fuse_charge for adding charges to operators

[Yue-Zhengyuan]

This PR exports one more utility function fuse_charge to add charges to operators.

(It can replace _fuse_ids currently in PEPSKit.)

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

Files with missing lines	Coverage Δ
src/TensorKitTensors.jl	100.00% <ø> (ø)
src/tjoperators.jl	97.91% <100.00%> (-0.05%)	⬇️
src/utils.jl	93.75% <100.00%> (+0.56%)	⬆️

🚀 New features to boost your workflow:

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

[lkdvos]

Definitely looks reasonable to me, maybe we can just brainstorm slightly about the name?

[Yue-Zhengyuan]

First I deliberately avoided add_physical_charge that are already in MPSKit. fuse_ids is not descriptive enough to me, which has no information on which space the id acts on.

To make hybrids, add_charge or fuse_charge both work for me.

[lkdvos]

Just getting some opinions here, I have a slight preference to avoid add and use fuse to stick a bit more to the categorical language, especially since adding really only makes sense in our particular parametrization of ZNIrrep or U1Irrep. shift also seems like a good candidate, since that is typically the purpose? In principle I'm also okay with migrating the MPSKit function here and then simply overloading it in MPSKit, but I do agree that is a bit more annoying to orchestrate.

[Yue-Zhengyuan]

@lkdvos How about you listing a few options and then we vote by emoji reactions? I now prefer fuse after reading your comment, since this it literally what we are actually doing.

[leburgel]

I like fuse_charge the best, in the generic case fusing in a sector doesn't necessarily add nor shift in a clear way. I like this better than the name of the original MPSKit method even.