I'm writing a scraper with puppeteer and I need to fill the fields of a form built with Aura js. The problem is I'm not able to reach the input and password fields because they're inside a custom element:
<lightning-input class="input sfdc_usernameinput slds-form-element" data-aura-rendered-by="116:0" lightning-input_input-host=""><span lightning-input_input="" data-aria="" class="slds-assistive-text"></span><label lightning-input_input="" for="input-5" class="slds-form-element__label slds-no-flex slds-assistive-text">Username</label>
<div lightning-input_input="" class="slds-form-element__control slds-grow">
<input lightning-input_input="" type="text" id="input-5" aria-invalid="false" placeholder="Username" class="slds-input">
</div>
</lightning-input>
And when I try:
await page.type('#input-5', 'myusername');
it throws the error:
Error: No node found for selector: #input-5
This is confirmed if I try document.getElementById('#input-5') in browser console.
So how do I get that input?
EDIT: Adding a screenshot that shows what happens here: the markup with children elements but no way to access to them.
The element you're looking for are encapsulated in the shadow DOM thus making them hidden from the normal DOM query methods.
You can still query them like this:
document.querySelector("lightning-badge").shadowRoot.querySelector('#input-5')
Example: on this Aura demo page I can get text of a lightning badge with this code:
document.querySelector("lightning-badge").shadowRoot.textContent
"Start Position"
This query can of course can be executed in a page.evaluate method with puppeteer as well.