-/* $Id: charmap.c,v 1.37 2005-06-14 12:42:49 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) 2004-2013 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
+*/
-/*
+/**
+ * \file charmap.c
+ * \brief character conversions (.chr)
+ *
* Support module to handle character-conversions into and out of the
* Zebra dictionary.
*/
+#if HAVE_CONFIG_H
+#include <config.h>
+#endif
#include <ctype.h>
#include <string.h>
#include <stdlib.h>
const char *CHR_UNKNOWN = "\001";
const char *CHR_SPACE = "\002";
const char *CHR_CUT = "\003";
-const char *CHR_BASE = "\005";
+const char *CHR_BASE = "\005"; /* CHECK CHR_BASE_CHAR as well */
struct chrmaptab_info
{
/*
* General argument structure for callback functions (internal use only)
*/
-typedef struct chrwork
+typedef struct chrwork
{
chrmaptab map;
char string[CHR_MAXSTR+1];
}
if (!len)
{
- if (!root->target || !root->target[0] || strcmp(root->target[0], to))
+ if (!root->target || !root->target[0] ||
+ strcmp((const char *) root->target[0], to))
{
- if (from_0 &&
+ if (from_0 &&
root->target && root->target[0] && root->target[0][0] &&
- strcmp (root->target[0], CHR_UNKNOWN))
+ strcmp((const char *) root->target[0], CHR_UNKNOWN))
{
- yaz_log (YLOG_WARN, "duplicate entry for charmap from '%s'",
- from_0);
+ yaz_log(YLOG_WARN, "duplicate entry for charmap from '%s'",
+ from_0);
}
root->target = (unsigned char **)
nmem_malloc(nmem, sizeof(*root->target)*2);
root->children[i] = 0;
}
if (!(root->children[(unsigned char) *from] =
- set_map_string(root->children[(unsigned char) *from], nmem,
- from + 1, len - 1, to, from_0)))
+ set_map_string(root->children[(unsigned char) *from], nmem,
+ from + 1, len - 1, to, from_0)))
return 0;
}
return root;
}
-static chr_t_entry *find_entry(chr_t_entry *t, const char **from, int len)
-{
- chr_t_entry *res;
-
- if (len && t->children && t->children[(unsigned char) **from])
- {
- const char *pos = *from;
-
- (*from)++;
- if ((res = find_entry(t->children[(unsigned char) *pos],
- from, len - 1)))
- return res;
- /* no match */
- *from = pos;
- }
- /* no children match. use ourselves, if we have a target */
- return t->target ? t : 0;
-}
-
static chr_t_entry *find_entry_x(chr_t_entry *t, const char **from, int *len, int first)
{
chr_t_entry *res;
else
res = 0;
/* otherwhise there was no match on beginning of field, move on */
- }
-
+ }
+
if (!res && t->children[(unsigned char) **from])
{
(*len)--;
chr_t_entry *t = maptab->q_input;
chr_t_entry *res;
int len_tmp[2];
-
+
len_tmp[0] = len;
len_tmp[1] = -1;
if (!(res = find_entry_x(t, from, len_tmp, first)))
const char *chr_map_output(chrmaptab maptab, const char **from, int len)
{
unsigned char c = ** (unsigned char **) from;
- (*from)++;
- return (const char*) maptab->output[c];
+ const char *out = (const char*) maptab->output[c];
+
+ if (out)
+ (*from)++;
+ return out;
}
static int zebra_ucs4_strlen(ucs4_t *s)
ucs4_t i = 0;
char fmtstr[8];
- yaz_log (YLOG_DEBUG, "prim_w %.3s", (char *) *s);
if (**s == '\\' && 1[*s])
{
(*s)++;
case 'n': c = '\n'; (*s)++; break;
case 't': c = '\t'; (*s)++; break;
case 's': c = ' '; (*s)++; break;
- case 'x':
+ case 'x':
if (zebra_ucs4_strlen(*s) >= 3)
{
fmtstr[0] = (*s)[1];
c = **s;
++(*s);
}
- yaz_log (YLOG_DEBUG, "out %d", c);
+ yaz_log(YLOG_DEBUG, "out %d", c);
return c;
}
{
chrmaptab tab = (chrmaptab) data;
char tmp[2];
-
+
tmp[0] = num; tmp[1] = '\0';
tab->input = set_map_string(tab->input, tab->nmem, s, strlen(s), tmp, 0);
tab->output[num + tab->base_uppercase] =
(unsigned char *) nmem_strdup(tab->nmem, s);
}
-/*
+/*
* Callback function.
* Add a space-entry to the value space.
*/
(char*) CHR_SPACE, 0);
}
-/*
+/*
* Callback function.
* Add a space-entry to the value space.
*/
static void fun_add_equivalent_string(const char *s, void *data, int num)
{
chr_equiv_work *arg = (chr_equiv_work *) data;
-
+
if (arg->no_eq == CHR_MAXEQUIV)
return;
arg->eq[arg->no_eq++] = nmem_strdup(arg->nmem, s);
chrwork *arg = (chrwork *) data;
assert(arg->map->input);
- yaz_log (YLOG_DEBUG, "set map %.*s", (int) strlen(s), s);
+ yaz_log(YLOG_DEBUG, "set map %.*s", (int) strlen(s), s);
set_map_string(arg->map->input, arg->map->nmem, s, strlen(s), arg->string,
0);
for (s = arg->string; *s; s++)
- yaz_log (YLOG_DEBUG, " %3d", (unsigned char) *s);
+ yaz_log(YLOG_DEBUG, " %3d", (unsigned char) *s);
}
-static int scan_to_utf8 (yaz_iconv_t t, ucs4_t *from, size_t inlen,
+static int scan_to_utf8(yaz_iconv_t t, ucs4_t *from, size_t inlen,
char *outbuf, size_t outbytesleft)
{
size_t inbytesleft = inlen * sizeof(ucs4_t);
char *inbuf = (char*) from;
size_t ret;
-
+
if (t == 0)
*outbuf++ = *from; /* ISO-8859-1 is OK here */
else
{
- ret = yaz_iconv (t, &inbuf, &inbytesleft, &outbuf, &outbytesleft);
+ ret = yaz_iconv(t, &inbuf, &inbytesleft, &outbuf, &outbytesleft);
+ if (ret != (size_t) (-1))
+ ret = yaz_iconv(t, 0, 0, &outbuf, &outbytesleft);
+
+
if (ret == (size_t) (-1))
{
yaz_log(YLOG_LOG, "from: %2X %2X %2X %2X",
from[0], from[1], from[2], from[3]);
- yaz_log (YLOG_WARN|YLOG_ERRNO, "bad unicode sequence");
+ yaz_log(YLOG_WARN|YLOG_ERRNO, "bad unicode sequence");
return -1;
}
}
ucs4_t arg[512];
ucs4_t arg_prim[512];
- ucs4_t *s0, *s = arg;
+ ucs4_t *s = arg;
ucs4_t c, begin, end;
size_t i;
char *inbuf = s_native;
size_t outbytesleft = sizeof(arg)-4;
size_t inbytesleft = strlen(s_native);
- size_t ret;
+ size_t ret;
ret = yaz_iconv(t_unicode, &inbuf, &inbytesleft,
&outbuf, &outbytesleft);
+ if (ret != (size_t)(-1))
+ ret = yaz_iconv(t_unicode, 0, 0, &outbuf, &outbytesleft);
+
if (ret == (size_t)(-1))
return -1;
i = (outbuf - (char*) arg)/sizeof(ucs4_t);
}
else
- {
+ {
for (i = 0; s_native[i]; i++)
arg[i] = s_native[i] & 255; /* ISO-8859-1 conversion */
}
s++;
for (c = begin; c <= end; c++)
{
- if (scan_to_utf8 (t_utf8, &c, 1, str, sizeof(str)-1))
+ if (scan_to_utf8(t_utf8, &c, 1, str, sizeof(str)-1))
return -1;
(*fun)(str, data, num ? (*num)++ : 0);
}
break;
case '(':
++s;
- s0 = s; i = 0;
+ i = 0;
while (*s != ')' || s[-1] == '\\')
+ {
+ if (*s == '\0')
+ {
+ yaz_log(YLOG_FATAL, "Missing ) in charmap");
+ return -1;
+ }
arg_prim[i++] = zebra_prim_w(&s);
+ }
arg_prim[i] = 0;
- if (scan_to_utf8 (t_utf8, arg_prim, zebra_ucs4_strlen(arg_prim), str, sizeof(str)-1))
+ if (scan_to_utf8(t_utf8, arg_prim, zebra_ucs4_strlen(arg_prim), str, sizeof(str)-1))
return -1;
(*fun)(str, data, num ? (*num)++ : 0);
s++;
break;
default:
c = zebra_prim_w(&s);
- if (scan_to_utf8 (t_utf8, &c, 1, str, sizeof(str)-1))
+ if (scan_to_utf8(t_utf8, &c, 1, str, sizeof(str)-1))
return -1;
(*fun)(str, data, num ? (*num)++ : 0);
}
return 0;
}
-chrmaptab chrmaptab_create(const char *tabpath, const char *name, int map_only,
- const char *tabroot)
+chrmaptab chrmaptab_create(const char *tabpath, const char *name,
+ const char *tabroot)
{
FILE *f;
char line[512], *argv[50];
chrmaptab res;
int lineno = 0;
+ int no_directives = 0;
int errors = 0;
int argc, num = (int) *CHR_BASE, i;
NMEM nmem;
unsigned endian = 31;
const char *ucs4_native = "UCS-4";
- if (*(char*) &endian == 31) /* little endian? */
- ucs4_native = "UCS-4LE";
-
- t_utf8 = yaz_iconv_open ("UTF-8", ucs4_native);
-
- yaz_log (YLOG_DEBUG, "maptab %s open", name);
+ yaz_log(YLOG_DEBUG, "maptab %s open", name);
if (!(f = yaz_fopen(tabpath, name, "r", tabroot)))
{
yaz_log(YLOG_WARN|YLOG_ERRNO, "%s", name);
return 0;
}
- nmem = nmem_create ();
+
+ if (*(char*) &endian == 31) /* little endian? */
+ ucs4_native = "UCS-4LE";
+
+ t_utf8 = yaz_iconv_open("UTF-8", ucs4_native);
+
+ nmem = nmem_create();
res = (chrmaptab) nmem_malloc(nmem, sizeof(*res));
res->nmem = nmem;
res->input = (chr_t_entry *) nmem_malloc(res->nmem, sizeof(*res->input));
nmem_malloc(res->nmem, sizeof(*res->input));
res->input->children[i]->children = 0;
res->input->children[i]->target = (unsigned char **)
- nmem_malloc (res->nmem, 2 * sizeof(unsigned char *));
+ nmem_malloc(res->nmem, 2 * sizeof(unsigned char *));
res->input->children[i]->target[1] = 0;
- if (map_only)
- {
- res->input->children[i]->target[0] = (unsigned char *)
- nmem_malloc (res->nmem, 2 * sizeof(unsigned char));
- res->input->children[i]->target[0][0] = i;
- res->input->children[i]->target[0][1] = 0;
- }
- else
- res->input->children[i]->target[0] = (unsigned char*) CHR_UNKNOWN;
+ res->input->children[i]->target[0] = (unsigned char*) CHR_UNKNOWN;
}
res->q_input = (chr_t_entry *)
nmem_malloc(res->nmem, sizeof(*res->q_input));
res->base_uppercase = 0;
while (!errors && (argc = readconf_line(f, &lineno, line, 512, argv, 50)))
- if (!map_only && !yaz_matchstr(argv[0], "lowercase"))
+ {
+ no_directives++;
+ if (!yaz_matchstr(argv[0], "lowercase"))
{
if (argc != 2)
{
res->output[(int) *CHR_UNKNOWN + num] = (unsigned char*) "@";
num = (int) *CHR_BASE;
}
- else if (!map_only && !yaz_matchstr(argv[0], "uppercase"))
+ else if (!yaz_matchstr(argv[0], "uppercase"))
{
if (!res->base_uppercase)
{
++errors;
}
}
- else if (!map_only && !yaz_matchstr(argv[0], "space"))
+ else if (!yaz_matchstr(argv[0], "space"))
{
if (argc != 2)
{
++errors;
}
}
- else if (!map_only && !yaz_matchstr(argv[0], "cut"))
+ else if (!yaz_matchstr(argv[0], "cut"))
{
if (argc != 2)
{
}
w.nmem = res->nmem;
w.no_eq = 0;
- if (scan_string(argv[1], t_unicode, t_utf8,
+ if (scan_string(argv[1], t_unicode, t_utf8,
fun_add_equivalent_string, &w, 0) < 0)
{
yaz_log(YLOG_FATAL, "equivalent: invalid string");
}
else if (!yaz_matchstr(argv[0], "encoding"))
{
- /*
- * Fix me. When t_unicode==0 and use encoding directive in *.chr file the beheviour of the
- * zebra need to comment next part of code.
- */
-
- /* Original code */
-#if 1
if (t_unicode != 0)
- yaz_iconv_close (t_unicode);
- t_unicode = yaz_iconv_open (ucs4_native, argv[1]);
-#endif
- /*
- * Fix me. It is additional staff for conversion of characters from local encoding
- * of *.chr file to UTF-8 (internal encoding).
- * NOTE: The derective encoding must be first directive in *.chr file.
- */
- /* For whatever reason Oleg enabled this.. */
-#if 0
- if (t_utf8 != 0)
- yaz_iconv_close(t_utf8);
- t_utf8 = yaz_iconv_open ("UTF-8", argv[1]);
-#endif
+ yaz_iconv_close(t_unicode);
+ t_unicode = yaz_iconv_open(ucs4_native, argv[1]);
}
else
{
yaz_log(YLOG_WARN, "Syntax error at '%s' in %s", line, name);
+ errors++;
}
-
+ }
yaz_fclose(f);
+ if (no_directives == 0)
+ {
+ yaz_log(YLOG_WARN, "No directives in '%s'", name);
+ errors++;
+ }
if (errors)
{
chrmaptab_destroy(res);
res = 0;
}
- yaz_log (YLOG_DEBUG, "maptab %s close %d errors", name, errors);
+ yaz_log(YLOG_DEBUG, "maptab %s num=%d close %d errors", name, num, errors);
if (t_utf8 != 0)
yaz_iconv_close(t_utf8);
if (t_unicode != 0)
void chrmaptab_destroy(chrmaptab tab)
{
if (tab)
- nmem_destroy (tab->nmem);
+ nmem_destroy(tab->nmem);
}
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * c-file-style: "Stroustrup"
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+