Unable to Access DIV element using Watir - html

Hi I am trying to access the DIV element using watir but I am unable to do that,I have tried in different ways but couldn't access it,may be I think it need to be access through some parent element can anyone help me out?
My system Configurations
IE-8
Windows 7
I tried with the below command
#ie.div(:text,'COMPOSE').click
the command gets execute with no errors but no action is performed on the UI

The best solution appears to be switching to Watir-Webdriver. With Watir-Webdriver, #ie.div(:text,'COMPOSE').click will work as expected.
Assuming that is not an option, there are a couple of reasons why that same command does not work with Watir(-Classic) v1.6.7:
The first problem is that #ie.div(:text,'COMPOSE').click will find the first div that contains this text. This would be one of the ancestors of the div you want. As a result, Watir will send the click event against the wrong element.
The second problem is that the div is not responding to the onclick event fired by Watir. I am not sure why this problem exists.
To solve the first problem, you will need to be more specific when locating the div. In this case, the "role" attribute can be used since none of the ancestor elements have this attribute. Watir-Classic does not support using the role attribute as a locator. As a result, you will need to create a custom locator using an element collection and the find method:
#ie.divs.find{ |div| div.attribute_value('role') == 'button' && div.text == 'COMPOSE' }
To solve the second problem, it turns out that double clicking does work. While newer versions of Watir-Classic have a double_click method implemented, it does not exist in 1.6.7. You can replicate the method by calling the fire_event method:
.fire_event('ondblclick')
Putting it all together, the following will click the compose button:
#ie.divs.find{ |div| div.attribute_value('role') == 'button' && div.text == 'COMPOSE' }.fire_event('ondblclick')

There may be more than one element on the page with the text 'COMPOSE', some may be hidden. Try:
#ie.divs(:text,'COMPOSE').size
That is divs with an s.
Then you can try something like the following and see if you get a change in the UI:
#ie.divs(:text,'COMPOSE').each { |b| b.fire_event('click') }
I remember that fire_event works better, but would recommend consulting the docs for the difference between .click and fire_event.

Related

Parent node in react-testing-library

The component that I have testing renders something this:
<div>Text<span>span text</span></div>
As it turns out for testing the only reliable text that I have is the 'span text' but I want to get the 'Text' part of the <div>. Using Jest and react-testing-library I can
await screen.findByText(spanText)
This returns an HTMLElement but it seems limited as I don't have any of the context around the element. For example HTML methods like parentNode and previousSibling return null or undefined. Ideally I would like to get the text content of the parent <div>. Any idea how I can do this with either Jest or react-testing-library?
A good solution for this is the closest function.
In description of closest function is written: Returns the first (starting at element) including ancestor that matches selectors, and null otherwise.
The solution would look like this:
screen.getByText("span text").closest("div")
Admittedly, Testing Library doesn't communicate clearly how to do this. It includes an eslint rule no-direct-node-access that says "Avoid direct Node access. Prefer using the methods from Testing Library". This gives the impression that TL exposes a method for a situation like this, but at the moment it does not.
It could be you don't want to use .closest(), either because your project enforces that eslint rule, or because it is not always a reliable selector. I've found two alternative ways to tackle a situation like you describe.
within():
If your element is inside another element that is selectable by a Testing Library method (like a footer or an element with unique text), you can use within() like:
within(screen.getByRole('footer')).getByText('Text');
find() within the element with a custom function:
screen.getAllByText('Text').find(div => div.innerHTML.includes('span text'));
Doesn't look the prettiest, but you can pass any JS function you want so it's very flexible and controllable.
Ps. if you use my second option depending on your TypeScript config you may need to make an undefined check before asserting on the element with Testing Library's expect(...).toBeDefined().
But I have used HTML methods a lot and there was no problem yet. What was your problem with HTML methods?
You can try this code.
const spanElement = screen.getElementByText('span text');
const parentDiv = spanElement.parentElement as HTMLElement;
within(parentDiv).getElementByText('...');

How to see if Chrome manipulates HTML by itself?

If you forget to close a HTML-Tag, Chrome will validate your code and try to fix problems like this.
I had a major problem because I forgot a closing Form-Tag, and instead of closing it correctly, Chrome deleted a following form, not the inputs, simply the Form-Tags.
When I looked at the Source Code itself, the Form-Tag was there, but not in the Elements-Tab in the console.
So at first, I thought it must have something to do with some JS deleting this DOM-Node and set a DOM-Breakpoint to find the script.
To cut a long story short, it took me hours to find out, that no JS deleted my form, but Chrome itself thought: There is a missing so I delete some other to fix that...
Is there any possibilty to see if Chrome automatically changes your DOM?
Thank You!
The browser Engine does indeed. They use string replace methods, although it happens internally.
<div>
</div>> // mistake
<div> //missing end tag
<div></div>
---------------------------------------------------
Methods
file=file.stringreplace('>>', '>')
an uneven count will add the missing div just after the next beginning div and conditionally if the missing is not found by the end of the file:
file=file.stringreplace('
<div>', '</div>
<div>')
The Parsing Engine after the missing and broken tags are repaired then parses the file and can then with a positive count set the screens GUI widgets by opening and closing tags as GUI Frames. It does this by adding tokens delimiters to the actual div tags making them easily distinguished from each other.
<div1s>
</div1e>
<div1s>//section columns
<div2s></div2e>
<div2s></div2e>
<div2s></div2e>
</div1e>
<div1s>Footer</div1e>
-----------------------------------------------------
The GUI Frame Tokens
for each "<dive1>"{
FrameCreate(CSS--ATTRIBUTES FROM ASSOCIATIVE ARRAYS--)
//the GUI Frame Widgets VERTICAL SECTIONS
}
//Next it finds the nested divs2 and embeds these into the thir parents above but with embedded Text Widgets also.
FrameTextBoxCreate(--CSS MATED ATTRIBUTES RULES--)
div3 etc------and so on.
In fact it is in the WebView GUI Widget Sets in its customized Mosaic Canvas Widget Sets in Chrome would be where they are repaired.

How to check nth-child in laravel dusk

$this->browse(function (Browser $browser) {
$browser->click('.md-button:nth-child(2)');
});
I want to click the class element 2. How to get nth-child class in laravel dusk.
Sometimes using Dusk to select certain elements can be tricky. It can be done by using the script method as a workaround which doesn't appear in the documentation as a far as I can see. You can then execute any JavaScript inside to select any element you like. jQuery makes this easy.
Try using:
$browser->script('$(".md-button:nth-child(2)").click();');
You can always do this by copying the css selector via your browser's dev tools. Where you right click on the element you want, then click inspect element, then when the developer tools window shows up, right click on the HTML element in the dev tools window, and select copy > copy selector, and then paste into your IDE where you're writing the test. This is a test I've run successfully as an example:
$this->browse(function (Browser $browser) {
$browser->visit(new UniversityCoursesListingPage)
->assertVisible('.course-list-content')
->assertVisible('.form-control')
->select('.course-filter select', 'title|desc')
->assertSeeIn('div:nth-child(1) > article > a > h3', 'Whitepaper Training')
->clickLink('My Account');}

How to use Laravel Dusk with Gmails dynamic selectors

I'm struggling to get consistent results using Laravel Dusk to send emails
via logging into Gmail and navigating to the "COMPOSE" button since the elements ID and class is being dynamically changed on page reload.
I have tried using inspect element and copying the selector with no luck.
The purpose of this exercise is just to get a better understanding of Laravel Dusk and have some fun while going about it.
My code below:
public function testGmailExample()
{
$this->browse(function (Browser $browser) {
$browser->visit('http://gmail.com')
->assertSee('Gmail')
->type('Email', 'myemail#gmail.com')
->pause(1000)
->press('#next')
->pause(1000)
->assertSee('myemail#gmail.com')
->type('#Passwd', 'myPassword')
->press('#signIn')
->waitFor('#\3a xy > div > div', 10)
->click('#\3a xy > div > div')
->pause(1000)
->type('to', 'an_excited_friends_email#gmail.com')
->type('subjectbox','Laravel Dusk is Awesome')
->click('Send')
->pause(3000);
//Only had one success with the above code, would prefer consistent results
});
}
Any help will be greatly appreciated.
You can use clickLink() for that :
$browser->clickLink('COMPOSE');
From the documentation :
To click a link, you may use the clickLink method on the browser instance. The clickLink method will click the link that has the given display text

Tracking Link Click on Google Tag Manager

I want to track clicks on the following button/link with Google Tag Manager. I created a trigger in Google Tag Manager that triggers when the element_id = 100. This works fine, except that when I click exactly on the text, it doesn't do anything, the link looks like a button, with the text in the middle of it. I can't change anything to the html or css, otherwise I can think of multiple things, so I need to find a solution without changing the html. Also, the 'myclass' class and the 'label' class get used in other elements.
<a class="myclass" id="100" href="http://www.url.com">
<span class="label">Text</span>
</a>
Anyone an idea?
Thanks a lot,
The following workaround worked:
Create trigger when element text contains "Text". This will trigger events on the button and the label on the button, of all buttons with "Text" as label.
Create tag for that trigger that checks with simple javascript if either the id of the current element = 100, which will happen when you click the button but not the label, or that the id of the parent = 100, which happens when you click the label. You can get the element that triggered the tag using the built-in variable "Click Element". Which you need to access the parent element.
Technically, you shouldn't have a CSS ID that starts with (or is) a number, so not sure if your code example is accurate or not. Whatever the case, you're probably better off using "matches CSS selector" so that you don't need to use any custom JS.
If indeed your HTML uses id="100", then the above will work. If it's anything else that doesn't start with a number, then you can use
#whatever > span