Web Inspector: Script ProfileViews should be searchable

[Joseph Pecoraro]

* SUMMARY
Script ProfileViews should be filterable

Should be able to search for a particular function in a ProfileView.

[Radar WebKit Bug Importer]

<rdar://problem/26228530>

[Radar WebKit Bug Importer]

<rdar://problem/26228605>

[Matt Baker]

Created attachment 278800 [details]
[Patch] Proposed Fix

[Joseph Pecoraro]

Comment on attachment 278800 [details]
[Patch] Proposed Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=278800&action=review

This looks good to me! We should stress test the performance of this briefly to ensure there are no unexpected side effects.

> Source/WebInspectorUI/UserInterface/Views/ScriptProfileTimelineView.js:139
> +            this.removeSubview(this._profileView);

This should probably removeEventListener to match the add below.

> Source/WebInspectorUI/UserInterface/Views/TimelineView.js:310
> +            if (!(node instanceof WebInspector.TimelineDataGridNode))
> +                return;

Nit: style

[Matt Baker]

Created attachment 278821 [details]
[Patch] Proposed Fix

[Matt Baker]

Comment on attachment 278821 [details]
[Patch] Proposed Fix

View in context: https://bugs.webkit.org/attachment.cgi?id=278821&action=review

> Source/WebInspectorUI/UserInterface/Views/ScriptClusterTimelineView.js:75
> +    selectRecord(record) { _contentViewContainer.currentContentView.selectRecord(record); }

Missing "this."

[Matt Baker]

Created attachment 278822 [details]
[Patch] Proposed Fix

[Matt Baker]

(In reply to comment #4)
> Comment on attachment 278800 [details]
> [Patch] Proposed Fix
> 
> View in context:
> https://bugs.webkit.org/attachment.cgi?id=278800&action=review
> 
> This looks good to me! We should stress test the performance of this briefly
> to ensure there are no unexpected side effects.

I did a few quick recordings comprised of multiple page navigations each (wired.com -> arstechnica.com -> cnn.com -> etc...), which resulted in > 45s of data and grids with > 50k records. Filtering the JavaScript & Events grids (details and call trees) had a noticeable delay of around 0.5-1s while typing.

My testing was far from rigorous, but indicates the perf impact is definitely acceptable for a first pass.

[Matt Baker]

(In reply to comment #8)
> (In reply to comment #4)
> > Comment on attachment 278800 [details]
> > [Patch] Proposed Fix
> > 
> > View in context:
> > https://bugs.webkit.org/attachment.cgi?id=278800&action=review
> > 
> > This looks good to me! We should stress test the performance of this briefly
> > to ensure there are no unexpected side effects.
> 
> I did a few quick recordings comprised of multiple page navigations each
> (wired.com -> arstechnica.com -> cnn.com -> etc...), which resulted in > 45s
> of data and grids with > 50k records. Filtering the JavaScript & Events
> grids (details and call trees) had a noticeable delay of around 0.5-1s while

Events and Call Trees

[Joseph Pecoraro]

Yeah, I just played with the patch. Performance is great for my basic tests so far, I think this would be really great to land now and improve later if we encounter issues!

[Joseph Pecoraro]

Comment on attachment 278822 [details]
[Patch] Proposed Fix

r=me

[WebKit Commit Bot]

Comment on attachment 278822 [details]
[Patch] Proposed Fix

Clearing flags on attachment: 278822

Committed r200873: <http://trac.webkit.org/changeset/200873>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.