Issue with RelaxedPrecision and OpCompositeConstruct

I have the following shader:

```
Capability Shader
Capability Sampled1D
1: ExtInstImport "GLSL.std.450"
MemoryModel Logical GLSL450
EntryPoint Fragment 4 "main" 9
ExecutionMode 4 OriginUpperLeft
Source GLSL 430
SourceExtension "GL_EXT_nonuniform_qualifier"
SourceExtension "GL_GOOGLE_cpp_style_line_directive"
SourceExtension "GL_GOOGLE_include_directive"
Name 4 "main"
Name 9 "fragColor"
Name 11 "gl_DefaultUniformBlock"
MemberName 11(gl_DefaultUniformBlock) 0 "uTDPass"
MemberName 11(gl_DefaultUniformBlock) 1 "uTDCurrentDepth"
MemberName 11(gl_DefaultUniformBlock) 2 "exampleUniform"
Name 13 ""
Decorate 9(fragColor) RelaxedPrecision
Decorate 9(fragColor) Location 0
MemberDecorate 11(gl_DefaultUniformBlock) 0 Offset 0
MemberDecorate 11(gl_DefaultUniformBlock) 1 Offset 4
MemberDecorate 11(gl_DefaultUniformBlock) 2 RelaxedPrecision
MemberDecorate 11(gl_DefaultUniformBlock) 2 Offset 8
Decorate 11(gl_DefaultUniformBlock) Block
Decorate 17 RelaxedPrecision
Decorate 18 RelaxedPrecision
2: TypeVoid
3: TypeFunction 2
6: TypeFloat 32
7: TypeVector 6(float) 4
8: TypePointer Output 7(fvec4)
9(fragColor): 8(ptr) Variable Output
10: TypeInt 32 1
11(gl_DefaultUniformBlock): TypeStruct 10(int) 10(int) 6(float)
12: TypePointer PushConstant 11(gl_DefaultUniformBlock)
13: 12(ptr) Variable PushConstant
14: 10(int) Constant 2
15: TypePointer PushConstant 6(float)
4(main): 2 Function None 3
5: Label
16: 15(ptr) AccessChain 13 14
17: 6(float) Load 16
18: 7(fvec4) CompositeConstruct 17 17 17 17
Store 9(fragColor) 18
Return
FunctionEnd
```

Reflection fails since it's trying to reflect the RelaxedPrecision on result_id 18, however ParseNodes() does not assign the result_id to OpCompositeConstruct nodes. Isn't OpCompositeConstruct a load/store operation, and should be treated as such? The SPIR-V is coming out of GLSLang, and passes the SPIR-V validator.

Thanks

Start by reading ParseNodes() and the reflection parser's handling of OpCompositeConstruct nodes. Reproduce the supplied shader case and verify that RelaxedPrecision attached to result_id 18 is reflected correctly without breaking validation.

Written by the indexing model from the issue text.