RDF[] for libsingular, delegate ideal comparison and .divides() to singular

[user202729]

1. in refuse to compare two ideals that we don't know how to compare #41040, it was decided that ideal comparison should compare them by set inclusion. Ideals of ZZ[x]'s comparison becomes unimplemented.

   This pull request implements it by delegating to Singular. This is similar to how inverse_mod currently delegate to Singular, see Make inverse_mod use Singular in a few cases #39743

2. Similarly, .divides() of polynomials over ℤ delegate to Singular.

3. Previously, RDF[] and CDF[] can be represented by Singular command-line interface. This represents these polynomial rings in libsingular too for feature parity.

   The magic constant 15 is just to match what you get using the singular interface.

---

The correctness of the code can be verified by comparing with singular's source code. For example shortfl.h contains

#if SIZEOF_DOUBLE == SIZEOF_LONG
#define SI_FLOAT double
#else
#define SI_FLOAT float
#endif

and later shortfl.cc:

union nf
{
  SI_FLOAT _f;
  number _n;

  nf(SI_FLOAT f): _f(f){};

  nf(number n): _n(n){};

  inline SI_FLOAT F() const {return _f;}
  inline number N() const {return _n;}
};

SI_FLOAT nrFloat(number n)
{
  return nf(n).F();
}

For n_long_C type, gnumpc.cc has e.g.

static long ngcInt(number &i, const coeffs r)
{
  assume( getCoeffType(r) == n_long_C );

  return ((gmp_complex*)i)->real();
}

indicating the correct thing to do with the number pointer is to cast them to gmp_complex*.

📝 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 d245ceb; changes) is ready! 🎉
This preview will update shortly after each push to this PR.