Add deprecation warning to algs module

[ElePT]

Summary

This PR shows a deprecation message proposal for the migration of algorithms to a separate community repo (see Qiskit/RFCs#44). Given that there is no module-level decorator, the message is not really standardized, but I believe that wording the message right is particularly important in this case, so please don't hesitate to comment/give suggestions (@1ucian0, @Eric-Arellano)

Details and comments

This PR is blocked by the actual creation of the new repository.

[Eric-Arellano]

Looks good! Thanks.

[coveralls]

Pull Request Test Coverage Report for Build 5611202474

• 2 of 2 (100.0%) changed or added relevant lines in 1 file are covered.
• 9 unchanged lines in 3 files lost coverage.
• Overall coverage increased (+0.005%) to 86.057%

---

Files with Coverage Reduction	New Missed Lines	%
crates/qasm2/src/expr.rs	1	93.76%
crates/qasm2/src/lex.rs	2	90.13%
crates/qasm2/src/parse.rs	6	97.11%

Totals
Change from base Build 5610052878:	0.005%
Covered Lines:	72898
Relevant Lines:	84709

---

💛 - Coveralls

[qiskit-bot]

One or more of the the following people are requested to review this:

• @ElePT
• @Qiskit/terra-core
• @woodsp-ibm

[woodsp-ibm]

I had mentioned the docs, migration guides in that we might want to add something there. What about release notes - again something can be added after but I imagine this ought to get noted at that level.

[mtreinish]

I added a release note in: e1a1c62 and it LGTM now. Feel free to change any of the text in it, I struggled a bit with what to say in there and I'm still not super satisfied with the exact text (but that's probably my bias against the decision to migrate it out of terra leaking through) we can also always tweak the exact release note contents until the final release. I'll hold off on adding to the merge queue until @woodsp-ibm and/or @Cryoris think it's ready to go

[woodsp-ibm]

It LGTM - the only comment/text I might make/add is that if users are still using the old deprecated algorithms/QuantumInstance they need to first migrate to the primitive based algos and then switch. We might want to point to the algorithm migration guide in the release note along with such a comment to aid them in this regard. If they are already using the primitive based algos then they should be good to use qiskit_algorithms and it should just be that simple import change.

[mtreinish]

This LGTM and I've verified from qiskit import algorithms raises the warnings as expected so I think we're good to go here and we can concentrate the efforts on getting the qiskit-algorithms package ready.