Fix RenderState Parameters Leaking Between Renders

[codex128]

This PR fixes shadows and other rendering issues caused by #2091 (github issue #2129), where render state parameters were leaking between render techniques because RenderState#set does not set all parameters.

Unfortunately, I was unable to get spotlight shadows to work in my test class, but TestSpotLightShadows worked fine, so I'm assuming I simply set up spotlight shadows wrong.

Any testing is greatly appreciated.

[Trass3r]

Looks like this fixed the issue 👍🏽.

[stephengold]

Unless there's further substantive discussion, I plan to integrate this PR in about 48 hours.

[stephengold]

Oh hey, I just noticed that this PR includes a new file "TestShadows.java".

That file needs a copy of the JME license at the top. It also needs javadoc describing the purpose of the test and how to determine whether the test passes or fails.

[codex128]

I plan on deleting that test, because it doesn't really test anything other tests don't already cover. I built it so I could focus on the shadow debug without a bunch of other code cluttering things up.

[stephengold]

back on track for integration in about 45 hours.

[capdevon]

Hi guys, what do you think about writing the copyFrom method using a static method like for java utility libraries (eg System.arraycopy() or Objects.equals(a, b) ? It might be useful to return the copied object.

//Usage:
mergedRenderState = RenderState.copy( RenderState.DEFAULT, mergedRenderState );
//or
RenderState.copy( RenderState.DEFAULT, mergedRenderState );

public static RenderState copy(RenderState from, RenderState to) {
        to.applyBlendMode = from.applyBlendMode;
        to.applyColorWrite = from.applyColorWrite;
        to.applyCullMode = from.applyCullMode;
        to.applyDepthFunc = from.applyDepthFunc;
        to.applyDepthTest = from.applyDepthTest;
        to.applyDepthWrite = from.applyDepthWrite;
        to.applyLineWidth = from.applyLineWidth;
        to.applyPolyOffset = from.applyPolyOffset;
        to.applyStencilTest = from.applyStencilTest;
        to.applyWireFrame = from.applyWireFrame;
        to.backStencilDepthFailOperation = from.backStencilDepthFailOperation;
        to.backStencilDepthPassOperation = from.backStencilDepthPassOperation;
        to.backStencilFunction = from.backStencilFunction;
        to.backStencilMask = from.backStencilMask;
        to.backStencilReference = from.backStencilReference;
        to.backStencilStencilFailOperation = from.backStencilStencilFailOperation;
        to.blendEquation = from.blendEquation;
        to.blendEquationAlpha = from.blendEquationAlpha;
        to.blendMode = from.blendMode;
        to.cachedHashCode = from.cachedHashCode;
        to.colorWrite = from.colorWrite;
        to.cullMode = from.cullMode;
        to.depthFunc = from.depthFunc;
        to.depthTest = from.depthTest;
        to.depthWrite = from.depthWrite;
        to.dfactorAlpha = from.dfactorAlpha;
        to.dfactorRGB = from.dfactorRGB;
        to.frontStencilDepthFailOperation = from.frontStencilDepthFailOperation;
        to.frontStencilDepthPassOperation = from.frontStencilDepthPassOperation;
        to.frontStencilFunction = from.frontStencilFunction;
        to.frontStencilMask = from.frontStencilMask;
        to.frontStencilReference = from.frontStencilReference;
        to.frontStencilStencilFailOperation = from.frontStencilStencilFailOperation;
        to.lineWidth = from.lineWidth;
        to.offsetEnabled = from.offsetEnabled;
        to.offsetFactor = from.offsetFactor;
        to.offsetUnits = from.offsetUnits;
        to.sfactorAlpha = from.sfactorAlpha;
        to.sfactorRGB = from.sfactorRGB;
        to.stencilTest = from.stencilTest;
        to.wireframe = from.wireframe;
        return to;
    }

[codex128]

I'm not opposed to it, but copyFrom (and even set, for that matter) is only used once throughout the whole of jme, so I don't think it's important enough to disrupt the PR to integrate.

[stephengold]

The proposed static copy() method would be error-prone, since it would be easy to unintentionally reverse the arguments. Let's stick with the current design.

[stephengold]

Unless there's further substantive discussion, I plan to integrate this PR in about 24 hours.

[stephengold]

Thank you @codex128 for your contribution to JME!