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
|
(********
Isabelle/HoTT: Dependent product (dependent function)
Feb 2019
********)
theory Prod
imports HoTT_Base HoTT_Methods
begin
section \<open>Basic type definitions\<close>
axiomatization
Prod :: "[t, t \<Rightarrow> t] \<Rightarrow> t" and
lam :: "[t, t \<Rightarrow> t] \<Rightarrow> t" and
app :: "[t, t] \<Rightarrow> t" ("(2_`/_)" [120, 121] 120)
\<comment> \<open>Application should bind tighter than abstraction.\<close>
syntax
"_Prod" :: "[idt, t, t] \<Rightarrow> t" ("(2\<Prod>'(_: _')./ _)" 30)
"_Prod'" :: "[idt, t, t] \<Rightarrow> t" ("(2\<Prod>_: _./ _)" 30)
"_lam" :: "[idt, t, t] \<Rightarrow> t" ("(2\<lambda>'(_: _')./ _)" 30)
"_lam'" :: "[idt, t, t] \<Rightarrow> t" ("(2\<lambda>_: _./ _)" 30)
translations
"\<Prod>x: A. B" \<rightleftharpoons> "(CONST Prod) A (\<lambda>x. B)"
"\<Prod>(x: A). B" \<rightleftharpoons> "(CONST Prod) A (\<lambda>x. B)"
"\<lambda>(x: A). b" \<rightleftharpoons> "(CONST lam) A (\<lambda>x. b)"
"\<lambda>x: A. b" \<rightleftharpoons> "(CONST lam) A (\<lambda>x. b)"
text \<open>
The syntax translations above bind the variable @{term x} in the expressions @{term B} and @{term b}.
\<close>
text \<open>Non-dependent functions are a special case:\<close>
abbreviation Fun :: "[t, t] \<Rightarrow> t" (infixr "\<rightarrow>" 40)
where "A \<rightarrow> B \<equiv> \<Prod>_: A. B"
axiomatization where
\<comment> \<open>Type rules\<close>
Prod_form: "\<lbrakk>A: U i; B: A \<leadsto> U i\<rbrakk> \<Longrightarrow> \<Prod>x: A. B x: U i" and
Prod_intro: "\<lbrakk>\<And>x. x: A \<Longrightarrow> b x: B x; A: U i\<rbrakk> \<Longrightarrow> \<lambda>x: A. b x: \<Prod>x: A. B x" and
Prod_elim: "\<lbrakk>f: \<Prod>x: A. B x; a: A\<rbrakk> \<Longrightarrow> f`a: B a" and
Prod_cmp: "\<lbrakk>\<And>x. x: A \<Longrightarrow> b x: B x; a: A\<rbrakk> \<Longrightarrow> (\<lambda>x: A. b x)`a \<equiv> b a" and
Prod_uniq: "f: \<Prod>x: A. B x \<Longrightarrow> \<lambda>x: A. f`x \<equiv> f" and
\<comment> \<open>Congruence rules\<close>
Prod_form_eq: "\<lbrakk>A: U i; B: A \<leadsto> U i; C: A \<leadsto> U i; \<And>x. x: A \<Longrightarrow> B x \<equiv> C x\<rbrakk>
\<Longrightarrow> \<Prod>x: A. B x \<equiv> \<Prod>x: A. C x" and
Prod_intro_eq: "\<lbrakk>\<And>x. x: A \<Longrightarrow> b x \<equiv> c x; A: U i\<rbrakk> \<Longrightarrow> \<lambda>x: A. b x \<equiv> \<lambda>x: A. c x"
text \<open>
The Pure rules for \<open>\<equiv>\<close> only let us judge strict syntactic equality of object lambda expressions.
The actual definitional equality rule in the type theory is @{thm Prod_intro_eq}.
Note that this is a separate rule from function extensionality.
\<close>
lemmas Prod_form [form]
lemmas Prod_routine [intro] = Prod_form Prod_intro Prod_elim
lemmas Prod_comp [comp] = Prod_cmp Prod_uniq
lemmas Prod_cong [cong] = Prod_form_eq Prod_intro_eq
section \<open>Function composition\<close>
definition compose :: "[t, t, t] \<Rightarrow> t"
where "compose A g f \<equiv> \<lambda>x: A. g`(f`x)"
syntax "_compose" :: "[t, t, t] \<Rightarrow> t" ("(2_ o[_]/ _)" [111, 0, 110] 110)
translations "g o[A] f" \<rightleftharpoons> "(CONST compose) A g f"
text \<open>The composition @{term "g o[A] f"} is annotated with the domain @{term A} of @{term f}.\<close>
syntax "_compose'" :: "[t, t] \<Rightarrow> t" (infixr "o" 110)
text \<open>Pretty-printing switch for composition; hides domain type information.\<close>
ML \<open>val pretty_compose = Attrib.setup_config_bool @{binding "pretty_compose"} (K true)\<close>
print_translation \<open>
let fun compose_tr' ctxt [A, g, f] =
if Config.get ctxt pretty_compose
then Syntax.const @{syntax_const "_compose'"} $ g $ f
else @{const compose} $ A $ g $ f
in
[(@{const_syntax compose}, compose_tr')]
end
\<close>
lemma compose_type:
assumes
"A: U i" and "B: U i" and "C: B \<leadsto> U i" and
"f: A \<rightarrow> B" and "g: \<Prod>x: B. C x"
shows "g o[A] f: \<Prod>x: A. C (f`x)"
unfolding compose_def by (derive lems: assms)
lemma compose_comp:
assumes "A: U i" and "\<And>x. x: A \<Longrightarrow> b x: B" and "\<And>x. x: B \<Longrightarrow> c x: C x"
shows "(\<lambda>x: B. c x) o[A] (\<lambda>x: A. b x) \<equiv> \<lambda>x: A. c (b x)"
unfolding compose_def by (derive lems: assms cong)
declare
compose_type [intro]
compose_comp [comp]
lemma compose_assoc:
assumes "A: U i" and "f: A \<rightarrow> B" and "g: B \<rightarrow> C" and "h: \<Prod>x: C. D x"
shows "(h o[B] g) o[A] f \<equiv> h o[A] g o[A] f"
unfolding compose_def by (derive lems: assms cong)
abbreviation id :: "t \<Rightarrow> t" ("(id _)" [115] 114) where "id A \<equiv> \<lambda>x: A. x"
lemma id_type [intro]: "\<And>A. A: U i \<Longrightarrow> id A: A \<rightarrow> A" by derive
section \<open>Universal quantification\<close>
text \<open>
It will often be useful to convert a proof goal asserting the inhabitation of a dependent product to one that instead uses Pure universal quantification.
Method @{theory_text quantify_all} converts the goal
@{text "t: \<Prod>x1: A1. ... \<Prod>xn: An. B x1 ... xn"},
where @{term B} is not a product, to
@{text "\<And>x1 ... xn . \<lbrakk>x1: A1; ...; xn: An\<rbrakk> \<Longrightarrow> ?b x1 ... xn: B x1 ... xn"}.
Method @{theory_text "quantify k"} does the same, but only for the first k unknowns.
\<close>
method quantify_all = (rule Prod_intro)+
method_setup quantify = \<open>repeat (fn ctxt => Method.rule_tac ctxt [@{thm Prod_intro}] [] 1)\<close>
end
|
