-/* $Id: dfa.c,v 1.34 2005-03-30 09:25:23 adam Exp $
- Copyright (C) 1995-2005
- Index Data ApS
-
-This file is part of the Zebra server.
+/* This file is part of the Zebra server.
+ Copyright (C) Index Data
Zebra is free software; you can redistribute it and/or modify it under
the terms of the GNU General Public License as published by the Free
for more details.
You should have received a copy of the GNU General Public License
-along with Zebra; see the file LICENSE.zebra. If not, write to the
-Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA
-02111-1307, USA.
+along with this program; if not, write to the Free Software
+Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
+
*/
+#if HAVE_CONFIG_H
+#include <config.h>
+#endif
#include <stdio.h>
#include <assert.h>
#include "dfap.h"
#include "imalloc.h"
-#define DFA_OPEN_RANGE 1
-
#define CAT 16000
#define OR 16001
#define STAR 16002
BSet charset);
static void term_Tnode (struct DFA_parse *parse_info);
-static void
- del_followpos (struct DFA_parse *parse_info),
- init_pos (struct DFA_parse *parse_info),
+static void
+ del_followpos (struct DFA_parse *parse_info),
+ init_pos (struct DFA_parse *parse_info),
del_pos (struct DFA_parse *parse_info),
mk_dfa_tran (struct DFA_parse *parse_info, struct DFA_states *dfas),
add_follow (struct DFA_parse *parse_info, DFASet lastpos, DFASet firstpos),
nextchar (struct DFA_parse *parse_info, int *esc),
read_charset (struct DFA_parse *parse_info);
-static const char
+static const char
*str_char (unsigned c);
#define L_LP 1
*expr_4 (struct DFA_parse *parse_info);
static struct Tnode *expr_1 (struct DFA_parse *parse_info)
-{
+{
struct Tnode *t1, *t2, *tn;
if (!(t1 = expr_2 (parse_info)))
lex (parse_info);
if (!(t2 = expr_2 (parse_info)))
return t2;
-
+
tn = mk_Tnode (parse_info);
tn->pos = OR;
tn->u.p[0] = t1;
static struct Tnode *expr_2 (struct DFA_parse *parse_info)
{
struct Tnode *t1, *t2, *tn;
-
+
if (!(t1 = expr_3 (parse_info)))
return t1;
while (parse_info->lookahead == L_WILD ||
{
if (!(t2 = expr_3 (parse_info)))
return t2;
-
+
tn = mk_Tnode (parse_info);
tn->pos = CAT;
tn->u.p[0] = t1;
t2 = mk_Tnode (parse_info);
t2->pos = ++parse_info->position;
t2->u.ch[1] = t2->u.ch[0] = '\n';
-
+
tn = mk_Tnode (parse_info);
tn->pos = CAT;
tn->u.p[0] = t1;
tn->u.p[1] = t2;
t1 = tn;
-
+
lex (parse_info);
}
}
t2->pos = ++parse_info->position;
t2->u.ch[0] = -(++parse_info->rule);
t2->u.ch[1] = start_anchor_flag ? 0 : -(parse_info->rule);
-
+
*tnp = mk_Tnode(parse_info);
(*tnp)->pos = CAT;
(*tnp)->u.p[0] = t1;
static int read_charset (struct DFA_parse *parse_info)
{
- int i, ch0, ch1, esc0, esc1, cc = 0;
+ int i, ch0, esc0, cc = 0;
parse_info->look_chars = mk_BSet (&parse_info->charset);
res_BSet (parse_info->charset, parse_info->look_chars);
cc = 1;
ch0 = nextchar_set (parse_info, &esc0);
}
+ /**
+ ch0 is last met character
+ ch1 is "next" char
+ */
while (ch0 != 0)
{
+ int ch1, esc1;
if (!esc0 && ch0 == ']')
break;
if (!esc0 && ch0 == '-')
}
else
{
- if (parse_info->cmap)
- {
- const char **mapto;
- char mapfrom[2];
- const char *mcp = mapfrom;
- mapfrom[0] = ch0;
- mapto = (*parse_info->cmap)(parse_info->cmap_data, &mcp, 1);
- assert (mapto);
- ch0 = mapto[0][0];
- }
+ if (ch0 == 1)
+ {
+ ch0 = nextchar(parse_info, &esc0);
+ }
+ else
+ {
+ if (parse_info->cmap)
+ {
+ const char **mapto;
+ char mapfrom[2];
+ const char *mcp = mapfrom;
+ mapfrom[0] = ch0;
+ mapto = parse_info->cmap(parse_info->cmap_data, &mcp, 1);
+ assert (mapto);
+ ch0 = mapto[0][0];
+ }
+ }
add_BSet (parse_info->charset, parse_info->look_chars, ch0);
ch1 = nextchar_set (parse_info, &esc1);
}
int open_range = 0;
if ((ch1 = nextchar_set (parse_info, &esc1)) == 0)
break;
-#if DFA_OPEN_RANGE
if (!esc1 && ch1 == ']')
{
ch1 = 255;
open_range = 1;
}
-#else
- if (!esc1 && ch1 == ']')
+ else if (ch1 == 1)
{
- add_BSet (parse_info->charset, parse_info->look_chars, '-');
- break;
+ ch1 = nextchar(parse_info, &esc1);
}
-#endif
- if (!open_range && parse_info->cmap)
+ else if (parse_info->cmap)
{
const char **mapto;
char mapfrom[2];
assert (mapto);
ch1 = mapto[0][0];
}
- for (i=ch0; ++i<=ch1;)
+ for (i = ch0; ++i <= ch1;)
add_BSet (parse_info->charset, parse_info->look_chars, i);
- if (!open_range)
- ch0 = nextchar_set (parse_info, &esc0);
- else
+
+ if (open_range)
break;
+ ch0 = nextchar_set (parse_info, &esc0);
}
else
{
mapto = (*parse_info->cmap) (parse_info->cmap_data, &cp0, len);
assert (mapto);
-
+
parse_info->expr_ptr = (const unsigned char *) cp0;
parse_info->look_ch = ((unsigned char **) mapto)[i][0];
yaz_log (YLOG_DEBUG, "map from %c to %d", parse_info->expr_ptr[-1], parse_info->look_ch);
return map_l_char (parse_info);
else if (parse_info->look_ch == '[')
return read_charset(parse_info);
- else
+ else
{
const int *cc;
for (cc = parse_info->charMap; *cc; cc += 2)
static void init_pos (struct DFA_parse *parse_info)
{
- parse_info->posar = (struct Tnode **) imalloc (sizeof(struct Tnode*)
+ parse_info->posar = (struct Tnode **) imalloc (sizeof(struct Tnode*)
* (1+parse_info->position));
}
{
while (lastpos)
{
- parse_info->followpos[lastpos->value] =
+ parse_info->followpos[lastpos->value] =
union_DFASet (parse_info->poset,
parse_info->followpos[lastpos->value], firstpos);
lastpos = lastpos->next;
- }
+ }
}
static void dfa_trav (struct DFA_parse *parse_info, struct Tnode *n)
{
struct Tnode **posar = parse_info->posar;
DFASetType poset = parse_info->poset;
-
+
switch (n->pos)
{
case CAT:
pos_i = pos;
j = i = 0;
for (tran_set = dfa_from->set; tran_set; tran_set = tran_set->next)
- if ((c = posar[tran_set->value]->u.ch[0]) >= 0 && c <= max_char)
+ if ((c = posar[tran_set->value]->u.ch[0]) >= 0 && c <= max_char)
*pos_i++ = tran_set->value;
else if (c < 0)
{
{
char_0 = max_char+1;
for (pos_i = pos; (i = *pos_i) != -1; ++pos_i)
- if (posar[i]->u.ch[1] >= char_1
+ if (posar[i]->u.ch[1] >= char_1
&& (c=posar[i]->u.ch[0]) < char_0)
{
if (c < char_1)
break;
char_1 = max_char;
-
+
tran_set = mk_DFASet (poset);
for (pos_i = pos; (i = *pos_i) != -1; ++pos_i)
{
{
long i, j;
int k;
- printf ("%d/%d tree nodes used, %d bytes each\n",
- parse_info->use_Tnode, parse_info->max_Tnode, sizeof (struct Tnode));
+ printf ("%d/%d tree nodes used, %ld bytes each\n",
+ parse_info->use_Tnode, parse_info->max_Tnode, (long) sizeof (struct Tnode));
k = inf_BSetHandle (parse_info->charset, &i, &j);
- printf ("%ld/%ld character sets, %d bytes each\n",
- i/k, j/k, k*sizeof(BSetWord));
+ printf ("%ld/%ld character sets, %ld bytes each\n",
+ i/k, j/k, (long) k*sizeof(BSetWord));
k = inf_DFASetType (parse_info->poset, &i, &j);
printf ("%ld/%ld poset items, %d bytes each\n", i, j, k);
printf ("%d DFA states\n", dfas->no);
{
while ((cc[0] = cc[2]))
{
- cc[1] = cc[3];
+ cc[1] = cc[3];
cc += 2;
}
break;
static struct DFA_parse *dfa_parse_init (void)
{
- struct DFA_parse *parse_info =
+ struct DFA_parse *parse_info =
(struct DFA_parse *) imalloc (sizeof (struct DFA_parse));
parse_info->charset = mk_BSetHandle (255, 20);
parse_info->rule = 0;
parse_info->root = NULL;
+ /* initialize the anyset which by default does not include \n */
parse_info->anyset = mk_BSet (&parse_info->charset);
res_BSet (parse_info->charset, parse_info->anyset);
+ add_BSet (parse_info->charset, parse_info->anyset, '\n');
com_BSet (parse_info->charset, parse_info->anyset);
+
parse_info->use_Tnode = parse_info->max_Tnode = 0;
parse_info->start = parse_info->end = NULL;
parse_info->charMap = NULL;
return dfa;
}
+void dfa_anyset_includes_nl(struct DFA *dfa)
+{
+ add_BSet (dfa->parse_info->charset, dfa->parse_info->anyset, '\n');
+}
+
void dfa_set_cmap (struct DFA *dfa, void *vp,
const char **(*cmap)(void *vp, const char **from, int len))
{
dfa->parse_info->cmap_data = vp;
}
+int dfa_get_last_rule (struct DFA *dfa)
+{
+ return dfa->parse_info->rule;
+}
+
int dfa_parse (struct DFA *dfa, const char **pattern)
{
struct Tnode *top;
assert (dfa->parse_info);
parse_info = dfa->parse_info;
- if (!parse_info->cmap)
- {
- res_BSet (parse_info->charset, parse_info->anyset);
- add_BSet (parse_info->charset, parse_info->anyset, '\n');
- com_BSet (parse_info->charset, parse_info->anyset);
- }
do_parse (parse_info, pattern, &top);
if (parse_info->err_code)
return parse_info->err_code;
ifree (*dfap);
*dfap = NULL;
}
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * c-file-style: "Stroustrup"
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+
projects / idzebra-moved-to-github.git / blobdiff