Refactor truncation module + add truncerror

[lkdvos]

This PR slightly refactors the truncation implementation, by separating out the interface from the implementation, more similar to the other parts of the library.

Additionally, I ported over truncerror (TruncationError), based on TensorKit.jl and TensorAlgebra.jl

I tentatively renamed it from truncerr to truncerror, I'm not sure how useful it is to shorten that name further. Happy to change that back if we don't like this.

[codecov]

Codecov Report

❌ Patch coverage is 81.92771% with 15 lines in your changes missing coverage. Please review.

Files with missing lines	Patch %	Lines
src/interface/truncation.jl	70.83%	14 Missing ⚠️
src/algorithms.jl	88.88%	1 Missing ⚠️

Files with missing lines	Coverage Δ
src/MatrixAlgebraKit.jl	100.00% <ø> (ø)
src/implementations/truncation.jl	98.57% <100.00%> (+10.15%)	⬆️
src/algorithms.jl	92.80% <88.88%> (-0.31%)	⬇️
src/interface/truncation.jl	70.83% <70.83%> (ø)

🚀 New features to boost your workflow:

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

[mtfishman]

Additionally, I ported over truncerror (TruncationError), based on TensorKit.jl and TensorAlgebra.jl

Great, it will be nice to have a shared implementation to use.

I tentatively renamed it from truncerr to truncerror, I'm not sure how useful it is to shorten that name further. Happy to change that back if we don't like this.

truncerror looks good to me.

[Jutho]

Great PR.

One thing it does not address is (not saying that it should) is that in the AD rules for truncated factorizations, I will need to reproduce the logic of truncate!(alg, out, strategy), i.e. manually call findtruncated(_sorted) and then to the actual truncation in the right way for that alg / out combination.