[Webkit-unassigned] [Bug 114639] [CSS Shaders] Extract validation logic out of CustomFilterValidatedProgram into a CustomFilterValidator class

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Tue May 7 17:04:10 PDT 2013


https://bugs.webkit.org/show_bug.cgi?id=114639


Max Vujovic <mvujovic at adobe.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #200998|                            |review?, commit-queue-
               Flag|                            |




--- Comment #1 from Max Vujovic <mvujovic at adobe.com>  2013-05-07 17:02:34 PST ---
Created an attachment (id=200998)
 --> (https://bugs.webkit.org/attachment.cgi?id=200998&action=review)
Patch

Some areas in the patch I'd like feedback on:

(1) CustomFilterValidator is a class with only static methods, and it should never be initialized. I've made its constructor private. I'm thinking I should additionally disallow the copy constructor and assignment. I could only find the macro "DISALLOW_IMPLICIT_CONSTRUCTORS" defined in wtf/dtoa/utils.h. It doesn't seem correct to include that file here, since the file is all about double to string conversion. I'm wondering if we have some convention in WebKit for static classes.

(2) I used only file static methods in CustomFilterValidator, since I try to keep things out of headers. Any thoughts on that? I could instead use class static methods or some mix between the two.

(3) Does the static local initialization of ANGLEWebKitBridge in CustomFilterValidator.cpp look ok? I didn't want to use the DEFINE_STATIC_LOCAL macro because it only supports sending arguments to the ANGLEWebKitBridge constructor. I need to both send arguments to the constructor *and* perform some initialization after running the constructor (see the createShaderValidator function).

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.


More information about the webkit-unassigned mailing list