Add ValueAlias for ValueType reconstruction

[ye-luo]

Proposed changes

Add a helper class for Value type reconstruction.

What type(s) of changes does this code introduce?

• New feature

Does this introduce a breaking change?

• No

What systems has this change been tested on?

laptop

Checklist

• Yes. This PR is up to date with current the current state of 'develop'
• Yes. Code added or changed in the PR has been clang-formatted

[PDoakORNL]

I don't see a problem with this but I guess I don't understand why its needed.

I always have assumed that classes and functions would be templated on Value since that basically determines everything except what the full precisions is.
If a derived type is complex definite and you have an indeterminant type I always assumed its type would be determined via

std::complex<RealAlias<T>> var;
// or
std::complex<FullPrecReal> var_full_prec;

What am I not seeing here?

[ye-luo]

I don't see a problem with this but I guess I don't understand why its needed.

I always have assumed that classes and functions would be templated on Value since that basically determines everything except what the full precisions is. If a derived type is complex definite and you have an indeterminant type I always assumed its type would be determined via

std::complex<RealAlias<T>> var;
// or
std::complex<FullPrecReal> var_full_prec;

What am I not seeing here?

Value is not complex definite. QTBase and QTFull relies on ifdef. To have full template, I tries not ot rely on QTFull but promote precision but to derive FullPrecValue based on Value using ValueAlias.

[PDoakORNL]

This is why I thought we would need Value and FullPrecValue or FullPrecReal for value/precision templating. I'm unclear on how if you have only a Real or a Value type coming into a function how you can discover whether it is reduced or full precision and what the full precision is without some global type alias.

This is why WaveFunctionTypes.hpp looks like it does, but I'll wait and see how this is used.

[ye-luo]

WaveFunctionTypes<VALUE, VALUE_FP> needs to consistently put value types. If one use WaveFunctionTypes<VALUE, REAL_FP> it breaks. With added reconstruct capability, it can be changed to WaveFunctionTypes<VALUE, FP_PREC> and internally derive VALUE_FP as ValueAlias<FP_PREC, VALUE>.

[PDoakORNL]

I get you now. I agree that <VALUE, FP_PREC> is better. And now I see what you are after here.

[prckent]

Test this please