lambdadelta

[helm.git] / matita / matita / contribs / lambdadelta / basic_2 / computation / cpxs_cpxs.ma

diff --git a/matita/matita/contribs/lambdadelta/basic_2/computation/cpxs_cpxs.ma b/matita/matita/contribs/lambdadelta/basic_2/computation/cpxs_cpxs.ma
index 08696ac9329f8e1db3a32a6a1ef1c9f48b4b3948..8017af6bfab8eacc9b0bb4d49c262252ef73b9ea 100644 (file)
--- a/matita/matita/contribs/lambdadelta/basic_2/computation/cpxs_cpxs.ma
+++ b/matita/matita/contribs/lambdadelta/basic_2/computation/cpxs_cpxs.ma
@@ -38,32 +38,65 @@ theorem cpxs_flat: ∀h,g,I,L,V1,V2,T1,T2. ⦃h, L⦄ ⊢ T1 ➡*[g] T2 →
 @(cpxs_trans … IHV1) -IHV1 /2 width=1/
 qed.
 
-theorem cpxs_beta: ∀h,g,a,L,V1,V2,W,T1,T2.
-                   ⦃h, L.ⓛW⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ V1 ➡*[g] V2 →
-                   ⦃h, L⦄ ⊢ ⓐV1.ⓛ{a}W.T1 ➡*[g] ⓓ{a}V2.T2.
-#h #g #a #L #V1 #V2 #W #T1 #T2 #HT12 #H @(cpxs_ind … H) -V2 /2 width=1/
+theorem cpxs_beta_rc: ∀h,g,a,L,V1,V2,W1,W2,T1,T2.
+                      ⦃h, L⦄ ⊢ V1 ➡[g] V2 → ⦃h, L.ⓛW1⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ W1 ➡*[g] W2 →
+                      ⦃h, L⦄ ⊢ ⓐV1.ⓛ{a}W1.T1 ➡*[g] ⓓ{a}ⓝW2.V2.T2.
+#h #g #a #L #V1 #V2 #W1 #W2 #T1 #T2 #HV12 #HT12 #H @(cpxs_ind … H) -W2 /2 width=1/
+#W #W2 #_ #HW2 #IHW1
+@(cpxs_trans … IHW1) -IHW1 /3 width=1/
+qed.
+
+theorem cpxs_beta: ∀h,g,a,L,V1,V2,W1,W2,T1,T2.
+                   ⦃h, L.ⓛW1⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ W1 ➡*[g] W2 → ⦃h, L⦄ ⊢ V1 ➡*[g] V2 →
+                   ⦃h, L⦄ ⊢ ⓐV1.ⓛ{a}W1.T1 ➡*[g] ⓓ{a}ⓝW2.V2.T2.
+#h #g #a #L #V1 #V2 #W1 #W2 #T1 #T2 #HT12 #HW12 #H @(cpxs_ind … H) -V2 /2 width=1/
 #V #V2 #_ #HV2 #IHV1
-@(cpxs_trans … IHV1) /2 width=1/
+@(cpxs_trans … IHV1) -IHV1 /3 width=1/
 qed.
 
 theorem cpxs_theta_rc: ∀h,g,a,L,V1,V,V2,W1,W2,T1,T2.
                        ⦃h, L⦄ ⊢ V1 ➡[g] V → ⇧[0, 1] V ≡ V2 →
-                      ⦃h, L.ⓓW1⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ W1 ➡*[g] W2 →
-                      ⦃h, L⦄ ⊢ ⓐV1.ⓓ{a}W1.T1 ➡*[g] ⓓ{a}W2.ⓐV2.T2.
+                       ⦃h, L.ⓓW1⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ W1 ➡*[g] W2 →
+                       ⦃h, L⦄ ⊢ ⓐV1.ⓓ{a}W1.T1 ➡*[g] ⓓ{a}W2.ⓐV2.T2.
 #h #g #a #L #V1 #V #V2 #W1 #W2 #T1 #T2 #HV1 #HV2 #HT12 #H elim H -W2 /2 width=3/
 #W #W2 #_ #HW2 #IHW1
-@(cpxs_trans … IHW1) /2 width=1/
+@(cpxs_trans … IHW1) -IHW1 /2 width=1/
 qed.
 
 theorem cpxs_theta: ∀h,g,a,L,V1,V,V2,W1,W2,T1,T2.
                     ⇧[0, 1] V ≡ V2 → ⦃h, L⦄ ⊢ W1 ➡*[g] W2 →
-                   ⦃h, L.ⓓW1⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ V1 ➡*[g] V →
-                   ⦃h, L⦄ ⊢ ⓐV1.ⓓ{a}W1.T1 ➡*[g] ⓓ{a}W2.ⓐV2.T2.
+                    ⦃h, L.ⓓW1⦄ ⊢ T1 ➡*[g] T2 → ⦃h, L⦄ ⊢ V1 ➡*[g] V →
+                    ⦃h, L⦄ ⊢ ⓐV1.ⓓ{a}W1.T1 ➡*[g] ⓓ{a}W2.ⓐV2.T2.
 #h #g #a #L #V1 #V #V2 #W1 #W2 #T1 #T2 #HV2 #HW12 #HT12 #H @(TC_ind_dx … V1 H) -V1 /2 width=3/
 #V1 #V0 #HV10 #_ #IHV0
-@(cpxs_trans … IHV0) /2 width=1/
+@(cpxs_trans … IHV0) -IHV0 /2 width=1/
 qed.
 
+(* Advanced inversion lemmas ************************************************)
+
+lemma cpxs_inv_appl1: ∀h,g,L,V1,T1,U2. ⦃h, L⦄ ⊢ ⓐV1.T1 ➡*[g] U2 →
+                      ∨∨ ∃∃V2,T2.       ⦃h, L⦄ ⊢ V1 ➡*[g] V2 & ⦃h, L⦄ ⊢ T1 ➡*[g] T2 &
+                                        U2 = ⓐV2. T2
+                       | ∃∃a,W,T.       ⦃h, L⦄ ⊢ T1 ➡*[g] ⓛ{a}W.T & ⦃h, L⦄ ⊢ ⓓ{a}ⓝW.V1.T ➡*[g] U2
+                       | ∃∃a,V0,V2,V,T. ⦃h, L⦄ ⊢ V1 ➡*[g] V0 & ⇧[0,1] V0 ≡ V2 &
+                                        ⦃h, L⦄ ⊢ T1 ➡*[g] ⓓ{a}V.T & ⦃h, L⦄ ⊢ ⓓ{a}V.ⓐV2.T ➡*[g] U2.
+#h #g #L #V1 #T1 #U2 #H @(cpxs_ind … H) -U2 [ /3 width=5/ ]
+#U #U2 #_ #HU2 * *
+[ #V0 #T0 #HV10 #HT10 #H destruct
+  elim (cpx_inv_appl1 … HU2) -HU2 *
+  [ #V2 #T2 #HV02 #HT02 #H destruct /4 width=5/
+  | #a #V2 #W #W2 #T #T2 #HV02 #HW2 #HT2 #H1 #H2 destruct
+    lapply (cpxs_strap1 … HV10 … HV02) -V0 #HV12
+    lapply (lsubx_cpx_trans … HT2 (L.ⓓⓝW.V1) ?) -HT2 /2 width=1/ #HT2
+    @or3_intro1 @(ex2_3_intro … HT10) -HT10 /3 width=1/ (**) (* explicit constructor. /5 width=8/ is too slow because TC_transitive gets in the way *)
+  | #a #V #V2 #W0 #W2 #T #T2 #HV0 #HV2 #HW02 #HT2 #H1 #H2 destruct
+    @or3_intro2 @(ex4_5_intro … HV2 HT10) /2 width=3/ /3 width=1/ (**) (* explicit constructor. /5 width=8/ is too slow because TC_transitive gets in the way *)
+  ]
+| /4 width=9/
+| /4 width=11/
+]
+qed-.
+
 (* Properties on sn extended parallel reduction for local environments ******)
 
 lemma lpx_cpx_trans: ∀h,g. s_r_trans … (cpx h g) (lpx h g).
@@ -77,11 +110,11 @@ lemma lpx_cpx_trans: ∀h,g. s_r_trans … (cpx h g) (lpx h g).
   lapply (cpxs_strap2 … HV10 … HV02) -V0 /2 width=7/
 | #a #I #L2 #V1 #V2 #T1 #T2 #_ #_ #IHV12 #IHT12 #L1 #HL12
   lapply (IHT12 (L1.ⓑ{I}V1) ?) -IHT12 /2 width=1/ /3 width=1/
-|5,7: /3 width=1/
+|5,7,8: /3 width=1/
 | #L2 #V2 #T1 #T #T2 #_ #HT2 #IHT1 #L1 #HL12
   lapply (IHT1 (L1.ⓓV2) ?) -IHT1 /2 width=1/ /2 width=3/
-| #a #L2 #V1 #V2 #W #T1 #T2 #_ #_ #IHV12 #IHT12 #L1 #HL12
-  lapply (IHT12 (L1.ⓛW) ?) -IHT12 /2 width=1/ /3 width=1/
+| #a #L2 #V1 #V2 #W1 #W2 #T1 #T2 #_ #_ #_ #IHV12 #IHW12 #IHT12 #L1 #HL12
+  lapply (IHT12 (L1.ⓛW1) ?) -IHT12 /2 width=1/ /3 width=1/
 | #a #L2 #V1 #V #V2 #W1 #W2 #T1 #T2 #_ #HV2 #_ #_ #IHV1 #IHW12 #IHT12 #L1 #HL12
   lapply (IHT12 (L1.ⓓW1) ?) -IHT12 /2 width=1/ /3 width=3/
 ]