Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

BG §15: Theorem 15.2, §14-gated conditional helpers

Definition
OddOrder_BG_Ch4_FamilyOfMaximal_S15_MF_Theorem152Helpers

by burkh4rt · Nov 12, 2026 · Mathlib 0df444a (Lean v4.33.1)

feit-thompsonfinite-groupsgroup-theoryishida-odd-order

The §14-gated conditional helpers for Theorem 15.2 of Bender–Glauberman, Local Analysis for the Odd Order Theorem, §15, split from the former monolithic §15 file. Lemmas include typeP1_msigma_eq_derivedInG, prime-manner action for type P\mathcal{P}P, existence of a Sylow subgroup equal to an OπO_{\pi}Oπ​-core, and a Frattini factorization.

The bundle consists of lemmas only.

The bundle also proves 17 supporting lemmas used by later statements.

Definition code
/-
Copyright (c) 2026 Yawara Ishida. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Yawara Ishida
-/
-- From github.com/yawara/odd-order @ 20a6fccbb63c (Apache-2.0; LICENSE at that commit):
--   OddOrder/BG/Ch4_FamilyOfMaximal/S15_MF/Theorem152Helpers.lean
-- Generated for Prove2Me by deleting declarations; the remaining commands are verbatim except:
-- `private` dropped where another file uses the declaration, and a solution's main theorem
-- renamed `_root_.solution` (platform convention).
import Definitions.Def_OddOrder_BG_Ch4_FamilyOfMaximal_S15_MF_SetupLemma151
import Definitions.Def_OddOrder_BG_Ch4_FamilyOfMaximal_S15_MF_TypeP1Forcing

namespace OddOrder.BG.Ch4.S15

open OddOrder.GroupTheory

open OddOrder.Isaacs

open OddOrder.BG.Ch3.S12

open OddOrder.BG.Ch4.S14

open scoped Pointwise

open scoped IsMulCommutative

open scoped commutatorElement

variable {G : Type*} [Group G]

/-- **BG Theorem 15.2, `M_σ = M'` for a type-`P₁` member** (mmd L4188 "`M_σ = M'`"): for a type-`P₁`
maximal subgroup `M` with Hall `κ(M)`-subgroup `K`, the `σ`-core equals the derived subgroup.

`M_σ ≤ M'` always (`Msigma_le_derived`).  For the reverse, compare orders: the derived subgroup
complements `K` in `M` (Theorem 14.7(h) duality, `typeP_duality`), so `|M'|·|K| = |M|`, while
`typeP1_card_eq` gives `|M| = |M_σ|·|K|`; cancelling `|K|` yields `|M'| = |M_σ|`, so `M_σ = M'`.
This is `§14`-gated only through `typeP_duality`/`typeP1_card_eq` (both sorry-free).  Supplies
conjunct `M_σ = M'` of Theorem 15.2, and the `M_σ ⋊ K` complement of its proof. -/
theorem typeP1_msigma_eq_derivedInG [Finite G]
    (hG : OddOrder.BG.IsMinimalSimpleOdd G) {M K Kstar : Subgroup G}
    (hM : M ∈ maximalSubgroups G) (hP1 : S14.IsTypeP1 M) (hKM : K ≤ M)
    (hK : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M))
    (hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G)) :
    OddOrder.BG.Ch3.S10.Msigma M = derivedInG M := by
  classical
  obtain ⟨hcomplD, _, _⟩ := S14.typeP_duality hG hM hP1.1 hKM hK hKstar
  have hM'le : derivedInG M ≤ M := Subgroup.map_subtype_le _
  have hcardM' : Nat.card ↥((derivedInG M).subgroupOf M) = Nat.card ↥(derivedInG M) :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hM'le).toEquiv
  have hcardK' : Nat.card ↥(K.subgroupOf M) = Nat.card ↥K :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hKM).toEquiv
  have hmul := hcomplD.card_mul_card
  rw [hcardM', hcardK'] at hmul
  have hcardM := S14.typeP1_card_eq hG hM hP1 hKM hK
  have heq : Nat.card ↥(derivedInG M) = Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M) :=
    Nat.eq_of_mul_eq_mul_right Nat.card_pos (by rw [hmul, hcardM])
  exact Subgroup.eq_of_le_of_card_ge (OddOrder.BG.Ch3.S10.Msigma_le_derived hG hM) heq.le

/-- **BG Theorem 15.2, the prime-manner action in `G`** (Proposition 14.2(a), mmd L4192 "`K` acts
in a prime manner on `M_σ`"): for a type-`P` maximal `M` with Hall `κ(M)`-subgroup `K`, every
nontrivial `x ∈ K` has `C_G(x) ⊓ M_σ = K* = C_{M_σ}(K)`.

Unpacks `ActsPrimeOn (M_σ) K` (`typeP_structure` conjunct (a)) — `C_{M_σ}(x) = C_{M_σ}(K)` for
`x ∈ K#` — and rewrites via `hKstar`.  The Hall `(κ ∪ σ)'`-subgroup `U` needed by `typeP_structure`
is built by Hall's theorem in the solvable `M` (`Ch03.hall_E_exists`).  Supplies the `hprime`
hypothesis of `centralizer_le_Msigma_of_primeManner`, `isFrobeniusGroup_DK_of_primeManner`, etc. -/
theorem actsPrimeManner_of_typeP [Finite G]
    (hG : OddOrder.BG.IsMinimalSimpleOdd G) {M K Kstar : Subgroup G}
    (hM : M ∈ maximalSubgroups G) (hP : S14.IsTypeP M) (hKM : K ≤ M)
    (hK : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M))
    (hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G)) :
    ∀ x ∈ K, x ≠ 1 →
      Subgroup.centralizer ({x} : Set G) ⊓ OddOrder.BG.Ch3.S10.Msigma M = Kstar := by
  classical
  have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
  obtain ⟨U', hU'⟩ := Ch03.hall_E_exists (G := ↥M)
    ((S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ)
  have hUeq : (U'.map M.subtype).subgroupOf M = U' :=
    Subgroup.comap_map_eq_self_of_injective M.subtype_injective U'
  have hU : Ch03.IsHallSubgroup ((S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ)
      ((U'.map M.subtype).subgroupOf M) := by rw [hUeq]; exact hU'
  have hprimeOn := (S14.typeP_structure hG hM hP hKM hK hKstar hU).1
  intro x hxK hx1
  have hpm := hprimeOn x hxK hx1
  rw [OddOrder.BG.Ch3.S13.fixedByElement_def, OddOrder.BG.Ch3.S13.fixedBy_def] at hpm
  rw [inf_comm, hpm, hKstar]

/-- **The prime-manner action in the `↥M`-internal form** (Theorem 15.2 step 2 plumbing): the
`↥M`-rephrasing of `actsPrimeManner_of_typeP`, supplying the `hcond2` hypothesis of
`kstar_le_opiCore_of_inputs`.  For `x ∈ (κ(M))^#` (as an `↥M`-element), the centralizer
`C_{↥M}({x}) ⊓ (M_σ ↾ M)` equals `C_{↥M}(K ↾ M) ⊓ (M_σ ↾ M)` — both restrict to `K* ↾ M` via
`centralizer_subgroupOf` (`C_{↥M}(T) = C_G(M.subtype '' T) ↾ M`) and the global prime-manner
identity `C_G(x) ⊓ M_σ = K* = M_σ ⊓ C_G(K)`. -/
theorem actsPrimeManner_subgroupOf_of_typeP [Finite G]
    (hG : OddOrder.BG.IsMinimalSimpleOdd G) {M K Kstar : Subgroup G}
    (hM : M ∈ maximalSubgroups G) (hP : S14.IsTypeP M) (hKM : K ≤ M)
    (hK : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M))
    (hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G)) :
    ∀ x ∈ (K.subgroupOf M : Set ↥M), x ≠ 1 →
      Subgroup.centralizer ({x} : Set ↥M) ⊓ (OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M
        = Subgroup.centralizer (K.subgroupOf M : Set ↥M)
            ⊓ (OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M := by
  classical
  have hprime := actsPrimeManner_of_typeP hG hM hP hKM hK hKstar
  intro x hxK hx1
  have hxKM : x ∈ K.subgroupOf M := hxK
  have hxG : (x : G) ∈ K := Subgroup.mem_subgroupOf.mp hxKM
  have hxG1 : (x : G) ≠ 1 := fun h => hx1 (Subtype.ext h)
  have himg_x : M.subtype '' ({x} : Set ↥M) = ({(x : G)} : Set G) := by
    rw [Set.image_singleton]; rfl
  have himg_K : M.subtype '' (K.subgroupOf M : Set ↥M) = (K : Set G) := by
    rw [← Subgroup.coe_map, Subgroup.map_subgroupOf_eq_of_le hKM]
  have sgInf : ∀ A B : Subgroup G, A.subgroupOf M ⊓ B.subgroupOf M = (A ⊓ B).subgroupOf M :=
    fun A B => (Subgroup.comap_inf A B M.subtype).symm
  -- both sides equal `Kstar.subgroupOf M`.
  have hL : Subgroup.centralizer ({x} : Set ↥M) ⊓ (OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M
      = Kstar.subgroupOf M := by
    rw [OddOrder.BG.Ch1.S03h.centralizer_subgroupOf ({x} : Set ↥M), himg_x, sgInf,
      hprime (x : G) hxG hxG1]
  have hR : Subgroup.centralizer (K.subgroupOf M : Set ↥M)
        ⊓ (OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M = Kstar.subgroupOf M := by
    rw [OddOrder.BG.Ch1.S03h.centralizer_subgroupOf (K.subgroupOf M : Set ↥M), himg_K, sgInf,
      inf_comm, ← hKstar]
  rw [hL, hR]

/-- **The normal `q`-Sylow of `↥M` underlying `Q = O_q(M)`** (Theorem 15.2 step 5 plumbing): when
`Q = O_q(M)` is the normal Sylow `q`-subgroup of `M` (a `q`-group with `q ∤ [M : Q]`), there is a
`Sylow q ↥M` whose image under `M.subtype` is `Q`.  `Q ↾ M` is a `q`-group (`comap_subtype`) of
`q'`-index, hence a Sylow (`IsPGroup.toSylow`); being normal (`Q ⊴ M`), its ambient image is the
singleton core `O_q(M) = Q` (`sylowMap_eq_opiCoreInG_singleton_of_normal`).  Supplies the Sylow
witness `(P, hQP)` of `D_centralizes_Q_of_not_mem_beta`. -/
theorem exists_sylow_eq_opiCore [Finite G] {M Q : Subgroup G} {q : ℕ} [Fact q.Prime]
    (hQ : Q = opiCoreInG ({q} : Set ℕ) M) (hQM : Q ≤ M)
    (hMnormQ : M ≤ Subgroup.normalizer (Q : Set G))
    (hQpg : IsPGroup q ↥Q) (hidx : ¬ q ∣ (Q.subgroupOf M).index) :
    ∃ P : Sylow q ↥M, Q = (P : Subgroup ↥M).map M.subtype := by
  have hQsubpg : IsPGroup q ↥(Q.subgroupOf M) := hQpg.comap_subtype
  refine ⟨hQsubpg.toSylow hidx, ?_⟩
  have hPnorm : (hQsubpg.toSylow hidx : Subgroup ↥M).Normal := by
    rw [hQsubpg.toSylow_coe hidx]
    exact (Subgroup.normal_subgroupOf_iff_le_normalizer hQM).mpr hMnormQ
  rw [OddOrder.BG.Ch3.S10.sylowMap_eq_opiCoreInG_singleton_of_normal _ hPnorm, ← hQ]

/-- **§14-independent Frattini factorization** (BG Corollary 15.3 proof, mmd L4213 "by the
Frattini argument").  If `Q ⊴ M`, `QH ⊴ M`, `Q ∩ H = 1`, `|Q|` and `|H|` are coprime, and `M`
is solvable, then `M = N_M(H)·Q`: every `m ∈ M` factors as `m = n·a` with `n ∈ N_G(H)` and
`a ∈ Q`.

Proof: inside `L = QH`, both `H` and the conjugate `H^{m⁻¹} = m⁻¹Hm` are complements of the
normal subgroup `Q` (coprime orders, so `Q` is a normal Hall subgroup of `L`), hence are
conjugate by some `q ∈ Q` (Schur–Zassenhaus conjugacy, `IsComplement'.exists_conj_of_coprime`):
`H^q = H^{m⁻¹}`.  Then `mq ∈ N_G(H)` and `m = (mq)·q⁻¹`.

This discharges the `hfratt` hypothesis of `mf_hall_centralizer_control_of_inputs` (Cor 15.3)
once Theorem 15.2 supplies the normal `q`-subgroup `Q` with `M_σ/Q` nilpotent (issue 8010). -/
theorem frattini_factorization [Finite G] {M Q H : Subgroup G}
    (hQM : Q ≤ M) (hHM : H ≤ M)
    (hQnorm : (Q.subgroupOf M).Normal)
    (hQHnorm : ((Q ⊔ H).subgroupOf M).Normal)
    (hdisj : Disjoint Q H)
    (hcop : Nat.Coprime (Nat.card ↥Q) (Nat.card ↥H))
    (hsolv : Group.IsSolvable ↥M) :
    ∀ m ∈ M, ∃ n a : G, n ∈ Subgroup.normalizer (H : Set G) ∧ a ∈ Q ∧ m = n * a := by
  classical
  have hQL : Q ≤ Q ⊔ H := le_sup_left
  have hHL : H ≤ Q ⊔ H := le_sup_right
  have hLM : Q ⊔ H ≤ M := sup_le hQM hHM
  have hMNQ : M ≤ Subgroup.normalizer (Q : Set G) :=
    (Subgroup.normal_subgroupOf_iff_le_normalizer hQM).mp hQnorm
  have hMNL : M ≤ Subgroup.normalizer ((Q ⊔ H : Subgroup G) : Set G) :=
    (Subgroup.normal_subgroupOf_iff_le_normalizer hLM).mp hQHnorm
  -- `Q ⊴ L` (as a subgroup of `↥(Q ⊔ H)`).
  have hQnL : (Q.subgroupOf (Q ⊔ H)).Normal :=
    (Subgroup.normal_subgroupOf_iff_le_normalizer hQL).mpr (le_trans hLM hMNQ)
  -- `↥(Q ⊔ H)` is solvable, hence so is its quotient by `Q.subgroupOf (Q ⊔ H)`.
  have hLsolv : Group.IsSolvable ↥(Q ⊔ H) :=
    Group.isSolvable_of_isSolvable_injective (Subgroup.inclusion_injective hLM)
  -- A complement-builder inside `↥(Q ⊔ H)`: a `K ≤ L` with `Q ⊓ K = ⊥` and `Q ⊔ K = Q ⊔ H`
  -- complements `Q` in `L`.
  have mk_compl : ∀ K : Subgroup G, K ≤ Q ⊔ H → (Q ⊓ K : Subgroup G) = ⊥ → Q ⊔ K = Q ⊔ H →
      Subgroup.IsComplement' (Q.subgroupOf (Q ⊔ H)) (K.subgroupOf (Q ⊔ H)) := by
    intro K hKL hQK hQKL
    apply Subgroup.isComplement'_of_disjoint_and_mul_eq_univ
    · rw [Subgroup.disjoint_def]
      intro x hxQ hxK
      rw [Subgroup.mem_subgroupOf] at hxQ hxK
      have hxQK : (x : G) ∈ Q ⊓ K := ⟨hxQ, hxK⟩
      rw [hQK, Subgroup.mem_bot] at hxQK
      exact Subtype.ext hxQK
    · have hsup : Q.subgroupOf (Q ⊔ H) ⊔ K.subgroupOf (Q ⊔ H) = ⊤ := by
        rw [← Subgroup.subgroupOf_sup hQL hKL, hQKL, Subgroup.subgroupOf_self]
      have := Subgroup.normal_mul (Q.subgroupOf (Q ⊔ H)) (K.subgroupOf (Q ⊔ H))
      rw [hsup, Subgroup.coe_top] at this
      exact this.symm
  intro m hmM
  have hminvM : m⁻¹ ∈ M := M.inv_mem hmM
  -- `conj m⁻¹` fixes `Q` and `L` (both normal in `M`, and `m⁻¹ ∈ M`).
  have hmiQ : MulAut.conj m⁻¹ • Q = Q := conj_smul_eq_self_of_mem_normalizer (hMNQ hminvM)
  have hmiL : MulAut.conj m⁻¹ • (Q ⊔ H) = Q ⊔ H :=
    conj_smul_eq_self_of_mem_normalizer (hMNL hminvM)
  -- `H' := m⁻¹Hm = conj m⁻¹ • H` is a complement of `Q` in `L`.
  have hH'L : MulAut.conj m⁻¹ • H ≤ Q ⊔ H := by
    rw [← hmiL]; exact Subgroup.pointwise_smul_le_pointwise_smul_iff.mpr hHL
  have hQH'disj : (Q ⊓ MulAut.conj m⁻¹ • H : Subgroup G) = ⊥ := by
    rw [← hmiQ, ← Subgroup.smul_inf, disjoint_iff.mp hdisj, Subgroup.smul_bot]
  have hQH'sup : Q ⊔ MulAut.conj m⁻¹ • H = Q ⊔ H := by
    rw [← hmiQ, ← Subgroup.smul_sup, hmiQ, hmiL]
  -- The two complements `H` and `H'`.
  have hcompl_H : Subgroup.IsComplement' (Q.subgroupOf (Q ⊔ H)) (H.subgroupOf (Q ⊔ H)) :=
    mk_compl H hHL (disjoint_iff.mp hdisj) rfl
  have hcompl_H' : Subgroup.IsComplement' (Q.subgroupOf (Q ⊔ H))
      ((MulAut.conj m⁻¹ • H).subgroupOf (Q ⊔ H)) :=
    mk_compl (MulAut.conj m⁻¹ • H) hH'L hQH'disj hQH'sup
  -- Coprimality `(|Q.subgroupOf L|, (Q.subgroupOf L).index)`: index `= |H|` by `hcompl_H`.
  have hcardQ : Nat.card ↥(Q.subgroupOf (Q ⊔ H)) = Nat.card ↥Q :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hQL).toEquiv
  have hcardH : Nat.card ↥(H.subgroupOf (Q ⊔ H)) = Nat.card ↥H :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hHL).toEquiv
  have hcop' : Nat.Coprime (Nat.card ↥(Q.subgroupOf (Q ⊔ H)))
      (Q.subgroupOf (Q ⊔ H)).index := by
    rw [hcardQ, hcompl_H.symm.index_eq_card, hcardH]; exact hcop
  -- Schur–Zassenhaus conjugacy: `H` and `H'` are conjugate by some `q' ∈ Q.subgroupOf L`.
  obtain ⟨q', hq'mem, hq'eq⟩ :=
    Subgroup.IsComplement'.exists_conj_of_coprime hcop' (Or.inr inferInstance) hcompl_H hcompl_H'
  -- Lift the conjugacy back to `G` along `L.subtype`: `conj q'.val • H = conj m⁻¹ • H`.
  set q : G := (q' : G) with hqdef
  have hqQ : q ∈ Q := Subgroup.mem_subgroupOf.mp hq'mem
  have hintertwine : (Q ⊔ H).subtype.comp (MulAut.conj q').toMonoidHom =
      ((MulAut.conj q).toMonoidHom).comp (Q ⊔ H).subtype := by
    ext x; rfl
  have hmapH : (H.subgroupOf (Q ⊔ H)).map (Q ⊔ H).subtype = H := by
    rw [Subgroup.subgroupOf, Subgroup.map_comap_eq, Subgroup.range_subtype, inf_of_le_right hHL]
  have hmapH' : ((MulAut.conj m⁻¹ • H).subgroupOf (Q ⊔ H)).map (Q ⊔ H).subtype =
      MulAut.conj m⁻¹ • H := by
    rw [Subgroup.subgroupOf, Subgroup.map_comap_eq, Subgroup.range_subtype, inf_of_le_right hH'L]
  have hlift : MulAut.conj q • H = MulAut.conj m⁻¹ • H := by
    have hlifted := congrArg (·.map (Q ⊔ H).subtype) hq'eq
    rw [Subgroup.map_map, hintertwine, ← Subgroup.map_map, hmapH, hmapH'] at hlifted
    rw [Subgroup.pointwise_smul_def]
    exact hlifted
  -- `n := m·q ∈ N_G(H)`, `a := q⁻¹ ∈ Q`, `m = n·a`.
  refine ⟨m * q, q⁻¹, ?_, Q.inv_mem hqQ, by group⟩
  apply mem_normalizer_of_conj_smul_eq_self
  rw [map_mul, mul_smul, hlift, ← mul_smul, ← map_mul, mul_inv_cancel, map_one, one_smul]

/-- **A Hall subgroup of a nilpotent `M_σ` is normal in `M`** (BG Corollary 15.3(b) entry, mmd
L4213 "Then `M_σ` is not nilpotent").  If `M_σ = O_σ(M)` is nilpotent and `H` is a Hall subgroup of
`M_σ`, then `H = O_{π(H)}(M_σ)` (the `π`-core of the nilpotent `M_σ`, hence characteristic), so the
`M`-normalizer of `M_σ` normalizes `H`; thus `H ⊴ M`.

Contrapositive: if `H ⋬ M` (the case `hfratt` handles), then `M_σ` is **not** nilpotent, i.e.
`M_F ≠ M_σ`, which lets Theorem 15.2 supply the normal `q`-subgroup `Q`. -/
theorem hall_subgroupOf_normal_of_msigma_nilpotent [Finite G]
    {M H : Subgroup G} (hHMσ : H ≤ OddOrder.BG.Ch3.S10.Msigma M)
    (hHhall : Ch03.IsHallSubgroup (S14.piSet H) (H.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)))
    (hnil : Group.IsNilpotent ↥(OddOrder.BG.Ch3.S10.Msigma M)) :
    (H.subgroupOf M).Normal := by
  classical
  set Mσ : Subgroup G := OddOrder.BG.Ch3.S10.Msigma M with hMσdef
  set π : Set ℕ := S14.piSet H with hπdef
  have hHM : H ≤ M := hHMσ.trans (OddOrder.BG.Ch3.S10.Msigma_le M)
  -- `H ≤ O_π(M_σ)` (a `π`-subgroup of the nilpotent `M_σ`).
  have hH_le_Oπ : H ≤ OddOrder.GroupTheory.opiCoreInG π Mσ :=
    OddOrder.BG.Ch3.S12.piGroup_le_opiCoreInG_of_nilpotent (fun r hr => hr) hHMσ
  -- `O_π(M_σ)` is a `π`-group, and `H` is Hall `π` in `M_σ`, so `|O_π| ∣ |H|`; with `H ≤ O_π`,
  -- `H = O_π(M_σ)`.
  have hOπMσ : OddOrder.GroupTheory.opiCoreInG π Mσ ≤ Mσ := OddOrder.GroupTheory.opiCoreInG_le _ _
  have hOπ_pi_Mσ : Ch03.Subgroup.IsPiGroup π
      ((OddOrder.GroupTheory.opiCoreInG π Mσ).subgroupOf Mσ) := by
    intro r hr
    rw [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hOπMσ).toEquiv] at hr
    exact OddOrder.GroupTheory.isPiSubgroup_opiCoreInG π Mσ r hr
  have hcard_dvd : Nat.card ↥((OddOrder.GroupTheory.opiCoreInG π Mσ).subgroupOf Mσ)
      ∣ Nat.card ↥(H.subgroupOf Mσ) := hHhall.card_dvd_of_isPiGroup hOπ_pi_Mσ
  have hcard_dvd' : Nat.card ↥(OddOrder.GroupTheory.opiCoreInG π Mσ) ∣ Nat.card ↥H := by
    rwa [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hOπMσ).toEquiv,
      Nat.card_congr (Subgroup.subgroupOfEquivOfLe hHMσ).toEquiv] at hcard_dvd
  have hHeq : H = OddOrder.GroupTheory.opiCoreInG π Mσ :=
    Subgroup.eq_of_le_of_card_ge hH_le_Oπ (Nat.le_of_dvd Nat.card_pos hcard_dvd')
  -- `M ≤ N_G(M_σ) ⟹ M ≤ N_G(O_π(M_σ)) = N_G(H)`, hence `H.subgroupOf M ⊴ M`.
  have hM_norm_Mσ : M ≤ Subgroup.normalizer (Mσ : Set G) := by
    rw [hMσdef, OddOrder.BG.Ch3.S10.Msigma]
    exact OddOrder.GroupTheory.le_normalizer_opiCoreInG (OddOrder.BG.Ch3.S10.sigma M) M
  have hM_norm_H : M ≤ Subgroup.normalizer (H : Set G) := by
    rw [hHeq]
    exact OddOrder.GroupTheory.le_normalizer_opiCoreInG_of_le_normalizer π hM_norm_Mσ
  exact (Subgroup.normal_subgroupOf_iff_le_normalizer hHM).mpr hM_norm_H

/-- **A Hall `π`-subgroup of a finite nilpotent group is its `π`-core** (hence characteristic).
General-`Γ` form of the core step inside `hall_subgroupOf_normal_of_msigma_nilpotent`: in a
nilpotent group every `π`-subgroup lies in `O_π(Γ)` (`isPiGroup_le_of_normal_isHallSubgroup`,
`O_π` being a normal Hall `π`-subgroup), and a Hall `π`-subgroup exhausts `O_π` by orders. -/
theorem isHallSubgroup_eq_oPiCore_of_nilpotent {Γ : Type*} [Group Γ] [Finite Γ]
    [Group.IsNilpotent Γ] {π : Set ℕ} {H : Subgroup Γ} (hHall : Ch03.IsHallSubgroup π H) :
    H = Ch03.oPiCore π Γ := by
  have hOHall : Ch03.IsHallSubgroup π (Ch03.oPiCore π Γ) :=
    OddOrder.BG.Ch3.S10.oPiCore_isHall_of_isNilpotent π
  have hOpi : Ch03.Subgroup.IsPiGroup π (Ch03.oPiCore π Γ) := Ch03.oPiCore.isPiGroup π
  have hHpi : Ch03.Subgroup.IsPiGroup π H := fun p hp => hHall.1 p hp
  have : (Ch03.oPiCore π Γ).Normal := inferInstance
  have hHle : H ≤ Ch03.oPiCore π Γ :=
    OddOrder.BG.Ch3.S10.isPiGroup_le_of_normal_isHallSubgroup hOHall hHpi
  have hdvd : Nat.card ↥(Ch03.oPiCore π Γ) ∣ Nat.card ↥H := hHall.card_dvd_of_isPiGroup hOpi
  exact Subgroup.eq_of_le_of_card_ge hHle (Nat.le_of_dvd Nat.card_pos hdvd)

/-- **The image of a Hall `π`-subgroup under a quotient map is Hall `π` in the quotient.**
`|H̄| ∣ |H|` keeps the `π`-group property; and `[Γ/N : H̄] = [Γ : HN]` divides `[Γ : H]`
(as `H ≤ HN`), a `π'`-number. -/
theorem isHallSubgroup_map_mk' {Γ : Type*} [Group Γ] [Finite Γ] {N : Subgroup Γ} [N.Normal]
    {π : Set ℕ} {H : Subgroup Γ} (hHall : Ch03.IsHallSubgroup π H) :
    Ch03.IsHallSubgroup π (H.map (QuotientGroup.mk' N)) := by
  refine ⟨?_, ?_⟩
  · -- `π`-group: `|H̄| ∣ |H|`.
    intro p hp
    have hdvd : Nat.card ↥(H.map (QuotientGroup.mk' N)) ∣ Nat.card ↥H :=
      Subgroup.card_map_dvd _ _
    rw [Nat.mem_primeFactors] at hp
    exact hHall.1 p (Nat.mem_primeFactors.mpr ⟨hp.1, hp.2.1.trans hdvd, Nat.card_pos.ne'⟩)
  · -- `π'`-index: `[Γ/N : H̄] = [Γ : NH] ∣ [Γ : H]`.
    intro p hp
    have hidx : (H.map (QuotientGroup.mk' N)).index = (N ⊔ H).index := by
      rw [← QuotientGroup.comap_map_mk' N H,
        Subgroup.index_comap_of_surjective _ (QuotientGroup.mk'_surjective N)]
    have hdvd : (N ⊔ H).index ∣ H.index := Subgroup.index_dvd_of_le le_sup_right
    rw [hidx, Nat.mem_primeFactors] at hp
    exact hHall.2 p (Nat.mem_primeFactors.mpr
      ⟨hp.1, hp.2.1.trans hdvd, Subgroup.index_ne_zero_of_finite⟩)

/-- **`N ⊔ H` is characteristic when `Γ/N` is nilpotent** (`N` characteristic, `H` a Hall
`π`-subgroup).  The image `H̄ = H.map (mk' N)` is the Hall `π`-subgroup of the nilpotent `Γ/N`,
hence equals `O_π(Γ/N)` (characteristic, `isHallSubgroup_eq_oPiCore_of_nilpotent`), and
`N ⊔ H = (mk' N)⁻¹(H̄)` is characteristic as the preimage of a characteristic subgroup under the
quotient by the characteristic `N` (`Subgroup.Characteristic.comap_quotient_mk`).

This is the BG Corollary 15.3 step "`QH ◁ M` because `M_σ/Q` is nilpotent" (mmd L4213), applied
with `Γ = ↥M_σ`, `N = Q.subgroupOf M_σ` (`= O_q(M_σ)`, characteristic), `H = H.subgroupOf M_σ`. -/
theorem characteristic_sup_hall_of_quotient_nilpotent {Γ : Type*} [Group Γ] [Finite Γ]
    {N : Subgroup Γ} [N.Characteristic] (hNil : Group.IsNilpotent (Γ ⧸ N)) {π : Set ℕ}
    {H : Subgroup Γ} (hHall : Ch03.IsHallSubgroup π H) : (N ⊔ H).Characteristic := by
  have := hNil
  have hHbar : Ch03.IsHallSubgroup π (H.map (QuotientGroup.mk' N)) := isHallSubgroup_map_mk' hHall
  have hHbar_eq : H.map (QuotientGroup.mk' N) = Ch03.oPiCore π (Γ ⧸ N) :=
    isHallSubgroup_eq_oPiCore_of_nilpotent hHbar
  have hHbar_char : (H.map (QuotientGroup.mk' N)).Characteristic := by
    rw [hHbar_eq]; exact Ch03.oPiCore.characteristic π (Γ ⧸ N)
  rw [← QuotientGroup.comap_map_mk' N H]
  exact Subgroup.Characteristic.comap_quotient_mk hHbar_char

/-- **`O_π(M) = O_π(N)`** when `N ◁ M` (in the ambient sense `M ≤ N_G(N)`) and `O_π(M) ≤ N`
(BG Corollary 15.3 plumbing).  `O_π(M)` is a normal `π`-subgroup of `N`, so `≤ O_π(N)`;
conversely `O_π(N)` is characteristic in `N`, hence (as `N ◁ M`) a normal `π`-subgroup of `M`,
so `≤ O_π(M)`.  Used to identify `Q = O_q(M)` (which lies in `M_σ`) with `O_q(M_σ)`, which makes
`Q.subgroupOf M_σ` characteristic in `↥M_σ`. -/
theorem opiCoreInG_eq_of_normal_le [Finite G] {π : Set ℕ} {M N : Subgroup G}
    (hNM : N ≤ M) (hMN : M ≤ Subgroup.normalizer (N : Set G))
    (hQN : opiCoreInG π M ≤ N) :
    opiCoreInG π M = opiCoreInG π N := by
  refine le_antisymm ?_ ?_
  · exact le_opiCoreInG_of_normal_of_isPiSubgroup hQN
      ((Subgroup.normal_subgroupOf_iff_le_normalizer hQN).mpr
        (hNM.trans (le_normalizer_opiCoreInG π M))) (isPiSubgroup_opiCoreInG π M)
  · have hON_le_M : opiCoreInG π N ≤ M := (opiCoreInG_le π N).trans hNM
    exact le_opiCoreInG_of_normal_of_isPiSubgroup hON_le_M
      ((Subgroup.normal_subgroupOf_iff_le_normalizer hON_le_M).mpr
        (le_normalizer_opiCoreInG_of_le_normalizer π hMN)) (isPiSubgroup_opiCoreInG π N)

/-- **A characteristic subgroup of an `M`-normal subgroup is `M`-normal** (relative version of the
private `normal_of_characteristic_subgroupOf`, for `N ◁ M` rather than `N ◁ G`).  If `N ≤ M` with
`M ≤ N_G(N)`, `A ≤ N`, and `A.subgroupOf N` is characteristic in `↥N`, then `A.subgroupOf M` is
normal in `↥M`.  Each `x ∈ M` restricts to an automorphism `ψ` of `↥N`; it fixes the characteristic
`A.subgroupOf N`, so conjugation by `x` fixes `A = (A.subgroupOf N).map N.subtype`.  This lifts the
`QH` characteristic-in-`M_σ` step (`characteristic_sup_hall_of_quotient_nilpotent`) to `QH ◁ M`. -/
theorem normal_subgroupOf_of_characteristic_subgroupOf_le {M N A : Subgroup G}
    (hNM : N ≤ M) (hMN : M ≤ Subgroup.normalizer (N : Set G)) (hAN : A ≤ N)
    (hchar : (A.subgroupOf N).Characteristic) : (A.subgroupOf M).Normal := by
  refine (Subgroup.normal_subgroupOf_iff_le_normalizer (hAN.trans hNM)).mpr ?_
  intro x hx
  rw [Subgroup.mem_normalizer_iff]
  intro y
  have hxN : x ∈ Subgroup.normalizer (N : Set G) := hMN hx
  let ψ : MulAut ↥N := N.normalizerMonoidHom ⟨x, hxN⟩
  have hAeq : (A.subgroupOf N).map N.subtype = A := Subgroup.map_subgroupOf_eq_of_le hAN
  have hfix := Subgroup.characteristic_iff_comap_eq.mp hchar ψ
  rw [← hAeq]
  constructor
  · rintro ⟨z, hz, rfl⟩
    refine ⟨ψ z, ?_, ?_⟩
    · have hzc : z ∈ (A.subgroupOf N).comap ψ.toMonoidHom := by rw [hfix]; exact hz
      exact Subgroup.mem_comap.mp hzc
    · change (ψ z : G) = x * (z : G) * x⁻¹
      rfl
  · intro hy
    obtain ⟨z, hz, hz_eq⟩ := hy
    have hyN : y ∈ N := (Subgroup.mem_normalizer_iff.mp hxN y).mpr (hz_eq ▸ z.2)
    have hzy : z = ψ.toMonoidHom ⟨y, hyN⟩ := by
      apply Subtype.ext; change (z : G) = x * y * x⁻¹; exact hz_eq
    have hyc : (⟨y, hyN⟩ : ↥N) ∈ (A.subgroupOf N).comap ψ.toMonoidHom := by
      rw [Subgroup.mem_comap, ← hzy]; exact hz
    rw [hfix] at hyc
    exact ⟨⟨y, hyN⟩, hyc, rfl⟩

/-- **A normal `π`-subgroup is contained in every Hall `π`-subgroup** (finite group).  `L ⊔ K`
is a `π`-group (`L` normal, so `↑L · ↑K = ↑(L ⊔ K)`), and `K` is a maximal `π`-subgroup (Hall),
so `L ⊔ K = K`.  Dual of `isPiGroup_le_of_normal_isHallSubgroup` (which assumes the *Hall*
subgroup normal); here the normal `π`-subgroup `L` need not be Hall and `K` need not be normal.
Used in BG Corollary 15.3(b) to place `Q = O_q(M_σ) ≤ H` when `q ∈ π(H)`. -/
theorem normal_isPiGroup_le_isHall {Γ : Type*} [Group Γ] [Finite Γ] {π : Set ℕ}
    {L K : Subgroup Γ} [L.Normal] (hL : Ch03.Subgroup.IsPiGroup π L)
    (hK : Ch03.IsHallSubgroup π K) : L ≤ K := by
  have hSup_pi : Ch03.Subgroup.IsPiGroup π (L ⊔ K : Subgroup Γ) := by
    intro r hr
    rw [Nat.mem_primeFactors] at hr
    obtain ⟨hr_prime, hr_dvd, _⟩ := hr
    have h_card_eq : Nat.card ↥(L ⊔ K : Subgroup Γ) * Nat.card ↥(L ⊓ K : Subgroup Γ)
        = Nat.card ↥L * Nat.card ↥K := by
      have h_hk := Subgroup.card_HK_mul_card_inf_eq_card_mul_card L K
      rwa [show (↑L * ↑K : Set Γ) = ↑(L ⊔ K : Subgroup Γ) from
        (Subgroup.normal_mul L K).symm] at h_hk
    have h_dvd_prod : r ∣ Nat.card ↥L * Nat.card ↥K := by
      rw [← h_card_eq]; exact hr_dvd.mul_right _
    rcases hr_prime.dvd_mul.mp h_dvd_prod with hL_dvd | hK_dvd
    · exact hL r (Nat.mem_primeFactors.mpr ⟨hr_prime, hL_dvd, Nat.card_pos.ne'⟩)
    · exact hK.1 r (Nat.mem_primeFactors.mpr ⟨hr_prime, hK_dvd, Nat.card_pos.ne'⟩)
  have h_card_dvd : Nat.card ↥(L ⊔ K : Subgroup Γ) ∣ Nat.card ↥K :=
    hK.card_dvd_of_isPiGroup hSup_pi
  have h_card_ge : Nat.card ↥K ≤ Nat.card ↥(L ⊔ K : Subgroup Γ) :=
    Nat.card_le_card_of_injective _ (Subgroup.inclusion_injective (le_sup_right (a := L)))
  have h_sup_eq : (L ⊔ K : Subgroup Γ) = K :=
    (Subgroup.eq_of_le_of_card_ge le_sup_right
      (Nat.le_antisymm (Nat.le_of_dvd Nat.card_pos h_card_dvd) h_card_ge).le).symm
  intro x hx
  have hx_sup : x ∈ (L ⊔ K : Subgroup Γ) := Subgroup.mem_sup_left hx
  rwa [h_sup_eq] at hx_sup

/-- **§14-independent assembly engine for BG Corollary 15.3** (mmd L4204).  Packages the
*logic* of Corollary 15.3 with its `§14`/`§15` inputs taken as named hypotheses, so that once
those land (Lane H), the wrapper `mf_hall_centralizer_control` discharges each by a single
citation and applies this skeleton (the gated-endpoint pattern, cf.
`typeP_kstar_in_mf_of_inputs`).  Hypothesis provenance (mmd L4209-4213):

* `ha` (the `C_M(H) = C_{M_σ}(H)·X` decomposition) ← Proposition 14.2(b1)(e) (`C_M(H)` is a
  `κ(M)'`-group) + Lemma 15.1(c) (the `X` cyclic-`τ₂` extraction);
* `hconj` (any `G`-conjugacy of `H`-elements is realized inside `M`) ← Theorem 14.4 (find
  `c ∈ C_G(x)` with `M^{gc} = M`) + self-normalizing `N_G(M) = M`
  (`normalizer_eq_self_of_mem_maximalSubgroups`); the witness is `m = gc`;
* `hfratt` (the Frattini factorization `M = N_M(H)·Q` when `H ⋬ M`) ← Theorem 15.2's normal
  `Q = O_q(M)` with `M_σ/Q` nilpotent (so `QH ⊴ M`, `Q ∩ H = 1`) + the Frattini argument.

The nontrivial step is the `H ⋬ M` glue: writing the realizing `m = n·a` (`n ∈ N_M(H)`,
`a ∈ Q`), the element `w = a x a⁻¹` lies in `H` (it is `n⁻¹ y n`) and `w x⁻¹ ∈ Q` (`Q ⊴ M`,
`x ∈ M`), so `w x⁻¹ ∈ Q ∩ H = 1`, whence `w = x` and `y = n x n⁻¹`. -/
theorem mf_hall_centralizer_control_of_inputs [Finite G]
    {M H : Subgroup G} (hHM : H ≤ M)
    (ha : ∃ X : Subgroup G, IsCyclic ↥X ∧ (↑(Nat.card ↥X).primeFactors ⊆ tau2 M) ∧
      Subgroup.centralizer (H : Set G) ⊓ M =
        (Subgroup.centralizer (H : Set G) ⊓ OddOrder.BG.Ch3.S10.Msigma M) ⊔ X)
    (hconj : ∀ x ∈ H, ∀ y ∈ H, ∀ g : G, y = g * x * g⁻¹ → ∃ m ∈ M, y = m * x * m⁻¹)
    (hfratt : ¬ (H.subgroupOf M).Normal → ∃ Q : Subgroup G, Q ≤ M ∧ (Q.subgroupOf M).Normal ∧
      Disjoint Q H ∧
      ∀ m ∈ M, ∃ n a : G, n ∈ Subgroup.normalizer (H : Set G) ∧ a ∈ Q ∧ m = n * a) :
    (∃ X : Subgroup G, IsCyclic ↥X ∧ (↑(Nat.card ↥X).primeFactors ⊆ tau2 M) ∧
      Subgroup.centralizer (H : Set G) ⊓ M =
        (Subgroup.centralizer (H : Set G) ⊓ OddOrder.BG.Ch3.S10.Msigma M) ⊔ X) ∧
    (∀ x ∈ H, ∀ y ∈ H, (∃ g : G, y = g * x * g⁻¹) →
      ∃ n ∈ Subgroup.normalizer (H : Set G), y = n * x * n⁻¹) := by
  refine ⟨ha, ?_⟩
  rintro x hx y hy ⟨g, hg⟩
  obtain ⟨m, hmM, hmy⟩ := hconj x hx y hy g hg
  by_cases hHnorm : (H.subgroupOf M).Normal
  · -- `H ⊴ M`:  `m ∈ M ⊆ N_G(H)`, so `n = m` works directly.
    exact ⟨m, ((Subgroup.normal_subgroupOf_iff_le_normalizer hHM).mp hHnorm) hmM, hmy⟩
  · -- `H ⋬ M`:  Frattini-factor `m = n·a` and run the commutator argument.
    obtain ⟨Q, hQM, hQnorm, hQH, hfact⟩ := hfratt hHnorm
    obtain ⟨n, a, hnN, haQ, hmna⟩ := hfact m hmM
    refine ⟨n, hnN, ?_⟩
    -- `y = n · (a x a⁻¹) · n⁻¹`.
    have hyw : y = n * (a * x * a⁻¹) * n⁻¹ := by rw [hmy, hmna]; group
    -- `w := a x a⁻¹ = n⁻¹ y n ∈ H`.
    have hwH : a * x * a⁻¹ ∈ H := by
      have hmem := (Subgroup.mem_normalizer_iff.mp
        ((Subgroup.normalizer (H : Set G)).inv_mem hnN) y).mp hy
      rw [inv_inv] at hmem
      have hweq : a * x * a⁻¹ = n⁻¹ * y * n := by rw [hyw]; group
      rw [hweq]; exact hmem
    -- `w x⁻¹ = a (x a⁻¹ x⁻¹) ∈ Q`  (`Q ⊴ M`, `x ∈ M`).
    have hwxQ : (a * x * a⁻¹) * x⁻¹ ∈ Q := by
      have hxnorm : x ∈ Subgroup.normalizer (Q : Set G) :=
        ((Subgroup.normal_subgroupOf_iff_le_normalizer hQM).mp hQnorm) (hHM hx)
      have hxax : x * a⁻¹ * x⁻¹ ∈ Q :=
        (Subgroup.mem_normalizer_iff.mp hxnorm a⁻¹).mp (Q.inv_mem haQ)
      have hrw : (a * x * a⁻¹) * x⁻¹ = a * (x * a⁻¹ * x⁻¹) := by group
      rw [hrw]; exact Subgroup.mul_mem _ haQ hxax
    -- `w x⁻¹ ∈ Q ∩ H = ⊥`, so `w = x` and `y = n x n⁻¹`.
    have hwxH : (a * x * a⁻¹) * x⁻¹ ∈ H := Subgroup.mul_mem _ hwH (H.inv_mem hx)
    have hwx1 : (a * x * a⁻¹) * x⁻¹ = 1 := by
      have : (a * x * a⁻¹) * x⁻¹ ∈ Q ⊓ H := Subgroup.mem_inf.mpr ⟨hwxQ, hwxH⟩
      rw [disjoint_iff.mp hQH] at this
      exact Subgroup.mem_bot.mp this
    have hwx : a * x * a⁻¹ = x := by
      have := mul_eq_one_iff_eq_inv.mp hwx1
      rw [this]; group
    rw [hyw, hwx]

/-- **BG Corollary 15.3(a) for `H = M_σ`** (mmd L4204-4209), *sorry-free*.  The centralizer
`C_M(M_σ)` decomposes as `(C_G(M_σ) ⊓ M_σ) ⊔ X` with `X` a cyclic `τ₂(M)`-subgroup.

This is exactly the `ha` input that `fitting_decomposition` (Corollary 15.5) consumes at
`H = M_σ`; using it in place of the then-sorried general `mf_hall_centralizer_control` made the
A(8) `FittingIsTI` chain axiom-clean (issue 8016; the general lemma has since been proved too).

Proof (BG L4209).  `C := C_M(M_σ)` is a `κ(M)'`-group
(`centralizer_msigma_isPiSubgroup_kappa_compl`, = Prop 14.2(b1)(e)).  Its normal Hall
`σ`-subgroup `N = M_σ ⊓ C` has a complement `X` by Schur–Zassenhaus: `[C : N] = M_σ.relIndex C`
divides `[M : M_σ]` (`relIndex_subgroupOf` + `relIndex_dvd_index_of_normal`, a `σ'`-number),
coprime to `|N| ∣ |M_σ|`.  Then `C = N ⊔ X` and `X` is a `(κ∪σ)'`-group (`|X| = [C:N]` is
`σ'`, and `X ≤ C` is `κ'`).  As `X ≤ M`, Hall's theorem (`hall_D`) embeds it in a Hall
`(κ∪σ)'`-subgroup `U` of `M`; `X` centralizes `M_σ` (so `C_{M_σ}(X) = M_σ ≠ 1`), and Lemma
15.1(c) (`typeP_hall_small_subgroup_cyclic_tau2`) makes `X` cyclic `τ₂`. -/
theorem mf_centralizer_msigma_decomp [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
    {M : Subgroup G} (hM : M ∈ maximalSubgroups G) :
    ∃ X : Subgroup G, IsCyclic ↥X ∧ (↑(Nat.card ↥X).primeFactors ⊆ tau2 M) ∧
      Subgroup.centralizer (OddOrder.BG.Ch3.S10.Msigma M : Set G) ⊓ M =
        (Subgroup.centralizer (OddOrder.BG.Ch3.S10.Msigma M : Set G) ⊓
          OddOrder.BG.Ch3.S10.Msigma M) ⊔ X := by
  classical
  have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
  set Mσ := OddOrder.BG.Ch3.S10.Msigma M with hMσdef
  set C : Subgroup G := Subgroup.centralizer (Mσ : Set G) ⊓ M with hCdef
  have hC_le_M : C ≤ M := inf_le_right
  have hC_le_cent : C ≤ Subgroup.centralizer (Mσ : Set G) := inf_le_left
  -- `M ≤ N_G(M_σ)`, so `C ≤ N_G(M_σ)`, hence `N := M_σ.subgroupOf C ⊴ C`.
  have hM_norm_Mσ : M ≤ Subgroup.normalizer (Mσ : Set G) := by
    rw [hMσdef, OddOrder.BG.Ch3.S10.Msigma]
    exact le_normalizer_opiCoreInG (OddOrder.BG.Ch3.S10.sigma M) M
  have hMσinfC_le_C : Mσ ⊓ C ≤ C := inf_le_right
  have hC_norm_MσinfC : C ≤ Subgroup.normalizer ((Mσ ⊓ C : Subgroup G) : Set G) := by
    have h1 : C ≤ Subgroup.normalizer (Mσ : Set G) := hC_le_M.trans hM_norm_Mσ
    exact (le_inf h1 Subgroup.le_normalizer).trans Subgroup.inf_normalizer_le_normalizer_inf
  have hN_normal : ((Mσ ⊓ C).subgroupOf C).Normal :=
    (Subgroup.normal_subgroupOf_iff_le_normalizer hMσinfC_le_C).mpr hC_norm_MσinfC
  set N : Subgroup ↥C := (Mσ ⊓ C).subgroupOf C with hNdef
  have hN_eq_Mσ : N = Mσ.subgroupOf C := by rw [hNdef, Subgroup.inf_subgroupOf_right]
  -- `|N| = |M_σ ⊓ C|` (a `σ`-number).
  have hNcard : Nat.card ↥N = Nat.card ↥(Mσ ⊓ C : Subgroup G) :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hMσinfC_le_C).toEquiv
  have hN_pi : ∀ p ∈ (Nat.card ↥N).primeFactors, p ∈ OddOrder.BG.Ch3.S10.sigma M := by
    intro p hp
    rw [hNcard] at hp
    exact OddOrder.BG.Ch3.S10.Msigma_isPiGroup M p
      (Nat.primeFactors_mono (Subgroup.card_dvd_of_le (inf_le_left : Mσ ⊓ C ≤ Mσ))
        Nat.card_pos.ne' hp)
  -- `N.index = M_σ.relIndex C ∣ (M_σ.subgroupOf M).index`, a `σ'`-number.
  have hMσM_normal : ((Mσ).subgroupOf M).Normal := by
    rw [hMσdef, OddOrder.BG.Ch3.S10.Msigma_subgroupOf]; infer_instance
  have hNindex_dvd : N.index ∣ (Mσ.subgroupOf M).index := by
    have hNi : N.index = Mσ.relIndex C := by rw [hN_eq_Mσ]; rfl
    rw [hNi, ← Subgroup.relIndex_subgroupOf hC_le_M]
    exact Subgroup.relIndex_dvd_index_of_normal (Mσ.subgroupOf M) (C.subgroupOf M)
  have hMσM_hall : Ch03.IsHallSubgroup (OddOrder.BG.Ch3.S10.sigma M) (Mσ.subgroupOf M) :=
    OddOrder.BG.Ch3.S10.Msigma_subgroupOf_isHall hG hM
  have hNindex_pi' : ∀ p ∈ N.index.primeFactors, p ∉ OddOrder.BG.Ch3.S10.sigma M := by
    intro p hp
    exact hMσM_hall.index_no_pi p (Nat.mem_primeFactors.mpr
      ⟨Nat.prime_of_mem_primeFactors hp,
        (Nat.dvd_of_mem_primeFactors hp).trans hNindex_dvd, Subgroup.index_ne_zero_of_finite⟩)
  -- Coprimality for Schur–Zassenhaus.
  have hcop : Nat.Coprime (Nat.card ↥N) N.index :=
    Ch03.Nat.coprime_of_isPiGroup_of_isPiGroup_compl
      (π := OddOrder.BG.Ch3.S10.sigma M) Nat.card_pos.ne' Subgroup.index_ne_zero_of_finite
      hN_pi hNindex_pi'
  -- Schur–Zassenhaus: complement `H''` of `N` in `↥C`; `X := H''.map C.subtype`.
  obtain ⟨H'', hH''⟩ := Subgroup.exists_right_complement'_of_coprime hcop
  set X : Subgroup G := H''.map C.subtype with hXdef
  have hX_le_C : X ≤ C := hXdef ▸ Subgroup.map_subtype_le H''
  have hX_le_M : X ≤ M := hX_le_C.trans hC_le_M
  -- `C = (M_σ ⊓ C) ⊔ X`.
  have hCeq0 : Mσ ⊓ C ⊔ X = C := by
    have htop : N ⊔ H'' = ⊤ := hH''.sup_eq_top
    have hmap := congrArg (Subgroup.map C.subtype) htop
    rw [Subgroup.map_sup, hNdef, Subgroup.subgroupOf_map_subtype, inf_eq_left.mpr hMσinfC_le_C,
      ← hXdef] at hmap
    rw [hmap, ← Subgroup.subgroupOf_self C, Subgroup.subgroupOf_map_subtype, inf_idem]
  -- `|X| = N.index` (complement card).
  have hXcard : Nat.card ↥X = N.index := by
    rw [hXdef, ← Nat.card_congr
        (Subgroup.equivMapOfInjective H'' C.subtype C.subtype_injective).toEquiv,
      (hH''.symm).index_eq_card]
  -- `X` is a `(κ ∪ σ)'`-group:  `κ'` (from `X ≤ C`) and `σ'` (from `|X| = N.index`).
  have hX_pi : ∀ p ∈ (Nat.card ↥X).primeFactors, p ∈ (S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ :=
      by
    intro p hp
    rw [Set.mem_compl_iff, Set.mem_union]
    push Not
    refine ⟨?_, ?_⟩
    · -- `p ∉ κ(M)`: `X ≤ C`, and `C` is a `κ'`-group.
      have hpC : p ∈ (Nat.card ↥C).primeFactors :=
        Nat.primeFactors_mono (Subgroup.card_dvd_of_le hX_le_C) Nat.card_pos.ne' hp
      have := centralizer_msigma_isPiSubgroup_kappa_compl hG hM
      exact (Set.mem_compl_iff _ _).mp (this p hpC)
    · -- `p ∉ σ(M)`: `|X| = N.index`, a `σ'`-number.
      rw [hXcard] at hp; exact hNindex_pi' p hp
  -- A Hall `(κ ∪ σ)'`-subgroup `U` of `M` containing `X` (Hall D).
  have hX_pi_M : ∀ p ∈ (Nat.card ↥(X.subgroupOf M)).primeFactors,
      p ∈ (S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ := by
    intro p hp
    exact hX_pi p (by rwa [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hX_le_M).toEquiv] at hp)
  obtain ⟨U₀, hU₀hall, hXU₀⟩ := Ch03.hall_D (G := ↥M) hX_pi_M
  set U : Subgroup G := U₀.map M.subtype with hUdef
  have hUof : U.subgroupOf M = U₀ :=
    Subgroup.comap_map_eq_self_of_injective M.subtype_injective U₀
  have hUM : U ≤ M := hUdef ▸ Subgroup.map_subtype_le U₀
  have hUhall : Ch03.IsHallSubgroup ((S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ)
      (U.subgroupOf M) := by rw [hUof]; exact hU₀hall
  have hXU : X ≤ U := by
    have h1 : X.subgroupOf M ≤ U.subgroupOf M := by rw [hUof]; exact hXU₀
    calc X = (X.subgroupOf M).map M.subtype := (Subgroup.map_subgroupOf_eq_of_le hX_le_M).symm
      _ ≤ (U.subgroupOf M).map M.subtype := Subgroup.map_mono h1
      _ = U := Subgroup.map_subgroupOf_eq_of_le hUM
  -- `X` centralizes `M_σ`, so `C_{M_σ}(X) = M_σ ≠ 1`.
  have hX_cent : X ≤ Subgroup.centralizer (Mσ : Set G) := hX_le_C.trans hC_le_cent
  have hMσinfCX : Mσ ⊓ Subgroup.centralizer (X : Set G) = Mσ := by
    rw [inf_eq_left]; exact Subgroup.le_centralizer_iff.mp hX_cent
  have hMσne : Mσ ≠ ⊥ := hMσdef ▸ OddOrder.BG.Ch3.S10.Msigma_ne_bot hG hM
  have hCXne : Mσ ⊓ Subgroup.centralizer (X : Set G) ≠ ⊥ := by rw [hMσinfCX]; exact hMσne
  -- Conclude.
  refine ⟨X, ?_, ?_, ?_⟩
  · -- `X` cyclic.
    by_cases hXbot : X = ⊥
    · rw [hXbot]; infer_instance
    · exact (typeP_hall_small_subgroup_cyclic_tau2 hG hM hUM hUhall hXU hXbot hCXne).2.1
  · -- `π(X) ⊆ τ₂`.
    by_cases hXbot : X = ⊥
    · rw [hXbot]; simp
    · exact (typeP_hall_small_subgroup_cyclic_tau2 hG hM hUM hUhall hXU hXbot hCXne).2.2
  · -- `C_G(M_σ) ⊓ M = (C_G(M_σ) ⊓ M_σ) ⊔ X`.
    have hMσinfC_eq : Mσ ⊓ C = Subgroup.centralizer (Mσ : Set G) ⊓ Mσ := by
      rw [hCdef, inf_comm (Subgroup.centralizer (Mσ : Set G)) M, ← inf_assoc,
        inf_eq_left.mpr (OddOrder.BG.Ch3.S10.Msigma_le M), inf_comm]
    rw [hMσinfC_eq] at hCeq0
    exact hCeq0.symm

/-- **BG Corollary 15.3(a), `ha` input for general Hall `H`** (mmd L4209), reducing to the single
`κ(M)'` fact.  Given that `C_M(H)` is a `κ(M)'`-group (the genuine residual, BG Prop 14.2(b1)/(e)),
the centralizer decomposes as `C_M(H) = C_{M_σ}(H) X` with `X` a cyclic `τ₂(M)`-subgroup.  This
generalizes `mf_centralizer_msigma_decomp` (the `H = M_σ` case): the proof is identical except the
`κ'` step now uses the hypothesis `hkappa` instead of `centralizer_msigma_isPiSubgroup_kappa_compl`,
and `C_{M_σ}(X) ≠ 1` follows from `H ≤ M_σ ⊓ C_G(X)` (as `X ≤ C_G(H)`) with `H ≠ ⊥`. -/
theorem mf_centralizer_hall_decomp_of_kappaCompl [Finite G]
    (hG : OddOrder.BG.IsMinimalSimpleOdd G) {M H : Subgroup G} (hM : M ∈ maximalSubgroups G)
    (hHMσ : H ≤ OddOrder.BG.Ch3.S10.Msigma M) (hHne : H ≠ ⊥)
    (hkappa : Subgroup.IsPiSubgroup (kappa M)ᶜ (Subgroup.centralizer (H : Set G) ⊓ M)) :
    ∃ X : Subgroup G, IsCyclic ↥X ∧ (↑(Nat.card ↥X).primeFactors ⊆ tau2 M) ∧
      Subgroup.centralizer (H : Set G) ⊓ M =
        (Subgroup.centralizer (H : Set G) ⊓ OddOrder.BG.Ch3.S10.Msigma M) ⊔ X := by
  classical
  have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
  set Mσ := OddOrder.BG.Ch3.S10.Msigma M with hMσdef
  set C : Subgroup G := Subgroup.centralizer (H : Set G) ⊓ M with hCdef
  have hC_le_M : C ≤ M := inf_le_right
  have hC_le_cent : C ≤ Subgroup.centralizer (H : Set G) := inf_le_left
  have hM_norm_Mσ : M ≤ Subgroup.normalizer (Mσ : Set G) := by
    rw [hMσdef, OddOrder.BG.Ch3.S10.Msigma]
    exact le_normalizer_opiCoreInG (OddOrder.BG.Ch3.S10.sigma M) M
  have hMσinfC_le_C : Mσ ⊓ C ≤ C := inf_le_right
  have hC_norm_MσinfC : C ≤ Subgroup.normalizer ((Mσ ⊓ C : Subgroup G) : Set G) := by
    have h1 : C ≤ Subgroup.normalizer (Mσ : Set G) := hC_le_M.trans hM_norm_Mσ
    exact (le_inf h1 Subgroup.le_normalizer).trans Subgroup.inf_normalizer_le_normalizer_inf
  have hN_normal : ((Mσ ⊓ C).subgroupOf C).Normal :=
    (Subgroup.normal_subgroupOf_iff_le_normalizer hMσinfC_le_C).mpr hC_norm_MσinfC
  set N : Subgroup ↥C := (Mσ ⊓ C).subgroupOf C with hNdef
  have hN_eq_Mσ : N = Mσ.subgroupOf C := by rw [hNdef, Subgroup.inf_subgroupOf_right]
  have hNcard : Nat.card ↥N = Nat.card ↥(Mσ ⊓ C : Subgroup G) :=
    Nat.card_congr (Subgroup.subgroupOfEquivOfLe hMσinfC_le_C).toEquiv
  have hN_pi : ∀ p ∈ (Nat.card ↥N).primeFactors, p ∈ OddOrder.BG.Ch3.S10.sigma M := by
    intro p hp
    rw [hNcard] at hp
    exact OddOrder.BG.Ch3.S10.Msigma_isPiGroup M p
      (Nat.primeFactors_mono (Subgroup.card_dvd_of_le (inf_le_left : Mσ ⊓ C ≤ Mσ))
        Nat.card_pos.ne' hp)
  have hMσM_normal : ((Mσ).subgroupOf M).Normal := by
    rw [hMσdef, OddOrder.BG.Ch3.S10.Msigma_subgroupOf]; infer_instance
  have hNindex_dvd : N.index ∣ (Mσ.subgroupOf M).index := by
    have hNi : N.index = Mσ.relIndex C := by rw [hN_eq_Mσ]; rfl
    rw [hNi, ← Subgroup.relIndex_subgroupOf hC_le_M]
    exact Subgroup.relIndex_dvd_index_of_normal (Mσ.subgroupOf M) (C.subgroupOf M)
  have hMσM_hall : Ch03.IsHallSubgroup (OddOrder.BG.Ch3.S10.sigma M) (Mσ.subgroupOf M) :=
    OddOrder.BG.Ch3.S10.Msigma_subgroupOf_isHall hG hM
  have hNindex_pi' : ∀ p ∈ N.index.primeFactors, p ∉ OddOrder.BG.Ch3.S10.sigma M := by
    intro p hp
    exact hMσM_hall.index_no_pi p (Nat.mem_primeFactors.mpr
      ⟨Nat.prime_of_mem_primeFactors hp,
        (Nat.dvd_of_mem_primeFactors hp).trans hNindex_dvd, Subgroup.index_ne_zero_of_finite⟩)
  have hcop : Nat.Coprime (Nat.card ↥N) N.index :=
    Ch03.Nat.coprime_of_isPiGroup_of_isPiGroup_compl
      (π := OddOrder.BG.Ch3.S10.sigma M) Nat.card_pos.ne' Subgroup.index_ne_zero_of_finite
      hN_pi hNindex_pi'
  obtain ⟨H'', hH''⟩ := Subgroup.exists_right_complement'_of_coprime hcop
  set X : Subgroup G := H''.map C.subtype with hXdef
  have hX_le_C : X ≤ C := hXdef ▸ Subgroup.map_subtype_le H''
  have hX_le_M : X ≤ M := hX_le_C.trans hC_le_M
  have hCeq0 : Mσ ⊓ C ⊔ X = C := by
    have htop : N ⊔ H'' = ⊤ := hH''.sup_eq_top
    have hmap := congrArg (Subgroup.map C.subtype) htop
    rw [Subgroup.map_sup, hNdef, Subgroup.subgroupOf_map_subtype, inf_eq_left.mpr hMσinfC_le_C,
      ← hXdef] at hmap
    rw [hmap, ← Subgroup.subgroupOf_self C, Subgroup.subgroupOf_map_subtype, inf_idem]
  have hXcard : Nat.card ↥X = N.index := by
    rw [hXdef, ← Nat.card_congr
        (Subgroup.equivMapOfInjective H'' C.subtype C.subtype_injective).toEquiv,
      (hH''.symm).index_eq_card]
  have hX_pi : ∀ p ∈ (Nat.card ↥X).primeFactors,
      p ∈ (S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ := by
    intro p hp
    rw [Set.mem_compl_iff, Set.mem_union]
    push Not
    refine ⟨?_, ?_⟩
    · have hpC : p ∈ (Nat.card ↥C).primeFactors :=
        Nat.primeFactors_mono (Subgroup.card_dvd_of_le hX_le_C) Nat.card_pos.ne' hp
      exact (Set.mem_compl_iff _ _).mp (hkappa p hpC)
    · rw [hXcard] at hp; exact hNindex_pi' p hp
  have hX_pi_M : ∀ p ∈ (Nat.card ↥(X.subgroupOf M)).primeFactors,
      p ∈ (S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ := by
    intro p hp
    exact hX_pi p (by rwa [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hX_le_M).toEquiv] at hp)
  obtain ⟨U₀, hU₀hall, hXU₀⟩ := Ch03.hall_D (G := ↥M) hX_pi_M
  set U : Subgroup G := U₀.map M.subtype with hUdef
  have hUof : U.subgroupOf M = U₀ :=
    Subgroup.comap_map_eq_self_of_injective M.subtype_injective U₀
  have hUM : U ≤ M := hUdef ▸ Subgroup.map_subtype_le U₀
  have hUhall : Ch03.IsHallSubgroup ((S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ)
      (U.subgroupOf M) := by rw [hUof]; exact hU₀hall
  have hXU : X ≤ U := by
    have h1 : X.subgroupOf M ≤ U.subgroupOf M := by rw [hUof]; exact hXU₀
    calc X = (X.subgroupOf M).map M.subtype := (Subgroup.map_subgroupOf_eq_of_le hX_le_M).symm
      _ ≤ (U.subgroupOf M).map M.subtype := Subgroup.map_mono h1
      _ = U := Subgroup.map_subgroupOf_eq_of_le hUM
  -- `X` centralizes `H ≤ M_σ`, so `M_σ ⊓ C_G(X) ⊇ H ≠ 1`.
  have hH_le_CX : H ≤ Subgroup.centralizer (X : Set G) := by
    rw [← Subgroup.le_centralizer_iff]
    exact hX_le_C.trans hC_le_cent
  have hCXne : Mσ ⊓ Subgroup.centralizer (X : Set G) ≠ ⊥ := by
    intro hbot
    exact hHne (le_bot_iff.mp (hbot ▸ le_inf hHMσ hH_le_CX))
  refine ⟨X, ?_, ?_, ?_⟩
  · by_cases hXbot : X = ⊥
    · rw [hXbot]; infer_instance
    · exact (typeP_hall_small_subgroup_cyclic_tau2 hG hM hUM hUhall hXU hXbot hCXne).2.1
  · by_cases hXbot : X = ⊥
    · rw [hXbot]; simp
    · exact (typeP_hall_small_subgroup_cyclic_tau2 hG hM hUM hUhall hXU hXbot hCXne).2.2
  · have hMσinfC_eq : Mσ ⊓ C = Subgroup.centralizer (H : Set G) ⊓ Mσ := by
      rw [hCdef, inf_comm (Subgroup.centralizer (H : Set G)) M, ← inf_assoc,
        inf_eq_left.mpr (OddOrder.BG.Ch3.S10.Msigma_le M), inf_comm]
    rw [hMσinfC_eq] at hCeq0
    exact hCeq0.symm

/-- **BG Corollary 15.3(a) for a general nontrivial Hall subgroup `H` of `M_σ`** (mmd L4209),
*sorry-free* and unconditional.  The `ha` input of `mf_hall_centralizer_control`, with the `κ(M)'`
hypothesis of `mf_centralizer_hall_decomp_of_kappaCompl` now discharged by
`centralizer_hall_isPiSubgroup_kappa_compl` (= BG Prop 14.2(b1)(e), the κ'-fact whose final gate was
Prop 14.2(e) `S ⊄ K*`).  Generalizes `mf_centralizer_msigma_decomp` (`H = M_σ`) to any Hall `H`. -/
theorem mf_centralizer_hall_decomp [Finite G]
    (hG : OddOrder.BG.IsMinimalSimpleOdd G) {M H : Subgroup G} (hM : M ∈ maximalSubgroups G)
    (hHMσ : H ≤ OddOrder.BG.Ch3.S10.Msigma M)
    (hHhall : Ch03.IsHallSubgroup (S14.piSet H) (H.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)))
    (hHne : H ≠ ⊥) :
    ∃ X : Subgroup G, IsCyclic ↥X ∧ (↑(Nat.card ↥X).primeFactors ⊆ tau2 M) ∧
      Subgroup.centralizer (H : Set G) ⊓ M =
        (Subgroup.centralizer (H : Set G) ⊓ OddOrder.BG.Ch3.S10.Msigma M) ⊔ X :=
  mf_centralizer_hall_decomp_of_kappaCompl hG hM hHMσ hHne
    (centralizer_hall_isPiSubgroup_kappa_compl hG hM hHMσ hHhall hHne)

/-- **§15 helper (§14-independent, reusable).**  The `G`-normalizer of `M_σ` is `M`:
`N_G(M_σ) = M`.  Since `M_σ = O_{σ(M)}(M)` is normal in `M` (`le_normalizer_opiCoreInG`),
`M ≤ N_G(M_σ)`; if the containment were proper, maximality would force `N_G(M_σ) = G`, so
`M_σ ⊴ G`, and simplicity would give `M_σ ∈ {⊥, ⊤}` — both excluded (`Msigma_ne_bot`,
`M_σ ≤ M ⊊ G`).  This turns the `N_G(M_σ)`-fusion of Corollary 15.3(b) at `H := M_σ` into the
`M`-fusion of BG Theorem D(1). -/
theorem normalizer_Msigma_eq_self [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
    {M : Subgroup G} (hM : M ∈ maximalSubgroups G) :
    Subgroup.normalizer (OddOrder.BG.Ch3.S10.Msigma M : Set G) = M := by
  have hle : M ≤ Subgroup.normalizer (OddOrder.BG.Ch3.S10.Msigma M : Set G) := by
    rw [OddOrder.BG.Ch3.S10.Msigma]
    exact le_normalizer_opiCoreInG (OddOrder.BG.Ch3.S10.sigma M) M
  refine le_antisymm ?_ hle
  rcases eq_or_lt_of_le hle with heq | hlt
  · exact le_of_eq heq.symm
  · have hnorm : (OddOrder.BG.Ch3.S10.Msigma M).Normal :=
      Subgroup.normalizer_eq_top_iff.mp ((mem_maximalSubgroups.mp hM).2 _ hlt)
    rcases hG.simple.eq_bot_or_eq_top_of_normal _ hnorm with hbot | htop
    · exact absurd hbot (OddOrder.BG.Ch3.S10.Msigma_ne_bot hG hM)
    · exact absurd (top_le_iff.mp (htop ▸ OddOrder.BG.Ch3.S10.Msigma_le M))
        (mem_maximalSubgroups.mp hM).1

end OddOrder.BG.Ch4.S15
Source
Yawara Ishida, Lean 4 formalization of the Feit–Thompson (odd order) theorem, github.com/yawara/odd-order @ 20a6fccbb63c (Apache-2.0), https://github.com/yawara/odd-order/blob/20a6fccbb63c0b138acda9f7fafd798c42f9441b/OddOrder/BG/Ch4_FamilyOfMaximal/S15_MF/Theorem152Helpers.lean

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me