Add a disamgiguating conversion

[kshyatt]

This seems to be needed for conversion of BlockTensorMaps backed by GPU arrays to CuTensorMap -- otherwise an ambiguity error arises.

[codecov]

Codecov Report

❌ Patch coverage is 0% with 30 lines in your changes missing coverage. Please review.
✅ Project coverage is 4.43%. Comparing base (0b3278f) to head (880e610).

Files with missing lines	Patch %	Lines
src/tensors/abstractblocktensor/conversion.jl	0.00%	24 Missing ⚠️
src/tensors/tensoroperations.jl	0.00%	6 Missing ⚠️

❗ There is a different number of reports uploaded between BASE (0b3278f) and HEAD (880e610). Click for more details.

HEAD has 12 uploads less than BASE

Flag	BASE (0b3278f)	HEAD (880e610)
	24	12

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #47       +/-   ##
==========================================
- Coverage   55.43%   4.43%   -51.01%     
==========================================
  Files          18      18               
  Lines        1407    1420       +13     
==========================================
- Hits          780      63      -717     
- Misses        627    1357      +730     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[lkdvos]

Is this not showing up in Aqua because the ambiguity only is there if CUDA is loaded?

[kshyatt]

Yes basically exactly that.

[kshyatt]

lol "amgiguating"

[kshyatt]

Annoying, more ambiguities to fix.

[kshyatt]

OK, I think this is ok to go (but we'll want to fix Project.toml) after QuantumKitHub/TensorKit.jl#375 gets into shape

[kshyatt]

OK, created a more specific TensorMap conversion and moved some common logic into a new function. Also got the storage type thing seemingly resolved...

[kshyatt]

I think everything got resolved here?

[lkdvos]

Suggested change

	end
	function TO.tensorscalar(t::AbstractBlockTensorMap{T, S, 0, 0}) where {T, S}
	return prod(TO.tensorscalar, nonzero_values(t))

Do you know if this would work?

[kshyatt]

Let me try it locally and get back to you