Bug 278687 – InvalidStateError is triggered for collapseToEnd() when switching the content to display:none

• Status: NEW
• Component: HTML Editing

[Karl Dubost]

Created attachment 472307 [details]
testcase

1. Load the attachment
2. Check the console.log for Safari, Firefox, Chrome

Safari is sending InvalidStateError while other browsers

Code from the testcase:

    <div id="test">
        <div contenteditable="true">
            <p>Some text</p>
        </div>
    </div>

    <p>Look at the console in Safari, Firefox and Chrome.</p>
    <p>Only Safari reports: <code>InvalidStateError: The object is in an invalid state.</code> for
        <code>collapseToEnd()</code></p>

    <script>
        function trigger_invalid_state(text_node) {
            text_node.parentNode.focus();
            var range = document.createRange();
            range.selectNode(text_node);
            range.setStart(text_node, 0);
            range.setEnd(text_node, 2);

            var sel = window.getSelection();
            sel.removeAllRanges();
            sel.addRange(range);
            window.getSelection().collapseToEnd();
        }


        var p_visible = document.querySelector('#test p');
        var text_node = p_visible.childNodes[0]
        console.log("visible: ", text_node);

        // no error
        trigger_invalid_state(text_node);

        // change it to display: none
        p_visible.parentNode.parentNode.style.display = "none";

        // call again
        trigger_invalid_state(text_node);

    </script>

I wonder if it's about focus on a display: none element

[Radar WebKit Bug Importer]

<rdar://problem/134744854>

[John A. Bilicki III]

Hi Karl,

I would technically argue that I would rather become aware of the issue than not. Though at the same time a good question would be: is this error message necessary in this instance? I can't commit to getting too involved right now though I just wanted to throw these thoughts in real quick since you had the courtesy to mention this in my bug report 278290, thank you.

[Karl Dubost]

Hi John,

This bug is about understanding the difference of behavior in between Safari on one side and Firefox/Chrome on the other side.

And help us determine which browsers act in accordance with the specifications and/or what makes sense.

[Ahmad Saleem]

Here - https://searchfox.org/wubkat/rev/b36cbce69fddb7da33823f316bd8ead5bebee970/Source/WebCore/page/DOMSelection.cpp#249

We can update our error message, if we want.

return Exception { ExceptionCode::InvalidStateError };

to

return Exception { ExceptionCode::InvalidStateError, "something something"_s };

__

Meanwhile I found something in Blink (closely touching or mentioning potential issue or could be totally different issue) - https://issues.chromium.org/issues/40277798

[Karl Dubost]

Ahmad, the bug to fix the error message is Bug 278290.
Not this one. :)