General defect prime, with no squarefree hypothesis
Provedgpy_defect_prime_generalnumber-theory
A compatible but not-good support pair has a defect prime: a prime dividing the combined lcm, either dividing the i-coordinate lcm or dividing another coordinate's lcm with a congruent shift. This holds in full generality, with no squarefree hypothesis, via a general Chinese remainder witness. This is an unconditional exact statement; it does not bound the defect.
Preamble
import Mathlib open Classical
Formal statement
theorem gpy_defect_prime_general {k : ℕ} {h d e : Fin k → ℕ} {i : Fin k}
(hm : ∀ j, Nat.lcm (d j) (e j) ≠ 0)
(hcompat : ∀ i ∈ Finset.univ, ∀ j ∈ Finset.univ,
h i ≡ h j [MOD Nat.gcd (Nat.lcm (d i) (e i)) (Nat.lcm (d j) (e j))])
(hnotgood : ¬ (∃ w, (∀ j, Nat.lcm (d j) (e j) ∣ w + h j) ∧
Nat.Coprime (w + h i)
(Finset.lcm Finset.univ (fun j => Nat.lcm (d j) (e j))))) :
∃ p, Nat.Prime p ∧
p ∣ Finset.lcm Finset.univ (fun j => Nat.lcm (d j) (e j)) ∧
(p ∣ Nat.lcm (d i) (e i) ∨
∃ j, j ≠ i ∧ p ∣ Nat.lcm (d j) (e j) ∧ h i ≡ h j [MOD p]) := by sorrySource
GPY defect smallness (GPYDefectSmall.lean), restated into Mathlib-only terms