1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
|
theory HoTT_Theorems
imports HoTT
begin
text "A bunch of theorems and other statements for sanity-checking, as well as things that should be automatically simplified.
Things that *should* be automated:
\<bullet> Checking that \<open>A\<close> is a well-formed type, when writing things like \<open>x : A\<close> and \<open>A : U\<close>.
\<bullet> Checking that the argument to a (dependent/non-dependent) function matches the type? Also the arguments to a pair?"
\<comment> \<open>Turn on trace for unification and the simplifier, for debugging.\<close>
declare[[unify_trace_simp, unify_trace_types, simp_trace, simp_trace_depth_limit=5]]
section \<open>\<Prod> type\<close>
subsection \<open>Typing functions\<close>
text "Declaring \<open>Prod_intro\<close> with the \<open>intro\<close> attribute (in HoTT.thy) enables \<open>standard\<close> to prove the following."
proposition assumes "A : U" shows "\<^bold>\<lambda>x:A. x : A\<rightarrow>A" using assms ..
proposition
assumes "A : U" and "A \<equiv> B"
shows "\<^bold>\<lambda>x:A. x : B\<rightarrow>A"
proof -
have "A\<rightarrow>A \<equiv> B\<rightarrow>A" using assms by simp
moreover have "\<^bold>\<lambda>x:A. x : A\<rightarrow>A" using assms(1) ..
ultimately show "\<^bold>\<lambda>x:A. x : B\<rightarrow>A" by simp
qed
proposition
assumes "A : U" and "B : U"
shows "\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B. x : A\<rightarrow>B\<rightarrow>A"
proof
fix x
assume "x : A"
with assms(2) show "\<^bold>\<lambda>y:B. x : B\<rightarrow>A" ..
qed (rule assms)
subsection \<open>Function application\<close>
proposition assumes "a : A" shows "(\<^bold>\<lambda>x:A. x)`a \<equiv> a" using assms by simp
text "Currying:"
lemma
assumes "a : A"
shows "(\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B(x). y)`a \<equiv> \<^bold>\<lambda>y:B(a). y"
proof
show "\<And>x. a : A \<Longrightarrow> x : A \<Longrightarrow> \<^bold>\<lambda>y:B x. y : B x \<rightarrow> B x"
lemma "\<lbrakk>a : A; b : B\<rbrakk> \<Longrightarrow> (\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B(x). y)`a`b \<equiv> b" by simp
lemma "a : A \<Longrightarrow> (\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B(x). f x y)`a \<equiv> \<^bold>\<lambda>y:B(a). f a y" by simp
lemma "\<lbrakk>a : A; b : B(a); c : C(a)(b)\<rbrakk> \<Longrightarrow> (\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B(x). \<^bold>\<lambda>z:C(x)(y). f x y z)`a`b`c \<equiv> f a b c" by simp
subsection \<open>Currying functions\<close>
proposition curried_function_formation:
fixes
A::Term and
B::"Term \<Rightarrow> Term" and
C::"Term \<Rightarrow> Term \<Rightarrow> Term"
assumes
"A : U" and
"B: A \<rightarrow> U" and
"\<And>x::Term. C(x): B(x) \<rightarrow> U"
shows "\<Prod>x:A. \<Prod>y:B(x). C x y : U"
proof
fix x::Term
assume *: "x : A"
show "\<Prod>y:B(x). C x y : U"
proof
show "B(x) : U" using * by (rule assms)
qed (rule assms)
qed (rule assms)
(**** GOOD CANDIDATE FOR AUTOMATION - EISBACH! ****)
proposition higher_order_currying_formation:
fixes
A::Term and
B::"Term \<Rightarrow> Term" and
C::"[Term, Term] \<Rightarrow> Term" and
D::"[Term, Term, Term] \<Rightarrow> Term"
assumes
"A : U" and
"B: A \<rightarrow> U" and
"\<And>x y::Term. y : B(x) \<Longrightarrow> C(x)(y) : U" and
"\<And>x y z::Term. z : C(x)(y) \<Longrightarrow> D(x)(y)(z) : U"
shows "\<Prod>x:A. \<Prod>y:B(x). \<Prod>z:C(x)(y). D(x)(y)(z) : U"
proof
fix x::Term assume 1: "x : A"
show "\<Prod>y:B(x). \<Prod>z:C(x)(y). D(x)(y)(z) : U"
proof
show "B(x) : U" using 1 by (rule assms)
fix y::Term assume 2: "y : B(x)"
show "\<Prod>z:C(x)(y). D(x)(y)(z) : U"
proof
show "C x y : U" using 2 by (rule assms)
show "\<And>z::Term. z : C(x)(y) \<Longrightarrow> D(x)(y)(z) : U" by (rule assms)
qed
qed
qed (rule assms)
(**** AND PROBABLY THIS TOO? ****)
lemma curried_type_judgment:
fixes
a b A::Term and
B::"Term \<Rightarrow> Term" and
f C::"[Term, Term] \<Rightarrow> Term"
assumes "\<And>x y::Term. \<lbrakk>x : A; y : B(x)\<rbrakk> \<Longrightarrow> f x y : C x y"
shows "\<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B(x). f x y : \<Prod>x:A. \<Prod>y:B(x). C x y"
proof
fix x::Term
assume *: "x : A"
show "\<^bold>\<lambda>y:B(x). f x y : \<Prod>y:B(x). C x y"
proof
fix y::Term
assume **: "y : B(x)"
show "f x y : C x y" using * ** by (rule assms)
qed
qed
text "Note that the propositions and proofs above often say nothing about the well-formedness of the types, or the well-typedness of the lambdas involved; one has to be very explicit and prove such things separately!
This is the result of the choices made regarding the premises of the type rules."
section \<open>\<Sum> type\<close>
text "The following shows that the dependent sum inductor has the type we expect it to have:"
lemma
assumes "C: \<Sum>x:A. B(x) \<rightarrow> U"
shows "indSum(C) : \<Prod>f:(\<Prod>x:A. \<Prod>y:B(x). C((x,y))). \<Prod>p:(\<Sum>x:A. B(x)). C(p)"
proof -
define F and P where
"F \<equiv> \<Prod>x:A. \<Prod>y:B(x). C((x,y))" and
"P \<equiv> \<Sum>x:A. B(x)"
have "\<^bold>\<lambda>f:F. \<^bold>\<lambda>p:P. indSum(C)`f`p : \<Prod>f:F. \<Prod>p:P. C(p)"
proof (rule curried_type_judgment)
fix f p::Term
assume "f : F" and "p : P"
with assms show "indSum(C)`f`p : C(p)" unfolding F_def P_def ..
qed
then show "indSum(C) : \<Prod>f:F. \<Prod>p:P. C(p)" by simp
qed
(**** AUTOMATION CANDIDATE ****)
text "Propositional uniqueness principle for dependent sums:"
text "We would like to eventually automate proving that 'a given type \<open>A\<close> is inhabited', i.e. search for an element \<open>a:A\<close>.
A good starting point would be to automate the application of elimination rules."
notepad begin
fix A B assume "A : U" and "B: A \<rightarrow> U"
define C where "C \<equiv> \<lambda>p. p =[\<Sum>x:A. B(x)] (fst[A,B]`p, snd[A,B]`p)"
have *: "C: \<Sum>x:A. B(x) \<rightarrow> U"
proof -
fix p assume "p : \<Sum>x:A. B(x)"
have "(fst[A,B]`p, snd[A,B]`p) : \<Sum>x:A. B(x)"
define f where "f \<equiv> \<^bold>\<lambda>x:A. \<^bold>\<lambda>y:B(x). refl((x,y))"
have "f`x`y : C((x,y))"
sorry
have "p : \<Sum>x:A. B(x) \<Longrightarrow> indSum(C)`f`p : C(p)" using * ** by (rule Sum_elim)
end
section \<open>Universes and polymorphism\<close>
text "Polymorphic identity function."
consts Ui::Term
definition Id where "Id \<equiv> \<^bold>\<lambda>A:Ui. \<^bold>\<lambda>x:A. x"
(*
section \<open>Natural numbers\<close>
text "Here's a dumb proof that 2 is a natural number."
proposition "succ(succ 0) : Nat"
proof -
have "0 : Nat" by (rule Nat_intro1)
from this have "(succ 0) : Nat" by (rule Nat_intro2)
thus "succ(succ 0) : Nat" by (rule Nat_intro2)
qed
text "We can of course iterate the above for as many applications of \<open>succ\<close> as we like.
The next thing to do is to implement induction to automate such proofs.
When we get more stuff working, I'd like to aim for formalizing the encode-decode method to be able to prove the only naturals are 0 and those obtained from it by \<open>succ\<close>."
*)
end
|
