author | blanchet |
Thu, 06 Feb 2014 23:42:36 +0100 | |
changeset 55350 | cf34abe28209 |
parent 55290 | 3951ced4156c |
child 55351 | 8d7031a35991 |
permissions | -rw-r--r-- |
52792 | 1 |
(* Title: Doc/Datatypes/Datatypes.thy |
2 |
Author: Jasmin Blanchette, TU Muenchen |
|
53617 | 3 |
Author: Lorenz Panny, TU Muenchen |
4 |
Author: Andrei Popescu, TU Muenchen |
|
5 |
Author: Dmitriy Traytel, TU Muenchen |
|
52792 | 6 |
|
7 |
Tutorial for (co)datatype definitions with the new package. |
|
8 |
*) |
|
9 |
||
10 |
theory Datatypes |
|
55073 | 11 |
imports |
12 |
Setup |
|
55075 | 13 |
"~~/src/HOL/Library/BNF_Decl" |
55350 | 14 |
"~~/src/HOL/Library/Cardinal_Notations" |
55129
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
15 |
"~~/src/HOL/Library/FSet" |
55073 | 16 |
"~~/src/HOL/Library/Simps_Case_Conv" |
52792 | 17 |
begin |
18 |
||
52828 | 19 |
section {* Introduction |
20 |
\label{sec:introduction} *} |
|
52792 | 21 |
|
22 |
text {* |
|
53028 | 23 |
The 2013 edition of Isabelle introduced a new definitional package for freely |
24 |
generated datatypes and codatatypes. The datatype support is similar to that |
|
25 |
provided by the earlier package due to Berghofer and Wenzel |
|
26 |
\cite{Berghofer-Wenzel:1999:TPHOL}, documented in the Isar reference manual |
|
53535 | 27 |
\cite{isabelle-isar-ref}; indeed, replacing the keyword \keyw{datatype} by |
53028 | 28 |
@{command datatype_new} is usually all that is needed to port existing theories |
29 |
to use the new package. |
|
52840 | 30 |
|
52841 | 31 |
Perhaps the main advantage of the new package is that it supports recursion |
53619 | 32 |
through a large class of non-datatypes, such as finite sets: |
52792 | 33 |
*} |
34 |
||
53644 | 35 |
datatype_new 'a tree\<^sub>f\<^sub>s = Node\<^sub>f\<^sub>s (lbl\<^sub>f\<^sub>s: 'a) (sub\<^sub>f\<^sub>s: "'a tree\<^sub>f\<^sub>s fset") |
52792 | 36 |
|
37 |
text {* |
|
52794 | 38 |
\noindent |
53025 | 39 |
Another strong point is the support for local definitions: |
52792 | 40 |
*} |
41 |
||
52805 | 42 |
context linorder |
43 |
begin |
|
52841 | 44 |
datatype_new flag = Less | Eq | Greater |
52805 | 45 |
end |
52792 | 46 |
|
47 |
text {* |
|
52794 | 48 |
\noindent |
54187 | 49 |
Furthermore, the package provides a lot of convenience, including automatically |
50 |
generated discriminators, selectors, and relators as well as a wealth of |
|
51 |
properties about them. |
|
52 |
||
53 |
In addition to inductive datatypes, the new package supports coinductive |
|
54 |
datatypes, or \emph{codatatypes}, which allow infinite values. For example, the |
|
55 |
following command introduces the type of lazy lists, which comprises both finite |
|
56 |
and infinite values: |
|
52794 | 57 |
*} |
58 |
||
53623 | 59 |
(*<*) |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
60 |
locale early |
54072 | 61 |
locale late |
53623 | 62 |
(*>*) |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
63 |
codatatype (*<*)(in early) (*>*)'a llist = LNil | LCons 'a "'a llist" |
52794 | 64 |
|
65 |
text {* |
|
66 |
\noindent |
|
52840 | 67 |
Mixed inductive--coinductive recursion is possible via nesting. Compare the |
53028 | 68 |
following four Rose tree examples: |
52794 | 69 |
*} |
70 |
||
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
71 |
datatype_new (*<*)(in early) (*>*)'a tree\<^sub>f\<^sub>f = Node\<^sub>f\<^sub>f 'a "'a tree\<^sub>f\<^sub>f list" |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
72 |
datatype_new (*<*)(in early) (*>*)'a tree\<^sub>f\<^sub>i = Node\<^sub>f\<^sub>i 'a "'a tree\<^sub>f\<^sub>i llist" |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
73 |
codatatype (*<*)(in early) (*>*)'a tree\<^sub>i\<^sub>f = Node\<^sub>i\<^sub>f 'a "'a tree\<^sub>i\<^sub>f list" |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
74 |
codatatype (*<*)(in early) (*>*)'a tree\<^sub>i\<^sub>i = Node\<^sub>i\<^sub>i 'a "'a tree\<^sub>i\<^sub>i llist" |
52792 | 75 |
|
52794 | 76 |
text {* |
54187 | 77 |
The first two tree types allow only paths of finite length, whereas the last two |
78 |
allow infinite paths. Orthogonally, the nodes in the first and third types have |
|
79 |
finitely many direct subtrees, whereas those of the second and fourth may have |
|
80 |
infinite branching. |
|
52840 | 81 |
|
55073 | 82 |
The package is part of @{theory Main}. Additional functionality is provided by |
55129
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
83 |
the theory @{theory BNF_Decl}, located in the directory |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
84 |
\verb|~~/src/HOL/Library|. |
55073 | 85 |
|
52805 | 86 |
The package, like its predecessor, fully adheres to the LCF philosophy |
87 |
\cite{mgordon79}: The characteristic theorems associated with the specified |
|
88 |
(co)datatypes are derived rather than introduced axiomatically.% |
|
53543 | 89 |
\footnote{If the @{text quick_and_dirty} option is enabled, some of the |
52841 | 90 |
internal constructions and most of the internal proof obligations are skipped.} |
52805 | 91 |
The package's metatheory is described in a pair of papers |
53552 | 92 |
\cite{traytel-et-al-2012,blanchette-et-al-wit}. The central notion is that of a |
93 |
\emph{bounded natural functor} (BNF)---a well-behaved type constructor for which |
|
94 |
nested (co)recursion is supported. |
|
52792 | 95 |
|
52794 | 96 |
This tutorial is organized as follows: |
97 |
||
52805 | 98 |
\begin{itemize} |
52822 | 99 |
\setlength{\itemsep}{0pt} |
100 |
||
52805 | 101 |
\item Section \ref{sec:defining-datatypes}, ``Defining Datatypes,'' |
52822 | 102 |
describes how to specify datatypes using the @{command datatype_new} command. |
52805 | 103 |
|
53018 | 104 |
\item Section \ref{sec:defining-recursive-functions}, ``Defining Recursive |
52805 | 105 |
Functions,'' describes how to specify recursive functions using |
53535 | 106 |
@{command primrec_new}, \keyw{fun}, and \keyw{function}. |
52805 | 107 |
|
108 |
\item Section \ref{sec:defining-codatatypes}, ``Defining Codatatypes,'' |
|
53829 | 109 |
describes how to specify codatatypes using the @{command codatatype} command. |
52805 | 110 |
|
53018 | 111 |
\item Section \ref{sec:defining-corecursive-functions}, ``Defining Corecursive |
52805 | 112 |
Functions,'' describes how to specify corecursive functions using the |
53826 | 113 |
@{command primcorec} and @{command primcorecursive} commands. |
52794 | 114 |
|
53018 | 115 |
\item Section \ref{sec:registering-bounded-natural-functors}, ``Registering |
53552 | 116 |
Bounded Natural Functors,'' explains how to use the @{command bnf} command |
117 |
to register arbitrary type constructors as BNFs. |
|
52805 | 118 |
|
53552 | 119 |
\item Section |
53617 | 120 |
\ref{sec:deriving-destructors-and-theorems-for-free-constructors}, |
121 |
``Deriving Destructors and Theorems for Free Constructors,'' explains how to |
|
53552 | 122 |
use the command @{command wrap_free_constructors} to derive destructor constants |
123 |
and theorems for freely generated types, as performed internally by @{command |
|
53829 | 124 |
datatype_new} and @{command codatatype}. |
52794 | 125 |
|
53617 | 126 |
%\item Section \ref{sec:standard-ml-interface}, ``Standard ML Interface,'' |
127 |
%describes the package's programmatic interface. |
|
52794 | 128 |
|
53617 | 129 |
%\item Section \ref{sec:interoperability}, ``Interoperability,'' |
130 |
%is concerned with the packages' interaction with other Isabelle packages and |
|
131 |
%tools, such as the code generator and the counterexample generators. |
|
52805 | 132 |
|
53617 | 133 |
%\item Section \ref{sec:known-bugs-and-limitations}, ``Known Bugs and |
134 |
%Limitations,'' concludes with known open issues at the time of writing. |
|
52805 | 135 |
\end{itemize} |
52822 | 136 |
|
137 |
||
138 |
\newbox\boxA |
|
54185 | 139 |
\setbox\boxA=\hbox{\texttt{NOSPAM}} |
140 |
||
141 |
\newcommand\authoremaili{\texttt{blan{\color{white}NOSPAM}\kern-\wd\boxA{}chette@\allowbreak |
|
52822 | 142 |
in.\allowbreak tum.\allowbreak de}} |
54185 | 143 |
\newcommand\authoremailii{\texttt{lore{\color{white}NOSPAM}\kern-\wd\boxA{}nz.panny@\allowbreak |
53028 | 144 |
\allowbreak tum.\allowbreak de}} |
54185 | 145 |
\newcommand\authoremailiii{\texttt{pope{\color{white}NOSPAM}\kern-\wd\boxA{}scua@\allowbreak |
52822 | 146 |
in.\allowbreak tum.\allowbreak de}} |
54185 | 147 |
\newcommand\authoremailiv{\texttt{tray{\color{white}NOSPAM}\kern-\wd\boxA{}tel@\allowbreak |
52822 | 148 |
in.\allowbreak tum.\allowbreak de}} |
149 |
||
53028 | 150 |
The commands @{command datatype_new} and @{command primrec_new} are expected to |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
151 |
replace \keyw{datatype} and \keyw{primrec} in a future release. Authors of new |
53535 | 152 |
theories are encouraged to use the new commands, and maintainers of older |
53028 | 153 |
theories may want to consider upgrading. |
52841 | 154 |
|
155 |
Comments and bug reports concerning either the tool or this tutorial should be |
|
53028 | 156 |
directed to the authors at \authoremaili, \authoremailii, \authoremailiii, |
157 |
and \authoremailiv. |
|
52794 | 158 |
*} |
159 |
||
53491 | 160 |
|
52827 | 161 |
section {* Defining Datatypes |
52805 | 162 |
\label{sec:defining-datatypes} *} |
163 |
||
164 |
text {* |
|
53646 | 165 |
Datatypes can be specified using the @{command datatype_new} command. |
52805 | 166 |
*} |
52792 | 167 |
|
52824 | 168 |
|
53617 | 169 |
subsection {* Introductory Examples |
170 |
\label{ssec:datatype-introductory-examples} *} |
|
52794 | 171 |
|
53646 | 172 |
text {* |
173 |
Datatypes are illustrated through concrete examples featuring different flavors |
|
174 |
of recursion. More examples can be found in the directory |
|
54185 | 175 |
\verb|~~/src/HOL/|\allowbreak\verb|BNF/Examples|. |
53646 | 176 |
*} |
53621 | 177 |
|
178 |
subsubsection {* Nonrecursive Types |
|
179 |
\label{sssec:datatype-nonrecursive-types} *} |
|
52794 | 180 |
|
52805 | 181 |
text {* |
53028 | 182 |
Datatypes are introduced by specifying the desired names and argument types for |
53491 | 183 |
their constructors. \emph{Enumeration} types are the simplest form of datatype. |
53028 | 184 |
All their constructors are nullary: |
52805 | 185 |
*} |
186 |
||
52828 | 187 |
datatype_new trool = Truue | Faalse | Perhaaps |
52805 | 188 |
|
189 |
text {* |
|
53028 | 190 |
\noindent |
53491 | 191 |
Here, @{const Truue}, @{const Faalse}, and @{const Perhaaps} have the type @{typ trool}. |
53028 | 192 |
|
193 |
Polymorphic types are possible, such as the following option type, modeled after |
|
194 |
its homologue from the @{theory Option} theory: |
|
52805 | 195 |
*} |
196 |
||
53025 | 197 |
(*<*) |
198 |
hide_const None Some |
|
54958
4933165fd112
do not use wrong constructor in auto-generated proof goal
panny
parents:
54955
diff
changeset
|
199 |
hide_type option |
53025 | 200 |
(*>*) |
201 |
datatype_new 'a option = None | Some 'a |
|
52805 | 202 |
|
203 |
text {* |
|
53028 | 204 |
\noindent |
53491 | 205 |
The constructors are @{text "None :: 'a option"} and |
206 |
@{text "Some :: 'a \<Rightarrow> 'a option"}. |
|
53028 | 207 |
|
208 |
The next example has three type parameters: |
|
52805 | 209 |
*} |
210 |
||
211 |
datatype_new ('a, 'b, 'c) triple = Triple 'a 'b 'c |
|
212 |
||
53028 | 213 |
text {* |
214 |
\noindent |
|
215 |
The constructor is |
|
53491 | 216 |
@{text "Triple :: 'a \<Rightarrow> 'b \<Rightarrow> 'c \<Rightarrow> ('a, 'b, 'c) triple"}. |
53028 | 217 |
Unlike in Standard ML, curried constructors are supported. The uncurried variant |
218 |
is also possible: |
|
219 |
*} |
|
220 |
||
221 |
datatype_new ('a, 'b, 'c) triple\<^sub>u = Triple\<^sub>u "'a * 'b * 'c" |
|
222 |
||
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
223 |
text {* |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
224 |
\noindent |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
225 |
Occurrences of nonatomic types on the right-hand side of the equal sign must be |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
226 |
enclosed in double quotes, as is customary in Isabelle. |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
227 |
*} |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
228 |
|
52824 | 229 |
|
53621 | 230 |
subsubsection {* Simple Recursion |
231 |
\label{sssec:datatype-simple-recursion} *} |
|
52794 | 232 |
|
52805 | 233 |
text {* |
53491 | 234 |
Natural numbers are the simplest example of a recursive type: |
52805 | 235 |
*} |
236 |
||
53332 | 237 |
datatype_new nat = Zero | Suc nat |
52805 | 238 |
|
239 |
text {* |
|
53491 | 240 |
\noindent |
54187 | 241 |
Lists were shown in the introduction. Terminated lists are a variant that |
242 |
stores a value of type @{typ 'b} at the very end: |
|
52805 | 243 |
*} |
244 |
||
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
245 |
datatype_new (*<*)(in early) (*>*)('a, 'b) tlist = TNil 'b | TCons 'a "('a, 'b) tlist" |
52805 | 246 |
|
52824 | 247 |
|
53621 | 248 |
subsubsection {* Mutual Recursion |
249 |
\label{sssec:datatype-mutual-recursion} *} |
|
52794 | 250 |
|
52805 | 251 |
text {* |
53552 | 252 |
\emph{Mutually recursive} types are introduced simultaneously and may refer to |
253 |
each other. The example below introduces a pair of types for even and odd |
|
254 |
natural numbers: |
|
52805 | 255 |
*} |
256 |
||
53623 | 257 |
datatype_new even_nat = Even_Zero | Even_Suc odd_nat |
258 |
and odd_nat = Odd_Suc even_nat |
|
52805 | 259 |
|
260 |
text {* |
|
53491 | 261 |
\noindent |
262 |
Arithmetic expressions are defined via terms, terms via factors, and factors via |
|
263 |
expressions: |
|
52805 | 264 |
*} |
265 |
||
52841 | 266 |
datatype_new ('a, 'b) exp = |
267 |
Term "('a, 'b) trm" | Sum "('a, 'b) trm" "('a, 'b) exp" |
|
52805 | 268 |
and ('a, 'b) trm = |
52841 | 269 |
Factor "('a, 'b) fct" | Prod "('a, 'b) fct" "('a, 'b) trm" |
270 |
and ('a, 'b) fct = |
|
271 |
Const 'a | Var 'b | Expr "('a, 'b) exp" |
|
52805 | 272 |
|
52824 | 273 |
|
53621 | 274 |
subsubsection {* Nested Recursion |
275 |
\label{sssec:datatype-nested-recursion} *} |
|
52794 | 276 |
|
52805 | 277 |
text {* |
53491 | 278 |
\emph{Nested recursion} occurs when recursive occurrences of a type appear under |
279 |
a type constructor. The introduction showed some examples of trees with nesting |
|
53675 | 280 |
through lists. A more complex example, that reuses our @{type option} type, |
53491 | 281 |
follows: |
52805 | 282 |
*} |
283 |
||
52843 | 284 |
datatype_new 'a btree = |
53025 | 285 |
BNode 'a "'a btree option" "'a btree option" |
52805 | 286 |
|
287 |
text {* |
|
53491 | 288 |
\noindent |
289 |
Not all nestings are admissible. For example, this command will fail: |
|
52805 | 290 |
*} |
291 |
||
54187 | 292 |
datatype_new 'a wrong = W1 | W2 (*<*)'a |
53542 | 293 |
typ (*>*)"'a wrong \<Rightarrow> 'a" |
52806 | 294 |
|
295 |
text {* |
|
53491 | 296 |
\noindent |
297 |
The issue is that the function arrow @{text "\<Rightarrow>"} allows recursion |
|
298 |
only through its right-hand side. This issue is inherited by polymorphic |
|
299 |
datatypes defined in terms of~@{text "\<Rightarrow>"}: |
|
300 |
*} |
|
301 |
||
55350 | 302 |
datatype_new ('a, 'b) fun_copy = Fun "'a \<Rightarrow> 'b" |
54187 | 303 |
datatype_new 'a also_wrong = W1 | W2 (*<*)'a |
55350 | 304 |
typ (*>*)"('a also_wrong, 'a) fun_copy" |
53491 | 305 |
|
306 |
text {* |
|
307 |
\noindent |
|
55129
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
308 |
The following definition of @{typ 'a}-branching trees is legal: |
53621 | 309 |
*} |
310 |
||
311 |
datatype_new 'a ftree = FTLeaf 'a | FTNode "'a \<Rightarrow> 'a ftree" |
|
312 |
||
313 |
text {* |
|
314 |
\noindent |
|
55129
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
315 |
And so is the definition of hereditarily finite sets: |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
316 |
*} |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
317 |
|
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
318 |
datatype_new hfset = HFSet "hfset fset" |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
319 |
|
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
320 |
text {* |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
321 |
\noindent |
53491 | 322 |
In general, type constructors @{text "('a\<^sub>1, \<dots>, 'a\<^sub>m) t"} |
323 |
allow recursion on a subset of their type arguments @{text 'a\<^sub>1}, \ldots, |
|
324 |
@{text 'a\<^sub>m}. These type arguments are called \emph{live}; the remaining |
|
325 |
type arguments are called \emph{dead}. In @{typ "'a \<Rightarrow> 'b"} and |
|
55350 | 326 |
@{typ "('a, 'b) fun_copy"}, the type variable @{typ 'a} is dead and |
327 |
@{typ 'b} is live. |
|
53552 | 328 |
|
53647 | 329 |
Type constructors must be registered as BNFs to have live arguments. This is |
330 |
done automatically for datatypes and codatatypes introduced by the @{command |
|
53829 | 331 |
datatype_new} and @{command codatatype} commands. |
53552 | 332 |
Section~\ref{sec:registering-bounded-natural-functors} explains how to register |
333 |
arbitrary type constructors as BNFs. |
|
54187 | 334 |
|
335 |
Here is another example that fails: |
|
52806 | 336 |
*} |
337 |
||
54187 | 338 |
datatype_new 'a pow_list = PNil 'a (*<*)'a |
339 |
datatype_new 'a pow_list' = PNil' 'a (*>*)| PCons "('a * 'a) pow_list" |
|
340 |
||
341 |
text {* |
|
342 |
\noindent |
|
343 |
This one features a different flavor of nesting, where the recursive call in the |
|
344 |
type specification occurs around (rather than inside) another type constructor. |
|
345 |
*} |
|
346 |
||
347 |
subsubsection {* Auxiliary Constants and Properties |
|
348 |
\label{sssec:datatype-auxiliary-constants-and-properties} *} |
|
52805 | 349 |
|
350 |
text {* |
|
53491 | 351 |
The @{command datatype_new} command introduces various constants in addition to |
53617 | 352 |
the constructors. With each datatype are associated set functions, a map |
353 |
function, a relator, discriminators, and selectors, all of which can be given |
|
54187 | 354 |
custom names. In the example below, the familiar names @{text null}, @{text hd}, |
355 |
@{text tl}, @{text set}, @{text map}, and @{text list_all2}, override the |
|
356 |
default names @{text is_Nil}, @{text un_Cons1}, @{text un_Cons2}, |
|
54491 | 357 |
@{text set_list}, @{text map_list}, and @{text rel_list}: |
52822 | 358 |
*} |
359 |
||
52841 | 360 |
(*<*) |
361 |
no_translations |
|
362 |
"[x, xs]" == "x # [xs]" |
|
363 |
"[x]" == "x # []" |
|
364 |
||
365 |
no_notation |
|
366 |
Nil ("[]") and |
|
367 |
Cons (infixr "#" 65) |
|
368 |
||
53543 | 369 |
hide_type list |
54494 | 370 |
hide_const Nil Cons hd tl set map list_all2 |
53025 | 371 |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
372 |
context early begin |
52841 | 373 |
(*>*) |
53025 | 374 |
datatype_new (set: 'a) list (map: map rel: list_all2) = |
52822 | 375 |
null: Nil (defaults tl: Nil) |
53025 | 376 |
| Cons (hd: 'a) (tl: "'a list") |
52822 | 377 |
|
378 |
text {* |
|
379 |
\noindent |
|
54187 | 380 |
|
381 |
\begin{tabular}{@ {}ll@ {}} |
|
382 |
Constructors: & |
|
383 |
@{text "Nil \<Colon> 'a list"} \\ |
|
384 |
& |
|
385 |
@{text "Cons \<Colon> 'a \<Rightarrow> 'a list \<Rightarrow> 'a list"} \\ |
|
386 |
Discriminator: & |
|
387 |
@{text "null \<Colon> 'a list \<Rightarrow> bool"} \\ |
|
388 |
Selectors: & |
|
389 |
@{text "hd \<Colon> 'a list \<Rightarrow> 'a"} \\ |
|
390 |
& |
|
391 |
@{text "tl \<Colon> 'a list \<Rightarrow> 'a list"} \\ |
|
392 |
Set function: & |
|
393 |
@{text "set \<Colon> 'a list \<Rightarrow> 'a set"} \\ |
|
394 |
Map function: & |
|
395 |
@{text "map \<Colon> ('a \<Rightarrow> 'b) \<Rightarrow> 'a list \<Rightarrow> 'b list"} \\ |
|
396 |
Relator: & |
|
397 |
@{text "list_all2 \<Colon> ('a \<Rightarrow> 'b \<Rightarrow> bool) \<Rightarrow> 'a list \<Rightarrow> 'b list \<Rightarrow> bool"} |
|
398 |
\end{tabular} |
|
399 |
||
400 |
The discriminator @{const null} and the selectors @{const hd} and @{const tl} |
|
401 |
are characterized as follows: |
|
52822 | 402 |
% |
53025 | 403 |
\[@{thm list.collapse(1)[of xs, no_vars]} |
404 |
\qquad @{thm list.collapse(2)[of xs, no_vars]}\] |
|
52822 | 405 |
% |
54187 | 406 |
For two-constructor datatypes, a single discriminator constant is sufficient. |
407 |
The discriminator associated with @{const Cons} is simply |
|
53491 | 408 |
@{term "\<lambda>xs. \<not> null xs"}. |
52822 | 409 |
|
53553 | 410 |
The @{text defaults} clause following the @{const Nil} constructor specifies a |
411 |
default value for selectors associated with other constructors. Here, it is used |
|
412 |
to ensure that the tail of the empty list is itself (instead of being left |
|
53535 | 413 |
unspecified). |
52822 | 414 |
|
53617 | 415 |
Because @{const Nil} is nullary, it is also possible to use |
53491 | 416 |
@{term "\<lambda>xs. xs = Nil"} as a discriminator. This is specified by |
53534 | 417 |
entering ``@{text "="}'' instead of the identifier @{const null}. Although this |
53535 | 418 |
may look appealing, the mixture of constructors and selectors in the |
53534 | 419 |
characteristic theorems can lead Isabelle's automation to switch between the |
420 |
constructor and the destructor view in surprising ways. |
|
52822 | 421 |
|
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
422 |
The usual mixfix syntax annotations are available for both types and |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
423 |
constructors. For example: |
52805 | 424 |
*} |
52794 | 425 |
|
53025 | 426 |
(*<*) |
427 |
end |
|
428 |
(*>*) |
|
53552 | 429 |
datatype_new ('a, 'b) prod (infixr "*" 20) = Pair 'a 'b |
430 |
||
431 |
text {* \blankline *} |
|
52822 | 432 |
|
52841 | 433 |
datatype_new (set: 'a) list (map: map rel: list_all2) = |
52822 | 434 |
null: Nil ("[]") |
52841 | 435 |
| Cons (hd: 'a) (tl: "'a list") (infixr "#" 65) |
436 |
||
437 |
text {* |
|
53535 | 438 |
\noindent |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
439 |
Incidentally, this is how the traditional syntax can be set up: |
52841 | 440 |
*} |
441 |
||
442 |
syntax "_list" :: "args \<Rightarrow> 'a list" ("[(_)]") |
|
443 |
||
53552 | 444 |
text {* \blankline *} |
445 |
||
52841 | 446 |
translations |
447 |
"[x, xs]" == "x # [xs]" |
|
448 |
"[x]" == "x # []" |
|
52822 | 449 |
|
52824 | 450 |
|
53617 | 451 |
subsection {* Command Syntax |
452 |
\label{ssec:datatype-command-syntax} *} |
|
453 |
||
454 |
||
53621 | 455 |
subsubsection {* \keyw{datatype\_new} |
456 |
\label{sssec:datatype-new} *} |
|
52794 | 457 |
|
52822 | 458 |
text {* |
53829 | 459 |
\begin{matharray}{rcl} |
460 |
@{command_def "datatype_new"} & : & @{text "local_theory \<rightarrow> local_theory"} |
|
461 |
\end{matharray} |
|
52822 | 462 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
463 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
464 |
@@{command datatype_new} target? @{syntax dt_options}? \<newline> |
52824 | 465 |
(@{syntax dt_name} '=' (@{syntax ctor} + '|') + @'and') |
52828 | 466 |
; |
54626 | 467 |
@{syntax_def dt_options}: '(' (('no_discs_sels' | 'no_code' | 'rep_compat') + ',') ')' |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
468 |
\<close>} |
52824 | 469 |
|
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
470 |
The syntactic entity \synt{target} can be used to specify a local |
53534 | 471 |
context---e.g., @{text "(in linorder)"}. It is documented in the Isar reference |
472 |
manual \cite{isabelle-isar-ref}. |
|
473 |
% |
|
54832 | 474 |
The optional target is potentially followed by datatype-specific options: |
52822 | 475 |
|
52824 | 476 |
\begin{itemize} |
477 |
\setlength{\itemsep}{0pt} |
|
478 |
||
479 |
\item |
|
53623 | 480 |
The @{text "no_discs_sels"} option indicates that no discriminators or selectors |
53543 | 481 |
should be generated. |
52822 | 482 |
|
52824 | 483 |
\item |
54626 | 484 |
The @{text "no_code"} option indicates that the datatype should not be |
485 |
registered for code generation. |
|
486 |
||
487 |
\item |
|
53644 | 488 |
The @{text "rep_compat"} option indicates that the generated names should |
489 |
contain optional (and normally not displayed) ``@{text "new."}'' components to |
|
490 |
prevent clashes with a later call to \keyw{rep\_datatype}. See |
|
52824 | 491 |
Section~\ref{ssec:datatype-compatibility-issues} for details. |
492 |
\end{itemize} |
|
52822 | 493 |
|
52827 | 494 |
The left-hand sides of the datatype equations specify the name of the type to |
53534 | 495 |
define, its type parameters, and additional information: |
52822 | 496 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
497 |
@{rail \<open> |
53534 | 498 |
@{syntax_def dt_name}: @{syntax tyargs}? name @{syntax map_rel}? mixfix? |
52824 | 499 |
; |
53534 | 500 |
@{syntax_def tyargs}: typefree | '(' ((name ':')? typefree + ',') ')' |
52824 | 501 |
; |
53534 | 502 |
@{syntax_def map_rel}: '(' ((('map' | 'rel') ':' name) +) ')' |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
503 |
\<close>} |
52822 | 504 |
|
52827 | 505 |
\noindent |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
506 |
The syntactic entity \synt{name} denotes an identifier, \synt{typefree} |
53534 | 507 |
denotes fixed type variable (@{typ 'a}, @{typ 'b}, \ldots), and \synt{mixfix} |
508 |
denotes the usual parenthesized mixfix notation. They are documented in the Isar |
|
509 |
reference manual \cite{isabelle-isar-ref}. |
|
52822 | 510 |
|
52827 | 511 |
The optional names preceding the type variables allow to override the default |
54491 | 512 |
names of the set functions (@{text set1_t}, \ldots, @{text setM_t}). |
53647 | 513 |
Inside a mutually recursive specification, all defined datatypes must |
514 |
mention exactly the same type variables in the same order. |
|
52822 | 515 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
516 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
517 |
@{syntax_def ctor}: (name ':')? name (@{syntax ctor_arg} * ) \<newline> |
53534 | 518 |
@{syntax dt_sel_defaults}? mixfix? |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
519 |
\<close>} |
52824 | 520 |
|
53535 | 521 |
\medskip |
522 |
||
52827 | 523 |
\noindent |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
524 |
The main constituents of a constructor specification are the name of the |
52827 | 525 |
constructor and the list of its argument types. An optional discriminator name |
53554 | 526 |
can be supplied at the front to override the default name |
527 |
(@{text t.is_C\<^sub>j}). |
|
52822 | 528 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
529 |
@{rail \<open> |
53534 | 530 |
@{syntax_def ctor_arg}: type | '(' name ':' type ')' |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
531 |
\<close>} |
52827 | 532 |
|
53535 | 533 |
\medskip |
534 |
||
52827 | 535 |
\noindent |
536 |
In addition to the type of a constructor argument, it is possible to specify a |
|
537 |
name for the corresponding selector to override the default name |
|
53554 | 538 |
(@{text un_C\<^sub>ji}). The same selector names can be reused for several |
539 |
constructors as long as they share the same type. |
|
52827 | 540 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
541 |
@{rail \<open> |
53621 | 542 |
@{syntax_def dt_sel_defaults}: '(' 'defaults' (name ':' term +) ')' |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
543 |
\<close>} |
52827 | 544 |
|
545 |
\noindent |
|
546 |
Given a constructor |
|
547 |
@{text "C \<Colon> \<sigma>\<^sub>1 \<Rightarrow> \<dots> \<Rightarrow> \<sigma>\<^sub>p \<Rightarrow> \<sigma>"}, |
|
548 |
default values can be specified for any selector |
|
549 |
@{text "un_D \<Colon> \<sigma> \<Rightarrow> \<tau>"} |
|
53535 | 550 |
associated with other constructors. The specified default value must be of type |
52828 | 551 |
@{text "\<sigma>\<^sub>1 \<Rightarrow> \<dots> \<Rightarrow> \<sigma>\<^sub>p \<Rightarrow> \<tau>"} |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
552 |
(i.e., it may depend on @{text C}'s arguments). |
52822 | 553 |
*} |
554 |
||
53617 | 555 |
|
53621 | 556 |
subsubsection {* \keyw{datatype\_new\_compat} |
557 |
\label{sssec:datatype-new-compat} *} |
|
53617 | 558 |
|
559 |
text {* |
|
53829 | 560 |
\begin{matharray}{rcl} |
561 |
@{command_def "datatype_new_compat"} & : & @{text "local_theory \<rightarrow> local_theory"} |
|
562 |
\end{matharray} |
|
563 |
||
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
564 |
@{rail \<open> |
53829 | 565 |
@@{command datatype_new_compat} names |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
566 |
\<close>} |
53829 | 567 |
|
568 |
\noindent |
|
53621 | 569 |
The old datatype package provides some functionality that is not yet replicated |
570 |
in the new package: |
|
571 |
||
572 |
\begin{itemize} |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
573 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
574 |
|
53621 | 575 |
\item It is integrated with \keyw{fun} and \keyw{function} |
576 |
\cite{isabelle-function}, Nitpick \cite{isabelle-nitpick}, Quickcheck, |
|
577 |
and other packages. |
|
578 |
||
579 |
\item It is extended by various add-ons, notably to produce instances of the |
|
580 |
@{const size} function. |
|
581 |
\end{itemize} |
|
582 |
||
583 |
\noindent |
|
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
584 |
New-style datatypes can in most cases be registered as old-style datatypes using |
53829 | 585 |
@{command datatype_new_compat}. The \textit{names} argument is a space-separated |
586 |
list of type names that are mutually recursive. For example: |
|
53621 | 587 |
*} |
588 |
||
53623 | 589 |
datatype_new_compat even_nat odd_nat |
53621 | 590 |
|
591 |
text {* \blankline *} |
|
592 |
||
53623 | 593 |
thm even_nat_odd_nat.size |
53621 | 594 |
|
595 |
text {* \blankline *} |
|
596 |
||
53623 | 597 |
ML {* Datatype_Data.get_info @{theory} @{type_name even_nat} *} |
53621 | 598 |
|
599 |
text {* |
|
53748 | 600 |
A few remarks concern nested recursive datatypes only: |
601 |
||
602 |
\begin{itemize} |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
603 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
604 |
|
53748 | 605 |
\item The old-style, nested-as-mutual induction rule, iterator theorems, and |
606 |
recursor theorems are generated under their usual names but with ``@{text |
|
607 |
"compat_"}'' prefixed (e.g., @{text compat_tree.induct}). |
|
608 |
||
609 |
\item All types through which recursion takes place must be new-style datatypes |
|
610 |
or the function type. In principle, it should be possible to support old-style |
|
611 |
datatypes as well, but the command does not support this yet (and there is |
|
612 |
currently no way to register old-style datatypes as new-style datatypes). |
|
54184 | 613 |
|
614 |
\item The recursor produced for types that recurse through functions has a |
|
615 |
different signature than with the old package. This makes it impossible to use |
|
616 |
the old \keyw{primrec} command. |
|
53748 | 617 |
\end{itemize} |
618 |
||
619 |
An alternative to @{command datatype_new_compat} is to use the old package's |
|
620 |
\keyw{rep\_datatype} command. The associated proof obligations must then be |
|
621 |
discharged manually. |
|
53617 | 622 |
*} |
623 |
||
624 |
||
625 |
subsection {* Generated Constants |
|
626 |
\label{ssec:datatype-generated-constants} *} |
|
627 |
||
628 |
text {* |
|
53623 | 629 |
Given a datatype @{text "('a\<^sub>1, \<dots>, 'a\<^sub>m) t"} |
53617 | 630 |
with $m > 0$ live type variables and $n$ constructors |
631 |
@{text "t.C\<^sub>1"}, \ldots, @{text "t.C\<^sub>n"}, the |
|
632 |
following auxiliary constants are introduced: |
|
633 |
||
634 |
\begin{itemize} |
|
635 |
\setlength{\itemsep}{0pt} |
|
636 |
||
54494 | 637 |
\item \relax{Case combinator}: @{text t.case_t} (rendered using the familiar |
53617 | 638 |
@{text case}--@{text of} syntax) |
639 |
||
640 |
\item \relax{Discriminators}: @{text "t.is_C\<^sub>1"}, \ldots, |
|
641 |
@{text "t.is_C\<^sub>n"} |
|
642 |
||
643 |
\item \relax{Selectors}: |
|
644 |
@{text t.un_C\<^sub>11}$, \ldots, @{text t.un_C\<^sub>1k\<^sub>1}, \\ |
|
645 |
\phantom{\relax{Selectors:}} \quad\vdots \\ |
|
646 |
\phantom{\relax{Selectors:}} @{text t.un_C\<^sub>n1}$, \ldots, @{text t.un_C\<^sub>nk\<^sub>n}. |
|
647 |
||
648 |
\item \relax{Set functions} (or \relax{natural transformations}): |
|
54491 | 649 |
@{text set1_t}, \ldots, @{text t.setm_t} |
650 |
||
651 |
\item \relax{Map function} (or \relax{functorial action}): @{text t.map_t} |
|
652 |
||
653 |
\item \relax{Relator}: @{text t.rel_t} |
|
654 |
||
54494 | 655 |
\item \relax{Iterator}: @{text t.fold_t} |
656 |
||
657 |
\item \relax{Recursor}: @{text t.rec_t} |
|
53617 | 658 |
|
659 |
\end{itemize} |
|
660 |
||
661 |
\noindent |
|
662 |
The case combinator, discriminators, and selectors are collectively called |
|
663 |
\emph{destructors}. The prefix ``@{text "t."}'' is an optional component of the |
|
54491 | 664 |
names and is normally hidden. |
53617 | 665 |
*} |
666 |
||
667 |
||
52840 | 668 |
subsection {* Generated Theorems |
669 |
\label{ssec:datatype-generated-theorems} *} |
|
52828 | 670 |
|
671 |
text {* |
|
53544 | 672 |
The characteristic theorems generated by @{command datatype_new} are grouped in |
53623 | 673 |
three broad categories: |
53535 | 674 |
|
53543 | 675 |
\begin{itemize} |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
676 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
677 |
|
53543 | 678 |
\item The \emph{free constructor theorems} are properties about the constructors |
679 |
and destructors that can be derived for any freely generated type. Internally, |
|
53542 | 680 |
the derivation is performed by @{command wrap_free_constructors}. |
53535 | 681 |
|
53552 | 682 |
\item The \emph{functorial theorems} are properties of datatypes related to |
683 |
their BNF nature. |
|
684 |
||
685 |
\item The \emph{inductive theorems} are properties of datatypes related to |
|
53544 | 686 |
their inductive nature. |
53552 | 687 |
|
53543 | 688 |
\end{itemize} |
53535 | 689 |
|
690 |
\noindent |
|
53542 | 691 |
The full list of named theorems can be obtained as usual by entering the |
53543 | 692 |
command \keyw{print\_theorems} immediately after the datatype definition. |
53542 | 693 |
This list normally excludes low-level theorems that reveal internal |
53552 | 694 |
constructions. To make these accessible, add the line |
53542 | 695 |
*} |
53535 | 696 |
|
53542 | 697 |
declare [[bnf_note_all]] |
698 |
(*<*) |
|
699 |
declare [[bnf_note_all = false]] |
|
700 |
(*>*) |
|
53535 | 701 |
|
53552 | 702 |
text {* |
703 |
\noindent |
|
704 |
to the top of the theory file. |
|
705 |
*} |
|
53535 | 706 |
|
53621 | 707 |
subsubsection {* Free Constructor Theorems |
708 |
\label{sssec:free-constructor-theorems} *} |
|
53535 | 709 |
|
53543 | 710 |
(*<*) |
53837 | 711 |
consts nonnull :: 'a |
53543 | 712 |
(*>*) |
713 |
||
53535 | 714 |
text {* |
54621 | 715 |
The free constructor theorems are partitioned in three subgroups. The first |
716 |
subgroup of properties is concerned with the constructors. They are listed below |
|
717 |
for @{typ "'a list"}: |
|
53543 | 718 |
|
53552 | 719 |
\begin{indentblock} |
53543 | 720 |
\begin{description} |
53544 | 721 |
|
53642 | 722 |
\item[@{text "t."}\hthm{inject} @{text "[iff, induct_simp]"}\rm:] ~ \\ |
53544 | 723 |
@{thm list.inject[no_vars]} |
724 |
||
53642 | 725 |
\item[@{text "t."}\hthm{distinct} @{text "[simp, induct_simp]"}\rm:] ~ \\ |
53543 | 726 |
@{thm list.distinct(1)[no_vars]} \\ |
727 |
@{thm list.distinct(2)[no_vars]} |
|
728 |
||
53642 | 729 |
\item[@{text "t."}\hthm{exhaust} @{text "[cases t, case_names C\<^sub>1 \<dots> C\<^sub>n]"}\rm:] ~ \\ |
53543 | 730 |
@{thm list.exhaust[no_vars]} |
731 |
||
53642 | 732 |
\item[@{text "t."}\hthm{nchotomy}\rm:] ~ \\ |
53543 | 733 |
@{thm list.nchotomy[no_vars]} |
734 |
||
735 |
\end{description} |
|
53552 | 736 |
\end{indentblock} |
53543 | 737 |
|
738 |
\noindent |
|
53621 | 739 |
In addition, these nameless theorems are registered as safe elimination rules: |
740 |
||
741 |
\begin{indentblock} |
|
742 |
\begin{description} |
|
743 |
||
54386 | 744 |
\item[@{text "t."}\hthm{distinct {\upshape[}THEN notE}@{text ", elim!"}\hthm{\upshape]}\rm:] ~ \\ |
53621 | 745 |
@{thm list.distinct(1)[THEN notE, elim!, no_vars]} \\ |
746 |
@{thm list.distinct(2)[THEN notE, elim!, no_vars]} |
|
747 |
||
748 |
\end{description} |
|
749 |
\end{indentblock} |
|
750 |
||
751 |
\noindent |
|
53543 | 752 |
The next subgroup is concerned with the case combinator: |
753 |
||
53552 | 754 |
\begin{indentblock} |
53543 | 755 |
\begin{description} |
53544 | 756 |
|
53798 | 757 |
\item[@{text "t."}\hthm{case} @{text "[simp, code]"}\rm:] ~ \\ |
53543 | 758 |
@{thm list.case(1)[no_vars]} \\ |
759 |
@{thm list.case(2)[no_vars]} |
|
760 |
||
53642 | 761 |
\item[@{text "t."}\hthm{case\_cong}\rm:] ~ \\ |
53543 | 762 |
@{thm list.case_cong[no_vars]} |
763 |
||
53642 | 764 |
\item[@{text "t."}\hthm{weak\_case\_cong} @{text "[cong]"}\rm:] ~ \\ |
53543 | 765 |
@{thm list.weak_case_cong[no_vars]} |
766 |
||
53642 | 767 |
\item[@{text "t."}\hthm{split}\rm:] ~ \\ |
53543 | 768 |
@{thm list.split[no_vars]} |
769 |
||
53642 | 770 |
\item[@{text "t."}\hthm{split\_asm}\rm:] ~ \\ |
53543 | 771 |
@{thm list.split_asm[no_vars]} |
772 |
||
53544 | 773 |
\item[@{text "t."}\hthm{splits} = @{text "split split_asm"}] |
53543 | 774 |
|
775 |
\end{description} |
|
53552 | 776 |
\end{indentblock} |
53543 | 777 |
|
778 |
\noindent |
|
54621 | 779 |
The third subgroup revolves around discriminators and selectors: |
53543 | 780 |
|
53552 | 781 |
\begin{indentblock} |
53543 | 782 |
\begin{description} |
53544 | 783 |
|
53694 | 784 |
\item[@{text "t."}\hthm{disc} @{text "[simp]"}\rm:] ~ \\ |
785 |
@{thm list.disc(1)[no_vars]} \\ |
|
786 |
@{thm list.disc(2)[no_vars]} |
|
787 |
||
53703 | 788 |
\item[@{text "t."}\hthm{discI}\rm:] ~ \\ |
789 |
@{thm list.discI(1)[no_vars]} \\ |
|
790 |
@{thm list.discI(2)[no_vars]} |
|
791 |
||
53805 | 792 |
\item[@{text "t."}\hthm{sel} @{text "[simp, code]"}\rm:] ~ \\ |
53694 | 793 |
@{thm list.sel(1)[no_vars]} \\ |
794 |
@{thm list.sel(2)[no_vars]} |
|
53543 | 795 |
|
53642 | 796 |
\item[@{text "t."}\hthm{collapse} @{text "[simp]"}\rm:] ~ \\ |
53543 | 797 |
@{thm list.collapse(1)[no_vars]} \\ |
798 |
@{thm list.collapse(2)[no_vars]} |
|
799 |
||
53837 | 800 |
\item[@{text "t."}\hthm{disc\_exclude} @{text "[dest]"}\rm:] ~ \\ |
53543 | 801 |
These properties are missing for @{typ "'a list"} because there is only one |
802 |
proper discriminator. Had the datatype been introduced with a second |
|
53837 | 803 |
discriminator called @{const nonnull}, they would have read thusly: \\[\jot] |
804 |
@{prop "null list \<Longrightarrow> \<not> nonnull list"} \\ |
|
805 |
@{prop "nonnull list \<Longrightarrow> \<not> null list"} |
|
53543 | 806 |
|
53642 | 807 |
\item[@{text "t."}\hthm{disc\_exhaust} @{text "[case_names C\<^sub>1 \<dots> C\<^sub>n]"}\rm:] ~ \\ |
53543 | 808 |
@{thm list.disc_exhaust[no_vars]} |
809 |
||
53916 | 810 |
\item[@{text "t."}\hthm{sel\_exhaust} @{text "[case_names C\<^sub>1 \<dots> C\<^sub>n]"}\rm:] ~ \\ |
811 |
@{thm list.sel_exhaust[no_vars]} |
|
812 |
||
53642 | 813 |
\item[@{text "t."}\hthm{expand}\rm:] ~ \\ |
53543 | 814 |
@{thm list.expand[no_vars]} |
815 |
||
53917 | 816 |
\item[@{text "t."}\hthm{sel\_split}\rm:] ~ \\ |
817 |
@{thm list.sel_split[no_vars]} |
|
818 |
||
819 |
\item[@{text "t."}\hthm{sel\_split\_asm}\rm:] ~ \\ |
|
820 |
@{thm list.sel_split_asm[no_vars]} |
|
821 |
||
54491 | 822 |
\item[@{text "t."}\hthm{case\_eq\_if}\rm:] ~ \\ |
823 |
@{thm list.case_eq_if[no_vars]} |
|
53543 | 824 |
|
825 |
\end{description} |
|
53552 | 826 |
\end{indentblock} |
54152 | 827 |
|
828 |
\noindent |
|
829 |
In addition, equational versions of @{text t.disc} are registered with the @{text "[code]"} |
|
830 |
attribute. |
|
53552 | 831 |
*} |
832 |
||
833 |
||
53621 | 834 |
subsubsection {* Functorial Theorems |
835 |
\label{sssec:functorial-theorems} *} |
|
53552 | 836 |
|
837 |
text {* |
|
54621 | 838 |
The functorial theorems are partitioned in two subgroups. The first subgroup |
839 |
consists of properties involving the constructors and either a set function, the |
|
840 |
map function, or the relator: |
|
53552 | 841 |
|
842 |
\begin{indentblock} |
|
843 |
\begin{description} |
|
844 |
||
53798 | 845 |
\item[@{text "t."}\hthm{set} @{text "[simp, code]"}\rm:] ~ \\ |
53694 | 846 |
@{thm list.set(1)[no_vars]} \\ |
847 |
@{thm list.set(2)[no_vars]} |
|
53552 | 848 |
|
53798 | 849 |
\item[@{text "t."}\hthm{map} @{text "[simp, code]"}\rm:] ~ \\ |
53552 | 850 |
@{thm list.map(1)[no_vars]} \\ |
851 |
@{thm list.map(2)[no_vars]} |
|
852 |
||
54146 | 853 |
\item[@{text "t."}\hthm{rel\_inject} @{text "[simp]"}\rm:] ~ \\ |
53552 | 854 |
@{thm list.rel_inject(1)[no_vars]} \\ |
855 |
@{thm list.rel_inject(2)[no_vars]} |
|
856 |
||
54146 | 857 |
\item[@{text "t."}\hthm{rel\_distinct} @{text "[simp]"}\rm:] ~ \\ |
53552 | 858 |
@{thm list.rel_distinct(1)[no_vars]} \\ |
859 |
@{thm list.rel_distinct(2)[no_vars]} |
|
860 |
||
861 |
\end{description} |
|
862 |
\end{indentblock} |
|
54146 | 863 |
|
864 |
\noindent |
|
865 |
In addition, equational versions of @{text t.rel_inject} and @{text |
|
866 |
rel_distinct} are registered with the @{text "[code]"} attribute. |
|
54621 | 867 |
|
868 |
The second subgroup consists of more abstract properties of the set functions, |
|
869 |
the map function, and the relator: |
|
870 |
||
871 |
\begin{indentblock} |
|
872 |
\begin{description} |
|
873 |
||
874 |
\item[@{text "t."}\hthm{map\_comp}\rm:] ~ \\ |
|
875 |
@{thm list.map_cong0[no_vars]} |
|
876 |
||
54624
36301c99ed26
revert making 'map_cong' a 'cong' -- it breaks too many proofs in the AFP
blanchet
parents:
54621
diff
changeset
|
877 |
\item[@{text "t."}\hthm{map\_cong} @{text "[fundef_cong]"}\rm:] ~ \\ |
54621 | 878 |
@{thm list.map_cong[no_vars]} |
879 |
||
880 |
\item[@{text "t."}\hthm{map\_id}\rm:] ~ \\ |
|
881 |
@{thm list.map_id[no_vars]} |
|
882 |
||
883 |
\item[@{text "t."}\hthm{rel\_compp}\rm:] ~ \\ |
|
884 |
@{thm list.rel_compp[no_vars]} |
|
885 |
||
886 |
\item[@{text "t."}\hthm{rel\_conversep}\rm:] ~ \\ |
|
887 |
@{thm list.rel_conversep[no_vars]} |
|
888 |
||
889 |
\item[@{text "t."}\hthm{rel\_eq}\rm:] ~ \\ |
|
890 |
@{thm list.rel_eq[no_vars]} |
|
891 |
||
892 |
\item[@{text "t."}\hthm{rel\_flip}\rm:] ~ \\ |
|
893 |
@{thm list.rel_flip[no_vars]} |
|
894 |
||
895 |
\item[@{text "t."}\hthm{rel\_mono}\rm:] ~ \\ |
|
896 |
@{thm list.rel_mono[no_vars]} |
|
897 |
||
898 |
\item[@{text "t."}\hthm{set\_map}\rm:] ~ \\ |
|
899 |
@{thm list.set_map[no_vars]} |
|
900 |
||
901 |
\end{description} |
|
902 |
\end{indentblock} |
|
53535 | 903 |
*} |
904 |
||
905 |
||
53621 | 906 |
subsubsection {* Inductive Theorems |
907 |
\label{sssec:inductive-theorems} *} |
|
53535 | 908 |
|
909 |
text {* |
|
53623 | 910 |
The inductive theorems are as follows: |
53544 | 911 |
|
53552 | 912 |
\begin{indentblock} |
53544 | 913 |
\begin{description} |
914 |
||
53642 | 915 |
\item[@{text "t."}\hthm{induct} @{text "[induct t, case_names C\<^sub>1 \<dots> C\<^sub>n]"}\rm:] ~ \\ |
53544 | 916 |
@{thm list.induct[no_vars]} |
917 |
||
53642 | 918 |
\item[@{text "t\<^sub>1_\<dots>_t\<^sub>m."}\hthm{induct} @{text "[case_names C\<^sub>1 \<dots> C\<^sub>n]"}\rm:] ~ \\ |
53544 | 919 |
Given $m > 1$ mutually recursive datatypes, this induction rule can be used to |
920 |
prove $m$ properties simultaneously. |
|
52828 | 921 |
|
53798 | 922 |
\item[@{text "t."}\hthm{fold} @{text "[simp, code]"}\rm:] ~ \\ |
53544 | 923 |
@{thm list.fold(1)[no_vars]} \\ |
924 |
@{thm list.fold(2)[no_vars]} |
|
925 |
||
53798 | 926 |
\item[@{text "t."}\hthm{rec} @{text "[simp, code]"}\rm:] ~ \\ |
53544 | 927 |
@{thm list.rec(1)[no_vars]} \\ |
928 |
@{thm list.rec(2)[no_vars]} |
|
929 |
||
930 |
\end{description} |
|
53552 | 931 |
\end{indentblock} |
53544 | 932 |
|
933 |
\noindent |
|
934 |
For convenience, @{command datatype_new} also provides the following collection: |
|
935 |
||
53552 | 936 |
\begin{indentblock} |
53544 | 937 |
\begin{description} |
938 |
||
939 |
\item[@{text "t."}\hthm{simps} = @{text t.inject} @{text t.distinct} @{text t.case} @{text t.rec} @{text t.fold} @{text t.map} @{text t.rel_inject}] ~ \\ |
|
53694 | 940 |
@{text t.rel_distinct} @{text t.set} |
53544 | 941 |
|
942 |
\end{description} |
|
53552 | 943 |
\end{indentblock} |
52828 | 944 |
*} |
945 |
||
52794 | 946 |
|
52827 | 947 |
subsection {* Compatibility Issues |
52824 | 948 |
\label{ssec:datatype-compatibility-issues} *} |
52794 | 949 |
|
52828 | 950 |
text {* |
53997 | 951 |
The command @{command datatype_new} has been designed to be highly compatible |
952 |
with the old \keyw{datatype}, to ease migration. There are nonetheless a few |
|
53647 | 953 |
incompatibilities that may arise when porting to the new package: |
954 |
||
955 |
\begin{itemize} |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
956 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
957 |
|
53647 | 958 |
\item \emph{The Standard ML interfaces are different.} Tools and extensions |
959 |
written to call the old ML interfaces will need to be adapted to the new |
|
960 |
interfaces. Little has been done so far in this direction. Whenever possible, it |
|
961 |
is recommended to use @{command datatype_new_compat} or \keyw{rep\_datatype} |
|
962 |
to register new-style datatypes as old-style datatypes. |
|
963 |
||
54537 | 964 |
\item \emph{The constants @{text t_case} and @{text t_rec} are now called |
965 |
@{text case_t} and @{text rec_t}.} |
|
966 |
||
967 |
\item \emph{The recursor @{text rec_t} has a different signature for nested |
|
54185 | 968 |
recursive datatypes.} In the old package, nested recursion through non-functions |
969 |
was internally reduced to mutual recursion. This reduction was visible in the |
|
970 |
type of the recursor, used by \keyw{primrec}. Recursion through functions was |
|
971 |
handled specially. In the new package, nested recursion (for functions and |
|
972 |
non-functions) is handled in a more modular fashion. The old-style recursor can |
|
973 |
be generated on demand using @{command primrec_new}, as explained in |
|
53647 | 974 |
Section~\ref{sssec:primrec-nested-as-mutual-recursion}, if the recursion is via |
975 |
new-style datatypes. |
|
976 |
||
54287 | 977 |
\item \emph{Accordingly, the induction rule is different for nested recursive |
978 |
datatypes.} Again, the old-style induction rule can be generated on demand using |
|
979 |
@{command primrec_new}, as explained in |
|
53647 | 980 |
Section~\ref{sssec:primrec-nested-as-mutual-recursion}, if the recursion is via |
981 |
new-style datatypes. |
|
52828 | 982 |
|
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
983 |
\item \emph{The internal constructions are completely different.} Proof texts |
53647 | 984 |
that unfold the definition of constants introduced by \keyw{datatype} will be |
985 |
difficult to port. |
|
986 |
||
987 |
\item \emph{A few theorems have different names.} |
|
53997 | 988 |
The properties @{text t.cases} and @{text t.recs} have been renamed |
53647 | 989 |
@{text t.case} and @{text t.rec}. For non-mutually recursive datatypes, |
990 |
@{text t.inducts} is available as @{text t.induct}. |
|
991 |
For $m > 1$ mutually recursive datatypes, |
|
53997 | 992 |
@{text "t\<^sub>1_\<dots>_t\<^sub>m.inducts(i)"} has been renamed |
53647 | 993 |
@{text "t\<^sub>i.induct"}. |
994 |
||
995 |
\item \emph{The @{text t.simps} collection has been extended.} |
|
996 |
Previously available theorems are available at the same index. |
|
997 |
||
998 |
\item \emph{Variables in generated properties have different names.} This is |
|
999 |
rarely an issue, except in proof texts that refer to variable names in the |
|
1000 |
@{text "[where \<dots>]"} attribute. The solution is to use the more robust |
|
1001 |
@{text "[of \<dots>]"} syntax. |
|
1002 |
\end{itemize} |
|
1003 |
||
1004 |
In the other direction, there is currently no way to register old-style |
|
1005 |
datatypes as new-style datatypes. If the goal is to define new-style datatypes |
|
1006 |
with nested recursion through old-style datatypes, the old-style |
|
1007 |
datatypes can be registered as a BNF |
|
1008 |
(Section~\ref{sec:registering-bounded-natural-functors}). If the goal is |
|
1009 |
to derive discriminators and selectors, this can be achieved using @{command |
|
1010 |
wrap_free_constructors} |
|
1011 |
(Section~\ref{sec:deriving-destructors-and-theorems-for-free-constructors}). |
|
52828 | 1012 |
*} |
1013 |
||
52792 | 1014 |
|
52827 | 1015 |
section {* Defining Recursive Functions |
52805 | 1016 |
\label{sec:defining-recursive-functions} *} |
1017 |
||
1018 |
text {* |
|
54183 | 1019 |
Recursive functions over datatypes can be specified using the @{command |
1020 |
primrec_new} command, which supports primitive recursion, or using the more |
|
1021 |
general \keyw{fun} and \keyw{function} commands. Here, the focus is on @{command |
|
53644 | 1022 |
primrec_new}; the other two commands are described in a separate tutorial |
53646 | 1023 |
\cite{isabelle-function}. |
52828 | 1024 |
|
53621 | 1025 |
%%% TODO: partial_function |
52805 | 1026 |
*} |
52792 | 1027 |
|
52805 | 1028 |
|
53617 | 1029 |
subsection {* Introductory Examples |
1030 |
\label{ssec:primrec-introductory-examples} *} |
|
52828 | 1031 |
|
53646 | 1032 |
text {* |
1033 |
Primitive recursion is illustrated through concrete examples based on the |
|
1034 |
datatypes defined in Section~\ref{ssec:datatype-introductory-examples}. More |
|
55075 | 1035 |
examples can be found in the directory \verb|~~/src/HOL/BNF_Examples|. |
53646 | 1036 |
*} |
1037 |
||
53621 | 1038 |
|
1039 |
subsubsection {* Nonrecursive Types |
|
1040 |
\label{sssec:primrec-nonrecursive-types} *} |
|
52828 | 1041 |
|
52841 | 1042 |
text {* |
53621 | 1043 |
Primitive recursion removes one layer of constructors on the left-hand side in |
1044 |
each equation. For example: |
|
52841 | 1045 |
*} |
1046 |
||
1047 |
primrec_new bool_of_trool :: "trool \<Rightarrow> bool" where |
|
53621 | 1048 |
"bool_of_trool Faalse \<longleftrightarrow> False" | |
1049 |
"bool_of_trool Truue \<longleftrightarrow> True" |
|
52841 | 1050 |
|
53621 | 1051 |
text {* \blankline *} |
52841 | 1052 |
|
53025 | 1053 |
primrec_new the_list :: "'a option \<Rightarrow> 'a list" where |
1054 |
"the_list None = []" | |
|
1055 |
"the_list (Some a) = [a]" |
|
52841 | 1056 |
|
53621 | 1057 |
text {* \blankline *} |
1058 |
||
53025 | 1059 |
primrec_new the_default :: "'a \<Rightarrow> 'a option \<Rightarrow> 'a" where |
1060 |
"the_default d None = d" | |
|
1061 |
"the_default _ (Some a) = a" |
|
52843 | 1062 |
|
53621 | 1063 |
text {* \blankline *} |
1064 |
||
52841 | 1065 |
primrec_new mirrror :: "('a, 'b, 'c) triple \<Rightarrow> ('c, 'b, 'a) triple" where |
1066 |
"mirrror (Triple a b c) = Triple c b a" |
|
1067 |
||
53621 | 1068 |
text {* |
1069 |
\noindent |
|
1070 |
The equations can be specified in any order, and it is acceptable to leave out |
|
1071 |
some cases, which are then unspecified. Pattern matching on the left-hand side |
|
1072 |
is restricted to a single datatype, which must correspond to the same argument |
|
1073 |
in all equations. |
|
1074 |
*} |
|
52828 | 1075 |
|
53621 | 1076 |
|
1077 |
subsubsection {* Simple Recursion |
|
1078 |
\label{sssec:primrec-simple-recursion} *} |
|
52828 | 1079 |
|
52841 | 1080 |
text {* |
53621 | 1081 |
For simple recursive types, recursive calls on a constructor argument are |
1082 |
allowed on the right-hand side: |
|
52841 | 1083 |
*} |
1084 |
||
53330
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1085 |
primrec_new replicate :: "nat \<Rightarrow> 'a \<Rightarrow> 'a list" where |
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1086 |
"replicate Zero _ = []" | |
53644 | 1087 |
"replicate (Suc n) x = x # replicate n x" |
52841 | 1088 |
|
53621 | 1089 |
text {* \blankline *} |
52843 | 1090 |
|
53332 | 1091 |
primrec_new at :: "'a list \<Rightarrow> nat \<Rightarrow> 'a" where |
53644 | 1092 |
"at (x # xs) j = |
52843 | 1093 |
(case j of |
53644 | 1094 |
Zero \<Rightarrow> x |
1095 |
| Suc j' \<Rightarrow> at xs j')" |
|
52843 | 1096 |
|
53621 | 1097 |
text {* \blankline *} |
1098 |
||
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1099 |
primrec_new (*<*)(in early) (*>*)tfold :: "('a \<Rightarrow> 'b \<Rightarrow> 'b) \<Rightarrow> ('a, 'b) tlist \<Rightarrow> 'b" where |
53644 | 1100 |
"tfold _ (TNil y) = y" | |
1101 |
"tfold f (TCons x xs) = f x (tfold f xs)" |
|
52841 | 1102 |
|
53025 | 1103 |
text {* |
53621 | 1104 |
\noindent |
54402 | 1105 |
Pattern matching is only available for the argument on which the recursion takes |
1106 |
place. Fortunately, it is easy to generate pattern-maching equations using the |
|
1107 |
\keyw{simps\_of\_case} command provided by the theory |
|
55290 | 1108 |
\verb|~~/src/HOL/|\allowbreak\verb|Library/|\allowbreak\verb|Simps_Case_Conv|. |
54402 | 1109 |
*} |
1110 |
||
1111 |
simps_of_case at_simps: at.simps |
|
1112 |
||
1113 |
text {* |
|
1114 |
This generates the lemma collection @{thm [source] at_simps}: |
|
1115 |
% |
|
1116 |
\[@{thm at_simps(1)[no_vars]} |
|
1117 |
\qquad @{thm at_simps(2)[no_vars]}\] |
|
1118 |
% |
|
54184 | 1119 |
The next example is defined using \keyw{fun} to escape the syntactic |
55254 | 1120 |
restrictions imposed on primitively recursive functions. The |
54184 | 1121 |
@{command datatype_new_compat} command is needed to register new-style datatypes |
1122 |
for use with \keyw{fun} and \keyw{function} |
|
53621 | 1123 |
(Section~\ref{sssec:datatype-new-compat}): |
53025 | 1124 |
*} |
52828 | 1125 |
|
53621 | 1126 |
datatype_new_compat nat |
1127 |
||
1128 |
text {* \blankline *} |
|
1129 |
||
1130 |
fun at_least_two :: "nat \<Rightarrow> bool" where |
|
1131 |
"at_least_two (Suc (Suc _)) \<longleftrightarrow> True" | |
|
1132 |
"at_least_two _ \<longleftrightarrow> False" |
|
1133 |
||
1134 |
||
1135 |
subsubsection {* Mutual Recursion |
|
1136 |
\label{sssec:primrec-mutual-recursion} *} |
|
52828 | 1137 |
|
52841 | 1138 |
text {* |
53621 | 1139 |
The syntax for mutually recursive functions over mutually recursive datatypes |
1140 |
is straightforward: |
|
52841 | 1141 |
*} |
1142 |
||
1143 |
primrec_new |
|
53623 | 1144 |
nat_of_even_nat :: "even_nat \<Rightarrow> nat" and |
1145 |
nat_of_odd_nat :: "odd_nat \<Rightarrow> nat" |
|
52841 | 1146 |
where |
53623 | 1147 |
"nat_of_even_nat Even_Zero = Zero" | |
1148 |
"nat_of_even_nat (Even_Suc n) = Suc (nat_of_odd_nat n)" | |
|
1149 |
"nat_of_odd_nat (Odd_Suc n) = Suc (nat_of_even_nat n)" |
|
52841 | 1150 |
|
53752 | 1151 |
text {* \blankline *} |
1152 |
||
52841 | 1153 |
primrec_new |
53330
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1154 |
eval\<^sub>e :: "('a \<Rightarrow> int) \<Rightarrow> ('b \<Rightarrow> int) \<Rightarrow> ('a, 'b) exp \<Rightarrow> int" and |
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1155 |
eval\<^sub>t :: "('a \<Rightarrow> int) \<Rightarrow> ('b \<Rightarrow> int) \<Rightarrow> ('a, 'b) trm \<Rightarrow> int" and |
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1156 |
eval\<^sub>f :: "('a \<Rightarrow> int) \<Rightarrow> ('b \<Rightarrow> int) \<Rightarrow> ('a, 'b) fct \<Rightarrow> int" |
52841 | 1157 |
where |
1158 |
"eval\<^sub>e \<gamma> \<xi> (Term t) = eval\<^sub>t \<gamma> \<xi> t" | |
|
1159 |
"eval\<^sub>e \<gamma> \<xi> (Sum t e) = eval\<^sub>t \<gamma> \<xi> t + eval\<^sub>e \<gamma> \<xi> e" | |
|
53330
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1160 |
"eval\<^sub>t \<gamma> \<xi> (Factor f) = eval\<^sub>f \<gamma> \<xi> f" | |
52841 | 1161 |
"eval\<^sub>t \<gamma> \<xi> (Prod f t) = eval\<^sub>f \<gamma> \<xi> f + eval\<^sub>t \<gamma> \<xi> t" | |
1162 |
"eval\<^sub>f \<gamma> _ (Const a) = \<gamma> a" | |
|
1163 |
"eval\<^sub>f _ \<xi> (Var b) = \<xi> b" | |
|
1164 |
"eval\<^sub>f \<gamma> \<xi> (Expr e) = eval\<^sub>e \<gamma> \<xi> e" |
|
1165 |
||
53621 | 1166 |
text {* |
1167 |
\noindent |
|
53647 | 1168 |
Mutual recursion is possible within a single type, using \keyw{fun}: |
53621 | 1169 |
*} |
52828 | 1170 |
|
53621 | 1171 |
fun |
1172 |
even :: "nat \<Rightarrow> bool" and |
|
1173 |
odd :: "nat \<Rightarrow> bool" |
|
1174 |
where |
|
1175 |
"even Zero = True" | |
|
1176 |
"even (Suc n) = odd n" | |
|
1177 |
"odd Zero = False" | |
|
1178 |
"odd (Suc n) = even n" |
|
1179 |
||
1180 |
||
1181 |
subsubsection {* Nested Recursion |
|
1182 |
\label{sssec:primrec-nested-recursion} *} |
|
1183 |
||
1184 |
text {* |
|
1185 |
In a departure from the old datatype package, nested recursion is normally |
|
1186 |
handled via the map functions of the nesting type constructors. For example, |
|
1187 |
recursive calls are lifted to lists using @{const map}: |
|
1188 |
*} |
|
52828 | 1189 |
|
52843 | 1190 |
(*<*) |
53644 | 1191 |
datatype_new 'a tree\<^sub>f\<^sub>f = Node\<^sub>f\<^sub>f (lbl\<^sub>f\<^sub>f: 'a) (sub\<^sub>f\<^sub>f: "'a tree\<^sub>f\<^sub>f list") |
52843 | 1192 |
(*>*) |
53028 | 1193 |
primrec_new at\<^sub>f\<^sub>f :: "'a tree\<^sub>f\<^sub>f \<Rightarrow> nat list \<Rightarrow> 'a" where |
1194 |
"at\<^sub>f\<^sub>f (Node\<^sub>f\<^sub>f a ts) js = |
|
52843 | 1195 |
(case js of |
1196 |
[] \<Rightarrow> a |
|
53028 | 1197 |
| j # js' \<Rightarrow> at (map (\<lambda>t. at\<^sub>f\<^sub>f t js') ts) j)" |
52843 | 1198 |
|
53025 | 1199 |
text {* |
53647 | 1200 |
\noindent |
53621 | 1201 |
The next example features recursion through the @{text option} type. Although |
53623 | 1202 |
@{text option} is not a new-style datatype, it is registered as a BNF with the |
54491 | 1203 |
map function @{const map_option}: |
53025 | 1204 |
*} |
52843 | 1205 |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1206 |
primrec_new (*<*)(in early) (*>*)sum_btree :: "('a\<Colon>{zero,plus}) btree \<Rightarrow> 'a" where |
52843 | 1207 |
"sum_btree (BNode a lt rt) = |
54491 | 1208 |
a + the_default 0 (map_option sum_btree lt) + |
1209 |
the_default 0 (map_option sum_btree rt)" |
|
52843 | 1210 |
|
53136 | 1211 |
text {* |
53621 | 1212 |
\noindent |
1213 |
The same principle applies for arbitrary type constructors through which |
|
1214 |
recursion is possible. Notably, the map function for the function type |
|
1215 |
(@{text \<Rightarrow>}) is simply composition (@{text "op \<circ>"}): |
|
53136 | 1216 |
*} |
52828 | 1217 |
|
54182 | 1218 |
primrec_new (*<*)(in early) (*>*)relabel_ft :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a ftree \<Rightarrow> 'a ftree" where |
1219 |
"relabel_ft f (FTLeaf x) = FTLeaf (f x)" | |
|
1220 |
"relabel_ft f (FTNode g) = FTNode (relabel_ft f \<circ> g)" |
|
1221 |
||
1222 |
text {* |
|
1223 |
\noindent |
|
1224 |
For convenience, recursion through functions can also be expressed using |
|
1225 |
$\lambda$-abstractions and function application rather than through composition. |
|
1226 |
For example: |
|
1227 |
*} |
|
1228 |
||
1229 |
primrec_new relabel_ft :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a ftree \<Rightarrow> 'a ftree" where |
|
1230 |
"relabel_ft f (FTLeaf x) = FTLeaf (f x)" | |
|
1231 |
"relabel_ft f (FTNode g) = FTNode (\<lambda>x. relabel_ft f (g x))" |
|
52828 | 1232 |
|
54183 | 1233 |
text {* \blankline *} |
1234 |
||
1235 |
primrec_new subtree_ft :: "'a \<Rightarrow> 'a ftree \<Rightarrow> 'a ftree" where |
|
1236 |
"subtree_ft x (FTNode g) = g x" |
|
1237 |
||
52843 | 1238 |
text {* |
53621 | 1239 |
\noindent |
54182 | 1240 |
For recursion through curried $n$-ary functions, $n$ applications of |
1241 |
@{term "op \<circ>"} are necessary. The examples below illustrate the case where |
|
1242 |
$n = 2$: |
|
53621 | 1243 |
*} |
1244 |
||
54182 | 1245 |
datatype_new 'a ftree2 = FTLeaf2 'a | FTNode2 "'a \<Rightarrow> 'a \<Rightarrow> 'a ftree2" |
1246 |
||
1247 |
text {* \blankline *} |
|
1248 |
||
1249 |
primrec_new (*<*)(in early) (*>*)relabel_ft2 :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a ftree2 \<Rightarrow> 'a ftree2" where |
|
1250 |
"relabel_ft2 f (FTLeaf2 x) = FTLeaf2 (f x)" | |
|
1251 |
"relabel_ft2 f (FTNode2 g) = FTNode2 (op \<circ> (op \<circ> (relabel_ft2 f)) g)" |
|
1252 |
||
1253 |
text {* \blankline *} |
|
1254 |
||
1255 |
primrec_new relabel_ft2 :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a ftree2 \<Rightarrow> 'a ftree2" where |
|
1256 |
"relabel_ft2 f (FTLeaf2 x) = FTLeaf2 (f x)" | |
|
1257 |
"relabel_ft2 f (FTNode2 g) = FTNode2 (\<lambda>x y. relabel_ft2 f (g x y))" |
|
54031 | 1258 |
|
54183 | 1259 |
text {* \blankline *} |
1260 |
||
1261 |
primrec_new subtree_ft2 :: "'a \<Rightarrow> 'a \<Rightarrow> 'a ftree2 \<Rightarrow> 'a ftree2" where |
|
1262 |
"subtree_ft2 x y (FTNode2 g) = g x y" |
|
1263 |
||
53621 | 1264 |
|
1265 |
subsubsection {* Nested-as-Mutual Recursion |
|
53644 | 1266 |
\label{sssec:primrec-nested-as-mutual-recursion} *} |
53621 | 1267 |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1268 |
(*<*) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1269 |
locale n2m begin |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1270 |
(*>*) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1271 |
|
53621 | 1272 |
text {* |
1273 |
For compatibility with the old package, but also because it is sometimes |
|
1274 |
convenient in its own right, it is possible to treat nested recursive datatypes |
|
1275 |
as mutually recursive ones if the recursion takes place though new-style |
|
1276 |
datatypes. For example: |
|
52843 | 1277 |
*} |
1278 |
||
53331
20440c789759
prove theorem in the right context (that knows about local variables)
traytel
parents:
53330
diff
changeset
|
1279 |
primrec_new |
53647 | 1280 |
at\<^sub>f\<^sub>f :: "'a tree\<^sub>f\<^sub>f \<Rightarrow> nat list \<Rightarrow> 'a" and |
1281 |
ats\<^sub>f\<^sub>f :: "'a tree\<^sub>f\<^sub>f list \<Rightarrow> nat \<Rightarrow> nat list \<Rightarrow> 'a" |
|
52843 | 1282 |
where |
53647 | 1283 |
"at\<^sub>f\<^sub>f (Node\<^sub>f\<^sub>f a ts) js = |
52843 | 1284 |
(case js of |
1285 |
[] \<Rightarrow> a |
|
53647 | 1286 |
| j # js' \<Rightarrow> ats\<^sub>f\<^sub>f ts j js')" | |
1287 |
"ats\<^sub>f\<^sub>f (t # ts) j = |
|
52843 | 1288 |
(case j of |
53647 | 1289 |
Zero \<Rightarrow> at\<^sub>f\<^sub>f t |
1290 |
| Suc j' \<Rightarrow> ats\<^sub>f\<^sub>f ts j')" |
|
52843 | 1291 |
|
53647 | 1292 |
text {* |
1293 |
\noindent |
|
54287 | 1294 |
Appropriate induction rules are generated as |
54031 | 1295 |
@{thm [source] at\<^sub>f\<^sub>f.induct}, |
1296 |
@{thm [source] ats\<^sub>f\<^sub>f.induct}, and |
|
54287 | 1297 |
@{thm [source] at\<^sub>f\<^sub>f_ats\<^sub>f\<^sub>f.induct}. The |
1298 |
induction rules and the underlying recursors are generated on a per-need basis |
|
1299 |
and are kept in a cache to speed up subsequent definitions. |
|
53647 | 1300 |
|
1301 |
Here is a second example: |
|
1302 |
*} |
|
53621 | 1303 |
|
53331
20440c789759
prove theorem in the right context (that knows about local variables)
traytel
parents:
53330
diff
changeset
|
1304 |
primrec_new |
53330
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1305 |
sum_btree :: "('a\<Colon>{zero,plus}) btree \<Rightarrow> 'a" and |
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1306 |
sum_btree_option :: "'a btree option \<Rightarrow> 'a" |
52843 | 1307 |
where |
1308 |
"sum_btree (BNode a lt rt) = |
|
53025 | 1309 |
a + sum_btree_option lt + sum_btree_option rt" | |
53330
77da8d3c46e0
fixed docs w.r.t. availability of "primrec_new" and friends
blanchet
parents:
53262
diff
changeset
|
1310 |
"sum_btree_option None = 0" | |
53025 | 1311 |
"sum_btree_option (Some t) = sum_btree t" |
52843 | 1312 |
|
1313 |
text {* |
|
53621 | 1314 |
% * can pretend a nested type is mutually recursive (if purely inductive) |
1315 |
% * avoids the higher-order map |
|
1316 |
% * e.g. |
|
1317 |
||
53617 | 1318 |
% * this can always be avoided; |
1319 |
% * e.g. in our previous example, we first mapped the recursive |
|
1320 |
% calls, then we used a generic at function to retrieve the result |
|
1321 |
% |
|
1322 |
% * there's no hard-and-fast rule of when to use one or the other, |
|
1323 |
% just like there's no rule when to use fold and when to use |
|
1324 |
% primrec_new |
|
1325 |
% |
|
1326 |
% * higher-order approach, considering nesting as nesting, is more |
|
1327 |
% compositional -- e.g. we saw how we could reuse an existing polymorphic |
|
53647 | 1328 |
% at or the_default, whereas @{const ats\<^sub>f\<^sub>f} is much more specific |
53617 | 1329 |
% |
1330 |
% * but: |
|
1331 |
% * is perhaps less intuitive, because it requires higher-order thinking |
|
1332 |
% * may seem inefficient, and indeed with the code generator the |
|
1333 |
% mutually recursive version might be nicer |
|
1334 |
% * is somewhat indirect -- must apply a map first, then compute a result |
|
1335 |
% (cannot mix) |
|
53647 | 1336 |
% * the auxiliary functions like @{const ats\<^sub>f\<^sub>f} are sometimes useful in own right |
53617 | 1337 |
% |
1338 |
% * impact on automation unclear |
|
1339 |
% |
|
52843 | 1340 |
*} |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1341 |
(*<*) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1342 |
end |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1343 |
(*>*) |
52843 | 1344 |
|
52824 | 1345 |
|
53617 | 1346 |
subsection {* Command Syntax |
1347 |
\label{ssec:primrec-command-syntax} *} |
|
1348 |
||
1349 |
||
53621 | 1350 |
subsubsection {* \keyw{primrec\_new} |
1351 |
\label{sssec:primrec-new} *} |
|
52828 | 1352 |
|
1353 |
text {* |
|
53829 | 1354 |
\begin{matharray}{rcl} |
1355 |
@{command_def "primrec_new"} & : & @{text "local_theory \<rightarrow> local_theory"} |
|
1356 |
\end{matharray} |
|
52794 | 1357 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
1358 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
1359 |
@@{command primrec_new} target? fixes \<newline> @'where' (@{syntax pr_equation} + '|') |
52840 | 1360 |
; |
53829 | 1361 |
@{syntax_def pr_equation}: thmdecl? prop |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
1362 |
\<close>} |
52828 | 1363 |
*} |
1364 |
||
52840 | 1365 |
|
53619 | 1366 |
(* |
52840 | 1367 |
subsection {* Generated Theorems |
1368 |
\label{ssec:primrec-generated-theorems} *} |
|
52824 | 1369 |
|
52828 | 1370 |
text {* |
53617 | 1371 |
% * synthesized nonrecursive definition |
1372 |
% * user specification is rederived from it, exactly as entered |
|
1373 |
% |
|
1374 |
% * induct |
|
1375 |
% * mutualized |
|
1376 |
% * without some needless induction hypotheses if not used |
|
1377 |
% * fold, rec |
|
1378 |
% * mutualized |
|
52828 | 1379 |
*} |
53619 | 1380 |
*) |
1381 |
||
52824 | 1382 |
|
52840 | 1383 |
subsection {* Recursive Default Values for Selectors |
53623 | 1384 |
\label{ssec:primrec-recursive-default-values-for-selectors} *} |
52827 | 1385 |
|
1386 |
text {* |
|
1387 |
A datatype selector @{text un_D} can have a default value for each constructor |
|
1388 |
on which it is not otherwise specified. Occasionally, it is useful to have the |
|
1389 |
default value be defined recursively. This produces a chicken-and-egg situation |
|
53621 | 1390 |
that may seem unsolvable, because the datatype is not introduced yet at the |
52827 | 1391 |
moment when the selectors are introduced. Of course, we can always define the |
1392 |
selectors manually afterward, but we then have to state and prove all the |
|
1393 |
characteristic theorems ourselves instead of letting the package do it. |
|
1394 |
||
1395 |
Fortunately, there is a fairly elegant workaround that relies on overloading and |
|
1396 |
that avoids the tedium of manual derivations: |
|
1397 |
||
1398 |
\begin{enumerate} |
|
1399 |
\setlength{\itemsep}{0pt} |
|
1400 |
||
1401 |
\item |
|
1402 |
Introduce a fully unspecified constant @{text "un_D\<^sub>0 \<Colon> 'a"} using |
|
1403 |
@{keyword consts}. |
|
1404 |
||
1405 |
\item |
|
53535 | 1406 |
Define the datatype, specifying @{text "un_D\<^sub>0"} as the selector's default |
1407 |
value. |
|
52827 | 1408 |
|
1409 |
\item |
|
53535 | 1410 |
Define the behavior of @{text "un_D\<^sub>0"} on values of the newly introduced |
1411 |
datatype using the \keyw{overloading} command. |
|
52827 | 1412 |
|
1413 |
\item |
|
1414 |
Derive the desired equation on @{text un_D} from the characteristic equations |
|
1415 |
for @{text "un_D\<^sub>0"}. |
|
1416 |
\end{enumerate} |
|
1417 |
||
53619 | 1418 |
\noindent |
52827 | 1419 |
The following example illustrates this procedure: |
1420 |
*} |
|
1421 |
||
1422 |
consts termi\<^sub>0 :: 'a |
|
1423 |
||
53619 | 1424 |
text {* \blankline *} |
1425 |
||
53491 | 1426 |
datatype_new ('a, 'b) tlist = |
52827 | 1427 |
TNil (termi: 'b) (defaults ttl: TNil) |
53491 | 1428 |
| TCons (thd: 'a) (ttl : "('a, 'b) tlist") (defaults termi: "\<lambda>_ xs. termi\<^sub>0 xs") |
52827 | 1429 |
|
53619 | 1430 |
text {* \blankline *} |
1431 |
||
52827 | 1432 |
overloading |
53491 | 1433 |
termi\<^sub>0 \<equiv> "termi\<^sub>0 \<Colon> ('a, 'b) tlist \<Rightarrow> 'b" |
52827 | 1434 |
begin |
53491 | 1435 |
primrec_new termi\<^sub>0 :: "('a, 'b) tlist \<Rightarrow> 'b" where |
53621 | 1436 |
"termi\<^sub>0 (TNil y) = y" | |
1437 |
"termi\<^sub>0 (TCons x xs) = termi\<^sub>0 xs" |
|
52827 | 1438 |
end |
1439 |
||
53619 | 1440 |
text {* \blankline *} |
1441 |
||
52827 | 1442 |
lemma terminal_TCons[simp]: "termi (TCons x xs) = termi xs" |
1443 |
by (cases xs) auto |
|
1444 |
||
1445 |
||
52828 | 1446 |
subsection {* Compatibility Issues |
53617 | 1447 |
\label{ssec:primrec-compatibility-issues} *} |
52828 | 1448 |
|
1449 |
text {* |
|
53997 | 1450 |
The command @{command primrec_new} has been designed to be highly compatible |
1451 |
with the old \keyw{primrec}, to ease migration. There is nonetheless at least |
|
1452 |
one incompatibility that may arise when porting to the new package: |
|
1453 |
||
1454 |
\begin{itemize} |
|
1455 |
\setlength{\itemsep}{0pt} |
|
1456 |
||
54185 | 1457 |
\item \emph{Some theorems have different names.} |
53997 | 1458 |
For $m > 1$ mutually recursive functions, |
54023
cede3c1d2417
minor doc fix (there is no guarantee that the equations for a given f_i are contiguous in the collection)
blanchet
parents:
54014
diff
changeset
|
1459 |
@{text "f\<^sub>1_\<dots>_f\<^sub>m.simps"} has been broken down into separate |
cede3c1d2417
minor doc fix (there is no guarantee that the equations for a given f_i are contiguous in the collection)
blanchet
parents:
54014
diff
changeset
|
1460 |
subcollections @{text "f\<^sub>i.simps"}. |
53997 | 1461 |
\end{itemize} |
52828 | 1462 |
*} |
52794 | 1463 |
|
1464 |
||
52827 | 1465 |
section {* Defining Codatatypes |
52805 | 1466 |
\label{sec:defining-codatatypes} *} |
1467 |
||
1468 |
text {* |
|
53829 | 1469 |
Codatatypes can be specified using the @{command codatatype} command. The |
53623 | 1470 |
command is first illustrated through concrete examples featuring different |
1471 |
flavors of corecursion. More examples can be found in the directory |
|
53997 | 1472 |
\verb|~~/src/HOL/|\allowbreak\verb|BNF/Examples|. The |
1473 |
\emph{Archive of Formal Proofs} also includes some useful codatatypes, notably |
|
1474 |
for lazy lists \cite{lochbihler-2010}. |
|
52805 | 1475 |
*} |
52792 | 1476 |
|
52824 | 1477 |
|
53617 | 1478 |
subsection {* Introductory Examples |
1479 |
\label{ssec:codatatype-introductory-examples} *} |
|
52794 | 1480 |
|
53623 | 1481 |
|
1482 |
subsubsection {* Simple Corecursion |
|
1483 |
\label{sssec:codatatype-simple-corecursion} *} |
|
1484 |
||
52805 | 1485 |
text {* |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
1486 |
Noncorecursive codatatypes coincide with the corresponding datatypes, so they |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
1487 |
are useless in practice. \emph{Corecursive codatatypes} have the same syntax |
53623 | 1488 |
as recursive datatypes, except for the command name. For example, here is the |
1489 |
definition of lazy lists: |
|
1490 |
*} |
|
1491 |
||
1492 |
codatatype (lset: 'a) llist (map: lmap rel: llist_all2) = |
|
1493 |
lnull: LNil (defaults ltl: LNil) |
|
1494 |
| LCons (lhd: 'a) (ltl: "'a llist") |
|
1495 |
||
1496 |
text {* |
|
1497 |
\noindent |
|
1498 |
Lazy lists can be infinite, such as @{text "LCons 0 (LCons 0 (\<dots>))"} and |
|
53647 | 1499 |
@{text "LCons 0 (LCons 1 (LCons 2 (\<dots>)))"}. Here is a related type, that of |
1500 |
infinite streams: |
|
1501 |
*} |
|
1502 |
||
1503 |
codatatype (sset: 'a) stream (map: smap rel: stream_all2) = |
|
1504 |
SCons (shd: 'a) (stl: "'a stream") |
|
1505 |
||
1506 |
text {* |
|
1507 |
\noindent |
|
1508 |
Another interesting type that can |
|
53623 | 1509 |
be defined as a codatatype is that of the extended natural numbers: |
1510 |
*} |
|
1511 |
||
53644 | 1512 |
codatatype enat = EZero | ESuc enat |
53623 | 1513 |
|
1514 |
text {* |
|
1515 |
\noindent |
|
1516 |
This type has exactly one infinite element, @{text "ESuc (ESuc (ESuc (\<dots>)))"}, |
|
1517 |
that represents $\infty$. In addition, it has finite values of the form |
|
1518 |
@{text "ESuc (\<dots> (ESuc EZero)\<dots>)"}. |
|
53675 | 1519 |
|
1520 |
Here is an example with many constructors: |
|
52805 | 1521 |
*} |
53623 | 1522 |
|
53675 | 1523 |
codatatype 'a process = |
1524 |
Fail |
|
1525 |
| Skip (cont: "'a process") |
|
1526 |
| Action (prefix: 'a) (cont: "'a process") |
|
1527 |
| Choice (left: "'a process") (right: "'a process") |
|
1528 |
||
53750 | 1529 |
text {* |
53829 | 1530 |
\noindent |
53750 | 1531 |
Notice that the @{const cont} selector is associated with both @{const Skip} |
54146 | 1532 |
and @{const Action}. |
53750 | 1533 |
*} |
1534 |
||
53623 | 1535 |
|
1536 |
subsubsection {* Mutual Corecursion |
|
1537 |
\label{sssec:codatatype-mutual-corecursion} *} |
|
1538 |
||
1539 |
text {* |
|
1540 |
\noindent |
|
1541 |
The example below introduces a pair of \emph{mutually corecursive} types: |
|
1542 |
*} |
|
1543 |
||
1544 |
codatatype even_enat = Even_EZero | Even_ESuc odd_enat |
|
1545 |
and odd_enat = Odd_ESuc even_enat |
|
1546 |
||
1547 |
||
1548 |
subsubsection {* Nested Corecursion |
|
1549 |
\label{sssec:codatatype-nested-corecursion} *} |
|
1550 |
||
1551 |
text {* |
|
1552 |
\noindent |
|
53675 | 1553 |
The next examples feature \emph{nested corecursion}: |
53623 | 1554 |
*} |
1555 |
||
53644 | 1556 |
codatatype 'a tree\<^sub>i\<^sub>i = Node\<^sub>i\<^sub>i (lbl\<^sub>i\<^sub>i: 'a) (sub\<^sub>i\<^sub>i: "'a tree\<^sub>i\<^sub>i llist") |
53675 | 1557 |
|
53752 | 1558 |
text {* \blankline *} |
1559 |
||
53644 | 1560 |
codatatype 'a tree\<^sub>i\<^sub>s = Node\<^sub>i\<^sub>s (lbl\<^sub>i\<^sub>s: 'a) (sub\<^sub>i\<^sub>s: "'a tree\<^sub>i\<^sub>s fset") |
52805 | 1561 |
|
53752 | 1562 |
text {* \blankline *} |
1563 |
||
55350 | 1564 |
codatatype 'a sm = SM (accept: bool) (trans: "'a \<Rightarrow> 'a sm") |
53675 | 1565 |
|
52824 | 1566 |
|
53617 | 1567 |
subsection {* Command Syntax |
1568 |
\label{ssec:codatatype-command-syntax} *} |
|
52805 | 1569 |
|
53619 | 1570 |
|
53621 | 1571 |
subsubsection {* \keyw{codatatype} |
1572 |
\label{sssec:codatatype} *} |
|
53619 | 1573 |
|
52824 | 1574 |
text {* |
53829 | 1575 |
\begin{matharray}{rcl} |
1576 |
@{command_def "codatatype"} & : & @{text "local_theory \<rightarrow> local_theory"} |
|
1577 |
\end{matharray} |
|
1578 |
||
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
1579 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
1580 |
@@{command codatatype} target? \<newline> |
53829 | 1581 |
(@{syntax dt_name} '=' (@{syntax ctor} + '|') + @'and') |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
1582 |
\<close>} |
53829 | 1583 |
|
1584 |
\noindent |
|
52827 | 1585 |
Definitions of codatatypes have almost exactly the same syntax as for datatypes |
53829 | 1586 |
(Section~\ref{ssec:datatype-command-syntax}). The @{text "no_discs_sels"} option |
1587 |
is not available, because destructors are a crucial notion for codatatypes. |
|
53623 | 1588 |
*} |
1589 |
||
1590 |
||
1591 |
subsection {* Generated Constants |
|
1592 |
\label{ssec:codatatype-generated-constants} *} |
|
1593 |
||
1594 |
text {* |
|
1595 |
Given a codatatype @{text "('a\<^sub>1, \<dots>, 'a\<^sub>m) t"} |
|
1596 |
with $m > 0$ live type variables and $n$ constructors @{text "t.C\<^sub>1"}, |
|
1597 |
\ldots, @{text "t.C\<^sub>n"}, the same auxiliary constants are generated as for |
|
1598 |
datatypes (Section~\ref{ssec:datatype-generated-constants}), except that the |
|
1599 |
iterator and the recursor are replaced by dual concepts: |
|
1600 |
||
1601 |
\begin{itemize} |
|
1602 |
\setlength{\itemsep}{0pt} |
|
1603 |
||
54494 | 1604 |
\item \relax{Coiterator}: @{text unfold_t} |
1605 |
||
1606 |
\item \relax{Corecursor}: @{text corec_t} |
|
53623 | 1607 |
|
1608 |
\end{itemize} |
|
1609 |
*} |
|
1610 |
||
1611 |
||
1612 |
subsection {* Generated Theorems |
|
1613 |
\label{ssec:codatatype-generated-theorems} *} |
|
1614 |
||
1615 |
text {* |
|
53829 | 1616 |
The characteristic theorems generated by @{command codatatype} are grouped in |
53623 | 1617 |
three broad categories: |
1618 |
||
1619 |
\begin{itemize} |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1620 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1621 |
|
53623 | 1622 |
\item The \emph{free constructor theorems} are properties about the constructors |
1623 |
and destructors that can be derived for any freely generated type. |
|
1624 |
||
1625 |
\item The \emph{functorial theorems} are properties of datatypes related to |
|
1626 |
their BNF nature. |
|
1627 |
||
1628 |
\item The \emph{coinductive theorems} are properties of datatypes related to |
|
1629 |
their coinductive nature. |
|
1630 |
\end{itemize} |
|
1631 |
||
1632 |
\noindent |
|
1633 |
The first two categories are exactly as for datatypes and are described in |
|
53642 | 1634 |
Sections |
1635 |
\ref{sssec:free-constructor-theorems}~and~\ref{sssec:functorial-theorems}. |
|
52824 | 1636 |
*} |
1637 |
||
53617 | 1638 |
|
53623 | 1639 |
subsubsection {* Coinductive Theorems |
1640 |
\label{sssec:coinductive-theorems} *} |
|
1641 |
||
1642 |
text {* |
|
54031 | 1643 |
The coinductive theorems are listed below for @{typ "'a llist"}: |
53623 | 1644 |
|
1645 |
\begin{indentblock} |
|
1646 |
\begin{description} |
|
1647 |
||
53643 | 1648 |
\item[\begin{tabular}{@ {}l@ {}} |
1649 |
@{text "t."}\hthm{coinduct} @{text "[coinduct t, consumes m, case_names t\<^sub>1 \<dots> t\<^sub>m,"} \\ |
|
1650 |
\phantom{@{text "t."}\hthm{coinduct} @{text "["}}@{text "case_conclusion D\<^sub>1 \<dots> D\<^sub>n]"}\rm: |
|
1651 |
\end{tabular}] ~ \\ |
|
53623 | 1652 |
@{thm llist.coinduct[no_vars]} |
53617 | 1653 |
|
53643 | 1654 |
\item[\begin{tabular}{@ {}l@ {}} |
1655 |
@{text "t."}\hthm{strong\_coinduct} @{text "[consumes m, case_names t\<^sub>1 \<dots> t\<^sub>m,"} \\ |
|
1656 |
\phantom{@{text "t."}\hthm{strong\_coinduct} @{text "["}}@{text "case_conclusion D\<^sub>1 \<dots> D\<^sub>n]"}\rm: |
|
1657 |
\end{tabular}] ~ \\ |
|
1658 |
@{thm llist.strong_coinduct[no_vars]} |
|
53617 | 1659 |
|
53643 | 1660 |
\item[\begin{tabular}{@ {}l@ {}} |
1661 |
@{text "t\<^sub>1_\<dots>_t\<^sub>m."}\hthm{coinduct} @{text "[case_names t\<^sub>1 \<dots> t\<^sub>m, case_conclusion D\<^sub>1 \<dots> D\<^sub>n]"} \\ |
|
1662 |
@{text "t\<^sub>1_\<dots>_t\<^sub>m."}\hthm{strong\_coinduct} @{text "[case_names t\<^sub>1 \<dots> t\<^sub>m,"} \\ |
|
1663 |
\phantom{@{text "t\<^sub>1_\<dots>_t\<^sub>m."}\hthm{strong\_coinduct} @{text "["}}@{text "case_conclusion D\<^sub>1 \<dots> D\<^sub>n]"}\rm: |
|
1664 |
\end{tabular}] ~ \\ |
|
1665 |
Given $m > 1$ mutually corecursive codatatypes, these coinduction rules can be |
|
1666 |
used to prove $m$ properties simultaneously. |
|
1667 |
||
54031 | 1668 |
\item[@{text "t."}\hthm{unfold}\rm:] ~ \\ |
53623 | 1669 |
@{thm llist.unfold(1)[no_vars]} \\ |
1670 |
@{thm llist.unfold(2)[no_vars]} |
|
1671 |
||
54031 | 1672 |
\item[@{text "t."}\hthm{corec}\rm:] ~ \\ |
53623 | 1673 |
@{thm llist.corec(1)[no_vars]} \\ |
1674 |
@{thm llist.corec(2)[no_vars]} |
|
1675 |
||
53703 | 1676 |
\item[@{text "t."}\hthm{disc\_unfold}\rm:] ~ \\ |
53643 | 1677 |
@{thm llist.disc_unfold(1)[no_vars]} \\ |
1678 |
@{thm llist.disc_unfold(2)[no_vars]} |
|
1679 |
||
53703 | 1680 |
\item[@{text "t."}\hthm{disc\_corec}\rm:] ~ \\ |
53643 | 1681 |
@{thm llist.disc_corec(1)[no_vars]} \\ |
1682 |
@{thm llist.disc_corec(2)[no_vars]} |
|
1683 |
||
1684 |
\item[@{text "t."}\hthm{disc\_unfold\_iff} @{text "[simp]"}\rm:] ~ \\ |
|
1685 |
@{thm llist.disc_unfold_iff(1)[no_vars]} \\ |
|
1686 |
@{thm llist.disc_unfold_iff(2)[no_vars]} |
|
1687 |
||
1688 |
\item[@{text "t."}\hthm{disc\_corec\_iff} @{text "[simp]"}\rm:] ~ \\ |
|
1689 |
@{thm llist.disc_corec_iff(1)[no_vars]} \\ |
|
1690 |
@{thm llist.disc_corec_iff(2)[no_vars]} |
|
1691 |
||
1692 |
\item[@{text "t."}\hthm{sel\_unfold} @{text "[simp]"}\rm:] ~ \\ |
|
1693 |
@{thm llist.sel_unfold(1)[no_vars]} \\ |
|
1694 |
@{thm llist.sel_unfold(2)[no_vars]} |
|
1695 |
||
1696 |
\item[@{text "t."}\hthm{sel\_corec} @{text "[simp]"}\rm:] ~ \\ |
|
1697 |
@{thm llist.sel_corec(1)[no_vars]} \\ |
|
1698 |
@{thm llist.sel_corec(2)[no_vars]} |
|
1699 |
||
53623 | 1700 |
\end{description} |
1701 |
\end{indentblock} |
|
1702 |
||
1703 |
\noindent |
|
53829 | 1704 |
For convenience, @{command codatatype} also provides the following collection: |
53623 | 1705 |
|
1706 |
\begin{indentblock} |
|
1707 |
\begin{description} |
|
1708 |
||
54031 | 1709 |
\item[@{text "t."}\hthm{simps} = @{text t.inject} @{text t.distinct} @{text t.case} @{text t.disc_corec} @{text t.disc_corec_iff}] ~ \\ |
1710 |
@{text t.sel_corec} @{text t.disc_unfold} @{text t.disc_unfold_iff} @{text t.sel_unfold} @{text t.map} \\ |
|
1711 |
@{text t.rel_inject} @{text t.rel_distinct} @{text t.set} |
|
53623 | 1712 |
|
1713 |
\end{description} |
|
1714 |
\end{indentblock} |
|
1715 |
*} |
|
52805 | 1716 |
|
1717 |
||
52827 | 1718 |
section {* Defining Corecursive Functions |
52805 | 1719 |
\label{sec:defining-corecursive-functions} *} |
1720 |
||
1721 |
text {* |
|
54183 | 1722 |
Corecursive functions can be specified using the @{command primcorec} and |
1723 |
\keyw{prim\-corec\-ursive} commands, which support primitive corecursion, or |
|
1724 |
using the more general \keyw{partial\_function} command. Here, the focus is on |
|
1725 |
the first two. More examples can be found in the directory |
|
55075 | 1726 |
\verb|~~/src/HOL/BNF_Examples|. |
53644 | 1727 |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1728 |
Whereas recursive functions consume datatypes one constructor at a time, |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1729 |
corecursive functions construct codatatypes one constructor at a time. |
53752 | 1730 |
Partly reflecting a lack of agreement among proponents of coalgebraic methods, |
1731 |
Isabelle supports three competing syntaxes for specifying a function $f$: |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1732 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1733 |
\begin{itemize} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1734 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1735 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1736 |
\abovedisplayskip=.5\abovedisplayskip |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1737 |
\belowdisplayskip=.5\belowdisplayskip |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1738 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1739 |
\item The \emph{destructor view} specifies $f$ by implications of the form |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1740 |
\[@{text "\<dots> \<Longrightarrow> is_C\<^sub>j (f x\<^sub>1 \<dots> x\<^sub>n)"}\] and |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1741 |
equations of the form |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1742 |
\[@{text "un_C\<^sub>ji (f x\<^sub>1 \<dots> x\<^sub>n) = \<dots>"}\] |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1743 |
This style is popular in the coalgebraic literature. |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1744 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1745 |
\item The \emph{constructor view} specifies $f$ by equations of the form |
54183 | 1746 |
\[@{text "\<dots> \<Longrightarrow> f x\<^sub>1 \<dots> x\<^sub>n = C\<^sub>j \<dots>"}\] |
53752 | 1747 |
This style is often more concise than the previous one. |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1748 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1749 |
\item The \emph{code view} specifies $f$ by a single equation of the form |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1750 |
\[@{text "f x\<^sub>1 \<dots> x\<^sub>n = \<dots>"}\] |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1751 |
with restrictions on the format of the right-hand side. Lazy functional |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1752 |
programming languages such as Haskell support a generalized version of this |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1753 |
style. |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1754 |
\end{itemize} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1755 |
|
53753
ae7f50e70c09
renamed "primcorec" to "primcorecursive", to open the door to a 'theory -> theory' command called "primcorec" (cf. "fun" vs. "function")
blanchet
parents:
53752
diff
changeset
|
1756 |
All three styles are available as input syntax. Whichever syntax is chosen, |
ae7f50e70c09
renamed "primcorec" to "primcorecursive", to open the door to a 'theory -> theory' command called "primcorec" (cf. "fun" vs. "function")
blanchet
parents:
53752
diff
changeset
|
1757 |
characteristic theorems for all three styles are generated. |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1758 |
|
52828 | 1759 |
%%% TODO: partial_function? E.g. for defining tail recursive function on lazy |
1760 |
%%% lists (cf. terminal0 in TLList.thy) |
|
52805 | 1761 |
*} |
1762 |
||
52824 | 1763 |
|
53617 | 1764 |
subsection {* Introductory Examples |
1765 |
\label{ssec:primcorec-introductory-examples} *} |
|
52805 | 1766 |
|
53646 | 1767 |
text {* |
1768 |
Primitive corecursion is illustrated through concrete examples based on the |
|
1769 |
codatatypes defined in Section~\ref{ssec:codatatype-introductory-examples}. More |
|
55075 | 1770 |
examples can be found in the directory \verb|~~/src/HOL/BNF_Examples|. The code |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1771 |
view is favored in the examples below. Sections |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1772 |
\ref{ssec:primrec-constructor-view} and \ref{ssec:primrec-destructor-view} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1773 |
present the same examples expressed using the constructor and destructor views. |
53646 | 1774 |
*} |
1775 |
||
53644 | 1776 |
subsubsection {* Simple Corecursion |
1777 |
\label{sssec:primcorec-simple-corecursion} *} |
|
1778 |
||
53646 | 1779 |
text {* |
53752 | 1780 |
Following the code view, corecursive calls are allowed on the right-hand side as |
1781 |
long as they occur under a constructor, which itself appears either directly to |
|
1782 |
the right of the equal sign or in a conditional expression: |
|
53646 | 1783 |
*} |
1784 |
||
53826 | 1785 |
primcorec literate :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a \<Rightarrow> 'a llist" where |
54072 | 1786 |
"literate g x = LCons x (literate g (g x))" |
53647 | 1787 |
|
53677 | 1788 |
text {* \blankline *} |
1789 |
||
53826 | 1790 |
primcorec siterate :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a \<Rightarrow> 'a stream" where |
54072 | 1791 |
"siterate g x = SCons x (siterate g (g x))" |
53644 | 1792 |
|
53646 | 1793 |
text {* |
1794 |
\noindent |
|
1795 |
The constructor ensures that progress is made---i.e., the function is |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1796 |
\emph{productive}. The above functions compute the infinite lazy list or stream |
54072 | 1797 |
@{text "[x, g x, g (g x), \<dots>]"}. Productivity guarantees that prefixes |
1798 |
@{text "[x, g x, g (g x), \<dots>, (g ^^ k) x]"} of arbitrary finite length |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1799 |
@{text k} can be computed by unfolding the code equation a finite number of |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
1800 |
times. |
53646 | 1801 |
|
53752 | 1802 |
Corecursive functions construct codatatype values, but nothing prevents them |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
1803 |
from also consuming such values. The following function drops every second |
53675 | 1804 |
element in a stream: |
1805 |
*} |
|
1806 |
||
53826 | 1807 |
primcorec every_snd :: "'a stream \<Rightarrow> 'a stream" where |
53675 | 1808 |
"every_snd s = SCons (shd s) (stl (stl s))" |
1809 |
||
1810 |
text {* |
|
53752 | 1811 |
\noindent |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1812 |
Constructs such as @{text "let"}---@{text "in"}, @{text |
53646 | 1813 |
"if"}---@{text "then"}---@{text "else"}, and @{text "case"}---@{text "of"} may |
1814 |
appear around constructors that guard corecursive calls: |
|
1815 |
*} |
|
1816 |
||
54072 | 1817 |
primcorec lappend :: "'a llist \<Rightarrow> 'a llist \<Rightarrow> 'a llist" where |
53644 | 1818 |
"lappend xs ys = |
1819 |
(case xs of |
|
1820 |
LNil \<Rightarrow> ys |
|
1821 |
| LCons x xs' \<Rightarrow> LCons x (lappend xs' ys))" |
|
1822 |
||
53646 | 1823 |
text {* |
53752 | 1824 |
\noindent |
54402 | 1825 |
Pattern matching is not supported by @{command primcorec}. Fortunately, it is |
1826 |
easy to generate pattern-maching equations using the \keyw{simps\_of\_case} |
|
1827 |
command provided by the theory \verb|~~/src/HOL/Library/Simps_Case_Conv|. |
|
1828 |
*} |
|
1829 |
||
1830 |
simps_of_case lappend_simps: lappend.code |
|
1831 |
||
1832 |
text {* |
|
1833 |
This generates the lemma collection @{thm [source] lappend_simps}: |
|
1834 |
% |
|
1835 |
\[@{thm lappend_simps(1)[no_vars]} |
|
1836 |
\qquad @{thm lappend_simps(2)[no_vars]}\] |
|
1837 |
% |
|
53646 | 1838 |
Corecursion is useful to specify not only functions but also infinite objects: |
1839 |
*} |
|
1840 |
||
53826 | 1841 |
primcorec infty :: enat where |
53644 | 1842 |
"infty = ESuc infty" |
1843 |
||
53646 | 1844 |
text {* |
53752 | 1845 |
\noindent |
1846 |
The example below constructs a pseudorandom process value. It takes a stream of |
|
53675 | 1847 |
actions (@{text s}), a pseudorandom function generator (@{text f}), and a |
1848 |
pseudorandom seed (@{text n}): |
|
1849 |
*} |
|
1850 |
||
54072 | 1851 |
primcorec |
53752 | 1852 |
random_process :: "'a stream \<Rightarrow> (int \<Rightarrow> int) \<Rightarrow> int \<Rightarrow> 'a process" |
1853 |
where |
|
53675 | 1854 |
"random_process s f n = |
1855 |
(if n mod 4 = 0 then |
|
1856 |
Fail |
|
1857 |
else if n mod 4 = 1 then |
|
1858 |
Skip (random_process s f (f n)) |
|
1859 |
else if n mod 4 = 2 then |
|
1860 |
Action (shd s) (random_process (stl s) f (f n)) |
|
1861 |
else |
|
1862 |
Choice (random_process (every_snd s) (f \<circ> f) (f n)) |
|
1863 |
(random_process (every_snd (stl s)) (f \<circ> f) (f (f n))))" |
|
1864 |
||
1865 |
text {* |
|
1866 |
\noindent |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1867 |
The main disadvantage of the code view is that the conditions are tested |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1868 |
sequentially. This is visible in the generated theorems. The constructor and |
53752 | 1869 |
destructor views offer nonsequential alternatives. |
53675 | 1870 |
*} |
1871 |
||
53644 | 1872 |
|
1873 |
subsubsection {* Mutual Corecursion |
|
1874 |
\label{sssec:primcorec-mutual-corecursion} *} |
|
1875 |
||
53647 | 1876 |
text {* |
1877 |
The syntax for mutually corecursive functions over mutually corecursive |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
1878 |
datatypes is unsurprising: |
53647 | 1879 |
*} |
1880 |
||
53826 | 1881 |
primcorec |
53644 | 1882 |
even_infty :: even_enat and |
1883 |
odd_infty :: odd_enat |
|
1884 |
where |
|
1885 |
"even_infty = Even_ESuc odd_infty" | |
|
1886 |
"odd_infty = Odd_ESuc even_infty" |
|
1887 |
||
1888 |
||
1889 |
subsubsection {* Nested Corecursion |
|
1890 |
\label{sssec:primcorec-nested-corecursion} *} |
|
1891 |
||
53647 | 1892 |
text {* |
1893 |
The next pair of examples generalize the @{const literate} and @{const siterate} |
|
1894 |
functions (Section~\ref{sssec:primcorec-nested-corecursion}) to possibly |
|
1895 |
infinite trees in which subnodes are organized either as a lazy list (@{text |
|
54072 | 1896 |
tree\<^sub>i\<^sub>i}) or as a finite set (@{text tree\<^sub>i\<^sub>s}). They rely on the map functions of |
1897 |
the nesting type constructors to lift the corecursive calls: |
|
53647 | 1898 |
*} |
1899 |
||
53826 | 1900 |
primcorec iterate\<^sub>i\<^sub>i :: "('a \<Rightarrow> 'a llist) \<Rightarrow> 'a \<Rightarrow> 'a tree\<^sub>i\<^sub>i" where |
54072 | 1901 |
"iterate\<^sub>i\<^sub>i g x = Node\<^sub>i\<^sub>i x (lmap (iterate\<^sub>i\<^sub>i g) (g x))" |
53644 | 1902 |
|
53677 | 1903 |
text {* \blankline *} |
1904 |
||
53826 | 1905 |
primcorec iterate\<^sub>i\<^sub>s :: "('a \<Rightarrow> 'a fset) \<Rightarrow> 'a \<Rightarrow> 'a tree\<^sub>i\<^sub>s" where |
54072 | 1906 |
"iterate\<^sub>i\<^sub>s g x = Node\<^sub>i\<^sub>s x (fimage (iterate\<^sub>i\<^sub>s g) (g x))" |
53644 | 1907 |
|
52805 | 1908 |
text {* |
53752 | 1909 |
\noindent |
54072 | 1910 |
Both examples follow the usual format for constructor arguments associated |
1911 |
with nested recursive occurrences of the datatype. Consider |
|
1912 |
@{const iterate\<^sub>i\<^sub>i}. The term @{term "g x"} constructs an @{typ "'a llist"} |
|
1913 |
value, which is turned into an @{typ "'a tree\<^sub>i\<^sub>i llist"} value using |
|
1914 |
@{const lmap}. |
|
1915 |
||
1916 |
This format may sometimes feel artificial. The following function constructs |
|
1917 |
a tree with a single, infinite branch from a stream: |
|
1918 |
*} |
|
1919 |
||
1920 |
primcorec tree\<^sub>i\<^sub>i_of_stream :: "'a stream \<Rightarrow> 'a tree\<^sub>i\<^sub>i" where |
|
1921 |
"tree\<^sub>i\<^sub>i_of_stream s = |
|
1922 |
Node\<^sub>i\<^sub>i (shd s) (lmap tree\<^sub>i\<^sub>i_of_stream (LCons (stl s) LNil))" |
|
1923 |
||
1924 |
text {* |
|
1925 |
\noindent |
|
54955
cf8d429dc24e
reintroduce recursive calls under constructors, taken out in 8dd0e0316881 mainly and in subsequent changes
blanchet
parents:
54832
diff
changeset
|
1926 |
A more natural syntax, also supported by Isabelle, is to move corecursive calls |
cf8d429dc24e
reintroduce recursive calls under constructors, taken out in 8dd0e0316881 mainly and in subsequent changes
blanchet
parents:
54832
diff
changeset
|
1927 |
under constructors: |
54072 | 1928 |
*} |
1929 |
||
54955
cf8d429dc24e
reintroduce recursive calls under constructors, taken out in 8dd0e0316881 mainly and in subsequent changes
blanchet
parents:
54832
diff
changeset
|
1930 |
primcorec (*<*)(in late) (*>*)tree\<^sub>i\<^sub>i_of_stream :: "'a stream \<Rightarrow> 'a tree\<^sub>i\<^sub>i" where |
55350 | 1931 |
"tree\<^sub>i\<^sub>i_of_stream s = |
1932 |
Node\<^sub>i\<^sub>i (shd s) (LCons (tree\<^sub>i\<^sub>i_of_stream (stl s)) LNil)" |
|
54072 | 1933 |
|
1934 |
text {* |
|
1935 |
The next example illustrates corecursion through functions, which is a bit |
|
1936 |
special. Deterministic finite automata (DFAs) are traditionally defined as |
|
1937 |
5-tuples @{text "(Q, \<Sigma>, \<delta>, q\<^sub>0, F)"}, where @{text Q} is a finite set of states, |
|
53675 | 1938 |
@{text \<Sigma>} is a finite alphabet, @{text \<delta>} is a transition function, @{text q\<^sub>0} |
1939 |
is an initial state, and @{text F} is a set of final states. The following |
|
55350 | 1940 |
function translates a DFA into a state machine: |
53675 | 1941 |
*} |
1942 |
||
55350 | 1943 |
primcorec (*<*)(in early) (*>*)sm_of_dfa :: "('q \<Rightarrow> 'a \<Rightarrow> 'q) \<Rightarrow> 'q set \<Rightarrow> 'q \<Rightarrow> 'a sm" where |
1944 |
"sm_of_dfa \<delta> F q = SM (q \<in> F) (sm_of_dfa \<delta> F \<circ> \<delta> q)" |
|
53675 | 1945 |
|
53751 | 1946 |
text {* |
1947 |
\noindent |
|
1948 |
The map function for the function type (@{text \<Rightarrow>}) is composition |
|
54181 | 1949 |
(@{text "op \<circ>"}). For convenience, corecursion through functions can |
54182 | 1950 |
also be expressed using $\lambda$-abstractions and function application rather |
54031 | 1951 |
than through composition. For example: |
53751 | 1952 |
*} |
1953 |
||
55350 | 1954 |
primcorec sm_of_dfa :: "('q \<Rightarrow> 'a \<Rightarrow> 'q) \<Rightarrow> 'q set \<Rightarrow> 'q \<Rightarrow> 'a sm" where |
1955 |
"sm_of_dfa \<delta> F q = SM (q \<in> F) (\<lambda>a. sm_of_dfa \<delta> F (\<delta> q a))" |
|
53752 | 1956 |
|
1957 |
text {* \blankline *} |
|
1958 |
||
55350 | 1959 |
primcorec empty_sm :: "'a sm" where |
1960 |
"empty_sm = SM False (\<lambda>_. empty_sm)" |
|
53751 | 1961 |
|
53752 | 1962 |
text {* \blankline *} |
1963 |
||
55350 | 1964 |
primcorec not_sm :: "'a sm \<Rightarrow> 'a sm" where |
1965 |
"not_sm M = SM (\<not> accept M) (\<lambda>a. not_sm (trans M a))" |
|
53751 | 1966 |
|
53752 | 1967 |
text {* \blankline *} |
1968 |
||
55350 | 1969 |
primcorec or_sm :: "'a sm \<Rightarrow> 'a sm \<Rightarrow> 'a sm" where |
1970 |
"or_sm M N = |
|
1971 |
SM (accept M \<or> accept N) (\<lambda>a. or_sm (trans M a) (trans N a))" |
|
53751 | 1972 |
|
54182 | 1973 |
text {* |
1974 |
\noindent |
|
1975 |
For recursion through curried $n$-ary functions, $n$ applications of |
|
1976 |
@{term "op \<circ>"} are necessary. The examples below illustrate the case where |
|
1977 |
$n = 2$: |
|
1978 |
*} |
|
1979 |
||
55350 | 1980 |
codatatype ('a, 'b) sm2 = |
1981 |
SM2 (accept2: bool) (trans2: "'a \<Rightarrow> 'b \<Rightarrow> ('a, 'b) sm2") |
|
54182 | 1982 |
|
1983 |
text {* \blankline *} |
|
1984 |
||
1985 |
primcorec |
|
55350 | 1986 |
(*<*)(in early) (*>*)sm2_of_dfa :: "('q \<Rightarrow> 'a \<Rightarrow> 'b \<Rightarrow> 'q) \<Rightarrow> 'q set \<Rightarrow> 'q \<Rightarrow> ('a, 'b) sm2" |
54182 | 1987 |
where |
55350 | 1988 |
"sm2_of_dfa \<delta> F q = SM2 (q \<in> F) (op \<circ> (op \<circ> (sm2_of_dfa \<delta> F)) (\<delta> q))" |
54182 | 1989 |
|
1990 |
text {* \blankline *} |
|
1991 |
||
1992 |
primcorec |
|
55350 | 1993 |
sm2_of_dfa :: "('q \<Rightarrow> 'a \<Rightarrow> 'b \<Rightarrow> 'q) \<Rightarrow> 'q set \<Rightarrow> 'q \<Rightarrow> ('a, 'b) sm2" |
54182 | 1994 |
where |
55350 | 1995 |
"sm2_of_dfa \<delta> F q = SM2 (q \<in> F) (\<lambda>a b. sm2_of_dfa \<delta> F (\<delta> q a b))" |
54182 | 1996 |
|
53644 | 1997 |
|
1998 |
subsubsection {* Nested-as-Mutual Corecursion |
|
1999 |
\label{sssec:primcorec-nested-as-mutual-corecursion} *} |
|
2000 |
||
53647 | 2001 |
text {* |
2002 |
Just as it is possible to recurse over nested recursive datatypes as if they |
|
2003 |
were mutually recursive |
|
2004 |
(Section~\ref{sssec:primrec-nested-as-mutual-recursion}), it is possible to |
|
53752 | 2005 |
pretend that nested codatatypes are mutually corecursive. For example: |
53647 | 2006 |
*} |
2007 |
||
54287 | 2008 |
(*<*) |
2009 |
context late |
|
2010 |
begin |
|
2011 |
(*>*) |
|
54072 | 2012 |
primcorec |
54287 | 2013 |
iterate\<^sub>i\<^sub>i :: "('a \<Rightarrow> 'a llist) \<Rightarrow> 'a \<Rightarrow> 'a tree\<^sub>i\<^sub>i" and |
53644 | 2014 |
iterates\<^sub>i\<^sub>i :: "('a \<Rightarrow> 'a llist) \<Rightarrow> 'a llist \<Rightarrow> 'a tree\<^sub>i\<^sub>i llist" |
2015 |
where |
|
54072 | 2016 |
"iterate\<^sub>i\<^sub>i g x = Node\<^sub>i\<^sub>i x (iterates\<^sub>i\<^sub>i g (g x))" | |
2017 |
"iterates\<^sub>i\<^sub>i g xs = |
|
53644 | 2018 |
(case xs of |
2019 |
LNil \<Rightarrow> LNil |
|
54072 | 2020 |
| LCons x xs' \<Rightarrow> LCons (iterate\<^sub>i\<^sub>i g x) (iterates\<^sub>i\<^sub>i g xs'))" |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2021 |
|
54287 | 2022 |
text {* |
2023 |
\noindent |
|
2024 |
Coinduction rules are generated as |
|
2025 |
@{thm [source] iterate\<^sub>i\<^sub>i.coinduct}, |
|
2026 |
@{thm [source] iterates\<^sub>i\<^sub>i.coinduct}, and |
|
2027 |
@{thm [source] iterate\<^sub>i\<^sub>i_iterates\<^sub>i\<^sub>i.coinduct} |
|
2028 |
and analogously for @{text strong_coinduct}. These rules and the |
|
2029 |
underlying corecursors are generated on a per-need basis and are kept in a cache |
|
2030 |
to speed up subsequent definitions. |
|
2031 |
*} |
|
2032 |
||
2033 |
(*<*) |
|
2034 |
end |
|
2035 |
(*>*) |
|
2036 |
||
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2037 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2038 |
subsubsection {* Constructor View |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2039 |
\label{ssec:primrec-constructor-view} *} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2040 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2041 |
(*<*) |
54182 | 2042 |
locale ctr_view |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2043 |
begin |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2044 |
(*>*) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2045 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2046 |
text {* |
53750 | 2047 |
The constructor view is similar to the code view, but there is one separate |
2048 |
conditional equation per constructor rather than a single unconditional |
|
2049 |
equation. Examples that rely on a single constructor, such as @{const literate} |
|
2050 |
and @{const siterate}, are identical in both styles. |
|
2051 |
||
2052 |
Here is an example where there is a difference: |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2053 |
*} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2054 |
|
53826 | 2055 |
primcorec lappend :: "'a llist \<Rightarrow> 'a llist \<Rightarrow> 'a llist" where |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2056 |
"lnull xs \<Longrightarrow> lnull ys \<Longrightarrow> lappend xs ys = LNil" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2057 |
"_ \<Longrightarrow> lappend xs ys = LCons (lhd (if lnull xs then ys else xs)) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2058 |
(if xs = LNil then ltl ys else lappend (ltl xs) ys)" |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2059 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2060 |
text {* |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2061 |
\noindent |
53752 | 2062 |
With the constructor view, we must distinguish between the @{const LNil} and |
2063 |
the @{const LCons} case. The condition for @{const LCons} is |
|
2064 |
left implicit, as the negation of that for @{const LNil}. |
|
53750 | 2065 |
|
2066 |
For this example, the constructor view is slighlty more involved than the |
|
2067 |
code equation. Recall the code view version presented in |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2068 |
Section~\ref{sssec:primcorec-simple-corecursion}. |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2069 |
% TODO: \[{thm code_view.lappend.code}\] |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2070 |
The constructor view requires us to analyze the second argument (@{term ys}). |
53752 | 2071 |
The code equation generated from the constructor view also suffers from this. |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2072 |
% TODO: \[{thm lappend.code}\] |
53750 | 2073 |
|
53752 | 2074 |
In contrast, the next example is arguably more naturally expressed in the |
2075 |
constructor view: |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2076 |
*} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2077 |
|
53831
80423b9080cf
support "of" syntax to disambiguate selector equations
panny
parents:
53829
diff
changeset
|
2078 |
primcorec |
53752 | 2079 |
random_process :: "'a stream \<Rightarrow> (int \<Rightarrow> int) \<Rightarrow> int \<Rightarrow> 'a process" |
2080 |
where |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2081 |
"n mod 4 = 0 \<Longrightarrow> random_process s f n = Fail" | |
53752 | 2082 |
"n mod 4 = 1 \<Longrightarrow> |
2083 |
random_process s f n = Skip (random_process s f (f n))" | |
|
2084 |
"n mod 4 = 2 \<Longrightarrow> |
|
2085 |
random_process s f n = Action (shd s) (random_process (stl s) f (f n))" | |
|
2086 |
"n mod 4 = 3 \<Longrightarrow> |
|
2087 |
random_process s f n = Choice (random_process (every_snd s) f (f n)) |
|
53826 | 2088 |
(random_process (every_snd (stl s)) f (f n))" |
2089 |
(*<*) |
|
53644 | 2090 |
end |
2091 |
(*>*) |
|
52805 | 2092 |
|
53750 | 2093 |
text {* |
53752 | 2094 |
\noindent |
53750 | 2095 |
Since there is no sequentiality, we can apply the equation for @{const Choice} |
53752 | 2096 |
without having first to discharge @{term "n mod (4\<Colon>int) \<noteq> 0"}, |
2097 |
@{term "n mod (4\<Colon>int) \<noteq> 1"}, and |
|
2098 |
@{term "n mod (4\<Colon>int) \<noteq> 2"}. |
|
53750 | 2099 |
The price to pay for this elegance is that we must discharge exclusivity proof |
2100 |
obligations, one for each pair of conditions |
|
53752 | 2101 |
@{term "(n mod (4\<Colon>int) = i, n mod (4\<Colon>int) = j)"} |
2102 |
with @{term "i < j"}. If we prefer not to discharge any obligations, we can |
|
2103 |
enable the @{text "sequential"} option. This pushes the problem to the users of |
|
2104 |
the generated properties. |
|
53750 | 2105 |
%Here are more examples to conclude: |
2106 |
*} |
|
2107 |
||
52824 | 2108 |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2109 |
subsubsection {* Destructor View |
53752 | 2110 |
\label{ssec:primrec-destructor-view} *} |
2111 |
||
2112 |
(*<*) |
|
54182 | 2113 |
locale dtr_view |
53752 | 2114 |
begin |
2115 |
(*>*) |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2116 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2117 |
text {* |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2118 |
The destructor view is in many respects dual to the constructor view. Conditions |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2119 |
determine which constructor to choose, and these conditions are interpreted |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2120 |
sequentially or not depending on the @{text "sequential"} option. |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2121 |
Consider the following examples: |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2122 |
*} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2123 |
|
53826 | 2124 |
primcorec literate :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a \<Rightarrow> 'a llist" where |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2125 |
"\<not> lnull (literate _ x)" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2126 |
"lhd (literate _ x) = x" | |
54072 | 2127 |
"ltl (literate g x) = literate g (g x)" |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2128 |
|
53752 | 2129 |
text {* \blankline *} |
2130 |
||
53826 | 2131 |
primcorec siterate :: "('a \<Rightarrow> 'a) \<Rightarrow> 'a \<Rightarrow> 'a stream" where |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2132 |
"shd (siterate _ x) = x" | |
54072 | 2133 |
"stl (siterate g x) = siterate g (g x)" |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2134 |
|
53752 | 2135 |
text {* \blankline *} |
2136 |
||
53826 | 2137 |
primcorec every_snd :: "'a stream \<Rightarrow> 'a stream" where |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2138 |
"shd (every_snd s) = shd s" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2139 |
"stl (every_snd s) = stl (stl s)" |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2140 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2141 |
text {* |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2142 |
\noindent |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2143 |
The first formula in the @{const literate} specification indicates which |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2144 |
constructor to choose. For @{const siterate} and @{const every_snd}, no such |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2145 |
formula is necessary, since the type has only one constructor. The last two |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2146 |
formulas are equations specifying the value of the result for the relevant |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2147 |
selectors. Corecursive calls appear directly to the right of the equal sign. |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2148 |
Their arguments are unrestricted. |
53750 | 2149 |
|
2150 |
The next example shows how to specify functions that rely on more than one |
|
2151 |
constructor: |
|
2152 |
*} |
|
2153 |
||
53826 | 2154 |
primcorec lappend :: "'a llist \<Rightarrow> 'a llist \<Rightarrow> 'a llist" where |
53750 | 2155 |
"lnull xs \<Longrightarrow> lnull ys \<Longrightarrow> lnull (lappend xs ys)" | |
2156 |
"lhd (lappend xs ys) = lhd (if lnull xs then ys else xs)" | |
|
2157 |
"ltl (lappend xs ys) = (if xs = LNil then ltl ys else lappend (ltl xs) ys)" |
|
2158 |
||
2159 |
text {* |
|
2160 |
\noindent |
|
2161 |
For a codatatype with $n$ constructors, it is sufficient to specify $n - 1$ |
|
2162 |
discriminator formulas. The command will then assume that the remaining |
|
2163 |
constructor should be taken otherwise. This can be made explicit by adding |
|
2164 |
*} |
|
2165 |
||
2166 |
(*<*) |
|
2167 |
end |
|
2168 |
||
54182 | 2169 |
locale dtr_view2 |
2170 |
begin |
|
2171 |
||
53826 | 2172 |
primcorec lappend :: "'a llist \<Rightarrow> 'a llist \<Rightarrow> 'a llist" where |
53750 | 2173 |
"lnull xs \<Longrightarrow> lnull ys \<Longrightarrow> lnull (lappend xs ys)" | |
55350 | 2174 |
"lhd (lappend xs ys) = lhd (if lnull xs then ys else xs)" | |
2175 |
"ltl (lappend xs ys) = (if xs = LNil then ltl ys else lappend (ltl xs) ys)" | |
|
53750 | 2176 |
(*>*) |
53752 | 2177 |
"_ \<Longrightarrow> \<not> lnull (lappend xs ys)" |
53750 | 2178 |
|
2179 |
text {* |
|
2180 |
\noindent |
|
53752 | 2181 |
to the specification. The generated selector theorems are conditional. |
2182 |
||
2183 |
The next example illustrates how to cope with selectors defined for several |
|
53750 | 2184 |
constructors: |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2185 |
*} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2186 |
|
53831
80423b9080cf
support "of" syntax to disambiguate selector equations
panny
parents:
53829
diff
changeset
|
2187 |
primcorec |
53752 | 2188 |
random_process :: "'a stream \<Rightarrow> (int \<Rightarrow> int) \<Rightarrow> int \<Rightarrow> 'a process" |
2189 |
where |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2190 |
"n mod 4 = 0 \<Longrightarrow> is_Fail (random_process s f n)" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2191 |
"n mod 4 = 1 \<Longrightarrow> is_Skip (random_process s f n)" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2192 |
"n mod 4 = 2 \<Longrightarrow> is_Action (random_process s f n)" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2193 |
"n mod 4 = 3 \<Longrightarrow> is_Choice (random_process s f n)" | |
53831
80423b9080cf
support "of" syntax to disambiguate selector equations
panny
parents:
53829
diff
changeset
|
2194 |
"cont (random_process s f n) = random_process s f (f n)" of Skip | |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2195 |
"prefix (random_process s f n) = shd s" | |
53831
80423b9080cf
support "of" syntax to disambiguate selector equations
panny
parents:
53829
diff
changeset
|
2196 |
"cont (random_process s f n) = random_process (stl s) f (f n)" of Action | |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2197 |
"left (random_process s f n) = random_process (every_snd s) f (f n)" | |
53831
80423b9080cf
support "of" syntax to disambiguate selector equations
panny
parents:
53829
diff
changeset
|
2198 |
"right (random_process s f n) = random_process (every_snd (stl s)) f (f n)" |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2199 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2200 |
text {* |
53750 | 2201 |
\noindent |
2202 |
Using the @{text "of"} keyword, different equations are specified for @{const |
|
2203 |
cont} depending on which constructor is selected. |
|
2204 |
||
2205 |
Here are more examples to conclude: |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2206 |
*} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2207 |
|
53826 | 2208 |
primcorec |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2209 |
even_infty :: even_enat and |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2210 |
odd_infty :: odd_enat |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2211 |
where |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2212 |
"\<not> is_Even_EZero even_infty" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2213 |
"un_Even_ESuc even_infty = odd_infty" | |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2214 |
"un_Odd_ESuc odd_infty = even_infty" |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2215 |
|
53752 | 2216 |
text {* \blankline *} |
2217 |
||
53826 | 2218 |
primcorec iterate\<^sub>i\<^sub>i :: "('a \<Rightarrow> 'a llist) \<Rightarrow> 'a \<Rightarrow> 'a tree\<^sub>i\<^sub>i" where |
54072 | 2219 |
"lbl\<^sub>i\<^sub>i (iterate\<^sub>i\<^sub>i g x) = x" | |
2220 |
"sub\<^sub>i\<^sub>i (iterate\<^sub>i\<^sub>i g x) = lmap (iterate\<^sub>i\<^sub>i g) (g x)" |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2221 |
(*<*) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2222 |
end |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2223 |
(*>*) |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2224 |
|
53750 | 2225 |
|
53617 | 2226 |
subsection {* Command Syntax |
2227 |
\label{ssec:primcorec-command-syntax} *} |
|
2228 |
||
2229 |
||
53826 | 2230 |
subsubsection {* \keyw{primcorec} and \keyw{primcorecursive} |
53753
ae7f50e70c09
renamed "primcorec" to "primcorecursive", to open the door to a 'theory -> theory' command called "primcorec" (cf. "fun" vs. "function")
blanchet
parents:
53752
diff
changeset
|
2231 |
\label{sssec:primcorecursive-and-primcorec} *} |
52840 | 2232 |
|
2233 |
text {* |
|
53829 | 2234 |
\begin{matharray}{rcl} |
2235 |
@{command_def "primcorec"} & : & @{text "local_theory \<rightarrow> local_theory"} \\ |
|
2236 |
@{command_def "primcorecursive"} & : & @{text "local_theory \<rightarrow> proof(prove)"} |
|
2237 |
\end{matharray} |
|
52840 | 2238 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2239 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
2240 |
(@@{command primcorec} | @@{command primcorecursive}) target? \<newline> |
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
2241 |
@{syntax pcr_option}? fixes @'where' |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2242 |
(@{syntax pcr_formula} + '|') |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2243 |
; |
53828 | 2244 |
@{syntax_def pcr_option}: '(' ('sequential' | 'exhaustive') ')' |
52840 | 2245 |
; |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2246 |
@{syntax_def pcr_formula}: thmdecl? prop (@'of' (term * ))? |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2247 |
\<close>} |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2248 |
|
54832 | 2249 |
The optional target is potentially followed by a corecursion-specific option: |
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2250 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2251 |
\begin{itemize} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2252 |
\setlength{\itemsep}{0pt} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2253 |
|
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2254 |
\item |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2255 |
The @{text "sequential"} option indicates that the conditions in specifications |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2256 |
expressed using the constructor or destructor view are to be interpreted |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2257 |
sequentially. |
53826 | 2258 |
|
2259 |
\item |
|
2260 |
The @{text "exhaustive"} option indicates that the conditions in specifications |
|
2261 |
expressed using the constructor or destructor view cover all possible cases. |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2262 |
\end{itemize} |
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2263 |
|
53826 | 2264 |
\noindent |
2265 |
The @{command primcorec} command is an abbreviation for @{command primcorecursive} with |
|
2266 |
@{text "by auto?"} to discharge any emerging proof obligations. |
|
52840 | 2267 |
*} |
52794 | 2268 |
|
52824 | 2269 |
|
53619 | 2270 |
(* |
52840 | 2271 |
subsection {* Generated Theorems |
2272 |
\label{ssec:primcorec-generated-theorems} *} |
|
53619 | 2273 |
*) |
52794 | 2274 |
|
2275 |
||
53623 | 2276 |
(* |
2277 |
subsection {* Recursive Default Values for Selectors |
|
2278 |
\label{ssec:primcorec-recursive-default-values-for-selectors} *} |
|
2279 |
||
2280 |
text {* |
|
2281 |
partial_function to the rescue |
|
2282 |
*} |
|
2283 |
*) |
|
2284 |
||
2285 |
||
52827 | 2286 |
section {* Registering Bounded Natural Functors |
52805 | 2287 |
\label{sec:registering-bounded-natural-functors} *} |
52792 | 2288 |
|
52805 | 2289 |
text {* |
53647 | 2290 |
The (co)datatype package can be set up to allow nested recursion through |
55350 | 2291 |
arbitrary type constructors, as long as they adhere to the BNF requirements |
2292 |
and are registered as BNFs. It is also possible to declare a BNF abstractly |
|
2293 |
without specifying its internal structure. |
|
52805 | 2294 |
*} |
2295 |
||
52824 | 2296 |
|
55350 | 2297 |
subsection {* Bounded Natural Functors |
2298 |
\label{ssec:bounded-natural-functors} *} |
|
2299 |
||
2300 |
text {* |
|
2301 |
Bounded natural functors (BNFs) are a semantic criterion for where |
|
2302 |
(co)re\-cur\-sion may appear on the right-hand side of an equation |
|
2303 |
\cite{traytel-et-al-2012,blanchette-et-al-wit}. |
|
2304 |
||
2305 |
An $n$-ary BNF is a type constructor equipped with a map function |
|
2306 |
(functorial action), $n$ set functions (natural transformations), |
|
2307 |
and an infinite cardinal bound that satisfy certain properties. |
|
2308 |
For example, @{typ "'a llist"} is a unary BNF. |
|
2309 |
Its relator @{text "llist_all2 \<Colon> |
|
2310 |
('a \<Rightarrow> 'b \<Rightarrow> bool) \<Rightarrow> |
|
2311 |
'a llist \<Rightarrow> 'b llist \<Rightarrow> bool"} |
|
2312 |
extends binary predicates over elements to binary predicates over parallel |
|
2313 |
lazy lists. The cardinal bound limits the number of elements returned by the |
|
2314 |
set function; it may not depend on the cardinality of @{typ 'a}. |
|
2315 |
||
2316 |
The type constructors introduced by @{command datatype_new} and |
|
2317 |
@{command codatatype} are automatically registered as BNFs. In addition, a |
|
2318 |
number of old-style datatypes and non-free types are preregistered. |
|
2319 |
||
2320 |
Given an $n$-ary BNF, the $n$ type variables associated with set functions, |
|
2321 |
and on which the map function acts, are \emph{live}; any other variables are |
|
2322 |
\emph{dead}. Nested (co)recursion can only take place through live variables. |
|
2323 |
*} |
|
2324 |
||
2325 |
||
2326 |
subsection {* Introductory Examples |
|
2327 |
\label{ssec:bnf-introductory-examples} *} |
|
52805 | 2328 |
|
2329 |
text {* |
|
55350 | 2330 |
The example below shows how to register a type as a BNF using the @{command bnf} |
2331 |
command. Some of the proof obligations are best viewed with the theory |
|
2332 |
@{theory Cardinal_Notations}, located in \verb|~~/src/HOL/Library|, |
|
2333 |
imported. |
|
2334 |
||
2335 |
The type is simply a copy of the function space @{typ "'d \<Rightarrow> 'a"}, where @{typ 'a} |
|
2336 |
is live and @{typ 'd} is dead. We introduce it together with its map function, |
|
2337 |
set function, and relator. |
|
52805 | 2338 |
*} |
55350 | 2339 |
|
2340 |
typedef ('d, 'a) fn = "UNIV \<Colon> ('d \<Rightarrow> 'a) set" |
|
2341 |
by simp |
|
2342 |
||
2343 |
text {* \blankline *} |
|
2344 |
||
2345 |
lemmas Abs_Rep_thms[simp] = |
|
2346 |
Abs_fn_inverse[OF UNIV_I] Rep_fn_inverse |
|
2347 |
||
2348 |
text {* \blankline *} |
|
2349 |
||
2350 |
definition map_fn :: "('a \<Rightarrow> 'b) \<Rightarrow> ('d, 'a) fn \<Rightarrow> ('d, 'b) fn" where |
|
2351 |
"map_fn f F = Abs_fn (\<lambda>x. f (Rep_fn F x))" |
|
2352 |
||
2353 |
text {* \blankline *} |
|
2354 |
||
2355 |
definition set_fn :: "('d, 'a) fn \<Rightarrow> 'a set" where |
|
2356 |
"set_fn F = range (Rep_fn F)" |
|
2357 |
||
2358 |
text {* \blankline *} |
|
2359 |
||
2360 |
definition |
|
2361 |
rel_fn :: "('a \<Rightarrow> 'b \<Rightarrow> bool) \<Rightarrow> ('d, 'a) fn \<Rightarrow> ('d, 'b) fn \<Rightarrow> bool" |
|
2362 |
where |
|
2363 |
"rel_fn R F G = fun_rel (op =) R (Rep_fn F) (Rep_fn G)" |
|
2364 |
||
2365 |
text {* \blankline *} |
|
2366 |
||
2367 |
axiomatization where cheat: "P" |
|
2368 |
||
2369 |
text {* \blankline *} |
|
2370 |
||
2371 |
bnf "('d, 'a) fn" |
|
2372 |
map: map_fn |
|
2373 |
sets: set_fn |
|
2374 |
bd: "natLeq +c |UNIV :: 'd set|" |
|
2375 |
rel: rel_fn |
|
2376 |
proof - |
|
2377 |
show "map_fn id = id" |
|
2378 |
by (auto simp add: map_fn_def[abs_def] id_comp) |
|
2379 |
next |
|
2380 |
fix F G show "map_fn (G \<circ> F) = map_fn G \<circ> map_fn F" |
|
2381 |
by (simp add: map_fn_def[abs_def] comp_def[abs_def]) |
|
2382 |
next |
|
2383 |
fix F f g |
|
2384 |
assume "\<And>x. x \<in> set_fn F \<Longrightarrow> f x = g x" |
|
2385 |
thus "map_fn f F = map_fn g F" |
|
2386 |
by (auto simp add: map_fn_def set_fn_def) |
|
2387 |
next |
|
2388 |
fix f show "set_fn \<circ> map_fn f = op ` f \<circ> set_fn" |
|
2389 |
by (auto simp add: set_fn_def map_fn_def comp_def) |
|
2390 |
next |
|
2391 |
show "card_order (natLeq +c |UNIV \<Colon> 'd set| )" |
|
2392 |
apply (rule card_order_csum) |
|
2393 |
apply (rule natLeq_card_order) |
|
2394 |
by (rule card_of_card_order_on) |
|
2395 |
next |
|
2396 |
show "cinfinite (natLeq +c |UNIV \<Colon> 'd set| )" |
|
2397 |
apply (rule cinfinite_csum) |
|
2398 |
apply (rule disjI1) |
|
2399 |
by (rule natLeq_cinfinite) |
|
2400 |
next |
|
2401 |
fix F :: "('d, 'a) fn" |
|
2402 |
have "|set_fn F| \<le>o |UNIV \<Colon> 'd set|" (is "_ \<le>o ?U") |
|
2403 |
unfolding set_fn_def by (rule card_of_image) |
|
2404 |
also have "?U \<le>o natLeq +c ?U" |
|
2405 |
by (rule ordLeq_csum2) (rule card_of_Card_order) |
|
2406 |
finally show "|set_fn F| \<le>o natLeq +c |UNIV \<Colon> 'd set|" . |
|
2407 |
next |
|
2408 |
fix R S |
|
2409 |
show "rel_fn R OO rel_fn S \<le> rel_fn (R OO S)" |
|
2410 |
by (auto simp add: rel_fn_def[abs_def] fun_rel_def) |
|
2411 |
next |
|
2412 |
fix R |
|
2413 |
show "rel_fn R = |
|
2414 |
(BNF_Util.Grp {x. set_fn x \<subseteq> Collect (split R)} (map_fn fst))\<inverse>\<inverse> OO |
|
2415 |
BNF_Util.Grp {x. set_fn x \<subseteq> Collect (split R)} (map_fn snd)" |
|
2416 |
unfolding set_fn_def rel_fn_def[abs_def] fun_rel_def Grp_def |
|
2417 |
fun_eq_iff relcompp.simps conversep.simps subset_iff image_iff |
|
2418 |
by (rule cheat) |
|
2419 |
qed |
|
2420 |
||
2421 |
text {* \blankline *} |
|
2422 |
||
2423 |
print_theorems |
|
2424 |
print_bnfs |
|
2425 |
||
2426 |
text {* |
|
2427 |
\noindent |
|
2428 |
Using \keyw{print\_theorems} and @{keyword print_bnfs}, we can contemplate and |
|
2429 |
show the world what we have achieved. |
|
2430 |
||
2431 |
This particular example does not need any nonemptiness witness, because the |
|
2432 |
one generated by default is good enough, but in general this would be |
|
2433 |
necessary. See \verb|~~/src/HOL/Basic_BNFs.thy|, |
|
2434 |
\verb|~~/src/HOL/Library/FSet.thy|, and \verb|~~/src/HOL/Library/Multiset.thy| |
|
2435 |
for further examples of BNF registration, some of which feature custom |
|
2436 |
witnesses. |
|
2437 |
||
2438 |
The next example declares a BNF axiomatically. The @{command bnf_decl} command |
|
2439 |
introduces a type @{text "('a, 'b, 'c) F"}, three set constants, a map |
|
2440 |
function, a relator, and a nonemptiness witness that depends only on |
|
2441 |
@{typ 'a}. (The type @{text "'a \<Rightarrow> ('a, 'b, 'c) F"} of |
|
2442 |
the witness can be read as an implication: If we have a witness for @{typ 'a}, |
|
2443 |
we can construct a witness for @{text "('a, 'b, 'c) F"}.) The BNF |
|
2444 |
properties are postulated as axioms. |
|
2445 |
*} |
|
2446 |
||
2447 |
bnf_decl (setA: 'a, setB: 'b, setC: 'c) F [wits: "'a \<Rightarrow> ('a, 'b, 'c) F"] |
|
2448 |
||
2449 |
text {* \blankline *} |
|
2450 |
||
2451 |
print_theorems |
|
2452 |
print_bnfs |
|
52794 | 2453 |
|
52824 | 2454 |
|
53617 | 2455 |
subsection {* Command Syntax |
2456 |
\label{ssec:bnf-command-syntax} *} |
|
2457 |
||
2458 |
||
53621 | 2459 |
subsubsection {* \keyw{bnf} |
2460 |
\label{sssec:bnf} *} |
|
52794 | 2461 |
|
53028 | 2462 |
text {* |
53829 | 2463 |
\begin{matharray}{rcl} |
2464 |
@{command_def "bnf"} & : & @{text "local_theory \<rightarrow> proof(prove)"} |
|
2465 |
\end{matharray} |
|
2466 |
||
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2467 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
2468 |
@@{command bnf} target? (name ':')? typ \<newline> |
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
2469 |
'map:' term ('sets:' (term +))? 'bd:' term \<newline> |
54421 | 2470 |
('wits:' (term +))? ('rel:' term)? |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2471 |
\<close>} |
53028 | 2472 |
*} |
52805 | 2473 |
|
53617 | 2474 |
|
54187 | 2475 |
subsubsection {* \keyw{bnf\_decl} |
2476 |
\label{sssec:bnf-decl} *} |
|
2477 |
||
2478 |
text {* |
|
2479 |
\begin{matharray}{rcl} |
|
55350 | 2480 |
@{command_def "bnf_decl"} & : & @{text "local_theory \<rightarrow> local_theory"} |
54187 | 2481 |
\end{matharray} |
2482 |
||
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2483 |
@{rail \<open> |
54602 | 2484 |
@@{command bnf_decl} target? @{syntax dt_name} |
2485 |
; |
|
55350 | 2486 |
@{syntax_def dt_name}: @{syntax tyargs}? name @{syntax map_rel}? \<newline> |
2487 |
@{syntax wit_types}? mixfix? |
|
54602 | 2488 |
; |
2489 |
@{syntax_def tyargs}: typefree | '(' (((name | '-') ':')? typefree + ',') ')' |
|
2490 |
; |
|
2491 |
@{syntax_def map_rel}: '(' ((('map' | 'rel') ':' name) +) ')' |
|
55350 | 2492 |
; |
2493 |
@{syntax_def wit_types}: '[' 'wits' ':' types ']' |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2494 |
\<close>} |
54602 | 2495 |
|
55350 | 2496 |
\noindent |
2497 |
The @{command bnf_decl} command declares a new type and associated constants |
|
2498 |
(map, set, relator, and cardinal bound) and asserts the BNF properties for |
|
2499 |
these constants as axioms. Type arguments are live by default; they can be |
|
2500 |
marked as dead by entering \texttt{-} (hyphen) instead of a name for the |
|
2501 |
corresponding set function. Witnesses can be specified by their types. |
|
2502 |
Otherwise, the syntax of @{command bnf_decl} is |
|
2503 |
identical to the left-hand side of a @{command datatype_new} or @{command |
|
2504 |
codatatype} definition. |
|
2505 |
||
2506 |
The command is useful to reason abstractly about BNFs. The axioms are safe |
|
2507 |
because there exists BNFs of arbitrary large arities. Applications must import |
|
2508 |
the theory @{theory BNF_Decl}, located in the directory |
|
2509 |
\verb|~~/src/HOL/Library|, to use this functionality. |
|
54187 | 2510 |
*} |
2511 |
||
2512 |
||
53621 | 2513 |
subsubsection {* \keyw{print\_bnfs} |
2514 |
\label{sssec:print-bnfs} *} |
|
53617 | 2515 |
|
2516 |
text {* |
|
53829 | 2517 |
\begin{matharray}{rcl} |
2518 |
@{command_def "print_bnfs"} & : & @{text "local_theory \<rightarrow>"} |
|
2519 |
\end{matharray} |
|
2520 |
||
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2521 |
@{rail \<open> |
53829 | 2522 |
@@{command print_bnfs} |
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2523 |
\<close>} |
53617 | 2524 |
*} |
2525 |
||
2526 |
||
2527 |
section {* Deriving Destructors and Theorems for Free Constructors |
|
2528 |
\label{sec:deriving-destructors-and-theorems-for-free-constructors} *} |
|
52794 | 2529 |
|
52805 | 2530 |
text {* |
53623 | 2531 |
The derivation of convenience theorems for types equipped with free constructors, |
53829 | 2532 |
as performed internally by @{command datatype_new} and @{command codatatype}, |
53623 | 2533 |
is available as a stand-alone command called @{command wrap_free_constructors}. |
52794 | 2534 |
|
53617 | 2535 |
% * need for this is rare but may arise if you want e.g. to add destructors to |
2536 |
% a type not introduced by ... |
|
2537 |
% |
|
2538 |
% * also useful for compatibility with old package, e.g. add destructors to |
|
2539 |
% old \keyw{datatype} |
|
2540 |
% |
|
2541 |
% * @{command wrap_free_constructors} |
|
54626 | 2542 |
% * @{text "no_discs_sels"}, @{text "no_code"}, @{text "rep_compat"} |
53617 | 2543 |
% * hack to have both co and nonco view via locale (cf. ext nats) |
54616 | 2544 |
% * code generator |
2545 |
% * eq, refl, simps |
|
52805 | 2546 |
*} |
52792 | 2547 |
|
52824 | 2548 |
|
53619 | 2549 |
(* |
53617 | 2550 |
subsection {* Introductory Example |
2551 |
\label{ssec:ctors-introductory-example} *} |
|
53619 | 2552 |
*) |
52794 | 2553 |
|
52824 | 2554 |
|
53617 | 2555 |
subsection {* Command Syntax |
2556 |
\label{ssec:ctors-command-syntax} *} |
|
2557 |
||
2558 |
||
53621 | 2559 |
subsubsection {* \keyw{wrap\_free\_constructors} |
53675 | 2560 |
\label{sssec:wrap-free-constructors} *} |
52828 | 2561 |
|
53018 | 2562 |
text {* |
53829 | 2563 |
\begin{matharray}{rcl} |
2564 |
@{command_def "wrap_free_constructors"} & : & @{text "local_theory \<rightarrow> proof(prove)"} |
|
2565 |
\end{matharray} |
|
53018 | 2566 |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2567 |
@{rail \<open> |
55029
61a6bf7d4b02
clarified @{rail} syntax: prefer explicit \<newline> symbol;
wenzelm
parents:
54958
diff
changeset
|
2568 |
@@{command wrap_free_constructors} target? @{syntax dt_options} \<newline> |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
2569 |
term_list name @{syntax wfc_discs_sels}? |
53018 | 2570 |
; |
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
2571 |
@{syntax_def wfc_discs_sels}: name_list (name_list_list name_term_list_list? )? |
53018 | 2572 |
; |
53534 | 2573 |
@{syntax_def name_term}: (name ':' term) |
54421 | 2574 |
; |
2575 |
X_list: '[' (X + ',') ']' |
|
55112
b1a5d603fd12
prefer rail cartouche -- avoid back-slashed quotes;
wenzelm
parents:
55029
diff
changeset
|
2576 |
\<close>} |
53018 | 2577 |
|
54626 | 2578 |
% options: no_discs_sels no_code rep_compat |
53028 | 2579 |
|
53829 | 2580 |
\noindent |
53542 | 2581 |
Section~\ref{ssec:datatype-generated-theorems} lists the generated theorems. |
53018 | 2582 |
*} |
52828 | 2583 |
|
52794 | 2584 |
|
53617 | 2585 |
(* |
52827 | 2586 |
section {* Standard ML Interface |
52805 | 2587 |
\label{sec:standard-ml-interface} *} |
52792 | 2588 |
|
52805 | 2589 |
text {* |
53623 | 2590 |
The package's programmatic interface. |
52805 | 2591 |
*} |
53617 | 2592 |
*) |
52794 | 2593 |
|
2594 |
||
53617 | 2595 |
(* |
52827 | 2596 |
section {* Interoperability |
52805 | 2597 |
\label{sec:interoperability} *} |
52794 | 2598 |
|
52805 | 2599 |
text {* |
53623 | 2600 |
The package's interaction with other Isabelle packages and tools, such as the |
2601 |
code generator and the counterexample generators. |
|
52805 | 2602 |
*} |
52794 | 2603 |
|
52824 | 2604 |
|
52828 | 2605 |
subsection {* Transfer and Lifting |
2606 |
\label{ssec:transfer-and-lifting} *} |
|
52794 | 2607 |
|
52824 | 2608 |
|
52828 | 2609 |
subsection {* Code Generator |
2610 |
\label{ssec:code-generator} *} |
|
52794 | 2611 |
|
52824 | 2612 |
|
52828 | 2613 |
subsection {* Quickcheck |
2614 |
\label{ssec:quickcheck} *} |
|
52794 | 2615 |
|
52824 | 2616 |
|
52828 | 2617 |
subsection {* Nitpick |
2618 |
\label{ssec:nitpick} *} |
|
52794 | 2619 |
|
52824 | 2620 |
|
52828 | 2621 |
subsection {* Nominal Isabelle |
2622 |
\label{ssec:nominal-isabelle} *} |
|
53617 | 2623 |
*) |
52794 | 2624 |
|
52805 | 2625 |
|
53617 | 2626 |
(* |
52827 | 2627 |
section {* Known Bugs and Limitations |
52805 | 2628 |
\label{sec:known-bugs-and-limitations} *} |
2629 |
||
2630 |
text {* |
|
53623 | 2631 |
Known open issues of the package. |
52805 | 2632 |
*} |
52794 | 2633 |
|
2634 |
text {* |
|
53753
ae7f50e70c09
renamed "primcorec" to "primcorecursive", to open the door to a 'theory -> theory' command called "primcorec" (cf. "fun" vs. "function")
blanchet
parents:
53752
diff
changeset
|
2635 |
%* primcorecursive and primcorec is unfinished |
53617 | 2636 |
% |
2637 |
%* slow n-ary mutual (co)datatype, avoid as much as possible (e.g. using nesting) |
|
2638 |
% |
|
2639 |
%* issues with HOL-Proofs? |
|
2640 |
% |
|
2641 |
%* partial documentation |
|
2642 |
% |
|
2643 |
%* no way to register "sum" and "prod" as (co)datatypes to enable N2M reduction for them |
|
2644 |
% (for @{command datatype_new_compat} and prim(co)rec) |
|
2645 |
% |
|
53619 | 2646 |
% * a fortiori, no way to register same type as both data- and codatatype |
53617 | 2647 |
% |
2648 |
%* no recursion through unused arguments (unlike with the old package) |
|
2649 |
% |
|
2650 |
%* in a locale, cannot use locally fixed types (because of limitation in typedef)? |
|
53619 | 2651 |
% |
2652 |
% *names of variables suboptimal |
|
52822 | 2653 |
*} |
53675 | 2654 |
*) |
52822 | 2655 |
|
2656 |
||
2657 |
text {* |
|
53863
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
2658 |
\section*{Acknowledgment} |
c7364dca96f2
textual improvements following Christian Sternagel's feedback
blanchet
parents:
53857
diff
changeset
|
2659 |
|
53749
b37db925b663
adapted primcorec documentation to reflect the three views
blanchet
parents:
53748
diff
changeset
|
2660 |
Tobias Nipkow and Makarius Wenzel encouraged us to implement the new |
53617 | 2661 |
(co)datatype package. Andreas Lochbihler provided lots of comments on earlier |
2662 |
versions of the package, especially for the coinductive part. Brian Huffman |
|
55129
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
2663 |
suggested major simplifications to the internal constructions, many of which |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
2664 |
have yet to be implemented. Florian Haftmann and Christian Urban provided |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
2665 |
general advice on Isabelle and package writing. Stefan Milius and Lutz |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
2666 |
Schr\"oder found an elegant proof that eliminated one of the BNF proof |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
2667 |
obligations. Andreas Lochbihler and Christian Sternagel suggested many textual |
26bd1cba3ab5
killed 'More_BNFs' by moving its various bits where they (now) belong
blanchet
parents:
55114
diff
changeset
|
2668 |
improvements to this tutorial. |
52794 | 2669 |
*} |
53617 | 2670 |
|
52792 | 2671 |
end |