Quotient-Paper/Paper-old.thy
author Christian Urban <christian dot urban at kcl dot ac dot uk>
Thu, 09 Jul 2015 09:12:44 +0100
changeset 3240 f80fa0d18d81
parent 2442 1f9360daf6e1
permissions -rw-r--r--
updated examples
Ignore whitespace changes - Everywhere: Within whitespace: At end of lines:
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     1
(*<*)
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     2
theory Paper
2183
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2182
diff changeset
     3
imports "Quotient"
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     4
        "LaTeXsugar"
2186
762a739c9eb4 added FSet to the correct paper
Christian Urban <urbanc@in.tum.de>
parents: 2183
diff changeset
     5
        "../Nominal/FSet"
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
     6
begin
1994
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1978
diff changeset
     7
2286
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
     8
(****
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
     9
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
    10
** things to do for the next version
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
    11
*
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
    12
* - what are quot_thms?
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
    13
* - what do all preservation theorems look like,
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
    14
    in particular preservation for quotient
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
    15
    compositions
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
    16
  - explain how Quotient R Abs Rep is proved (j-version)
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
    17
  - give an example where precise specification helps (core Haskell in nominal?)
2374
0321e89e66a3 quote for a new paper
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
    18
0321e89e66a3 quote for a new paper
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
    19
  - Quote from Peter:
0321e89e66a3 quote for a new paper
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
    20
0321e89e66a3 quote for a new paper
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
    21
    One might think quotient have been studied to death, but
2414
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2413
diff changeset
    22
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2413
diff changeset
    23
  - Mention Andreas Lochbiler in Acknowledgements and 'desceding'.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2413
diff changeset
    24
2286
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
    25
*)
e7bc2ae30faf added a few points that need to be looked at the next version of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2283
diff changeset
    26
1994
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1978
diff changeset
    27
notation (latex output)
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
    28
  rel_conj ("_ \<circ>\<circ>\<circ> _" [53, 53] 52) and
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
    29
  pred_comp ("_ \<circ>\<circ> _" [1, 1] 30) and
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
    30
  "op -->" (infix "\<longrightarrow>" 100) and
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
    31
  "==>" (infix "\<Longrightarrow>" 100) and
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
    32
  fun_map ("_ \<^raw:\mbox{\singlearr}> _" 51) and
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
    33
  fun_rel ("_ \<^raw:\mbox{\doublearr}> _" 51) and
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
    34
  list_eq (infix "\<approx>" 50) and (* Not sure if we want this notation...? *)
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
    35
  fempty ("\<emptyset>") and
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
    36
  funion ("_ \<union> _") and
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
    37
  finsert ("{_} \<union> _") and 
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
    38
  Cons ("_::_") and
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
    39
  concat ("flat") and
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
    40
  fconcat ("\<Union>")
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    41
 
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
    42
  
1994
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1978
diff changeset
    43
2182
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    44
ML {*
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    45
fun nth_conj n (_, r) = nth (HOLogic.dest_conj r) n;
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
    46
2182
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    47
fun style_lhs_rhs proj = Scan.succeed (fn ctxt => fn t =>
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    48
  let
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    49
    val concl =
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    50
      Object_Logic.drop_judgment (ProofContext.theory_of ctxt) (Logic.strip_imp_concl t)
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    51
  in
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    52
    case concl of (_ $ l $ r) => proj (l, r)
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    53
    | _ => error ("Binary operator expected in term: " ^ Syntax.string_of_term ctxt concl)
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    54
  end);
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    55
*}
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
    56
2182
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    57
setup {*
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    58
  Term_Style.setup "rhs1" (style_lhs_rhs (nth_conj 0)) #>
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    59
  Term_Style.setup "rhs2" (style_lhs_rhs (nth_conj 1)) #>
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    60
  Term_Style.setup "rhs3" (style_lhs_rhs (nth_conj 2))
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
    61
*}
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
    62
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
    63
(*>*)
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
    64
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
    65
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
    66
section {* Introduction *}
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
    67
2102
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
    68
text {* 
2205
69b4eb4b12c6 added larry's quote
Christian Urban <urbanc@in.tum.de>
parents: 2199
diff changeset
    69
   \begin{flushright}
69b4eb4b12c6 added larry's quote
Christian Urban <urbanc@in.tum.de>
parents: 2199
diff changeset
    70
  {\em ``Not using a [quotient] package has its advantages: we do not have to\\ 
69b4eb4b12c6 added larry's quote
Christian Urban <urbanc@in.tum.de>
parents: 2199
diff changeset
    71
    collect all the theorems we shall ever want into one giant list;''}\\
2213
231a20534950 improved abstract, some tuning
Christian Urban <urbanc@in.tum.de>
parents: 2212
diff changeset
    72
    Larry Paulson \cite{Paulson06}
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
    73
  \end{flushright}
2103
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
    74
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
    75
  \noindent
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    76
  Isabelle is a popular generic theorem prover in which many logics can be
2214
02e03d4287ec a bit more in the introduction and abstract
Christian Urban <urbanc@in.tum.de>
parents: 2213
diff changeset
    77
  implemented. The most widely used one, however, is Higher-Order Logic
02e03d4287ec a bit more in the introduction and abstract
Christian Urban <urbanc@in.tum.de>
parents: 2213
diff changeset
    78
  (HOL). This logic consists of a small number of axioms and inference rules
02e03d4287ec a bit more in the introduction and abstract
Christian Urban <urbanc@in.tum.de>
parents: 2213
diff changeset
    79
  over a simply-typed term-language. Safe reasoning in HOL is ensured by two
02e03d4287ec a bit more in the introduction and abstract
Christian Urban <urbanc@in.tum.de>
parents: 2213
diff changeset
    80
  very restricted mechanisms for extending the logic: one is the definition of
02e03d4287ec a bit more in the introduction and abstract
Christian Urban <urbanc@in.tum.de>
parents: 2213
diff changeset
    81
  new constants in terms of existing ones; the other is the introduction of
02e03d4287ec a bit more in the introduction and abstract
Christian Urban <urbanc@in.tum.de>
parents: 2213
diff changeset
    82
  new types by identifying non-empty subsets in existing types. It is well
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
    83
  understood how to use both mechanisms for dealing with quotient
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    84
  constructions in HOL (see \cite{Homeier05,Paulson06}).  For example the
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    85
  integers in Isabelle/HOL are constructed by a quotient construction over the
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    86
  type @{typ "nat \<times> nat"} and the equivalence relation
2102
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
    87
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
    88
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
    89
  @{text "(n\<^isub>1, n\<^isub>2) \<approx> (m\<^isub>1, m\<^isub>2) \<equiv> n\<^isub>1 + m\<^isub>2 = m\<^isub>1 + n\<^isub>2"}\hfill\numbered{natpairequiv}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
    90
  \end{isabelle}
2102
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
    91
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
    92
  \noindent
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
    93
  This constructions yields the new type @{typ int} and definitions for @{text
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    94
  "0"} and @{text "1"} of type @{typ int} can be given in terms of pairs of
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    95
  natural numbers (namely @{text "(0, 0)"} and @{text "(1, 0)"}). Operations
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    96
  such as @{text "add"} with type @{typ "int \<Rightarrow> int \<Rightarrow> int"} can be defined in
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
    97
  terms of operations on pairs of natural numbers (namely @{text
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
    98
  "add_pair (n\<^isub>1, m\<^isub>1) (n\<^isub>2,
2222
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
    99
  m\<^isub>2) \<equiv> (n\<^isub>1 + n\<^isub>2, m\<^isub>1 + m\<^isub>2)"}).
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
   100
  Similarly one can construct the type of finite sets, written @{term "\<alpha> fset"}, 
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   101
  by quotienting the type @{text "\<alpha> list"} according to the equivalence relation
2103
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
   102
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   103
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   104
  @{text "xs \<approx> ys \<equiv> (\<forall>x. memb x xs \<longleftrightarrow> memb x ys)"}\hfill\numbered{listequiv}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   105
  \end{isabelle}
2103
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
   106
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
   107
  \noindent
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   108
  which states that two lists are equivalent if every element in one list is
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   109
  also member in the other. The empty finite set, written @{term "{||}"}, can
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   110
  then be defined as the empty list and the union of two finite sets, written
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   111
  @{text "\<union>"}, as list append.
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   112
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   113
  Quotients are important in a variety of areas, but they are really ubiquitous in
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   114
  the area of reasoning about programming language calculi. A simple example
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   115
  is the lambda-calculus, whose raw terms are defined as
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   116
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   117
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   118
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   119
  @{text "t ::= x | t t | \<lambda>x.t"}\hfill\numbered{lambda}
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   120
  \end{isabelle}
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
   121
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   122
  \noindent
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   123
  The problem with this definition arises, for instance, when one attempts to
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   124
  prove formally the substitution lemma \cite{Barendregt81} by induction
2222
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
   125
  over the structure of terms. This can be fiendishly complicated (see
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   126
  \cite[Pages 94--104]{CurryFeys58} for some ``rough'' sketches of a proof
2226
36c9d9e658c7 some slight tuning of the intro
Christian Urban <urbanc@in.tum.de>
parents: 2225
diff changeset
   127
  about raw lambda-terms). In contrast, if we reason about
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   128
  $\alpha$-equated lambda-terms, that means terms quotient according to
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   129
  $\alpha$-equivalence, then the reasoning infrastructure provided, 
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   130
  for example, by Nominal Isabelle \cite{UrbanKaliszyk11} makes the formal 
2222
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
   131
  proof of the substitution lemma almost trivial. 
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   132
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   133
  The difficulty is that in order to be able to reason about integers, finite
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   134
  sets or $\alpha$-equated lambda-terms one needs to establish a reasoning
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   135
  infrastructure by transferring, or \emph{lifting}, definitions and theorems
2226
36c9d9e658c7 some slight tuning of the intro
Christian Urban <urbanc@in.tum.de>
parents: 2225
diff changeset
   136
  from the raw type @{typ "nat \<times> nat"} to the quotient type @{typ int}
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   137
  (similarly for finite sets and $\alpha$-equated lambda-terms). This lifting
2222
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
   138
  usually requires a \emph{lot} of tedious reasoning effort \cite{Paulson06}.  
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   139
  It is feasible to do this work manually, if one has only a few quotient
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   140
  constructions at hand. But if they have to be done over and over again, as in 
2222
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
   141
  Nominal Isabelle, then manual reasoning is not an option.
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   142
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   143
  The purpose of a \emph{quotient package} is to ease the lifting of theorems
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   144
  and automate the reasoning as much as possible. In the
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   145
  context of HOL, there have been a few quotient packages already
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   146
  \cite{harrison-thesis,Slotosch97}. The most notable one is by Homeier
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   147
  \cite{Homeier05} implemented in HOL4.  The fundamental construction these
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   148
  quotient packages perform can be illustrated by the following picture:
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
   149
2417
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   150
%%% FIXME: Referee 1 says:
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   151
%%% Diagram is unclear.  Firstly, isn't an existing type a "set (not sets) of raw elements"?
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   152
%%% Secondly, isn't the _set of_ equivalence classes mapped to and from the new type?
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   153
%%% Thirdly, what do the words "non-empty subset" refer to ?
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   154
2423
f5cbf74d4ec5 Clarifications to FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2422
diff changeset
   155
%%% Cezary: I like the diagram, maybe 'new type' could be outside, but otherwise
f5cbf74d4ec5 Clarifications to FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2422
diff changeset
   156
%%% I wouldn't change it.
f5cbf74d4ec5 Clarifications to FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2422
diff changeset
   157
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
   158
  \begin{center}
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   159
  \mbox{}\hspace{20mm}\begin{tikzpicture}
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   160
  %%\draw[step=2mm] (-4,-1) grid (4,1);
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   161
  
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   162
  \draw[very thick] (0.7,0.3) circle (4.85mm);
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   163
  \draw[rounded corners=1mm, very thick] ( 0.0,-0.9) rectangle ( 1.8, 0.9);
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   164
  \draw[rounded corners=1mm, very thick] (-1.95,0.8) rectangle (-2.9,-0.195);
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   165
  
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   166
  \draw (-2.0, 0.8) --  (0.7,0.8);
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   167
  \draw (-2.0,-0.195)  -- (0.7,-0.195);
2103
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
   168
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   169
  \draw ( 0.7, 0.23) node {\begin{tabular}{@ {}c@ {}}equiv-\\[-1mm]clas.\end{tabular}};
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   170
  \draw (-2.45, 0.35) node {\begin{tabular}{@ {}c@ {}}new\\[-1mm]type\end{tabular}};
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   171
  \draw (1.8, 0.35) node[right=-0.1mm]
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   172
    {\begin{tabular}{@ {}l@ {}}existing\\[-1mm] type\\ (sets of raw elements)\end{tabular}};
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   173
  \draw (0.9, -0.55) node {\begin{tabular}{@ {}l@ {}}non-empty\\[-1mm]subset\end{tabular}};
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   174
  
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   175
  \draw[->, very thick] (-1.8, 0.36) -- (-0.1,0.36);
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   176
  \draw[<-, very thick] (-1.8, 0.16) -- (-0.1,0.16);
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   177
  \draw (-0.95, 0.26) node[above=0.4mm] {@{text Rep}};
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   178
  \draw (-0.95, 0.26) node[below=0.4mm] {@{text Abs}};
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   179
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   180
  \end{tikzpicture}
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   181
  \end{center}
2217
fc5bfd0cc1cd more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2215
diff changeset
   182
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   183
  \noindent
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   184
  The starting point is an existing type, to which we refer as the
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   185
  \emph{raw type} and over which an equivalence relation given by the user is
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   186
  defined. With this input the package introduces a new type, to which we
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   187
  refer as the \emph{quotient type}. This type comes with an
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   188
  \emph{abstraction} and a \emph{representation} function, written @{text Abs}
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   189
  and @{text Rep}.\footnote{Actually slightly more basic functions are given;
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   190
  the functions @{text Abs} and @{text Rep} need to be derived from them. We
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   191
  will show the details later. } They relate elements in the
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   192
  existing type to elements in the new type and vice versa, and can be uniquely
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   193
  identified by their quotient type. For example for the integer quotient construction
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   194
  the types of @{text Abs} and @{text Rep} are
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   195
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   196
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   197
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   198
  @{text "Abs :: nat \<times> nat \<Rightarrow> int"}\hspace{10mm}@{text "Rep :: int \<Rightarrow> nat \<times> nat"}
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   199
  \end{isabelle}
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   200
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   201
  \noindent
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   202
  We therefore often write @{text Abs_int} and @{text Rep_int} if the
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   203
  typing information is important. 
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   204
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   205
  Every abstraction and representation function stands for an isomorphism
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   206
  between the non-empty subset and elements in the new type. They are
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   207
  necessary for making definitions involving the new type. For example @{text
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   208
  "0"} and @{text "1"} of type @{typ int} can be defined as
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   209
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   210
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   211
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   212
  @{text "0 \<equiv> Abs_int (0, 0)"}\hspace{10mm}@{text "1 \<equiv> Abs_int (1, 0)"}
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   213
  \end{isabelle}
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   214
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   215
  \noindent
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   216
  Slightly more complicated is the definition of @{text "add"} having type 
2222
973649d612f8 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2221
diff changeset
   217
  @{typ "int \<Rightarrow> int \<Rightarrow> int"}. Its definition is as follows
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   218
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   219
   \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   220
  @{text "add n m \<equiv> Abs_int (add_pair (Rep_int n) (Rep_int m))"}
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   221
  \hfill\numbered{adddef}
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   222
  \end{isabelle}
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   223
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   224
  \noindent
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   225
  where we take the representation of the arguments @{text n} and @{text m},
2226
36c9d9e658c7 some slight tuning of the intro
Christian Urban <urbanc@in.tum.de>
parents: 2225
diff changeset
   226
  add them according to the function @{text "add_pair"} and then take the
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   227
  abstraction of the result.  This is all straightforward and the existing
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
   228
  quotient packages can deal with such definitions. But what is surprising is
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   229
  that none of them can deal with slightly more complicated definitions involving
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   230
  \emph{compositions} of quotients. Such compositions are needed for example
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   231
  in case of quotienting lists to yield finite sets and the operator that 
2226
36c9d9e658c7 some slight tuning of the intro
Christian Urban <urbanc@in.tum.de>
parents: 2225
diff changeset
   232
  flattens lists of lists, defined as follows
2102
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
   233
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   234
  @{thm [display, indent=10] concat.simps(1) concat.simps(2)[no_vars]}
2183
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2182
diff changeset
   235
2103
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
   236
  \noindent
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   237
  We expect that the corresponding operator on finite sets, written @{term "fconcat"},
2248
Christian Urban <urbanc@in.tum.de>
parents: 2247 2246
diff changeset
   238
  builds finite unions of finite sets:
2188
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   239
2220
2c4c0d93daa6 more to the introduction of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2217
diff changeset
   240
  @{thm [display, indent=10] fconcat_empty[no_vars] fconcat_insert[no_vars]}
2102
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
   241
2103
e08e3c29dbc0 a bit for the introduction of the q-paper
Christian Urban <urbanc@in.tum.de>
parents: 2102
diff changeset
   242
  \noindent
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   243
  The quotient package should automatically provide us with a definition for @{text "\<Union>"} in
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   244
  terms of @{text flat}, @{text Rep_fset} and @{text Abs_fset}. The problem is 
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   245
  that the method  used in the existing quotient
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   246
  packages of just taking the representation of the arguments and then taking
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   247
  the abstraction of the result is \emph{not} enough. The reason is that in case
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   248
  of @{text "\<Union>"} we obtain the incorrect definition
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   249
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   250
  @{text [display, indent=10] "\<Union> S \<equiv> Abs_fset (flat (Rep_fset S))"}
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   251
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   252
  \noindent
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   253
  where the right-hand side is not even typable! This problem can be remedied in the
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   254
  existing quotient packages by introducing an intermediate step and reasoning
2226
36c9d9e658c7 some slight tuning of the intro
Christian Urban <urbanc@in.tum.de>
parents: 2225
diff changeset
   255
  about flattening of lists of finite sets. However, this remedy is rather
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   256
  cumbersome and inelegant in light of our work, which can deal with such
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   257
  definitions directly. The solution is that we need to build aggregate
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   258
  representation and abstraction functions, which in case of @{text "\<Union>"}
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   259
  generate the following definition
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   260
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   261
  @{text [display, indent=10] "\<Union> S \<equiv> Abs_fset (flat ((map_list Rep_fset \<circ> Rep_fset) S))"}
2221
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   262
e749cefbf66c even more on the qpaper (intro almost done)
Christian Urban <urbanc@in.tum.de>
parents: 2220
diff changeset
   263
  \noindent
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   264
  where @{term map_list} is the usual mapping function for lists. In this paper we
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   265
  will present a formal definition of our aggregate abstraction and
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   266
  representation functions (this definition was omitted in \cite{Homeier05}). 
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   267
  They generate definitions, like the one above for @{text "\<Union>"}, 
2226
36c9d9e658c7 some slight tuning of the intro
Christian Urban <urbanc@in.tum.de>
parents: 2225
diff changeset
   268
  according to the type of the raw constant and the type
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
   269
  of the quotient constant. This means we also have to extend the notions
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   270
  of \emph{aggregate equivalence relation}, \emph{respectfulness} and \emph{preservation}
2231
01d08af79f01 qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2230
diff changeset
   271
  from Homeier \cite{Homeier05}.
2223
c474186439bd more intro
Christian Urban <urbanc@in.tum.de>
parents: 2222
diff changeset
   272
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   273
  In addition we are able to address the criticism by Paulson \cite{Paulson06} cited
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   274
  at the beginning of this section about having to collect theorems that are
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   275
  lifted from the raw level to the quotient level into one giant list. Homeier's and
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   276
  also our quotient package are modular so that they allow lifting
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   277
  theorems separately. This has the advantage for the user of being able to develop a
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   278
  formal theory interactively as a natural progression. A pleasing side-result of
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   279
  the modularity is that we are able to clearly specify what is involved
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   280
  in the lifting process (this was only hinted at in \cite{Homeier05} and
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   281
  implemented as a ``rough recipe'' in ML-code).
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   282
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   283
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   284
  The paper is organised as follows: Section \ref{sec:prelims} presents briefly
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   285
  some necessary preliminaries; Section \ref{sec:type} describes the definitions 
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   286
  of quotient types and shows how definitions of constants can be made over 
2277
816204c76e90 Answer questions in comments
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2276
diff changeset
   287
  quotient types. Section \ref{sec:resp} introduces the notions of respectfulness
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   288
  and preservation; Section \ref{sec:lift} describes the lifting of theorems;
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   289
  Section \ref{sec:examples} presents some examples
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   290
  and Section \ref{sec:conc} concludes and compares our results to existing 
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   291
  work.
2102
200954544cae added some of the quotient literature; a bit more to the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 1994
diff changeset
   292
*}
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
   293
2257
Christian Urban <urbanc@in.tum.de>
parents: 2256
diff changeset
   294
section {* Preliminaries and General Quotients\label{sec:prelims} *}
1978
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
   295
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
   296
text {*
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   297
  We give in this section a crude overview of HOL and describe the main
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   298
  definitions given by Homeier for quotients \cite{Homeier05}.
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   299
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   300
  At its core, HOL is based on a simply-typed term language, where types are 
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   301
  recorded in Church-style fashion (that means, we can always infer the type of 
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   302
  a term and its subterms without any additional information). The grammars
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   303
  for types and terms are as follows
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   304
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   305
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   306
  \begin{tabular}{@ {}rl@ {\hspace{3mm}}l@ {}}
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   307
  @{text "\<sigma>, \<tau> ::="} & @{text "\<alpha> | (\<sigma>,\<dots>, \<sigma>) \<kappa>"} & (type variables and type constructors)\\
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   308
  @{text "t, s ::="} & @{text "x\<^isup>\<sigma> | c\<^isup>\<sigma> | t t | \<lambda>x\<^isup>\<sigma>. t"} & 
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   309
  (variables, constants, applications and abstractions)\\
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   310
  \end{tabular}
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   311
  \end{isabelle}
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   312
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   313
  \noindent
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   314
  We often write just @{text \<kappa>} for @{text "() \<kappa>"}, and use @{text "\<alpha>s"} and
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   315
  @{text "\<sigma>s"} to stand for collections of type variables and types,
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   316
  respectively.  The type of a term is often made explicit by writing @{text
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   317
  "t :: \<sigma>"}. HOL includes a type @{typ bool} for booleans and the function
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   318
  type, written @{text "\<sigma> \<Rightarrow> \<tau>"}. HOL also contains many primitive and defined
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   319
  constants; for example, a primitive constant is equality, with type @{text "= :: \<sigma> \<Rightarrow> \<sigma> \<Rightarrow>
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   320
  bool"}, and the identity function with type @{text "id :: \<sigma> \<Rightarrow> \<sigma>"} is
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   321
  defined as @{text "\<lambda>x\<^sup>\<sigma>. x\<^sup>\<sigma>"}.
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   322
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   323
  An important point to note is that theorems in HOL can be seen as a subset
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   324
  of terms that are constructed specially (namely through axioms and proof
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   325
  rules). As a result we are able to define automatic proof
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   326
  procedures showing that one theorem implies another by decomposing the term
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   327
  underlying the first theorem.
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   328
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   329
  Like Homeier's, our work relies on map-functions defined for every type
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   330
  constructor taking some arguments, for example @{text map_list} for lists. Homeier
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   331
  describes in \cite{Homeier05} map-functions for products, sums, options and
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   332
  also the following map for function types
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   333
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   334
  @{thm [display, indent=10] fun_map_def[no_vars, THEN eq_reflection]}
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   335
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   336
  \noindent
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   337
  Using this map-function, we can give the following, equivalent, but more 
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   338
  uniform definition for @{text add} shown in \eqref{adddef}:
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   339
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   340
  @{text [display, indent=10] "add \<equiv> (Rep_int \<singlearr> Rep_int \<singlearr> Abs_int) add_pair"}
2182
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   341
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   342
  \noindent
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   343
  Using extensionality and unfolding the definition of @{text "\<singlearr>"}, 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   344
  we can get back to \eqref{adddef}. 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   345
  In what follows we shall use the convention to write @{text "map_\<kappa>"} for a map-function 
2413
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   346
  of the type-constructor @{text \<kappa>}. For a type @{text \<kappa>} with arguments @{text "\<alpha>\<^isub>1\<^isub>\<dots>\<^isub>n"} the
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   347
  type of @{text "map_\<kappa>"} has to be @{text "\<alpha>\<^isub>1\<Rightarrow>\<dots>\<Rightarrow>\<alpha>\<^isub>n\<Rightarrow>\<alpha>\<^isub>1\<dots>\<alpha>\<^isub>n \<kappa>"}. For example @{text "map_list"}
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   348
  has to have the type @{text "\<alpha>\<Rightarrow>\<alpha> list"}.
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   349
  In our implementation we maintain
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   350
  a database of these map-functions that can be dynamically extended.
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   351
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   352
  It will also be necessary to have operators, referred to as @{text "rel_\<kappa>"},
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   353
  which define equivalence relations in terms of constituent equivalence
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   354
  relations. For example given two equivalence relations @{text "R\<^isub>1"}
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   355
  and @{text "R\<^isub>2"}, we can define an equivalence relations over 
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   356
  products as follows
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   357
  %
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   358
  @{text [display, indent=10] "(R\<^isub>1 \<tripple> R\<^isub>2) (x\<^isub>1, x\<^isub>2) (y\<^isub>1, y\<^isub>2) \<equiv> R\<^isub>1 x\<^isub>1 y\<^isub>1 \<and> R\<^isub>2 x\<^isub>2 y\<^isub>2"}
1978
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
   359
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   360
  \noindent
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   361
  Homeier gives also the following operator for defining equivalence 
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   362
  relations over function types
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   363
  %
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   364
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   365
  @{thm fun_rel_def[of "R\<^isub>1" "R\<^isub>2", no_vars, THEN eq_reflection]}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   366
  \hfill\numbered{relfun}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   367
  \end{isabelle}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   368
  
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   369
  \noindent
2319
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   370
  In the context of quotients, the following two notions from \cite{Homeier05} 
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   371
  are needed later on.
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   372
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   373
  \begin{definition}[Respects]\label{def:respects}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   374
  An element @{text "x"} respects a relation @{text "R"} provided @{text "R x x"}.
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   375
  \end{definition}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   376
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   377
  \begin{definition}[Bounded Quantification and Bounded Abstractions]\label{def:babs}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   378
  @{text "\<forall>x \<in> S. P x"} holds if for all @{text x}, @{text "x \<in> S"} implies @{text "P x"};
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   379
  and @{text "(\<lambda>x \<in> S. f x) = f x"} provided @{text "x \<in> S"}.
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   380
  \end{definition}
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   381
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   382
  The central definition in Homeier's work \cite{Homeier05} relates equivalence 
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   383
  relations, abstraction and representation functions:
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   384
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   385
  \begin{definition}[Quotient Types]
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   386
  Given a relation $R$, an abstraction function $Abs$
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   387
  and a representation function $Rep$, the predicate @{term "Quotient R Abs Rep"}
2413
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   388
  holds if and only if
2182
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   389
  \begin{enumerate}
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   390
  \item @{thm (rhs1) Quotient_def[of "R", no_vars]}
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   391
  \item @{thm (rhs2) Quotient_def[of "R", no_vars]}
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   392
  \item @{thm (rhs3) Quotient_def[of "R", no_vars]}
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   393
  \end{enumerate}
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   394
  \end{definition}
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   395
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   396
  \noindent
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   397
  The value of this definition lies in the fact that validity of @{text "Quotient R Abs Rep"} can 
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   398
  often be proved in terms of the validity of @{text "Quotient"} over the constituent 
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   399
  types of @{text "R"}, @{text Abs} and @{text Rep}. 
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   400
  For example Homeier proves the following property for higher-order quotient
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   401
  types:
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   402
 
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   403
  \begin{proposition}\label{funquot}
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   404
  @{thm[mode=IfThen] fun_quotient[where ?R1.0="R\<^isub>1" and ?R2.0="R\<^isub>2" 
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   405
      and ?abs1.0="Abs\<^isub>1" and ?abs2.0="Abs\<^isub>2" and ?rep1.0="Rep\<^isub>1" and ?rep2.0="Rep\<^isub>2"]}
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   406
  \end{proposition}
2195
0c1dcdefb515 Functionalized the ABS/REP definition.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2194
diff changeset
   407
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   408
  \noindent
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   409
  As a result, Homeier is able to build an automatic prover that can nearly
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   410
  always discharge a proof obligation involving @{text "Quotient"}. Our quotient
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   411
  package makes heavy 
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   412
  use of this part of Homeier's work including an extension 
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   413
  for dealing with compositions of equivalence relations defined as follows:
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   414
2417
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   415
%%% FIXME Referee 2 claims that composition-of-relations means OO, and this is also
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   416
%%% what wikipedia says. Any idea for a different name? Conjugation of Relations?
fc12e208a9e2 Add 2 FIXMEs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2416
diff changeset
   417
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   418
  \begin{definition}[Composition of Relations]
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   419
  @{abbrev "rel_conj R\<^isub>1 R\<^isub>2"} where @{text "\<circ>\<circ>"} is the predicate
2319
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   420
  composition defined by 
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   421
  @{thm (concl) pred_compI[of "R\<^isub>1" "x" "y" "R\<^isub>2" "z"]}
2366
46be4145b922 changes suggested by Peter Homeier
Christian Urban <urbanc@in.tum.de>
parents: 2334
diff changeset
   422
  holds if and only if there exists a @{text y} such that @{thm (prem 1) pred_compI[of "R\<^isub>1" "x" "y" "R\<^isub>2" "z"]} and
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   423
  @{thm (prem 2) pred_compI[of "R\<^isub>1" "x" "y" "R\<^isub>2" "z"]}.
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   424
  \end{definition}
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   425
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   426
  \noindent
2413
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   427
  Unfortunately a general quotient theorem for @{text "\<circ>\<circ>\<circ>"}, analogous to the one
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   428
  for @{text "\<singlearr>"} given in Proposition \ref{funquot}, would not be true
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   429
  in general. It cannot even be stated inside HOL, because of restrictions on types.
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   430
  However, we can prove specific instances of a
2282
fab7f09dda22 qpaper / address FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2281
diff changeset
   431
  quotient theorem for composing particular quotient relations.
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   432
  For example, to lift theorems involving @{term flat} the quotient theorem for 
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   433
  composing @{text "\<approx>\<^bsub>list\<^esub>"} will be necessary: given @{term "Quotient R Abs Rep"} 
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   434
  with @{text R} being an equivalence relation, then
2282
fab7f09dda22 qpaper / address FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2281
diff changeset
   435
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   436
  @{text [display, indent=2] "Quotient (rel_list R \<circ>\<circ>\<circ> \<approx>\<^bsub>list\<^esub>) (Abs_fset \<circ> map_list Abs) (map_list Rep \<circ> Rep_fset)"}
2282
fab7f09dda22 qpaper / address FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2281
diff changeset
   437
fab7f09dda22 qpaper / address FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2281
diff changeset
   438
  \vspace{-.5mm}
2195
0c1dcdefb515 Functionalized the ABS/REP definition.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2194
diff changeset
   439
*}
0c1dcdefb515 Functionalized the ABS/REP definition.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2194
diff changeset
   440
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   441
section {* Quotient Types and Quotient Definitions\label{sec:type} *}
1978
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
   442
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   443
text {*
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   444
  The first step in a quotient construction is to take a name for the new
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   445
  type, say @{text "\<kappa>\<^isub>q"}, and an equivalence relation, say @{text R},
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   446
  defined over a raw type, say @{text "\<sigma>"}. The type of the equivalence
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   447
  relation must be @{text "\<sigma> \<Rightarrow> \<sigma> \<Rightarrow> bool"}. The user-visible part of
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   448
  the quotient type declaration is therefore
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   449
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   450
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   451
  \isacommand{quotient\_type}~~@{text "\<alpha>s \<kappa>\<^isub>q = \<sigma> / R"}\hfill\numbered{typedecl}
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   452
  \end{isabelle}
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   453
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   454
  \noindent
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   455
  and a proof that @{text "R"} is indeed an equivalence relation. Two concrete
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   456
  examples are
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   457
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   458
  
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   459
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   460
  \begin{tabular}{@ {}l}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   461
  \isacommand{quotient\_type}~~@{text "int = nat \<times> nat / \<approx>\<^bsub>nat \<times> nat\<^esub>"}\\
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   462
  \isacommand{quotient\_type}~~@{text "\<alpha> fset = \<alpha> list / \<approx>\<^bsub>list\<^esub>"}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   463
  \end{tabular}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   464
  \end{isabelle}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   465
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   466
  \noindent
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   467
  which introduce the type of integers and of finite sets using the
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   468
  equivalence relations @{text "\<approx>\<^bsub>nat \<times> nat\<^esub>"} and @{text
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   469
  "\<approx>\<^bsub>list\<^esub>"} defined in \eqref{natpairequiv} and
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   470
  \eqref{listequiv}, respectively (the proofs about being equivalence
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   471
  relations is omitted).  Given this data, we define for declarations shown in
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   472
  \eqref{typedecl} the quotient types internally as
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   473
  
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   474
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   475
  \isacommand{typedef}~~@{text "\<alpha>s \<kappa>\<^isub>q = {c. \<exists>x. c = R x}"}
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   476
  \end{isabelle}
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   477
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   478
  \noindent
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   479
  where the right-hand side is the (non-empty) set of equivalence classes of
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   480
  @{text "R"}. The constraint in this declaration is that the type variables
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   481
  in the raw type @{text "\<sigma>"} must be included in the type variables @{text
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   482
  "\<alpha>s"} declared for @{text "\<kappa>\<^isub>q"}. HOL will then provide us with the following
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   483
  abstraction and representation functions 
2182
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2152
diff changeset
   484
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   485
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   486
  @{text "abs_\<kappa>\<^isub>q :: \<sigma> set \<Rightarrow> \<alpha>s \<kappa>\<^isub>q"}\hspace{10mm}@{text "rep_\<kappa>\<^isub>q :: \<alpha>s \<kappa>\<^isub>q \<Rightarrow> \<sigma> set"}
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   487
  \end{isabelle}
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   488
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   489
  \noindent 
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   490
  As can be seen from the type, they relate the new quotient type and equivalence classes of the raw
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   491
  type. However, as Homeier \cite{Homeier05} noted, it is much more convenient
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   492
  to work with the following derived abstraction and representation functions
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   493
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   494
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   495
  @{text "Abs_\<kappa>\<^isub>q x \<equiv> abs_\<kappa>\<^isub>q (R x)"}\hspace{10mm}@{text "Rep_\<kappa>\<^isub>q x \<equiv> \<epsilon> (rep_\<kappa>\<^isub>q x)"}
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   496
  \end{isabelle}
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   497
  
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   498
  \noindent
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   499
  on the expense of having to use Hilbert's choice operator @{text "\<epsilon>"} in the
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   500
  definition of @{text "Rep_\<kappa>\<^isub>q"}. These derived notions relate the
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   501
  quotient type and the raw type directly, as can be seen from their type,
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   502
  namely @{text "\<sigma> \<Rightarrow> \<alpha>s \<kappa>\<^isub>q"} and @{text "\<alpha>s \<kappa>\<^isub>q \<Rightarrow> \<sigma>"},
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   503
  respectively.  Given that @{text "R"} is an equivalence relation, the
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   504
  following property holds  for every quotient type 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   505
  (for the proof see \cite{Homeier05}).
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   506
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   507
  \begin{proposition}
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   508
  @{text "Quotient R Abs_\<kappa>\<^isub>q Rep_\<kappa>\<^isub>q"}.
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   509
  \end{proposition}
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   510
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   511
  The next step in a quotient construction is to introduce definitions of new constants
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   512
  involving the quotient type. These definitions need to be given in terms of concepts
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   513
  of the raw type (remember this is the only way how to extend HOL
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   514
  with new definitions). For the user the visible part of such definitions is the declaration
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   515
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   516
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   517
  \isacommand{quotient\_definition}~~@{text "c :: \<tau>"}~~\isacommand{is}~~@{text "t :: \<sigma>"}
2235
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   518
  \end{isabelle}
ad725de6e39b more on the constant lifting section
Christian Urban <urbanc@in.tum.de>
parents: 2234
diff changeset
   519
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   520
  \noindent
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   521
  where @{text t} is the definiens (its type @{text \<sigma>} can always be inferred)
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   522
  and @{text "c"} is the name of definiendum, whose type @{text "\<tau>"} needs to be
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   523
  given explicitly (the point is that @{text "\<tau>"} and @{text "\<sigma>"} can only differ 
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   524
  in places where a quotient and raw type is involved). Two concrete examples are
2188
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   525
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   526
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   527
  \begin{tabular}{@ {}l}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   528
  \isacommand{quotient\_definition}~~@{text "0 :: int"}~~\isacommand{is}~~@{text "(0::nat, 0::nat)"}\\
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   529
  \isacommand{quotient\_definition}~~@{text "\<Union> :: (\<alpha> fset) fset \<Rightarrow> \<alpha> fset"}~~%
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   530
  \isacommand{is}~~@{text "flat"} 
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   531
  \end{tabular}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   532
  \end{isabelle}
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   533
  
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   534
  \noindent
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   535
  The first one declares zero for integers and the second the operator for
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   536
  building unions of finite sets (@{text "flat"} having the type 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   537
  @{text "(\<alpha> list) list \<Rightarrow> \<alpha> list"}). 
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   538
2413
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
   539
  From such declarations given by the user, the quotient package needs to derive proper
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   540
  definitions using @{text "Abs"} and @{text "Rep"}. The data we rely on is the given quotient type
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   541
  @{text "\<tau>"} and the raw type @{text "\<sigma>"}.  They allow us to define \emph{aggregate
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   542
  abstraction} and \emph{representation functions} using the functions @{text "ABS (\<sigma>,
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   543
  \<tau>)"} and @{text "REP (\<sigma>, \<tau>)"} whose clauses we shall give below. The idea behind
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   544
  these two functions is to simultaneously descend into the raw types @{text \<sigma>} and 
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   545
  quotient types @{text \<tau>}, and generate the appropriate
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   546
  @{text "Abs"} and @{text "Rep"} in places where the types differ. Therefore
2269
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   547
  we generate just the identity whenever the types are equal. On the ``way'' down,
e4699a240d2c tuned everytinh up to section 4
Christian Urban <urbanc@in.tum.de>
parents: 2265
diff changeset
   548
  however we might have to use map-functions to let @{text Abs} and @{text Rep} act
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   549
  over the appropriate types. In what follows we use the short-hand notation 
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   550
  @{text "ABS (\<sigma>s, \<tau>s)"} to mean @{text "ABS (\<sigma>\<^isub>1, \<tau>\<^isub>1)\<dots>ABS (\<sigma>\<^isub>n, \<tau>\<^isub>n)"}; similarly 
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   551
  for @{text REP}.
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   552
  %
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   553
  \begin{center}
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   554
  \hfill
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   555
  \begin{tabular}{rcl}
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   556
  \multicolumn{3}{@ {\hspace{-4mm}}l}{equal types:}\\ 
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   557
  @{text "ABS (\<sigma>, \<sigma>)"} & $\dn$ & @{text "id :: \<sigma> \<Rightarrow> \<sigma>"}\\
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   558
  @{text "REP (\<sigma>, \<sigma>)"} & $\dn$ & @{text "id :: \<sigma> \<Rightarrow> \<sigma>"}\smallskip\\
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   559
  \multicolumn{3}{@ {\hspace{-4mm}}l}{function types:}\\ 
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   560
  @{text "ABS (\<sigma>\<^isub>1 \<Rightarrow> \<sigma>\<^isub>2, \<tau>\<^isub>1 \<Rightarrow> \<tau>\<^isub>2)"} & $\dn$ & @{text "REP (\<sigma>\<^isub>1, \<tau>\<^isub>1) \<singlearr> ABS (\<sigma>\<^isub>2, \<tau>\<^isub>2)"}\\
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   561
  @{text "REP (\<sigma>\<^isub>1 \<Rightarrow> \<sigma>\<^isub>2, \<tau>\<^isub>1 \<Rightarrow> \<tau>\<^isub>2)"} & $\dn$ & @{text "ABS (\<sigma>\<^isub>1, \<tau>\<^isub>1) \<singlearr> REP (\<sigma>\<^isub>2, \<tau>\<^isub>2)"}\smallskip\\
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   562
  \multicolumn{3}{@ {\hspace{-4mm}}l}{equal type constructors:}\\ 
2232
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   563
  @{text "ABS (\<sigma>s \<kappa>, \<tau>s \<kappa>)"} & $\dn$ & @{text "map_\<kappa> (ABS (\<sigma>s, \<tau>s))"}\\
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   564
  @{text "REP (\<sigma>s \<kappa>, \<tau>s \<kappa>)"} & $\dn$ & @{text "map_\<kappa> (REP (\<sigma>s, \<tau>s))"}\smallskip\\
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   565
  \multicolumn{3}{@ {\hspace{-4mm}}l}{unequal type constructors with @{text "\<alpha>s
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   566
  \<kappa>\<^isub>q"} being the quotient of the raw type @{text "\<rho>s \<kappa>"}:}\\
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   567
  @{text "ABS (\<sigma>s \<kappa>, \<tau>s \<kappa>\<^isub>q)"} & $\dn$ & @{text "Abs_\<kappa>\<^isub>q \<circ> (MAP(\<rho>s \<kappa>) (ABS (\<sigma>s', \<tau>s)))"}\\
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   568
  @{text "REP (\<sigma>s \<kappa>, \<tau>s \<kappa>\<^isub>q)"} & $\dn$ & @{text "(MAP(\<rho>s \<kappa>) (REP (\<sigma>s', \<tau>s))) \<circ> Rep_\<kappa>\<^isub>q"}
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   569
  \end{tabular}\hfill\numbered{ABSREP}
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   570
  \end{center}
2234
8035515bbbc6 something about the quotient ype definitions
Christian Urban <urbanc@in.tum.de>
parents: 2233
diff changeset
   571
  %
2232
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   572
  \noindent
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   573
  In the last two clauses we rely on the fact that the type @{text "\<alpha>s
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   574
  \<kappa>\<^isub>q"} is the quotient of the raw type @{text "\<rho>s \<kappa>"} (for example
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   575
  @{text "int"} and @{text "nat \<times> nat"}, or @{text "\<alpha> fset"} and @{text "\<alpha>
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   576
  list"}). The quotient construction ensures that the type variables in @{text
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   577
  "\<rho>s \<kappa>"} must be among the @{text "\<alpha>s"}. The @{text "\<sigma>s'"} are given by the
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   578
  substitutions for the @{text "\<alpha>s"} when matching  @{text "\<sigma>s \<kappa>"} against
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   579
  @{text "\<rho>s \<kappa>"}.  The
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   580
  function @{text "MAP"} calculates an \emph{aggregate map-function} for a raw
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   581
  type as follows:
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   582
  %
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   583
  \begin{center}
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   584
  \begin{tabular}{rcl}
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   585
  @{text "MAP' (\<alpha>)"} & $\dn$ & @{text "a\<^sup>\<alpha>"}\\
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   586
  @{text "MAP' (\<kappa>)"} & $\dn$ & @{text "id :: \<kappa> \<Rightarrow> \<kappa>"}\\
2232
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   587
  @{text "MAP' (\<sigma>s \<kappa>)"} & $\dn$ & @{text "map_\<kappa> (MAP'(\<sigma>s))"}\smallskip\\
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   588
  @{text "MAP (\<sigma>)"} & $\dn$ & @{text "\<lambda>as. MAP'(\<sigma>)"}  
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   589
  \end{tabular}
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   590
  \end{center}
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   591
  %
2232
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   592
  \noindent
2366
46be4145b922 changes suggested by Peter Homeier
Christian Urban <urbanc@in.tum.de>
parents: 2334
diff changeset
   593
  In this definition we rely on the fact that in the first clause we can interpret type-variables @{text \<alpha>} as 
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   594
  term variables @{text a}. In the last clause we build an abstraction over all
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   595
  term-variables of the map-function generated by the auxiliary function 
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   596
  @{text "MAP'"}.
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   597
  The need for aggregate map-functions can be seen in cases where we build quotients, 
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   598
  say @{text "(\<alpha>, \<beta>) \<kappa>\<^isub>q"}, out of compound raw types, say @{text "(\<alpha> list) \<times> \<beta>"}. 
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   599
  In this case @{text MAP} generates  the 
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   600
  aggregate map-function:
2232
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   601
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   602
%%% FIXME: Reviewer 2 asks: last two lines defining ABS and REP for
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   603
%%% unequal type constructors: How are the $\varrho$s defined? The
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   604
%%% following paragraph mentions them, but this paragraph is unclear,
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   605
%%% since it then mentions $\alpha$s, which do not seem to be defined
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   606
%%% either. As a result, I do not understand the first two sentences
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   607
%%% in this paragraph. I can imagine roughly what the following
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   608
%%% sentence `The $\sigma$s' are given by the matchers for the
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   609
%%% $\alpha$s$ when matching $\varrho$s $\kappa$ against $\sigma$s
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   610
%%% $\kappa$.' means, but also think that it is too vague.
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   611
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   612
  @{text [display, indent=10] "\<lambda>a b. map_prod (map_list a) b"}
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   613
  
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   614
  \noindent
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   615
  which is essential in order to define the corresponding aggregate 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   616
  abstraction and representation functions.
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   617
  
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   618
  To see how these definitions pan out in practise, let us return to our
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   619
  example about @{term "concat"} and @{term "fconcat"}, where we have the raw type
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   620
  @{text "(\<alpha> list) list \<Rightarrow> \<alpha> list"} and the quotient type @{text "(\<alpha> fset) fset \<Rightarrow> \<alpha>
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   621
  fset"}. Feeding these types into @{text ABS} gives us (after some @{text "\<beta>"}-simplifications)
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   622
  the abstraction function
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   623
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   624
  @{text [display, indent=10] "(map_list (map_list id \<circ> Rep_fset) \<circ> Rep_fset) \<singlearr> Abs_fset \<circ> map_list id"}
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   625
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   626
  \noindent
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   627
  In our implementation we further
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   628
  simplify this function by rewriting with the usual laws about @{text
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   629
  "map"}s and @{text "id"}, for example @{term "map_list id = id"} and @{text "f \<circ> id =
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   630
  id \<circ> f = f"}. This gives us the simpler abstraction function
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   631
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   632
  @{text [display, indent=10] "(map_list Rep_fset \<circ> Rep_fset) \<singlearr> Abs_fset"}
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   633
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   634
  \noindent
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   635
  which we can use for defining @{term "fconcat"} as follows
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   636
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   637
  @{text [display, indent=10] "\<Union> \<equiv> ((map_list Rep_fset \<circ> Rep_fset) \<singlearr> Abs_fset) flat"}
2232
f49b5dfabd59 improved definition of ABS and REP
Christian Urban <urbanc@in.tum.de>
parents: 2231
diff changeset
   638
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   639
  \noindent
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   640
  Note that by using the operator @{text "\<singlearr>"} and special clauses
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   641
  for function types in \eqref{ABSREP}, we do not have to 
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   642
  distinguish between arguments and results, but can deal with them uniformly.
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   643
  Consequently, all definitions in the quotient package 
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   644
  are of the general form
2188
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   645
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   646
  @{text [display, indent=10] "c \<equiv> ABS (\<sigma>, \<tau>) t"}
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   647
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   648
  \noindent
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   649
  where @{text \<sigma>} is the type of the definiens @{text "t"} and @{text "\<tau>"} the
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   650
  type of the defined quotient constant @{text "c"}. This data can be easily
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   651
  generated from the declaration given by the user.
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   652
  To increase the confidence in this way of making definitions, we can prove 
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   653
  that the terms involved are all typable.
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   654
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   655
  \begin{lemma}
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   656
  If @{text "ABS (\<sigma>, \<tau>)"} returns some abstraction function @{text "Abs"} 
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   657
  and @{text "REP (\<sigma>, \<tau>)"} some representation function @{text "Rep"}, 
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   658
  then @{text "Abs"} is of type @{text "\<sigma> \<Rightarrow> \<tau>"} and @{text "Rep"} of type
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   659
  @{text "\<tau> \<Rightarrow> \<sigma>"}.
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
   660
  \end{lemma}
2233
22c6b6144abd added some examples
Christian Urban <urbanc@in.tum.de>
parents: 2232
diff changeset
   661
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   662
  \begin{proof}
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   663
  By mutual induction and analysing the definitions of @{text "ABS"} and @{text "REP"}. 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   664
  The cases of equal types and function types are
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   665
  straightforward (the latter follows from @{text "\<singlearr>"} having the
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   666
  type @{text "(\<alpha> \<Rightarrow> \<beta>) \<Rightarrow> (\<gamma> \<Rightarrow> \<delta>) \<Rightarrow> (\<beta> \<Rightarrow> \<gamma>) \<Rightarrow> (\<alpha> \<Rightarrow> \<delta>)"}). In case of equal type
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   667
  constructors we can observe that a map-function after applying the functions
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   668
  @{text "ABS (\<sigma>s, \<tau>s)"} produces a term of type @{text "\<sigma>s \<kappa> \<Rightarrow> \<tau>s \<kappa>"}.  The
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   669
  interesting case is the one with unequal type constructors. Since we know
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   670
  the quotient is between @{text "\<alpha>s \<kappa>\<^isub>q"} and @{text "\<rho>s \<kappa>"}, we have
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   671
  that @{text "Abs_\<kappa>\<^isub>q"} is of type @{text "\<rho>s \<kappa> \<Rightarrow> \<alpha>s
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   672
  \<kappa>\<^isub>q"}. This type can be more specialised to @{text "\<rho>s[\<tau>s] \<kappa> \<Rightarrow> \<tau>s
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   673
  \<kappa>\<^isub>q"} where the type variables @{text "\<alpha>s"} are instantiated with the
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   674
  @{text "\<tau>s"}. The complete type can be calculated by observing that @{text
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   675
  "MAP (\<rho>s \<kappa>)"}, after applying the functions @{text "ABS (\<sigma>s', \<tau>s)"} to it,
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   676
  returns a term of type @{text "\<rho>s[\<sigma>s'] \<kappa> \<Rightarrow> \<rho>s[\<tau>s] \<kappa>"}. This type is
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   677
  equivalent to @{text "\<sigma>s \<kappa> \<Rightarrow> \<rho>s[\<tau>s] \<kappa>"}, which we just have to compose with
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   678
  @{text "\<rho>s[\<tau>s] \<kappa> \<Rightarrow> \<tau>s \<kappa>\<^isub>q"} according to the type of @{text "\<circ>"}.\qed
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
   679
  \end{proof}
2188
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   680
*}
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   681
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   682
section {* Respectfulness and Preservation \label{sec:resp} *}
2188
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   683
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   684
text {*
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   685
  The main point of the quotient package is to automatically ``lift'' theorems
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   686
  involving constants over the raw type to theorems involving constants over
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   687
  the quotient type. Before we can describe this lifting process, we need to impose 
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   688
  two restrictions in form of proof obligations that arise during the
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   689
  lifting. The reason is that even if definitions for all raw constants 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   690
  can be given, \emph{not} all theorems can be lifted to the quotient type. Most 
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   691
  notable is the bound variable function, that is the constant @{text bn}, defined 
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   692
  for raw lambda-terms as follows
2188
57972032e20e qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2186
diff changeset
   693
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   694
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
2252
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   695
  @{text "bn (x) \<equiv> \<emptyset>"}\hspace{4mm}
Christian Urban <urbanc@in.tum.de>
parents: 2248
diff changeset
   696
  @{text "bn (t\<^isub>1 t\<^isub>2) \<equiv> bn (t\<^isub>1) \<union> bn (t\<^isub>2)"}\hspace{4mm}
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   697
  @{text "bn (\<lambda>x. t) \<equiv> {x} \<union> bn (t)"}
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   698
  \end{isabelle}
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   699
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   700
  \noindent
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   701
  We can generate a definition for this constant using @{text ABS} and @{text REP}.
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   702
  But this constant does \emph{not} respect @{text "\<alpha>"}-equivalence and 
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   703
  consequently no theorem involving this constant can be lifted to @{text
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   704
  "\<alpha>"}-equated lambda terms. Homeier formulates the restrictions in terms of
2277
816204c76e90 Answer questions in comments
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2276
diff changeset
   705
  the properties of \emph{respectfulness} and \emph{preservation}. We have
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   706
  to slightly extend Homeier's definitions in order to deal with quotient
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   707
  compositions. 
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   708
2422
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   709
%%% FIXME: Reviewer 3 asks why are the definitions that follow enough to deal
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   710
%%% with quotient composition.
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   711
2247
084b2b7df98a some tuning and start work on section 4
Christian Urban <urbanc@in.tum.de>
parents: 2243
diff changeset
   712
  To formally define what respectfulness is, we have to first define 
2419
13d93ac68b07 Intuition behind REL
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2417
diff changeset
   713
  the notion of \emph{aggregate equivalence relations} using the function @{text "REL(\<sigma>, \<tau>)"}
13d93ac68b07 Intuition behind REL
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2417
diff changeset
   714
  The idea behind this function is to simultaneously descend into the raw types
13d93ac68b07 Intuition behind REL
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2417
diff changeset
   715
  @{text \<sigma>} and quotient types @{text \<tau>}, and generate the appropriate
13d93ac68b07 Intuition behind REL
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2417
diff changeset
   716
  quotient equivalence relations in places where the types differ and equalities
13d93ac68b07 Intuition behind REL
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2417
diff changeset
   717
  elsewhere.
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   718
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   719
  \begin{center}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   720
  \hfill
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   721
  \begin{tabular}{rcl}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   722
  \multicolumn{3}{@ {\hspace{-4mm}}l}{equal types:}\\ 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   723
  @{text "REL (\<sigma>, \<sigma>)"} & $\dn$ & @{text "= :: \<sigma> \<Rightarrow> \<sigma> \<Rightarrow> bool"}\smallskip\\
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   724
   \multicolumn{3}{@ {\hspace{-4mm}}l}{equal type constructors:}\\ 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   725
  @{text "REL (\<sigma>s \<kappa>, \<tau>s \<kappa>)"} & $\dn$ & @{text "rel_\<kappa> (REL (\<sigma>s, \<tau>s))"}\smallskip\\
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   726
  \multicolumn{3}{@ {\hspace{-4mm}}l}{unequal type constructors with @{text "\<alpha>s
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   727
  \<kappa>\<^isub>q"} being the quotient of the raw type @{text "\<rho>s \<kappa>"}:}\smallskip\\
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   728
  @{text "REL (\<sigma>s \<kappa>, \<tau>s \<kappa>\<^isub>q)"} & $\dn$ & @{text "rel_\<kappa>\<^isub>q (REL (\<sigma>s', \<tau>s))"}\\
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   729
  \end{tabular}\hfill\numbered{REL}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   730
  \end{center}
2238
8ddf1330f2ed completed proof and started section about respectfulness and preservation
Christian Urban <urbanc@in.tum.de>
parents: 2237
diff changeset
   731
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   732
  \noindent
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   733
  The @{text "\<sigma>s'"} in the last clause are calculated as in \eqref{ABSREP}:
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   734
  we know that type @{text "\<alpha>s \<kappa>\<^isub>q"} is the quotient of the raw type 
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   735
  @{text "\<rho>s \<kappa>"}. The @{text "\<sigma>s'"} are the substitutions for @{text "\<alpha>s"} obtained by matching 
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   736
  @{text "\<rho>s \<kappa>"} and @{text "\<sigma>s \<kappa>"}.
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   737
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   738
  Let us return to the lifting procedure of theorems. Assume we have a theorem
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   739
  that contains the raw constant @{text "c\<^isub>r :: \<sigma>"} and which we want to
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   740
  lift to a theorem where @{text "c\<^isub>r"} is replaced by the corresponding
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   741
  constant @{text "c\<^isub>q :: \<tau>"} defined over a quotient type. In this situation 
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   742
  we generate the following proof obligation
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   743
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   744
  @{text [display, indent=10] "REL (\<sigma>, \<tau>) c\<^isub>r c\<^isub>r"}
2258
72ce58b76c3b finished preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2257
diff changeset
   745
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   746
  \noindent
2277
816204c76e90 Answer questions in comments
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2276
diff changeset
   747
  Homeier calls these proof obligations \emph{respectfulness
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   748
  theorems}. However, unlike his quotient package, we might have several
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   749
  respectfulness theorems for one constant---he has at most one.
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   750
  The reason is that because of our quotient compositions, the types
2416
12283a96e851 The type does determine respectfulness, the constant without an instantiated type does not.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2414
diff changeset
   751
  @{text \<sigma>} and @{text \<tau>} are not completely determined by @{text "c\<^bsub>r\<^esub>"}.
12283a96e851 The type does determine respectfulness, the constant without an instantiated type does not.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2414
diff changeset
   752
  And for every instantiation of the types, a corresponding
12283a96e851 The type does determine respectfulness, the constant without an instantiated type does not.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2414
diff changeset
   753
  respectfulness theorem is necessary.
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   754
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   755
  Before lifting a theorem, we require the user to discharge
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   756
  respectfulness proof obligations. In case of @{text bn}
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   757
  this obligation is as follows
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   758
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   759
  @{text [display, indent=10] "(\<approx>\<^isub>\<alpha> \<doublearr> =) bn bn"}
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   760
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   761
  \noindent
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   762
  and the point is that the user cannot discharge it: because it is not true. To see this,
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   763
  we can just unfold the definition of @{text "\<doublearr>"} \eqref{relfun} 
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
   764
  using extensionality to obtain the false statement
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   765
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
   766
  @{text [display, indent=10] "\<forall>t\<^isub>1 t\<^isub>2. if t\<^isub>1 \<approx>\<^isub>\<alpha> t\<^isub>2 then bn(t\<^isub>1) = bn(t\<^isub>2)"}
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   767
 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   768
  \noindent
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   769
  In contrast, if we lift a theorem about @{text "append"} to a theorem describing 
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   770
  the union of finite sets, then we need to discharge the proof obligation
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   771
2275
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   772
  @{text [display, indent=10] "(\<approx>\<^bsub>list\<^esub> \<doublearr> \<approx>\<^bsub>list\<^esub> \<doublearr> \<approx>\<^bsub>list\<^esub>) append append"}
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   773
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   774
  \noindent
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   775
  To do so, we have to establish
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   776
  
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   777
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %%%
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   778
  if @{text "xs \<approx>\<^bsub>list\<^esub> ys"} and  @{text "us \<approx>\<^bsub>list\<^esub> vs"}
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   779
  then @{text "xs @ us \<approx>\<^bsub>list\<^esub> ys @ vs"}
2275
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   780
  \end{isabelle}
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   781
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   782
  \noindent
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   783
  which is straightforward given the definition shown in \eqref{listequiv}.
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   784
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   785
  The second restriction we have to impose arises from non-lifted polymorphic
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   786
  constants, which are instantiated to a type being quotient. For example,
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   787
  take the @{term "cons"}-constructor to add a pair of natural numbers to a
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   788
  list, whereby we assume the pair of natural numbers turns into an integer in
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   789
  the quotient construction. The point is that we still want to use @{text
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   790
  cons} for adding integers to lists---just with a different type. To be able
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   791
  to lift such theorems, we need a \emph{preservation property} for @{text
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   792
  cons}. Assuming we have a polymorphic raw constant @{text "c\<^isub>r :: \<sigma>"}
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   793
  and a corresponding quotient constant @{text "c\<^isub>q :: \<tau>"}, then a
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   794
  preservation property is as follows
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
   795
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   796
%%% FIXME: Reviewer 2 asks: You say what a preservation theorem is,
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   797
%%% but not which preservation theorems you assume. Do you generate a
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   798
%%% proof obligation for a preservation theorem for each raw constant
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   799
%%% and its corresponding lifted constant?
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   800
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   801
%%% Cezary: I think this would be a nice thing to do but we have not
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   802
%%% done it, the theorems need to be 'guessed' from the remaining obligations
2275
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   803
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   804
  @{text [display, indent=10] "Quotient R\<^bsub>\<alpha>s\<^esub> Abs\<^bsub>\<alpha>s\<^esub> Rep\<^bsub>\<alpha>s\<^esub> implies  ABS (\<sigma>, \<tau>) c\<^isub>r = c\<^isub>r"}
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   805
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   806
  \noindent
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   807
  where the @{text "\<alpha>s"} stand for the type variables in the type of @{text "c\<^isub>r"}.
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   808
  In case of @{text cons} (which has type @{text "\<alpha> \<Rightarrow> \<alpha> list \<Rightarrow> \<alpha> list"}) we have 
2275
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   809
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   810
  @{text [display, indent=10] "(Rep ---> map_list Rep ---> map_list Abs) cons = cons"}
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   811
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   812
  \noindent
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   813
  under the assumption @{text "Quotient R Abs Rep"}. Interestingly, if we have
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   814
  an instance of @{text cons} where the type variable @{text \<alpha>} is instantiated
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   815
  with @{text "nat \<times> nat"} and we also quotient this type to yield integers,
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   816
  then we need to show the corresponding preservation property.
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
   817
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   818
  %%%@ {thm [display, indent=10] insert_preserve2[no_vars]}
2189
029bd37d010a qpaper..
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2188
diff changeset
   819
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   820
  %Given two quotients, one of which quotients a container, and the
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   821
  %other quotients the type in the container, we can write the
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   822
  %composition of those quotients. To compose two quotient theorems
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   823
  %we compose the relations with relation composition as defined above
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   824
  %and the abstraction and relation functions are the ones of the sub
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   825
  %quotients composed with the usual function composition.
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   826
  %The @ {term "Rep"} and @ {term "Abs"} functions that we obtain agree
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   827
  %with the definition of aggregate Abs/Rep functions and the
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   828
  %relation is the same as the one given by aggregate relations.
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   829
  %This becomes especially interesting
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   830
  %when we compose the quotient with itself, as there is no simple
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   831
  %intermediate step.
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   832
  %
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   833
  %Lets take again the example of @ {term flat}. To be able to lift
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   834
  %theorems that talk about it we provide the composition quotient
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   835
  %theorem which allows quotienting inside the container:
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   836
  %
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   837
  %If @ {term R} is an equivalence relation and @ {term "Quotient R Abs Rep"}
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   838
  %then
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   839
  % 
2367
34af7f2ca490 some minor changes
Christian Urban <urbanc@in.tum.de>
parents: 2366
diff changeset
   840
  %@ {text [display, indent=10] "Quotient (list_rel R \<circ>\<circ>\<circ> \<approx>\<^bsub>list\<^esub>) (abs_fset \<circ> map_list Abs) (map_list Rep o rep_fset)"}
2275
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   841
  %%%
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   842
  %%%\noindent
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   843
  %%%this theorem will then instantiate the quotients needed in the
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   844
  %%%injection and cleaning proofs allowing the lifting procedure to
69b80ad616c5 finished section 4, but put some things I do not understand on comment
Christian Urban <urbanc@in.tum.de>
parents: 2274
diff changeset
   845
  %%%proceed in an unchanged way.
2192
87024a9a9d89 fixed compile error
Christian Urban <urbanc@in.tum.de>
parents: 2191
diff changeset
   846
*}
87024a9a9d89 fixed compile error
Christian Urban <urbanc@in.tum.de>
parents: 2191
diff changeset
   847
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
   848
section {* Lifting of Theorems\label{sec:lift} *}
1978
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
   849
2194
a52499e125ce qpaper / lifting introduction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2193
diff changeset
   850
text {*
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
   851
2422
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   852
%%% FIXME Reviewer 3 asks: Section 5 shows the technicalities of
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   853
%%% lifting theorems. But there is no clarification about the
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   854
%%% correctness. A reader would also be interested in seeing some
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   855
%%% discussions about the generality and limitation of the approach
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   856
%%% proposed there
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
   857
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   858
  The main benefit of a quotient package is to lift automatically theorems over raw
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   859
  types to theorems over quotient types. We will perform this lifting in
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   860
  three phases, called \emph{regularization},
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   861
  \emph{injection} and \emph{cleaning} according to procedures in Homeier's ML-code.
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
   862
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
   863
  The purpose of regularization is to change the quantifiers and abstractions
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
   864
  in a ``raw'' theorem to quantifiers over variables that respect their respective relations
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   865
  (Definition \ref{def:respects} states what respects means). The purpose of injection is to add @{term Rep}
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
   866
  and @{term Abs} of appropriate types in front of constants and variables
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   867
  of the raw type so that they can be replaced by the corresponding constants from the
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   868
  quotient type. The purpose of cleaning is to bring the theorem derived in the
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   869
  first two phases into the form the user has specified. Abstractly, our
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   870
  package establishes the following three proof steps:
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   871
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   872
%%% FIXME: Reviewer 1 complains that the reader needs to guess the
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   873
%%% meaning of reg_thm and inj_thm, as well as the arguments of REG
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   874
%%% which are given above. I wouldn't change it.
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   875
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   876
  \begin{center}
2369
3aeb58524131 minor things
Christian Urban <urbanc@in.tum.de>
parents: 2367
diff changeset
   877
  \begin{tabular}{l@ {\hspace{4mm}}l}
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   878
  1.) Regularization & @{text "raw_thm \<longrightarrow> reg_thm"}\\
2369
3aeb58524131 minor things
Christian Urban <urbanc@in.tum.de>
parents: 2367
diff changeset
   879
  2.) Injection & @{text "reg_thm \<longleftrightarrow> inj_thm"}\\
3aeb58524131 minor things
Christian Urban <urbanc@in.tum.de>
parents: 2367
diff changeset
   880
  3.) Cleaning & @{text "inj_thm \<longleftrightarrow> quot_thm"}\\
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   881
  \end{tabular}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   882
  \end{center}
2193
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2192
diff changeset
   883
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   884
  \noindent
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   885
  which means, stringed together, the raw theorem implies the quotient theorem.
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   886
  In contrast to other quotient packages, our package requires that the user specifies 
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   887
  both, the @{text "raw_thm"} (as theorem) and the \emph{term} of the @{text "quot_thm"}.\footnote{Though we
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   888
  also provide a fully automated mode, where the @{text "quot_thm"} is guessed
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   889
  from the form of @{text "raw_thm"}.} As a result, the user has fine control
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   890
  over which parts of a raw theorem should be lifted. 
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   891
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   892
  The second and third proof step performed in package will always succeed if the appropriate
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   893
  respectfulness and preservation theorems are given. In contrast, the first
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   894
  proof step can fail: a theorem given by the user does not always
2319
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   895
  imply a regularized version and a stronger one needs to be proved. An example
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   896
  for this kind of failure is the simple statement for integers @{text "0 \<noteq> 1"}.
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   897
  One might hope that it can be proved by lifting @{text "(0, 0) \<noteq> (1, 0)"},  
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   898
  but this raw theorem only shows that two particular elements in the
2319
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   899
  equivalence classes are not equal. In order to obtain @{text "0 \<noteq> 1"}, a  
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   900
  more general statement stipulating that the equivalence classes are not 
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   901
  equal is necessary.  This kind of failure is beyond the scope where the 
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   902
  quotient package can help: the user has to provide a raw theorem that
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   903
  can be regularized automatically, or has to provide an explicit proof
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   904
  for the first proof step.
2193
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2192
diff changeset
   905
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
   906
  In the following we will first define the statement of the
2280
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
   907
  regularized theorem based on @{text "raw_thm"} and
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   908
  @{text "quot_thm"}. Then we define the statement of the injected theorem, based
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
   909
  on @{text "reg_thm"} and @{text "quot_thm"}. We then show the three proof steps,
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   910
  which can all be performed independently from each other.
2197
3a6afcb187ec qpaper / regularize
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2196
diff changeset
   911
2373
4cc4390d5d25 corrected lambda-preservation theorem
Christian Urban <urbanc@in.tum.de>
parents: 2371
diff changeset
   912
  We first define the function @{text REG}, which takes the terms of the 
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   913
  @{text "raw_thm"} and @{text "quot_thm"} as input and returns
2373
4cc4390d5d25 corrected lambda-preservation theorem
Christian Urban <urbanc@in.tum.de>
parents: 2371
diff changeset
   914
  @{text "reg_thm"}. The idea
2207
ea7c3f21d6df Qpaper/more.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2206
diff changeset
   915
  behind this function is that it replaces quantifiers and
ea7c3f21d6df Qpaper/more.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2206
diff changeset
   916
  abstractions involving raw types by bounded ones, and equalities
2373
4cc4390d5d25 corrected lambda-preservation theorem
Christian Urban <urbanc@in.tum.de>
parents: 2371
diff changeset
   917
  involving raw types by appropriate aggregate
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   918
  equivalence relations. It is defined by simultaneously recursing on 
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   919
  the structure of  @{text "raw_thm"} and @{text "quot_thm"} as follows:
1994
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1978
diff changeset
   920
2244
e907165b953b qpaper / REG
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2243
diff changeset
   921
  \begin{center}
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   922
  \begin{tabular}{rcl}
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   923
  \multicolumn{3}{@ {}l}{abstractions:}\smallskip\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   924
  @{text "REG (\<lambda>x\<^sup>\<sigma>. t, \<lambda>x\<^sup>\<tau>. s)"} & $\dn$ & 
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   925
  $\begin{cases}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   926
  @{text "\<lambda>x\<^sup>\<sigma>. REG (t, s)"} \quad\mbox{provided @{text "\<sigma> = \<tau>"}}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   927
  @{text "\<lambda>x\<^sup>\<sigma> \<in> Respects (REL (\<sigma>, \<tau>)). REG (t, s)"}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   928
  \end{cases}$\smallskip\\
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   929
  \\
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   930
  \multicolumn{3}{@ {}l}{universal quantifiers:}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   931
  @{text "REG (\<forall>x\<^sup>\<sigma>. t, \<forall>x\<^sup>\<tau>. s)"} & $\dn$ & 
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   932
  $\begin{cases}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   933
  @{text "\<forall>x\<^sup>\<sigma>. REG (t, s)"} \quad\mbox{provided @{text "\<sigma> = \<tau>"}}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   934
  @{text "\<forall>x\<^sup>\<sigma> \<in> Respects (REL (\<sigma>, \<tau>)). REG (t, s)"}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   935
  \end{cases}$\smallskip\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   936
  \multicolumn{3}{@ {}l}{equality:}\smallskip\\
2280
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
   937
  %% REL of two equal types is the equality so we do not need a separate case
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
   938
  @{text "REG (=\<^bsup>\<sigma>\<Rightarrow>\<sigma>\<Rightarrow>bool\<^esup>, =\<^bsup>\<tau>\<Rightarrow>\<tau>\<Rightarrow>bool\<^esup>)"} & $\dn$ & @{text "REL (\<sigma>, \<tau>)"}\\\smallskip\\
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   939
  \multicolumn{3}{@ {}l}{applications, variables and constants:}\\
2244
e907165b953b qpaper / REG
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2243
diff changeset
   940
  @{text "REG (t\<^isub>1 t\<^isub>2, s\<^isub>1 s\<^isub>2)"} & $\dn$ & @{text "REG (t\<^isub>1, s\<^isub>1) REG (t\<^isub>2, s\<^isub>2)"}\\
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   941
  @{text "REG (x\<^isub>1, x\<^isub>2)"} & $\dn$ & @{text "x\<^isub>1"}\\
2371
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   942
  @{text "REG (c\<^isub>1, c\<^isub>2)"} & $\dn$ & @{text "c\<^isub>1"}\\
86c73a06ba4b minor things on the paper
Christian Urban <urbanc@in.tum.de>
parents: 2369
diff changeset
   943
  \end{tabular}
2244
e907165b953b qpaper / REG
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2243
diff changeset
   944
  \end{center}
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   945
  %
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   946
  \noindent
2230
fec38b7ceeb3 some spelling
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2229
diff changeset
   947
  In the above definition we omitted the cases for existential quantifiers
2207
ea7c3f21d6df Qpaper/more.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2206
diff changeset
   948
  and unique existential quantifiers, as they are very similar to the cases
2376
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
   949
  for the universal quantifier. 
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   950
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   951
  Next we define the function @{text INJ} which takes as argument
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   952
  @{text "reg_thm"} and @{text "quot_thm"} (both as
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   953
  terms) and returns @{text "inj_thm"}:
2198
8fe1a706ade7 qpaper / injection statement
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2197
diff changeset
   954
2245
280b92df6a8b qpaper / INJ
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2244
diff changeset
   955
  \begin{center}
280b92df6a8b qpaper / INJ
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2244
diff changeset
   956
  \begin{tabular}{rcl}
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   957
  \multicolumn{3}{@ {\hspace{-4mm}}l}{abstractions:}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   958
  @{text "INJ (\<lambda>x. t :: \<sigma>, \<lambda>x. s :: \<tau>) "} & $\dn$ & 
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   959
  $\begin{cases}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   960
  @{text "\<lambda>x. INJ (t, s)"} \quad\mbox{provided @{text "\<sigma> = \<tau>"}}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   961
  @{text "REP (\<sigma>, \<tau>) (ABS (\<sigma>, \<tau>) (\<lambda>x. INJ (t, s)))"}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   962
  \end{cases}$\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   963
  @{text "INJ (\<lambda>x \<in> R. t :: \<sigma>, \<lambda>x. s :: \<tau>) "} & $\dn$ 
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   964
  & @{text "REP (\<sigma>, \<tau>) (ABS (\<sigma>, \<tau>) (\<lambda>x \<in> R. INJ (t, s)))"}\smallskip\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   965
  \multicolumn{3}{@ {\hspace{-4mm}}l}{universal quantifiers:}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   966
  @{text "INJ (\<forall> t, \<forall> s) "} & $\dn$ & @{text "\<forall> INJ (t, s)"}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   967
  @{text "INJ (\<forall> t \<in> R, \<forall> s) "} & $\dn$ & @{text "\<forall> INJ (t, s) \<in> R"}\smallskip\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   968
  \multicolumn{3}{@ {\hspace{-4mm}}l}{applications, variables and constants:}\smallskip\\
2245
280b92df6a8b qpaper / INJ
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2244
diff changeset
   969
  @{text "INJ (t\<^isub>1 t\<^isub>2, s\<^isub>1 s\<^isub>2) "} & $\dn$ & @{text " INJ (t\<^isub>1, s\<^isub>1) INJ (t\<^isub>2, s\<^isub>2)"}\\
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   970
  @{text "INJ (x\<^isub>1\<^sup>\<sigma>, x\<^isub>2\<^sup>\<tau>) "} & $\dn$ & 
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   971
  $\begin{cases}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   972
  @{text "x\<^isub>1"} \quad\mbox{provided @{text "\<sigma> = \<tau>"}}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   973
  @{text "REP (\<sigma>, \<tau>) (ABS (\<sigma>, \<tau>) x\<^isub>1)"}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   974
  \end{cases}$\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   975
  @{text "INJ (c\<^isub>1\<^sup>\<sigma>, c\<^isub>2\<^sup>\<tau>) "} & $\dn$ & 
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   976
  $\begin{cases}
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   977
  @{text "c\<^isub>1"} \quad\mbox{provided @{text "\<sigma> = \<tau>"}}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   978
  @{text "REP (\<sigma>, \<tau>) (ABS (\<sigma>, \<tau>) c\<^isub>1)"}\\
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
   979
  \end{cases}$\\
2245
280b92df6a8b qpaper / INJ
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2244
diff changeset
   980
  \end{tabular}
280b92df6a8b qpaper / INJ
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2244
diff changeset
   981
  \end{center}
2198
8fe1a706ade7 qpaper / injection statement
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2197
diff changeset
   982
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   983
  \noindent 
2319
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   984
  In this definition we again omitted the cases for existential and unique existential
2376
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
   985
  quantifiers. 
2208
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2207
diff changeset
   986
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   987
%%% FIXME: Reviewer2 citing following sentence: You mention earlier
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   988
%%% that this implication may fail to be true. Does that meant that
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   989
%%% the `first proof step' is a heuristic that proves the implication
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   990
%%% raw_thm \implies reg_thm in some instances, but fails in others?
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   991
%%% You should clarify under which circumstances the implication is
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   992
%%% being proved here.
2423
f5cbf74d4ec5 Clarifications to FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2422
diff changeset
   993
%%% Cezary: It would be nice to cite Homeiers discussions in the
f5cbf74d4ec5 Clarifications to FIXMEs.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2422
diff changeset
   994
%%% Quotient Package manual from HOL (the longer paper), do you agree?
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
   995
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   996
  In the first proof step, establishing @{text "raw_thm \<longrightarrow> reg_thm"}, we always 
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
   997
  start with an implication. Isabelle provides \emph{mono} rules that can split up 
2319
7c8783d2dcd0 further post-submission tuning
Christian Urban <urbanc@in.tum.de>
parents: 2287
diff changeset
   998
  the implications into simpler implicational subgoals. This succeeds for every
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
   999
  monotone connective, except in places where the function @{text REG} replaced,
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1000
  for instance, a quantifier by a bounded quantifier. In this case we have 
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1001
  rules of the form
2208
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2207
diff changeset
  1002
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1003
  @{text [display, indent=10] "(\<forall>x. R x \<longrightarrow> (P x \<longrightarrow> Q x)) \<longrightarrow> (\<forall>x. P x \<longrightarrow> \<forall>x \<in> R. Q x)"}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1004
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1005
  \noindent
2280
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
  1006
  They decompose a bounded quantifier on the right-hand side. We can decompose a
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
  1007
  bounded quantifier anywhere if R is an equivalence relation or
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
  1008
  if it is a relation over function types with the range being an equivalence
229660b9f2fc Fix regularization. Two "FIXME" left in introduction. Minor spellings.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2279
diff changeset
  1009
  relation. If @{text R} is an equivalence relation we can prove that
2261
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2260
diff changeset
  1010
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1011
  @{text [display, indent=10] "\<forall>x \<in> Respects R. P x = \<forall>x. P x"}    
2199
6ce64fb5cbd9 qpaper / lemmas used in proofs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2198
diff changeset
  1012
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1013
  \noindent
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1014
  If @{term R\<^isub>2} is an equivalence relation, we can prove that for any predicate @{term P}
2231
01d08af79f01 qpaper.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2230
diff changeset
  1015
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1016
%%% FIXME Reviewer 1 claims the theorem is obviously false so maybe we
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1017
%%% should include a proof sketch?
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1018
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
  1019
  @{thm [display, indent=10] (concl) ball_reg_eqv_range[of R\<^isub>1 R\<^isub>2, no_vars]}
2209
5952b0f28261 Qpaper/regularization proof.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2208
diff changeset
  1020
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1021
  \noindent
2242
3f480e33d8df qpaper/various
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2241
diff changeset
  1022
  The last theorem is new in comparison with Homeier's package. There the
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1023
  injection procedure would be used to prove such goals and 
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1024
  the assumption about the equivalence relation would be used. We use the above theorem directly,
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1025
  because this allows us to completely separate the first and the second
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1026
  proof step into two independent ``units''.
2206
2d6cada7d5e0 Qpaper/Minor
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2205
diff changeset
  1027
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1028
  The second proof step, establishing @{text "reg_thm \<longleftrightarrow> inj_thm"},  starts with an equality
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1029
  between the terms of the regularized theorem and the injected theorem.
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1030
  The proof again follows the structure of the
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1031
  two underlying terms and is defined for a goal being a relation between these two terms.
2199
6ce64fb5cbd9 qpaper / lemmas used in proofs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2198
diff changeset
  1032
2211
9d0673c319d1 qpaper / injection proof.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2210
diff changeset
  1033
  \begin{itemize}
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1034
  \item For two constants an appropriate respectfulness theorem is applied.
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1035
  \item For two variables, we use the assumptions proved in the regularization step.
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1036
  \item For two abstractions, we @{text "\<eta>"}-expand and @{text "\<beta>"}-reduce them.
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1037
  \item For two applications, we check that the right-hand side is an application of
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1038
    @{term Rep} to an @{term Abs} and @{term "Quotient R Rep Abs"} holds. If yes then we
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1039
    can apply the theorem:
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1040
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1041
    @{term [display, indent=10] "R x y \<longrightarrow> R x (Rep (Abs y))"}
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1042
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1043
    Otherwise we introduce an appropriate relation between the subterms
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1044
    and continue with two subgoals using the lemma:
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1045
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
  1046
    @{text [display, indent=10] "(R\<^isub>1 \<doublearr> R\<^isub>2) f g \<longrightarrow> R\<^isub>1 x y \<longrightarrow> R\<^isub>2 (f x) (g y)"}
2211
9d0673c319d1 qpaper / injection proof.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2210
diff changeset
  1047
  \end{itemize}
2199
6ce64fb5cbd9 qpaper / lemmas used in proofs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2198
diff changeset
  1048
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1049
  We defined the theorem @{text "inj_thm"} in such a way that
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1050
  establishing the equivalence @{text "inj_thm \<longleftrightarrow> quot_thm"} can be
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1051
  achieved by rewriting @{text "inj_thm"} with the preservation theorems and quotient
2437
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2423
diff changeset
  1052
  definitions. First the definitions of all lifted constants
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1053
  are used to fold the @{term Rep} with the raw constants. Next for
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1054
  all abstractions and quantifiers the lambda and
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1055
  quantifier preservation theorems are used to replace the
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1056
  variables that include raw types with respects by quantifiers
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1057
  over variables that include quotient types. We show here only
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1058
  the lambda preservation theorem. Given
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1059
  @{term "Quotient R\<^isub>1 Abs\<^isub>1 Rep\<^isub>1"} and @{term "Quotient R\<^isub>2 Abs\<^isub>2 Rep\<^isub>2"}, we have:
2211
9d0673c319d1 qpaper / injection proof.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2210
diff changeset
  1060
2373
4cc4390d5d25 corrected lambda-preservation theorem
Christian Urban <urbanc@in.tum.de>
parents: 2371
diff changeset
  1061
  @{thm [display, indent=10] (concl) lambda_prs[of _ "Abs\<^isub>1" "Rep\<^isub>1" _ "Abs\<^isub>2" "Rep\<^isub>2", no_vars]}
2199
6ce64fb5cbd9 qpaper / lemmas used in proofs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2198
diff changeset
  1062
2243
5e98b3f231a0 qpaper / minor
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2242
diff changeset
  1063
  \noindent
2412
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1064
  Next, relations over lifted types can be rewritten to equalities
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1065
  over lifted type. Rewriting is performed with the following theorem,
63f0e7f914dd fixes for referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2377
diff changeset
  1066
  which has been shown by Homeier~\cite{Homeier05}:
2211
9d0673c319d1 qpaper / injection proof.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2210
diff changeset
  1067
2376
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
  1068
  @{thm [display, indent=10] (concl) Quotient_rel_rep[no_vars]}
2199
6ce64fb5cbd9 qpaper / lemmas used in proofs
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2198
diff changeset
  1069
2271
c0c5bc4ee8cb qpaper/Rewrite section5
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2268
diff changeset
  1070
  \noindent
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1071
  Finally, we rewrite with the preservation theorems. This will result
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1072
  in two equal terms that can be solved by reflexivity.
2246
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2245
diff changeset
  1073
  *}
1994
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1978
diff changeset
  1074
2376
Christian Urban <urbanc@in.tum.de>
parents: 2373
diff changeset
  1075
2274
99cf23602d36 4 almost finished
Christian Urban <urbanc@in.tum.de>
parents: 2273
diff changeset
  1076
section {* Examples \label{sec:examples} *}
1994
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1978
diff changeset
  1077
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1078
text {*
2206
2d6cada7d5e0 Qpaper/Minor
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2205
diff changeset
  1079
2421
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1080
%%% FIXME Reviewer 1 would like an example of regularized and injected
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1081
%%% statements. He asks for the examples twice, but I would still ignore
4ef4661be815 few remaining remarks as fixme's.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2419
diff changeset
  1082
%%% it due to lack of space...
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1083
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1084
  In this section we will show a sequence of declarations for defining the 
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1085
  type of integers by quotienting pairs of natural numbers, and
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1086
  lifting one theorem. 
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1087
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1088
  A user of our quotient package first needs to define a relation on
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1089
  the raw type with which the quotienting will be performed. We give
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1090
  the same integer relation as the one presented in \eqref{natpairequiv}:
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1091
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1092
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1093
  \begin{tabular}{@ {}l}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1094
  \isacommand{fun}~~@{text "int_rel :: (nat \<times> nat) \<Rightarrow> (nat \<times> nat) \<Rightarrow> (nat \<times> nat)"}\\
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1095
  \isacommand{where}~~@{text "int_rel (m, n) (p, q) = (m + q = n + p)"}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1096
  \end{tabular}
2239
ff997de1bd73 qpaper/examples
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2238
diff changeset
  1097
  \end{isabelle}
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1098
2239
ff997de1bd73 qpaper/examples
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2238
diff changeset
  1099
  \noindent
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1100
  Next the quotient type must be defined. This generates a proof obligation that the
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1101
  relation is an equivalence relation, which is solved automatically using the
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1102
  definition of equivalence and extensionality:
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1103
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1104
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1105
  \begin{tabular}{@ {}l}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1106
  \isacommand{quotient\_type}~~@{text "int"}~~\isacommand{=}~~@{text "(nat \<times> nat)"}~~\isacommand{/}~~@{text "int_rel"}\\
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1107
  \hspace{5mm}@{text "by (auto simp add: equivp_def expand_fun_eq)"}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1108
  \end{tabular}
2239
ff997de1bd73 qpaper/examples
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2238
diff changeset
  1109
  \end{isabelle}
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1110
2239
ff997de1bd73 qpaper/examples
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2238
diff changeset
  1111
  \noindent
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1112
  The user can then specify the constants on the quotient type:
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1113
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1114
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1115
  \begin{tabular}{@ {}l}
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1116
  \isacommand{quotient\_definition}~~@{text "0 :: int"}~~\isacommand{is}~~@{text "(0 :: nat, 0 :: nat)"}\\[3mm]
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1117
  \isacommand{fun}~~@{text "add_pair"}~~\isacommand{where}~~%
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1118
  @{text "add_pair (m, n) (p, q) \<equiv> (m + p :: nat, n + q :: nat)"}\\
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
  1119
  \isacommand{quotient\_definition}~~@{text "+ :: int \<Rightarrow> int \<Rightarrow> int"}~~%
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1120
  \isacommand{is}~~@{text "add_pair"}\\
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1121
  \end{tabular}
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1122
  \end{isabelle}
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1123
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1124
  \noindent
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1125
  The following theorem about addition on the raw level can be proved.
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1126
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1127
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1128
  \isacommand{lemma}~~@{text "add_pair_zero: int_rel (add_pair (0, 0) x) x"}
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1129
  \end{isabelle}
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1130
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1131
  \noindent
2413
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
  1132
  If the user lifts this theorem, the quotient package performs all the lifting
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
  1133
  automatically leaving the respectfulness proof for the constant @{text "add_pair"}
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
  1134
  as the only remaining proof obligation. This property needs to be proved by the user:
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1135
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1136
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1137
  \begin{tabular}{@ {}l}
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1138
  \isacommand{lemma}~~@{text "[quot_respect]:"}\\ 
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1139
  @{text "(int_rel \<doublearr> int_rel \<doublearr> int_rel) add_pair add_pair"}
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1140
  \end{tabular}
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1141
  \end{isabelle}
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1142
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1143
  \noindent
2413
1341a2d7570f further comments from the referees
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2412
diff changeset
  1144
  It can be discharged automatically by Isabelle when hinting to unfold the definition
2273
d62c082cb56b cleaned up definitions
Christian Urban <urbanc@in.tum.de>
parents: 2272
diff changeset
  1145
  of @{text "\<doublearr>"}.
2332
9a560e489c64 polished paper again (and took out some claims about Homeier's package)
Christian Urban <urbanc@in.tum.de>
parents: 2319
diff changeset
  1146
  After this, the user can prove the lifted lemma as follows:
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1147
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1148
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1149
  \isacommand{lemma}~~@{text "0 + (x :: int) = x"}~~\isacommand{by}~~@{text "lifting add_pair_zero"}
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1150
  \end{isabelle}
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1151
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1152
  \noindent
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1153
  or by using the completely automated mode stating just:
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1154
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1155
  \begin{isabelle}\ \ \ \ \ \ \ \ \ \ %
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1156
  \isacommand{thm}~~@{text "add_pair_zero[quot_lifted]"}
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1157
  \end{isabelle}
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1158
2240
6c4b54482396 qpaper/more on example
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2239
diff changeset
  1159
  \noindent
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1160
  Both methods give the same result, namely
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1161
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1162
  @{text [display, indent=10] "0 + x = x"}
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1163
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1164
  \noindent
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1165
  where @{text x} is of type integer.
2279
2cbcdaba795a polished everything and submitted
Christian Urban <urbanc@in.tum.de>
parents: 2278
diff changeset
  1166
  Although seemingly simple, arriving at this result without the help of a quotient
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1167
  package requires a substantial reasoning effort (see \cite{Paulson06}).
2210
6aaec9dd0c62 qpaper / example interaction
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2209
diff changeset
  1168
*}
2206
2d6cada7d5e0 Qpaper/Minor
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2205
diff changeset
  1169
2256
f5f21feaa168 some slight tuning of the preliminary section
Christian Urban <urbanc@in.tum.de>
parents: 2255
diff changeset
  1170
section {* Conclusion and Related Work\label{sec:conc}*}
1978
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
  1171
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
  1172
text {*
2243
5e98b3f231a0 qpaper / minor
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2242
diff changeset
  1173
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1174
  The code of the quotient package and the examples described here are already
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1175
  included in the standard distribution of Isabelle.\footnote{Available from
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1176
  \href{http://isabelle.in.tum.de/}{http://isabelle.in.tum.de/}.} The package is
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1177
  heavily used in the new version of Nominal Isabelle, which provides a
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1178
  convenient reasoning infrastructure for programming language calculi
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1179
  involving general binders.  To achieve this, it builds types representing
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1180
  @{text \<alpha>}-equivalent terms.  Earlier versions of Nominal Isabelle have been
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1181
  used successfully in formalisations of an equivalence checking algorithm for
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1182
  LF \cite{UrbanCheneyBerghofer08}, Typed
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1183
  Scheme~\cite{TobinHochstadtFelleisen08}, several calculi for concurrency
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1184
  \cite{BengtsonParow09} and a strong normalisation result for cut-elimination
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1185
  in classical logic \cite{UrbanZhu08}.
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1186
2237
d1ab5d2d6926 more on the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2236
diff changeset
  1187
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1188
  There is a wide range of existing literature for dealing with quotients
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1189
  in theorem provers.  Slotosch~\cite{Slotosch97} implemented a mechanism that
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1190
  automatically defines quotient types for Isabelle/HOL. But he did not
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1191
  include theorem lifting.  Harrison's quotient package~\cite{harrison-thesis}
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1192
  is the first one that is able to automatically lift theorems, however only
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1193
  first-order theorems (that is theorems where abstractions, quantifiers and
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1194
  variables do not involve functions that include the quotient type). There is
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1195
  also some work on quotient types in non-HOL based systems and logical
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1196
  frameworks, including theory interpretations in
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1197
  PVS~\cite{PVS:Interpretations}, new types in MetaPRL~\cite{Nogin02}, and
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1198
  setoids in Coq \cite{ChicliPS02}.  Paulson showed a construction of
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1199
  quotients that does not require the Hilbert Choice operator, but also only
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1200
  first-order theorems can be lifted~\cite{Paulson06}.  The most related work
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1201
  to our package is the package for HOL4 by Homeier~\cite{Homeier05}.  He
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1202
  introduced most of the abstract notions about quotients and also deals with
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1203
  lifting of higher-order theorems. However, he cannot deal with quotient
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1204
  compositions (needed for lifting theorems about @{text flat}). Also, a
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1205
  number of his definitions, like @{text ABS}, @{text REP} and @{text INJ} etc
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1206
  only exist in \cite{Homeier05} as ML-code, not included in the paper.
2334
0d10196364aa added comment about partial equivalence relations
Christian Urban <urbanc@in.tum.de>
parents: 2333
diff changeset
  1207
  Like Homeier's, our quotient package can deal with partial equivalence
0d10196364aa added comment about partial equivalence relations
Christian Urban <urbanc@in.tum.de>
parents: 2333
diff changeset
  1208
  relations, but for lack of space we do not describe the mechanisms
0d10196364aa added comment about partial equivalence relations
Christian Urban <urbanc@in.tum.de>
parents: 2333
diff changeset
  1209
  needed for this kind of quotient constructions.
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
  1210
2422
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
  1211
%%% FIXME Reviewer 3 would like to know more about the lifting in Coq and PVS,
cd694a9988bc Finished adding remarks from the reviewers.
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2421
diff changeset
  1212
%%% and some comparison. I don't think we have the space for any additions...
2333
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1213
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1214
  One feature of our quotient package is that when lifting theorems, the user
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1215
  can precisely specify what the lifted theorem should look like. This feature
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1216
  is necessary, for example, when lifting an induction principle for two
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1217
  lists.  Assuming this principle has as the conclusion a predicate of the
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1218
  form @{text "P xs ys"}, then we can precisely specify whether we want to
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1219
  quotient @{text "xs"} or @{text "ys"}, or both. We found this feature very
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1220
  useful in the new version of Nominal Isabelle, where such a choice is
a0fecce8b244 even further polishing of the qpaper
Christian Urban <urbanc@in.tum.de>
parents: 2332
diff changeset
  1221
  required to generate a reasoning infrastructure for alpha-equated terms.
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1222
%%
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1223
%% give an example for this
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1224
%%
2278
337569f85398 conclusion done
Christian Urban <urbanc@in.tum.de>
parents: 2277
diff changeset
  1225
  \medskip
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
  1226
2263
d2ca79475103 qpaper/ackno
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2261
diff changeset
  1227
  \noindent
2287
adb5b1349280 some post-submission polishing
Christian Urban <urbanc@in.tum.de>
parents: 2286
diff changeset
  1228
  {\bf Acknowledgements:} We would like to thank Peter Homeier for the many
2277
816204c76e90 Answer questions in comments
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2276
diff changeset
  1229
  discussions about his HOL4 quotient package and explaining to us
2283
5c603b0945ac small addition to the acknowledgement
Christian Urban <urbanc@in.tum.de>
parents: 2282
diff changeset
  1230
  some of its finer points in the implementation. Without his patient
5c603b0945ac small addition to the acknowledgement
Christian Urban <urbanc@in.tum.de>
parents: 2282
diff changeset
  1231
  help, this work would have been impossible.
2263
d2ca79475103 qpaper/ackno
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 2261
diff changeset
  1232
2224
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
  1233
*}
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
  1234
f5b6f9d8a882 completed the intro (except minor things)
Christian Urban <urbanc@in.tum.de>
parents: 2223
diff changeset
  1235
2227
42d576c54704 polishing of ABS/REP
Christian Urban <urbanc@in.tum.de>
parents: 2226
diff changeset
  1236
1975
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
  1237
(*<*)
b1281a0051ae added stub for quotient paper; call with isabelle make qpaper
Christian Urban <urbanc@in.tum.de>
parents:
diff changeset
  1238
end
1978
8feedc0d4ea8 quotient paper
Cezary Kaliszyk <kaliszyk@in.tum.de>
parents: 1975
diff changeset
  1239
(*>*)