author | paulson |
Thu, 05 Sep 1996 10:29:52 +0200 | |
changeset 1951 | f2b8005bdc6e |
parent 1801 | 927a31ba4346 |
child 1955 | 5309416236b6 |
permissions | -rw-r--r-- |
1460 | 1 |
(* Title: tactic |
0 | 2 |
ID: $Id$ |
1460 | 3 |
Author: Lawrence C Paulson, Cambridge University Computer Laboratory |
0 | 4 |
Copyright 1991 University of Cambridge |
5 |
||
6 |
Tactics |
|
7 |
*) |
|
8 |
||
9 |
signature TACTIC = |
|
1501 | 10 |
sig |
0 | 11 |
val ares_tac: thm list -> int -> tactic |
12 |
val asm_rewrite_goal_tac: |
|
214
ed6a3e2b1a33
added new parameter to the simplification tactics which indicates if
nipkow
parents:
191
diff
changeset
|
13 |
bool*bool -> (meta_simpset -> tactic) -> meta_simpset -> int -> tactic |
0 | 14 |
val assume_tac: int -> tactic |
15 |
val atac: int ->tactic |
|
670 | 16 |
val bimatch_from_nets_tac: |
1501 | 17 |
(int*(bool*thm)) Net.net * (int*(bool*thm)) Net.net -> int -> tactic |
0 | 18 |
val bimatch_tac: (bool*thm)list -> int -> tactic |
670 | 19 |
val biresolve_from_nets_tac: |
1501 | 20 |
(int*(bool*thm)) Net.net * (int*(bool*thm)) Net.net -> int -> tactic |
0 | 21 |
val biresolve_tac: (bool*thm)list -> int -> tactic |
1501 | 22 |
val build_net: thm list -> (int*thm) Net.net |
23 |
val build_netpair: (int*(bool*thm)) Net.net * (int*(bool*thm)) Net.net -> |
|
24 |
(bool*thm)list -> (int*(bool*thm)) Net.net * (int*(bool*thm)) Net.net |
|
0 | 25 |
val compose_inst_tac: (string*string)list -> (bool*thm*int) -> int -> tactic |
26 |
val compose_tac: (bool * thm * int) -> int -> tactic |
|
27 |
val cut_facts_tac: thm list -> int -> tactic |
|
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
28 |
val cut_inst_tac: (string*string)list -> thm -> int -> tactic |
0 | 29 |
val dmatch_tac: thm list -> int -> tactic |
30 |
val dresolve_tac: thm list -> int -> tactic |
|
31 |
val dres_inst_tac: (string*string)list -> thm -> int -> tactic |
|
32 |
val dtac: thm -> int ->tactic |
|
33 |
val etac: thm -> int ->tactic |
|
34 |
val eq_assume_tac: int -> tactic |
|
35 |
val ematch_tac: thm list -> int -> tactic |
|
36 |
val eresolve_tac: thm list -> int -> tactic |
|
37 |
val eres_inst_tac: (string*string)list -> thm -> int -> tactic |
|
38 |
val filter_thms: (term*term->bool) -> int*term*thm list -> thm list |
|
39 |
val filt_resolve_tac: thm list -> int -> int -> tactic |
|
40 |
val flexflex_tac: tactic |
|
41 |
val fold_goals_tac: thm list -> tactic |
|
42 |
val fold_tac: thm list -> tactic |
|
43 |
val forward_tac: thm list -> int -> tactic |
|
44 |
val forw_inst_tac: (string*string)list -> thm -> int -> tactic |
|
947 | 45 |
val freeze: thm -> thm |
1077
c2df11ae8b55
Renamed insert_kbrl to insert_tagged_brl and exported it. Now
lcp
parents:
952
diff
changeset
|
46 |
val insert_tagged_brl: ('a*(bool*thm)) * |
1501 | 47 |
(('a*(bool*thm))Net.net * ('a*(bool*thm))Net.net) -> |
48 |
('a*(bool*thm))Net.net * ('a*(bool*thm))Net.net |
|
1801 | 49 |
val delete_tagged_brl: (bool*thm) * |
50 |
((int*(bool*thm))Net.net * (int*(bool*thm))Net.net) -> |
|
51 |
(int*(bool*thm))Net.net * (int*(bool*thm))Net.net |
|
0 | 52 |
val is_fact: thm -> bool |
53 |
val lessb: (bool * thm) * (bool * thm) -> bool |
|
54 |
val lift_inst_rule: thm * int * (string*string)list * thm -> thm |
|
55 |
val make_elim: thm -> thm |
|
1501 | 56 |
val match_from_net_tac: (int*thm) Net.net -> int -> tactic |
0 | 57 |
val match_tac: thm list -> int -> tactic |
58 |
val metacut_tac: thm -> int -> tactic |
|
59 |
val net_bimatch_tac: (bool*thm) list -> int -> tactic |
|
60 |
val net_biresolve_tac: (bool*thm) list -> int -> tactic |
|
61 |
val net_match_tac: thm list -> int -> tactic |
|
62 |
val net_resolve_tac: thm list -> int -> tactic |
|
63 |
val PRIMITIVE: (thm -> thm) -> tactic |
|
64 |
val PRIMSEQ: (thm -> thm Sequence.seq) -> tactic |
|
65 |
val prune_params_tac: tactic |
|
66 |
val rename_tac: string -> int -> tactic |
|
67 |
val rename_last_tac: string -> string list -> int -> tactic |
|
1501 | 68 |
val resolve_from_net_tac: (int*thm) Net.net -> int -> tactic |
0 | 69 |
val resolve_tac: thm list -> int -> tactic |
70 |
val res_inst_tac: (string*string)list -> thm -> int -> tactic |
|
71 |
val rewrite_goals_tac: thm list -> tactic |
|
72 |
val rewrite_tac: thm list -> tactic |
|
73 |
val rewtac: thm -> tactic |
|
1209 | 74 |
val rotate_tac: int -> int -> tactic |
0 | 75 |
val rtac: thm -> int -> tactic |
76 |
val rule_by_tactic: tactic -> thm -> thm |
|
439 | 77 |
val subgoal_tac: string -> int -> tactic |
78 |
val subgoals_tac: string list -> int -> tactic |
|
0 | 79 |
val subgoals_of_brl: bool * thm -> int |
80 |
val trace_goalno_tac: (int -> tactic) -> int -> tactic |
|
1501 | 81 |
end; |
0 | 82 |
|
83 |
||
1501 | 84 |
structure Tactic : TACTIC = |
0 | 85 |
struct |
86 |
||
1501 | 87 |
(*Discover which goal is chosen: SOMEGOAL(trace_goalno_tac tac) *) |
88 |
fun trace_goalno_tac tac i st = |
|
89 |
case Sequence.pull(tac i st) of |
|
1460 | 90 |
None => Sequence.null |
0 | 91 |
| seqcell => (prs("Subgoal " ^ string_of_int i ^ " selected\n"); |
1501 | 92 |
Sequence.seqof(fn()=> seqcell)); |
0 | 93 |
|
94 |
fun string_of (a,0) = a |
|
95 |
| string_of (a,i) = a ^ "_" ^ string_of_int i; |
|
96 |
||
947 | 97 |
(*convert all Vars in a theorem to Frees*) |
0 | 98 |
fun freeze th = |
99 |
let val fth = freezeT th |
|
100 |
val {prop,sign,...} = rep_thm fth |
|
101 |
fun mk_inst (Var(v,T)) = |
|
1460 | 102 |
(cterm_of sign (Var(v,T)), |
103 |
cterm_of sign (Free(string_of v, T))) |
|
0 | 104 |
val insts = map mk_inst (term_vars prop) |
105 |
in instantiate ([],insts) fth end; |
|
106 |
||
107 |
(*Makes a rule by applying a tactic to an existing rule*) |
|
1501 | 108 |
fun rule_by_tactic tac rl = |
109 |
case Sequence.pull(tac (freeze (standard rl))) of |
|
1460 | 110 |
None => raise THM("rule_by_tactic", 0, [rl]) |
0 | 111 |
| Some(rl',_) => standard rl'; |
112 |
||
113 |
(*** Basic tactics ***) |
|
114 |
||
115 |
(*Makes a tactic whose effect on a state is given by thmfun: thm->thm seq.*) |
|
1501 | 116 |
fun PRIMSEQ thmfun st = thmfun st handle THM _ => Sequence.null; |
0 | 117 |
|
118 |
(*Makes a tactic whose effect on a state is given by thmfun: thm->thm.*) |
|
119 |
fun PRIMITIVE thmfun = PRIMSEQ (Sequence.single o thmfun); |
|
120 |
||
121 |
(*** The following fail if the goal number is out of range: |
|
122 |
thus (REPEAT (resolve_tac rules i)) stops once subgoal i disappears. *) |
|
123 |
||
124 |
(*Solve subgoal i by assumption*) |
|
125 |
fun assume_tac i = PRIMSEQ (assumption i); |
|
126 |
||
127 |
(*Solve subgoal i by assumption, using no unification*) |
|
128 |
fun eq_assume_tac i = PRIMITIVE (eq_assumption i); |
|
129 |
||
130 |
(** Resolution/matching tactics **) |
|
131 |
||
132 |
(*The composition rule/state: no lifting or var renaming. |
|
133 |
The arg = (bires_flg, orule, m) ; see bicompose for explanation.*) |
|
134 |
fun compose_tac arg i = PRIMSEQ (bicompose false arg i); |
|
135 |
||
136 |
(*Converts a "destruct" rule like P&Q==>P to an "elimination" rule |
|
137 |
like [| P&Q; P==>R |] ==> R *) |
|
138 |
fun make_elim rl = zero_var_indexes (rl RS revcut_rl); |
|
139 |
||
140 |
(*Attack subgoal i by resolution, using flags to indicate elimination rules*) |
|
141 |
fun biresolve_tac brules i = PRIMSEQ (biresolution false brules i); |
|
142 |
||
143 |
(*Resolution: the simple case, works for introduction rules*) |
|
144 |
fun resolve_tac rules = biresolve_tac (map (pair false) rules); |
|
145 |
||
146 |
(*Resolution with elimination rules only*) |
|
147 |
fun eresolve_tac rules = biresolve_tac (map (pair true) rules); |
|
148 |
||
149 |
(*Forward reasoning using destruction rules.*) |
|
150 |
fun forward_tac rls = resolve_tac (map make_elim rls) THEN' assume_tac; |
|
151 |
||
152 |
(*Like forward_tac, but deletes the assumption after use.*) |
|
153 |
fun dresolve_tac rls = eresolve_tac (map make_elim rls); |
|
154 |
||
155 |
(*Shorthand versions: for resolution with a single theorem*) |
|
1460 | 156 |
fun rtac rl = resolve_tac [rl]; |
157 |
fun etac rl = eresolve_tac [rl]; |
|
158 |
fun dtac rl = dresolve_tac [rl]; |
|
0 | 159 |
val atac = assume_tac; |
160 |
||
161 |
(*Use an assumption or some rules ... A popular combination!*) |
|
162 |
fun ares_tac rules = assume_tac ORELSE' resolve_tac rules; |
|
163 |
||
164 |
(*Matching tactics -- as above, but forbid updating of state*) |
|
165 |
fun bimatch_tac brules i = PRIMSEQ (biresolution true brules i); |
|
166 |
fun match_tac rules = bimatch_tac (map (pair false) rules); |
|
167 |
fun ematch_tac rules = bimatch_tac (map (pair true) rules); |
|
168 |
fun dmatch_tac rls = ematch_tac (map make_elim rls); |
|
169 |
||
170 |
(*Smash all flex-flex disagreement pairs in the proof state.*) |
|
171 |
val flexflex_tac = PRIMSEQ flexflex_rule; |
|
172 |
||
173 |
(*Lift and instantiate a rule wrt the given state and subgoal number *) |
|
1501 | 174 |
fun lift_inst_rule (st, i, sinsts, rule) = |
175 |
let val {maxidx,sign,...} = rep_thm st |
|
176 |
val (_, _, Bi, _) = dest_state(st,i) |
|
1460 | 177 |
val params = Logic.strip_params Bi (*params of subgoal i*) |
0 | 178 |
val params = rev(rename_wrt_term Bi params) (*as they are printed*) |
179 |
val paramTs = map #2 params |
|
180 |
and inc = maxidx+1 |
|
181 |
fun liftvar (Var ((a,j), T)) = Var((a, j+inc), paramTs---> incr_tvar inc T) |
|
182 |
| liftvar t = raise TERM("Variable expected", [t]); |
|
183 |
fun liftterm t = list_abs_free (params, |
|
1460 | 184 |
Logic.incr_indexes(paramTs,inc) t) |
0 | 185 |
(*Lifts instantiation pair over params*) |
230 | 186 |
fun liftpair (cv,ct) = (cterm_fun liftvar cv, cterm_fun liftterm ct) |
0 | 187 |
fun lifttvar((a,i),ctyp) = |
1460 | 188 |
let val {T,sign} = rep_ctyp ctyp |
189 |
in ((a,i+inc), ctyp_of sign (incr_tvar inc T)) end |
|
1501 | 190 |
val rts = types_sorts rule and (types,sorts) = types_sorts st |
0 | 191 |
fun types'(a,~1) = (case assoc(params,a) of None => types(a,~1) | sm => sm) |
192 |
| types'(ixn) = types ixn; |
|
949
83c588d6fee9
Changed treatment of during type inference internally generated type
nipkow
parents:
947
diff
changeset
|
193 |
val used = add_term_tvarnames |
1501 | 194 |
(#prop(rep_thm st) $ #prop(rep_thm rule),[]) |
949
83c588d6fee9
Changed treatment of during type inference internally generated type
nipkow
parents:
947
diff
changeset
|
195 |
val (Tinsts,insts) = read_insts sign rts (types',sorts) used sinsts |
0 | 196 |
in instantiate (map lifttvar Tinsts, map liftpair insts) |
1501 | 197 |
(lift_rule (st,i) rule) |
0 | 198 |
end; |
199 |
||
200 |
||
201 |
(*** Resolve after lifting and instantation; may refer to parameters of the |
|
202 |
subgoal. Fails if "i" is out of range. ***) |
|
203 |
||
204 |
(*compose version: arguments are as for bicompose.*) |
|
205 |
fun compose_inst_tac sinsts (bires_flg, rule, nsubgoal) i = |
|
1501 | 206 |
STATE ( fn st => |
207 |
compose_tac (bires_flg, lift_inst_rule (st, i, sinsts, rule), |
|
1460 | 208 |
nsubgoal) i |
209 |
handle TERM (msg,_) => (writeln msg; no_tac) |
|
210 |
| THM (msg,_,_) => (writeln msg; no_tac) ); |
|
0 | 211 |
|
761 | 212 |
(*"Resolve" version. Note: res_inst_tac cannot behave sensibly if the |
213 |
terms that are substituted contain (term or type) unknowns from the |
|
214 |
goal, because it is unable to instantiate goal unknowns at the same time. |
|
215 |
||
952
9e10962866b0
Removed an old bug which made some simultaneous instantiations fail if they
nipkow
parents:
949
diff
changeset
|
216 |
The type checker is instructed not to freezes flexible type vars that |
9e10962866b0
Removed an old bug which made some simultaneous instantiations fail if they
nipkow
parents:
949
diff
changeset
|
217 |
were introduced during type inference and still remain in the term at the |
9e10962866b0
Removed an old bug which made some simultaneous instantiations fail if they
nipkow
parents:
949
diff
changeset
|
218 |
end. This increases flexibility but can introduce schematic type vars in |
9e10962866b0
Removed an old bug which made some simultaneous instantiations fail if they
nipkow
parents:
949
diff
changeset
|
219 |
goals. |
761 | 220 |
*) |
0 | 221 |
fun res_inst_tac sinsts rule i = |
222 |
compose_inst_tac sinsts (false, rule, nprems_of rule) i; |
|
223 |
||
1501 | 224 |
(*eresolve elimination version*) |
0 | 225 |
fun eres_inst_tac sinsts rule i = |
226 |
compose_inst_tac sinsts (true, rule, nprems_of rule) i; |
|
227 |
||
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
228 |
(*For forw_inst_tac and dres_inst_tac. Preserve Var indexes of rl; |
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
229 |
increment revcut_rl instead.*) |
0 | 230 |
fun make_elim_preserve rl = |
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
231 |
let val {maxidx,...} = rep_thm rl |
922
196ca0973a6d
added CPure (curried functions) and ProtoPure (ancestor of Pure and CPure)
clasohm
parents:
761
diff
changeset
|
232 |
fun cvar ixn = cterm_of Sign.proto_pure (Var(ixn,propT)); |
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
233 |
val revcut_rl' = |
1460 | 234 |
instantiate ([], [(cvar("V",0), cvar("V",maxidx+1)), |
235 |
(cvar("W",0), cvar("W",maxidx+1))]) revcut_rl |
|
0 | 236 |
val arg = (false, rl, nprems_of rl) |
237 |
val [th] = Sequence.list_of_s (bicompose false arg 1 revcut_rl') |
|
238 |
in th end |
|
239 |
handle Bind => raise THM("make_elim_preserve", 1, [rl]); |
|
240 |
||
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
241 |
(*instantiate and cut -- for a FACT, anyway...*) |
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
242 |
fun cut_inst_tac sinsts rule = res_inst_tac sinsts (make_elim_preserve rule); |
0 | 243 |
|
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
244 |
(*forward tactic applies a RULE to an assumption without deleting it*) |
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
245 |
fun forw_inst_tac sinsts rule = cut_inst_tac sinsts rule THEN' assume_tac; |
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
246 |
|
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
247 |
(*dresolve tactic applies a RULE to replace an assumption*) |
0 | 248 |
fun dres_inst_tac sinsts rule = eres_inst_tac sinsts (make_elim_preserve rule); |
249 |
||
1951 | 250 |
(*Deletion of an assumption*) |
251 |
fun thin_tac s = eres_inst_tac [("V",s)] thin_rl; |
|
252 |
||
270
d506ea00c825
tactic/make_elim_preserve: recoded to avoid using lift_inst_rule. Instead
lcp
parents:
230
diff
changeset
|
253 |
(*** Applications of cut_rl ***) |
0 | 254 |
|
255 |
(*Used by metacut_tac*) |
|
256 |
fun bires_cut_tac arg i = |
|
1460 | 257 |
resolve_tac [cut_rl] i THEN biresolve_tac arg (i+1) ; |
0 | 258 |
|
259 |
(*The conclusion of the rule gets assumed in subgoal i, |
|
260 |
while subgoal i+1,... are the premises of the rule.*) |
|
261 |
fun metacut_tac rule = bires_cut_tac [(false,rule)]; |
|
262 |
||
263 |
(*Recognizes theorems that are not rules, but simple propositions*) |
|
264 |
fun is_fact rl = |
|
265 |
case prems_of rl of |
|
1460 | 266 |
[] => true | _::_ => false; |
0 | 267 |
|
268 |
(*"Cut" all facts from theorem list into the goal as assumptions. *) |
|
269 |
fun cut_facts_tac ths i = |
|
270 |
EVERY (map (fn th => metacut_tac th i) (filter is_fact ths)); |
|
271 |
||
272 |
(*Introduce the given proposition as a lemma and subgoal*) |
|
273 |
fun subgoal_tac sprop = res_inst_tac [("psi", sprop)] cut_rl; |
|
274 |
||
439 | 275 |
(*Introduce a list of lemmas and subgoals*) |
276 |
fun subgoals_tac sprops = EVERY' (map subgoal_tac sprops); |
|
277 |
||
0 | 278 |
|
279 |
(**** Indexing and filtering of theorems ****) |
|
280 |
||
281 |
(*Returns the list of potentially resolvable theorems for the goal "prem", |
|
1460 | 282 |
using the predicate could(subgoal,concl). |
0 | 283 |
Resulting list is no longer than "limit"*) |
284 |
fun filter_thms could (limit, prem, ths) = |
|
285 |
let val pb = Logic.strip_assums_concl prem; (*delete assumptions*) |
|
286 |
fun filtr (limit, []) = [] |
|
1460 | 287 |
| filtr (limit, th::ths) = |
288 |
if limit=0 then [] |
|
289 |
else if could(pb, concl_of th) then th :: filtr(limit-1, ths) |
|
290 |
else filtr(limit,ths) |
|
0 | 291 |
in filtr(limit,ths) end; |
292 |
||
293 |
||
294 |
(*** biresolution and resolution using nets ***) |
|
295 |
||
296 |
(** To preserve the order of the rules, tag them with increasing integers **) |
|
297 |
||
298 |
(*insert tags*) |
|
299 |
fun taglist k [] = [] |
|
300 |
| taglist k (x::xs) = (k,x) :: taglist (k+1) xs; |
|
301 |
||
302 |
(*remove tags and suppress duplicates -- list is assumed sorted!*) |
|
303 |
fun untaglist [] = [] |
|
304 |
| untaglist [(k:int,x)] = [x] |
|
305 |
| untaglist ((k,x) :: (rest as (k',x')::_)) = |
|
306 |
if k=k' then untaglist rest |
|
307 |
else x :: untaglist rest; |
|
308 |
||
309 |
(*return list elements in original order*) |
|
310 |
val orderlist = untaglist o sort (fn(x,y)=> #1 x < #1 y); |
|
311 |
||
312 |
(*insert one tagged brl into the pair of nets*) |
|
1077
c2df11ae8b55
Renamed insert_kbrl to insert_tagged_brl and exported it. Now
lcp
parents:
952
diff
changeset
|
313 |
fun insert_tagged_brl (kbrl as (k,(eres,th)), (inet,enet)) = |
0 | 314 |
if eres then |
1460 | 315 |
case prems_of th of |
316 |
prem::_ => (inet, Net.insert_term ((prem,kbrl), enet, K false)) |
|
317 |
| [] => error"insert_tagged_brl: elimination rule with no premises" |
|
0 | 318 |
else (Net.insert_term ((concl_of th, kbrl), inet, K false), enet); |
319 |
||
320 |
(*build a pair of nets for biresolution*) |
|
670 | 321 |
fun build_netpair netpair brls = |
1077
c2df11ae8b55
Renamed insert_kbrl to insert_tagged_brl and exported it. Now
lcp
parents:
952
diff
changeset
|
322 |
foldr insert_tagged_brl (taglist 1 brls, netpair); |
0 | 323 |
|
1801 | 324 |
(*delete one kbrl from the pair of nets; |
325 |
we don't know the value of k, so we use 0 and ignore it in the comparison*) |
|
326 |
local |
|
327 |
fun eq_kbrl ((k,(eres,th)), (k',(eres',th'))) = eq_thm (th,th') |
|
328 |
in |
|
329 |
fun delete_tagged_brl (brl as (eres,th), (inet,enet)) = |
|
330 |
if eres then |
|
331 |
case prems_of th of |
|
332 |
prem::_ => (inet, Net.delete_term ((prem, (0,brl)), enet, eq_kbrl)) |
|
333 |
| [] => error"delete_brl: elimination rule with no premises" |
|
334 |
else (Net.delete_term ((concl_of th, (0,brl)), inet, eq_kbrl), enet); |
|
335 |
end; |
|
336 |
||
337 |
||
0 | 338 |
(*biresolution using a pair of nets rather than rules*) |
339 |
fun biresolution_from_nets_tac match (inet,enet) = |
|
340 |
SUBGOAL |
|
341 |
(fn (prem,i) => |
|
342 |
let val hyps = Logic.strip_assums_hyp prem |
|
343 |
and concl = Logic.strip_assums_concl prem |
|
344 |
val kbrls = Net.unify_term inet concl @ |
|
345 |
flat (map (Net.unify_term enet) hyps) |
|
346 |
in PRIMSEQ (biresolution match (orderlist kbrls) i) end); |
|
347 |
||
348 |
(*versions taking pre-built nets*) |
|
349 |
val biresolve_from_nets_tac = biresolution_from_nets_tac false; |
|
350 |
val bimatch_from_nets_tac = biresolution_from_nets_tac true; |
|
351 |
||
352 |
(*fast versions using nets internally*) |
|
670 | 353 |
val net_biresolve_tac = |
354 |
biresolve_from_nets_tac o build_netpair(Net.empty,Net.empty); |
|
355 |
||
356 |
val net_bimatch_tac = |
|
357 |
bimatch_from_nets_tac o build_netpair(Net.empty,Net.empty); |
|
0 | 358 |
|
359 |
(*** Simpler version for resolve_tac -- only one net, and no hyps ***) |
|
360 |
||
361 |
(*insert one tagged rl into the net*) |
|
362 |
fun insert_krl (krl as (k,th), net) = |
|
363 |
Net.insert_term ((concl_of th, krl), net, K false); |
|
364 |
||
365 |
(*build a net of rules for resolution*) |
|
366 |
fun build_net rls = |
|
367 |
foldr insert_krl (taglist 1 rls, Net.empty); |
|
368 |
||
369 |
(*resolution using a net rather than rules; pred supports filt_resolve_tac*) |
|
370 |
fun filt_resolution_from_net_tac match pred net = |
|
371 |
SUBGOAL |
|
372 |
(fn (prem,i) => |
|
373 |
let val krls = Net.unify_term net (Logic.strip_assums_concl prem) |
|
374 |
in |
|
1460 | 375 |
if pred krls |
0 | 376 |
then PRIMSEQ |
1460 | 377 |
(biresolution match (map (pair false) (orderlist krls)) i) |
0 | 378 |
else no_tac |
379 |
end); |
|
380 |
||
381 |
(*Resolve the subgoal using the rules (making a net) unless too flexible, |
|
382 |
which means more than maxr rules are unifiable. *) |
|
383 |
fun filt_resolve_tac rules maxr = |
|
384 |
let fun pred krls = length krls <= maxr |
|
385 |
in filt_resolution_from_net_tac false pred (build_net rules) end; |
|
386 |
||
387 |
(*versions taking pre-built nets*) |
|
388 |
val resolve_from_net_tac = filt_resolution_from_net_tac false (K true); |
|
389 |
val match_from_net_tac = filt_resolution_from_net_tac true (K true); |
|
390 |
||
391 |
(*fast versions using nets internally*) |
|
392 |
val net_resolve_tac = resolve_from_net_tac o build_net; |
|
393 |
val net_match_tac = match_from_net_tac o build_net; |
|
394 |
||
395 |
||
396 |
(*** For Natural Deduction using (bires_flg, rule) pairs ***) |
|
397 |
||
398 |
(*The number of new subgoals produced by the brule*) |
|
1077
c2df11ae8b55
Renamed insert_kbrl to insert_tagged_brl and exported it. Now
lcp
parents:
952
diff
changeset
|
399 |
fun subgoals_of_brl (true,rule) = nprems_of rule - 1 |
c2df11ae8b55
Renamed insert_kbrl to insert_tagged_brl and exported it. Now
lcp
parents:
952
diff
changeset
|
400 |
| subgoals_of_brl (false,rule) = nprems_of rule; |
0 | 401 |
|
402 |
(*Less-than test: for sorting to minimize number of new subgoals*) |
|
403 |
fun lessb (brl1,brl2) = subgoals_of_brl brl1 < subgoals_of_brl brl2; |
|
404 |
||
405 |
||
406 |
(*** Meta-Rewriting Tactics ***) |
|
407 |
||
408 |
fun result1 tacf mss thm = |
|
1501 | 409 |
case Sequence.pull(tacf mss thm) of |
0 | 410 |
None => None |
411 |
| Some(thm,_) => Some(thm); |
|
412 |
||
413 |
(*Rewrite subgoal i only *) |
|
214
ed6a3e2b1a33
added new parameter to the simplification tactics which indicates if
nipkow
parents:
191
diff
changeset
|
414 |
fun asm_rewrite_goal_tac mode prover_tac mss i = |
ed6a3e2b1a33
added new parameter to the simplification tactics which indicates if
nipkow
parents:
191
diff
changeset
|
415 |
PRIMITIVE(rewrite_goal_rule mode (result1 prover_tac) mss i); |
0 | 416 |
|
69
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
417 |
(*Rewrite throughout proof state. *) |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
418 |
fun rewrite_tac defs = PRIMITIVE(rewrite_rule defs); |
0 | 419 |
|
420 |
(*Rewrite subgoals only, not main goal. *) |
|
69
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
421 |
fun rewrite_goals_tac defs = PRIMITIVE (rewrite_goals_rule defs); |
0 | 422 |
|
1460 | 423 |
fun rewtac def = rewrite_goals_tac [def]; |
0 | 424 |
|
425 |
||
1501 | 426 |
(*** for folding definitions, handling critical pairs ***) |
69
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
427 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
428 |
(*The depth of nesting in a term*) |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
429 |
fun term_depth (Abs(a,T,t)) = 1 + term_depth t |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
430 |
| term_depth (f$t) = 1 + max [term_depth f, term_depth t] |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
431 |
| term_depth _ = 0; |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
432 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
433 |
val lhs_of_thm = #1 o Logic.dest_equals o #prop o rep_thm; |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
434 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
435 |
(*folding should handle critical pairs! E.g. K == Inl(0), S == Inr(Inl(0)) |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
436 |
Returns longest lhs first to avoid folding its subexpressions.*) |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
437 |
fun sort_lhs_depths defs = |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
438 |
let val keylist = make_keylist (term_depth o lhs_of_thm) defs |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
439 |
val keys = distinct (sort op> (map #2 keylist)) |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
440 |
in map (keyfilter keylist) keys end; |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
441 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
442 |
fun fold_tac defs = EVERY |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
443 |
(map rewrite_tac (sort_lhs_depths (map symmetric defs))); |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
444 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
445 |
fun fold_goals_tac defs = EVERY |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
446 |
(map rewrite_goals_tac (sort_lhs_depths (map symmetric defs))); |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
447 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
448 |
|
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
449 |
(*** Renaming of parameters in a subgoal |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
450 |
Names may contain letters, digits or primes and must be |
e7588b53d6b0
tactic/fold_tac,fold_goals_tac: now handle critical pairs on the
lcp
parents:
0
diff
changeset
|
451 |
separated by blanks ***) |
0 | 452 |
|
453 |
(*Calling this will generate the warning "Same as previous level" since |
|
454 |
it affects nothing but the names of bound variables!*) |
|
455 |
fun rename_tac str i = |
|
456 |
let val cs = explode str |
|
457 |
in |
|
458 |
if !Logic.auto_rename |
|
459 |
then (writeln"Note: setting Logic.auto_rename := false"; |
|
1460 | 460 |
Logic.auto_rename := false) |
0 | 461 |
else (); |
462 |
case #2 (take_prefix (is_letdig orf is_blank) cs) of |
|
463 |
[] => PRIMITIVE (rename_params_rule (scanwords is_letdig cs, i)) |
|
464 |
| c::_ => error ("Illegal character: " ^ c) |
|
465 |
end; |
|
466 |
||
1501 | 467 |
(*Rename recent parameters using names generated from a and the suffixes, |
468 |
provided the string a, which represents a term, is an identifier. *) |
|
0 | 469 |
fun rename_last_tac a sufs i = |
470 |
let val names = map (curry op^ a) sufs |
|
471 |
in if Syntax.is_identifier a |
|
472 |
then PRIMITIVE (rename_params_rule (names,i)) |
|
473 |
else all_tac |
|
474 |
end; |
|
475 |
||
476 |
(*Prunes all redundant parameters from the proof state by rewriting*) |
|
477 |
val prune_params_tac = rewrite_tac [triv_forall_equality]; |
|
478 |
||
1501 | 479 |
(*rotate_tac n i: rotate the assumptions of subgoal i by n positions, from |
480 |
right to left if n is positive, and from left to right if n is negative.*) |
|
1209 | 481 |
fun rotate_tac n = |
482 |
let fun rot(n) = EVERY'(replicate n (dtac asm_rl)); |
|
483 |
in if n >= 0 then rot n |
|
484 |
else SUBGOAL (fn (t,i) => rot(length(Logic.strip_assums_hyp t)+n) i) |
|
485 |
end; |
|
486 |
||
0 | 487 |
end; |
1501 | 488 |
|
489 |
open Tactic; |