src/HOL/Integ/Presburger.thy

dropped debug cmd

+
−(*  Title:      HOL/Integ/Presburger.thy
+
−    ID:         $Id$
+
−    Author:     Amine Chaieb, Tobias Nipkow and Stefan Berghofer, TU Muenchen
+
−
+
−File containing necessary theorems for the proof
+
−generation for Cooper Algorithm  
+
−*)
+
−
+
−header {* Presburger Arithmetic: Cooper's Algorithm *}
+
−
+
−theory Presburger
+
−imports NatSimprocs
+
−uses
+
−  ("cooper_dec.ML") ("cooper_proof.ML") ("qelim.ML") 
+
−  ("reflected_presburger.ML") ("reflected_cooper.ML") ("presburger.ML")
+
−begin
+
−
+
−text {* Theorem for unitifying the coeffitients of @{text x} in an existential formula*}
+
−
+
−theorem unity_coeff_ex: "(\<exists>x::int. P (l * x)) = (\<exists>x. l dvd (1*x+0) \<and> P x)"
+
−  apply (rule iffI)
+
−  apply (erule exE)
+
−  apply (rule_tac x = "l * x" in exI)
+
−  apply simp
+
−  apply (erule exE)
+
−  apply (erule conjE)
+
−  apply (erule dvdE)
+
−  apply (rule_tac x = k in exI)
+
−  apply simp
+
−  done
+
−
+
−lemma uminus_dvd_conv: "(d dvd (t::int)) = (-d dvd t)"
+
−apply(unfold dvd_def)
+
−apply(rule iffI)
+
−apply(clarsimp)
+
−apply(rename_tac k)
+
−apply(rule_tac x = "-k" in exI)
+
−apply simp
+
−apply(clarsimp)
+
−apply(rename_tac k)
+
−apply(rule_tac x = "-k" in exI)
+
−apply simp
+
−done
+
−
+
−lemma uminus_dvd_conv': "(d dvd (t::int)) = (d dvd -t)"
+
−apply(unfold dvd_def)
+
−apply(rule iffI)
+
−apply(clarsimp)
+
−apply(rule_tac x = "-k" in exI)
+
−apply simp
+
−apply(clarsimp)
+
−apply(rule_tac x = "-k" in exI)
+
−apply simp
+
−done
+
−
+
−
+
−
+
−text {*Theorems for the combination of proofs of the equality of @{text P} and @{text P_m} for integers @{text x} less than some integer @{text z}.*}
+
−
+
−theorem eq_minf_conjI: "\<exists>z1::int. \<forall>x. x < z1 \<longrightarrow> (A1 x = A2 x) \<Longrightarrow>
+
−  \<exists>z2::int. \<forall>x. x < z2 \<longrightarrow> (B1 x = B2 x) \<Longrightarrow>
+
−  \<exists>z::int. \<forall>x. x < z \<longrightarrow> ((A1 x \<and> B1 x) = (A2 x \<and> B2 x))"
+
−  apply (erule exE)+
+
−  apply (rule_tac x = "min z1 z2" in exI)
+
−  apply simp
+
−  done
+
−
+
−
+
−theorem eq_minf_disjI: "\<exists>z1::int. \<forall>x. x < z1 \<longrightarrow> (A1 x = A2 x) \<Longrightarrow>
+
−  \<exists>z2::int. \<forall>x. x < z2 \<longrightarrow> (B1 x = B2 x) \<Longrightarrow>
+
−  \<exists>z::int. \<forall>x. x < z \<longrightarrow> ((A1 x \<or> B1 x) = (A2 x \<or> B2 x))"
+
−
+
−  apply (erule exE)+
+
−  apply (rule_tac x = "min z1 z2" in exI)
+
−  apply simp
+
−  done
+
−
+
−
+
−text {*Theorems for the combination of proofs of the equality of @{text P} and @{text P_m} for integers @{text x} greather than some integer @{text z}.*}
+
−
+
−theorem eq_pinf_conjI: "\<exists>z1::int. \<forall>x. z1 < x \<longrightarrow> (A1 x = A2 x) \<Longrightarrow>
+
−  \<exists>z2::int. \<forall>x. z2 < x \<longrightarrow> (B1 x = B2 x) \<Longrightarrow>
+
−  \<exists>z::int. \<forall>x. z < x \<longrightarrow> ((A1 x \<and> B1 x) = (A2 x \<and> B2 x))"
+
−  apply (erule exE)+
+
−  apply (rule_tac x = "max z1 z2" in exI)
+
−  apply simp
+
−  done
+
−
+
−
+
−theorem eq_pinf_disjI: "\<exists>z1::int. \<forall>x. z1 < x \<longrightarrow> (A1 x = A2 x) \<Longrightarrow>
+
−  \<exists>z2::int. \<forall>x. z2 < x \<longrightarrow> (B1 x = B2 x) \<Longrightarrow>
+
−  \<exists>z::int. \<forall>x. z < x  \<longrightarrow> ((A1 x \<or> B1 x) = (A2 x \<or> B2 x))"
+
−  apply (erule exE)+
+
−  apply (rule_tac x = "max z1 z2" in exI)
+
−  apply simp
+
−  done
+
−
+
−text {*
+
−  \medskip Theorems for the combination of proofs of the modulo @{text
+
−  D} property for @{text "P plusinfinity"}
+
−
+
−  FIXME: This is THE SAME theorem as for the @{text minusinf} version,
+
−  but with @{text "+k.."} instead of @{text "-k.."} In the future
+
−  replace these both with only one. *}
+
−
+
−theorem modd_pinf_conjI: "\<forall>(x::int) k. A x = A (x+k*d) \<Longrightarrow>
+
−  \<forall>(x::int) k. B x = B (x+k*d) \<Longrightarrow>
+
−  \<forall>(x::int) (k::int). (A x \<and> B x) = (A (x+k*d) \<and> B (x+k*d))"
+
−  by simp
+
−
+
−theorem modd_pinf_disjI: "\<forall>(x::int) k. A x = A (x+k*d) \<Longrightarrow>
+
−  \<forall>(x::int) k. B x = B (x+k*d) \<Longrightarrow>
+
−  \<forall>(x::int) (k::int). (A x \<or> B x) = (A (x+k*d) \<or> B (x+k*d))"
+
−  by simp
+
−
+
−text {*
+
−  This is one of the cases where the simplifed formula is prooved to
+
−  habe some property (in relation to @{text P_m}) but we need to prove
+
−  the property for the original formula (@{text P_m})
+
−
+
−  FIXME: This is exaclty the same thm as for @{text minusinf}. *}
+
−
+
−lemma pinf_simp_eq: "ALL x. P(x) = Q(x) ==> (EX (x::int). P(x)) --> (EX (x::int). F(x))  ==> (EX (x::int). Q(x)) --> (EX (x::int). F(x)) "
+
−  by blast
+
−
+
−
+
−text {*
+
−  \medskip Theorems for the combination of proofs of the modulo @{text D}
+
−  property for @{text "P minusinfinity"} *}
+
−
+
−theorem modd_minf_conjI: "\<forall>(x::int) k. A x = A (x-k*d) \<Longrightarrow>
+
−  \<forall>(x::int) k. B x = B (x-k*d) \<Longrightarrow>
+
−  \<forall>(x::int) (k::int). (A x \<and> B x) = (A (x-k*d) \<and> B (x-k*d))"
+
−  by simp
+
−
+
−theorem modd_minf_disjI: "\<forall>(x::int) k. A x = A (x-k*d) \<Longrightarrow>
+
−  \<forall>(x::int) k. B x = B (x-k*d) \<Longrightarrow>
+
−  \<forall>(x::int) (k::int). (A x \<or> B x) = (A (x-k*d) \<or> B (x-k*d))"
+
−  by simp
+
−
+
−text {*
+
−  This is one of the cases where the simplifed formula is prooved to
+
−  have some property (in relation to @{text P_m}) but we need to
+
−  prove the property for the original formula (@{text P_m}). *}
+
−
+
−lemma minf_simp_eq: "ALL x. P(x) = Q(x) ==> (EX (x::int). P(x)) --> (EX (x::int). F(x))  ==> (EX (x::int). Q(x)) --> (EX (x::int). F(x)) "
+
−  by blast
+
−
+
−text {*
+
−  Theorem needed for proving at runtime divide properties using the
+
−  arithmetic tactic (which knows only about modulo = 0). *}
+
−
+
−lemma zdvd_iff_zmod_eq_0: "(m dvd n) = (n mod m = (0::int))"
+
−  by(simp add:dvd_def zmod_eq_0_iff)
+
−
+
−text {*
+
−  \medskip Theorems used for the combination of proof for the
+
−  backwards direction of Cooper's Theorem. They rely exclusively on
+
−  Predicate calculus.*}
+
−
+
−lemma not_ast_p_disjI: "(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> P1(x) --> P1(x + d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> P2(x) --> P2(x + d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) -->(P1(x) \<or> P2(x)) --> (P1(x + d) \<or> P2(x + d))) "
+
−  by blast
+
−
+
−
+
−lemma not_ast_p_conjI: "(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a- j)) --> P1(x) --> P1(x + d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> P2(x) --> P2(x + d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) -->(P1(x) \<and> P2(x)) --> (P1(x + d)
+
−\<and> P2(x + d))) "
+
−  by blast
+
−
+
−lemma not_ast_p_Q_elim: "
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) -->P(x) --> P(x + d))
+
−==> ( P = Q )
+
−==> (ALL x. ~(EX (j::int) : {1..d}. EX (a::int) : A. P(a - j)) -->P(x) --> P(x + d))"
+
−  by blast
+
−
+
−text {*
+
−  \medskip Theorems used for the combination of proof for the
+
−  backwards direction of Cooper's Theorem. They rely exclusively on
+
−  Predicate calculus.*}
+
−
+
−lemma not_bst_p_disjI: "(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> P1(x) --> P1(x - d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> P2(x) --> P2(x - d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) -->(P1(x) \<or> P2(x)) --> (P1(x - d)
+
−\<or> P2(x-d))) "
+
−  by blast
+
−
+
−lemma not_bst_p_conjI: "(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> P1(x) --> P1(x - d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> P2(x) --> P2(x - d))
+
−==>
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) -->(P1(x) \<and> P2(x)) --> (P1(x - d)
+
−\<and> P2(x-d))) "
+
−  by blast
+
−
+
−lemma not_bst_p_Q_elim: "
+
−(ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) -->P(x) --> P(x - d)) 
+
−==> ( P = Q )
+
−==> (ALL x. ~(EX (j::int) : {1..d}. EX (b::int) : B. P(b+j)) -->P(x) --> P(x - d))"
+
−  by blast
+
−
+
−text {* \medskip This is the first direction of Cooper's Theorem. *}
+
−lemma cooper_thm: "(R --> (EX x::int. P x))  ==> (Q -->(EX x::int.  P x )) ==> ((R|Q) --> (EX x::int. P x )) "
+
−  by blast
+
−
+
−text {*
+
−  \medskip The full Cooper's Theorem in its equivalence Form. Given
+
−  the premises it is trivial too, it relies exclusively on prediacte calculus.*}
+
−lemma cooper_eq_thm: "(R --> (EX x::int. P x))  ==> (Q -->(EX x::int.  P x )) ==> ((~Q)
+
−--> (EX x::int. P x ) --> R) ==> (EX x::int. P x) = R|Q "
+
−  by blast
+
−
+
−text {*
+
−  \medskip Some of the atomic theorems generated each time the atom
+
−  does not depend on @{text x}, they are trivial.*}
+
−
+
−lemma  fm_eq_minf: "EX z::int. ALL x. x < z --> (P = P) "
+
−  by blast
+
−
+
−lemma  fm_modd_minf: "ALL (x::int). ALL (k::int). (P = P)"
+
−  by blast
+
−
+
−lemma not_bst_p_fm: "ALL (x::int). Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> fm --> fm"
+
−  by blast
+
−
+
−lemma  fm_eq_pinf: "EX z::int. ALL x. z < x --> (P = P) "
+
−  by blast
+
−
+
−text {* The next two thms are the same as the @{text minusinf} version. *}
+
−
+
−lemma  fm_modd_pinf: "ALL (x::int). ALL (k::int). (P = P)"
+
−  by blast
+
−
+
−lemma not_ast_p_fm: "ALL (x::int). Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> fm --> fm"
+
−  by blast
+
−
+
−text {* Theorems to be deleted from simpset when proving simplified formulaes. *}
+
−
+
−lemma P_eqtrue: "(P=True) = P"
+
−  by iprover
+
−
+
−lemma P_eqfalse: "(P=False) = (~P)"
+
−  by iprover
+
−
+
−text {*
+
−  \medskip Theorems for the generation of the bachwards direction of
+
−  Cooper's Theorem.
+
−
+
−  These are the 6 interesting atomic cases which have to be proved relying on the
+
−  properties of B-set and the arithmetic and contradiction proofs. *}
+
−
+
−lemma not_bst_p_lt: "0 < (d::int) ==>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> ( 0 < -x + a) --> (0 < -(x - d) + a )"
+
−  by arith
+
−
+
−lemma not_bst_p_gt: "\<lbrakk> (g::int) \<in> B; g = -a \<rbrakk> \<Longrightarrow>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> (0 < (x) + a) --> ( 0 < (x - d) + a)"
+
−apply clarsimp
+
−apply(rule ccontr)
+
−apply(drule_tac x = "x+a" in bspec)
+
−apply(simp add:atLeastAtMost_iff)
+
−apply(drule_tac x = "-a" in bspec)
+
−apply assumption
+
−apply(simp)
+
−done
+
−
+
−lemma not_bst_p_eq: "\<lbrakk> 0 < d; (g::int) \<in> B; g = -a - 1 \<rbrakk> \<Longrightarrow>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> (0 = x + a) --> (0 = (x - d) + a )"
+
−apply clarsimp
+
−apply(subgoal_tac "x = -a")
+
− prefer 2 apply arith
+
−apply(drule_tac x = "1" in bspec)
+
−apply(simp add:atLeastAtMost_iff)
+
−apply(drule_tac x = "-a- 1" in bspec)
+
−apply assumption
+
−apply(simp)
+
−done
+
−
+
−
+
−lemma not_bst_p_ne: "\<lbrakk> 0 < d; (g::int) \<in> B; g = -a \<rbrakk> \<Longrightarrow>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> ~(0 = x + a) --> ~(0 = (x - d) + a)"
+
−apply clarsimp
+
−apply(subgoal_tac "x = -a+d")
+
− prefer 2 apply arith
+
−apply(drule_tac x = "d" in bspec)
+
−apply(simp add:atLeastAtMost_iff)
+
−apply(drule_tac x = "-a" in bspec)
+
−apply assumption
+
−apply(simp)
+
−done
+
−
+
−
+
−lemma not_bst_p_dvd: "(d1::int) dvd d ==>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> d1 dvd (x + a) --> d1 dvd ((x - d) + a )"
+
−apply(clarsimp simp add:dvd_def)
+
−apply(rename_tac m)
+
−apply(rule_tac x = "m - k" in exI)
+
−apply(simp add:int_distrib)
+
−done
+
−
+
−lemma not_bst_p_ndvd: "(d1::int) dvd d ==>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (b::int) : B. Q(b+j)) --> ~(d1 dvd (x + a)) --> ~(d1 dvd ((x - d) + a ))"
+
−apply(clarsimp simp add:dvd_def)
+
−apply(rename_tac m)
+
−apply(erule_tac x = "m + k" in allE)
+
−apply(simp add:int_distrib)
+
−done
+
−
+
−text {*
+
−  \medskip Theorems for the generation of the bachwards direction of
+
−  Cooper's Theorem.
+
−
+
−  These are the 6 interesting atomic cases which have to be proved
+
−  relying on the properties of A-set ant the arithmetic and
+
−  contradiction proofs. *}
+
−
+
−lemma not_ast_p_gt: "0 < (d::int) ==>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> ( 0 < x + t) --> (0 < (x + d) + t )"
+
−  by arith
+
−
+
−lemma not_ast_p_lt: "\<lbrakk>0 < d ;(t::int) \<in> A \<rbrakk> \<Longrightarrow>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> (0 < -x + t) --> ( 0 < -(x + d) + t)"
+
−  apply clarsimp
+
−  apply (rule ccontr)
+
−  apply (drule_tac x = "t-x" in bspec)
+
−  apply simp
+
−  apply (drule_tac x = "t" in bspec)
+
−  apply assumption
+
−  apply simp
+
−  done
+
−
+
−lemma not_ast_p_eq: "\<lbrakk> 0 < d; (g::int) \<in> A; g = -t + 1 \<rbrakk> \<Longrightarrow>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> (0 = x + t) --> (0 = (x + d) + t )"
+
−  apply clarsimp
+
−  apply (drule_tac x="1" in bspec)
+
−  apply simp
+
−  apply (drule_tac x="- t + 1" in bspec)
+
−  apply assumption
+
−  apply(subgoal_tac "x = -t")
+
−  prefer 2 apply arith
+
−  apply simp
+
−  done
+
−
+
−lemma not_ast_p_ne: "\<lbrakk> 0 < d; (g::int) \<in> A; g = -t \<rbrakk> \<Longrightarrow>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> ~(0 = x + t) --> ~(0 = (x + d) + t)"
+
−  apply clarsimp
+
−  apply (subgoal_tac "x = -t-d")
+
−  prefer 2 apply arith
+
−  apply (drule_tac x = "d" in bspec)
+
−  apply simp
+
−  apply (drule_tac x = "-t" in bspec)
+
−  apply assumption
+
−  apply simp
+
−  done
+
−
+
−lemma not_ast_p_dvd: "(d1::int) dvd d ==>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> d1 dvd (x + t) --> d1 dvd ((x + d) + t )"
+
−  apply(clarsimp simp add:dvd_def)
+
−  apply(rename_tac m)
+
−  apply(rule_tac x = "m + k" in exI)
+
−  apply(simp add:int_distrib)
+
−  done
+
−
+
−lemma not_ast_p_ndvd: "(d1::int) dvd d ==>
+
− ALL x. Q(x::int) \<and> ~(EX (j::int) : {1..d}. EX (a::int) : A. Q(a - j)) --> ~(d1 dvd (x + t)) --> ~(d1 dvd ((x + d) + t ))"
+
−  apply(clarsimp simp add:dvd_def)
+
−  apply(rename_tac m)
+
−  apply(erule_tac x = "m - k" in allE)
+
−  apply(simp add:int_distrib)
+
−  done
+
−
+
−text {*
+
−  \medskip These are the atomic cases for the proof generation for the
+
−  modulo @{text D} property for @{text "P plusinfinity"}
+
−
+
−  They are fully based on arithmetics. *}
+
−
+
−lemma  dvd_modd_pinf: "((d::int) dvd d1) ==>
+
− (ALL (x::int). ALL (k::int). (((d::int) dvd (x + t)) = (d dvd (x+k*d1 + t))))"
+
−  apply(clarsimp simp add:dvd_def)
+
−  apply(rule iffI)
+
−  apply(clarsimp)
+
−  apply(rename_tac n m)
+
−  apply(rule_tac x = "m + n*k" in exI)
+
−  apply(simp add:int_distrib)
+
−  apply(clarsimp)
+
−  apply(rename_tac n m)
+
−  apply(rule_tac x = "m - n*k" in exI)
+
−  apply(simp add:int_distrib mult_ac)
+
−  done
+
−
+
−lemma  not_dvd_modd_pinf: "((d::int) dvd d1) ==>
+
− (ALL (x::int). ALL k. (~((d::int) dvd (x + t))) = (~(d dvd (x+k*d1 + t))))"
+
−  apply(clarsimp simp add:dvd_def)
+
−  apply(rule iffI)
+
−  apply(clarsimp)
+
−  apply(rename_tac n m)
+
−  apply(erule_tac x = "m - n*k" in allE)
+
−  apply(simp add:int_distrib mult_ac)
+
−  apply(clarsimp)
+
−  apply(rename_tac n m)
+
−  apply(erule_tac x = "m + n*k" in allE)
+
−  apply(simp add:int_distrib mult_ac)
+
−  done
+
−
+
−text {*
+
−  \medskip These are the atomic cases for the proof generation for the
+
−  equivalence of @{text P} and @{text "P plusinfinity"} for integers
+
−  @{text x} greater than some integer @{text z}.
+
−
+
−  They are fully based on arithmetics. *}
+
−
+
−lemma  eq_eq_pinf: "EX z::int. ALL x. z < x --> (( 0 = x +t ) = False )"
+
−  apply(rule_tac x = "-t" in exI)
+
−  apply simp
+
−  done
+
−
+
−lemma  neq_eq_pinf: "EX z::int. ALL x.  z < x --> ((~( 0 = x +t )) = True )"
+
−  apply(rule_tac x = "-t" in exI)
+
−  apply simp
+
−  done
+
−
+
−lemma  le_eq_pinf: "EX z::int. ALL x.  z < x --> ( 0 < x +t  = True )"
+
−  apply(rule_tac x = "-t" in exI)
+
−  apply simp
+
−  done
+
−
+
−lemma  len_eq_pinf: "EX z::int. ALL x. z < x  --> (0 < -x +t  = False )"
+
−  apply(rule_tac x = "t" in exI)
+
−  apply simp
+
−  done
+
−
+
−lemma  dvd_eq_pinf: "EX z::int. ALL x.  z < x --> ((d dvd (x + t)) = (d dvd (x + t))) "
+
−  by simp
+
−
+
−lemma  not_dvd_eq_pinf: "EX z::int. ALL x. z < x  --> ((~(d dvd (x + t))) = (~(d dvd (x + t)))) "
+
−  by simp
+
−
+
−text {*
+
−  \medskip These are the atomic cases for the proof generation for the
+
−  modulo @{text D} property for @{text "P minusinfinity"}.
+
−
+
−  They are fully based on arithmetics. *}
+
−
+
−lemma  dvd_modd_minf: "((d::int) dvd d1) ==>
+
− (ALL (x::int). ALL (k::int). (((d::int) dvd (x + t)) = (d dvd (x-k*d1 + t))))"
+
−apply(clarsimp simp add:dvd_def)
+
−apply(rule iffI)
+
−apply(clarsimp)
+
−apply(rename_tac n m)
+
−apply(rule_tac x = "m - n*k" in exI)
+
−apply(simp add:int_distrib)
+
−apply(clarsimp)
+
−apply(rename_tac n m)
+
−apply(rule_tac x = "m + n*k" in exI)
+
−apply(simp add:int_distrib mult_ac)
+
−done
+
−
+
−
+
−lemma  not_dvd_modd_minf: "((d::int) dvd d1) ==>
+
− (ALL (x::int). ALL k. (~((d::int) dvd (x + t))) = (~(d dvd (x-k*d1 + t))))"
+
−apply(clarsimp simp add:dvd_def)
+
−apply(rule iffI)
+
−apply(clarsimp)
+
−apply(rename_tac n m)
+
−apply(erule_tac x = "m + n*k" in allE)
+
−apply(simp add:int_distrib mult_ac)
+
−apply(clarsimp)
+
−apply(rename_tac n m)
+
−apply(erule_tac x = "m - n*k" in allE)
+
−apply(simp add:int_distrib mult_ac)
+
−done
+
−
+
−text {*
+
−  \medskip These are the atomic cases for the proof generation for the
+
−  equivalence of @{text P} and @{text "P minusinfinity"} for integers
+
−  @{text x} less than some integer @{text z}.
+
−
+
−  They are fully based on arithmetics. *}
+
−
+
−lemma  eq_eq_minf: "EX z::int. ALL x. x < z --> (( 0 = x +t ) = False )"
+
−apply(rule_tac x = "-t" in exI)
+
−apply simp
+
−done
+
−
+
−lemma  neq_eq_minf: "EX z::int. ALL x. x < z --> ((~( 0 = x +t )) = True )"
+
−apply(rule_tac x = "-t" in exI)
+
−apply simp
+
−done
+
−
+
−lemma  le_eq_minf: "EX z::int. ALL x. x < z --> ( 0 < x +t  = False )"
+
−apply(rule_tac x = "-t" in exI)
+
−apply simp
+
−done
+
−
+
−
+
−lemma  len_eq_minf: "EX z::int. ALL x. x < z --> (0 < -x +t  = True )"
+
−apply(rule_tac x = "t" in exI)
+
−apply simp
+
−done
+
−
+
−lemma  dvd_eq_minf: "EX z::int. ALL x. x < z --> ((d dvd (x + t)) = (d dvd (x + t))) "
+
−  by simp
+
−
+
−lemma  not_dvd_eq_minf: "EX z::int. ALL x. x < z --> ((~(d dvd (x + t))) = (~(d dvd (x + t)))) "
+
−  by simp
+
−
+
−text {*
+
−  \medskip This Theorem combines whithnesses about @{text "P
+
−  minusinfinity"} to show one component of the equivalence proof for
+
−  Cooper's Theorem.
+
−
+
−  FIXME: remove once they are part of the distribution. *}
+
−
+
−theorem int_ge_induct[consumes 1,case_names base step]:
+
−  assumes ge: "k \<le> (i::int)" and
+
−        base: "P(k)" and
+
−        step: "\<And>i. \<lbrakk>k \<le> i; P i\<rbrakk> \<Longrightarrow> P(i+1)"
+
−  shows "P i"
+
−proof -
+
−  { fix n have "\<And>i::int. n = nat(i-k) \<Longrightarrow> k <= i \<Longrightarrow> P i"
+
−    proof (induct n)
+
−      case 0
+
−      hence "i = k" by arith
+
−      thus "P i" using base by simp
+
−    next
+
−      case (Suc n)
+
−      hence "n = nat((i - 1) - k)" by arith
+
−      moreover
+
−      have ki1: "k \<le> i - 1" using Suc.prems by arith
+
−      ultimately
+
−      have "P(i - 1)" by(rule Suc.hyps)
+
−      from step[OF ki1 this] show ?case by simp
+
−    qed
+
−  }
+
−  from this ge show ?thesis by fast
+
−qed
+
−
+
−theorem int_gr_induct[consumes 1,case_names base step]:
+
−  assumes gr: "k < (i::int)" and
+
−        base: "P(k+1)" and
+
−        step: "\<And>i. \<lbrakk>k < i; P i\<rbrakk> \<Longrightarrow> P(i+1)"
+
−  shows "P i"
+
−apply(rule int_ge_induct[of "k + 1"])
+
−  using gr apply arith
+
− apply(rule base)
+
−apply(rule step)
+
− apply simp+
+
−done
+
−
+
−lemma decr_lemma: "0 < (d::int) \<Longrightarrow> x - (abs(x-z)+1) * d < z"
+
−apply(induct rule: int_gr_induct)
+
− apply simp
+
−apply (simp add:int_distrib)
+
−done
+
−
+
−lemma incr_lemma: "0 < (d::int) \<Longrightarrow> z < x + (abs(x-z)+1) * d"
+
−apply(induct rule: int_gr_induct)
+
− apply simp
+
−apply (simp add:int_distrib)
+
−done
+
−
+
−lemma  minusinfinity:
+
−  assumes "0 < d" and
+
−    P1eqP1: "ALL x k. P1 x = P1(x - k*d)" and
+
−    ePeqP1: "EX z::int. ALL x. x < z \<longrightarrow> (P x = P1 x)"
+
−  shows "(EX x. P1 x) \<longrightarrow> (EX x. P x)"
+
−proof
+
−  assume eP1: "EX x. P1 x"
+
−  then obtain x where P1: "P1 x" ..
+
−  from ePeqP1 obtain z where P1eqP: "ALL x. x < z \<longrightarrow> (P x = P1 x)" ..
+
−  let ?w = "x - (abs(x-z)+1) * d"
+
−  show "EX x. P x"
+
−  proof
+
−    have w: "?w < z" by(rule decr_lemma)
+
−    have "P1 x = P1 ?w" using P1eqP1 by blast
+
−    also have "\<dots> = P(?w)" using w P1eqP by blast
+
−    finally show "P ?w" using P1 by blast
+
−  qed
+
−qed
+
−
+
−text {*
+
−  \medskip This Theorem combines whithnesses about @{text "P
+
−  minusinfinity"} to show one component of the equivalence proof for
+
−  Cooper's Theorem. *}
+
−
+
−lemma plusinfinity:
+
−  assumes "0 < d" and
+
−    P1eqP1: "ALL (x::int) (k::int). P1 x = P1 (x + k * d)" and
+
−    ePeqP1: "EX z::int. ALL x. z < x  --> (P x = P1 x)"
+
−  shows "(EX x::int. P1 x) --> (EX x::int. P x)"
+
−proof
+
−  assume eP1: "EX x. P1 x"
+
−  then obtain x where P1: "P1 x" ..
+
−  from ePeqP1 obtain z where P1eqP: "ALL x. z < x \<longrightarrow> (P x = P1 x)" ..
+
−  let ?w = "x + (abs(x-z)+1) * d"
+
−  show "EX x. P x"
+
−  proof
+
−    have w: "z < ?w" by(rule incr_lemma)
+
−    have "P1 x = P1 ?w" using P1eqP1 by blast
+
−    also have "\<dots> = P(?w)" using w P1eqP by blast
+
−    finally show "P ?w" using P1 by blast
+
−  qed
+
−qed
+
− 
+
−text {*
+
−  \medskip Theorem for periodic function on discrete sets. *}
+
−
+
−lemma minf_vee:
+
−  assumes dpos: "(0::int) < d" and modd: "ALL x k. P x = P(x - k*d)"
+
−  shows "(EX x. P x) = (EX j : {1..d}. P j)"
+
−  (is "?LHS = ?RHS")
+
−proof
+
−  assume ?LHS
+
−  then obtain x where P: "P x" ..
+
−  have "x mod d = x - (x div d)*d"
+
−    by(simp add:zmod_zdiv_equality mult_ac eq_diff_eq)
+
−  hence Pmod: "P x = P(x mod d)" using modd by simp
+
−  show ?RHS
+
−  proof (cases)
+
−    assume "x mod d = 0"
+
−    hence "P 0" using P Pmod by simp
+
−    moreover have "P 0 = P(0 - (-1)*d)" using modd by blast
+
−    ultimately have "P d" by simp
+
−    moreover have "d : {1..d}" using dpos by(simp add:atLeastAtMost_iff)
+
−    ultimately show ?RHS ..
+
−  next
+
−    assume not0: "x mod d \<noteq> 0"
+
−    have "P(x mod d)" using dpos P Pmod by(simp add:pos_mod_sign pos_mod_bound)
+
−    moreover have "x mod d : {1..d}"
+
−    proof -
+
−      have "0 \<le> x mod d" by(rule pos_mod_sign)
+
−      moreover have "x mod d < d" by(rule pos_mod_bound)
+
−      ultimately show ?thesis using not0 by(simp add:atLeastAtMost_iff)
+
−    qed
+
−    ultimately show ?RHS ..
+
−  qed
+
−next
+
−  assume ?RHS thus ?LHS by blast
+
−qed
+
−
+
−text {*
+
−  \medskip Theorem for periodic function on discrete sets. *}
+
−
+
−lemma pinf_vee:
+
−  assumes dpos: "0 < (d::int)" and modd: "ALL (x::int) (k::int). P x = P (x+k*d)"
+
−  shows "(EX x::int. P x) = (EX (j::int) : {1..d} . P j)"
+
−  (is "?LHS = ?RHS")
+
−proof
+
−  assume ?LHS
+
−  then obtain x where P: "P x" ..
+
−  have "x mod d = x + (-(x div d))*d"
+
−    by(simp add:zmod_zdiv_equality mult_ac eq_diff_eq)
+
−  hence Pmod: "P x = P(x mod d)" using modd by (simp only:)
+
−  show ?RHS
+
−  proof (cases)
+
−    assume "x mod d = 0"
+
−    hence "P 0" using P Pmod by simp
+
−    moreover have "P 0 = P(0 + 1*d)" using modd by blast
+
−    ultimately have "P d" by simp
+
−    moreover have "d : {1..d}" using dpos by(simp add:atLeastAtMost_iff)
+
−    ultimately show ?RHS ..
+
−  next
+
−    assume not0: "x mod d \<noteq> 0"
+
−    have "P(x mod d)" using dpos P Pmod by(simp add:pos_mod_sign pos_mod_bound)
+
−    moreover have "x mod d : {1..d}"
+
−    proof -
+
−      have "0 \<le> x mod d" by(rule pos_mod_sign)
+
−      moreover have "x mod d < d" by(rule pos_mod_bound)
+
−      ultimately show ?thesis using not0 by(simp add:atLeastAtMost_iff)
+
−    qed
+
−    ultimately show ?RHS ..
+
−  qed
+
−next
+
−  assume ?RHS thus ?LHS by blast
+
−qed
+
−
+
−lemma decr_mult_lemma:
+
−  assumes dpos: "(0::int) < d" and
+
−          minus: "ALL x::int. P x \<longrightarrow> P(x - d)" and
+
−          knneg: "0 <= k"
+
−  shows "ALL x. P x \<longrightarrow> P(x - k*d)"
+
−using knneg
+
−proof (induct rule:int_ge_induct)
+
−  case base thus ?case by simp
+
−next
+
−  case (step i)
+
−  show ?case
+
−  proof
+
−    fix x
+
−    have "P x \<longrightarrow> P (x - i * d)" using step.hyps by blast
+
−    also have "\<dots> \<longrightarrow> P(x - (i + 1) * d)"
+
−      using minus[THEN spec, of "x - i * d"]
+
−      by (simp add:int_distrib OrderedGroup.diff_diff_eq[symmetric])
+
−    ultimately show "P x \<longrightarrow> P(x - (i + 1) * d)" by blast
+
−  qed
+
−qed
+
−
+
−lemma incr_mult_lemma:
+
−  assumes dpos: "(0::int) < d" and
+
−          plus: "ALL x::int. P x \<longrightarrow> P(x + d)" and
+
−          knneg: "0 <= k"
+
−  shows "ALL x. P x \<longrightarrow> P(x + k*d)"
+
−using knneg
+
−proof (induct rule:int_ge_induct)
+
−  case base thus ?case by simp
+
−next
+
−  case (step i)
+
−  show ?case
+
−  proof
+
−    fix x
+
−    have "P x \<longrightarrow> P (x + i * d)" using step.hyps by blast
+
−    also have "\<dots> \<longrightarrow> P(x + (i + 1) * d)"
+
−      using plus[THEN spec, of "x + i * d"]
+
−      by (simp add:int_distrib zadd_ac)
+
−    ultimately show "P x \<longrightarrow> P(x + (i + 1) * d)" by blast
+
−  qed
+
−qed
+
−
+
−lemma cpmi_eq: "0 < D \<Longrightarrow> (EX z::int. ALL x. x < z --> (P x = P1 x))
+
−==> ALL x.~(EX (j::int) : {1..D}. EX (b::int) : B. P(b+j)) --> P (x) --> P (x - D) 
+
−==> (ALL (x::int). ALL (k::int). ((P1 x)= (P1 (x-k*D))))
+
−==> (EX (x::int). P(x)) = ((EX (j::int) : {1..D} . (P1(j))) | (EX (j::int) : {1..D}. EX (b::int) : B. P (b+j)))"
+
−apply(rule iffI)
+
−prefer 2
+
−apply(drule minusinfinity)
+
−apply assumption+
+
−apply(fastsimp)
+
−apply clarsimp
+
−apply(subgoal_tac "!!k. 0<=k \<Longrightarrow> !x. P x \<longrightarrow> P (x - k*D)")
+
−apply(frule_tac x = x and z=z in decr_lemma)
+
−apply(subgoal_tac "P1(x - (\<bar>x - z\<bar> + 1) * D)")
+
−prefer 2
+
−apply(subgoal_tac "0 <= (\<bar>x - z\<bar> + 1)")
+
−prefer 2 apply arith
+
− apply fastsimp
+
−apply(drule (1) minf_vee)
+
−apply blast
+
−apply(blast dest:decr_mult_lemma)
+
−done
+
−
+
−text {* Cooper Theorem, plus infinity version. *}
+
−lemma cppi_eq: "0 < D \<Longrightarrow> (EX z::int. ALL x. z < x --> (P x = P1 x))
+
−==> ALL x.~(EX (j::int) : {1..D}. EX (a::int) : A. P(a - j)) --> P (x) --> P (x + D) 
+
−==> (ALL (x::int). ALL (k::int). ((P1 x)= (P1 (x+k*D))))
+
−==> (EX (x::int). P(x)) = ((EX (j::int) : {1..D} . (P1(j))) | (EX (j::int) : {1..D}. EX (a::int) : A. P (a - j)))"
+
−  apply(rule iffI)
+
−  prefer 2
+
−  apply(drule plusinfinity)
+
−  apply assumption+
+
−  apply(fastsimp)
+
−  apply clarsimp
+
−  apply(subgoal_tac "!!k. 0<=k \<Longrightarrow> !x. P x \<longrightarrow> P (x + k*D)")
+
−  apply(frule_tac x = x and z=z in incr_lemma)
+
−  apply(subgoal_tac "P1(x + (\<bar>x - z\<bar> + 1) * D)")
+
−  prefer 2
+
−  apply(subgoal_tac "0 <= (\<bar>x - z\<bar> + 1)")
+
−  prefer 2 apply arith
+
−  apply fastsimp
+
−  apply(drule (1) pinf_vee)
+
−  apply blast
+
−  apply(blast dest:incr_mult_lemma)
+
−  done
+
−
+
−
+
−text {*
+
−  \bigskip Theorems for the quantifier elminination Functions. *}
+
−
+
−lemma qe_ex_conj: "(EX (x::int). A x) = R
+
−		==> (EX (x::int). P x) = (Q & (EX x::int. A x))
+
−		==> (EX (x::int). P x) = (Q & R)"
+
−by blast
+
−
+
−lemma qe_ex_nconj: "(EX (x::int). P x) = (True & Q)
+
−		==> (EX (x::int). P x) = Q"
+
−by blast
+
−
+
−lemma qe_conjI: "P1 = P2 ==> Q1 = Q2 ==> (P1 & Q1) = (P2 & Q2)"
+
−by blast
+
−
+
−lemma qe_disjI: "P1 = P2 ==> Q1 = Q2 ==> (P1 | Q1) = (P2 | Q2)"
+
−by blast
+
−
+
−lemma qe_impI: "P1 = P2 ==> Q1 = Q2 ==> (P1 --> Q1) = (P2 --> Q2)"
+
−by blast
+
−
+
−lemma qe_eqI: "P1 = P2 ==> Q1 = Q2 ==> (P1 = Q1) = (P2 = Q2)"
+
−by blast
+
−
+
−lemma qe_Not: "P = Q ==> (~P) = (~Q)"
+
−by blast
+
−
+
−lemma qe_ALL: "(EX x. ~P x) = R ==> (ALL x. P x) = (~R)"
+
−by blast
+
−
+
−text {* \bigskip Theorems for proving NNF *}
+
−
+
−lemma nnf_im: "((~P) = P1) ==> (Q=Q1) ==> ((P --> Q) = (P1 | Q1))"
+
−by blast
+
−
+
−lemma nnf_eq: "((P & Q) = (P1 & Q1)) ==> (((~P) & (~Q)) = (P2 & Q2)) ==> ((P = Q) = ((P1 & Q1)|(P2 & Q2)))"
+
−by blast
+
−
+
−lemma nnf_nn: "(P = Q) ==> ((~~P) = Q)"
+
−  by blast
+
−lemma nnf_ncj: "((~P) = P1) ==> ((~Q) = Q1) ==> ((~(P & Q)) = (P1 | Q1))"
+
−by blast
+
−
+
−lemma nnf_ndj: "((~P) = P1) ==> ((~Q) = Q1) ==> ((~(P | Q)) = (P1 & Q1))"
+
−by blast
+
−lemma nnf_nim: "(P = P1) ==> ((~Q) = Q1) ==> ((~(P --> Q)) = (P1 & Q1))"
+
−by blast
+
−lemma nnf_neq: "((P & (~Q)) = (P1 & Q1)) ==> (((~P) & Q) = (P2 & Q2)) ==> ((~(P = Q)) = ((P1 & Q1)|(P2 & Q2)))"
+
−by blast
+
−lemma nnf_sdj: "((A & (~B)) = (A1 & B1)) ==> ((C & (~D)) = (C1 & D1)) ==> (A = (~C)) ==> ((~((A & B) | (C & D))) = ((A1 & B1) | (C1 & D1)))"
+
−by blast
+
−
+
−
+
−lemma qe_exI2: "A = B ==> (EX (x::int). A(x)) = (EX (x::int). B(x))"
+
−  by simp
+
−
+
−lemma qe_exI: "(!!x::int. A x = B x) ==> (EX (x::int). A(x)) = (EX (x::int). B(x))"
+
−  by iprover
+
−
+
−lemma qe_ALLI: "(!!x::int. A x = B x) ==> (ALL (x::int). A(x)) = (ALL (x::int). B(x))"
+
−  by iprover
+
−
+
−lemma cp_expand: "(EX (x::int). P (x)) = (EX (j::int) : {1..d}. EX (b::int) : B. (P1 (j) | P(b+j)))
+
−==>(EX (x::int). P (x)) = (EX (j::int) : {1..d}. EX (b::int) : B. (P1 (j) | P(b+j))) "
+
−by blast
+
−
+
−lemma cppi_expand: "(EX (x::int). P (x)) = (EX (j::int) : {1..d}. EX (a::int) : A. (P1 (j) | P(a - j)))
+
−==>(EX (x::int). P (x)) = (EX (j::int) : {1..d}. EX (a::int) : A. (P1 (j) | P(a - j))) "
+
−by blast
+
−
+
−
+
−lemma simp_from_to: "{i..j::int} = (if j < i then {} else insert i {i+1..j})"
+
−apply(simp add:atLeastAtMost_def atLeast_def atMost_def)
+
−apply(fastsimp)
+
−done
+
−
+
−text {* \bigskip Theorems required for the @{text adjustcoeffitienteq} *}
+
−
+
−lemma ac_dvd_eq: assumes not0: "0 ~= (k::int)"
+
−shows "((m::int) dvd (c*n+t)) = (k*m dvd ((k*c)*n+(k*t)))" (is "?P = ?Q")
+
−proof
+
−  assume ?P
+
−  thus ?Q
+
−    apply(simp add:dvd_def)
+
−    apply clarify
+
−    apply(rename_tac d)
+
−    apply(drule_tac f = "op * k" in arg_cong)
+
−    apply(simp only:int_distrib)
+
−    apply(rule_tac x = "d" in exI)
+
−    apply(simp only:mult_ac)
+
−    done
+
−next
+
−  assume ?Q
+
−  then obtain d where "k * c * n + k * t = (k*m)*d" by(fastsimp simp:dvd_def)
+
−  hence "(c * n + t) * k = (m*d) * k" by(simp add:int_distrib mult_ac)
+
−  hence "((c * n + t) * k) div k = ((m*d) * k) div k" by(rule arg_cong[of _ _ "%t. t div k"])
+
−  hence "c*n+t = m*d" by(simp add: zdiv_zmult_self1[OF not0[symmetric]])
+
−  thus ?P by(simp add:dvd_def)
+
−qed
+
−
+
−lemma ac_lt_eq: assumes gr0: "0 < (k::int)"
+
−shows "((m::int) < (c*n+t)) = (k*m <((k*c)*n+(k*t)))" (is "?P = ?Q")
+
−proof
+
−  assume P: ?P
+
−  show ?Q using zmult_zless_mono2[OF P gr0] by(simp add: int_distrib mult_ac)
+
−next
+
−  assume ?Q
+
−  hence "0 < k*(c*n + t - m)" by(simp add: int_distrib mult_ac)
+
−  with gr0 have "0 < (c*n + t - m)" by(simp add: zero_less_mult_iff)
+
−  thus ?P by(simp)
+
−qed
+
−
+
−lemma ac_eq_eq : assumes not0: "0 ~= (k::int)" shows "((m::int) = (c*n+t)) = (k*m =((k*c)*n+(k*t)) )" (is "?P = ?Q")
+
−proof
+
−  assume ?P
+
−  thus ?Q
+
−    apply(drule_tac f = "op * k" in arg_cong)
+
−    apply(simp only:int_distrib)
+
−    done
+
−next
+
−  assume ?Q
+
−  hence "m * k = (c*n + t) * k" by(simp add:int_distrib mult_ac)
+
−  hence "((m) * k) div k = ((c*n + t) * k) div k" by(rule arg_cong[of _ _ "%t. t div k"])
+
−  thus ?P by(simp add: zdiv_zmult_self1[OF not0[symmetric]])
+
−qed
+
−
+
−lemma ac_pi_eq: assumes gr0: "0 < (k::int)" shows "(~((0::int) < (c*n + t))) = (0 < ((-k)*c)*n + ((-k)*t + k))"
+
−proof -
+
−  have "(~ (0::int) < (c*n + t)) = (0<1-(c*n + t))" by arith
+
−  also have  "(1-(c*n + t)) = (-1*c)*n + (-t+1)" by(simp add: int_distrib mult_ac)
+
−  also have "0<(-1*c)*n + (-t+1) = (0 < (k*(-1*c)*n) + (k*(-t+1)))" by(rule ac_lt_eq[of _ 0,OF gr0,simplified])
+
−  also have "(k*(-1*c)*n) + (k*(-t+1)) = ((-k)*c)*n + ((-k)*t + k)" by(simp add: int_distrib mult_ac)
+
−  finally show ?thesis .
+
−qed
+
−
+
−lemma binminus_uminus_conv: "(a::int) - b = a + (-b)"
+
−by arith
+
−
+
−lemma  linearize_dvd: "(t::int) = t1 ==> (d dvd t) = (d dvd t1)"
+
−by simp
+
−
+
−lemma lf_lt: "(l::int) = ll ==> (r::int) = lr ==> (l < r) =(ll < lr)"
+
−by simp
+
−
+
−lemma lf_eq: "(l::int) = ll ==> (r::int) = lr ==> (l = r) =(ll = lr)"
+
−by simp
+
−
+
−lemma lf_dvd: "(l::int) = ll ==> (r::int) = lr ==> (l dvd r) =(ll dvd lr)"
+
−by simp
+
−
+
−text {* \bigskip Theorems for transforming predicates on nat to predicates on @{text int}*}
+
−
+
−theorem all_nat: "(\<forall>x::nat. P x) = (\<forall>x::int. 0 <= x \<longrightarrow> P (nat x))"
+
−  by (simp split add: split_nat)
+
−
+
−theorem ex_nat: "(\<exists>x::nat. P x) = (\<exists>x::int. 0 <= x \<and> P (nat x))"
+
−  apply (simp split add: split_nat)
+
−  apply (rule iffI)
+
−  apply (erule exE)
+
−  apply (rule_tac x = "int x" in exI)
+
−  apply simp
+
−  apply (erule exE)
+
−  apply (rule_tac x = "nat x" in exI)
+
−  apply (erule conjE)
+
−  apply (erule_tac x = "nat x" in allE)
+
−  apply simp
+
−  done
+
−
+
−theorem zdiff_int_split: "P (int (x - y)) =
+
−  ((y \<le> x \<longrightarrow> P (int x - int y)) \<and> (x < y \<longrightarrow> P 0))"
+
−  apply (case_tac "y \<le> x")
+
−  apply (simp_all add: zdiff_int)
+
−  done
+
−
+
−theorem zdvd_int: "(x dvd y) = (int x dvd int y)"
+
−  apply (simp only: dvd_def ex_nat int_int_eq [symmetric] zmult_int [symmetric]
+
−    nat_0_le cong add: conj_cong)
+
−  apply (rule iffI)
+
−  apply iprover
+
−  apply (erule exE)
+
−  apply (case_tac "x=0")
+
−  apply (rule_tac x=0 in exI)
+
−  apply simp
+
−  apply (case_tac "0 \<le> k")
+
−  apply iprover
+
−  apply (simp add: linorder_not_le)
+
−  apply (drule mult_strict_left_mono_neg [OF iffD2 [OF zero_less_int_conv]])
+
−  apply assumption
+
−  apply (simp add: mult_ac)
+
−  done
+
−
+
−theorem number_of1: "(0::int) <= number_of n \<Longrightarrow> (0::int) <= number_of (n BIT b)"
+
−  by simp
+
−
+
−theorem number_of2: "(0::int) <= Numeral0" by simp
+
−
+
−theorem Suc_plus1: "Suc n = n + 1" by simp
+
−
+
−text {*
+
−  \medskip Specific instances of congruence rules, to prevent
+
−  simplifier from looping. *}
+
−
+
−theorem imp_le_cong: "(0 <= x \<Longrightarrow> P = P') \<Longrightarrow> (0 <= (x::int) \<longrightarrow> P) = (0 <= x \<longrightarrow> P')"
+
−  by simp
+
−
+
−theorem conj_le_cong: "(0 <= x \<Longrightarrow> P = P') \<Longrightarrow> (0 <= (x::int) \<and> P) = (0 <= x \<and> P')"
+
−  by (simp cong: conj_cong)
+
−
+
−    (* Theorems used in presburger.ML for the computation simpset*)
+
−    (* FIXME: They are present in Float.thy, so may be Float.thy should be lightened.*)
+
−
+
−lemma lift_bool: "x \<Longrightarrow> x=True"
+
−  by simp
+
−
+
−lemma nlift_bool: "~x \<Longrightarrow> x=False"
+
−  by simp
+
−
+
−lemma not_false_eq_true: "(~ False) = True" by simp
+
−
+
−lemma not_true_eq_false: "(~ True) = False" by simp
+
−
+
−
+
−lemma int_eq_number_of_eq:
+
−  "(((number_of v)::int) = (number_of w)) = iszero ((number_of (v + (uminus w)))::int)"
+
−  by simp
+
−lemma int_iszero_number_of_Pls: "iszero (Numeral0::int)" 
+
−  by (simp only: iszero_number_of_Pls)
+
−
+
−lemma int_nonzero_number_of_Min: "~(iszero ((-1)::int))"
+
−  by simp
+
−
+
−lemma int_iszero_number_of_0: "iszero ((number_of (w BIT bit.B0))::int) = iszero ((number_of w)::int)"
+
−  by simp
+
−
+
−lemma int_iszero_number_of_1: "\<not> iszero ((number_of (w BIT bit.B1))::int)" 
+
−  by simp
+
−
+
−lemma int_less_number_of_eq_neg: "(((number_of x)::int) < number_of y) = neg ((number_of (x + (uminus y)))::int)"
+
−  by simp
+
−
+
−lemma int_not_neg_number_of_Pls: "\<not> (neg (Numeral0::int))" 
+
−  by simp
+
−
+
−lemma int_neg_number_of_Min: "neg (-1::int)"
+
−  by simp
+
−
+
−lemma int_neg_number_of_BIT: "neg ((number_of (w BIT x))::int) = neg ((number_of w)::int)"
+
−  by simp
+
−
+
−lemma int_le_number_of_eq: "(((number_of x)::int) \<le> number_of y) = (\<not> neg ((number_of (y + (uminus x)))::int))"
+
−  by simp
+
−lemma int_number_of_add_sym: "((number_of v)::int) + number_of w = number_of (v + w)"
+
−  by simp
+
−
+
−lemma int_number_of_diff_sym:
+
−  "((number_of v)::int) - number_of w = number_of (v + (uminus w))"
+
−  by simp
+
−
+
−lemma int_number_of_mult_sym:
+
−  "((number_of v)::int) * number_of w = number_of (v * w)"
+
−  by simp
+
−
+
−lemma int_number_of_minus_sym: "- ((number_of v)::int) = number_of (uminus v)"
+
−  by simp
+
−lemma add_left_zero: "0 + a = (a::'a::comm_monoid_add)"
+
−  by simp
+
−
+
−lemma add_right_zero: "a + 0 = (a::'a::comm_monoid_add)"
+
−  by simp
+
−
+
−lemma mult_left_one: "1 * a = (a::'a::semiring_1)"
+
−  by simp
+
−
+
−lemma mult_right_one: "a * 1 = (a::'a::semiring_1)"
+
−  by simp
+
−
+
−lemma int_pow_0: "(a::int)^(Numeral0) = 1"
+
−  by simp
+
−
+
−lemma int_pow_1: "(a::int)^(Numeral1) = a"
+
−  by simp
+
−
+
−lemma zero_eq_Numeral0_nring: "(0::'a::number_ring) = Numeral0"
+
−  by simp
+
−
+
−lemma one_eq_Numeral1_nring: "(1::'a::number_ring) = Numeral1"
+
−  by simp
+
−
+
−lemma zero_eq_Numeral0_nat: "(0::nat) = Numeral0"
+
−  by simp
+
−
+
−lemma one_eq_Numeral1_nat: "(1::nat) = Numeral1"
+
−  by simp
+
−
+
−lemma zpower_Pls: "(z::int)^Numeral0 = Numeral1"
+
−  by simp
+
−
+
−lemma zpower_Min: "(z::int)^((-1)::nat) = Numeral1"
+
−proof -
+
−  have 1:"((-1)::nat) = 0"
+
−    by simp
+
−  show ?thesis by (simp add: 1)
+
−qed
+
−
+
−use "cooper_dec.ML"
+
−use "reflected_presburger.ML" 
+
−use "reflected_cooper.ML"
+
−oracle
+
−  presburger_oracle ("term") = ReflectedCooper.presburger_oracle
+
−
+
−use "cooper_proof.ML"
+
−use "qelim.ML"
+
−use "presburger.ML"
+
−
+
−setup "Presburger.setup"
+
−
+
−text {* Code generator setup *}
+
−
+
−text {*
+
−  Presburger arithmetic is necessary (at least convenient) to prove some
+
−  of the following code lemmas on integer numerals.
+
−*}
+
−
+
−lemma eq_number_of [code func]:
+
−  "((number_of k)\<Colon>int) = number_of l \<longleftrightarrow> k = l"
+
−  unfolding number_of_is_id ..
+
−
+
−lemma less_eq_number_of [code func]:
+
−  "((number_of k)\<Colon>int) <= number_of l \<longleftrightarrow> k <= l"
+
−  unfolding number_of_is_id ..
+
−
+
−lemma eq_Pls_Pls:
+
−  "Numeral.Pls = Numeral.Pls" ..
+
−
+
−lemma eq_Pls_Min:
+
−  "Numeral.Pls \<noteq> Numeral.Min"
+
−  unfolding Pls_def Min_def by auto
+
−
+
−lemma eq_Pls_Bit0:
+
−  "Numeral.Pls = Numeral.Bit k bit.B0 \<longleftrightarrow> Numeral.Pls = k"
+
−  unfolding Pls_def Bit_def bit.cases by auto
+
−
+
−lemma eq_Pls_Bit1:
+
−  "Numeral.Pls \<noteq> Numeral.Bit k bit.B1"
+
−  unfolding Pls_def Bit_def bit.cases by arith
+
−
+
−lemma eq_Min_Pls:
+
−  "Numeral.Min \<noteq> Numeral.Pls"
+
−  unfolding Pls_def Min_def by auto
+
−
+
−lemma eq_Min_Min:
+
−  "Numeral.Min = Numeral.Min" ..
+
−
+
−lemma eq_Min_Bit0:
+
−  "Numeral.Min \<noteq> Numeral.Bit k bit.B0"
+
−  unfolding Min_def Bit_def bit.cases by arith
+
−
+
−lemma eq_Min_Bit1:
+
−  "Numeral.Min = Numeral.Bit k bit.B1 \<longleftrightarrow> Numeral.Min = k"
+
−  unfolding Min_def Bit_def bit.cases by auto
+
−
+
−lemma eq_Bit0_Pls:
+
−  "Numeral.Bit k bit.B0 = Numeral.Pls \<longleftrightarrow> Numeral.Pls = k"
+
−  unfolding Pls_def Bit_def bit.cases by auto
+
−
+
−lemma eq_Bit1_Pls:
+
−  "Numeral.Bit k bit.B1 \<noteq>  Numeral.Pls"
+
−  unfolding Pls_def Bit_def bit.cases by arith
+
−
+
−lemma eq_Bit0_Min:
+
−  "Numeral.Bit k bit.B0 \<noteq> Numeral.Min"
+
−  unfolding Min_def Bit_def bit.cases by arith
+
−
+
−lemma eq_Bit1_Min:
+
−  "(Numeral.Bit k bit.B1) = Numeral.Min \<longleftrightarrow> Numeral.Min = k"
+
−  unfolding Min_def Bit_def bit.cases by auto
+
−
+
−lemma eq_Bit_Bit:
+
−  "Numeral.Bit k1 v1 = Numeral.Bit k2 v2 \<longleftrightarrow>
+
−    v1 = v2 \<and> k1 = k2"
+
−  unfolding Bit_def
+
−  apply (cases v1)
+
−  apply (cases v2)
+
−  apply auto
+
−  apply arith
+
−  apply (cases v2)
+
−  apply auto
+
−  apply arith
+
−  apply (cases v2)
+
−  apply auto
+
−done
+
−
+
−lemmas eq_numeral_code [code func] =
+
−  eq_Pls_Pls eq_Pls_Min eq_Pls_Bit0 eq_Pls_Bit1
+
−  eq_Min_Pls eq_Min_Min eq_Min_Bit0 eq_Min_Bit1
+
−  eq_Bit0_Pls eq_Bit1_Pls eq_Bit0_Min eq_Bit1_Min eq_Bit_Bit
+
−
+
−lemma less_eq_Pls_Pls:
+
−  "Numeral.Pls \<le> Numeral.Pls" ..
+
−
+
−lemma less_eq_Pls_Min:
+
−  "\<not> (Numeral.Pls \<le> Numeral.Min)"
+
−  unfolding Pls_def Min_def by auto
+
−
+
−lemma less_eq_Pls_Bit:
+
−  "Numeral.Pls \<le> Numeral.Bit k v \<longleftrightarrow> Numeral.Pls \<le> k"
+
−  unfolding Pls_def Bit_def by (cases v) auto
+
−
+
−lemma less_eq_Min_Pls:
+
−  "Numeral.Min \<le> Numeral.Pls"
+
−  unfolding Pls_def Min_def by auto
+
−
+
−lemma less_eq_Min_Min:
+
−  "Numeral.Min \<le> Numeral.Min" ..
+
−
+
−lemma less_eq_Min_Bit0:
+
−  "Numeral.Min \<le> Numeral.Bit k bit.B0 \<longleftrightarrow> Numeral.Min < k"
+
−  unfolding Min_def Bit_def by auto
+
−
+
−lemma less_eq_Min_Bit1:
+
−  "Numeral.Min \<le> Numeral.Bit k bit.B1 \<longleftrightarrow> Numeral.Min \<le> k"
+
−  unfolding Min_def Bit_def by auto
+
−
+
−lemma less_eq_Bit0_Pls:
+
−  "Numeral.Bit k bit.B0 \<le> Numeral.Pls \<longleftrightarrow> k \<le> Numeral.Pls"
+
−  unfolding Pls_def Bit_def by simp
+
−
+
−lemma less_eq_Bit1_Pls:
+
−  "Numeral.Bit k bit.B1 \<le> Numeral.Pls \<longleftrightarrow> k < Numeral.Pls"
+
−  unfolding Pls_def Bit_def by auto
+
−
+
−lemma less_eq_Bit_Min:
+
−  "Numeral.Bit k v \<le> Numeral.Min \<longleftrightarrow> k \<le> Numeral.Min"
+
−  unfolding Min_def Bit_def by (cases v) auto
+
−
+
−lemma less_eq_Bit0_Bit:
+
−  "Numeral.Bit k1 bit.B0 \<le> Numeral.Bit k2 v \<longleftrightarrow> k1 \<le> k2"
+
−  unfolding Min_def Bit_def bit.cases by (cases v) auto
+
−
+
−lemma less_eq_Bit_Bit1:
+
−  "Numeral.Bit k1 v \<le> Numeral.Bit k2 bit.B1 \<longleftrightarrow> k1 \<le> k2"
+
−  unfolding Min_def Bit_def bit.cases by (cases v) auto
+
−
+
−lemmas less_eq_numeral_code [code func] = less_eq_Pls_Pls less_eq_Pls_Min less_eq_Pls_Bit
+
−  less_eq_Min_Pls less_eq_Min_Min less_eq_Min_Bit0 less_eq_Min_Bit1
+
−  less_eq_Bit0_Pls less_eq_Bit1_Pls less_eq_Bit_Min less_eq_Bit0_Bit less_eq_Bit_Bit1
+
−
+
−end