返回里最重要的就是sessionId，因为后面所有跟浏览器的交互都是基于该id进行的。 ### 总结 * selenium里，selenium client先打开chromedriver * chromedirver创建session时打开了浏览器，所以浏览器的打开跟selenium无关，完全是chromedriver的能力 ### 更多 其实上面的例子里我们手工调用了[webdriver协议](https://w3c.github.io/webdriver/)里的[new session](https://w3c.github.io/webdriver/#new-session)协议，创建了webdriver session。具体的细节大家可以参考协议了解更多。 原始封面 ![课程图片](https://images.unsplash.com/photo-1503435980610-a51f3ddfee50?w=300)