Bug 99567 – Web Inspector: Shadow DOM: Node removal doesn't reflect

• Status: RESOLVED
• Resolution: FIXED
• Component: Web Inspector (Deprecated)

[Kent Tamura]

Created attachment 169129 [details]
Screenshot

Environment:
  Google Chrome 24.0.1298.0 canary

Steps to reproduce
0. Enable "Show Shadow DOM" in "General" tab of DevTools setting
1. Open data:text/html,<input type=time>
2. "Inspect Element" on the time field
3. In the Elements panel, expand all of shadow nodes of the time field.  You'll see two or three <span>s with some aria attributes.
4. Make sure selecting the time field in the Elements panel
5. Open JavaScript console
6. Type   $0.value = "23:59";
7. Observe the shadow nodes

Expected:
 Shadow nodes in the Elements panel are replaced for the value "23:59"

Actual:
 Shadow nodes for "23:59" are appended in the Elements panel.  See attached screenshot.

[Shinya Kawanaka]

I'm taking a look at this.

[Shinya Kawanaka]

Created attachment 169327 [details]
Patch

[Kent Tamura]

Comment on attachment 169327 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=169327&action=review

> LayoutTests/inspector/elements/update-time-value-with-shadow.html:39
> +    <input id="input" type="time" value="14:00" />

input[type=time] with this structure is available only on desktop Chromium for now. We had better test with an element which is available on any platforms. e.g.  Prepare <input type=text>, then change the type like input.type = "checkbox"

[Shinya Kawanaka]

(In reply to comment #3)
> (From update of attachment 169327 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=169327&action=review
> 
> > LayoutTests/inspector/elements/update-time-value-with-shadow.html:39
> > +    <input id="input" type="time" value="14:00" />
> 
> input[type=time] with this structure is available only on desktop Chromium for now. We had better test with an element which is available on any platforms. e.g.  Prepare <input type=text>, then change the type like input.type = "checkbox"

OK. I'll change the test case soon.

[Shinya Kawanaka]

(In reply to comment #4)
> (In reply to comment #3)
> > (From update of attachment 169327 [details] [details])
> > View in context: https://bugs.webkit.org/attachment.cgi?id=169327&action=review
> > 
> > > LayoutTests/inspector/elements/update-time-value-with-shadow.html:39
> > > +    <input id="input" type="time" value="14:00" />
> > 
> > input[type=time] with this structure is available only on desktop Chromium for now. We had better test with an element which is available on any platforms. e.g.  Prepare <input type=text>, then change the type like input.type = "checkbox"
> 
> OK. I'll change the test case soon.

It turned out that this issue does not happen when we change the type.
Hmm...

[Shinya Kawanaka]

Is it OK if I don't use <input />? This issue happens even if we don't use <input />.
I'll just use #shadow-root and appendChild/removeChild.

[Hajime Morrita]

(In reply to comment #6)
> Is it OK if I don't use <input />? This issue happens even if we don't use <input />.
> I'll just use #shadow-root and appendChild/removeChild.
This sounds even better. It clarifies what the problem is.

[Shinya Kawanaka]

Created attachment 169335 [details]
Patch

[Build Bot]

Comment on attachment 169335 [details]
Patch

Attachment 169335 [details] did not pass mac-ews (mac):
Output: http://queues.webkit.org/results/14391875

New failing tests:
inspector/elements/update-shadowdom.html

[Shinya Kawanaka]

Ah, we have to skip mac...

[Shinya Kawanaka]

Created attachment 169349 [details]
Patch

[Pavel Feldman]

Comment on attachment 169349 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=169349&action=review

> LayoutTests/inspector/elements/update-shadowdom.html:9
> +    shadowRoot.appendChild(document.createElement('span'));

Do you want to capture both states? I.e. that the addition was reflected in the inspector as well?

[Shinya Kawanaka]

Created attachment 169529 [details]
Patch

[Shinya Kawanaka]

> Do you want to capture both states? I.e. that the addition was reflected in the inspector as well?

Ah... It's my mistake. Removed.

[WebKit Review Bot]

Comment on attachment 169529 [details]
Patch

Rejecting attachment 169529 [details] from commit-queue.

Failed to run "['/mnt/git/webkit-commit-queue/Tools/Scripts/webkit-patch', '--status-host=queues.webkit.org', '-..." exit_code: 1

ERROR: /mnt/git/webkit-commit-queue/Source/WebCore/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive).

Full output: http://queues.webkit.org/results/14466692

[Shinya Kawanaka]

Created attachment 169820 [details]
Patch for landing

[WebKit Review Bot]

Comment on attachment 169820 [details]
Patch for landing

Clearing flags on attachment: 169820

Committed r132024: <http://trac.webkit.org/changeset/132024>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.