1 |
/* Eliot */ |
/* Eliot */ |
2 |
/* Copyright (C) 1999 antoine.fraboulet */ |
/* Copyright (C) 2005 Antoine Fraboulet */ |
|
/* antoine.fraboulet@free.fr */ |
|
3 |
/* */ |
/* */ |
4 |
/* This program is free software; you can redistribute it and/or modify */ |
/* This file is part of Eliot. */ |
5 |
|
/* */ |
6 |
|
/* Eliot is free software; you can redistribute it and/or modify */ |
7 |
/* it under the terms of the GNU General Public License as published by */ |
/* it under the terms of the GNU General Public License as published by */ |
8 |
/* the Free Software Foundation; either version 2 of the License, or */ |
/* the Free Software Foundation; either version 2 of the License, or */ |
9 |
/* (at your option) any later version. */ |
/* (at your option) any later version. */ |
10 |
/* */ |
/* */ |
11 |
/* This program is distributed in the hope that it will be useful, */ |
/* Elit is distributed in the hope that it will be useful, */ |
12 |
/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ |
/* but WITHOUT ANY WARRANTY; without even the implied warranty of */ |
13 |
/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ |
/* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the */ |
14 |
/* GNU General Public License for more details. */ |
/* GNU General Public License for more details. */ |
16 |
/* You should have received a copy of the GNU General Public License */ |
/* You should have received a copy of the GNU General Public License */ |
17 |
/* along with this program; if not, write to the Free Software */ |
/* along with this program; if not, write to the Free Software */ |
18 |
/* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ |
/* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */ |
19 |
|
|
20 |
/* |
/* |
21 |
* $Id$ |
* $Id$ |
22 |
*/ |
*/ |
23 |
|
|
24 |
#include "config.h" |
#include "config.h" |
25 |
|
#include <assert.h> |
26 |
#include <string.h> |
#include <string.h> |
27 |
#include <stdlib.h> |
#include <stdlib.h> |
28 |
#include <stdio.h> |
#include <stdio.h> |
31 |
# include <sys/wait.h> |
# include <sys/wait.h> |
32 |
#endif |
#endif |
33 |
#include <unistd.h> |
#include <unistd.h> |
34 |
|
|
35 |
#include "dic.h" |
#include "dic.h" |
36 |
#include "regexp.h" |
#include "regexp.h" |
37 |
|
#include "alist.h" |
38 |
#include "automaton.h" |
#include "automaton.h" |
39 |
|
|
40 |
#ifndef PDBG |
#ifdef DEBUG_AUTOMATON__ |
41 |
#ifdef DEBUG_RE2 |
#define DMSG(a) a |
|
# define PDBG(s) s |
|
42 |
#else |
#else |
43 |
# define PDBG(s) |
#define DMSG(a) |
44 |
#endif |
#endif |
45 |
|
|
46 |
|
#define MAX_TRANSITION_LETTERS 256 |
47 |
|
|
48 |
|
typedef struct automaton_state_t *astate; |
49 |
|
typedef struct Automaton_t *Automaton; |
50 |
|
|
51 |
|
/* ************************************************** * |
52 |
|
exported functions for static automata |
53 |
|
* ************************************************** */ |
54 |
|
|
55 |
|
automaton automaton_build (int init_state, int *ptl, int *PS, struct search_RegE_list_t *list); |
56 |
|
void automaton_delete (automaton a); |
57 |
|
int automaton_get_nstate (automaton a); |
58 |
|
int automaton_get_init (automaton a); |
59 |
|
int automaton_get_accept (automaton a, int state); |
60 |
|
int automaton_get_next_state (automaton a, int start, char l); |
61 |
|
void automaton_dump (automaton a, char* filename); |
62 |
|
|
63 |
|
|
64 |
|
/* ************************************************** * |
65 |
|
static functions for dynamic automata |
66 |
|
* ************************************************** */ |
67 |
|
|
68 |
|
static Automaton s_automaton_create (); |
69 |
|
static void s_automaton_delete (Automaton a); |
70 |
|
|
71 |
|
static alist s_automaton_id_create (int id); |
72 |
|
static char* s_automaton_id_to_str (alist id); |
73 |
|
|
74 |
|
static astate s_automaton_state_create (alist id); |
75 |
|
|
76 |
|
static void s_automaton_add_state (Automaton a, astate s); |
77 |
|
static astate s_automaton_get_state (Automaton a, alist id); |
78 |
|
|
79 |
|
static Automaton s_automaton_PS_to_NFA (int init_state, int *ptl, int *PS); |
80 |
|
static Automaton s_automaton_NFA_to_DFA (Automaton a, struct search_RegE_list_t *list); |
81 |
|
static automaton s_automaton_finalize (Automaton a); |
82 |
|
#ifdef DEBUG_AUTOMATON |
83 |
|
static void s_automaton_dump (Automaton a, char* filename); |
84 |
#endif |
#endif |
85 |
|
|
86 |
automaton |
/* ************************************************** * |
87 |
automaton_build(int init_state, int *ptl, int *PS) |
data types |
88 |
|
* ************************************************** */ |
89 |
|
|
90 |
|
struct automaton_state_t { |
91 |
|
alist id; // alist of int |
92 |
|
int accept; |
93 |
|
int id_static; |
94 |
|
astate next[MAX_TRANSITION_LETTERS]; |
95 |
|
}; |
96 |
|
|
97 |
|
struct Automaton_t { |
98 |
|
int nstates; |
99 |
|
astate init_state; |
100 |
|
alist states; // alist of alist of int |
101 |
|
}; |
102 |
|
|
103 |
|
struct automaton_t { |
104 |
|
int nstates; |
105 |
|
int init; |
106 |
|
int *accept; |
107 |
|
int **trans; |
108 |
|
}; |
109 |
|
|
110 |
|
/* ************************************************** * |
111 |
|
exported functions for static automata |
112 |
|
* ************************************************** */ |
113 |
|
|
114 |
|
automaton |
115 |
|
automaton_build(int init_state, int *ptl, int *PS, struct search_RegE_list_t *list) |
116 |
{ |
{ |
117 |
/* int init_state; == root->PP */ |
Automaton nfa,dfa; |
118 |
/* int *ptl; mapping postition -> lettre */ |
automaton final; |
|
/* int *PS; Position Suivante [ 1 << (position-1)] = \cup { 1 << (p-1) | p \in position acceptée } */ |
|
|
|
|
|
int i,l,pos,letter,ens; |
|
|
int state,plist; |
|
|
int *state_list; |
|
|
char used_letter[256]; |
|
|
automaton a; |
|
119 |
|
|
120 |
a = (automaton)malloc(sizeof(struct _automaton)); |
nfa = s_automaton_PS_to_NFA(init_state,ptl,PS); |
121 |
|
DMSG(printf("\n non deterministic automaton OK \n\n")); |
122 |
a->nterm = PS[0]; |
DMSG(s_automaton_dump(nfa,"auto_nfa")); |
123 |
a->nstate = 1; |
|
124 |
a->init = init_state; |
dfa = s_automaton_NFA_to_DFA(nfa, list); |
125 |
a->accept = (int*) calloc(1 << PS[0], sizeof(int)); // #{states} |
DMSG(printf("\n deterministic automaton OK \n\n")); |
126 |
a->marque = (int*) calloc(1 << PS[0], sizeof(int)); // #{states} |
DMSG(s_automaton_dump(dfa,"auto_dfa")); |
127 |
a->Dtrans = (int**)calloc(1 << PS[0], sizeof(int*)); // #{states} * #{letters} |
|
128 |
a->callocsize = 1 << PS[0]; |
final = s_automaton_finalize(dfa); |
129 |
|
DMSG(printf("\n final automaton OK \n\n")); |
130 |
|
DMSG(automaton_dump(final,"auto_fin")); |
131 |
|
|
132 |
|
s_automaton_delete(nfa); |
133 |
|
s_automaton_delete(dfa); |
134 |
|
return final; |
135 |
|
} |
136 |
|
|
137 |
|
void |
138 |
|
automaton_delete(automaton a) |
139 |
|
{ |
140 |
|
int i; |
141 |
|
free(a->accept); |
142 |
|
for(i=0; i <= a->nstates; i++) |
143 |
|
free(a->trans[i]); |
144 |
|
free(a->trans); |
145 |
|
free(a); |
146 |
|
} |
147 |
|
|
148 |
|
inline int |
149 |
|
automaton_get_nstates(automaton a) |
150 |
|
{ |
151 |
|
return a->nstates; |
152 |
|
} |
153 |
|
|
154 |
|
inline int |
155 |
|
automaton_get_init(automaton a) |
156 |
|
{ |
157 |
|
return a->init; |
158 |
|
} |
159 |
|
|
160 |
|
inline int |
161 |
|
automaton_get_accept(automaton a, int state) |
162 |
|
{ |
163 |
|
assert(0 <= state && state <= a->nstates); |
164 |
|
return a->accept[state]; |
165 |
|
} |
166 |
|
|
167 |
|
inline int |
168 |
|
automaton_get_next_state(automaton a, int state, char l) |
169 |
|
{ |
170 |
|
assert(0 <= state && state <= a->nstates); |
171 |
|
return a->trans[state][(int)l]; |
172 |
|
} |
173 |
|
|
174 |
|
void |
175 |
|
automaton_dump(automaton a, char* filename) |
176 |
|
{ |
177 |
|
int i,l; |
178 |
|
FILE* f; |
179 |
|
pid_t pid; |
180 |
|
if (a == NULL) |
181 |
|
return ; |
182 |
|
f=fopen(filename,"w"); |
183 |
|
fprintf(f,"digraph automaton {\n"); |
184 |
|
for(i=1; i<=a->nstates; i++) |
185 |
|
{ |
186 |
|
fprintf(f,"\t%d [label = \"%d\"",i,i); |
187 |
|
if (i == a->init) |
188 |
|
fprintf(f,", style = filled, color=lightgrey"); |
189 |
|
if (a->accept[i]) |
190 |
|
fprintf(f,", shape = doublecircle"); |
191 |
|
fprintf(f,"];\n"); |
192 |
|
} |
193 |
|
fprintf(f,"\n"); |
194 |
|
for(i=1; i<=a->nstates; i++) |
195 |
|
for(l=0; l < MAX_TRANSITION_LETTERS; l++) |
196 |
|
if (a->trans[i][l]) |
197 |
|
{ |
198 |
|
fprintf(f,"\t%d -> %d [label = \"",i,a->trans[i][l]); |
199 |
|
regexp_print_letter(f,l); |
200 |
|
fprintf(f,"\"];\n"); |
201 |
|
} |
202 |
|
fprintf(f,"fontsize=20;\n"); |
203 |
|
fprintf(f,"}\n"); |
204 |
|
fclose(f); |
205 |
|
|
206 |
|
#ifdef HAVE_SYS_WAIT_H |
207 |
|
pid = fork (); |
208 |
|
if (pid > 0) { |
209 |
|
wait(NULL); |
210 |
|
} else if (pid == 0) { |
211 |
|
execlp("dotty","dotty",filename,NULL); |
212 |
|
printf("exec dotty failed\n"); |
213 |
|
exit(1); |
214 |
|
} |
215 |
|
#endif |
216 |
|
} |
217 |
|
|
218 |
|
/* ************************************************** * |
219 |
|
* ************************************************** * |
220 |
|
* ************************************************** */ |
221 |
|
|
222 |
|
void |
223 |
|
state_delete_fun(void* ps) |
224 |
|
{ |
225 |
|
astate s = ps; |
226 |
|
alist_delete(s->id); |
227 |
|
free(s); |
228 |
|
} |
229 |
|
|
230 |
|
static Automaton |
231 |
|
s_automaton_create() |
232 |
|
{ |
233 |
|
Automaton a; |
234 |
|
a = (Automaton)malloc(sizeof(struct Automaton_t)); |
235 |
|
a->nstates = 0; |
236 |
|
a->init_state = NULL; |
237 |
|
a->states = alist_create(); |
238 |
|
alist_set_delete(a->states,state_delete_fun); |
239 |
|
return a; |
240 |
|
} |
241 |
|
|
242 |
|
|
243 |
|
static void |
244 |
|
s_automaton_delete(Automaton a) |
245 |
|
{ |
246 |
|
alist_delete(a->states); |
247 |
|
free(a); |
248 |
|
} |
249 |
|
|
250 |
|
static alist |
251 |
|
s_automaton_id_create(int id) |
252 |
|
{ |
253 |
|
alist a = alist_create(); |
254 |
|
alist_add(a,(void*)id); |
255 |
|
return a; |
256 |
|
} |
257 |
|
|
258 |
for(i=0; i < (1 << PS[0]); i++) |
static char* s_automaton_id_to_str(alist id) |
259 |
|
{ |
260 |
|
static char s[250]; |
261 |
|
memset(s,0,sizeof(s)); |
262 |
|
alist_elt ptr; |
263 |
|
for(ptr = alist_get_first(id); ptr ; ptr = alist_get_next(id,ptr)) |
264 |
{ |
{ |
265 |
// 256 different letters max |
char tmp[50]; |
266 |
a->Dtrans[i] = (int*)calloc(256, sizeof(int)); |
sprintf(tmp,"%d ",(int)alist_elt_get_value(ptr)); |
267 |
|
strcat(s,tmp); |
268 |
} |
} |
269 |
|
return s; |
270 |
|
} |
271 |
|
|
272 |
|
static astate |
273 |
|
s_automaton_state_create(alist id) |
274 |
|
{ |
275 |
|
astate s; |
276 |
|
s = (astate)malloc(sizeof(struct automaton_state_t)); |
277 |
|
s->id = id; |
278 |
|
s->accept = 0; |
279 |
|
memset(s->next,0,sizeof(astate)*MAX_TRANSITION_LETTERS); |
280 |
|
DMSG(printf("** state %s creation\n",s_automaton_id_to_str(id))); |
281 |
|
return s; |
282 |
|
} |
283 |
|
|
284 |
|
static void |
285 |
|
s_automaton_add_state(Automaton a, astate s) |
286 |
|
{ |
287 |
|
a->nstates ++; |
288 |
|
alist_add(a->states,(void*)s); |
289 |
|
DMSG(printf("** state %s added to automaton\n",s_automaton_id_to_str(s->id))); |
290 |
|
} |
291 |
|
|
292 |
|
static astate |
293 |
|
s_automaton_get_state(Automaton a, alist id) |
294 |
|
{ |
295 |
|
astate s; |
296 |
|
alist_elt ptr; |
297 |
|
for(ptr = alist_get_first(a->states) ; ptr ; ptr = alist_get_next(a->states,ptr)) |
298 |
|
{ |
299 |
|
s = alist_elt_get_value(ptr); |
300 |
|
if (alist_equal(s->id,id)) |
301 |
|
{ |
302 |
|
//DMSG(printf("** get state %s ok\n",s_automaton_id_to_str(s->id))); |
303 |
|
return s; |
304 |
|
} |
305 |
|
} |
306 |
|
return NULL; |
307 |
|
} |
308 |
|
|
309 |
|
/* ************************************************** * |
310 |
|
* ************************************************** * |
311 |
|
* ************************************************** */ |
312 |
|
|
313 |
state_list = (int*)calloc(1 << PS[0], sizeof(int)); // #{states} |
Automaton |
314 |
|
s_automaton_PS_to_NFA(int init_state_id, int *ptl, int *PS) |
315 |
|
{ |
316 |
|
int p; |
317 |
|
int maxpos = PS[0]; |
318 |
|
Automaton nfa = NULL; |
319 |
|
alist temp_id; |
320 |
|
alist_elt ptr; |
321 |
|
astate temp_state,current_state; |
322 |
|
alist L; |
323 |
|
char used_letter[MAX_TRANSITION_LETTERS]; |
324 |
|
|
325 |
|
nfa = s_automaton_create(); |
326 |
|
L = alist_create(); |
327 |
|
|
328 |
/* 1: init_state = root->PP */ |
/* 1: init_state = root->PP */ |
329 |
plist = 0; |
temp_id = s_automaton_id_create(init_state_id); |
330 |
state_list[plist++] = a->init; |
temp_state = s_automaton_state_create(temp_id); |
331 |
|
nfa->init_state = temp_state; |
332 |
|
s_automaton_add_state(nfa,temp_state); |
333 |
|
alist_add(L,temp_state); |
334 |
/* 2: while \exist state \in state_list */ |
/* 2: while \exist state \in state_list */ |
335 |
while (plist) |
while (! alist_is_empty(L)) |
336 |
{ |
{ |
337 |
state = state_list[--plist]; |
current_state = (astate)alist_pop_first_value(L); |
338 |
PDBG(fprintf(stdout,"** traitement état 0x%08x\n",state)); |
DMSG(printf("** current state = %s\n",s_automaton_id_to_str(current_state->id))); |
339 |
memset(used_letter,0,sizeof(used_letter)); |
memset(used_letter,0,sizeof(used_letter)); |
340 |
/* 3: \foreach l in \sigma | l \neq # */ |
/* 3: \foreach l in \sigma | l \neq # */ |
341 |
for(l=1; l < PS[0]; l++) |
for(p=1; p < maxpos; p++) |
342 |
{ |
{ |
343 |
letter = ptl[l]; |
int current_letter = ptl[p]; |
344 |
if (used_letter[letter] == 0) |
if (used_letter[current_letter] == 0) |
345 |
{ |
{ |
346 |
/* 4: int ens = \cup { PS(pos) | pos \in state \wedge pos == l } */ |
/* 4: int set = \cup { PS(pos) | pos \in state \wedge pos == l } */ |
347 |
ens = 0; |
int pos, ens = 0; |
348 |
for(pos = 1; pos <= PS[0]; pos++) |
for(pos = 1; pos <= maxpos; pos++) |
349 |
{ |
{ |
350 |
if (ptl[pos] == letter && (state & (1 << (pos - 1)))) |
if (ptl[pos] == current_letter && |
351 |
|
(int)alist_elt_get_value(alist_get_first(current_state->id)) & (1 << (pos - 1))) |
352 |
ens |= PS[pos]; |
ens |= PS[pos]; |
353 |
} |
} |
354 |
/* 5: */ |
/* 5: transition from current_state to temp_state */ |
355 |
if (ens && a->marque[ens] == 0) |
if (ens) |
356 |
{ |
{ |
357 |
state_list[plist++] = ens; |
temp_id = s_automaton_id_create(ens); |
358 |
a->Dtrans[state][letter] = ens; |
temp_state = s_automaton_get_state(nfa,temp_id); |
359 |
PDBG(fprintf(stdout," adding %x +",state)); |
if (temp_state == NULL) |
|
PDBG(regexp_print_letter(stdout,letter)); |
|
|
PDBG(fprintf(stdout,"> %x (queue %x)\n",ens,ens)); |
|
|
if (ens != state) |
|
360 |
{ |
{ |
361 |
a->nstate = a->nstate + 1; |
temp_state = s_automaton_state_create(temp_id); |
362 |
|
s_automaton_add_state (nfa,temp_state); |
363 |
|
current_state->next[current_letter] = temp_state; |
364 |
|
alist_add(L,temp_state); |
365 |
|
} |
366 |
|
else |
367 |
|
{ |
368 |
|
alist_delete(temp_id); |
369 |
|
current_state->next[current_letter] = temp_state; |
370 |
} |
} |
371 |
} |
} |
372 |
/* 6: */ |
used_letter[current_letter] = 1; |
|
if (ens && a->marque[ens] == 1) |
|
|
{ |
|
|
a->Dtrans[state][letter] = ens; |
|
|
PDBG(fprintf(stdout," adding %x -",state)); |
|
|
PDBG(regexp_print_letter(stdout,letter)); |
|
|
PDBG(fprintf(stdout,"> %x\n",ens)); |
|
|
} |
|
|
a->marque[state] = 1; |
|
|
used_letter[letter] = 1; |
|
373 |
} |
} |
374 |
} |
} |
375 |
} |
} |
376 |
|
|
377 |
PDBG(fprintf(stdout,"** accept : ")); |
alist_delete(L); |
378 |
for(i=0; i < (1 << PS[0]); i++) |
|
379 |
|
for(ptr = alist_get_first(nfa->states); ptr ; ptr = alist_get_next(nfa->states,ptr)) |
380 |
|
{ |
381 |
|
astate s = (astate)alist_elt_get_value(ptr); |
382 |
|
if ((int)alist_elt_get_value(alist_get_first(s->id)) & (1 << (maxpos - 1))) |
383 |
|
s->accept = 1; |
384 |
|
} |
385 |
|
|
386 |
|
return nfa; |
387 |
|
} |
388 |
|
|
389 |
|
/* ************************************************** * |
390 |
|
* ************************************************** * |
391 |
|
* ************************************************** */ |
392 |
|
|
393 |
|
static alist |
394 |
|
s_automaton_successor(alist S, int letter, Automaton nfa, struct search_RegE_list_t *list) |
395 |
|
{ |
396 |
|
alist R,r; |
397 |
|
alist_elt ptr; |
398 |
|
R = alist_create(); /* R = \empty */ |
399 |
|
/* \forall y \in S */ |
400 |
|
for(ptr = alist_get_first(S); ptr ; ptr = alist_get_next(S,ptr)) |
401 |
{ |
{ |
402 |
if (a->marque[i] && (i & (1 << (PS[0] - 1)))) |
int i; |
403 |
|
alist t, Ry; astate y,z; |
404 |
|
|
405 |
|
i = (int)alist_elt_get_value(ptr); |
406 |
|
t = s_automaton_id_create(i); |
407 |
|
assert(y = s_automaton_get_state(nfa,t)); |
408 |
|
alist_delete(t); |
409 |
|
|
410 |
|
Ry = alist_create(); /* Ry = \empty */ |
411 |
|
|
412 |
|
if ((z = y->next[letter]) != NULL) /* \delta (y,z) = l */ |
413 |
|
{ |
414 |
|
r = s_automaton_successor(z->id,RE_EPSILON,nfa, list); |
415 |
|
alist_insert(Ry,r); |
416 |
|
alist_delete(r); |
417 |
|
alist_insert(Ry,z->id); /* Ry = Ry \cup succ(z) */ |
418 |
|
} |
419 |
|
|
420 |
|
#if 0 |
421 |
|
if ((z = y->next[RE_EPSILON]) != NULL) /* \delta (y,z) = \epsilon */ |
422 |
{ |
{ |
423 |
a->accept[i] = 1; |
r = s_automaton_successor(z->id,letter,nfa, list); |
424 |
PDBG(fprintf(stdout,"%x ",i)); |
alist_insert(Ry,r); /* Ry = Ry \cup succ(z) */ |
425 |
|
alist_delete(r); |
426 |
} |
} |
427 |
|
#endif |
428 |
|
|
429 |
|
if (letter < RE_FINAL_TOK) |
430 |
|
{ |
431 |
|
for(i = 0 ; i < DIC_SEARCH_REGE_LIST ; i++) |
432 |
|
if (list->valid[i]) |
433 |
|
{ |
434 |
|
if (list->letters[i][letter] && (z = y->next[(int)list->symbl[i]]) != NULL) |
435 |
|
{ |
436 |
|
DMSG(printf("*** letter ")); |
437 |
|
DMSG(regexp_print_letter(stdout,letter)); |
438 |
|
DMSG(printf("is in ")); |
439 |
|
DMSG(regexp_print_letter(stdout,i)); |
440 |
|
|
441 |
|
r = s_automaton_successor(z->id,RE_EPSILON,nfa, list); |
442 |
|
alist_insert(Ry,r); |
443 |
|
alist_delete(r); |
444 |
|
alist_insert(Ry,z->id); |
445 |
|
} |
446 |
|
} |
447 |
|
} |
448 |
|
|
449 |
|
// if (alist_is_empty(Ry)) /* Ry = \empty */ |
450 |
|
// return Ry; |
451 |
|
|
452 |
|
alist_insert(R,Ry); /* R = R \cup Ry */ |
453 |
|
alist_delete(Ry); |
454 |
} |
} |
|
PDBG(fprintf(stdout,"\n")); |
|
455 |
|
|
456 |
free(state_list); |
return R; |
|
return a; |
|
457 |
} |
} |
458 |
|
|
459 |
////////////////////////////////////////////////// |
static void |
460 |
////////////////////////////////////////////////// |
s_automaton_node_set_accept(astate s, Automaton nfa) |
461 |
|
{ |
462 |
|
void* idx; |
463 |
|
alist_elt ptr; |
464 |
|
|
465 |
void |
DMSG(printf("=== setting accept for node (%s) :",s_automaton_id_to_str(s->id))); |
466 |
automaton_delete(automaton a) |
for(ptr = alist_get_first(nfa->states) ; ptr ; ptr = alist_get_next(nfa->states,ptr)) |
467 |
|
{ |
468 |
|
astate ns = (astate)alist_elt_get_value(ptr); |
469 |
|
idx = alist_elt_get_value(alist_get_first(ns->id)); |
470 |
|
DMSG(printf("%s ",s_automaton_id_to_str(ns->id))); |
471 |
|
if (ns->accept && alist_is_in(s->id,idx)) |
472 |
|
{ |
473 |
|
DMSG(printf("(ok) ")); |
474 |
|
s->accept = 1; |
475 |
|
} |
476 |
|
} |
477 |
|
DMSG(printf("\n")); |
478 |
|
} |
479 |
|
|
480 |
|
static Automaton |
481 |
|
s_automaton_NFA_to_DFA(Automaton nfa, struct search_RegE_list_t *list) |
482 |
{ |
{ |
483 |
int i; |
Automaton dfa = NULL; |
484 |
free(a->accept); |
alist temp_id; |
485 |
free(a->marque); |
alist_elt ptr; |
486 |
for(i=0; i < a->callocsize; i++) |
astate temp_state, current_state; |
487 |
|
alist L; |
488 |
|
int letter; |
489 |
|
|
490 |
|
dfa = s_automaton_create(); |
491 |
|
L = alist_create(); |
492 |
|
|
493 |
|
temp_id = alist_clone(nfa->init_state->id); |
494 |
|
temp_state = s_automaton_state_create(temp_id); |
495 |
|
dfa->init_state = temp_state; |
496 |
|
s_automaton_add_state(dfa,temp_state); |
497 |
|
alist_add(L,temp_state); |
498 |
|
while (! alist_is_empty(L)) |
499 |
{ |
{ |
500 |
free(a->Dtrans[i]); |
current_state = (astate)alist_pop_first_value(L); |
501 |
|
DMSG(printf("** current state = %s\n",s_automaton_id_to_str(current_state->id))); |
502 |
|
for(letter = 1; letter < DIC_LETTERS; letter++) |
503 |
|
{ |
504 |
|
// DMSG(printf("*** start successor of %s\n",s_automaton_id_to_str(current_state->id))); |
505 |
|
|
506 |
|
temp_id = s_automaton_successor(current_state->id,letter,nfa,list); |
507 |
|
|
508 |
|
if (! alist_is_empty(temp_id)) |
509 |
|
{ |
510 |
|
|
511 |
|
DMSG(printf("*** successor of %s for ",s_automaton_id_to_str(current_state->id))); |
512 |
|
DMSG(regexp_print_letter(stdout,letter)); |
513 |
|
DMSG(printf(" = %s\n", s_automaton_id_to_str(temp_id))); |
514 |
|
|
515 |
|
temp_state = s_automaton_get_state(dfa,temp_id); |
516 |
|
|
517 |
|
// DMSG(printf("*** automaton get state -%s- ok\n",s_automaton_id_to_str(temp_id))); |
518 |
|
|
519 |
|
if (temp_state == NULL) |
520 |
|
{ |
521 |
|
temp_state = s_automaton_state_create(temp_id); |
522 |
|
s_automaton_add_state(dfa,temp_state); |
523 |
|
current_state->next[letter] = temp_state; |
524 |
|
alist_add(L,temp_state); |
525 |
|
} |
526 |
|
else |
527 |
|
{ |
528 |
|
alist_delete(temp_id); |
529 |
|
current_state->next[letter] = temp_state; |
530 |
|
} |
531 |
|
} |
532 |
|
else |
533 |
|
{ |
534 |
|
alist_delete(temp_id); |
535 |
|
} |
536 |
|
} |
537 |
} |
} |
538 |
free(a->Dtrans); |
|
539 |
free(a); |
for(ptr = alist_get_first(dfa->states) ; ptr ; ptr = alist_get_next(dfa->states,ptr)) |
540 |
|
{ |
541 |
|
astate s = (astate)alist_elt_get_value(ptr); |
542 |
|
s_automaton_node_set_accept(s,nfa); |
543 |
|
} |
544 |
|
|
545 |
|
alist_delete(L); |
546 |
|
return dfa; |
547 |
} |
} |
548 |
|
|
549 |
////////////////////////////////////////////////// |
/* ************************************************** * |
550 |
////////////////////////////////////////////////// |
* ************************************************** * |
551 |
|
* ************************************************** */ |
552 |
|
|
553 |
#if 0 |
static automaton |
554 |
void |
s_automaton_finalize(Automaton a) |
|
automaton_minimize(automaton a, automaton b) |
|
555 |
{ |
{ |
556 |
int i; |
int i,l; |
557 |
b->nterm = a->nterm; |
automaton fa = NULL; |
558 |
b->nstate = a->nstate; |
alist_elt ptr; |
559 |
b->init = a->init; |
astate s; |
560 |
b->accept = (int*) calloc(a->nstate, sizeof(int )); // #{etats} |
|
561 |
b->marque = (int*) calloc(a->nstate, sizeof(int )); // #{etats} |
if (a == NULL) |
562 |
b->Dtrans = (int**)calloc(a->nstate, sizeof(int*)); // #{etats} * #{lettres} |
return NULL; |
563 |
for(i=0; i < a->nstate; i++) |
|
564 |
|
/* creation */ |
565 |
|
fa = (automaton)malloc(sizeof(struct automaton_t)); |
566 |
|
fa->nstates = a->nstates; |
567 |
|
fa->accept = (int*) malloc((fa->nstates + 1)*sizeof(int)); |
568 |
|
memset(fa->accept,0,(fa->nstates + 1)*sizeof(int)); |
569 |
|
fa->trans = (int**)malloc((fa->nstates + 1)*sizeof(int*)); |
570 |
|
for(i=0; i <= fa->nstates; i++) |
571 |
{ |
{ |
572 |
b->Dtrans[i] = (int*)calloc(256, sizeof(int)); |
fa->trans[i] = (int*)malloc(MAX_TRANSITION_LETTERS * sizeof(int)); |
573 |
|
memset(fa->trans[i],0,MAX_TRANSITION_LETTERS * sizeof(int)); |
574 |
} |
} |
575 |
|
|
576 |
/* NOT DONE */ |
/* create new id for states */ |
577 |
|
for(i = 1 , ptr = alist_get_first(a->states); ptr ; ptr = alist_get_next(a->states,ptr), i++) |
578 |
|
{ |
579 |
|
s = (astate)alist_elt_get_value(ptr); |
580 |
|
s->id_static = i; |
581 |
|
} |
582 |
|
|
583 |
|
/* build new automaton */ |
584 |
|
for(ptr = alist_get_first(a->states); ptr ; ptr = alist_get_next(a->states,ptr)) |
585 |
|
{ |
586 |
|
s = (astate)alist_elt_get_value(ptr); |
587 |
|
i = s->id_static; |
588 |
|
|
589 |
|
if (s == a->init_state) |
590 |
|
fa->init = i; |
591 |
|
if (s->accept == 1) |
592 |
|
fa->accept[i] = 1; |
593 |
|
|
594 |
|
for(l=0; l < MAX_TRANSITION_LETTERS; l++) |
595 |
|
if (s->next[l]) |
596 |
|
fa->trans[i][l] = s->next[l]->id_static; |
597 |
|
} |
598 |
|
|
599 |
|
return fa; |
600 |
} |
} |
|
#endif |
|
601 |
|
|
|
////////////////////////////////////////////////// |
|
|
////////////////////////////////////////////////// |
|
602 |
|
|
603 |
#ifdef DEBUG_RE |
/* ************************************************** * |
604 |
|
* ************************************************** * |
605 |
|
* ************************************************** */ |
606 |
|
|
607 |
static void |
static void |
608 |
print_automaton_nodes(FILE* f, automaton a) |
s_automaton_print_nodes(FILE* f, Automaton a) |
609 |
{ |
{ |
610 |
int state; |
char * sid; |
611 |
for(state=0; state < (1 << a->nterm); state++) |
astate s; |
612 |
|
alist_elt ptr; |
613 |
|
for(ptr = alist_get_first(a->states) ; ptr != NULL ; ptr = alist_get_next(a->states,ptr)) |
614 |
{ |
{ |
615 |
if (a->marque[state]) |
s = alist_elt_get_value(ptr); |
616 |
{ |
sid = s_automaton_id_to_str(s->id); |
617 |
fprintf(f,"\t%d [label = \"%x\"",state,state); |
fprintf(f,"\t\"%s\" [label = \"%s\"",sid,sid); |
618 |
if (a->init == state) |
if (s == a->init_state) |
619 |
{ |
{ |
620 |
fprintf(f,", style = filled, color=lightgrey"); |
fprintf(f,", style = filled, color=lightgrey"); |
621 |
} |
} |
622 |
if (a->accept[state]) |
if (s->accept) |
623 |
{ |
{ |
624 |
fprintf(f,", shape = doublecircle"); |
fprintf(f,", shape = doublecircle"); |
|
} |
|
|
fprintf(f,"];\n"); |
|
625 |
} |
} |
626 |
|
fprintf(f,"];\n"); |
627 |
} |
} |
628 |
fprintf(f,"\n"); |
fprintf(f,"\n"); |
629 |
} |
} |
|
#endif |
|
630 |
|
|
|
#ifdef DEBUG_RE |
|
631 |
static void |
static void |
632 |
print_automaton_edges(FILE* f, automaton a) |
s_automaton_print_edges(FILE* f, Automaton a) |
633 |
{ |
{ |
634 |
int state,letter; |
int letter; |
635 |
for(state=0; state < (1 << a->nterm); state++) |
char * sid; |
636 |
|
astate s; |
637 |
|
alist_elt ptr; |
638 |
|
for(ptr = alist_get_first(a->states) ; ptr != NULL ; ptr = alist_get_next(a->states,ptr)) |
639 |
{ |
{ |
640 |
if (a->marque[state]) |
s = (astate)alist_elt_get_value(ptr); |
641 |
|
for(letter=0; letter < 255; letter++) |
642 |
{ |
{ |
643 |
for(letter=0; letter < 255; letter++) |
if (s->next[letter]) |
644 |
{ |
{ |
645 |
if (a->Dtrans[state][letter]) |
sid = s_automaton_id_to_str(s->id); |
646 |
{ |
fprintf(f,"\t\"%s\" -> ",sid); |
647 |
fprintf(f,"\t%d -> %d [label = \"",state,a->Dtrans[state][letter]); |
sid = s_automaton_id_to_str(s->next[letter]->id); |
648 |
regexp_print_letter(f,letter); |
fprintf(f,"\"%s\" [label = \"",sid); |
649 |
fprintf(f,"\"];\n"); |
regexp_print_letter(f,letter); |
650 |
} |
fprintf(f,"\"];\n"); |
651 |
} |
} |
652 |
} |
} |
653 |
} |
} |
654 |
} |
} |
|
#endif |
|
655 |
|
|
656 |
#ifdef DEBUG_RE |
static void |
657 |
void |
s_automaton_dump(Automaton a, char* filename) |
|
automaton_dump(automaton a, char* filename) |
|
658 |
{ |
{ |
659 |
FILE* f; |
FILE* f; |
660 |
pid_t pid; |
pid_t pid; |
661 |
|
|
662 |
|
if (a == NULL) |
663 |
|
return; |
664 |
f=fopen(filename,"w"); |
f=fopen(filename,"w"); |
665 |
fprintf(f,"digraph automaton {\n"); |
fprintf(f,"digraph automaton {\n"); |
666 |
print_automaton_nodes(f,a); |
s_automaton_print_nodes(f,a); |
667 |
print_automaton_edges(f,a); |
s_automaton_print_edges(f,a); |
668 |
fprintf(f,"fontsize=20;\n"); |
fprintf(f,"fontsize=20;\n"); |
669 |
fprintf(f,"}\n"); |
fprintf(f,"}\n"); |
670 |
fclose(f); |
fclose(f); |
680 |
} |
} |
681 |
#endif |
#endif |
682 |
} |
} |
|
#endif |
|
683 |
|
|
684 |
////////////////////////////////////////////////// |
/* ************************************************** * |
685 |
////////////////////////////////////////////////// |
* ************************************************** * |
686 |
|
* ************************************************** */ |
687 |
|
|