Add checks and explicit errors when cart buttons are missing data

[tomalec]

Changes proposed in this Pull Request:

Add checks and explicit errors when cart buttons are missing data
to help debugging issues with other extensions manipulating WC Core add & remove from cart buttons or events.

This would help identifying issues like the one reported at https://wordpress.org/support/topic/js-error-on-add-to-cart-when-logged-out/

Checks:

• Does your code follow the WordPress coding standards?
• Have you written new tests for your changes, as applicable?
• Have you successfully run tests with your changes locally?
• Have you checked to ensure there aren't other open Pull Requests for the same update/change?

Screenshots:

Detailed test instructions:

1. Build the branch
2. open shop page
3. Run synthetic/malformed cart events

jQuery( document.body ).trigger( 'added_to_cart' );
jQuery( document.body ).trigger( 'removed_from_cart' );

8. Check for errors in the console
9. Add an item to the cart using UI.
10. Remove all product ids document.querySelectorAll('[data-product_id]').forEach((e)=>e.removeAttribute('data-product_id')). 8. Remove item from mini cart using UI
11. Check for errors in the console
12. Smoke test not tempered events and buttons for adding to & removing from cart, for regression

Additional details:

Changelog entry

Tweak - Add checks and explicit errors when cart buttons are missing data.

[eason9487]

Thanks for the improvement. LGTM.

Commented on two minor concerns, and I believe it won't need another round even if there are related changes.

[eason9487]

If the purpose is to help debugging, perhaps it also needs to consider the case where the button parameter is an empty array/jQuery collection.

jQuery( document.body ).trigger( 'added_to_cart', [ , , jQuery('match-nothing') ] );

[eason9487]

Although parseInt won't return a type other than number or NaN, it would be better to avoid using isNaN when possible because it has some confusing special-case behavior that leads to semantics ambiguity.

Maybe it could be replaced with Number.isNaN.

[tomalec]

Thanks @eason9487 for the detailed review :)