Refactor generic ideal method + avoid computing principal generator of ideal of PolynomialRing on construction

[user202729]

• use keyword arguments.
• delete the special case logic where naive .gcd() is ran for univariate polynomial ring.

Reason:

There's a doctest

K.<a> = QuadraticField(4569)
j = 46969655/32768
E = EllipticCurve(j=K(j))
C = E.isogeny_class()

that now takes very long to run. With this change, the runtime is cut in half. (Finding a single generator for a number field ideal is usually harder than finding a HNF basis of the ideal, and the latter is all pari needs to do relevant computations.)

Note: this is kind of a breaking change, in that ZZ.ideal(1, 2, 3).gens() will now return (1, 2, 3). I think it is preferable (users who want reduced generators can run gens_reduced).

Indeed, some tests break https://github.com/sagemath/sage/actions/runs/20863923668/job/59950182623 --- but ideally they should be refactored to avoid relying specifically on this behavior.

📝 Checklist

• The title is concise and informative.
• The description explains in detail what this PR is about.
• I have linked a relevant issue or discussion.
• I have created tests covering the changes.
• I have updated the documentation and checked the documentation preview.

⌛ Dependencies

[github-actions]

Documentation preview for this PR (built with commit 0e6f5fe; changes) is ready! 🎉
This preview will update shortly after each push to this PR.