public sealed override object Invoke(BindingFlags invokeAttr, Binder binder, object[] parameters, CultureInfo culture)
{
#if ENABLE_REFLECTION_TRACE
if (ReflectionTrace.Enabled)
{
ReflectionTrace.ConstructorInfo_Invoke(this, parameters);
}
#endif
if (invokeAttr != BindingFlags.Default || binder != null || culture != null)
{
throw new NotImplementedException();
}
if (parameters == null)
{
parameters = Array.Empty <Object>();
}
// Most objects are allocated by NewObject and their constructors return "void". But in many frameworks,
// there are "weird" cases (e.g. String) where the constructor must do both the allocation and initialization.
// Reflection.Core does not hardcode these special cases. It's up to the ExecutionEnvironment to steer
// us the right way by coordinating the implementation of NewObject and MethodInvoker.
Object newObject = ReflectionCoreExecution.ExecutionEnvironment.NewObject(this.DeclaringType.TypeHandle);
Object ctorAllocatedObject = this.MethodInvoker.Invoke(newObject, parameters);
return(newObject != null ? newObject : ctorAllocatedObject);
}
public sealed override Object Invoke(Object[] parameters)
{
if (ReflectionTrace.Enabled)
{
ReflectionTrace.ConstructorInfo_Invoke(this, parameters);
}
if (parameters == null)
{
parameters = Array.Empty <Object>();
}
// Most objects are allocated by NewObject and their constructors return "void". But in many frameworks,
// there are "weird" cases (e.g. String) where the constructor must do both the allocation and initialization.
// Reflection.Core does not hardcode these special cases. It's up to the ExecutionEnvironment to steer
// us the right way by coordinating the implementation of NewObject and MethodInvoker.
Object newObject = ReflectionCoreExecution.ExecutionEnvironment.NewObject(this.DeclaringType.TypeHandle);
Object ctorAllocatedObject = this.MethodInvoker.Invoke(newObject, parameters);
return(newObject != null ? newObject : ctorAllocatedObject);
}