Further development of Fock states

[apkille]

• Adds a documentation page for using Fock states.
• Moved predefined Fock objects from predefined.jl to a new file predefined_fock.jl.
• Added displacement and phase shift operators, DisplaceOp and PhaseShiftOp, respectively.
• Added simplification rules and rewriter with tests for Fock states.
• Updated express functionality for Fock objects.

[apkille]

@Krastanov you might disagree with the following changes, which is OK and we can discuss them:

• I replaced ContinuousCoherentState and DiscreteCoherentState with simply CoherentState, as that distinction didn't seem too important to include in this package, particularly since QuantumOptics.jl only has coherentstate.
• In express_nolookup, before for Fock states we would change the dimension of the Fock basis to 2 with the warning that the cutoff wasn't defined, but of course this isn't always the case, as we can define a finite cutoff. So, if the cutoff is infinite, then we change the dimension to 2.

[codecov]

Codecov Report

Attention: Patch coverage is 96.15385% with 1 line in your changes missing coverage. Please review.

Project coverage is 74.65%. Comparing base (f291b77) to head (3f9fe6d).

Files	Patch %	Lines
src/QSymbolicsBase/predefined_fock.jl	93.75%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #69      +/-   ##
==========================================
+ Coverage   72.94%   74.65%   +1.70%     
==========================================
  Files          18       19       +1     
  Lines         791      789       -2     
==========================================
+ Hits          577      589      +12     
+ Misses        214      200      -14     

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

[Krastanov]

I have a bunch of questions related to some of the details of how numerical expressions happen. I think I prefer we follow different semantics for the cutoff. Let's consider these notes as a first round of discussion on the topic.

[Krastanov]

This is pretty good. I have the following requests:

• bring back the warning on default cutoff
• make the cutoff a keyword arg in the constructor (to make it easy to extend in the future)
• remove one of the exports
• check whether the naming change affects QuantumSavory and make sure the breaking change is mentioned in a changelong and an appropriate version bump is made

Feel free to merge without another review after this is done.

[Krastanov]

let me know if I am missing something that is making the requests unreasonable

[apkille]

All of your requests sound good to me.

@Krastanov Edit: actually, there's one small issue related to the first request, which I mentioned earlier in the first round of review:

The problem with raising a warning in the constructor is that it creates allocations and increases the runtime. This causes tests in test_express_opt.jl to not pass. I instead added a warning to the docs page

[Krastanov]

ah, noted! Sounds good, let's keep it as a warning to the docs for now

[Krastanov]

I agree to the plan about adding a warning to the docs.

[apkille]

@Krastanov copying this down below: in your commit you modified QuantumOpticsRepr so that it has optional arguments, yet in your review you said that it should have a keyword argument. Could you please clarify?

[Krastanov]

My bad. I meant kwarg. Maybe with @kwdef

permission to merge