wenzelm@7135
|
1 |
|
wenzelm@13048
|
2 |
\chapter{Generic tools and packages}\label{ch:gen-tools}
|
wenzelm@7167
|
3 |
|
wenzelm@12621
|
4 |
\section{Theory specification commands}
|
wenzelm@12618
|
5 |
|
wenzelm@12618
|
6 |
\subsection{Axiomatic type classes}\label{sec:axclass}
|
wenzelm@7167
|
7 |
|
wenzelm@8517
|
8 |
\indexisarcmd{axclass}\indexisarcmd{instance}\indexisarmeth{intro-classes}
|
wenzelm@7167
|
9 |
\begin{matharray}{rcl}
|
wenzelm@8517
|
10 |
\isarcmd{axclass} & : & \isartrans{theory}{theory} \\
|
wenzelm@8517
|
11 |
\isarcmd{instance} & : & \isartrans{theory}{proof(prove)} \\
|
wenzelm@8517
|
12 |
intro_classes & : & \isarmeth \\
|
wenzelm@7167
|
13 |
\end{matharray}
|
wenzelm@7167
|
14 |
|
wenzelm@8517
|
15 |
Axiomatic type classes are provided by Isabelle/Pure as a \emph{definitional}
|
wenzelm@8517
|
16 |
interface to type classes (cf.~\S\ref{sec:classes}). Thus any object logic
|
wenzelm@8547
|
17 |
may make use of this light-weight mechanism of abstract theories
|
wenzelm@8901
|
18 |
\cite{Wenzel:1997:TPHOL}. There is also a tutorial on using axiomatic type
|
wenzelm@13024
|
19 |
classes in Isabelle \cite{isabelle-axclass} that is part of the standard
|
wenzelm@8901
|
20 |
Isabelle documentation.
|
wenzelm@8517
|
21 |
|
wenzelm@7167
|
22 |
\begin{rail}
|
wenzelm@12879
|
23 |
'axclass' classdecl (axmdecl prop +)
|
wenzelm@8517
|
24 |
;
|
wenzelm@14605
|
25 |
'instance' (nameref ('<' | subseteq) nameref | nameref '::' arity)
|
wenzelm@7167
|
26 |
;
|
wenzelm@7167
|
27 |
\end{rail}
|
wenzelm@7167
|
28 |
|
wenzelm@7167
|
29 |
\begin{descr}
|
wenzelm@17274
|
30 |
|
wenzelm@13024
|
31 |
\item [$\AXCLASS~c \subseteq \vec c~~axms$] defines an axiomatic type class as
|
wenzelm@11100
|
32 |
the intersection of existing classes, with additional axioms holding. Class
|
wenzelm@10223
|
33 |
axioms may not contain more than one type variable. The class axioms (with
|
wenzelm@10223
|
34 |
implicit sort constraints added) are bound to the given names. Furthermore
|
wenzelm@17274
|
35 |
a class introduction rule is generated (being bound as
|
wenzelm@17274
|
36 |
$c_class{\dtt}intro$); this rule is employed by method $intro_classes$ to
|
wenzelm@17274
|
37 |
support instantiation proofs of this class.
|
wenzelm@17274
|
38 |
|
wenzelm@12976
|
39 |
The ``axioms'' are stored as theorems according to the given name
|
wenzelm@13039
|
40 |
specifications, adding the class name $c$ as name space prefix; the same
|
wenzelm@17274
|
41 |
facts are also stored collectively as $c_class{\dtt}axioms$.
|
wenzelm@14605
|
42 |
|
wenzelm@14605
|
43 |
\item [$\INSTANCE~c@1 \subseteq c@2$ and $\INSTANCE~t :: (\vec s)s$] setup a
|
wenzelm@11100
|
44 |
goal stating a class relation or type arity. The proof would usually
|
wenzelm@11100
|
45 |
proceed by $intro_classes$, and then establish the characteristic theorems
|
wenzelm@11100
|
46 |
of the type classes involved. After finishing the proof, the theory will be
|
wenzelm@11100
|
47 |
augmented by a type signature declaration corresponding to the resulting
|
wenzelm@11100
|
48 |
theorem.
|
wenzelm@13041
|
49 |
|
wenzelm@8517
|
50 |
\item [$intro_classes$] repeatedly expands all class introduction rules of
|
wenzelm@10858
|
51 |
this theory. Note that this method usually needs not be named explicitly,
|
wenzelm@13040
|
52 |
as it is already included in the default proof step (of $\PROOFNAME$ etc.).
|
wenzelm@13040
|
53 |
In particular, instantiation of trivial (syntactic) classes may be performed
|
wenzelm@13040
|
54 |
by a single ``$\DDOT$'' proof step.
|
wenzelm@13027
|
55 |
|
wenzelm@7167
|
56 |
\end{descr}
|
wenzelm@7167
|
57 |
|
wenzelm@7315
|
58 |
|
wenzelm@12618
|
59 |
\subsection{Locales and local contexts}\label{sec:locale}
|
wenzelm@12618
|
60 |
|
wenzelm@13040
|
61 |
Locales are named local contexts, consisting of a list of declaration elements
|
wenzelm@13041
|
62 |
that are modeled after the Isar proof context commands (cf.\
|
wenzelm@13040
|
63 |
\S\ref{sec:proof-context}).
|
wenzelm@12618
|
64 |
|
wenzelm@13048
|
65 |
|
wenzelm@12976
|
66 |
\subsubsection{Localized commands}
|
wenzelm@12967
|
67 |
|
wenzelm@12976
|
68 |
Existing locales may be augmented later on by adding new facts. Note that the
|
wenzelm@12976
|
69 |
actual context definition may not be changed! Several theory commands that
|
wenzelm@12976
|
70 |
produce facts in some way are available in ``localized'' versions, referring
|
wenzelm@12976
|
71 |
to a named locale instead of the global theory context.
|
wenzelm@12976
|
72 |
|
wenzelm@12976
|
73 |
\indexouternonterm{locale}
|
wenzelm@12967
|
74 |
\begin{rail}
|
wenzelm@12967
|
75 |
locale: '(' 'in' name ')'
|
wenzelm@12967
|
76 |
;
|
wenzelm@12976
|
77 |
\end{rail}
|
wenzelm@12967
|
78 |
|
wenzelm@12976
|
79 |
Emerging facts of localized commands are stored in two versions, both in the
|
wenzelm@12976
|
80 |
target locale and the theory (after export). The latter view produces a
|
wenzelm@12976
|
81 |
qualified binding, using the locale name as a name space prefix.
|
wenzelm@12976
|
82 |
|
wenzelm@12976
|
83 |
For example, ``$\LEMMAS~(\IN~loc)~a = \vec b$'' retrieves facts $\vec b$ from
|
wenzelm@12976
|
84 |
the locale context of $loc$ and augments its body by an appropriate
|
wenzelm@12976
|
85 |
``$\isarkeyword{notes}$'' element (see below). The exported view of $a$,
|
wenzelm@12976
|
86 |
after discharging the locale context, is stored as $loc{.}a$ within the global
|
wenzelm@13041
|
87 |
theory. A localized goal ``$\LEMMANAME~(\IN~loc)~a:~\phi$'' works similarly,
|
wenzelm@13041
|
88 |
only that the fact emerges through the subsequent proof, which may refer to
|
wenzelm@13041
|
89 |
the full infrastructure of the locale context (covering local parameters with
|
wenzelm@13041
|
90 |
typing and concrete syntax, assumptions, definitions etc.). Most notably,
|
wenzelm@13411
|
91 |
fact declarations of the locale are active during the proof as well (e.g.\
|
wenzelm@13041
|
92 |
local $simp$ rules).
|
wenzelm@12976
|
93 |
|
wenzelm@13411
|
94 |
As a general principle, results exported from a locale context acquire
|
wenzelm@13411
|
95 |
additional premises according to the specification. Usually this is only a
|
wenzelm@13411
|
96 |
single predicate according to the standard ``closed'' view of locale
|
wenzelm@13411
|
97 |
specifications.
|
wenzelm@13411
|
98 |
|
wenzelm@12976
|
99 |
|
wenzelm@12976
|
100 |
\subsubsection{Locale specifications}
|
wenzelm@12976
|
101 |
|
wenzelm@12976
|
102 |
\indexisarcmd{locale}\indexisarcmd{print-locale}\indexisarcmd{print-locales}
|
wenzelm@12976
|
103 |
\begin{matharray}{rcl}
|
wenzelm@12976
|
104 |
\isarcmd{locale} & : & \isarkeep{theory} \\
|
wenzelm@12976
|
105 |
\isarcmd{print_locale}^* & : & \isarkeep{theory~|~proof} \\
|
wenzelm@12976
|
106 |
\isarcmd{print_locales}^* & : & \isarkeep{theory~|~proof} \\
|
wenzelm@12976
|
107 |
\end{matharray}
|
wenzelm@12976
|
108 |
|
wenzelm@12976
|
109 |
\indexouternonterm{contextexpr}\indexouternonterm{contextelem}
|
wenzelm@12976
|
110 |
|
wenzelm@12976
|
111 |
\railalias{printlocale}{print\_locale}
|
wenzelm@12976
|
112 |
\railterm{printlocale}
|
wenzelm@12976
|
113 |
|
wenzelm@12976
|
114 |
\begin{rail}
|
wenzelm@13411
|
115 |
'locale' ('(open)')? name ('=' localeexpr)?
|
wenzelm@12976
|
116 |
;
|
ballarin@17228
|
117 |
printlocale '!'? localeexpr
|
wenzelm@12976
|
118 |
;
|
wenzelm@12976
|
119 |
localeexpr: ((contextexpr '+' (contextelem+)) | contextexpr | (contextelem+))
|
wenzelm@12976
|
120 |
;
|
wenzelm@12976
|
121 |
|
wenzelm@12976
|
122 |
contextexpr: nameref | '(' contextexpr ')' |
|
ballarin@16102
|
123 |
(contextexpr (name mixfix? +)) | (contextexpr + '+')
|
wenzelm@12976
|
124 |
;
|
ballarin@16168
|
125 |
contextelem: fixes | constrains | assumes | defines | notes | includes
|
wenzelm@12976
|
126 |
;
|
wenzelm@12976
|
127 |
fixes: 'fixes' (name ('::' type)? structmixfix? + 'and')
|
wenzelm@12976
|
128 |
;
|
ballarin@16168
|
129 |
constrains: 'constrains' (name '::' type + 'and')
|
ballarin@16168
|
130 |
;
|
wenzelm@12976
|
131 |
assumes: 'assumes' (thmdecl? props + 'and')
|
wenzelm@12976
|
132 |
;
|
wenzelm@12976
|
133 |
defines: 'defines' (thmdecl? prop proppat? + 'and')
|
wenzelm@12976
|
134 |
;
|
wenzelm@12976
|
135 |
notes: 'notes' (thmdef? thmrefs + 'and')
|
wenzelm@12976
|
136 |
;
|
wenzelm@12976
|
137 |
includes: 'includes' contextexpr
|
wenzelm@12976
|
138 |
;
|
wenzelm@12967
|
139 |
\end{rail}
|
wenzelm@12618
|
140 |
|
wenzelm@12976
|
141 |
\begin{descr}
|
wenzelm@13411
|
142 |
|
wenzelm@13411
|
143 |
\item [$\LOCALE~loc~=~import~+~body$] defines a new locale $loc$ as a context
|
wenzelm@12976
|
144 |
consisting of a certain view of existing locales ($import$) plus some
|
wenzelm@12976
|
145 |
additional elements ($body$). Both $import$ and $body$ are optional; the
|
wenzelm@13024
|
146 |
degenerate form $\LOCALE~loc$ defines an empty locale, which may still be
|
wenzelm@13024
|
147 |
useful to collect declarations of facts later on. Type-inference on locale
|
wenzelm@12976
|
148 |
expressions automatically takes care of the most general typing that the
|
wenzelm@12976
|
149 |
combined context elements may acquire.
|
wenzelm@13041
|
150 |
|
wenzelm@12976
|
151 |
The $import$ consists of a structured context expression, consisting of
|
wenzelm@12976
|
152 |
references to existing locales, renamed contexts, or merged contexts.
|
ballarin@16102
|
153 |
Renaming uses positional notation: $c~\vec x$ means that (a prefix of) the
|
wenzelm@12976
|
154 |
fixed parameters of context $c$ are named according to $\vec x$; a
|
ballarin@16102
|
155 |
``\texttt{_}'' (underscore) \indexisarthm{_@\texttt{_}} means to skip that
|
ballarin@16102
|
156 |
position. Renaming by default deletes existing syntax. Optionally,
|
ballarin@16102
|
157 |
new syntax may by specified with a mixfix annotation. Note that the
|
ballarin@16102
|
158 |
special syntax declared with ``$(structure)$'' (see below) is
|
ballarin@16102
|
159 |
neither deleted nor can it be changed.
|
wenzelm@13041
|
160 |
Merging proceeds from left-to-right, suppressing any duplicates stemming
|
wenzelm@13041
|
161 |
from different paths through the import hierarchy.
|
wenzelm@13041
|
162 |
|
wenzelm@12976
|
163 |
The $body$ consists of basic context elements, further context expressions
|
wenzelm@12976
|
164 |
may be included as well.
|
wenzelm@12976
|
165 |
|
wenzelm@12976
|
166 |
\begin{descr}
|
wenzelm@13041
|
167 |
|
wenzelm@12976
|
168 |
\item [$\FIXES{~x::\tau~(mx)}$] declares a local parameter of type $\tau$
|
wenzelm@12976
|
169 |
and mixfix annotation $mx$ (both are optional). The special syntax
|
wenzelm@13027
|
170 |
declaration ``$(structure)$'' means that $x$ may be referenced
|
wenzelm@13027
|
171 |
implicitly in this context.
|
wenzelm@13041
|
172 |
|
ballarin@16168
|
173 |
\item [$\CONSTRAINS{~x::\tau}$] introduces a type constraint $\tau$
|
ballarin@16168
|
174 |
on the local parameter $x$.
|
ballarin@16168
|
175 |
|
wenzelm@12976
|
176 |
\item [$\ASSUMES{a}{\vec\phi}$] introduces local premises, similar to
|
wenzelm@12976
|
177 |
$\ASSUMENAME$ within a proof (cf.\ \S\ref{sec:proof-context}).
|
wenzelm@13041
|
178 |
|
wenzelm@12976
|
179 |
\item [$\DEFINES{a}{x \equiv t}$] defines a previously declared parameter.
|
wenzelm@13041
|
180 |
This is close to $\DEFNAME$ within a proof (cf.\
|
wenzelm@12976
|
181 |
\S\ref{sec:proof-context}), but $\DEFINESNAME$ takes an equational
|
wenzelm@13041
|
182 |
proposition instead of variable-term pair. The left-hand side of the
|
wenzelm@13041
|
183 |
equation may have additional arguments, e.g.\ ``$\DEFINES{}{f~\vec x
|
wenzelm@13041
|
184 |
\equiv t}$''.
|
wenzelm@13041
|
185 |
|
wenzelm@12976
|
186 |
\item [$\NOTES{a}{\vec b}$] reconsiders facts within a local context. Most
|
wenzelm@12976
|
187 |
notably, this may include arbitrary declarations in any attribute
|
wenzelm@12976
|
188 |
specifications included here, e.g.\ a local $simp$ rule.
|
wenzelm@13041
|
189 |
|
wenzelm@12976
|
190 |
\item [$\INCLUDES{c}$] copies the specified context in a statically scoped
|
ballarin@15763
|
191 |
manner. Only available in the long goal format of \S\ref{sec:goals}.
|
wenzelm@13041
|
192 |
|
wenzelm@12976
|
193 |
In contrast, the initial $import$ specification of a locale expression
|
wenzelm@12976
|
194 |
maintains a dynamic relation to the locales being referenced (benefiting
|
wenzelm@12976
|
195 |
from any later fact declarations in the obvious manner).
|
wenzelm@12976
|
196 |
\end{descr}
|
wenzelm@13411
|
197 |
|
wenzelm@13041
|
198 |
Note that ``$\IS{p}$'' patterns given in the syntax of $\ASSUMESNAME$ and
|
wenzelm@13411
|
199 |
$\DEFINESNAME$ above are illegal in locale definitions. In the long goal
|
wenzelm@13411
|
200 |
format of \S\ref{sec:goals}, term bindings may be included as expected,
|
wenzelm@13411
|
201 |
though.
|
wenzelm@13411
|
202 |
|
wenzelm@13411
|
203 |
\medskip By default, locale specifications are ``closed up'' by turning the
|
wenzelm@13411
|
204 |
given text into a predicate definition $loc_axioms$ and deriving the
|
wenzelm@13411
|
205 |
original assumptions as local lemmas (modulo local definitions). The
|
wenzelm@13411
|
206 |
predicate statement covers only the newly specified assumptions, omitting
|
wenzelm@13411
|
207 |
the content of included locale expressions. The full cumulative view is
|
wenzelm@13411
|
208 |
only provided on export, involving another predicate $loc$ that refers to
|
wenzelm@13411
|
209 |
the complete specification text.
|
wenzelm@13411
|
210 |
|
wenzelm@13411
|
211 |
In any case, the predicate arguments are those locale parameters that
|
wenzelm@13411
|
212 |
actually occur in the respective piece of text. Also note that these
|
wenzelm@13411
|
213 |
predicates operate at the meta-level in theory, but the locale packages
|
wenzelm@13411
|
214 |
attempts to internalize statements according to the object-logic setup
|
wenzelm@13411
|
215 |
(e.g.\ replacing $\Forall$ by $\forall$, and $\Imp$ by $\imp$ in HOL; see
|
wenzelm@13411
|
216 |
also \S\ref{sec:object-logic}). Separate introduction rules
|
wenzelm@13411
|
217 |
$loc_axioms.intro$ and $loc.intro$ are declared as well.
|
wenzelm@13411
|
218 |
|
wenzelm@13411
|
219 |
The $(open)$ option of a locale specification prevents both the current
|
wenzelm@13411
|
220 |
$loc_axioms$ and cumulative $loc$ predicate constructions. Predicates are
|
wenzelm@13411
|
221 |
also omitted for empty specification texts.
|
wenzelm@12976
|
222 |
|
wenzelm@12976
|
223 |
\item [$\isarkeyword{print_locale}~import~+~body$] prints the specified locale
|
wenzelm@12976
|
224 |
expression in a flattened form. The notable special case
|
wenzelm@12976
|
225 |
$\isarkeyword{print_locale}~loc$ just prints the contents of the named
|
wenzelm@12976
|
226 |
locale, but keep in mind that type-inference will normalize type variables
|
ballarin@17228
|
227 |
according to the usual alphabetical order. The command omits
|
ballarin@17228
|
228 |
$\isarkeyword{notes}$ elements by default. Use
|
ballarin@17228
|
229 |
$\isarkeyword{print_locale}!$ to get them included.
|
wenzelm@13041
|
230 |
|
wenzelm@12976
|
231 |
\item [$\isarkeyword{print_locales}$] prints the names of all locales of the
|
wenzelm@12976
|
232 |
current theory.
|
wenzelm@12976
|
233 |
|
wenzelm@12976
|
234 |
\end{descr}
|
wenzelm@12976
|
235 |
|
wenzelm@12618
|
236 |
|
ballarin@15763
|
237 |
\subsubsection{Interpretation of locales}
|
ballarin@15763
|
238 |
|
ballarin@15763
|
239 |
Locale expressions (more precisely, \emph{context expressions}) may be
|
ballarin@15763
|
240 |
instantiated, and the instantiated facts added to the current context.
|
ballarin@15763
|
241 |
This requires a proof of the instantiated specification and is called
|
ballarin@15763
|
242 |
\emph{locale interpretation}. Interpretation is possible in theories
|
ballarin@17043
|
243 |
and locales
|
ballarin@17043
|
244 |
(command $\isarcmd{interpretation}$) and also in proof contexts
|
ballarin@15763
|
245 |
($\isarcmd{interpret}$).
|
ballarin@15763
|
246 |
|
ballarin@15763
|
247 |
\indexisarcmd{interpretation}\indexisarcmd{interpret}
|
ballarin@15763
|
248 |
\indexisarcmd{print-interps}
|
ballarin@15763
|
249 |
\begin{matharray}{rcl}
|
ballarin@15763
|
250 |
\isarcmd{interpretation} & : & \isartrans{theory}{proof(prove)} \\
|
ballarin@15763
|
251 |
\isarcmd{interpret} & : & \isartrans{proof(state) ~|~ proof(chain)}{proof(prove)} \\
|
ballarin@15763
|
252 |
\isarcmd{print_interps}^* & : & \isarkeep{theory~|~proof} \\
|
ballarin@15763
|
253 |
\end{matharray}
|
ballarin@15763
|
254 |
|
ballarin@15763
|
255 |
\indexouternonterm{interp}
|
ballarin@15763
|
256 |
|
ballarin@15763
|
257 |
\railalias{printinterps}{print\_interps}
|
ballarin@15763
|
258 |
\railterm{printinterps}
|
ballarin@15763
|
259 |
|
ballarin@15763
|
260 |
\begin{rail}
|
ballarin@17043
|
261 |
'interpretation' (interp | name ('<' | subseteq) contextexp)
|
ballarin@15763
|
262 |
;
|
ballarin@15763
|
263 |
'interpret' interp
|
ballarin@15763
|
264 |
;
|
ballarin@17139
|
265 |
printinterps '!'? name
|
ballarin@15763
|
266 |
;
|
ballarin@15763
|
267 |
interp: thmdecl? contextexpr ('[' (inst+) ']')?
|
ballarin@15763
|
268 |
;
|
ballarin@15763
|
269 |
\end{rail}
|
ballarin@15763
|
270 |
|
ballarin@17043
|
271 |
|
ballarin@15763
|
272 |
\begin{descr}
|
ballarin@15763
|
273 |
|
ballarin@15763
|
274 |
\item [$\isarcmd{interpretation}~expr~insts$]
|
ballarin@17043
|
275 |
|
ballarin@17043
|
276 |
The first form of $\isarcmd{interpretation}$ interprets $expr$
|
ballarin@17043
|
277 |
in the theory. The instantiation is given as a list of
|
ballarin@17043
|
278 |
terms $insts$ and is positional.
|
ballarin@15763
|
279 |
All parameters must receive an instantiation term --- with the
|
ballarin@15763
|
280 |
exception of defined parameters. These are, if omitted, derived
|
ballarin@15763
|
281 |
from the defining equation and other instantiations. Use ``\_'' to
|
ballarin@15763
|
282 |
omit an instantiation term. Free variables are automatically
|
ballarin@15763
|
283 |
generalized.
|
ballarin@15763
|
284 |
|
ballarin@17043
|
285 |
The command generates proof obligations for the instantiated
|
ballarin@17043
|
286 |
specifications (assumes and defines elements). Once these are
|
ballarin@17043
|
287 |
discharged by the user, instantiated facts are added to the theory in
|
ballarin@17043
|
288 |
a post-processing phase.
|
ballarin@15763
|
289 |
|
ballarin@15763
|
290 |
The command is aware of interpretations already active in the
|
ballarin@15763
|
291 |
theory. No proof obligations are generated for those, neither is
|
ballarin@15763
|
292 |
post-processing applied to their facts. This avoids duplication of
|
ballarin@15763
|
293 |
interpreted facts, in particular. Note that, in the case of a
|
ballarin@15763
|
294 |
locale with import, parts of the interpretation may already be
|
ballarin@15763
|
295 |
active. The command will only generate proof obligations and add
|
ballarin@15763
|
296 |
facts for new parts.
|
ballarin@15763
|
297 |
|
ballarin@17043
|
298 |
The context expression may be preceded by a name and/or attributes.
|
ballarin@17043
|
299 |
These take effect in the post-processing of facts. The name is used
|
ballarin@17043
|
300 |
to prefix fact names, for example to avoid accidental hiding of
|
ballarin@17043
|
301 |
other facts. Attributes are applied after attributes of the
|
ballarin@17043
|
302 |
interpreted facts.
|
ballarin@17043
|
303 |
|
ballarin@15763
|
304 |
Adding facts to locales has the
|
ballarin@15763
|
305 |
effect of adding interpreted facts to the theory for all active
|
ballarin@17043
|
306 |
interpretations also. That is, interpretations dynamically
|
ballarin@17043
|
307 |
participate in any facts added to locales.
|
ballarin@17043
|
308 |
|
ballarin@17043
|
309 |
\item [$\isarcmd{interpretation}~name~\subseteq~expr$]
|
ballarin@17043
|
310 |
|
ballarin@17043
|
311 |
This form of the command interprets $expr$ in the locale $name$. It
|
ballarin@17043
|
312 |
requires a proof that the specification of $name$ implies the
|
ballarin@17043
|
313 |
specification of $expr$. As in the localized version of the theorem
|
ballarin@17043
|
314 |
command, the proof is in the context of $name$. After the proof
|
ballarin@17043
|
315 |
obligation has been dischared, the facts of $expr$
|
ballarin@17043
|
316 |
become part of locale $name$ as \emph{derived} context elements and
|
ballarin@17043
|
317 |
are available when the context $name$ is subsequently entered.
|
ballarin@17043
|
318 |
Note that, like import, this is dynamic: facts added to a locale
|
ballarin@17139
|
319 |
part of $expr$ after interpretation become also available in
|
ballarin@17043
|
320 |
$name$. Like facts
|
ballarin@17043
|
321 |
of renamed context elements, facts obtained by interpretation may be
|
ballarin@17043
|
322 |
accessed by prefixing with the parameter renaming (where the parameters
|
ballarin@17043
|
323 |
are separated by `\_').
|
ballarin@17043
|
324 |
|
ballarin@17043
|
325 |
Unlike interpretation in theories, instantiation is confined to the
|
ballarin@17043
|
326 |
renaming of parameters, which may be specified as part of the context
|
ballarin@17043
|
327 |
expression $expr$. Using defined parameters in $name$ one may
|
ballarin@17043
|
328 |
achieve an effect similar to instantiation, though.
|
ballarin@17043
|
329 |
|
ballarin@17043
|
330 |
Only specification fragments of $expr$ that are not already part of
|
ballarin@17043
|
331 |
$name$ (be it imported, derived or a derived fragment of the import)
|
ballarin@17043
|
332 |
are considered by interpretation. This enables circular
|
ballarin@17043
|
333 |
interpretations.
|
ballarin@17043
|
334 |
|
ballarin@17139
|
335 |
If interpretations of $name$ exist in the current theory, the
|
ballarin@17139
|
336 |
command adds interpretations for $expr$ as well, with the same
|
ballarin@17139
|
337 |
prefix and attributes, although only for fragments of $expr$ that
|
ballarin@17139
|
338 |
are not interpreted in the theory already.
|
ballarin@17139
|
339 |
|
ballarin@15763
|
340 |
\item [$\isarcmd{interpret}~expr~insts$]
|
ballarin@15763
|
341 |
interprets $expr$ in the proof context and is otherwise similar to
|
ballarin@17043
|
342 |
interpretation in theories. Free variables in instantiations are not
|
ballarin@15763
|
343 |
generalized, however.
|
ballarin@15763
|
344 |
|
ballarin@15763
|
345 |
\item [$\isarcmd{print_interps}~loc$]
|
ballarin@15763
|
346 |
prints the interpretations of a particular locale $loc$ that are
|
ballarin@17139
|
347 |
active in the current context, either theory or proof context. The
|
ballarin@17139
|
348 |
exclamation point argument causes triggers printing of
|
ballarin@17139
|
349 |
\emph{witness} theorems justifying interpretations. These are
|
ballarin@17139
|
350 |
normally omitted from the output.
|
ballarin@17139
|
351 |
|
ballarin@15763
|
352 |
|
ballarin@15763
|
353 |
\end{descr}
|
ballarin@15763
|
354 |
|
ballarin@15837
|
355 |
\begin{warn}
|
ballarin@15837
|
356 |
Since attributes are applied to interpreted theorems, interpretation
|
ballarin@15837
|
357 |
may modify the current simpset and claset. Take this into
|
ballarin@15837
|
358 |
account when choosing attributes for local theorems.
|
ballarin@15837
|
359 |
\end{warn}
|
ballarin@15837
|
360 |
|
ballarin@16168
|
361 |
\begin{warn}
|
ballarin@17043
|
362 |
An interpretation in a theory may subsume previous interpretations.
|
ballarin@17043
|
363 |
This happens if the same specification fragment is interpreted twice
|
ballarin@17043
|
364 |
and the instantiation of the second interpretation is more general
|
ballarin@17043
|
365 |
than the interpretation of the first. A warning
|
ballarin@16168
|
366 |
is issued, since it is likely that these could have been generalized
|
ballarin@16168
|
367 |
in the first place. The locale package does not attempt to remove
|
ballarin@16168
|
368 |
subsumed interpretations. This situation is normally harmless, but
|
ballarin@16168
|
369 |
note that $blast$ gets confused by the presence of multiple axclass
|
ballarin@17139
|
370 |
instances of a rule.
|
ballarin@16168
|
371 |
\end{warn}
|
ballarin@16168
|
372 |
|
ballarin@15763
|
373 |
|
wenzelm@12621
|
374 |
\section{Derived proof schemes}
|
wenzelm@12621
|
375 |
|
wenzelm@12621
|
376 |
\subsection{Generalized elimination}\label{sec:obtain}
|
wenzelm@12621
|
377 |
|
wenzelm@17864
|
378 |
\indexisarcmd{obtain}\indexisarcmd{guess}
|
wenzelm@12621
|
379 |
\begin{matharray}{rcl}
|
wenzelm@12621
|
380 |
\isarcmd{obtain} & : & \isartrans{proof(state)}{proof(prove)} \\
|
wenzelm@17864
|
381 |
\isarcmd{guess}^* & : & \isartrans{proof(state)}{proof(prove)} \\
|
wenzelm@12621
|
382 |
\end{matharray}
|
wenzelm@12621
|
383 |
|
wenzelm@12621
|
384 |
Generalized elimination means that additional elements with certain properties
|
wenzelm@13041
|
385 |
may be introduced in the current context, by virtue of a locally proven
|
wenzelm@12621
|
386 |
``soundness statement''. Technically speaking, the $\OBTAINNAME$ language
|
wenzelm@12621
|
387 |
element is like a declaration of $\FIXNAME$ and $\ASSUMENAME$ (see also see
|
wenzelm@12621
|
388 |
\S\ref{sec:proof-context}), together with a soundness proof of its additional
|
wenzelm@12621
|
389 |
claim. According to the nature of existential reasoning, assumptions get
|
wenzelm@12621
|
390 |
eliminated from any result exported from the context later, provided that the
|
wenzelm@12621
|
391 |
corresponding parameters do \emph{not} occur in the conclusion.
|
wenzelm@12621
|
392 |
|
wenzelm@12621
|
393 |
\begin{rail}
|
wenzelm@12879
|
394 |
'obtain' (vars + 'and') 'where' (props + 'and')
|
wenzelm@12621
|
395 |
;
|
wenzelm@17864
|
396 |
'guess' (vars + 'and')
|
wenzelm@17864
|
397 |
;
|
wenzelm@12621
|
398 |
\end{rail}
|
wenzelm@12621
|
399 |
|
wenzelm@12621
|
400 |
$\OBTAINNAME$ is defined as a derived Isar command as follows, where $\vec b$
|
wenzelm@12621
|
401 |
shall refer to (optional) facts indicated for forward chaining.
|
wenzelm@12621
|
402 |
\begin{matharray}{l}
|
wenzelm@12621
|
403 |
\langle facts~\vec b\rangle \\
|
wenzelm@12621
|
404 |
\OBTAIN{\vec x}{a}{\vec \phi}~~\langle proof\rangle \equiv {} \\[1ex]
|
wenzelm@13041
|
405 |
\quad \HAVE{}{\All{thesis} (\All{\vec x} \vec\phi \Imp thesis) \Imp thesis} \\
|
wenzelm@13041
|
406 |
\quad \PROOF{succeed} \\
|
wenzelm@12621
|
407 |
\qquad \FIX{thesis} \\
|
wenzelm@13041
|
408 |
\qquad \ASSUME{that~[intro?]}{\All{\vec x} \vec\phi \Imp thesis} \\
|
wenzelm@13042
|
409 |
\qquad \THUS{}{thesis} \\
|
wenzelm@13042
|
410 |
\quad\qquad \APPLY{-} \\
|
wenzelm@13041
|
411 |
\quad\qquad \USING{\vec b}~~\langle proof\rangle \\
|
wenzelm@13041
|
412 |
\quad \QED{} \\
|
wenzelm@12621
|
413 |
\quad \FIX{\vec x}~\ASSUMENAME^\ast~a\colon~\vec\phi \\
|
wenzelm@12621
|
414 |
\end{matharray}
|
wenzelm@12621
|
415 |
|
wenzelm@12621
|
416 |
Typically, the soundness proof is relatively straight-forward, often just by
|
wenzelm@13048
|
417 |
canonical automated tools such as ``$\BY{simp}$'' or ``$\BY{blast}$''.
|
wenzelm@13048
|
418 |
Accordingly, the ``$that$'' reduction above is declared as simplification and
|
wenzelm@13048
|
419 |
introduction rule.
|
wenzelm@12621
|
420 |
|
wenzelm@12621
|
421 |
In a sense, $\OBTAINNAME$ represents at the level of Isar proofs what would be
|
wenzelm@12621
|
422 |
meta-logical existential quantifiers and conjunctions. This concept has a
|
wenzelm@13041
|
423 |
broad range of useful applications, ranging from plain elimination (or
|
wenzelm@17864
|
424 |
introduction) of object-level existential and conjunctions, to elimination
|
wenzelm@12621
|
425 |
over results of symbolic evaluation of recursive definitions, for example.
|
wenzelm@12621
|
426 |
Also note that $\OBTAINNAME$ without parameters acts much like $\HAVENAME$,
|
wenzelm@13041
|
427 |
where the result is treated as a genuine assumption.
|
wenzelm@12621
|
428 |
|
wenzelm@17864
|
429 |
\medskip
|
wenzelm@17864
|
430 |
|
wenzelm@17864
|
431 |
The improper variant $\isarkeyword{guess}$ is similar to $\OBTAINNAME$, but
|
wenzelm@17864
|
432 |
derives the obtained statement from the course of reasoning! The proof starts
|
wenzelm@17864
|
433 |
with a fixed goal $thesis$. The subsequent proof may refine this to anything
|
wenzelm@17864
|
434 |
of the form like $\All{\vec x} \vec\phi \Imp thesis$, but must not introduce
|
wenzelm@17864
|
435 |
new subgoals. The final goal state is then used as reduction rule for the
|
wenzelm@17864
|
436 |
obtain scheme described above. Obtained parameters $\vec x$ are marked as
|
wenzelm@17864
|
437 |
internal by default, which prevents the proof context from being polluted by
|
wenzelm@17864
|
438 |
ad-hoc variables. The variable names and type constraints given as arguments
|
wenzelm@17864
|
439 |
for $\isarkeyword{guess}$ specify a prefix of obtained parameters explicitly
|
wenzelm@17864
|
440 |
in the text.
|
wenzelm@17864
|
441 |
|
wenzelm@17864
|
442 |
It is important to note that the facts introduced by $\OBTAINNAME$ and
|
wenzelm@17864
|
443 |
$\isarkeyword{guess}$ may not be polymorphic: any type-variables occurring
|
wenzelm@17864
|
444 |
here are fixed in the present context!
|
wenzelm@17864
|
445 |
|
wenzelm@12621
|
446 |
|
wenzelm@12621
|
447 |
\subsection{Calculational reasoning}\label{sec:calculation}
|
wenzelm@7315
|
448 |
|
wenzelm@8619
|
449 |
\indexisarcmd{also}\indexisarcmd{finally}
|
wenzelm@8619
|
450 |
\indexisarcmd{moreover}\indexisarcmd{ultimately}
|
wenzelm@12976
|
451 |
\indexisarcmd{print-trans-rules}
|
wenzelm@12976
|
452 |
\indexisaratt{trans}\indexisaratt{sym}\indexisaratt{symmetric}
|
wenzelm@7315
|
453 |
\begin{matharray}{rcl}
|
wenzelm@7315
|
454 |
\isarcmd{also} & : & \isartrans{proof(state)}{proof(state)} \\
|
wenzelm@7315
|
455 |
\isarcmd{finally} & : & \isartrans{proof(state)}{proof(chain)} \\
|
wenzelm@8619
|
456 |
\isarcmd{moreover} & : & \isartrans{proof(state)}{proof(state)} \\
|
wenzelm@8619
|
457 |
\isarcmd{ultimately} & : & \isartrans{proof(state)}{proof(chain)} \\
|
wenzelm@10154
|
458 |
\isarcmd{print_trans_rules}^* & : & \isarkeep{theory~|~proof} \\
|
wenzelm@7315
|
459 |
trans & : & \isaratt \\
|
wenzelm@12976
|
460 |
sym & : & \isaratt \\
|
wenzelm@12976
|
461 |
symmetric & : & \isaratt \\
|
wenzelm@7315
|
462 |
\end{matharray}
|
wenzelm@7315
|
463 |
|
wenzelm@7315
|
464 |
Calculational proof is forward reasoning with implicit application of
|
oheimb@11332
|
465 |
transitivity rules (such those of $=$, $\leq$, $<$). Isabelle/Isar maintains
|
wenzelm@7391
|
466 |
an auxiliary register $calculation$\indexisarthm{calculation} for accumulating
|
wenzelm@7897
|
467 |
results obtained by transitivity composed with the current result. Command
|
wenzelm@7897
|
468 |
$\ALSO$ updates $calculation$ involving $this$, while $\FINALLY$ exhibits the
|
wenzelm@7897
|
469 |
final $calculation$ by forward chaining towards the next goal statement. Both
|
wenzelm@7897
|
470 |
commands require valid current facts, i.e.\ may occur only after commands that
|
wenzelm@7897
|
471 |
produce theorems such as $\ASSUMENAME$, $\NOTENAME$, or some finished proof of
|
wenzelm@8619
|
472 |
$\HAVENAME$, $\SHOWNAME$ etc. The $\MOREOVER$ and $\ULTIMATELY$ commands are
|
wenzelm@8619
|
473 |
similar to $\ALSO$ and $\FINALLY$, but only collect further results in
|
wenzelm@8619
|
474 |
$calculation$ without applying any rules yet.
|
wenzelm@7315
|
475 |
|
wenzelm@13041
|
476 |
Also note that the implicit term abbreviation ``$\dots$'' has its canonical
|
wenzelm@13041
|
477 |
application with calculational proofs. It refers to the argument of the
|
wenzelm@13041
|
478 |
preceding statement. (The argument of a curried infix expression happens to be
|
wenzelm@13041
|
479 |
its right-hand side.)
|
wenzelm@7315
|
480 |
|
wenzelm@7315
|
481 |
Isabelle/Isar calculations are implicitly subject to block structure in the
|
wenzelm@7315
|
482 |
sense that new threads of calculational reasoning are commenced for any new
|
wenzelm@7315
|
483 |
block (as opened by a local goal, for example). This means that, apart from
|
wenzelm@7315
|
484 |
being able to nest calculations, there is no separate \emph{begin-calculation}
|
wenzelm@7315
|
485 |
command required.
|
wenzelm@7315
|
486 |
|
wenzelm@8619
|
487 |
\medskip
|
wenzelm@8619
|
488 |
|
wenzelm@13041
|
489 |
The Isar calculation proof commands may be defined as follows:\footnote{We
|
wenzelm@13041
|
490 |
suppress internal bookkeeping such as proper handling of block-structure.}
|
wenzelm@8619
|
491 |
\begin{matharray}{rcl}
|
wenzelm@8619
|
492 |
\ALSO@0 & \equiv & \NOTE{calculation}{this} \\
|
wenzelm@9606
|
493 |
\ALSO@{n+1} & \equiv & \NOTE{calculation}{trans~[OF~calculation~this]} \\[0.5ex]
|
wenzelm@8619
|
494 |
\FINALLY & \equiv & \ALSO~\FROM{calculation} \\
|
wenzelm@8619
|
495 |
\MOREOVER & \equiv & \NOTE{calculation}{calculation~this} \\
|
wenzelm@8619
|
496 |
\ULTIMATELY & \equiv & \MOREOVER~\FROM{calculation} \\
|
wenzelm@8619
|
497 |
\end{matharray}
|
wenzelm@8619
|
498 |
|
wenzelm@7315
|
499 |
\begin{rail}
|
wenzelm@13024
|
500 |
('also' | 'finally') ('(' thmrefs ')')?
|
wenzelm@8619
|
501 |
;
|
wenzelm@8507
|
502 |
'trans' (() | 'add' | 'del')
|
wenzelm@7315
|
503 |
;
|
wenzelm@7315
|
504 |
\end{rail}
|
wenzelm@7315
|
505 |
|
wenzelm@7315
|
506 |
\begin{descr}
|
wenzelm@13041
|
507 |
|
wenzelm@8547
|
508 |
\item [$\ALSO~(\vec a)$] maintains the auxiliary $calculation$ register as
|
wenzelm@7315
|
509 |
follows. The first occurrence of $\ALSO$ in some calculational thread
|
wenzelm@7905
|
510 |
initializes $calculation$ by $this$. Any subsequent $\ALSO$ on the same
|
wenzelm@7335
|
511 |
level of block-structure updates $calculation$ by some transitivity rule
|
wenzelm@7458
|
512 |
applied to $calculation$ and $this$ (in that order). Transitivity rules are
|
wenzelm@11095
|
513 |
picked from the current context, unless alternative rules are given as
|
wenzelm@11095
|
514 |
explicit arguments.
|
wenzelm@9614
|
515 |
|
wenzelm@8547
|
516 |
\item [$\FINALLY~(\vec a)$] maintaining $calculation$ in the same way as
|
wenzelm@7315
|
517 |
$\ALSO$, and concludes the current calculational thread. The final result
|
wenzelm@7315
|
518 |
is exhibited as fact for forward chaining towards the next goal. Basically,
|
wenzelm@7987
|
519 |
$\FINALLY$ just abbreviates $\ALSO~\FROM{calculation}$. Note that
|
wenzelm@7987
|
520 |
``$\FINALLY~\SHOW{}{\Var{thesis}}~\DOT$'' and
|
wenzelm@7987
|
521 |
``$\FINALLY~\HAVE{}{\phi}~\DOT$'' are typical idioms for concluding
|
wenzelm@7987
|
522 |
calculational proofs.
|
wenzelm@9614
|
523 |
|
wenzelm@8619
|
524 |
\item [$\MOREOVER$ and $\ULTIMATELY$] are analogous to $\ALSO$ and $\FINALLY$,
|
wenzelm@8619
|
525 |
but collect results only, without applying rules.
|
wenzelm@13041
|
526 |
|
wenzelm@13024
|
527 |
\item [$\isarkeyword{print_trans_rules}$] prints the list of transitivity
|
wenzelm@13024
|
528 |
rules (for calculational commands $\ALSO$ and $\FINALLY$) and symmetry rules
|
wenzelm@13024
|
529 |
(for the $symmetric$ operation and single step elimination patters) of the
|
wenzelm@13024
|
530 |
current context.
|
wenzelm@13041
|
531 |
|
wenzelm@8547
|
532 |
\item [$trans$] declares theorems as transitivity rules.
|
wenzelm@13041
|
533 |
|
wenzelm@13024
|
534 |
\item [$sym$] declares symmetry rules.
|
wenzelm@13041
|
535 |
|
wenzelm@12976
|
536 |
\item [$symmetric$] resolves a theorem with some rule declared as $sym$ in the
|
wenzelm@12976
|
537 |
current context. For example, ``$\ASSUME{[symmetric]}{x = y}$'' produces a
|
wenzelm@12976
|
538 |
swapped fact derived from that assumption.
|
wenzelm@13041
|
539 |
|
wenzelm@13024
|
540 |
In structured proof texts it is often more appropriate to use an explicit
|
wenzelm@13024
|
541 |
single-step elimination proof, such as ``$\ASSUME{}{x = y}~\HENCE{}{y =
|
wenzelm@13041
|
542 |
x}~\DDOT$''. The very same rules known to $symmetric$ are declared as
|
wenzelm@13041
|
543 |
$elim?$ as well.
|
wenzelm@13027
|
544 |
|
wenzelm@7315
|
545 |
\end{descr}
|
wenzelm@7315
|
546 |
|
wenzelm@7315
|
547 |
|
wenzelm@13041
|
548 |
\section{Proof tools}
|
wenzelm@8517
|
549 |
|
wenzelm@12618
|
550 |
\subsection{Miscellaneous methods and attributes}\label{sec:misc-meth-att}
|
wenzelm@8517
|
551 |
|
wenzelm@9606
|
552 |
\indexisarmeth{unfold}\indexisarmeth{fold}\indexisarmeth{insert}
|
wenzelm@8517
|
553 |
\indexisarmeth{erule}\indexisarmeth{drule}\indexisarmeth{frule}
|
wenzelm@8517
|
554 |
\indexisarmeth{fail}\indexisarmeth{succeed}
|
wenzelm@8517
|
555 |
\begin{matharray}{rcl}
|
wenzelm@8517
|
556 |
unfold & : & \isarmeth \\
|
wenzelm@10741
|
557 |
fold & : & \isarmeth \\
|
wenzelm@10741
|
558 |
insert & : & \isarmeth \\[0.5ex]
|
wenzelm@8517
|
559 |
erule^* & : & \isarmeth \\
|
wenzelm@8517
|
560 |
drule^* & : & \isarmeth \\
|
wenzelm@13024
|
561 |
frule^* & : & \isarmeth \\
|
wenzelm@8517
|
562 |
succeed & : & \isarmeth \\
|
wenzelm@8517
|
563 |
fail & : & \isarmeth \\
|
wenzelm@8517
|
564 |
\end{matharray}
|
wenzelm@8517
|
565 |
|
wenzelm@8517
|
566 |
\begin{rail}
|
wenzelm@10741
|
567 |
('fold' | 'unfold' | 'insert') thmrefs
|
wenzelm@10741
|
568 |
;
|
wenzelm@10741
|
569 |
('erule' | 'drule' | 'frule') ('('nat')')? thmrefs
|
wenzelm@7135
|
570 |
;
|
wenzelm@7135
|
571 |
\end{rail}
|
wenzelm@7135
|
572 |
|
wenzelm@7167
|
573 |
\begin{descr}
|
wenzelm@13041
|
574 |
|
wenzelm@13024
|
575 |
\item [$unfold~\vec a$ and $fold~\vec a$] expand (or fold back again) the
|
wenzelm@13024
|
576 |
given meta-level definitions throughout all goals; any chained facts
|
wenzelm@13024
|
577 |
provided are inserted into the goal and subject to rewriting as well.
|
wenzelm@13041
|
578 |
|
wenzelm@10741
|
579 |
\item [$insert~\vec a$] inserts theorems as facts into all goals of the proof
|
wenzelm@10741
|
580 |
state. Note that current facts indicated for forward chaining are ignored.
|
wenzelm@13024
|
581 |
|
wenzelm@8547
|
582 |
\item [$erule~\vec a$, $drule~\vec a$, and $frule~\vec a$] are similar to the
|
wenzelm@8547
|
583 |
basic $rule$ method (see \S\ref{sec:pure-meth-att}), but apply rules by
|
wenzelm@8517
|
584 |
elim-resolution, destruct-resolution, and forward-resolution, respectively
|
wenzelm@10741
|
585 |
\cite{isabelle-ref}. The optional natural number argument (default $0$)
|
wenzelm@13041
|
586 |
specifies additional assumption steps to be performed here.
|
wenzelm@13041
|
587 |
|
wenzelm@10741
|
588 |
Note that these methods are improper ones, mainly serving for
|
wenzelm@10741
|
589 |
experimentation and tactic script emulation. Different modes of basic rule
|
wenzelm@10741
|
590 |
application are usually expressed in Isar at the proof language level,
|
wenzelm@10741
|
591 |
rather than via implicit proof state manipulations. For example, a proper
|
wenzelm@13041
|
592 |
single-step elimination would be done using the plain $rule$ method, with
|
wenzelm@10741
|
593 |
forward chaining of current facts.
|
wenzelm@13024
|
594 |
|
wenzelm@8517
|
595 |
\item [$succeed$] yields a single (unchanged) result; it is the identity of
|
wenzelm@8517
|
596 |
the ``\texttt{,}'' method combinator (cf.\ \S\ref{sec:syn-meth}).
|
wenzelm@13024
|
597 |
|
wenzelm@8517
|
598 |
\item [$fail$] yields an empty result sequence; it is the identity of the
|
wenzelm@8517
|
599 |
``\texttt{|}'' method combinator (cf.\ \S\ref{sec:syn-meth}).
|
wenzelm@13024
|
600 |
|
wenzelm@7167
|
601 |
\end{descr}
|
wenzelm@7135
|
602 |
|
wenzelm@10318
|
603 |
\indexisaratt{tagged}\indexisaratt{untagged}
|
wenzelm@9614
|
604 |
\indexisaratt{THEN}\indexisaratt{COMP}
|
ballarin@14175
|
605 |
\indexisaratt{unfolded}\indexisaratt{folded}
|
wenzelm@13027
|
606 |
\indexisaratt{standard}\indexisarattof{Pure}{elim-format}
|
wenzelm@13024
|
607 |
\indexisaratt{no-vars}
|
wenzelm@8517
|
608 |
\begin{matharray}{rcl}
|
wenzelm@9905
|
609 |
tagged & : & \isaratt \\
|
wenzelm@9905
|
610 |
untagged & : & \isaratt \\[0.5ex]
|
wenzelm@9614
|
611 |
THEN & : & \isaratt \\
|
wenzelm@8517
|
612 |
COMP & : & \isaratt \\[0.5ex]
|
wenzelm@9905
|
613 |
unfolded & : & \isaratt \\
|
wenzelm@9905
|
614 |
folded & : & \isaratt \\[0.5ex]
|
wenzelm@9941
|
615 |
elim_format & : & \isaratt \\
|
wenzelm@13041
|
616 |
standard^* & : & \isaratt \\
|
wenzelm@9936
|
617 |
no_vars^* & : & \isaratt \\
|
wenzelm@8517
|
618 |
\end{matharray}
|
wenzelm@8517
|
619 |
|
wenzelm@8517
|
620 |
\begin{rail}
|
wenzelm@9905
|
621 |
'tagged' (nameref+)
|
wenzelm@8517
|
622 |
;
|
wenzelm@9905
|
623 |
'untagged' name
|
wenzelm@8517
|
624 |
;
|
wenzelm@10154
|
625 |
('THEN' | 'COMP') ('[' nat ']')? thmref
|
wenzelm@8517
|
626 |
;
|
wenzelm@9905
|
627 |
('unfolded' | 'folded') thmrefs
|
wenzelm@8517
|
628 |
;
|
wenzelm@8517
|
629 |
\end{rail}
|
wenzelm@8517
|
630 |
|
wenzelm@8517
|
631 |
\begin{descr}
|
wenzelm@13041
|
632 |
|
wenzelm@9905
|
633 |
\item [$tagged~name~args$ and $untagged~name$] add and remove $tags$ of some
|
wenzelm@8517
|
634 |
theorem. Tags may be any list of strings that serve as comment for some
|
wenzelm@8517
|
635 |
tools (e.g.\ $\LEMMANAME$ causes the tag ``$lemma$'' to be added to the
|
wenzelm@8517
|
636 |
result). The first string is considered the tag name, the rest its
|
wenzelm@8517
|
637 |
arguments. Note that untag removes any tags of the same name.
|
wenzelm@13041
|
638 |
|
wenzelm@13041
|
639 |
\item [$THEN~a$ and $COMP~a$] compose rules by resolution. $THEN$ resolves
|
wenzelm@13041
|
640 |
with the first premise of $a$ (an alternative position may be also
|
wenzelm@13041
|
641 |
specified); the $COMP$ version skips the automatic lifting process that is
|
wenzelm@13041
|
642 |
normally intended (cf.\ \texttt{RS} and \texttt{COMP} in
|
wenzelm@8547
|
643 |
\cite[\S5]{isabelle-ref}).
|
wenzelm@13041
|
644 |
|
wenzelm@9905
|
645 |
\item [$unfolded~\vec a$ and $folded~\vec a$] expand and fold back again the
|
wenzelm@9905
|
646 |
given meta-level definitions throughout a rule.
|
wenzelm@13041
|
647 |
|
wenzelm@13027
|
648 |
\item [$elim_format$] turns a destruction rule into elimination rule format,
|
wenzelm@13027
|
649 |
by resolving with the rule $\PROP A \Imp (\PROP A \Imp \PROP B) \Imp \PROP
|
wenzelm@13027
|
650 |
B$.
|
wenzelm@13048
|
651 |
|
wenzelm@13048
|
652 |
Note that the Classical Reasoner (\S\ref{sec:classical}) provides its own
|
wenzelm@13048
|
653 |
version of this operation.
|
wenzelm@13041
|
654 |
|
wenzelm@13041
|
655 |
\item [$standard$] puts a theorem into the standard form of object-rules at
|
wenzelm@13041
|
656 |
the outermost theory level. Note that this operation violates the local
|
wenzelm@13041
|
657 |
proof context (including active locales).
|
wenzelm@13041
|
658 |
|
wenzelm@9232
|
659 |
\item [$no_vars$] replaces schematic variables by free ones; this is mainly
|
wenzelm@9232
|
660 |
for tuning output of pretty printed theorems.
|
wenzelm@13027
|
661 |
|
wenzelm@8517
|
662 |
\end{descr}
|
wenzelm@7135
|
663 |
|
wenzelm@7135
|
664 |
|
wenzelm@12621
|
665 |
\subsection{Further tactic emulations}\label{sec:tactics}
|
wenzelm@9606
|
666 |
|
wenzelm@9606
|
667 |
The following improper proof methods emulate traditional tactics. These admit
|
wenzelm@9606
|
668 |
direct access to the goal state, which is normally considered harmful! In
|
wenzelm@9606
|
669 |
particular, this may involve both numbered goal addressing (default 1), and
|
wenzelm@9606
|
670 |
dynamic instantiation within the scope of some subgoal.
|
wenzelm@9606
|
671 |
|
wenzelm@9606
|
672 |
\begin{warn}
|
ballarin@14175
|
673 |
Dynamic instantiations refer to universally quantified parameters of
|
ballarin@14175
|
674 |
a subgoal (the dynamic context) rather than fixed variables and term
|
ballarin@14175
|
675 |
abbreviations of a (static) Isar context.
|
wenzelm@9606
|
676 |
\end{warn}
|
wenzelm@9606
|
677 |
|
ballarin@14175
|
678 |
Tactic emulation methods, unlike their ML counterparts, admit
|
ballarin@14175
|
679 |
simultaneous instantiation from both dynamic and static contexts. If
|
ballarin@14175
|
680 |
names occur in both contexts goal parameters hide locally fixed
|
ballarin@14175
|
681 |
variables. Likewise, schematic variables refer to term abbreviations,
|
ballarin@14175
|
682 |
if present in the static context. Otherwise the schematic variable is
|
ballarin@14175
|
683 |
interpreted as a schematic variable and left to be solved by unification
|
ballarin@14175
|
684 |
with certain parts of the subgoal.
|
ballarin@14175
|
685 |
|
wenzelm@9606
|
686 |
Note that the tactic emulation proof methods in Isabelle/Isar are consistently
|
ballarin@14175
|
687 |
named $foo_tac$. Note also that variable names occurring on left hand sides
|
ballarin@14212
|
688 |
of instantiations must be preceded by a question mark if they coincide with
|
ballarin@14212
|
689 |
a keyword or contain dots.
|
ballarin@14175
|
690 |
This is consistent with the attribute $where$ (see \S\ref{sec:pure-meth-att}).
|
wenzelm@9606
|
691 |
|
wenzelm@9606
|
692 |
\indexisarmeth{rule-tac}\indexisarmeth{erule-tac}
|
wenzelm@9606
|
693 |
\indexisarmeth{drule-tac}\indexisarmeth{frule-tac}
|
wenzelm@9606
|
694 |
\indexisarmeth{cut-tac}\indexisarmeth{thin-tac}
|
wenzelm@9642
|
695 |
\indexisarmeth{subgoal-tac}\indexisarmeth{rename-tac}
|
wenzelm@9614
|
696 |
\indexisarmeth{rotate-tac}\indexisarmeth{tactic}
|
wenzelm@9606
|
697 |
\begin{matharray}{rcl}
|
wenzelm@9606
|
698 |
rule_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
699 |
erule_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
700 |
drule_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
701 |
frule_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
702 |
cut_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
703 |
thin_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
704 |
subgoal_tac^* & : & \isarmeth \\
|
wenzelm@9614
|
705 |
rename_tac^* & : & \isarmeth \\
|
wenzelm@9614
|
706 |
rotate_tac^* & : & \isarmeth \\
|
wenzelm@9606
|
707 |
tactic^* & : & \isarmeth \\
|
wenzelm@9606
|
708 |
\end{matharray}
|
wenzelm@9606
|
709 |
|
wenzelm@9606
|
710 |
\railalias{ruletac}{rule\_tac}
|
wenzelm@9606
|
711 |
\railterm{ruletac}
|
wenzelm@9606
|
712 |
|
wenzelm@9606
|
713 |
\railalias{eruletac}{erule\_tac}
|
wenzelm@9606
|
714 |
\railterm{eruletac}
|
wenzelm@9606
|
715 |
|
wenzelm@9606
|
716 |
\railalias{druletac}{drule\_tac}
|
wenzelm@9606
|
717 |
\railterm{druletac}
|
wenzelm@9606
|
718 |
|
wenzelm@9606
|
719 |
\railalias{fruletac}{frule\_tac}
|
wenzelm@9606
|
720 |
\railterm{fruletac}
|
wenzelm@9606
|
721 |
|
wenzelm@9606
|
722 |
\railalias{cuttac}{cut\_tac}
|
wenzelm@9606
|
723 |
\railterm{cuttac}
|
wenzelm@9606
|
724 |
|
wenzelm@9606
|
725 |
\railalias{thintac}{thin\_tac}
|
wenzelm@9606
|
726 |
\railterm{thintac}
|
wenzelm@9606
|
727 |
|
wenzelm@9606
|
728 |
\railalias{subgoaltac}{subgoal\_tac}
|
wenzelm@9606
|
729 |
\railterm{subgoaltac}
|
wenzelm@9606
|
730 |
|
wenzelm@9614
|
731 |
\railalias{renametac}{rename\_tac}
|
wenzelm@9614
|
732 |
\railterm{renametac}
|
wenzelm@9614
|
733 |
|
wenzelm@9614
|
734 |
\railalias{rotatetac}{rotate\_tac}
|
wenzelm@9614
|
735 |
\railterm{rotatetac}
|
wenzelm@9614
|
736 |
|
wenzelm@9606
|
737 |
\begin{rail}
|
wenzelm@9606
|
738 |
( ruletac | eruletac | druletac | fruletac | cuttac | thintac ) goalspec?
|
wenzelm@9606
|
739 |
( insts thmref | thmrefs )
|
wenzelm@9606
|
740 |
;
|
wenzelm@9606
|
741 |
subgoaltac goalspec? (prop +)
|
wenzelm@9606
|
742 |
;
|
wenzelm@9614
|
743 |
renametac goalspec? (name +)
|
wenzelm@9614
|
744 |
;
|
wenzelm@9614
|
745 |
rotatetac goalspec? int?
|
wenzelm@9614
|
746 |
;
|
wenzelm@9606
|
747 |
'tactic' text
|
wenzelm@9606
|
748 |
;
|
wenzelm@9606
|
749 |
|
wenzelm@9606
|
750 |
insts: ((name '=' term) + 'and') 'in'
|
wenzelm@9606
|
751 |
;
|
wenzelm@9606
|
752 |
\end{rail}
|
wenzelm@9606
|
753 |
|
wenzelm@9606
|
754 |
\begin{descr}
|
wenzelm@13041
|
755 |
|
wenzelm@9606
|
756 |
\item [$rule_tac$ etc.] do resolution of rules with explicit instantiation.
|
wenzelm@9606
|
757 |
This works the same way as the ML tactics \texttt{res_inst_tac} etc. (see
|
wenzelm@9606
|
758 |
\cite[\S3]{isabelle-ref}).
|
wenzelm@13041
|
759 |
|
wenzelm@13041
|
760 |
Multiple rules may be only given if there is no instantiation; then
|
wenzelm@9606
|
761 |
$rule_tac$ is the same as \texttt{resolve_tac} in ML (see
|
wenzelm@9606
|
762 |
\cite[\S3]{isabelle-ref}).
|
wenzelm@13041
|
763 |
|
wenzelm@9606
|
764 |
\item [$cut_tac$] inserts facts into the proof state as assumption of a
|
wenzelm@9606
|
765 |
subgoal, see also \texttt{cut_facts_tac} in \cite[\S3]{isabelle-ref}. Note
|
wenzelm@13027
|
766 |
that the scope of schematic variables is spread over the main goal
|
wenzelm@13027
|
767 |
statement. Instantiations may be given as well, see also ML tactic
|
wenzelm@9606
|
768 |
\texttt{cut_inst_tac} in \cite[\S3]{isabelle-ref}.
|
wenzelm@13041
|
769 |
|
wenzelm@9606
|
770 |
\item [$thin_tac~\phi$] deletes the specified assumption from a subgoal; note
|
wenzelm@9606
|
771 |
that $\phi$ may contain schematic variables. See also \texttt{thin_tac} in
|
wenzelm@9606
|
772 |
\cite[\S3]{isabelle-ref}.
|
wenzelm@13041
|
773 |
|
wenzelm@9606
|
774 |
\item [$subgoal_tac~\phi$] adds $\phi$ as an assumption to a subgoal. See
|
wenzelm@9606
|
775 |
also \texttt{subgoal_tac} and \texttt{subgoals_tac} in
|
wenzelm@9606
|
776 |
\cite[\S3]{isabelle-ref}.
|
wenzelm@13041
|
777 |
|
wenzelm@9614
|
778 |
\item [$rename_tac~\vec x$] renames parameters of a goal according to the list
|
wenzelm@9614
|
779 |
$\vec x$, which refers to the \emph{suffix} of variables.
|
wenzelm@13041
|
780 |
|
wenzelm@9614
|
781 |
\item [$rotate_tac~n$] rotates the assumptions of a goal by $n$ positions:
|
wenzelm@9614
|
782 |
from right to left if $n$ is positive, and from left to right if $n$ is
|
wenzelm@9614
|
783 |
negative; the default value is $1$. See also \texttt{rotate_tac} in
|
wenzelm@9614
|
784 |
\cite[\S3]{isabelle-ref}.
|
wenzelm@13041
|
785 |
|
wenzelm@9606
|
786 |
\item [$tactic~text$] produces a proof method from any ML text of type
|
wenzelm@9606
|
787 |
\texttt{tactic}. Apart from the usual ML environment and the current
|
wenzelm@9606
|
788 |
implicit theory context, the ML code may refer to the following locally
|
wenzelm@9606
|
789 |
bound values:
|
wenzelm@9606
|
790 |
|
wenzelm@9606
|
791 |
{\footnotesize\begin{verbatim}
|
wenzelm@9606
|
792 |
val ctxt : Proof.context
|
wenzelm@9606
|
793 |
val facts : thm list
|
wenzelm@9606
|
794 |
val thm : string -> thm
|
wenzelm@9606
|
795 |
val thms : string -> thm list
|
wenzelm@9606
|
796 |
\end{verbatim}}
|
wenzelm@9606
|
797 |
Here \texttt{ctxt} refers to the current proof context, \texttt{facts}
|
wenzelm@9606
|
798 |
indicates any current facts for forward-chaining, and
|
wenzelm@9606
|
799 |
\texttt{thm}~/~\texttt{thms} retrieve named facts (including global
|
wenzelm@9606
|
800 |
theorems) from the context.
|
wenzelm@9606
|
801 |
\end{descr}
|
wenzelm@9606
|
802 |
|
wenzelm@9606
|
803 |
|
wenzelm@12621
|
804 |
\subsection{The Simplifier}\label{sec:simplifier}
|
wenzelm@7135
|
805 |
|
wenzelm@13048
|
806 |
\subsubsection{Simplification methods}
|
wenzelm@12618
|
807 |
|
wenzelm@8483
|
808 |
\indexisarmeth{simp}\indexisarmeth{simp-all}
|
wenzelm@7315
|
809 |
\begin{matharray}{rcl}
|
wenzelm@7315
|
810 |
simp & : & \isarmeth \\
|
wenzelm@8483
|
811 |
simp_all & : & \isarmeth \\
|
wenzelm@7315
|
812 |
\end{matharray}
|
wenzelm@7315
|
813 |
|
wenzelm@8483
|
814 |
\railalias{simpall}{simp\_all}
|
wenzelm@8483
|
815 |
\railterm{simpall}
|
wenzelm@8483
|
816 |
|
wenzelm@8704
|
817 |
\railalias{noasm}{no\_asm}
|
wenzelm@8704
|
818 |
\railterm{noasm}
|
wenzelm@8704
|
819 |
|
wenzelm@8704
|
820 |
\railalias{noasmsimp}{no\_asm\_simp}
|
wenzelm@8704
|
821 |
\railterm{noasmsimp}
|
wenzelm@8704
|
822 |
|
wenzelm@8704
|
823 |
\railalias{noasmuse}{no\_asm\_use}
|
wenzelm@8704
|
824 |
\railterm{noasmuse}
|
wenzelm@8704
|
825 |
|
berghofe@13617
|
826 |
\railalias{asmlr}{asm\_lr}
|
berghofe@13617
|
827 |
\railterm{asmlr}
|
berghofe@13617
|
828 |
|
wenzelm@11128
|
829 |
\indexouternonterm{simpmod}
|
wenzelm@7315
|
830 |
\begin{rail}
|
wenzelm@13027
|
831 |
('simp' | simpall) ('!' ?) opt? (simpmod *)
|
wenzelm@7315
|
832 |
;
|
wenzelm@7315
|
833 |
|
berghofe@13617
|
834 |
opt: '(' (noasm | noasmsimp | noasmuse | asmlr) ')'
|
wenzelm@8704
|
835 |
;
|
wenzelm@9711
|
836 |
simpmod: ('add' | 'del' | 'only' | 'cong' (() | 'add' | 'del') |
|
wenzelm@9847
|
837 |
'split' (() | 'add' | 'del')) ':' thmrefs
|
wenzelm@7315
|
838 |
;
|
wenzelm@7315
|
839 |
\end{rail}
|
wenzelm@7315
|
840 |
|
wenzelm@7321
|
841 |
\begin{descr}
|
wenzelm@13015
|
842 |
|
wenzelm@8547
|
843 |
\item [$simp$] invokes Isabelle's simplifier, after declaring additional rules
|
wenzelm@8594
|
844 |
according to the arguments given. Note that the \railtterm{only} modifier
|
wenzelm@8547
|
845 |
first removes all other rewrite rules, congruences, and looper tactics
|
wenzelm@8594
|
846 |
(including splits), and then behaves like \railtterm{add}.
|
wenzelm@13041
|
847 |
|
wenzelm@9711
|
848 |
\medskip The \railtterm{cong} modifiers add or delete Simplifier congruence
|
wenzelm@9711
|
849 |
rules (see also \cite{isabelle-ref}), the default is to add.
|
wenzelm@13041
|
850 |
|
wenzelm@9711
|
851 |
\medskip The \railtterm{split} modifiers add or delete rules for the
|
wenzelm@9711
|
852 |
Splitter (see also \cite{isabelle-ref}), the default is to add. This works
|
wenzelm@9711
|
853 |
only if the Simplifier method has been properly setup to include the
|
wenzelm@9711
|
854 |
Splitter (all major object logics such HOL, HOLCF, FOL, ZF do this already).
|
wenzelm@13041
|
855 |
|
wenzelm@13015
|
856 |
\item [$simp_all$] is similar to $simp$, but acts on all goals (backwards from
|
wenzelm@13015
|
857 |
the last to the first one).
|
wenzelm@13015
|
858 |
|
wenzelm@7321
|
859 |
\end{descr}
|
wenzelm@7321
|
860 |
|
wenzelm@13015
|
861 |
By default the Simplifier methods take local assumptions fully into account,
|
wenzelm@13015
|
862 |
using equational assumptions in the subsequent normalization process, or
|
wenzelm@13024
|
863 |
simplifying assumptions themselves (cf.\ \texttt{asm_full_simp_tac} in
|
wenzelm@13015
|
864 |
\cite[\S10]{isabelle-ref}). In structured proofs this is usually quite well
|
wenzelm@13015
|
865 |
behaved in practice: just the local premises of the actual goal are involved,
|
wenzelm@13041
|
866 |
additional facts may be inserted via explicit forward-chaining (using $\THEN$,
|
wenzelm@13015
|
867 |
$\FROMNAME$ etc.). The full context of assumptions is only included if the
|
wenzelm@13015
|
868 |
``$!$'' (bang) argument is given, which should be used with some care, though.
|
wenzelm@7321
|
869 |
|
wenzelm@13015
|
870 |
Additional Simplifier options may be specified to tune the behavior further
|
wenzelm@13041
|
871 |
(mostly for unstructured scripts with many accidental local facts):
|
wenzelm@13041
|
872 |
``$(no_asm)$'' means assumptions are ignored completely (cf.\
|
wenzelm@13041
|
873 |
\texttt{simp_tac}), ``$(no_asm_simp)$'' means assumptions are used in the
|
wenzelm@13041
|
874 |
simplification of the conclusion but are not themselves simplified (cf.\
|
wenzelm@13041
|
875 |
\texttt{asm_simp_tac}), and ``$(no_asm_use)$'' means assumptions are
|
wenzelm@13041
|
876 |
simplified but are not used in the simplification of each other or the
|
wenzelm@13041
|
877 |
conclusion (cf.\ \texttt{full_simp_tac}).
|
berghofe@13617
|
878 |
For compatibility reasons, there is also an option ``$(asm_lr)$'',
|
berghofe@13617
|
879 |
which means that an assumption is only used for simplifying assumptions
|
berghofe@13617
|
880 |
which are to the right of it (cf.\ \texttt{asm_lr_simp_tac}).
|
wenzelm@8704
|
881 |
|
wenzelm@8704
|
882 |
\medskip
|
wenzelm@8704
|
883 |
|
wenzelm@8704
|
884 |
The Splitter package is usually configured to work as part of the Simplifier.
|
wenzelm@9711
|
885 |
The effect of repeatedly applying \texttt{split_tac} can be simulated by
|
wenzelm@13041
|
886 |
``$(simp~only\colon~split\colon~\vec a)$''. There is also a separate $split$
|
wenzelm@13041
|
887 |
method available for single-step case splitting.
|
wenzelm@8483
|
888 |
|
wenzelm@8483
|
889 |
|
wenzelm@12621
|
890 |
\subsubsection{Declaring rules}
|
wenzelm@8483
|
891 |
|
wenzelm@8667
|
892 |
\indexisarcmd{print-simpset}
|
wenzelm@8638
|
893 |
\indexisaratt{simp}\indexisaratt{split}\indexisaratt{cong}
|
wenzelm@7321
|
894 |
\begin{matharray}{rcl}
|
wenzelm@13024
|
895 |
\isarcmd{print_simpset}^* & : & \isarkeep{theory~|~proof} \\
|
wenzelm@7321
|
896 |
simp & : & \isaratt \\
|
wenzelm@9711
|
897 |
cong & : & \isaratt \\
|
wenzelm@8483
|
898 |
split & : & \isaratt \\
|
wenzelm@7321
|
899 |
\end{matharray}
|
wenzelm@7321
|
900 |
|
wenzelm@7321
|
901 |
\begin{rail}
|
wenzelm@9711
|
902 |
('simp' | 'cong' | 'split') (() | 'add' | 'del')
|
wenzelm@7321
|
903 |
;
|
wenzelm@7321
|
904 |
\end{rail}
|
wenzelm@7321
|
905 |
|
wenzelm@7321
|
906 |
\begin{descr}
|
wenzelm@13024
|
907 |
|
wenzelm@13024
|
908 |
\item [$\isarcmd{print_simpset}$] prints the collection of rules declared to
|
wenzelm@13024
|
909 |
the Simplifier, which is also known as ``simpset'' internally
|
wenzelm@8667
|
910 |
\cite{isabelle-ref}. This is a diagnostic command; $undo$ does not apply.
|
wenzelm@13024
|
911 |
|
wenzelm@8547
|
912 |
\item [$simp$] declares simplification rules.
|
wenzelm@13024
|
913 |
|
wenzelm@8638
|
914 |
\item [$cong$] declares congruence rules.
|
wenzelm@13024
|
915 |
|
wenzelm@9711
|
916 |
\item [$split$] declares case split rules.
|
wenzelm@13024
|
917 |
|
wenzelm@7321
|
918 |
\end{descr}
|
wenzelm@7319
|
919 |
|
wenzelm@7315
|
920 |
|
wenzelm@12621
|
921 |
\subsubsection{Forward simplification}
|
wenzelm@12621
|
922 |
|
wenzelm@9905
|
923 |
\indexisaratt{simplified}
|
wenzelm@7315
|
924 |
\begin{matharray}{rcl}
|
wenzelm@9905
|
925 |
simplified & : & \isaratt \\
|
wenzelm@7315
|
926 |
\end{matharray}
|
wenzelm@7315
|
927 |
|
wenzelm@9905
|
928 |
\begin{rail}
|
wenzelm@13015
|
929 |
'simplified' opt? thmrefs?
|
wenzelm@9905
|
930 |
;
|
wenzelm@7905
|
931 |
|
wenzelm@9905
|
932 |
opt: '(' (noasm | noasmsimp | noasmuse) ')'
|
wenzelm@9905
|
933 |
;
|
wenzelm@9905
|
934 |
\end{rail}
|
wenzelm@9905
|
935 |
|
wenzelm@9905
|
936 |
\begin{descr}
|
wenzelm@13048
|
937 |
|
wenzelm@13015
|
938 |
\item [$simplified~\vec a$] causes a theorem to be simplified, either by
|
wenzelm@13015
|
939 |
exactly the specified rules $\vec a$, or the implicit Simplifier context if
|
wenzelm@13015
|
940 |
no arguments are given. The result is fully simplified by default,
|
wenzelm@13015
|
941 |
including assumptions and conclusion; the options $no_asm$ etc.\ tune the
|
wenzelm@13048
|
942 |
Simplifier in the same way as the for the $simp$ method.
|
wenzelm@13041
|
943 |
|
wenzelm@13015
|
944 |
Note that forward simplification restricts the simplifier to its most basic
|
wenzelm@13015
|
945 |
operation of term rewriting; solver and looper tactics \cite{isabelle-ref}
|
wenzelm@13015
|
946 |
are \emph{not} involved here. The $simplified$ attribute should be only
|
wenzelm@13015
|
947 |
rarely required under normal circumstances.
|
wenzelm@13015
|
948 |
|
wenzelm@9905
|
949 |
\end{descr}
|
wenzelm@7315
|
950 |
|
wenzelm@7315
|
951 |
|
wenzelm@13048
|
952 |
\subsubsection{Low-level equational reasoning}
|
wenzelm@9614
|
953 |
|
wenzelm@12976
|
954 |
\indexisarmeth{subst}\indexisarmeth{hypsubst}\indexisarmeth{split}
|
wenzelm@9614
|
955 |
\begin{matharray}{rcl}
|
wenzelm@13015
|
956 |
subst^* & : & \isarmeth \\
|
wenzelm@9614
|
957 |
hypsubst^* & : & \isarmeth \\
|
wenzelm@13015
|
958 |
split^* & : & \isarmeth \\
|
wenzelm@9614
|
959 |
\end{matharray}
|
wenzelm@9614
|
960 |
|
wenzelm@9614
|
961 |
\begin{rail}
|
nipkow@15995
|
962 |
'subst' ('(' 'asm' ')')? ('(' (nat+) ')')? thmref
|
wenzelm@9614
|
963 |
;
|
wenzelm@9799
|
964 |
'split' ('(' 'asm' ')')? thmrefs
|
wenzelm@9703
|
965 |
;
|
wenzelm@9614
|
966 |
\end{rail}
|
wenzelm@9614
|
967 |
|
wenzelm@13015
|
968 |
These methods provide low-level facilities for equational reasoning that are
|
wenzelm@13015
|
969 |
intended for specialized applications only. Normally, single step
|
wenzelm@13015
|
970 |
calculations would be performed in a structured text (see also
|
wenzelm@13015
|
971 |
\S\ref{sec:calculation}), while the Simplifier methods provide the canonical
|
wenzelm@13015
|
972 |
way for automated normalization (see \S\ref{sec:simplifier}).
|
wenzelm@9614
|
973 |
|
wenzelm@9614
|
974 |
\begin{descr}
|
wenzelm@13041
|
975 |
|
nipkow@15995
|
976 |
\item [$subst~eq$] performs a single substitution step using rule $eq$, which
|
wenzelm@13041
|
977 |
may be either a meta or object equality.
|
wenzelm@13041
|
978 |
|
nipkow@15995
|
979 |
\item [$subst~(asm)~eq$] substitutes in an assumption.
|
nipkow@15995
|
980 |
|
nipkow@15995
|
981 |
\item [$subst~(i \dots j)~eq$] performs several substitutions in the
|
nipkow@15995
|
982 |
conclusion. The numbers $i$ to $j$ indicate the positions to substitute at.
|
nipkow@15995
|
983 |
Positions are ordered from the top of the term tree moving down from left to
|
nipkow@15995
|
984 |
right. For example, in $(a+b)+(c+d)$ there are three positions where
|
nipkow@15995
|
985 |
commutativity of $+$ is applicable: 1 refers to the whole term, 2 to $a+b$
|
nipkow@15995
|
986 |
and 3 to $c+d$. If the positions in the list $(i \dots j)$ are
|
nipkow@15995
|
987 |
non-overlapping (e.g. $(2~3)$ in $(a+b)+(c+d)$) you may assume all
|
nipkow@15995
|
988 |
substitutions are performed simultaneously. Otherwise the behaviour of
|
nipkow@15995
|
989 |
$subst$ is not specified.
|
nipkow@15995
|
990 |
|
nipkow@15995
|
991 |
\item [$subst~(asm)~(i \dots j)~eq$] performs the substitutions in the
|
nipkow@16010
|
992 |
assumptions. Positions $1 \dots i@1$ refer
|
nipkow@16010
|
993 |
to assumption 1, positions $i@1+1 \dots i@2$ to assumption 2, and so on.
|
nipkow@15995
|
994 |
|
wenzelm@13041
|
995 |
\item [$hypsubst$] performs substitution using some assumption; this only
|
wenzelm@13041
|
996 |
works for equations of the form $x = t$ where $x$ is a free or bound
|
wenzelm@13041
|
997 |
variable.
|
wenzelm@13041
|
998 |
|
wenzelm@13041
|
999 |
\item [$split~\vec a$] performs single-step case splitting using rules $thms$.
|
wenzelm@9799
|
1000 |
By default, splitting is performed in the conclusion of a goal; the $asm$
|
wenzelm@9799
|
1001 |
option indicates to operate on assumptions instead.
|
wenzelm@13048
|
1002 |
|
wenzelm@9703
|
1003 |
Note that the $simp$ method already involves repeated application of split
|
wenzelm@13048
|
1004 |
rules as declared in the current context.
|
wenzelm@9614
|
1005 |
\end{descr}
|
wenzelm@9614
|
1006 |
|
wenzelm@9614
|
1007 |
|
wenzelm@12621
|
1008 |
\subsection{The Classical Reasoner}\label{sec:classical}
|
wenzelm@7135
|
1009 |
|
wenzelm@13048
|
1010 |
\subsubsection{Basic methods}
|
wenzelm@7315
|
1011 |
|
wenzelm@13024
|
1012 |
\indexisarmeth{rule}\indexisarmeth{default}\indexisarmeth{contradiction}
|
wenzelm@13024
|
1013 |
\indexisarmeth{intro}\indexisarmeth{elim}
|
wenzelm@7321
|
1014 |
\begin{matharray}{rcl}
|
wenzelm@7321
|
1015 |
rule & : & \isarmeth \\
|
wenzelm@13024
|
1016 |
contradiction & : & \isarmeth \\
|
wenzelm@7321
|
1017 |
intro & : & \isarmeth \\
|
wenzelm@7321
|
1018 |
elim & : & \isarmeth \\
|
wenzelm@7321
|
1019 |
\end{matharray}
|
wenzelm@7321
|
1020 |
|
wenzelm@7321
|
1021 |
\begin{rail}
|
wenzelm@8547
|
1022 |
('rule' | 'intro' | 'elim') thmrefs?
|
wenzelm@7321
|
1023 |
;
|
wenzelm@7321
|
1024 |
\end{rail}
|
wenzelm@7321
|
1025 |
|
wenzelm@7321
|
1026 |
\begin{descr}
|
wenzelm@13041
|
1027 |
|
wenzelm@7466
|
1028 |
\item [$rule$] as offered by the classical reasoner is a refinement over the
|
wenzelm@13024
|
1029 |
primitive one (see \S\ref{sec:pure-meth-att}). Both versions essentially
|
wenzelm@13024
|
1030 |
work the same, but the classical version observes the classical rule context
|
wenzelm@13041
|
1031 |
in addition to that of Isabelle/Pure.
|
wenzelm@13041
|
1032 |
|
wenzelm@13041
|
1033 |
Common object logics (HOL, ZF, etc.) declare a rich collection of classical
|
wenzelm@13041
|
1034 |
rules (even if these would qualify as intuitionistic ones), but only few
|
wenzelm@13041
|
1035 |
declarations to the rule context of Isabelle/Pure
|
wenzelm@13041
|
1036 |
(\S\ref{sec:pure-meth-att}).
|
wenzelm@13041
|
1037 |
|
wenzelm@13024
|
1038 |
\item [$contradiction$] solves some goal by contradiction, deriving any result
|
wenzelm@13041
|
1039 |
from both $\neg A$ and $A$. Chained facts, which are guaranteed to
|
wenzelm@13041
|
1040 |
participate, may appear in either order.
|
wenzelm@9614
|
1041 |
|
wenzelm@7466
|
1042 |
\item [$intro$ and $elim$] repeatedly refine some goal by intro- or
|
wenzelm@13041
|
1043 |
elim-resolution, after having inserted any chained facts. Exactly the rules
|
wenzelm@13041
|
1044 |
given as arguments are taken into account; this allows fine-tuned
|
wenzelm@13041
|
1045 |
decomposition of a proof problem, in contrast to common automated tools.
|
wenzelm@13041
|
1046 |
|
wenzelm@7321
|
1047 |
\end{descr}
|
wenzelm@7321
|
1048 |
|
wenzelm@7321
|
1049 |
|
wenzelm@13048
|
1050 |
\subsubsection{Automated methods}
|
wenzelm@7321
|
1051 |
|
wenzelm@9799
|
1052 |
\indexisarmeth{blast}\indexisarmeth{fast}\indexisarmeth{slow}
|
wenzelm@9799
|
1053 |
\indexisarmeth{best}\indexisarmeth{safe}\indexisarmeth{clarify}
|
wenzelm@7321
|
1054 |
\begin{matharray}{rcl}
|
wenzelm@9780
|
1055 |
blast & : & \isarmeth \\
|
wenzelm@9780
|
1056 |
fast & : & \isarmeth \\
|
wenzelm@9799
|
1057 |
slow & : & \isarmeth \\
|
wenzelm@9780
|
1058 |
best & : & \isarmeth \\
|
wenzelm@9780
|
1059 |
safe & : & \isarmeth \\
|
wenzelm@9780
|
1060 |
clarify & : & \isarmeth \\
|
wenzelm@7321
|
1061 |
\end{matharray}
|
wenzelm@7321
|
1062 |
|
wenzelm@11128
|
1063 |
\indexouternonterm{clamod}
|
wenzelm@7321
|
1064 |
\begin{rail}
|
wenzelm@13027
|
1065 |
'blast' ('!' ?) nat? (clamod *)
|
wenzelm@7321
|
1066 |
;
|
wenzelm@13027
|
1067 |
('fast' | 'slow' | 'best' | 'safe' | 'clarify') ('!' ?) (clamod *)
|
wenzelm@7321
|
1068 |
;
|
wenzelm@7321
|
1069 |
|
wenzelm@9408
|
1070 |
clamod: (('intro' | 'elim' | 'dest') ('!' | () | '?') | 'del') ':' thmrefs
|
wenzelm@7321
|
1071 |
;
|
wenzelm@7321
|
1072 |
\end{rail}
|
wenzelm@7321
|
1073 |
|
wenzelm@7321
|
1074 |
\begin{descr}
|
wenzelm@7321
|
1075 |
\item [$blast$] refers to the classical tableau prover (see \texttt{blast_tac}
|
wenzelm@7335
|
1076 |
in \cite[\S11]{isabelle-ref}). The optional argument specifies a
|
wenzelm@10858
|
1077 |
user-supplied search bound (default 20).
|
wenzelm@9799
|
1078 |
\item [$fast$, $slow$, $best$, $safe$, and $clarify$] refer to the generic
|
wenzelm@9799
|
1079 |
classical reasoner. See \texttt{fast_tac}, \texttt{slow_tac},
|
wenzelm@9799
|
1080 |
\texttt{best_tac}, \texttt{safe_tac}, and \texttt{clarify_tac} in
|
wenzelm@9799
|
1081 |
\cite[\S11]{isabelle-ref} for more information.
|
wenzelm@7321
|
1082 |
\end{descr}
|
wenzelm@7321
|
1083 |
|
wenzelm@13041
|
1084 |
Any of the above methods support additional modifiers of the context of
|
wenzelm@13041
|
1085 |
classical rules. Their semantics is analogous to the attributes given before.
|
wenzelm@13041
|
1086 |
Facts provided by forward chaining are inserted into the goal before
|
wenzelm@13041
|
1087 |
commencing proof search. The ``!''~argument causes the full context of
|
wenzelm@13041
|
1088 |
assumptions to be included as well.
|
wenzelm@7321
|
1089 |
|
wenzelm@7315
|
1090 |
|
wenzelm@12621
|
1091 |
\subsubsection{Combined automated methods}\label{sec:clasimp}
|
wenzelm@7315
|
1092 |
|
wenzelm@9799
|
1093 |
\indexisarmeth{auto}\indexisarmeth{force}\indexisarmeth{clarsimp}
|
wenzelm@9799
|
1094 |
\indexisarmeth{fastsimp}\indexisarmeth{slowsimp}\indexisarmeth{bestsimp}
|
wenzelm@7321
|
1095 |
\begin{matharray}{rcl}
|
wenzelm@9606
|
1096 |
auto & : & \isarmeth \\
|
wenzelm@7321
|
1097 |
force & : & \isarmeth \\
|
wenzelm@9438
|
1098 |
clarsimp & : & \isarmeth \\
|
wenzelm@9606
|
1099 |
fastsimp & : & \isarmeth \\
|
wenzelm@9799
|
1100 |
slowsimp & : & \isarmeth \\
|
wenzelm@9799
|
1101 |
bestsimp & : & \isarmeth \\
|
wenzelm@7321
|
1102 |
\end{matharray}
|
wenzelm@7315
|
1103 |
|
wenzelm@11128
|
1104 |
\indexouternonterm{clasimpmod}
|
wenzelm@7321
|
1105 |
\begin{rail}
|
wenzelm@13027
|
1106 |
'auto' '!'? (nat nat)? (clasimpmod *)
|
wenzelm@9780
|
1107 |
;
|
wenzelm@13027
|
1108 |
('force' | 'clarsimp' | 'fastsimp' | 'slowsimp' | 'bestsimp') '!'? (clasimpmod *)
|
wenzelm@7321
|
1109 |
;
|
wenzelm@7315
|
1110 |
|
wenzelm@9711
|
1111 |
clasimpmod: ('simp' (() | 'add' | 'del' | 'only') |
|
wenzelm@10031
|
1112 |
('cong' | 'split') (() | 'add' | 'del') |
|
wenzelm@10031
|
1113 |
'iff' (((() | 'add') '?'?) | 'del') |
|
wenzelm@9408
|
1114 |
(('intro' | 'elim' | 'dest') ('!' | () | '?') | 'del')) ':' thmrefs
|
wenzelm@7321
|
1115 |
\end{rail}
|
wenzelm@7135
|
1116 |
|
wenzelm@7321
|
1117 |
\begin{descr}
|
wenzelm@9799
|
1118 |
\item [$auto$, $force$, $clarsimp$, $fastsimp$, $slowsimp$, and $bestsimp$]
|
wenzelm@9799
|
1119 |
provide access to Isabelle's combined simplification and classical reasoning
|
wenzelm@9799
|
1120 |
tactics. These correspond to \texttt{auto_tac}, \texttt{force_tac},
|
wenzelm@9799
|
1121 |
\texttt{clarsimp_tac}, and Classical Reasoner tactics with the Simplifier
|
wenzelm@9799
|
1122 |
added as wrapper, see \cite[\S11]{isabelle-ref} for more information. The
|
wenzelm@13048
|
1123 |
modifier arguments correspond to those given in \S\ref{sec:simplifier} and
|
wenzelm@13048
|
1124 |
\S\ref{sec:classical}. Just note that the ones related to the Simplifier
|
wenzelm@13048
|
1125 |
are prefixed by \railtterm{simp} here.
|
wenzelm@9614
|
1126 |
|
wenzelm@7987
|
1127 |
Facts provided by forward chaining are inserted into the goal before doing
|
wenzelm@7987
|
1128 |
the search. The ``!''~argument causes the full context of assumptions to be
|
wenzelm@7987
|
1129 |
included as well.
|
wenzelm@7321
|
1130 |
\end{descr}
|
wenzelm@7135
|
1131 |
|
wenzelm@7987
|
1132 |
|
wenzelm@13048
|
1133 |
\subsubsection{Declaring rules}
|
wenzelm@7135
|
1134 |
|
wenzelm@8667
|
1135 |
\indexisarcmd{print-claset}
|
wenzelm@7391
|
1136 |
\indexisaratt{intro}\indexisaratt{elim}\indexisaratt{dest}
|
wenzelm@9936
|
1137 |
\indexisaratt{iff}\indexisaratt{rule}
|
wenzelm@7321
|
1138 |
\begin{matharray}{rcl}
|
wenzelm@13024
|
1139 |
\isarcmd{print_claset}^* & : & \isarkeep{theory~|~proof} \\
|
wenzelm@7321
|
1140 |
intro & : & \isaratt \\
|
wenzelm@7321
|
1141 |
elim & : & \isaratt \\
|
wenzelm@7321
|
1142 |
dest & : & \isaratt \\
|
wenzelm@9936
|
1143 |
rule & : & \isaratt \\
|
wenzelm@7391
|
1144 |
iff & : & \isaratt \\
|
wenzelm@7321
|
1145 |
\end{matharray}
|
wenzelm@7321
|
1146 |
|
wenzelm@7321
|
1147 |
\begin{rail}
|
wenzelm@9408
|
1148 |
('intro' | 'elim' | 'dest') ('!' | () | '?')
|
wenzelm@7321
|
1149 |
;
|
wenzelm@9936
|
1150 |
'rule' 'del'
|
wenzelm@9936
|
1151 |
;
|
wenzelm@10031
|
1152 |
'iff' (((() | 'add') '?'?) | 'del')
|
wenzelm@9936
|
1153 |
;
|
wenzelm@7321
|
1154 |
\end{rail}
|
wenzelm@7321
|
1155 |
|
wenzelm@7321
|
1156 |
\begin{descr}
|
wenzelm@13024
|
1157 |
|
wenzelm@13024
|
1158 |
\item [$\isarcmd{print_claset}$] prints the collection of rules declared to
|
wenzelm@13024
|
1159 |
the Classical Reasoner, which is also known as ``simpset'' internally
|
wenzelm@8667
|
1160 |
\cite{isabelle-ref}. This is a diagnostic command; $undo$ does not apply.
|
wenzelm@13024
|
1161 |
|
wenzelm@8517
|
1162 |
\item [$intro$, $elim$, and $dest$] declare introduction, elimination, and
|
oheimb@11332
|
1163 |
destruction rules, respectively. By default, rules are considered as
|
wenzelm@9408
|
1164 |
\emph{unsafe} (i.e.\ not applied blindly without backtracking), while a
|
wenzelm@13041
|
1165 |
single ``!'' classifies as \emph{safe}. Rule declarations marked by ``?''
|
wenzelm@13041
|
1166 |
coincide with those of Isabelle/Pure, cf.\ \S\ref{sec:pure-meth-att} (i.e.\
|
wenzelm@13041
|
1167 |
are only applied in single steps of the $rule$ method).
|
wenzelm@13024
|
1168 |
|
oheimb@11332
|
1169 |
\item [$rule~del$] deletes introduction, elimination, or destruction rules from
|
wenzelm@9936
|
1170 |
the context.
|
wenzelm@13041
|
1171 |
|
wenzelm@13041
|
1172 |
\item [$iff$] declares logical equivalences to the Simplifier and the
|
wenzelm@13024
|
1173 |
Classical reasoner at the same time. Non-conditional rules result in a
|
wenzelm@13024
|
1174 |
``safe'' introduction and elimination pair; conditional ones are considered
|
wenzelm@13024
|
1175 |
``unsafe''. Rules with negative conclusion are automatically inverted
|
wenzelm@13041
|
1176 |
(using $\neg$ elimination internally).
|
wenzelm@13041
|
1177 |
|
wenzelm@13041
|
1178 |
The ``?'' version of $iff$ declares rules to the Isabelle/Pure context only,
|
wenzelm@13041
|
1179 |
and omits the Simplifier declaration.
|
wenzelm@13041
|
1180 |
|
wenzelm@7321
|
1181 |
\end{descr}
|
wenzelm@7135
|
1182 |
|
wenzelm@8203
|
1183 |
|
wenzelm@13048
|
1184 |
\subsubsection{Classical operations}
|
wenzelm@13027
|
1185 |
|
wenzelm@13027
|
1186 |
\indexisaratt{elim-format}\indexisaratt{swapped}
|
wenzelm@13027
|
1187 |
|
wenzelm@13027
|
1188 |
\begin{matharray}{rcl}
|
wenzelm@13027
|
1189 |
elim_format & : & \isaratt \\
|
wenzelm@13027
|
1190 |
swapped & : & \isaratt \\
|
wenzelm@13027
|
1191 |
\end{matharray}
|
wenzelm@13027
|
1192 |
|
wenzelm@13027
|
1193 |
\begin{descr}
|
wenzelm@13041
|
1194 |
|
wenzelm@13027
|
1195 |
\item [$elim_format$] turns a destruction rule into elimination rule format;
|
wenzelm@13027
|
1196 |
this operation is similar to the the intuitionistic version
|
wenzelm@13027
|
1197 |
(\S\ref{sec:misc-meth-att}), but each premise of the resulting rule acquires
|
wenzelm@13041
|
1198 |
an additional local fact of the negated main thesis; according to the
|
wenzelm@13027
|
1199 |
classical principle $(\neg A \Imp A) \Imp A$.
|
wenzelm@13041
|
1200 |
|
wenzelm@13027
|
1201 |
\item [$swapped$] turns an introduction rule into an elimination, by resolving
|
wenzelm@13027
|
1202 |
with the classical swap principle $(\neg B \Imp A) \Imp (\neg A \Imp B)$.
|
wenzelm@13027
|
1203 |
|
wenzelm@13027
|
1204 |
\end{descr}
|
wenzelm@13027
|
1205 |
|
wenzelm@13027
|
1206 |
|
wenzelm@12621
|
1207 |
\subsection{Proof by cases and induction}\label{sec:cases-induct}
|
wenzelm@11691
|
1208 |
|
wenzelm@13048
|
1209 |
\subsubsection{Rule contexts}
|
wenzelm@12618
|
1210 |
|
wenzelm@12618
|
1211 |
\indexisarcmd{case}\indexisarcmd{print-cases}
|
wenzelm@12618
|
1212 |
\indexisaratt{case-names}\indexisaratt{params}\indexisaratt{consumes}
|
wenzelm@12618
|
1213 |
\begin{matharray}{rcl}
|
wenzelm@12618
|
1214 |
\isarcmd{case} & : & \isartrans{proof(state)}{proof(state)} \\
|
wenzelm@12618
|
1215 |
\isarcmd{print_cases}^* & : & \isarkeep{proof} \\
|
wenzelm@12618
|
1216 |
case_names & : & \isaratt \\
|
wenzelm@12618
|
1217 |
params & : & \isaratt \\
|
wenzelm@12618
|
1218 |
consumes & : & \isaratt \\
|
wenzelm@12618
|
1219 |
\end{matharray}
|
wenzelm@12618
|
1220 |
|
wenzelm@12618
|
1221 |
Basically, Isar proof contexts are built up explicitly using commands like
|
wenzelm@12618
|
1222 |
$\FIXNAME$, $\ASSUMENAME$ etc.\ (see \S\ref{sec:proof-context}). In typical
|
wenzelm@12618
|
1223 |
verification tasks this can become hard to manage, though. In particular, a
|
wenzelm@12618
|
1224 |
large number of local contexts may emerge from case analysis or induction over
|
wenzelm@12618
|
1225 |
inductive sets and types.
|
wenzelm@12618
|
1226 |
|
wenzelm@12618
|
1227 |
\medskip
|
wenzelm@12618
|
1228 |
|
wenzelm@12618
|
1229 |
The $\CASENAME$ command provides a shorthand to refer to certain parts of
|
wenzelm@12618
|
1230 |
logical context symbolically. Proof methods may provide an environment of
|
wenzelm@12618
|
1231 |
named ``cases'' of the form $c\colon \vec x, \vec \phi$. Then the effect of
|
wenzelm@13041
|
1232 |
``$\CASE{c}$'' is that of ``$\FIX{\vec x}~\ASSUME{c}{\vec\phi}$''. Term
|
wenzelm@13041
|
1233 |
bindings may be covered as well, such as $\Var{case}$ for the intended
|
wenzelm@13041
|
1234 |
conclusion.
|
wenzelm@12618
|
1235 |
|
wenzelm@13027
|
1236 |
Normally the ``terminology'' of a case value (i.e.\ the parameters $\vec x$)
|
wenzelm@13041
|
1237 |
are marked as hidden. Using the explicit form ``$\CASE{(c~\vec x)}$'' enables
|
wenzelm@13041
|
1238 |
proof writers to choose their own names for the subsequent proof text.
|
wenzelm@12618
|
1239 |
|
wenzelm@12618
|
1240 |
\medskip
|
wenzelm@12618
|
1241 |
|
wenzelm@13027
|
1242 |
It is important to note that $\CASENAME$ does \emph{not} provide direct means
|
wenzelm@13027
|
1243 |
to peek at the current goal state, which is generally considered
|
wenzelm@13027
|
1244 |
non-observable in Isar. The text of the cases basically emerge from standard
|
wenzelm@13027
|
1245 |
elimination or induction rules, which in turn are derived from previous theory
|
wenzelm@13041
|
1246 |
specifications in a canonical way (say from $\isarkeyword{inductive}$
|
wenzelm@13041
|
1247 |
definitions).
|
wenzelm@13027
|
1248 |
|
wenzelm@12618
|
1249 |
Named cases may be exhibited in the current proof context only if both the
|
wenzelm@12618
|
1250 |
proof method and the rules involved support this. Case names and parameters
|
wenzelm@12618
|
1251 |
of basic rules may be declared by hand as well, by using appropriate
|
wenzelm@12618
|
1252 |
attributes. Thus variant versions of rules that have been derived manually
|
wenzelm@12618
|
1253 |
may be used in advanced case analysis later.
|
wenzelm@12618
|
1254 |
|
wenzelm@12618
|
1255 |
\railalias{casenames}{case\_names}
|
wenzelm@12618
|
1256 |
\railterm{casenames}
|
wenzelm@12618
|
1257 |
|
wenzelm@12618
|
1258 |
\begin{rail}
|
wenzelm@13041
|
1259 |
'case' (caseref | '(' caseref ((name | underscore) +) ')')
|
wenzelm@12618
|
1260 |
;
|
wenzelm@13024
|
1261 |
caseref: nameref attributes?
|
wenzelm@13024
|
1262 |
;
|
wenzelm@13024
|
1263 |
|
wenzelm@13027
|
1264 |
casenames (name +)
|
wenzelm@12618
|
1265 |
;
|
wenzelm@13027
|
1266 |
'params' ((name *) + 'and')
|
wenzelm@12618
|
1267 |
;
|
wenzelm@12618
|
1268 |
'consumes' nat?
|
wenzelm@12618
|
1269 |
;
|
wenzelm@12618
|
1270 |
\end{rail}
|
wenzelm@12618
|
1271 |
|
wenzelm@12618
|
1272 |
\begin{descr}
|
wenzelm@13041
|
1273 |
|
wenzelm@13041
|
1274 |
\item [$\CASE{(c~\vec x)}$] invokes a named local context $c\colon \vec x,
|
wenzelm@13041
|
1275 |
\vec \phi$, as provided by an appropriate proof method (such as $cases$ and
|
wenzelm@13041
|
1276 |
$induct$, see \S\ref{sec:cases-induct-meth}). The command ``$\CASE{(c~\vec
|
wenzelm@13041
|
1277 |
x)}$'' abbreviates ``$\FIX{\vec x}~\ASSUME{c}{\vec\phi}$''.
|
wenzelm@13041
|
1278 |
|
wenzelm@12618
|
1279 |
\item [$\isarkeyword{print_cases}$] prints all local contexts of the current
|
wenzelm@12618
|
1280 |
state, using Isar proof language notation. This is a diagnostic command;
|
wenzelm@12618
|
1281 |
$undo$ does not apply.
|
wenzelm@13041
|
1282 |
|
wenzelm@12618
|
1283 |
\item [$case_names~\vec c$] declares names for the local contexts of premises
|
wenzelm@12618
|
1284 |
of some theorem; $\vec c$ refers to the \emph{suffix} of the list of
|
wenzelm@12618
|
1285 |
premises.
|
wenzelm@13041
|
1286 |
|
wenzelm@12618
|
1287 |
\item [$params~\vec p@1 \dots \vec p@n$] renames the innermost parameters of
|
wenzelm@12618
|
1288 |
premises $1, \dots, n$ of some theorem. An empty list of names may be given
|
wenzelm@12618
|
1289 |
to skip positions, leaving the present parameters unchanged.
|
wenzelm@13041
|
1290 |
|
wenzelm@12618
|
1291 |
Note that the default usage of case rules does \emph{not} directly expose
|
wenzelm@12618
|
1292 |
parameters to the proof context (see also \S\ref{sec:cases-induct-meth}).
|
wenzelm@13041
|
1293 |
|
wenzelm@12618
|
1294 |
\item [$consumes~n$] declares the number of ``major premises'' of a rule,
|
wenzelm@12618
|
1295 |
i.e.\ the number of facts to be consumed when it is applied by an
|
wenzelm@12618
|
1296 |
appropriate proof method (cf.\ \S\ref{sec:cases-induct-meth}). The default
|
wenzelm@12618
|
1297 |
value of $consumes$ is $n = 1$, which is appropriate for the usual kind of
|
wenzelm@13041
|
1298 |
cases and induction rules for inductive sets (cf.\
|
wenzelm@12618
|
1299 |
\S\ref{sec:hol-inductive}). Rules without any $consumes$ declaration given
|
wenzelm@12618
|
1300 |
are treated as if $consumes~0$ had been specified.
|
wenzelm@13041
|
1301 |
|
wenzelm@12618
|
1302 |
Note that explicit $consumes$ declarations are only rarely needed; this is
|
wenzelm@12618
|
1303 |
already taken care of automatically by the higher-level $cases$ and $induct$
|
wenzelm@12618
|
1304 |
declarations, see also \S\ref{sec:cases-induct-att}.
|
wenzelm@13027
|
1305 |
|
wenzelm@12618
|
1306 |
\end{descr}
|
wenzelm@12618
|
1307 |
|
wenzelm@12618
|
1308 |
|
wenzelm@12621
|
1309 |
\subsubsection{Proof methods}\label{sec:cases-induct-meth}
|
wenzelm@11691
|
1310 |
|
wenzelm@11691
|
1311 |
\indexisarmeth{cases}\indexisarmeth{induct}
|
wenzelm@11691
|
1312 |
\begin{matharray}{rcl}
|
wenzelm@11691
|
1313 |
cases & : & \isarmeth \\
|
wenzelm@11691
|
1314 |
induct & : & \isarmeth \\
|
wenzelm@11691
|
1315 |
\end{matharray}
|
wenzelm@11691
|
1316 |
|
wenzelm@11691
|
1317 |
The $cases$ and $induct$ methods provide a uniform interface to case analysis
|
wenzelm@11691
|
1318 |
and induction over datatypes, inductive sets, and recursive functions. The
|
wenzelm@11691
|
1319 |
corresponding rules may be specified and instantiated in a casual manner.
|
wenzelm@11691
|
1320 |
Furthermore, these methods provide named local contexts that may be invoked
|
wenzelm@13048
|
1321 |
via the $\CASENAME$ proof command within the subsequent proof text. This
|
wenzelm@13048
|
1322 |
accommodates compact proof texts even when reasoning about large
|
wenzelm@13048
|
1323 |
specifications.
|
wenzelm@11691
|
1324 |
|
wenzelm@11691
|
1325 |
\begin{rail}
|
wenzelm@11691
|
1326 |
'cases' spec
|
wenzelm@11691
|
1327 |
;
|
wenzelm@11691
|
1328 |
'induct' spec
|
wenzelm@11691
|
1329 |
;
|
wenzelm@11691
|
1330 |
|
wenzelm@13041
|
1331 |
spec: open? args rule?
|
wenzelm@11691
|
1332 |
;
|
wenzelm@11691
|
1333 |
open: '(' 'open' ')'
|
wenzelm@11691
|
1334 |
;
|
wenzelm@13041
|
1335 |
args: (insts * 'and')
|
wenzelm@11691
|
1336 |
;
|
wenzelm@11691
|
1337 |
rule: ('type' | 'set') ':' nameref | 'rule' ':' thmref
|
wenzelm@11691
|
1338 |
;
|
wenzelm@11691
|
1339 |
\end{rail}
|
wenzelm@11691
|
1340 |
|
wenzelm@11691
|
1341 |
\begin{descr}
|
wenzelm@13041
|
1342 |
|
wenzelm@13041
|
1343 |
\item [$cases~insts~R$] applies method $rule$ with an appropriate case
|
wenzelm@11691
|
1344 |
distinction theorem, instantiated to the subjects $insts$. Symbolic case
|
wenzelm@11691
|
1345 |
names are bound according to the rule's local contexts.
|
wenzelm@13041
|
1346 |
|
wenzelm@11691
|
1347 |
The rule is determined as follows, according to the facts and arguments
|
wenzelm@11691
|
1348 |
passed to the $cases$ method:
|
wenzelm@11691
|
1349 |
\begin{matharray}{llll}
|
wenzelm@11691
|
1350 |
\Text{facts} & & \Text{arguments} & \Text{rule} \\\hline
|
wenzelm@11691
|
1351 |
& cases & & \Text{classical case split} \\
|
wenzelm@11691
|
1352 |
& cases & t & \Text{datatype exhaustion (type of $t$)} \\
|
wenzelm@11691
|
1353 |
\edrv a \in A & cases & \dots & \Text{inductive set elimination (of $A$)} \\
|
wenzelm@11691
|
1354 |
\dots & cases & \dots ~ R & \Text{explicit rule $R$} \\
|
wenzelm@11691
|
1355 |
\end{matharray}
|
wenzelm@13041
|
1356 |
|
wenzelm@11691
|
1357 |
Several instantiations may be given, referring to the \emph{suffix} of
|
wenzelm@11691
|
1358 |
premises of the case rule; within each premise, the \emph{prefix} of
|
wenzelm@11691
|
1359 |
variables is instantiated. In most situations, only a single term needs to
|
wenzelm@11691
|
1360 |
be specified; this refers to the first variable of the last premise (it is
|
wenzelm@11691
|
1361 |
usually the same for all cases).
|
wenzelm@13041
|
1362 |
|
wenzelm@13041
|
1363 |
The ``$(open)$'' option causes the parameters of the new local contexts to
|
wenzelm@13041
|
1364 |
be exposed to the current proof context. Thus local variables stemming from
|
wenzelm@11691
|
1365 |
distant parts of the theory development may be introduced in an implicit
|
wenzelm@11691
|
1366 |
manner, which can be quite confusing to the reader. Furthermore, this
|
wenzelm@11691
|
1367 |
option may cause unwanted hiding of existing local variables, resulting in
|
wenzelm@11691
|
1368 |
less robust proof texts.
|
wenzelm@13041
|
1369 |
|
wenzelm@13041
|
1370 |
\item [$induct~insts~R$] is analogous to the $cases$ method, but refers to
|
wenzelm@11691
|
1371 |
induction rules, which are determined as follows:
|
wenzelm@11691
|
1372 |
\begin{matharray}{llll}
|
wenzelm@11691
|
1373 |
\Text{facts} & & \Text{arguments} & \Text{rule} \\\hline
|
wenzelm@11691
|
1374 |
& induct & P ~ x ~ \dots & \Text{datatype induction (type of $x$)} \\
|
wenzelm@11691
|
1375 |
\edrv x \in A & induct & \dots & \Text{set induction (of $A$)} \\
|
wenzelm@11691
|
1376 |
\dots & induct & \dots ~ R & \Text{explicit rule $R$} \\
|
wenzelm@11691
|
1377 |
\end{matharray}
|
wenzelm@13041
|
1378 |
|
wenzelm@11691
|
1379 |
Several instantiations may be given, each referring to some part of a mutual
|
wenzelm@11691
|
1380 |
inductive definition or datatype --- only related partial induction rules
|
wenzelm@11691
|
1381 |
may be used together, though. Any of the lists of terms $P, x, \dots$
|
wenzelm@11691
|
1382 |
refers to the \emph{suffix} of variables present in the induction rule.
|
wenzelm@11691
|
1383 |
This enables the writer to specify only induction variables, or both
|
wenzelm@11691
|
1384 |
predicates and variables, for example.
|
wenzelm@13041
|
1385 |
|
wenzelm@13041
|
1386 |
The ``$(open)$'' option works the same way as for $cases$.
|
wenzelm@13027
|
1387 |
|
wenzelm@11691
|
1388 |
\end{descr}
|
wenzelm@11691
|
1389 |
|
wenzelm@13048
|
1390 |
Above methods produce named local contexts, as determined by the instantiated
|
wenzelm@13048
|
1391 |
rule as specified in the text. Beyond that, the $induct$ method guesses
|
wenzelm@13048
|
1392 |
further instantiations from the goal specification itself. Any persisting
|
wenzelm@13048
|
1393 |
unresolved schematic variables of the resulting rule will render the the
|
wenzelm@13048
|
1394 |
corresponding case invalid. The term binding $\Var{case}$\indexisarvar{case}
|
wenzelm@13048
|
1395 |
for the conclusion will be provided with each case, provided that term is
|
wenzelm@13048
|
1396 |
fully specified.
|
wenzelm@11691
|
1397 |
|
wenzelm@13048
|
1398 |
The $\isarkeyword{print_cases}$ command prints all named cases present in the
|
wenzelm@13048
|
1399 |
current proof state.
|
wenzelm@11691
|
1400 |
|
wenzelm@11691
|
1401 |
\medskip
|
wenzelm@11691
|
1402 |
|
wenzelm@11691
|
1403 |
It is important to note that there is a fundamental difference of the $cases$
|
wenzelm@11691
|
1404 |
and $induct$ methods in handling of non-atomic goal statements: $cases$ just
|
wenzelm@11691
|
1405 |
applies a certain rule in backward fashion, splitting the result into new
|
wenzelm@11691
|
1406 |
goals with the local contexts being augmented in a purely monotonic manner.
|
wenzelm@11691
|
1407 |
|
nipkow@13622
|
1408 |
In contrast, $induct$ passes the full goal statement through the
|
nipkow@13622
|
1409 |
``recursive'' course involved in the induction. Thus the original statement
|
nipkow@13622
|
1410 |
is basically replaced by separate copies, corresponding to the induction
|
nipkow@13622
|
1411 |
hypotheses and conclusion; the original goal context is no longer available.
|
nipkow@13622
|
1412 |
This behavior allows \emph{strengthened induction predicates} to be expressed
|
nipkow@13622
|
1413 |
concisely as meta-level rule statements, i.e.\ $\All{\vec x} \vec\phi \Imp
|
nipkow@13622
|
1414 |
\psi$ to indicate ``variable'' parameters $\vec x$ and ``recursive''
|
nipkow@13622
|
1415 |
assumptions $\vec\phi$. Note that ``$\isarcmd{case}~c$'' already performs
|
nipkow@13622
|
1416 |
``$\FIX{\vec x}$''. Also note that local definitions may be expressed as
|
nipkow@13622
|
1417 |
$\All{\vec x} n \equiv t[\vec x] \Imp \phi[n]$, with induction over $n$.
|
nipkow@13622
|
1418 |
|
wenzelm@11691
|
1419 |
|
wenzelm@13425
|
1420 |
In induction proofs, local assumptions introduced by cases are split into two
|
wenzelm@13425
|
1421 |
different kinds: $hyps$ stemming from the rule and $prems$ from the goal
|
wenzelm@13425
|
1422 |
statement. This is reflected in the extracted cases accordingly, so invoking
|
wenzelm@13425
|
1423 |
``$\isarcmd{case}~c$'' will provide separate facts $c\mathord.hyps$ and
|
wenzelm@13425
|
1424 |
$c\mathord.prems$, as well as fact $c$ to hold the all-inclusive list.
|
wenzelm@13425
|
1425 |
|
wenzelm@11691
|
1426 |
\medskip
|
wenzelm@11691
|
1427 |
|
wenzelm@11691
|
1428 |
Facts presented to either method are consumed according to the number of
|
wenzelm@12618
|
1429 |
``major premises'' of the rule involved (see also \S\ref{sec:cases-induct}),
|
wenzelm@13041
|
1430 |
which is usually $0$ for plain cases and induction rules of datatypes etc.\
|
wenzelm@12618
|
1431 |
and $1$ for rules of inductive sets and the like. The remaining facts are
|
wenzelm@12618
|
1432 |
inserted into the goal verbatim before the actual $cases$ or $induct$ rule is
|
wenzelm@12618
|
1433 |
applied (thus facts may be even passed through an induction).
|
wenzelm@11691
|
1434 |
|
wenzelm@11691
|
1435 |
|
wenzelm@12621
|
1436 |
\subsubsection{Declaring rules}\label{sec:cases-induct-att}
|
wenzelm@11691
|
1437 |
|
wenzelm@11691
|
1438 |
\indexisarcmd{print-induct-rules}\indexisaratt{cases}\indexisaratt{induct}
|
wenzelm@11691
|
1439 |
\begin{matharray}{rcl}
|
wenzelm@11691
|
1440 |
\isarcmd{print_induct_rules}^* & : & \isarkeep{theory~|~proof} \\
|
wenzelm@11691
|
1441 |
cases & : & \isaratt \\
|
wenzelm@11691
|
1442 |
induct & : & \isaratt \\
|
wenzelm@11691
|
1443 |
\end{matharray}
|
wenzelm@11691
|
1444 |
|
wenzelm@11691
|
1445 |
\begin{rail}
|
wenzelm@11691
|
1446 |
'cases' spec
|
wenzelm@11691
|
1447 |
;
|
wenzelm@11691
|
1448 |
'induct' spec
|
wenzelm@11691
|
1449 |
;
|
wenzelm@11691
|
1450 |
|
wenzelm@11691
|
1451 |
spec: ('type' | 'set') ':' nameref
|
wenzelm@11691
|
1452 |
;
|
wenzelm@11691
|
1453 |
\end{rail}
|
wenzelm@11691
|
1454 |
|
wenzelm@13024
|
1455 |
\begin{descr}
|
wenzelm@13041
|
1456 |
|
wenzelm@13024
|
1457 |
\item [$\isarkeyword{print_induct_rules}$] prints cases and induct rules for
|
wenzelm@13024
|
1458 |
sets and types of the current context.
|
wenzelm@13041
|
1459 |
|
wenzelm@13024
|
1460 |
\item [$cases$ and $induct$] (as attributes) augment the corresponding context
|
wenzelm@13024
|
1461 |
of rules for reasoning about inductive sets and types, using the
|
wenzelm@13024
|
1462 |
corresponding methods of the same name. Certain definitional packages of
|
wenzelm@13024
|
1463 |
object-logics usually declare emerging cases and induction rules as
|
wenzelm@13024
|
1464 |
expected, so users rarely need to intervene.
|
wenzelm@13048
|
1465 |
|
wenzelm@13024
|
1466 |
Manual rule declarations usually include the the $case_names$ and $ps$
|
wenzelm@13024
|
1467 |
attributes to adjust names of cases and parameters of a rule (see
|
wenzelm@13048
|
1468 |
\S\ref{sec:cases-induct}); the $consumes$ declaration is taken care of
|
wenzelm@13024
|
1469 |
automatically: $consumes~0$ is specified for ``type'' rules and $consumes~1$
|
wenzelm@13024
|
1470 |
for ``set'' rules.
|
wenzelm@13041
|
1471 |
|
wenzelm@13024
|
1472 |
\end{descr}
|
wenzelm@11691
|
1473 |
|
wenzelm@9614
|
1474 |
%%% Local Variables:
|
wenzelm@7135
|
1475 |
%%% mode: latex
|
wenzelm@7135
|
1476 |
%%% TeX-master: "isar-ref"
|
wenzelm@9614
|
1477 |
%%% End:
|