Factor a polynomial in two variables, by Kronecker's substitution

[Rafael-SOWNet]

Part of #746 tier 1, item 43 — and it is the case that item names.

MathS.Polynomials.Factor("x ^ 2 - y ^ 2", "x")   // 2.3.0: null.  Now: (x + y) * (x - y)

How

A factor of a polynomial of degree d in x has degree at most d in x. So with s = d + 1, the
map x^i y^j → t^(i + s*j) is injective on every monomial that can appear in the polynomial or in
any of its factors: i is the remainder and j the quotient of the exponent by s, and neither
can be confused with another pair. The one-variable image is handed to the factoriser that already
exists, and each subset of its irreducible factors names a candidate two-variable factor.

	2.3.0	now
Factor("x ^ 2 - y ^ 2", "x")	null	(x + y) * (x - y)
Factor("x ^ 2 + 2 * x * y + y ^ 2", "x")	null	(x + y) ^ 2
Factor("x ^ 3 - y ^ 3", "x")	null	(x - y) * (x ^ 2 + x * y + y ^ 2)
Factor("x ^ 4 - y ^ 4", "x")	null	(x + y) * (x ^ 2 + y ^ 2) * (x - y)
Factor("x ^ 2 * y ^ 2 - 1", "x")	null	(x * y + 1) * (x * y - 1)
Factor("x ^ 2 - y ^ 2 + 2 * x + 1", "x")	null	(x + y + 1) * (x - y + 1)
Factor("x ^ 2 + y ^ 2", "x")	null	null — irreducible over ℚ
Factor("x * y + z", "x")	null	null — three variables

It cannot answer wrongly

This is the part worth reviewing. The substitution is injective on monomials but not on
factorisations, so the image may factor further than the polynomial does and every candidate really
is a guess. Two checks stand between a guess and an answer:

1. each candidate must divide the remaining polynomial exactly, or it is discarded;
2. the assembled factors are multiplied and divided back into the input before anything is returned.

So the failure mode is a refusal, never a wrong factorisation.

What it refuses

The image has degree d + s*e for degree e in the second variable, and the one-variable factoriser
stops at 32 — so this reaches bidegrees like (2, 10), (3, 7) and (5, 4) and refuses past them. The
recombination is over subsets, so the image's factor count is capped at 12.

Lifting that ceiling means Hensel lifting with an evaluation homomorphism. This is not a step
towards that — it is the classical method, exact inside its bound and honest outside it. Item 43's
full ask is still open.

Blast radius

MathS.Polynomials.Factor has no caller inside the library — checked with a grep, not assumed —
so no simplification, solution or integral moves with this.

• Suite: 8574 passed, 0 failed, 14 skipped.
• casbench and simpsweep re-run against this branch are byte-identical to master's committed
  reports.
• Repeated factors are collected into a power, as the one-variable path does, so one question does
  not get two shapes of answer depending on which path took it.
• Tests assert the number of distinct factors as well as numeric agreement, so a partial
  factorisation reported as a whole one fails rather than passing quietly.