Problem/Motivation

When copying an email address with unicode character U00A0 in an email field on a node, the validation message is triggered. U00A0 is a non-breaking space character.
However, when saving the node, it bypasses validation, and the unicode character is saved in the database.

Steps to reproduce

Create an email field on a node type.
Copy an email string, containing an unicode U00A0 character

Proposed resolution

Don't proceed to the server when the client side validation fails.

CommentFileSizeAuthor
#2 roles-select-list-not-display.patch14.23 KBadityarawat_ln

Comments

jurgenr created an issue. See original summary.

adityarawat_ln’s picture

adityarawat_ln commented
Assigned: Unassigned » adityarawat_ln
Status: Active » Needs review
StatusFileSize
new roles-select-list-not-display.patch14.23 KB

Hi @jurgenr,

I reproduced the issue and investigated the client-side validation behavior.

The invalid email containing the Unicode U+00A0 (non-breaking space) was correctly detected initially. However, when submitting the form, the client-side validation process was interrupted by a JavaScript exception:

Step attribute on input type time is not supported.

The form contained a type="time" field with a step attribute, which caused the jQuery Validation library to throw an exception during full form validation. Because the validation process was interrupted, the invalid email could bypass client-side validation and be submitted.

I applied a fix to handle the step validation for unsupported input types such as time, preventing the JavaScript exception and allowing the remaining client-side validation to run correctly.

After applying the fix:

Invalid emails containing U+00A0 are prevented from being submitted.
Valid emails submit successfully.
The JavaScript exception no longer occurs.

I have created and attached a patch for review and testing. Please review the patch and let me know your feedback.