/*
- * Copyright (C) 1994-1998, Index Data I/S
+ * Copyright (C) 1994-1998, Index Data
* All rights reserved.
* Sebastian Hammer, Adam Dickmeiss
*
* $Log: regxread.c,v $
- * Revision 1.14 1998-03-05 08:41:00 adam
+ * Revision 1.19 1998-11-03 10:22:39 adam
+ * Fixed memory leak that could occur for when large data1 node were
+ * concatenated. Data-type data1_nodes may have multiple nodes.
+ *
+ * Revision 1.18 1998/10/15 13:11:47 adam
+ * Added support for option -record for "end element". When specified
+ * end element will mark end-of-record when at outer-level.
+ *
+ * Revision 1.17 1998/07/01 10:13:51 adam
+ * Minor fix.
+ *
+ * Revision 1.16 1998/06/30 15:15:09 adam
+ * Tags are trimmed: white space removed before- and after the tag.
+ *
+ * Revision 1.15 1998/06/30 12:55:45 adam
+ * Bug fix.
+ *
+ * Revision 1.14 1998/03/05 08:41:00 adam
* Implemented rule contexts.
*
* Revision 1.13 1997/12/12 06:33:58 adam
#include <stdio.h>
#include <assert.h>
#include <string.h>
+#include <ctype.h>
#include <tpath.h>
#include <zebrautl.h>
parent = d1_stack[*d1_level -1];
assert (parent);
- if ((res=d1_stack[*d1_level]) && res->which == DATA1N_data)
- {
- if (elen + res->u.data.len <= DATA1_LOCALDATA)
- memcpy (res->u.data.data + res->u.data.len, ebuf, elen);
- else
- {
- char *nb = xmalloc (elen + res->u.data.len);
- memcpy (nb, res->u.data.data, res->u.data.len);
- memcpy (nb + res->u.data.len, ebuf, elen);
- if (res->u.data.len > DATA1_LOCALDATA)
- xfree (res->u.data.data);
- res->u.data.data = nb;
- res->destroy = destroy_data;
- }
- res->u.data.len += elen;
- }
+
+ res = data1_mk_node (spec->dh, spec->m);
+ res->parent = parent;
+ res->which = DATA1N_data;
+ res->u.data.what = DATA1I_text;
+ res->u.data.len = elen;
+ res->u.data.formatted_text = formatted_text;
+ if (elen > DATA1_LOCALDATA)
+ res->u.data.data = nmem_malloc (spec->m, elen);
else
- {
- res = data1_mk_node (spec->dh, spec->m);
- res->parent = parent;
- res->which = DATA1N_data;
- res->u.data.what = DATA1I_text;
- res->u.data.len = elen;
- res->u.data.formatted_text = formatted_text;
- if (elen > DATA1_LOCALDATA)
- res->u.data.data = nmem_malloc (spec->m, elen);
- else
- res->u.data.data = res->lbuf;
- memcpy (res->u.data.data, ebuf, elen);
- res->root = parent->root;
-
- parent->last_child = res;
- if (d1_stack[*d1_level])
- d1_stack[*d1_level]->next = res;
- else
- parent->child = res;
- d1_stack[*d1_level] = res;
- }
+ res->u.data.data = res->lbuf;
+ memcpy (res->u.data.data, ebuf, elen);
+ res->root = parent->root;
+
+ parent->last_child = res;
+ if (d1_stack[*d1_level])
+ d1_stack[*d1_level]->next = res;
+ else
+ parent->child = res;
+ d1_stack[*d1_level] = res;
}
static void execDataP (struct lexSpec *spec,
d1_stack[++(*d1_level)] = NULL;
}
+static void tagStrip (const char **tag, int *len)
+{
+ int i;
+
+ for (i = *len; i > 0 && isspace((*tag)[i-1]); --i)
+ ;
+ *len = i;
+ for (i = 0; i < *len && isspace((*tag)[i]); i++)
+ ;
+ *tag += i;
+ *len -= i;
+}
+
static void tagBegin (struct lexSpec *spec,
data1_node **d1_stack, int *d1_level,
const char *tag, int len)
logf (LOG_WARN, "in element begin. No record type defined");
return ;
}
-
+ tagStrip (&tag, &len);
+
res = data1_mk_node (spec->dh, spec->m);
res->parent = parent;
res->which = DATA1N_tag;
res->u.tag.get_bytes = -1;
if (len >= DATA1_LOCALDATA)
- len = DATA1_LOCALDATA-1;
- memcpy (res->lbuf, tag, len);
- res->lbuf[len] = '\0';
- res->u.tag.tag = res->lbuf;
+ res->u.tag.tag = nmem_malloc (spec->m, len+1);
+ else
+ res->u.tag.tag = res->lbuf;
+
+ memcpy (res->u.tag.tag, tag, len);
+ res->u.tag.tag[len] = '\0';
#if REGX_DEBUG
logf (LOG_DEBUG, "begin tag %s (%d)", res->u.tag.tag, *d1_level);
}
static void tagEnd (struct lexSpec *spec,
- data1_node **d1_stack, int *d1_level,
+ data1_node **d1_stack, int *d1_level, int min_level,
const char *tag, int len)
{
- while (*d1_level > 1)
+ tagStrip (&tag, &len);
+ while (*d1_level > min_level)
{
(*d1_level)--;
+ if (*d1_level == 0)
+ break;
if ((d1_stack[*d1_level]->which == DATA1N_tag) &&
(!tag ||
(strlen(d1_stack[*d1_level]->u.tag.tag) == (size_t) len &&
}
else if (!strcmp (p, "element"))
{
- r = execTok (spec, &s, arg_no, arg_start, arg_end,
- &cmd_str, &cmd_len);
-#if 0
- if (*d1_level == 1)
- {
- *d1_level = 0;
- returnCode = 0;
- }
-#endif
+ int min_level = 1;
+ while ((r = execTok (spec, &s, arg_no, arg_start, arg_end,
+ &cmd_str, &cmd_len)) == 3)
+ {
+ if (cmd_len==7 && !memcmp ("-record", cmd_str, cmd_len))
+ min_level = 0;
+ }
if (r > 2)
{
- tagEnd (spec, d1_stack, d1_level, cmd_str, cmd_len);
+ tagEnd (spec, d1_stack, d1_level, min_level,
+ cmd_str, cmd_len);
r = execTok (spec, &s, arg_no, arg_start, arg_end,
&cmd_str, &cmd_len);
}
else
- tagEnd (spec, d1_stack, d1_level, NULL, 0);
+ tagEnd (spec, d1_stack, d1_level, min_level, NULL, 0);
+ if (*d1_level == 0)
+ {
+#if REGX_DEBUG
+ logf (LOG_DEBUG, "end element end records");
+#endif
+ returnCode = 0;
+ }
+
}
else if (!strcmp (p, "context"))
{
&cmd_str, &cmd_len);
} while (r > 1);
if (element_str)
- tagEnd (spec, d1_stack, d1_level, NULL, 0);
+ tagEnd (spec, d1_stack, d1_level, 1, NULL, 0);
}
else if (!strcmp (p, "unread"))
{
return NULL;
}
}
+ curLexSpec->dh = p->dh;
if (!p->offset)
{
curLexSpec->f_win_start = 0;
projects / idzebra-moved-to-github.git / blobdiff