Move SkPaint mangement for 2D canvas into CanvasRenderingContext2DState

Source/core/html/canvas/CanvasStyle.cpp

 /*
  * Copyright (C) 2006, 2008 Apple Inc. All rights reserved.
  * Copyright (C) 2008, 2010 Nokia Corporation and/or its subsidiary(-ies)
  * Copyright (C) 2007 Alp Toker <alp@atoker.com>
  * Copyright (C) 2008 Eric Seidel <eric@webkit.org>
  *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions
  * are met:
  * 1. Redistributions of source code must retain the above copyright
@@ skipping 127 matching lines @@
         break;
     case Gradient:
         context->setFillGradient(canvasGradient()->gradient());
         break;
     case ImagePattern:
         context->setFillPattern(canvasPattern()->pattern());
         break;
     }
 }
 
+SkShader* CanvasStyle::shader() const
+{
+    switch (m_type) {
+    case ColorRGBA:
+        return nullptr;
+    case Gradient:
+        return canvasGradient()->gradient()->shader();
+    case ImagePattern:
+        return canvasPattern()->pattern()->shader();
+    default:
+        ASSERT_NOT_REACHED();
+    }
+    return nullptr;
+}
+
+RGBA32 CanvasStyle::paintColor() const
+{
+    if (m_type == ColorRGBA)
+        return m_rgba;
+    ASSERT(m_type == Gradient || m_type == ImagePattern);
+    return SK_ColorBLACK;

[Stephen White, 2015/03/18 22:09:35]

Nit: this is an RGBA32 (Blink type) but you're returning a Skia constant
(SK_ColorBLACK).

[Justin Novosad, 2015/03/19 14:32:21]

Yeah... RGBA32 and SkColor are completely interchangeable.  The only reason I
can think of for RGBA32 to exist is to prevent blink::Color from depending on
SkColor, but perhaps that is futile? I think we should consider removing RGBA32
from the code entirely and just use SkColor. Anyways... for this CL, I will use
Color::black here.

+}
+
 DEFINE_TRACE(CanvasStyle)
 {
     visitor->trace(m_gradient);
     visitor->trace(m_pattern);
 }
 
 }