I'm running a Test Project in visual studios that is running test steps for a web page.
The web page itself is coded poorly. it has something like this:
<a id ="GenericButton" </a>
<div id="UniqueID" class="ThisDiv">
<a id ="GenericButton" </a>
</div>
The issue that I'm having is that I need to click the link inside the div "UniqueID". I don't want to click the first web button.
Using the code:
HtmlHyperlink myLink = new HtmlHyperlink(this.IE);
myLink.SearchProperties[HtmlHyperlink.PropertyNames.Id] = "GenericButton";
Mouse.Click(myLink)
Will result in clicking the first button instead of the button inside of the div "UniqueID"
I can get the div using:
HtmlDiv myDiv = new HtmlDiv(this.IE);
myDiv.SearchProperties[HtmlCheckBox.PropertyNames.Class] = "ThisDiv";
myDiv.SearchProperties[HtmlCheckBox.PropertyNames.Id] = "UniqueID";
//The below line is later referenced
int myDivTag = myDiv.TagInstance;
But how can I then capture the a frame inside of this div?
Furthermore, when debugging, the object myDiv doesn't seem to properly set itself in Visual Studios until running the line "int myDivTag = myDiv.TagInstance;". Why doesn't the debugger know which object I am referring to until after that happens?
I am using Visual Studio libraries for these operations instead of something like WatiN.
If the web page is poorly code, then the tests for the web page should fail for the bits that have been implemented poorly - simple.
In HTML, two elements should not have the same ID. So write your tests expecting only the single element. Since the input markup to your test will provide two elements with the same ID, the test should fail and remain failed until the problem is actually fixed on the other end.
There's no point hacking your tests to work with bad implements as it defeats the point of testing.
Update: I have found this answer which may give a solution to finding the nested a element.
Adrian gave the helpful link to the fix: How to handle testing an HTML control that lacks sufficiently unique attributes?
HtmlDiv myDiv = new HtmlDiv(this.IE);
myDiv.SearchProperties[HtmlDiv.PropertyNames.Class] = "ThisDiv";
myDiv.SearchProperties[HtmlDiv.PropertyNames.Id] = "UniqueID";
HtmlHyperlink myLink = new HtmlHyperlink(myDiv);
myLink.SearchProperties[HtmlHyperlink.PropertyNames.id] = "GenericButton";
Mouse.Click(myLink);
Related
I know there are existing questions on this topic, but none of them seems to help me with this:
I've got a lightbox with several elements.
I can find and access all of these elements, except ONE, using the XPath.
These are the items:
Text header: No problem
Text: No problem
Input field: No problem
Text: No problem
Text: No problem
Button (upload file): THIS IS SEEMINGLY IMPOSSIBLE FOR Selenium TO FIND
Button (cancel): No problem
Button (send): No problem
The XPaths for all the elements:
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[1] /content-placeholder/h1
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/ul[1]/li[1]/label/span
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/ul[1]/li[1]/div/div/input
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/label
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/span
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/input
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[5]/content-placeholder/button[1]
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[5]/content-placeholder/button[2]
The problematic element is this:
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/input
As far as I can see, there's no reason why it should be different from the other elements (textfield, button, text)?
I'm accessing all these elements with an implicit wait, to check that they've all loaded before continuing.
GCDriver.WaitForVisible("//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/input");
From the GCDriver (Selenium Driver) class:
public static void WaitForVisible (string xpath) {
var wait = new WebDriverWait(GCDriver.Instance,
TimeSpan.FromSeconds(10));
wait.Until(driver =>
driver.FindElement(By.XPath(xpath)).Displayed);
}
Now, as mentioned, this works for all the other elements, as well as accessing them directly. For this, the wait times out with WebDriverTimeoutEsception:
Result Message:
Test method Tests.Regression_tests.VerifyOverlays.Verify_Update_Ticket_OverlayContent threw exception:
OpenQA.Selenium.WebDriverTimeoutException: Timed out after 10 seconds
Also, of course, trying to ACCESS the button with .Click() also fails:
GCDriver.Instance.FindElement(By.XPath(".//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/input")).Click();
Result Message:
Test method Tests.Regression_tests.VerifyOverlays.Verify_Update_Ticket_OverlayContent threw exception:
System.InvalidOperationException: unknown error: Element is not clickable at point (-208, 307)
Here's the html code for the element:
<a class="btn btn-grey file-input-container btn-small" data-bind="enable: !uploading() "
style="margin-top: 10px; padding: 7px 12px; "data-tooltipped=""
aria-describedby="tippy-tooltip-32"
data-original-title="Add Attachment">
<i class="fa fa-cloud-upload"/>
<span class="mq-for-small-hide">
<span localize-me="">Add Attachment</span>
</span>
<input data-bind="upload: addAttachments,
enable: !uploading()" type="file"/>
</a>
I've tried some other ways of getting the element, but since this is quite (imo) "messy" html, with no unique ID's or good class names, I've been unable to figure out how.
And it REALLY bugs me that I cannot find it by the XPath. There are 8 elements on the page, all visible and accessible, but this ONE element is impossible to find with Selenium.
The element is there; I can manually click the button on the page while Selenium runs it.
UPDATE:
I also tried using .Enabled instead of .Displayed. Same result.
UPDATE 2:
There are two answers below, and I have to select one as the "winner".
Shubham Jain gives an answer that, while not the exact thing I was trying to to, is a very good work-around. By using JavaScriptExecutor to try clicking the button, it also checks if the button is visible. However, the answer given doesn't do what it tries to do; Clicking doesn't work quite that way. See Solution below to see the correct/working code to click a button using JavaScriptExectutor.
smit9234's answer is exactly what I'm trying to do, although clicking doesn't work that way. To click the button, JS is necessary in this case. However, the question was how to check .Displayed, and that works with the modified XPath he gave me from the code excerpt.
Solution
The XPath of the element (button) is, according to FirePath:
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/input
This, however, doesn't work. Selenium simply cannot find it, even though it's clearly there.
THIS XPath, however, does work:
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/span/span
However, it works with reagards to the .Displayed check. It does NOT work with Click(). To be able to click the button, I began with Shubham Jain's code example and created this method in the Driver class, to be able to use JavaScript (with Selenium's JavaScriptExecutor) to click the button:
using OpenQA.Selenium.Interactions;
public static void JSClick (string xpath) {
IWebElement icon = Instance.FindElement(By.XPath(xpath));
Actions ob = new Actions(Instance);
ob.Click(icon);
IAction action = ob.Build();
action.Perform();
}
Looking at the html snippet you posted, it seems like this is a file attachment function. Based on the html structure of the snippet, try using the following xpath:
.//*[#id='overlays']/overlay--master/div/div/overlay-lightbox/div/div[3]/content-placeholder/a/span/span
You should then be able to use the click(); method to click the "Add Attachments"
I assume that clicking on the input doesn't do anything, however you should be able to use the sendKeys(); method for sending the "file path" to the input element.
Use below XPath :-
//input[#type='file' and contains(#data-bind,'upload: addAttachments')]
You can use javascriptexecutorof selenium to click on button. It operated directly on JS of page.
In java :-
WebElement element = driver.findElement(By.id("gbqfd"));
JavascriptExecutor executor = (JavascriptExecutor)driver;
executor.executeScript("arguments[0].click();", element);
I don't know more about c# but I believe it something like
IWebElement clicks = driver.FindElement(By.Id("gbqfq"));
IJavaScriptExecutor js = (IJavaScriptExecutor)driver;
js.ExecuteScript("arguments[0].click();", clicks);
Change the locator in above elements as per your convenience.
Below you will find more details of javascriptexecutor
https://seleniumhq.github.io/selenium/docs/api/java/org/openqa/selenium/JavascriptExecutor.html
Hope it will help you :)
May be input element is not visible on the page. you may not use displayed function for that element and try with enabled as given below.
public static void WaitForEnabled (string xpath) {
var wait = new WebDriverWait(GCDriver.Instance,
TimeSpan.FromSeconds(10));
wait.Until(driver =>
driver.FindElement(By.XPath(xpath)).Enabled);
}
if the above is not working, you try to click on anchor tag instead of input.
It seems like it's not clickable. It looks like there's some javascript on the page with a function called "uploading()".
since you're button has this on it
enable: !uploading()
just a test to verify if this is actually the cause, put a breakpoint before your click. on the browser dev tools stick a breakpoint in the uploading() function on the javascript file and see what it's returning.
If this is the case you'll have to use the javascript executor to bypass this.
I'm using Selenium from C# to try and periodically do a search on a website and pull in updated data. Unfortunately the website needs a login I can't make public so I can't actually give any kind of reproducible test case.
Using the Selenium IDE I've made a script that successfully logs in and pulls the data, so the task is possible. However when exporting that script to C#, there's a failure when trying to use an iframe that appears during the process.
By reading around I've discovered I need to add code to explicitly find and switch to the iframe, which I've done;
driver.SwitchTo().Frame(driver.FindElement(By.Id("popup_iframe")));
Inside that frame is an area to the left with items to click. Each item when clicked changes the visible controls to the right. When I break at this point and click manually, everything works as expected. If I call the .Click on the IWebElement representing the object, nothing happens so the next step in the program fails because it gets another object, tries to click it and an exception's thrown because the object isn't visible.
I'm leaving 5s delays between operations in case delays are needed.
The code on the website in this area is;
<div id="ctl00_popup_workarea_left_container" class="popup-workarea-left-container" style="overflow: auto; height: 337px; display: block;">
<a id="LBTN1" class="navbutton navbutton_Disabled_Selected">Basic Search</a>
<a id="LBTN2" class="navbutton">Advanced Search</a>
<a id="LBTN3" class="navbutton">Permit Holder Search</a>
<a id="LBTN4" class="navbutton">Assessor Search</a>
</div>
The code I'm using to get and click the IWebElement is
driver.FindElement(By.Id("LBTN3")).Click();
I've also tried
Actions act = new Actions(driver);
act.MoveToElement(driver.FindElement(By.Id("LBTN3"))).Click().Perform();
Aside from the fact these don't throw exceptions, I've confirmed I'm getting an IWebElement interactively;
driver.FindElement(By.Id("LBTN3"))
{Element (id = f42435b5-139c-4334-be12-4a1f9f48221f)}
Coordinates: {OpenQA.Selenium.Remote.RemoteCoordinates}
Displayed: true
Enabled: true
Id: "f42435b5-139c-4334-be12-4a1f9f48221f"
Location: {X = 0 Y = 112}
LocationOnScreenOnceScrolledIntoView: {X = 0 Y = 112}
Selected: false
Size: {Width = 200 Height = 29}
TagName: "a"
Text: "Permit Holder Search"
WrappedDriver: {OpenQA.Selenium.Firefox.FirefoxDriver}
So I'm getting the element. The HTML appears to show that it really is the element- I'm not accidentally using the ID of something containing it. The original script created in the Selenium IDE works fine- I just can't automate it so it kicks off unattended in the early hours. I know this is vague because I can't give people access to play to reproduce it, but does anyone have an idea of what could be going wrong and how to fix?
For completeness I thought I'd better put an answer here.
I never identified why none of the attempts I made worked, but I discovered that injecting Javascript into the page to click the element I wanted to click did work. So assuming .FindElement() is able to get the element to click, the following snippet worked for me in a number of places where IWebElement.Click() proved uncooperative.
IWebElement ibtn3 = driver.FindElement(By.Id("LBTN3"));
IJavaScriptExecutor executor = (IJavaScriptExecutor)driver;
executor.ExecuteScript("arguments[0].click();", ibtn3);
best to show this by code example:
html
<div runat="server" id="PI"> </div>
c#
protected void addNewProject_Click(object sender, EventArgs e)
{
PI.attributes.add("z-index", "0");
}
basically gives me an error saying "the name PI does not exist int he current context."
Any idea why it says that?
Make sure that your .designer.* file has been updated with a variable for PI. Some source control systems lock the file and prevent Visual Studio from automatically updating the file.
If the variable hasn't been created for you, you can always go back to your .designer file and add the new variable yourself. Just follow the patter for your other ASP.NET controls.
You're looking for a partial class with the same name as your code behind class.
This SO question addresses your issue as well.
I found an alternative solution, instead of making the image switch z-indexes, I am making the asp.net panel switch z-indexes. For some reason when I try to declare my object in designer.cs file...it registers in code behind. This is good, however right when I test it out on a web-browser, the designer file auto-regenerates and gets rid of it again. So instead of playing around with the html controls I played around with what I know works and used the asp.net panel controls.
I always declare them in my codebehind file, something like this:
var myDiv = new HtmlGenericControl("div");
Then you just treat it the same as any other control on your page.
I have a C# Windows Forms Application with a webBrowser within it. It visits a Wikia chatroom, where it then locates the element which contains the chat output. This element looks like:
<div style="" id="Chat_XXXXX" class="Chat">
<ul>
(chat text)
</ul>
</div>
The purpose of the program is to retrieve the chat text every once in a while for logging purposes. This is easily done by parsing the "InnerHtml" of the "Chat_XXXXX" element. However, I also need to clear the text from the window when I do this (for various reasons, I cannot leave the text in the window). I figured I would just erase the chat text portion of the element, as this is how it is done with a handy javascript file called "chat hacks" for Wiki chat (here). Or at least, I think that's how it does it. If you look at the function "clearWindow" in that file, you can see what it does:
NodeChatController.prototype.clearWindow = function() {
this.viewDiscussion.chatUL.html('');
this.inlineAlert(i18n['cleared']);
}
I have tried setting the InnerHtml of "Chat_XXXXX" using the following three strings (not all at the same time, of course):
HtmlDocument document = webBrowser1.Document;
document.GetElementById("Chat_XXXXX").InnerHtml = ""
document.GetElementById("Chat_XXXXX").InnerHtml = "<ul></ul>"
document.GetElementById("Chat_XXXXX").InnerHtml = "<ul><li class=\"inline-alert\"> Window cleared. </li></ul>"
However, although these clear the window (and in the case of the last one prints a message), the chat no longer updates as new messages show up. The only fix is to reload the page, which isn't an option, because reloading the page brings in a whole load of chat history (which I'm trying to avoid). I've also tried importing that javascript mentioned above into the page using:
HtmlElement head = webBrowser1.Document.GetElementsByTagName("head")[0];
HtmlElement scriptEl = webBrowser1.Document.CreateElement("script");
scriptEl.SetAttribute("type", "text/javascript");
scriptEl.SetAttribute("src", "https://db.tt/66q8UQbY");
head.AppendChild(scriptEl);
This javascript creates a button which clears the chat window. The button clears the window just fine, but again, the chat no longer updates. I know this button works correctly without stopping further incoming chat in "regular" browsers (Chrome, Firefox, Opera, etc). I've used it many times, and the script itself is quite popular in the Wikia community. I've already followed the steps here: "WPF WebBrowser Control - position:fixed Element jumps while scrolling (Windows 8)" to get the browser to act as close as it can (?) to Internet Explorer. I've already checked the Body field of the document after altering the InnerHtml to make sure that my replacements didn't alter anything important. Just for clarity, here's an example of what is contained in the (chat text) portion of my original example:
<li class="you" data-user="UserNameHere" id="entry-c812">
...avatars and junk...
<span class="message">hello</span>
</li>
I honestly have no idea what could be causing the chat element to stop updating after it has been edited (especially since it works outside of this program), so I don't know what information to include. Whatever you need, I'll provide it. Here's the javascript from Wikia which generates the chat output window: chat_js2. Look for "Chat_" to find the part which originally generates the window. I don't know where the output is updated in that file though.
Without posting pages of C# code and markup, has anyone got a reason why this code
var link = _driver.FindElement(By.Id(field + "Field"));
var id = link.GetAttribute("id");
var text = link.Text;
given this markup
<a id="ForenameField" href="/MyUrl/MyFolder/MyId">3 errors</a>
Assigns an empty string to the text variable, but if I put a breakpoint on the second or third line and inspect the link variable, I can see the inner text of the element against the Text property on the inspector, it reads "3 errors", but the value of text is an empty string. It is not hidden, I can see the text if I add a watch or use quickview, any ideas?
Ok, it's my bad. Using jquery to toggle class on the div that contains the html in the question, meant that although users see the div appearing, the class that hides the div is still in the tag. A bit like this
<div class="hideThis showThis"><!-- my elements /--></div>
This makes it so that Selenium is right not give me a text value. It is strange however that the Visual studio debugger thinks that there should be a value. Visual Studio seems to go with what I can see, but Selenium is more pedantic about the hideThis class being there.
I go with the idea that if you can't see it you can't interact with it, so it is worth looking up the html graph from the element you expect to have a value to see if any class is present which would hide your element.
Feel free to recommend that I delete this rather obvious wisdom.
I know this was posted over a year ago, but I had this exact problem too and came across this thread. I was able to solve it by just waiting for the DOM to load--some elements aren't visible until the DOM is updated. So just putting Thread.Sleep(6000) or whatever after navigating to the page got it to work for me.