import Mathlib namespace OAI namespace ExactQuantumFactoring open scoped BigOperators /-- The five primitive gates exactly as in Section 2. -/ inductive Primitive where | not | cnot | toffoli | hadamard | phase deriving DecidableEq def Primitive.arity : Primitive → ℕ | .not | .hadamard | .phase => 1 | .cnot => 2 | .toffoli => 3 def Primitive.code : Primitive → ℕ | .not => 0 | .cnot => 1 | .toffoli => 2 | .hadamard => 3 | .phase => 4 /-- An inverse and/or one added control. This is a fixed set of 20 gate names, not an unbounded-control or arbitrary-angle gate model. -/ structure Gate where primitive : Primitive inverse : Bool controlled : Bool deriving DecidableEq def Gate.arity (g : Gate) : ℕ := g.primitive.arity + if g.controlled then 1 else 0 /-- Matrix entries are indexed by lists in wire order (row first, column second). Entries on ill-sized lists are irrelevant and set to zero. -/ noncomputable def Primitive.matrix : Primitive → List Bool → List Bool → ℂ | .not, [a], [b] => if a = !b then 1 else 0 | .cnot, [a,b], [c,d] => if a = c ∧ b = Bool.xor c d then 1 else 0 | .toffoli, [a,b,c], [d,e,f] => if a = d ∧ b = e ∧ c = Bool.xor f (d && e) then 1 else 0 | .hadamard, [a], [b] => (if a && b then -1 else 1) / (Real.sqrt 2 : ℂ) | .phase, [a], [b] => if a = b then (if b then Complex.I else 1) else 0 | _, _, _ => 0 noncomputable def Gate.baseMatrix (g : Gate) (a b : List Bool) : ℂ := if g.inverse then star (g.primitive.matrix b a) else g.primitive.matrix a b noncomputable def Gate.matrix (g : Gate) (a b : List Bool) : ℂ := if g.controlled then match a, b with | false :: a', false :: b' => if a' = b' then 1 else 0 | true :: a', true :: b' => g.baseMatrix a' b' | _, _ => 0 else g.baseMatrix a b /-- A primitive is placed on distinct actual wires, with no operation of unbounded arity hidden in a circuit instruction. -/ structure Instruction (q : ℕ) where gate : Gate wire : Fin gate.arity → Fin q distinct : Function.Injective wire abbrev Basis (q : ℕ) := Fin q → Bool abbrev State (q : ℕ) := Basis q → ℂ noncomputable def Instruction.matrix {q : ℕ} (o : Instruction q) : Matrix (Basis q) (Basis q) ℂ := by classical exact fun a b => if ∀ i : Fin q, (∀ j, o.wire j ≠ i) → a i = b i then o.gate.matrix (List.ofFn (a ∘ o.wire)) (List.ofFn (b ∘ o.wire)) else 0 structure Circuit where qubits : ℕ instructions : List (Instruction qubits) noncomputable def Circuit.apply (c : Circuit) (ψ : State c.qubits) : State c.qubits := c.instructions.foldl (fun v o => o.matrix.mulVec v) ψ /-- Input wires are first, least significant bit first; all other wires start zero. -/ def inputBasis (q ℓ N : ℕ) : Basis q := fun i => if i.val < ℓ then N.testBit i.val else false noncomputable def Circuit.outputState (c : Circuit) (ℓ N : ℕ) : State c.qubits := by classical exact c.apply (fun x => if x = inputBasis c.qubits ℓ N then 1 else 0) /-- The cutoff and padding are the standing parameters in Section 2. -/ def paddedLength (ℓ : ℕ) : ℕ := max 128 ℓ def readBit {q : ℕ} (x : Basis q) (i : ℕ) : Bool := if h : i < q then x ⟨i,h⟩ else false /-- n entries of n bits each, in the n² wires immediately after the input. -/ def outputList {q : ℕ} (ℓ : ℕ) (x : Basis q) : List ℕ := let n := paddedLength ℓ List.ofFn (fun a : Fin n => ∑ j ∈ Finset.range n, if readBit x (ℓ + a.val * n + j) then 2^j else 0) /-- Complete nondecreasing prime factorization, multiplicities included, padded by zeros, as in Section 6. There is no primality oracle in the circuit. -/ def CorrectEncoding (N n : ℕ) (xs : List ℕ) : Prop := ∃ ps : List ℕ, (∀ p ∈ ps, Nat.Prime p) ∧ ps.Pairwise (· ≤ ·) ∧ ps.prod = N ∧ ps.length ≤ n ∧ xs = ps ++ List.replicate (n - ps.length) 0 /-- Unnormalized Born sum on correct outputs; success is exactly one, not positivity, postselection, or a limiting high probability. -/ noncomputable def Circuit.correctProbability (c : Circuit) (ℓ N : ℕ) : ℝ := by classical exact ∑ x : Basis c.qubits, if CorrectEncoding N (paddedLength ℓ) (outputList ℓ x) then Complex.normSq (c.outputState ℓ N x) else 0 /-- Unary natural words with delimiters make the circuit description explicit. Unary wire addresses cause at most polynomial overhead when width is polynomial. -/ def natWord (k : ℕ) : List Bool := List.replicate k true ++ [false] def Instruction.encode {q : ℕ} (o : Instruction q) : List Bool := natWord o.gate.primitive.code ++ [o.gate.inverse, o.gate.controlled] ++ (List.ofFn (fun i => o.wire i |>.val)).flatMap natWord def Circuit.encode (c : Circuit) : List Bool := natWord c.qubits ++ natWord c.instructions.length ++ c.instructions.flatMap Instruction.encode /-- A single finite-alphabet classical generator receives only the INPUT LENGTH. All stack alphabets, states and labels are finite: no large-integer unit-cost oracle is hidden in TM2 push/branch functions. -/ def Uniform (family : ℕ → Circuit) : Prop := ∃ generator : Turing.TM2ComputableInPolyTime (fun ℓ => List.replicate ℓ true) Circuit.encode family, ∀ k, Finite (generator.tm.Γ k) def PolynomialResources (family : ℕ → Circuit) : Prop := ∃ bound : Polynomial ℕ, ∀ ℓ, (family ℓ).qubits ≤ bound.eval ℓ ∧ (family ℓ).instructions.length ≤ bound.eval ℓ /-- Literal main endpoint: one length-uniform family, exact output for EVERY binary integer N≥2, and worst-case polynomial gates AND qubits. -/ def MainTheorem : Prop := ∃ family : ℕ → Circuit, Uniform family ∧ PolynomialResources family ∧ ∀ (ℓ N : ℕ), 2 ≤ N → N.size = ℓ → ℓ + (paddedLength ℓ)^2 ≤ (family ℓ).qubits ∧ (family ℓ).correctProbability ℓ N = 1 end ExactQuantumFactoring end OAI