Selenium:滚动到页面末尾

时间:2015-10-13 05:24:00

标签: javascript node.js selenium selenium-webdriver

硒:

我是WebDriverJS的新手。我在Java中尝试过这种方法。

Long repaeted = 0l, scrollHeight = 0l, returnHeight = 0l;
while(true){
    if (repaeted == 0) {
        returnHeight = (Long) jse.executeScript("var scroll =document.documentElement.scrollHeight;window.scrollTo(0, scroll); return scroll;");
         System.out.println("Height : "+scrollHeight +"\t Chnage : "+returnHeight+ "\t Repeated : "+repaeted);
         scrollHeight = returnHeight;
     }else {
         returnHeight = (Long) jse.executeScript("var scroll =  document.documentElement.scrollHeight;window.scrollTo(0, scroll); return scroll;");
         System.out.println("Height : "+scrollHeight +"\t Chnage : "+returnHeight+ "\t Repeated : "+repaeted);
         if (scrollHeight.intValue() == returnHeight.intValue()) {
             System.out.println("Break.."+ returnHeight);
             break;
         } else { scrollHeight = returnHeight; }
     }
            repaeted++;
 } 

但是我在迭代循环时遇到webdriverjs的问题。

var webdriver = require('..'),
    By = webdriver.By,
    until = webdriver.until;
// make sure chromedriver can be found on your system PATH
var driver = new webdriver.Builder()
    .forBrowser('chrome')
    .withCapabilities(webdriver.Capabilities.chrome())
    .build();


driver.get('https://in.yahoo.com/').then(function(){
        var window = new webdriver.WebDriver.Window(driver);
        window.maximize();
        driver.manage().timeouts().implicitlyWait(1000 * 3);
    })
    .then(function(){
        console.log('Entered');
        var check = 0, count = 0
        for(var i = 0; i< 50; i++){
        //driver.sleep(1000 * 2);
driver.executeScript('var dynamicscroll = document.documentElement.scrollHeight;window.scrollTo(0, dynamicscroll);return dynamicscroll;').then(function(height){
        console.log('Check : '+check+'  Height : '+height +'  Repeated : '+(count++));
        if(check === 0 || check !== height){console.log('continue'); check = height; }
        else { console.log('break'); i = 100; }
            });
        }
        })
    .then(null, function(err) {
      console.error("An error was thrown! By Promise..." + err);
    });

driver.quit();

在我的代码中我已经硬编码循环迭代直到50次,我想在滚动高度到达结束时退出/中断循环。在这种方法中,我想删除像java代码这样的硬编码,因为我不知道为滚动保持动态增加的其他应用程序迭代多少次。 例如,Facebook应用程序,雅虎新闻......

3 个答案:

答案 0 :(得分:5)

滚动到动态页面的底部可能具有挑战性,具体取决于页面的实现方式。

首先,您必须使用滚动条找到容器,因为它可能与链接到window.scrollTo的容器不同。

然后通过增加scrollTop来滚动容器,直到scrollHeight变得稳定而没有待处理的请求。要检查是否存在待处理请求,请在页面具有JQuery时评估jQuery.active,或者XMLHttpRequest来监控send上的呼叫。

这是一个使用泛型函数滚动到页面底部多次或直到结束的示例:

var webdriver = require('selenium-webdriver');

var driver = new webdriver.Builder().forBrowser('chrome').build();

driver.get('https://groups.google.com/forum/#!search/webdriverjs');

 // scroll to the bottom 3 times
driver.executeAsyncScript(scrollBottom, 3)
  .then(n => console.log(`scrolled ${n} time(s)`));

 // scroll to the bottom until the end
driver.executeAsyncScript(scrollBottom)
  .then(n => console.log(`scrolled ${n} time(s)`));
function scrollBottom(){
  var count = arguments[arguments.length - 2] || 0x7fffffff;
  var callback = arguments[arguments.length - 1];

  /* get the scrollable container */
  var elm = document.elementFromPoint(window.innerWidth - 25, window.innerHeight / 2);
  for ( ;elm && (++elm.scrollTop, !elm.scrollTop); elm=elm.parentElement);
  elm = elm || document.documentElement;

  /* hook XMLHttpRequest to monitor Ajax requests */
  if (!('idle' in XMLHttpRequest)) (function(){
    var n = 0, t = Date.now(), send = XMLHttpRequest.prototype.send;
    var dispose = function(){ --n; t = Date.now(); };
    var loadend = function(){ setTimeout(dispose, 1) };
    XMLHttpRequest.idle = function() { return n > 0 ? 0 : Date.now() - t; };
    XMLHttpRequest.prototype.send = function(){
      ++n;
      this.addEventListener('loadend', loadend);
      send.apply(this, arguments);
    };
  })();

  /* scroll until steady scrollHeight or count of scroll and no pending request */
  var i = 0, scrollHeight = -1, scrollTop = -1;
  (function scroll(){
    if ((scrollHeight === elm.scrollHeight || i === count) && XMLHttpRequest.idle() > 60)
      return callback(i);
    scrollTop = elm.scrollTop;
    scrollHeight = elm.scrollHeight;
    if (i < count)
      i += (elm.scrollTop = 0x7fffffff, scrollTop !== elm.scrollTop);
    setTimeout(scroll, 100);
  })();
}

或者通过滚动直到高度在特定时间(此处为5秒)内不再增加:

function scrollBottom(){
  var count = arguments[arguments.length - 2] || 0x7fffffff;
  var callback = arguments[arguments.length - 1];
  var timeout = 5000;  /* 5 seconds timeout */
  var i = 0;

  /* get the scrollable container */
  var elm = document.elementFromPoint(window.innerWidth - 25, window.innerHeight / 2);
  for ( ;elm && (++elm.scrollTop, !elm.scrollTop); elm=elm.parentElement);
  elm = elm || document.documentElement;

  /* scroll while the height is increasing or until timeout */
  (function scroll(){
    var endtime = Date.now() + timeout;
    var height = elm.scrollHeight;
    elm.scrollTop = 0x7fffffff;  /* scroll */

    setTimeout(function check(){
      if (Date.now() > endtime)            /* returns if waited more than 5 sec */
        callback(i);
      else if (elm.scrollHeight == height) /* wait again if same height  */
        setTimeout(check, 60);
      else if (++i === count)              /* returns if scrolled the expected count */
        callback(i);
      else                                 /* scroll again */
        setTimeout(scroll, 60);
    }, 250);
  })();
}

答案 1 :(得分:1)

纯JavaScript:

在JavaScript中,我们可以使用setTimeout()函数。它将在您指定的时间延迟后递归调用指定的函数。

我测试了google群组应用,其div标签垂直滚动动态增加。要加载内容我使用了5000的时间延迟。您可以在浏览器的控制台中使用以下网址测试此代码:https://groups.google.com/forum/#!search/webdrierjs

var i = 0, height = 0, check = 0, t = null;
flow();

function run(arg){
var objDiv = document.querySelector('div.IVILX2C-b-F');
objDiv.scrollTop = objDiv.scrollHeight;
return objDiv.scrollHeight;
}

function flow() {
i++;
    switch(i){ 
        case 0:     height  = run(i);
                    sleep(5000);
                break;
        case -1:    run(i);
                    clearTimeout(t); //stops flow
                break;
        default:    check = run(i);
                    console.log('Return Height : '+check +'  Count : '+i);
                    if(check === height){ i = -2;
                    console.log('Break message : '+i);
                    }else {
                    console.log('Changed...');
                    height = check;
                    }
                sleep(5000);
                break;
    }
}

function sleep(delay) { t=setTimeout("flow()",delay);} //starts flow control again after time specified.
//function sleep(delay) {  var start = new Date().getTime();     while (new Date().getTime() < start + delay);  flow(); } // stops execution and then continues.

但即使我无法使用WebDriver / WebDriverJS运行此脚本,因为它不会在延迟时间内调用递归函数。

答案 2 :(得分:1)

根据经验,滚动到页面末尾的最快方法是查找页脚元素和movetoit,通常是#footer.footer或只是footer选择器会执行此操作。 E.g:

footer = driver.findElement({id: "footer"});
driver.executeScript("arguments[0].scrollIntoView(false);", footer);

在无休止的情况下像Facebook,Twitter等流可能会在达到限制时阻止你,所以可以递归地将最大迭代与window.scrollTo(0, 300);结合起来,并在每次滚动后等待几秒钟。