Meta issue: algebras, orders and lattices

[thofma]

Orders

• Cleanup, fix the TODOs from feat: arbitrary base rings for AlgAssAbsOrd #1728.
• Rename. The implementation in feat: arbitrary base rings for AlgAssAbsOrd #1728 requires the base ring to be PID/Hermite ring. Maybe AlgebraOrderPID?

Lattices

• Rename ideal(...) to lattice(...).
• Rename the type AlgAssAbsOrdIdl to AlgAssAbsOrdLat, or whatever the prefix will be.
• Decide whether (fractional) ideals of orders should be of the same type as lattices, distinguished only by a (runtime) flag.

Algebras

• Add radical and splitting into simple algebras for function fields over finite fields.
• Add hom(A, B, ::Vector) constructor.
• Morphisms defined by hom(A, B, data) do not check whether they are morphisms.
• Think about antihomomorphisms (separate type)? Example is $A \mapsto A^t$ on matrix algebras or $g \mapsto g^{-1}$ for group algberas.

Misc & Performance

• Make sure that integral_split is fast for ZZ/QQ.
• Decide on API for _, d = integral_split(x, R). Should x be "integral" over R if and only if is_one(d)? Or is_unit(d)?