BG §15: Theorem 15.2, §14-gated conditional helpers
DefinitionOddOrder_BG_Ch4_FamilyOfMaximal_S15_MF_Theorem152Helpersfeit-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 , existence of a Sylow subgroup equal to an -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