-/*
- * Copyright (C) 1994-1999, Index Data
- * All rights reserved.
- * Sebastian Hammer, Adam Dickmeiss
- *
- * $Log: zebramap.c,v $
- * Revision 1.18 1999-10-15 08:27:46 adam
- * Fixed replace handler. 8-bit fix.
- *
- * Revision 1.17 1999/09/08 12:13:21 adam
- * Fixed minor bug "replace"-mappings. Removed some logging messages.
- *
- * Revision 1.16 1999/09/07 07:19:21 adam
- * Work on character mapping. Implemented replace rules.
- *
- * Revision 1.15 1999/05/26 07:49:14 adam
- * C++ compilation.
- *
- * Revision 1.14 1999/02/19 10:37:40 adam
- * Minor fix.
- *
- * Revision 1.13 1999/02/18 15:01:04 adam
- * Structure=key uses register type 0.
- *
- * Revision 1.12 1999/02/12 13:29:25 adam
- * Implemented position-flag for registers.
- *
- * Revision 1.11 1998/10/13 20:09:19 adam
- * Changed call to readconf_line.
- *
- * Revision 1.10 1998/06/23 15:33:37 adam
- * Added feature to specify sort criteria in query (type 7 specifies
- * sort flags).
- *
- * Revision 1.9 1998/04/02 14:35:30 adam
- * First version of Zebra that works with compiled ASN.1.
- *
- * Revision 1.8 1998/03/05 08:42:44 adam
- * Minor changes to zebramap data structures. Query mapping rules changed.
- *
- * Revision 1.7 1998/02/10 12:03:07 adam
- * Implemented Sort.
- *
- * Revision 1.6 1998/01/29 13:36:01 adam
- * Structure word-list, free-form-text and document-text all
- * trigger ranked search.
- *
- * Revision 1.5 1997/11/19 10:22:14 adam
- * Bug fix (introduced by previous commit).
- *
- * Revision 1.4 1997/11/18 10:05:08 adam
- * Changed character map facility so that admin can specify character
- * mapping files for each register type, w, p, etc.
- *
- * Revision 1.3 1997/11/17 15:35:26 adam
- * Bug fix. Relation=relevance wasn't observed.
- *
- * Revision 1.2 1997/10/31 12:39:30 adam
- * Changed log message.
- *
- * Revision 1.1 1997/10/27 14:33:06 adam
- * Moved towards generic character mapping depending on "structure"
- * field in abstract syntax file. Fixed a few memory leaks. Fixed
- * bug with negative integers when doing searches with relational
- * operators.
- *
- */
+/* This file is part of the Zebra server.
+ Copyright (C) 1994-2011 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
+Software Foundation; either version 2, or (at your option) any later
+version.
+
+Zebra is distributed in the hope that it will be useful, but WITHOUT ANY
+WARRANTY; without even the implied warranty of MERCHANTABILITY or
+FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License
+for more details.
+You should have received a copy of the GNU General Public License
+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 <assert.h>
+#include <stdlib.h>
#include <ctype.h>
-#include <yaz-util.h>
#include <charmap.h>
+#include <attrfind.h>
+#include <yaz/yaz-util.h>
+
+#if YAZ_HAVE_ICU
+#include <yaz/icu.h>
+#endif
#include <zebramap.h>
#define ZEBRA_MAP_TYPE_SORT 1
#define ZEBRA_MAP_TYPE_INDEX 2
+#define ZEBRA_MAP_TYPE_STATICRANK 3
#define ZEBRA_REPLACE_ANY 300
-struct zm_token {
- int *token_from;
- char *token_to;
- int token_min;
- struct zm_token *next;
-};
-
struct zebra_map {
- unsigned reg_id;
+ const char *id;
int completeness;
int positioned;
+ int alwaysmatches;
+ int first_in_field;
int type;
+ int use_chain;
+ int debug;
union {
struct {
- int dummy;
- } index;
- struct {
int entry_size;
} sort;
} u;
chrmaptab maptab;
const char *maptab_name;
+ zebra_maps_t zebra_maps;
+#if YAZ_HAVE_XML2
+ xmlDocPtr doc;
+#endif
+#if YAZ_HAVE_ICU
+ struct icu_chain *icu_chain;
+#endif
+ WRBUF input_str;
+ WRBUF print_str;
+ size_t simple_off;
struct zebra_map *next;
- struct zm_token *replace_tokens;
};
-struct zebra_maps {
+struct zebra_maps_s {
char *tabpath;
+ char *tabroot;
NMEM nmem;
- struct zebra_map *map_list;
char temp_map_str[2];
const char *temp_map_ptr[2];
- struct zebra_map **lookup_array;
- WRBUF wrbuf_1, wrbuf_2;
+ WRBUF wrbuf_1;
+ int no_files_read;
+ zebra_map_t map_list;
+ zebra_map_t last_map;
};
-void zebra_maps_close (ZebraMaps zms)
+void zebra_maps_close(zebra_maps_t zms)
{
struct zebra_map *zm = zms->map_list;
while (zm)
{
if (zm->maptab)
- chrmaptab_destroy (zm->maptab);
+ chrmaptab_destroy(zm->maptab);
+#if YAZ_HAVE_ICU
+ if (zm->icu_chain)
+ icu_chain_destroy(zm->icu_chain);
+#endif
+#if YAZ_HAVE_XML2
+ xmlFreeDoc(zm->doc);
+#endif
+ wrbuf_destroy(zm->input_str);
+ wrbuf_destroy(zm->print_str);
zm = zm->next;
}
- wrbuf_free (zms->wrbuf_1, 1);
- wrbuf_free (zms->wrbuf_2, 1);
- nmem_destroy (zms->nmem);
- xfree (zms);
+ wrbuf_destroy(zms->wrbuf_1);
+ nmem_destroy(zms->nmem);
+ xfree(zms);
}
-static void zebra_map_read (ZebraMaps zms, const char *name)
+zebra_map_t zebra_add_map(zebra_maps_t zms, const char *index_type,
+ int map_type)
{
- FILE *f;
- char line[512];
- char *argv[10];
- int argc;
- int lineno = 0;
- struct zebra_map **zm = 0, *zp;
+ zebra_map_t zm = (zebra_map_t) nmem_malloc(zms->nmem, sizeof(*zm));
+
+ zm->zebra_maps = zms;
+ zm->id = nmem_strdup(zms->nmem, index_type);
+ zm->maptab_name = 0;
+ zm->use_chain = 0;
+ zm->debug = 0;
+ zm->maptab = 0;
+ zm->type = map_type;
+ zm->completeness = 0;
+ zm->positioned = 0;
+ zm->alwaysmatches = 0;
+ zm->first_in_field = 0;
+
+ if (zms->last_map)
+ zms->last_map->next = zm;
+ else
+ zms->map_list = zm;
+ zms->last_map = zm;
+ zm->next = 0;
+#if YAZ_HAVE_ICU
+ zm->icu_chain = 0;
+#endif
+#if YAZ_HAVE_XML2
+ zm->doc = 0;
+#endif
+ zm->input_str = wrbuf_alloc();
+ zm->print_str = wrbuf_alloc();
+ return zm;
+}
- if (!(f = yaz_path_fopen(zms->tabpath, name, "r")))
+static int parse_command(zebra_maps_t zms, int argc, char **argv,
+ const char *fname, int lineno)
+{
+ zebra_map_t zm = zms->last_map;
+ if (argc == 1)
{
- logf(LOG_WARN|LOG_ERRNO, "%s", name);
- return ;
+ yaz_log(YLOG_WARN, "%s:%d: Missing arguments for '%s'",
+ fname, lineno, argv[0]);
+ return -1;
}
- while ((argc = readconf_line(f, &lineno, line, 512, argv, 10)))
+ if (argc > 2)
{
- if (!yaz_matchstr (argv[0], "index") && argc == 2)
- {
- if (!zm)
- zm = &zms->map_list;
- else
- zm = &(*zm)->next;
- *zm = (struct zebra_map *) nmem_malloc (zms->nmem, sizeof(**zm));
- (*zm)->reg_id = argv[1][0];
- (*zm)->maptab_name = NULL;
- (*zm)->maptab = NULL;
- (*zm)->type = ZEBRA_MAP_TYPE_INDEX;
- (*zm)->completeness = 0;
- (*zm)->positioned = 1;
- (*zm)->replace_tokens = 0;
- }
- else if (!yaz_matchstr (argv[0], "sort") && argc == 2)
- {
- if (!zm)
- zm = &zms->map_list;
- else
- zm = &(*zm)->next;
- *zm = (struct zebra_map *) nmem_malloc (zms->nmem, sizeof(**zm));
- (*zm)->reg_id = argv[1][0];
- (*zm)->maptab_name = NULL;
- (*zm)->type = ZEBRA_MAP_TYPE_SORT;
- (*zm)->u.sort.entry_size = 80;
- (*zm)->maptab = NULL;
- (*zm)->completeness = 0;
- (*zm)->positioned = 0;
- (*zm)->replace_tokens = 0;
- }
- else if (zm && !yaz_matchstr (argv[0], "charmap") && argc == 2)
- {
- (*zm)->maptab_name = nmem_strdup (zms->nmem, argv[1]);
- }
- else if (zm && !yaz_matchstr (argv[0], "completeness") && argc == 2)
- {
- (*zm)->completeness = atoi (argv[1]);
- }
- else if (zm && !yaz_matchstr (argv[0], "position") && argc == 2)
- {
- (*zm)->positioned = atoi (argv[1]);
- }
- else if (zm && !yaz_matchstr (argv[0], "entrysize") && argc == 2)
+ yaz_log(YLOG_WARN, "%s:%d: Too many arguments for '%s'",
+ fname, lineno, argv[0]);
+ return -1;
+ }
+ if (!yaz_matchstr(argv[0], "index"))
+ {
+ zm = zebra_add_map(zms, argv[1], ZEBRA_MAP_TYPE_INDEX);
+ zm->positioned = 1;
+ }
+ else if (!yaz_matchstr(argv[0], "sort"))
+ {
+ zm = zebra_add_map(zms, argv[1], ZEBRA_MAP_TYPE_SORT);
+ zm->u.sort.entry_size = 80;
+ }
+ else if (!yaz_matchstr(argv[0], "staticrank"))
+ {
+ zm = zebra_add_map(zms, argv[1], ZEBRA_MAP_TYPE_STATICRANK);
+ zm->completeness = 1;
+ }
+ else if (!zm)
+ {
+ yaz_log(YLOG_WARN, "%s:%d: Missing sort/index before '%s'",
+ fname, lineno, argv[0]);
+ return -1;
+ }
+ else if (!yaz_matchstr(argv[0], "charmap") && argc == 2)
+ {
+ if (zm->type != ZEBRA_MAP_TYPE_STATICRANK)
+ zm->maptab_name = nmem_strdup(zms->nmem, argv[1]);
+ else
+ {
+ yaz_log(YLOG_WARN|YLOG_FATAL, "%s:%d: charmap for "
+ "staticrank is invalid", fname, lineno);
+ yaz_log(YLOG_LOG, "Type is %d", zm->type);
+ return -1;
+ }
+ }
+ else if (!yaz_matchstr(argv[0], "completeness") && argc == 2)
+ {
+ zm->completeness = atoi(argv[1]);
+ }
+ else if (!yaz_matchstr(argv[0], "position") && argc == 2)
+ {
+ zm->positioned = atoi(argv[1]);
+ }
+ else if (!yaz_matchstr(argv[0], "alwaysmatches") && argc == 2)
+ {
+ if (zm->type != ZEBRA_MAP_TYPE_STATICRANK)
+ zm->alwaysmatches = atoi(argv[1]);
+ else
{
- if ((*zm)->type == ZEBRA_MAP_TYPE_SORT)
- (*zm)->u.sort.entry_size = atoi (argv[1]);
+ yaz_log(YLOG_WARN|YLOG_FATAL, "%s:%d: alwaysmatches for "
+ "staticrank is invalid", fname, lineno);
+ return -1;
}
- else if (zm && !yaz_matchstr (argv[0], "replace") && argc >= 2)
+ }
+ else if (!yaz_matchstr(argv[0], "firstinfield") && argc == 2)
+ {
+ zm->first_in_field = atoi(argv[1]);
+ }
+ else if (!yaz_matchstr(argv[0], "entrysize") && argc == 2)
+ {
+ if (zm->type == ZEBRA_MAP_TYPE_SORT)
+ zm->u.sort.entry_size = atoi(argv[1]);
+ else
{
- struct zm_token *token = nmem_malloc (zms->nmem, sizeof(*token));
- token->next = (*zm)->replace_tokens;
- (*zm)->replace_tokens = token;
-#if 0
- logf (LOG_LOG, "replace %s", argv[1]);
+ yaz_log(YLOG_WARN,
+ "%s:%d: entrysize only valid in sort section",
+ fname, lineno);
+ return -1;
+ }
+ }
+ else if (!yaz_matchstr(argv[0], "simplechain"))
+ {
+ zm->use_chain = 1;
+#if YAZ_HAVE_ICU
+ zm->icu_chain = 0;
#endif
- token->token_from = 0;
- if (argc >= 2)
+ }
+ else if (!yaz_matchstr(argv[0], "icuchain"))
+ {
+ char full_path[1024];
+ if (!yaz_filepath_resolve(argv[1], zms->tabpath, zms->tabroot,
+ full_path))
+ {
+ yaz_log(YLOG_WARN, "%s:%d: Could not locate icuchain config '%s'",
+ fname, lineno, argv[1]);
+ return -1;
+ }
+#if YAZ_HAVE_XML2
+ zm->doc = xmlParseFile(full_path);
+ if (!zm->doc)
+ {
+ yaz_log(YLOG_WARN, "%s:%d: Could not load icuchain config '%s'",
+ fname, lineno, argv[1]);
+ return -1;
+ }
+ else
+ {
+#if YAZ_HAVE_ICU
+ UErrorCode status;
+ xmlNode *xml_node = xmlDocGetRootElement(zm->doc);
+ zm->icu_chain =
+ icu_chain_xml_config(xml_node,
+/* not sure about sort for this function yet.. */
+#if 1
+ 1,
+#else
+ zm->type == ZEBRA_MAP_TYPE_SORT,
+#endif
+ &status);
+ if (!zm->icu_chain)
{
- char *cp = argv[1];
- int *dp = token->token_from = (int *)
- nmem_malloc (zms->nmem, (1+strlen(cp))*sizeof(int));
- while (*cp)
- if (*cp == '$')
- {
- *dp++ = ' ';
- cp++;
- }
- else if (*cp == '.')
- {
- *dp++ = ZEBRA_REPLACE_ANY;
- cp++;
- }
- else
- {
- *dp++ = zebra_prim(&cp);
-#if 0
- logf (LOG_LOG, " char %2X %c", dp[-1], dp[-1]);
+ yaz_log(YLOG_WARN, "%s:%d: Failed to load ICU chain %s",
+ fname, lineno, argv[1]);
+ }
+ zm->use_chain = 1;
+#else
+ yaz_log(YLOG_WARN, "%s:%d: ICU support unavailable",
+ fname, lineno);
+ return -1;
#endif
- }
- *dp = '\0';
- }
- if (argc >= 3)
- {
- char *cp = argv[2];
- char *dp = token->token_to =
- nmem_malloc (zms->nmem, strlen(cp)+1);
- while (*cp)
- if (*cp == '$')
- {
- *dp++ = ' ';
- cp++;
- }
- else
- *dp++ = zebra_prim(&cp);
- *dp = '\0';
- }
- else
- token->token_to = 0;
}
+#else
+ yaz_log(YLOG_WARN, "%s:%d: XML support unavailable",
+ fname, lineno);
+ return -1;
+#endif
}
- if (zm)
- (*zm)->next = NULL;
- fclose (f);
-
- for (zp = zms->map_list; zp; zp = zp->next)
- zms->lookup_array[zp->reg_id] = zp;
+ else if (!yaz_matchstr(argv[0], "debug") && argc == 2)
+ {
+ zm->debug = atoi(argv[1]);
+ }
+ else
+ {
+ yaz_log(YLOG_WARN, "%s:%d: Unrecognized directive '%s'",
+ fname, lineno, argv[0]);
+ return -1;
+ }
+ return 0;
}
-static void zms_map_handle (void *p, const char *name, const char *value)
+ZEBRA_RES zebra_maps_read_file(zebra_maps_t zms, const char *fname)
{
- ZebraMaps zms = (ZebraMaps) p;
-
- zebra_map_read (zms, value);
+ FILE *f;
+ char line[512];
+ char *argv[10];
+ int argc;
+ int lineno = 0;
+ int failures = 0;
+
+ if (!(f = yaz_fopen(zms->tabpath, fname, "r", zms->tabroot)))
+ {
+ yaz_log(YLOG_ERRNO|YLOG_FATAL, "%s", fname);
+ return ZEBRA_FAIL;
+ }
+ while ((argc = readconf_line(f, &lineno, line, 512, argv, 10)))
+ {
+ int r = parse_command(zms, argc, argv, fname, lineno);
+ if (r)
+ failures++;
+ }
+ yaz_fclose(f);
+
+ if (failures)
+ return ZEBRA_FAIL;
+
+ (zms->no_files_read)++;
+ return ZEBRA_OK;
}
-ZebraMaps zebra_maps_open (Res res)
+zebra_maps_t zebra_maps_open(Res res, const char *base_path,
+ const char *profile_path)
{
- ZebraMaps zms = (ZebraMaps) xmalloc (sizeof(*zms));
- int i;
+ zebra_maps_t zms = (zebra_maps_t) xmalloc(sizeof(*zms));
- zms->nmem = nmem_create ();
- zms->tabpath = nmem_strdup (zms->nmem,
- res_get_def (res, "profilePath", "."));
- zms->map_list = NULL;
+ zms->nmem = nmem_create();
+ zms->tabpath = profile_path ? nmem_strdup(zms->nmem, profile_path) : 0;
+ zms->tabroot = 0;
+ if (base_path)
+ zms->tabroot = nmem_strdup(zms->nmem, base_path);
+ zms->map_list = 0;
+ zms->last_map = 0;
zms->temp_map_str[0] = '\0';
zms->temp_map_str[1] = '\0';
zms->temp_map_ptr[0] = zms->temp_map_str;
zms->temp_map_ptr[1] = NULL;
- zms->lookup_array = (struct zebra_map**)
- nmem_malloc (zms->nmem, sizeof(*zms->lookup_array)*256);
- for (i = 0; i<256; i++)
- zms->lookup_array[i] = 0;
- if (!res || !res_trav (res, "index", zms, zms_map_handle))
- zebra_map_read (zms, "default.idx");
-
zms->wrbuf_1 = wrbuf_alloc();
- zms->wrbuf_2 = wrbuf_alloc();
+
+ zms->no_files_read = 0;
return zms;
}
-struct zebra_map *zebra_map_get (ZebraMaps zms, unsigned reg_id)
+void zebra_maps_define_default_sort(zebra_maps_t zms)
{
- return zms->lookup_array[reg_id];
+ zebra_map_t zm = zebra_add_map(zms, "s", ZEBRA_MAP_TYPE_SORT);
+ zm->u.sort.entry_size = 80;
}
-chrmaptab zebra_charmap_get (ZebraMaps zms, unsigned reg_id)
+zebra_map_t zebra_map_get(zebra_maps_t zms, const char *id)
{
- struct zebra_map *zm = zebra_map_get (zms, reg_id);
+ zebra_map_t zm;
+ for (zm = zms->map_list; zm; zm = zm->next)
+ if (!strcmp(zm->id, id))
+ break;
+ return zm;
+}
+
+zebra_map_t zebra_map_get_or_add(zebra_maps_t zms, const char *id)
+{
+ struct zebra_map *zm = zebra_map_get(zms, id);
if (!zm)
{
- zm = (struct zebra_map *) nmem_malloc (zms->nmem, sizeof(*zm));
- logf (LOG_WARN, "Unknown register type: %c", reg_id);
+ zm = zebra_add_map(zms, id, ZEBRA_MAP_TYPE_INDEX);
+
+ /* no reason to warn if no maps are read from file */
+ if (zms->no_files_read)
+ yaz_log(YLOG_WARN, "Unknown register type: %s", id);
- zm->reg_id = reg_id;
- zm->maptab_name = NULL;
- zm->maptab = NULL;
- zm->type = ZEBRA_MAP_TYPE_INDEX;
+ zm->maptab_name = nmem_strdup(zms->nmem, "@");
zm->completeness = 0;
- zm->next = zms->map_list;
- zms->map_list = zm->next;
-
- zms->lookup_array[zm->reg_id & 255] = zm;
+ zm->positioned = 1;
}
+ return zm;
+}
+
+chrmaptab zebra_charmap_get(zebra_map_t zm)
+{
if (!zm->maptab)
{
- if (!zm->maptab_name || !yaz_matchstr (zm->maptab_name, "@"))
+ if (!zm->maptab_name || !yaz_matchstr(zm->maptab_name, "@"))
return NULL;
- if (!(zm->maptab = chrmaptab_create (zms->tabpath,
- zm->maptab_name, 0)))
- logf(LOG_WARN, "Failed to read character table %s",
- zm->maptab_name);
+ if (!(zm->maptab = chrmaptab_create(zm->zebra_maps->tabpath,
+ zm->maptab_name,
+ zm->zebra_maps->tabroot)))
+ yaz_log(YLOG_WARN, "Failed to read character table %s",
+ zm->maptab_name);
else
- logf(LOG_DEBUG, "Read character table %s", zm->maptab_name);
+ yaz_log(YLOG_DEBUG, "Read character table %s", zm->maptab_name);
}
return zm->maptab;
}
-const char **zebra_maps_input (ZebraMaps zms, unsigned reg_id,
- const char **from, int len)
+const char **zebra_maps_input(zebra_map_t zm,
+ const char **from, int len, int first)
{
- chrmaptab maptab;
-
- maptab = zebra_charmap_get (zms, reg_id);
+ chrmaptab maptab = zebra_charmap_get(zm);
if (maptab)
- return chr_map_input(maptab, from, len);
+ return chr_map_input(maptab, from, len, first);
- zms->temp_map_str[0] = **from;
+ zm->zebra_maps->temp_map_str[0] = **from;
(*from)++;
- return zms->temp_map_ptr;
+ return zm->zebra_maps->temp_map_ptr;
}
-#if 0
-int zebra_maps_input_tokens (ZebraMaps zms, unsigned reg_id,
- const char *input_str, int input_len,
- WRBUF wrbuf)
+const char **zebra_maps_search(zebra_map_t zm,
+ const char **from, int len, int *q_map_match)
{
- chrmaptab maptab = zebra_charmap_get (zms, reg_id);
- int len[4];
- char *str[3];
- int input_i = 0;
- int first = 1;
- const char **out;
-
- if (!maptab)
- {
- wrbuf_write (wrbuf, input_str, input_len);
- return -1;
- }
- str[0] = " ";
- len[0] = 1;
- str[1] = input_str;
- len[1] = input_len;
- str[2] = " ";
- len[2] = 1;
- len[3] = -1;
+ chrmaptab maptab;
- out = chr_map_input (maptab, str, len);
- while (len[1] > 0)
+ *q_map_match = 0;
+ maptab = zebra_charmap_get(zm);
+ if (maptab)
{
- while (out && *out && **out == *CHR_SPACE)
- out = chr_map_input (maptab, str, len);
+ const char **map;
+ map = chr_map_q_input(maptab, from, len, 0);
+ if (map && map[0])
+ {
+ *q_map_match = 1;
+ return map;
+ }
+ map = chr_map_input(maptab, from, len, 0);
+ if (map)
+ return map;
}
+ zm->zebra_maps->temp_map_str[0] = **from;
+
+ (*from)++;
+ return zm->zebra_maps->temp_map_ptr;
}
-#endif
-const char *zebra_maps_output(ZebraMaps zms, unsigned reg_id,
+const char *zebra_maps_output(zebra_map_t zm,
const char **from)
{
- chrmaptab maptab = zebra_charmap_get (zms, reg_id);
+ chrmaptab maptab = zebra_charmap_get(zm);
if (!maptab)
return 0;
- return chr_map_output (maptab, from, 1);
+ return chr_map_output(maptab, from, 1);
}
/* ------------------------------------ */
-typedef struct {
- int type;
- int major;
- int minor;
- Z_AttributeElement **attributeList;
- int num_attributes;
-} AttrType;
+int zebra_maps_is_complete(zebra_map_t zm)
+{
+ if (zm)
+ return zm->completeness;
+ return 0;
+}
-static int attr_find (AttrType *src, oid_value *attributeSetP)
+int zebra_maps_is_positioned(zebra_map_t zm)
{
- while (src->major < src->num_attributes)
- {
- Z_AttributeElement *element;
-
- element = src->attributeList[src->major];
- if (src->type == *element->attributeType)
- {
- switch (element->which)
- {
- case Z_AttributeValue_numeric:
- ++(src->major);
- if (element->attributeSet && attributeSetP)
- {
- oident *attrset;
-
- attrset = oid_getentbyoid (element->attributeSet);
- *attributeSetP = attrset->value;
- }
- return *element->value.numeric;
- break;
- case Z_AttributeValue_complex:
- if (src->minor >= element->value.complex->num_list ||
- element->value.complex->list[src->minor]->which !=
- Z_StringOrNumeric_numeric)
- break;
- ++(src->minor);
- if (element->attributeSet && attributeSetP)
- {
- oident *attrset;
-
- attrset = oid_getentbyoid (element->attributeSet);
- *attributeSetP = attrset->value;
- }
- return *element->value.complex->list[src->minor-1]->u.numeric;
- default:
- assert (0);
- }
- }
- ++(src->major);
- }
- return -1;
+ if (zm)
+ return zm->positioned;
+ return 0;
}
-static void attr_init_APT (AttrType *src, Z_AttributesPlusTerm *zapt, int type)
+int zebra_maps_is_index(zebra_map_t zm)
{
-#ifdef ASN_COMPILED
- src->attributeList = zapt->attributes->attributes;
- src->num_attributes = zapt->attributes->num_attributes;
-#else
- src->attributeList = zapt->attributeList;
- src->num_attributes = zapt->num_attributes;
-#endif
- src->type = type;
- src->major = 0;
- src->minor = 0;
+ if (zm)
+ return zm->type == ZEBRA_MAP_TYPE_INDEX;
+ return 0;
}
-static void attr_init_AttrList (AttrType *src, Z_AttributeList *list, int type)
+int zebra_maps_is_staticrank(zebra_map_t zm)
{
- src->attributeList = list->attributes;
- src->num_attributes = list->num_attributes;
- src->type = type;
- src->major = 0;
- src->minor = 0;
+ if (zm)
+ return zm->type == ZEBRA_MAP_TYPE_STATICRANK;
+ return 0;
}
-
-/* ------------------------------------ */
-
-int zebra_maps_is_complete (ZebraMaps zms, unsigned reg_id)
-{
- struct zebra_map *zm = zebra_map_get (zms, reg_id);
+
+int zebra_maps_is_sort(zebra_map_t zm)
+{
if (zm)
- return zm->completeness;
+ return zm->type == ZEBRA_MAP_TYPE_SORT;
return 0;
}
-int zebra_maps_is_positioned (ZebraMaps zms, unsigned reg_id)
+int zebra_maps_is_alwaysmatches(zebra_map_t zm)
{
- struct zebra_map *zm = zebra_map_get (zms, reg_id);
if (zm)
- return zm->positioned;
+ return zm->alwaysmatches;
return 0;
}
-
-int zebra_maps_is_sort (ZebraMaps zms, unsigned reg_id)
+
+int zebra_maps_is_first_in_field(zebra_map_t zm)
{
- struct zebra_map *zm = zebra_map_get (zms, reg_id);
if (zm)
- return zm->type == ZEBRA_MAP_TYPE_SORT;
+ return zm->first_in_field;
return 0;
}
-int zebra_maps_sort (ZebraMaps zms, Z_SortAttributes *sortAttributes)
+int zebra_maps_sort(zebra_maps_t zms, Z_SortAttributes *sortAttributes,
+ int *numerical)
{
AttrType use;
- attr_init_AttrList (&use, sortAttributes->list, 1);
-
- return attr_find (&use, NULL);
+ AttrType structure;
+ int structure_value;
+ attr_init_AttrList(&use, sortAttributes->list, 1);
+ attr_init_AttrList(&structure, sortAttributes->list, 4);
+
+ *numerical = 0;
+ structure_value = attr_find(&structure, 0);
+ if (structure_value == 109)
+ *numerical = 1;
+ return attr_find(&use, NULL);
}
-int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt,
- unsigned *reg_id, char **search_type, char **rank_type,
- int *complete_flag, int *sort_flag)
+int zebra_maps_attr(zebra_maps_t zms, Z_AttributesPlusTerm *zapt,
+ const char **index_type, char **search_type, char *rank_type,
+ int *complete_flag, int *sort_flag)
{
AttrType completeness;
AttrType structure;
AttrType relation;
AttrType sort_relation;
+ AttrType weight;
+ AttrType use;
int completeness_value;
int structure_value;
+ const char *structure_str = 0;
int relation_value;
int sort_relation_value;
-
- attr_init_APT (&structure, zapt, 4);
- attr_init_APT (&completeness, zapt, 6);
- attr_init_APT (&relation, zapt, 2);
- attr_init_APT (&sort_relation, zapt, 7);
-
- completeness_value = attr_find (&completeness, NULL);
- structure_value = attr_find (&structure, NULL);
- relation_value = attr_find (&relation, NULL);
- sort_relation_value = attr_find (&sort_relation, NULL);
+ int weight_value;
+ int use_value;
+
+ attr_init_APT(&structure, zapt, 4);
+ attr_init_APT(&completeness, zapt, 6);
+ attr_init_APT(&relation, zapt, 2);
+ attr_init_APT(&sort_relation, zapt, 7);
+ attr_init_APT(&weight, zapt, 9);
+ attr_init_APT(&use, zapt, 1);
+
+ completeness_value = attr_find(&completeness, NULL);
+ structure_value = attr_find_ex(&structure, NULL, &structure_str);
+ relation_value = attr_find(&relation, NULL);
+ sort_relation_value = attr_find(&sort_relation, NULL);
+ weight_value = attr_find(&weight, NULL);
+ use_value = attr_find(&use, NULL);
if (completeness_value == 2 || completeness_value == 3)
*complete_flag = 1;
else
*complete_flag = 0;
- *reg_id = 0;
+ *index_type = 0;
- *sort_flag = (sort_relation_value > 0) ? 1 : 0;
+ *sort_flag =(sort_relation_value > 0) ? 1 : 0;
*search_type = "phrase";
- *rank_type = "void";
+ strcpy(rank_type, "void");
if (relation_value == 102)
- *rank_type = "rank";
-
+ {
+ if (weight_value == -1)
+ weight_value = 34;
+ sprintf(rank_type, "rank,w=%d,u=%d", weight_value, use_value);
+ }
if (*complete_flag)
- *reg_id = 'p';
+ *index_type = "p";
else
- *reg_id = 'w';
+ *index_type = "w";
switch (structure_value)
{
case 6: /* word list */
break;
case 107: /* local-number */
*search_type = "local";
- *reg_id = 0;
+ *index_type = 0;
break;
case 109: /* numeric string */
- *reg_id = 'n';
+ *index_type = "n";
*search_type = "numeric";
break;
case 104: /* urx */
- *reg_id = 'u';
+ *index_type = "u";
*search_type = "phrase";
break;
case 3: /* key */
- *reg_id = '0';
+ *index_type = "0";
+ *search_type = "phrase";
+ break;
+ case 4: /* year */
+ *index_type = "y";
*search_type = "phrase";
break;
+ case 5: /* date */
+ *index_type = "d";
+ *search_type = "phrase";
+ break;
+ case -2:
+ if (structure_str && *structure_str)
+ *index_type = structure_str;
+ else
+ return -1;
+ break;
default:
return -1;
}
return 0;
}
-int zebra_replace_sub(ZebraMaps zms, unsigned reg_id, const char *ex_list,
- const char *input_str, int input_len, WRBUF wrbuf);
-
-WRBUF zebra_replace(ZebraMaps zms, unsigned reg_id, const char *ex_list,
+WRBUF zebra_replace(zebra_map_t zm, const char *ex_list,
const char *input_str, int input_len)
{
- struct zebra_map *zm = zebra_map_get (zms, reg_id);
-
- wrbuf_rewind(zms->wrbuf_1);
- wrbuf_write(zms->wrbuf_1, input_str, input_len);
- if (!zm->replace_tokens)
- return zms->wrbuf_1;
-
-#if 0
- logf (LOG_LOG, "in:%.*s:", wrbuf_len(zms->wrbuf_1),
- wrbuf_buf(zms->wrbuf_1));
-#endif
- for (;;)
+ wrbuf_rewind(zm->zebra_maps->wrbuf_1);
+ wrbuf_write(zm->zebra_maps->wrbuf_1, input_str, input_len);
+ return zm->zebra_maps->wrbuf_1;
+}
+
+#define SE_CHARS ";,.()-/?<> \r\n\t"
+
+static int tokenize_simple(zebra_map_t zm,
+ const char **result_buf, size_t *result_len)
+{
+ char *buf = wrbuf_buf(zm->input_str);
+ size_t len = wrbuf_len(zm->input_str);
+ size_t i = zm->simple_off;
+ size_t start;
+
+ while (i < len && strchr(SE_CHARS, buf[i]))
+ i++;
+ start = i;
+ while (i < len && !strchr(SE_CHARS, buf[i]))
+ {
+ if (buf[i] > 32 && buf[i] < 127)
+ buf[i] = tolower(buf[i]);
+ i++;
+ }
+
+ zm->simple_off = i;
+ if (start != i)
{
- if (!zebra_replace_sub(zms, reg_id, ex_list, wrbuf_buf(zms->wrbuf_1),
- wrbuf_len(zms->wrbuf_1), zms->wrbuf_2))
- return zms->wrbuf_2;
- if (!zebra_replace_sub(zms, reg_id, ex_list, wrbuf_buf(zms->wrbuf_2),
- wrbuf_len(zms->wrbuf_2), zms->wrbuf_1))
- return zms->wrbuf_1;
+ *result_buf = buf + start;
+ *result_len = i - start;
+ return 1;
}
return 0;
-}
+ }
+
-int zebra_replace_sub(ZebraMaps zms, unsigned reg_id, const char *ex_list,
- const char *input_str, int input_len, WRBUF wrbuf)
+int zebra_map_tokenize_next(zebra_map_t zm,
+ const char **result_buf, size_t *result_len,
+ const char **display_buf, size_t *display_len)
{
- int i = -1;
- int no_replaces = 0;
- struct zebra_map *zm = zebra_map_get (zms, reg_id);
+ assert(zm->use_chain);
- wrbuf_rewind(wrbuf);
- for (i = -1; i <= input_len; )
+#if YAZ_HAVE_ICU
+ if (!zm->icu_chain)
+ return tokenize_simple(zm, result_buf, result_len);
+ else
{
- struct zm_token *token;
- char replace_string[128];
- int replace_out;
- int replace_in = 0;
+ UErrorCode status;
+ while (icu_chain_next_token(zm->icu_chain, &status))
+ {
+ if (!U_SUCCESS(status))
+ return 0;
+ *result_buf = icu_chain_token_sortkey(zm->icu_chain);
+ assert(*result_buf);
- for (token = zm->replace_tokens; !replace_in && token;
- token = token->next)
- {
- int j = 0;
- int replace_done = 0;
- replace_out = 0;
- for (;; j++)
- {
- int c;
- if (!token->token_from[j])
- {
- replace_in = j;
- break;
- }
- if (ex_list && strchr (ex_list, token->token_from[j]))
- break;
- if (i+j < 0 || j+i >= input_len)
- c = ' ';
- else
- c = input_str[j+i] & 255;
- if (token->token_from[j] == ZEBRA_REPLACE_ANY)
- {
- if (c == ' ')
- break;
- replace_string[replace_out++] = c;
- }
- else
- {
- if (c != token->token_from[j])
- {
- break;
- }
- if (!replace_done)
- {
- const char *cp = token->token_to;
- replace_done = 1;
- for (; cp && *cp; cp++)
- replace_string[replace_out++] = *cp;
- }
- }
- }
- }
- if (!replace_in)
- {
- if (i >= 0 && i < input_len)
- wrbuf_putc(wrbuf, input_str[i]);
- i++;
- }
- else
- {
- no_replaces++;
- if (replace_out)
- wrbuf_write(wrbuf, replace_string, replace_out);
- i += replace_in;
- }
+ *result_len = strlen(*result_buf);
+
+ if (display_buf)
+ {
+ *display_buf = icu_chain_token_display(zm->icu_chain);
+ if (display_len)
+ *display_len = strlen(*display_buf);
+ }
+ if (zm->debug)
+ {
+ wrbuf_rewind(zm->print_str);
+ wrbuf_write_escaped(zm->print_str, *result_buf, *result_len);
+ yaz_log(YLOG_LOG, "output %s", wrbuf_cstr(zm->print_str));
+ }
+
+ if (**result_buf != '\0')
+ return 1;
+ }
}
-#if 0
- logf (LOG_LOG, "out:%.*s:", wrbuf_len(wrbuf), wrbuf_buf(wrbuf));
+ return 0;
+#else
+ return tokenize_simple(zm, result_buf, result_len);
#endif
- return no_replaces;
}
+
+int zebra_map_tokenize_start(zebra_map_t zm,
+ const char *buf, size_t len)
+{
+#if YAZ_HAVE_ICU
+ int ret;
+#endif
+ assert(zm->use_chain);
+
+ wrbuf_rewind(zm->input_str);
+ wrbuf_write(zm->input_str, buf, len);
+ zm->simple_off = 0;
+#if YAZ_HAVE_ICU
+ if (zm->icu_chain)
+ {
+ UErrorCode status;
+ if (zm->debug)
+ {
+ wrbuf_rewind(zm->print_str);
+ wrbuf_write_escaped(zm->print_str, wrbuf_buf(zm->input_str),
+ wrbuf_len(zm->input_str));
+
+ yaz_log(YLOG_LOG, "input %s",
+ wrbuf_cstr(zm->print_str));
+ }
+ ret = icu_chain_assign_cstr(zm->icu_chain,
+ wrbuf_cstr(zm->input_str), &status);
+ if (!ret && !U_SUCCESS(status))
+ {
+ if (zm->debug)
+ {
+ yaz_log(YLOG_WARN, "bad encoding for input");
+ }
+ return -1;
+ }
+ }
+#endif
+ return 0;
+}
+
+int zebra_maps_is_icu(zebra_map_t zm)
+{
+ assert(zm);
+#if YAZ_HAVE_ICU
+ return zm->use_chain;
+#else
+ return 0;
+#endif
+}
+
+
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * c-file-style: "Stroustrup"
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+