-/* $Id: charmap.c,v 1.27 2003-01-13 10:53:16 oleg Exp $
- Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002
- Index Data Aps
+/* $Id: charmap.c,v 1.44 2007-01-15 15:10:26 adam Exp $
+ Copyright (C) 1995-2007
+ Index Data ApS
This file is part of the Zebra server.
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
+
*/
#include <ctype.h>
#include <string.h>
+#include <stdlib.h>
#include <assert.h>
typedef unsigned ucs4_t;
-#include <yaz/yaz-util.h>
#include <charmap.h>
+#include <yaz/yaz-util.h>
#define CHR_MAXSTR 1024
#define CHR_MAXEQUIV 32
+const unsigned char CHR_FIELD_BEGIN = '^';
+
const char *CHR_UNKNOWN = "\001";
const char *CHR_SPACE = "\002";
-const char *CHR_BASE = "\003";
+const char *CHR_CUT = "\003";
+const char *CHR_BASE = "\005"; /* CHECK CHR_BASE_CHAR as well */
struct chrmaptab_info
{
} chrwork;
/*
+ * Callback for equivalent stuff
+ */
+typedef struct
+{
+ NMEM nmem;
+ int no_eq;
+ char *eq[CHR_MAXEQUIV];
+} chr_equiv_work;
+/*
* Add an entry to the character map.
*/
static chr_t_entry *set_map_string(chr_t_entry *root, NMEM nmem,
}
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 &&
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 (LOG_WARN, "duplicate entry for charmap from '%s'",
+ yaz_log (YLOG_WARN, "duplicate entry for charmap from '%s'",
from_0);
}
root->target = (unsigned char **)
return t->target ? t : 0;
}
-static chr_t_entry *find_entry_x(chr_t_entry *t, const char **from, int *len)
+static chr_t_entry *find_entry_x(chr_t_entry *t, const char **from, int *len, int first)
{
chr_t_entry *res;
from++;
len++;
}
- if (*len > 0 && t->children && t->children[(unsigned char) **from])
+ if (*len > 0 && t->children)
{
const char *old_from = *from;
int old_len = *len;
+
+ res = 0;
+
+ if (first && t->children[CHR_FIELD_BEGIN])
+ {
+ if ((res = find_entry_x(t->children[CHR_FIELD_BEGIN], from, len, 0)) && res != t->children[CHR_FIELD_BEGIN])
+ return res;
+ else
+ res = 0;
+ /* otherwhise there was no match on beginning of field, move on */
+ }
- (*len)--;
- (*from)++;
- if ((res = find_entry_x(t->children[(unsigned char) *old_from],
- from, len)))
- return res;
- /* no match */
- *len = old_len;
- *from = old_from;
+ if (!res && t->children[(unsigned char) **from])
+ {
+ (*len)--;
+ (*from)++;
+ if ((res = find_entry_x(t->children[(unsigned char) *old_from],
+ from, len, 0)))
+ return res;
+ /* no match */
+ *len = old_len;
+ *from = old_from;
+ }
}
/* no children match. use ourselves, if we have a target */
return t->target ? t : 0;
}
-const char **chr_map_input_x(chrmaptab maptab, const char **from, int *len)
+const char **chr_map_input_x(chrmaptab maptab, const char **from, int *len, int first)
{
chr_t_entry *t = maptab->input;
chr_t_entry *res;
- if (!(res = find_entry_x(t, from, len)))
+ if (!(res = find_entry_x(t, from, len, first)))
abort();
return (const char **) (res->target);
}
-const char **chr_map_input(chrmaptab maptab, const char **from, int len)
+const char **chr_map_input(chrmaptab maptab, const char **from, int len, int first)
{
chr_t_entry *t = maptab->input;
chr_t_entry *res;
len_tmp[0] = len;
len_tmp[1] = -1;
- if (!(res = find_entry_x(t, from, len_tmp)))
+ if (!(res = find_entry_x(t, from, len_tmp, first)))
abort();
return (const char **) (res->target);
}
+const char **chr_map_q_input(chrmaptab maptab,
+ const char **from, int len, int first)
+{
+ 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)))
+ return 0;
+ return (const char **) (res->target);
+}
+
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;
}
-unsigned char zebra_prim(char **s)
+static int zebra_ucs4_strlen(ucs4_t *s)
{
- unsigned char c;
- unsigned int i = 0;
-
- yaz_log (LOG_DEBUG, "prim %.3s", *s);
- if (**s == '\\')
- {
- (*s)++;
- c = **s;
- switch (c)
- {
- case '\\': c = '\\'; (*s)++; break;
- case 'r': c = '\r'; (*s)++; break;
- case 'n': c = '\n'; (*s)++; break;
- case 't': c = '\t'; (*s)++; break;
- case 's': c = ' '; (*s)++; break;
- case 'x': sscanf(*s, "x%2x", &i); c = i; *s += 3; break;
- case '0':
- case '1':
- case '2':
- case '3':
- case '4':
- case '5':
- case '6':
- case '7':
- case '8':
- case '9':
- sscanf(*s, "%3o", &i);
- c = i;
- *s += 3;
- break;
- default:
- (*s)++;
- }
- }
- else
- {
- c = **s;
- ++(*s);
- }
- return c;
+ int i = 0;
+ while (*s++)
+ i++;
+ return i;
}
ucs4_t zebra_prim_w(ucs4_t **s)
ucs4_t i = 0;
char fmtstr[8];
- yaz_log (LOG_DEBUG, "prim %.3s", (char *) *s);
- if (**s == '\\')
+ yaz_log (YLOG_DEBUG, "prim_w %.3s", (char *) *s);
+ if (**s == '\\' && 1[*s])
{
(*s)++;
c = **s;
case 't': c = '\t'; (*s)++; break;
case 's': c = ' '; (*s)++; break;
case 'x':
- fmtstr[0] = (*s)[0];
- fmtstr[1] = (*s)[1];
- fmtstr[2] = (*s)[2];
- fmtstr[3] = 0;
- sscanf(fmtstr, "x%2x", &i);
- c = i;
- *s += 3; break;
+ if (zebra_ucs4_strlen(*s) >= 3)
+ {
+ fmtstr[0] = (*s)[1];
+ fmtstr[1] = (*s)[2];
+ fmtstr[2] = 0;
+ sscanf(fmtstr, "%x", &i);
+ c = i;
+ *s += 3;
+ }
+ break;
case '0':
case '1':
case '2':
case '7':
case '8':
case '9':
- fmtstr[0] = (*s)[0];
- fmtstr[1] = (*s)[1];
- fmtstr[2] = (*s)[2];
- fmtstr[3] = 0;
- sscanf(fmtstr, "%3o", &i);
- c = i;
- *s += 3;
+ if (zebra_ucs4_strlen(*s) >= 3)
+ {
+ fmtstr[0] = (*s)[0];
+ fmtstr[1] = (*s)[1];
+ fmtstr[2] = (*s)[2];
+ fmtstr[3] = 0;
+ sscanf(fmtstr, "%o", &i);
+ c = i;
+ *s += 3;
+ }
break;
+ case 'L':
+ if (zebra_ucs4_strlen(*s) >= 5)
+ {
+ fmtstr[0] = (*s)[1];
+ fmtstr[1] = (*s)[2];
+ fmtstr[2] = (*s)[3];
+ fmtstr[3] = (*s)[4];
+ fmtstr[4] = 0;
+ sscanf(fmtstr, "%x", &i);
+ c = i;
+ *s += 5;
+ }
+ break;
default:
(*s)++;
}
c = **s;
++(*s);
}
- yaz_log (LOG_DEBUG, "out %d", c);
+ yaz_log (YLOG_DEBUG, "out %d", c);
return c;
}
(char*) CHR_SPACE, 0);
}
+/*
+ * Callback function.
+ * Add a space-entry to the value space.
+ */
+static void fun_addcut(const char *s, void *data, int num)
+{
+ chrmaptab tab = (chrmaptab) data;
+ tab->input = set_map_string(tab->input, tab->nmem, s, strlen(s),
+ (char*) CHR_CUT, 0);
+}
+
/*
* Create a string containing the mapped characters provided.
*/
chrwork *arg = (chrwork *) data;
const char **res, *p = s;
- res = chr_map_input(arg->map, &s, strlen(s));
+ res = chr_map_input(arg->map, &s, strlen(s), 0);
if (*res == (char*) CHR_UNKNOWN)
- logf(LOG_WARN, "Map: '%s' has no mapping", p);
+ yaz_log(YLOG_WARN, "Map: '%s' has no mapping", p);
strncat(arg->string, *res, CHR_MAXSTR - strlen(arg->string));
arg->string[CHR_MAXSTR] = '\0';
}
/*
- * Add a map to the string contained in the argument.
+ * Create an unmodified string (scan_string handler).
*/
-static void fun_add_map(const char *s, void *data, int num)
+static void fun_add_equivalent_string(const char *s, void *data, int num)
{
- chrwork *arg = (chrwork *) data;
-
- assert(arg->map->input);
- logf (LOG_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++)
- logf (LOG_DEBUG, " %3d", (unsigned char) *s);
+ 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);
}
/*
- * Add a query map to the string contained in the argument.
+ * Add a map to the string contained in the argument.
*/
-static void fun_add_qmap(const char *s, void *data, int num)
+static void fun_add_map(const char *s, void *data, int num)
{
chrwork *arg = (chrwork *) data;
- assert(arg->map->q_input);
- logf (LOG_DEBUG, "set qmap %.*s", (int) strlen(s), s);
- set_map_string(arg->map->q_input, arg->map->nmem, s,
- strlen(s), arg->string, 0);
+ assert(arg->map->input);
+ 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++)
- logf (LOG_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,
ret = yaz_iconv (t, &inbuf, &inbytesleft, &outbuf, &outbytesleft);
if (ret == (size_t) (-1))
{
- yaz_log (LOG_WARN|LOG_ERRNO, "bad unicode sequence");
+ 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");
return -1;
}
}
char str[1024];
ucs4_t arg[512];
+ ucs4_t arg_prim[512];
ucs4_t *s0, *s = arg;
ucs4_t c, begin, end;
size_t i;
begin = zebra_prim_w(&s);
if (*s != '-')
{
- logf(LOG_FATAL, "Bad range in char-map");
+ yaz_log(YLOG_FATAL, "Bad range in char-map");
return -1;
}
s++;
end = zebra_prim_w(&s);
if (end <= begin)
{
- logf(LOG_FATAL, "Bad range in char-map");
+ yaz_log(YLOG_FATAL, "Bad range in char-map");
return -1;
}
s++;
(*fun)(str, data, num ? (*num)++ : 0);
}
break;
- case '[': s++; abort(); break;
case '(':
++s;
- s0 = s;
- while (*s != ')' || s[-1] == '\\')
- s++;
- *s = 0;
- if (scan_to_utf8 (t_utf8, s0, s - s0, str, sizeof(str)-1))
+ s0 = s; i = 0;
+ while (*s != ')' || s[-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))
return -1;
(*fun)(str, data, num ? (*num)++ : 0);
s++;
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];
unsigned endian = 31;
const char *ucs4_native = "UCS-4";
+ 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;
+ }
+
if (*(char*) &endian == 31) /* little endian? */
ucs4_native = "UCS-4LE";
t_utf8 = yaz_iconv_open ("UTF-8", ucs4_native);
- logf (LOG_DEBUG, "maptab %s open", name);
- if (!(f = yaz_fopen(tabpath, name, "r", tabroot)))
- {
- logf(LOG_WARN|LOG_ERRNO, "%s", name);
- return 0;
- }
nmem = nmem_create ();
res = (chrmaptab) nmem_malloc(nmem, sizeof(*res));
res->nmem = nmem;
res->input->children[i]->target = (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"))
+ if (!yaz_matchstr(argv[0], "lowercase"))
{
if (argc != 2)
{
- logf(LOG_FATAL, "Syntax error in charmap");
+ yaz_log(YLOG_FATAL, "Syntax error in charmap");
++errors;
}
if (scan_string(argv[1], t_unicode, t_utf8, fun_addentry,
res, &num) < 0)
{
- logf(LOG_FATAL, "Bad value-set specification");
+ yaz_log(YLOG_FATAL, "Bad value-set specification");
++errors;
}
res->base_uppercase = num;
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)
{
- logf(LOG_FATAL, "Uppercase directive with no lowercase set");
+ yaz_log(YLOG_FATAL, "Uppercase directive with no lowercase set");
++errors;
}
if (argc != 2)
{
- logf(LOG_FATAL, "Missing arg for uppercase directive");
+ yaz_log(YLOG_FATAL, "Missing arg for uppercase directive");
++errors;
}
if (scan_string(argv[1], t_unicode, t_utf8, fun_addentry,
res, &num) < 0)
{
- logf(LOG_FATAL, "Bad value-set specification");
+ yaz_log(YLOG_FATAL, "Bad value-set specification");
++errors;
}
}
- else if (!map_only && !yaz_matchstr(argv[0], "space"))
+ else if (!yaz_matchstr(argv[0], "space"))
{
if (argc != 2)
{
- logf(LOG_FATAL, "Syntax error in charmap");
+ yaz_log(YLOG_FATAL, "Syntax error in charmap for space");
++errors;
}
if (scan_string(argv[1], t_unicode, t_utf8,
fun_addspace, res, 0) < 0)
{
- logf(LOG_FATAL, "Bad space specification");
+ yaz_log(YLOG_FATAL, "Bad space specification");
+ ++errors;
+ }
+ }
+ else if (!yaz_matchstr(argv[0], "cut"))
+ {
+ if (argc != 2)
+ {
+ yaz_log(YLOG_FATAL, "Syntax error in charmap for cut");
+ ++errors;
+ }
+ if (scan_string(argv[1], t_unicode, t_utf8,
+ fun_addcut, res, 0) < 0)
+ {
+ yaz_log(YLOG_FATAL, "Bad cut specification");
++errors;
}
}
if (argc != 3)
{
- logf(LOG_FATAL, "charmap directive map requires 2 args");
+ yaz_log(YLOG_FATAL, "charmap directive map requires 2 args");
++errors;
}
buf.map = res;
if (scan_string(argv[2], t_unicode, t_utf8,
fun_mkstring, &buf, 0) < 0)
{
- logf(LOG_FATAL, "Bad map target");
+ yaz_log(YLOG_FATAL, "Bad map target");
++errors;
}
if (scan_string(argv[1], t_unicode, t_utf8,
fun_add_map, &buf, 0) < 0)
{
- logf(LOG_FATAL, "Bad map source");
+ yaz_log(YLOG_FATAL, "Bad map source");
++errors;
}
}
- else if (!yaz_matchstr(argv[0], "qmap"))
+ else if (!yaz_matchstr(argv[0], "equivalent"))
{
- chrwork buf;
+ chr_equiv_work w;
- if (argc != 3)
+ if (argc != 2)
{
- logf(LOG_FATAL, "charmap directive qmap requires 2 args");
+ yaz_log(YLOG_FATAL, "equivalent requires 1 argument");
++errors;
}
- buf.map = res;
- buf.string[0] = '\0';
- if (scan_string(argv[2], t_unicode, t_utf8,
- fun_mkstring, &buf, 0) < 0)
+ w.nmem = res->nmem;
+ w.no_eq = 0;
+ if (scan_string(argv[1], t_unicode, t_utf8,
+ fun_add_equivalent_string, &w, 0) < 0)
{
- logf(LOG_FATAL, "Bad qmap target");
+ yaz_log(YLOG_FATAL, "equivalent: invalid string");
++errors;
}
- if (scan_string(argv[1], t_unicode, t_utf8,
- fun_add_qmap, &buf, 0) < 0)
+ else if (w.no_eq == 0)
{
- logf(LOG_FATAL, "Bad qmap source");
+ yaz_log(YLOG_FATAL, "equivalent: no strings");
++errors;
}
+ else
+ {
+ char *result_str;
+ int i, slen = 5;
+
+ /* determine length of regular expression */
+ for (i = 0; i<w.no_eq; i++)
+ slen += strlen(w.eq[i]) + 1;
+ result_str = nmem_malloc(res->nmem, slen + 5);
+
+ /* build the regular expression */
+ *result_str = '\0';
+ slen = 0;
+ for (i = 0; i<w.no_eq; i++)
+ {
+ result_str[slen++] = i ? '|' : '(';
+ strcpy(result_str + slen, w.eq[i]);
+ slen += strlen(w.eq[i]);
+ }
+ result_str[slen++] = ')';
+ result_str[slen] = '\0';
+
+ /* each eq will map to this regular expression */
+ for (i = 0; i<w.no_eq; i++)
+ {
+ set_map_string(res->q_input, res->nmem,
+ w.eq[i], strlen(w.eq[i]),
+ result_str, 0);
+ }
+ }
}
else if (!yaz_matchstr(argv[0], "encoding"))
{
* 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
}
else
{
- logf(LOG_WARN, "Syntax error at '%s' in %s", line, name);
+ yaz_log(YLOG_WARN, "Syntax error at '%s' in %s", line, name);
}
yaz_fclose(f);
chrmaptab_destroy(res);
res = 0;
}
- logf (LOG_DEBUG, "maptab %s close %d errors", name, errors);
+ yaz_log (YLOG_DEBUG, "maptab %s close %d errors", name, errors);
if (t_utf8 != 0)
yaz_iconv_close(t_utf8);
if (t_unicode != 0)
}
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+