fa *mkdfa(const char *s, int anchor) /* does the real work of making a dfa */
/* anchor = 1 for anchored matches, else 0 */
{
Node *p, *p1;
fa *f;
p = reparse(s);
p1 = op2(CAT, op2(STAR, op2(ALL, NIL, NIL), NIL), p);
/* put ALL STAR in front of reg. exp. */
p1 = op2(CAT, p1, op2(FINAL, NIL, NIL));
/* put FINAL after reg. exp. */
poscnt = 0;
penter(p1); /* enter parent pointers and leaf indices */
if ((f = (fa *) calloc(1, sizeof(fa) + poscnt*sizeof(rrow))) == NULL)
overflo("out of space for fa");
f->accept = poscnt-1; /* penter has computed number of positions in re */
cfoll(f, p1); /* set up follow sets */
freetr(p1);
if ((f->posns[0] = (int *) calloc(1, *(f->re[0].lfollow)*sizeof(int))) == NULL)
overflo("out of space in makedfa");
if ((f->posns[1] = (int *) calloc(1, sizeof(int))) == NULL)
overflo("out of space in makedfa");
*f->posns[1] = 0;
f->initstat = makeinit(f, anchor);
f->anchor = anchor;
f->restr = (uschar *) tostring(s);
return f;
}
void cfoll(fa *f, Node *v) /* enter follow set of each leaf of vertex v into lfollow[leaf] */
{
int i;
int *p;
switch (type(v)) {
ELEAF
LEAF
f->re[info(v)].ltype = type(v);
f->re[info(v)].lval.np = right(v);
while (f->accept >= maxsetvec) { /* guessing here! */
setvec = reallocarray(setvec, maxsetvec,
4 * sizeof(int));
tmpset = reallocarray(tmpset, maxsetvec,
4 * sizeof(int));
if (setvec == 0 || tmpset == 0)
overflo("out of space in cfoll()");
maxsetvec *= 4;
}
for (i = 0; i <= f->accept; i++)
setvec[i] = 0;
setcnt = 0;
follow(v); /* computes setvec and setcnt */
if ((p = (int *) calloc(setcnt+1, sizeof(int))) == NULL)
overflo("out of space building follow set");
f->re[info(v)].lfollow = p;
*p = setcnt;
for (i = f->accept; i >= 0; i--)
if (setvec[i] == 1)
*++p = i;
break;
UNARY
cfoll(f,left(v));
break;
case CAT:
case OR:
cfoll(f,left(v));
cfoll(f,right(v));
break;
default: /* can't happen */
FATAL("can't happen: unknown type %d in cfoll", type(v));
}
}
void
main(int argc, char **argv)
{
int i;
ARGBEGIN {
# ifdef DEBUG
case 'd': debug++; break;
case 'y': yydebug = TRUE; break;
# endif
case 't': case 'T':
Binit(&fout, 1, OWRITE);
errorf= 2;
foutopen = 1;
break;
case 'v': case 'V':
report = 1;
break;
case 'n': case 'N':
report = 0;
break;
case '9':
nine = 1;
break;
default:
warning("Unknown option %c", ARGC());
} ARGEND
sargc = argc;
sargv = argv;
if (argc > 0){
yyfile = argv[fptr++];
fin = Bopen(yyfile, OREAD);
if(fin == 0)
error ("%s - can't open file: %r", yyfile);
sargc--;
sargv++;
}
else {
yyfile = "/fd/0";
fin = myalloc(sizeof(Biobuf), 1);
if(fin == 0)
exits("core");
Binit(fin, 0, OREAD);
}
if(Bgetc(fin) == Beof) /* no input */
exits(0);
Bseek(fin, 0, 0);
gch();
/* may be gotten: def, subs, sname, stchar, ccl, dchar */
get1core();
/* may be gotten: name, left, right, nullstr, parent, ptr */
strcpy((char*)sp, "INITIAL");
sname[0] = sp;
sp += strlen("INITIAL") + 1;
sname[1] = 0;
if(yyparse()) exits("error"); /* error return code */
/* may be disposed of: def, subs, dchar */
free1core();
/* may be gotten: tmpstat, foll, positions, gotof, nexts, nchar, state, atable, sfall, cpackflg */
get2core();
ptail();
mkmatch();
# ifdef DEBUG
if(debug) pccl();
# endif
sect = ENDSECTION;
if(tptr>0)cfoll(tptr-1);
# ifdef DEBUG
if(debug)pfoll();
# endif
cgoto();
# ifdef DEBUG
if(debug){
print("Print %d states:\n",stnum+1);
for(i=0;i<=stnum;i++)stprt(i);
}
# endif
/* may be disposed of: positions, tmpstat, foll, state, name, left, right, parent, ccl, stchar, sname */
/* may be gotten: verify, advance, stoff */
free2core();
get3core();
layout();
/* may be disposed of: verify, advance, stoff, nexts, nchar,
gotof, atable, ccpackflg, sfall */
# ifdef DEBUG
free3core();
# endif
fother = Bopen(cname,OREAD);
if(fother == 0)
error("Lex driver missing, file %s: %r",cname);
while ( (i=Bgetc(fother)) != Beof)
Bputc(&fout, i);
Bterm(fother);
Bterm(&fout);
if(
# ifdef DEBUG
debug ||
# endif
report == 1)statistics();
if (fin)
Bterm(fin);
exits(0); /* success return code */
}
int
main(int argc, char **argv)
{
int i;
int c;
char *apath = NULL;
char *ypath;
Boolean eoption = 0, woption = 0;
sargv = argv;
sargc = argc;
(void) setlocale(LC_ALL, "");
#ifdef DEBUG
while ((c = getopt(argc, argv, "dyctvnewVQ:Y:")) != EOF) {
#else
while ((c = getopt(argc, argv, "ctvnewVQ:Y:")) != EOF) {
#endif
switch (c) {
#ifdef DEBUG
case 'd':
debug++;
break;
case 'y':
yydebug = TRUE;
break;
#endif
case 'V':
(void) fprintf(stderr, "lex: %s %s\n",
(const char *)SGU_PKG,
(const char *)SGU_REL);
break;
case 'Q':
v_stmp = optarg;
if (*v_stmp != 'y' && *v_stmp != 'n')
error(
"lex: -Q should be followed by [y/n]");
break;
case 'Y':
apath = (char *)malloc(strlen(optarg) +
sizeof ("/nceucform") + 1);
if (apath == NULL)
error("No available memory "
"for directory name.");
else
apath = strcpy(apath, optarg);
break;
case 'c':
ratfor = FALSE;
break;
case 't':
fout = stdout;
break;
case 'v':
report = 1;
break;
case 'n':
report = 0;
break;
case 'w':
case 'W':
woption = 1;
handleeuc = 1;
widecio = 1;
break;
case 'e':
case 'E':
eoption = 1;
handleeuc = 1;
widecio = 0;
break;
default:
(void) fprintf(stderr,
"Usage: lex [-ewctvnV] [-Y directory] "
"[-Q(y/n)] [file]\n");
exit(1);
}
}
if (woption && eoption) {
error(
"You may not specify both -w and -e simultaneously.");
}
no_input = argc - optind;
if (no_input) {
/* XCU4: recognize "-" file operand for stdin */
if (strcmp(argv[optind], "-") == 0)
fin = stdin;
else {
fin = fopen(argv[optind], "r");
if (fin == NULL)
error(
"Can't open input file -- %s", argv[optind]);
}
} else
fin = stdin;
/* may be gotten: def, subs, sname, schar, ccl, dchar */
(void) gch();
/* may be gotten: name, left, right, nullstr, parent */
get1core();
scopy(L_INITIAL, sp);
sname[0] = sp;
sp += slength(L_INITIAL) + 1;
sname[1] = 0;
/* XCU4: %x exclusive start */
exclusive[0] = 0;
if (!handleeuc) {
/*
* Set ZCH and ncg to their default values
* as they may be needed to handle %t directive.
*/
ZCH = ncg = NCH; /* ncg behaves as constant in this mode. */
}
/* may be disposed of: def, subs, dchar */
if (yyparse())
exit(1); /* error return code */
if (handleeuc) {
ncg = ncgidtbl * 2;
ZCH = ncg;
if (ncg >= MAXNCG)
error(
"Too complex rules -- requires too many char groups.");
sortcgidtbl();
}
repbycgid(); /* Call this even in ASCII compat. mode. */
/*
* maybe get:
* tmpstat, foll, positions, gotof, nexts,
* nchar, state, atable, sfall, cpackflg
*/
free1core();
get2core();
ptail();
mkmatch();
#ifdef DEBUG
if (debug)
pccl();
#endif
sect = ENDSECTION;
if (tptr > 0)
cfoll(tptr-1);
#ifdef DEBUG
if (debug)
pfoll();
#endif
cgoto();
#ifdef DEBUG
if (debug) {
(void) printf("Print %d states:\n", stnum + 1);
for (i = 0; i <= stnum; i++)
stprt(i);
}
#endif
/*
* may be disposed of:
* positions, tmpstat, foll, state, name,
* left, right, parent, ccl, schar, sname
* maybe get: verify, advance, stoff
*/
free2core();
get3core();
layout();
/*
* may be disposed of:
* verify, advance, stoff, nexts, nchar,
* gotof, atable, ccpackflg, sfall
*/
#ifdef DEBUG
free3core();
#endif
if (handleeuc) {
if (ratfor)
error("Ratfor is not supported by -w or -e option.");
ypath = EUCNAME;
}
else
ypath = ratfor ? RATNAME : CNAME;
if (apath != NULL)
ypath = strcat(apath, strrchr(ypath, '/'));
fother = fopen(ypath, "r");
if (fother == NULL)
error("Lex driver missing, file %s", ypath);
while ((i = getc(fother)) != EOF)
(void) putc((char)i, fout);
(void) fclose(fother);
(void) fclose(fout);
free(apath);
if (report == 1)
statistics();
(void) fclose(stdout);
(void) fclose(stderr);
return (0); /* success return code */
}
static void
get1core(void)
{
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
ccptr = ccl = (CHR *)myalloc(CCLSIZE, sizeof (*ccl));
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
pcptr = pchar = (CHR *)myalloc(pchlen, sizeof (*pchar));
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
def = (CHR **)myalloc(DEFSIZE, sizeof (*def));
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
subs = (CHR **)myalloc(DEFSIZE, sizeof (*subs));
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
dp = dchar = (CHR *)myalloc(DEFCHAR, sizeof (*dchar));
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
sname = (CHR **)myalloc(STARTSIZE, sizeof (*sname));
/* XCU4: exclusive start array */
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
exclusive = (int *)myalloc(STARTSIZE, sizeof (*exclusive));
/*LINTED: E_BAD_PTR_CAST_ALIGN*/
sp = schar = (CHR *)myalloc(STARTCHAR, sizeof (*schar));
if (ccl == 0 || def == 0 ||
pchar == 0 || subs == 0 || dchar == 0 ||
sname == 0 || exclusive == 0 || schar == 0)
error("Too little core to begin");
}
void
cfoll(int v)
{
int i,j,k;
uint8_t *p;
i = name[v];
if(i < NCH) i = 1; /* character */
switch(i){
case 1: case RSTR: case RCCL: case RNCCL: case RNULLS:
for(j=0;j<tptr;j++)
tmpstat[j] = FALSE;
count = 0;
follow(v);
# ifdef PP
padd(foll,v); /* packing version */
# endif
# ifndef PP
add(foll,v); /* no packing version */
# endif
if(i == RSTR) cfoll(left[v]);
else if(i == RCCL || i == RNCCL){ /* compress ccl list */
for(j=1; j<NCH;j++)
symbol[j] = (i==RNCCL);
p = ptr[v];
while(*p)
symbol[*p++] = (i == RCCL);
p = pcptr;
for(j=1;j<NCH;j++)
if(symbol[j]){
for(k=0;p+k < pcptr; k++)
if(cindex[j] == *(p+k))
break;
if(p+k >= pcptr)*pcptr++ = cindex[j];
}
*pcptr++ = 0;
if(pcptr > pchar + pchlen)
error("Too many packed character classes");
ptr[v] = p;
name[v] = RCCL; /* RNCCL eliminated */
# ifdef DEBUG
if(debug && *p){
print("ccl %d: %d",v,*p++);
while(*p)
print(", %d",*p++);
print("\n");
}
# endif
}
break;
case CARAT:
cfoll(left[v]);
break;
case STAR: case PLUS: case QUEST: case RSCON:
cfoll(left[v]);
break;
case BAR: case RCAT: case DIV: case RNEWE:
cfoll(left[v]);
cfoll(right[v]);
break;
# ifdef DEBUG
case FINAL:
case S1FINAL:
case S2FINAL:
break;
default:
warning("bad switch cfoll %d",v);
# endif
}
}
