Bug 127899 – Web Inspector: Add the model objects for the new profile data

• Status: RESOLVED
• Resolution: FIXED
• Component: Web Inspector

[Timothy Hatcher]

Add proper model objects for the profile data.

[Radar WebKit Bug Importer]

<rdar://problem/15944957>

[Timothy Hatcher]

Created attachment 222649 [details]
Patch

[Joseph Pecoraro]

Comment on attachment 222649 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=222649&action=review

r=me

> Source/WebInspectorUI/UserInterface/ProfileNode.js:148
> +        rangeStartTime = !isNaN(rangeStartTime) ? rangeStartTime : 0;
> +        rangeEndTime = !isNaN(rangeEndTime) ? rangeEndTime : Infinity;

This reads strange to me. I'd prefer a positive condition "isNaN(foo) ? 0 : foo".

Or the longer but more straightforward:

if (isNaN(foo))
  foo = ..;

> Source/WebInspectorUI/UserInterface/ProfileNode.js:207
> +        cookie[WebInspector.ProfileNode.SourceCodeURLCookieKey] = this._sourceCodeLocation ? this._sourceCodeLocation.sourceCode.url ? this._sourceCodeLocation.sourceCode.url.hash : null : null;

Ew! Replace:

    a ? a.b ? a.b.c : null : null;

With:

    a && a.b ? a.b.c : null

> Source/WebInspectorUI/UserInterface/ProfileNodeCall.js:30
> +    console.assert(startTime);

So, startTime cannot be 0? Should this be a type check, or is 0 not allowed?

    console.assert(typeof startTime === "number");

> Source/WebInspectorUI/UserInterface/TimelineManager.js:375
> +        // Iterate over the node tree using a stack. Doing this recursively can easily cause a stack overflow.
> +        // We traverse the profile in post-order and convert the payloads in place until we get back to the root.

This is so wild. I feel like there must be a better way. But I can't find any logic issues with it.

> Source/WebInspectorUI/UserInterface/TimelineManager.js:396
> +        return new WebInspector.Profile(rootNodes, payload.idleTime);

Assert "idleTime" in payload near the function entry?

[Timothy Hatcher]

https://trac.webkit.org/r163141