X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=util%2Fzebramap.c;h=8fd2227563287055e271474accae0030f5f3c021;hb=3db15cf883ad80b6a4b0e960f3c3b59c86b59c6b;hp=a925c7909fca2cdd26aea1ceae9ae8bc8000f589;hpb=3d01abbe1238c82d62e4c01eb7cac760d3d53377;p=idzebra-moved-to-github.git diff --git a/util/zebramap.c b/util/zebramap.c index a925c79..8fd2227 100644 --- a/util/zebramap.c +++ b/util/zebramap.c @@ -1,10 +1,30 @@ /* - * Copyright (C) 1994-1997, Index Data I/S + * Copyright (C) 1994-1998, Index Data I/S * All rights reserved. * Sebastian Hammer, Adam Dickmeiss * * $Log: zebramap.c,v $ - * Revision 1.1 1997-10-27 14:33:06 adam + * 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 @@ -21,7 +41,10 @@ struct zebra_map { int reg_type; + int completeness; + int sort_flag; chrmaptab maptab; + const char *maptab_name; struct zebra_map *next; }; @@ -29,6 +52,9 @@ struct zebra_maps { char *tabpath; NMEM nmem; struct zebra_map *map_list; + char temp_map_str[2]; + const char *temp_map_ptr[2]; + struct zebra_map **lookup_array; }; void zebra_maps_close (ZebraMaps zms) @@ -36,85 +62,143 @@ void zebra_maps_close (ZebraMaps zms) struct zebra_map *zm = zms->map_list; while (zm) { - struct zebra_map *zm_next = zm->next; - - chrmaptab_destroy (zm->maptab); - xfree (zm); - zm = zm_next; + if (zm->maptab) + chrmaptab_destroy (zm->maptab); + zm = zm->next; } nmem_destroy (zms->nmem); xfree (zms); } -ZebraMaps zebra_maps_open (const char *tabpath) +static void zebra_map_read (ZebraMaps zms, const char *name) +{ + FILE *f; + char line[512]; + char *argv[10]; + int argc; + struct zebra_map **zm = 0, *zp; + + if (!(f = yaz_path_fopen(zms->tabpath, name, "r"))) + { + logf(LOG_WARN|LOG_ERRNO, "%s", name); + return ; + } + while ((argc = readconf_line(f, line, 512, argv, 10))) + { + if (!strcmp (argv[0], "index") && argc == 2) + { + if (!zm) + zm = &zms->map_list; + else + zm = &(*zm)->next; + *zm = nmem_malloc (zms->nmem, sizeof(**zm)); + (*zm)->reg_type = argv[1][0]; + (*zm)->maptab_name = NULL; + (*zm)->maptab = NULL; + (*zm)->sort_flag = 0; + (*zm)->completeness = 0; + } + else if (!strcmp (argv[0], "sort") && argc == 2) + { + if (!zm) + zm = &zms->map_list; + else + zm = &(*zm)->next; + *zm = nmem_malloc (zms->nmem, sizeof(**zm)); + (*zm)->reg_type = argv[1][0]; + (*zm)->maptab_name = NULL; + (*zm)->sort_flag = 1; + (*zm)->maptab = NULL; + (*zm)->completeness = 0; + } + else if (zm && !strcmp (argv[0], "charmap") && argc == 2) + { + (*zm)->maptab_name = nmem_strdup (zms->nmem, argv[1]); + } + else if (zm && !strcmp (argv[0], "completeness") && argc == 2) + { + (*zm)->completeness = atoi (argv[1]); + } + } + if (zm) + (*zm)->next = NULL; + fclose (f); + + for (zp = zms->map_list; zp; zp = zp->next) + zms->lookup_array[zp->reg_type & 255] = zp; +} + +static void zms_map_handle (void *p, const char *name, const char *value) +{ + ZebraMaps zms = p; + + zebra_map_read (zms, value); +} + +ZebraMaps zebra_maps_open (const char *tabpath, Res res) { ZebraMaps zms = xmalloc (sizeof(*zms)); + int i; zms->nmem = nmem_create (); zms->tabpath = nmem_strdup (zms->nmem, tabpath); zms->map_list = NULL; + + 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 = + nmem_malloc (zms->nmem, sizeof(*zms->lookup_array)*256); + for (i = 0; i<256; i++) + zms->lookup_array[i] = 0; + if (!res_trav (res, "index", zms, zms_map_handle)) + zebra_map_read (zms, "default.idx"); return zms; } -chrmaptab zebra_map_get (ZebraMaps zms, int reg_type) +struct zebra_map *zebra_map_get (ZebraMaps zms, int reg_type) { - char name[512]; - struct zebra_map *zm; + return zms->lookup_array[reg_type]; +} - for (zm = zms->map_list; zm; zm = zm->next) +chrmaptab zebra_charmap_get (ZebraMaps zms, int reg_type) +{ + struct zebra_map *zm = zebra_map_get (zms, reg_type); + if (!zm) { - if (reg_type == zm->reg_type) - return zm->maptab; + logf (LOG_WARN, "Unknown register type: %c", reg_type); + return NULL; } - *name = '\0'; - switch (reg_type) + if (!zm->maptab) { - case 'w': - case 'p': - strcat (name, "string"); - break; - case 'n': - strcat (name, "numeric"); - break; - case 'u': - strcat (name, "urx"); - break; - default: - strcat (name, "null"); + if (!strcmp (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); + else + logf(LOG_DEBUG, "Read character table %s", zm->maptab_name); } - strcat (name, ".chr"); - - zm = xmalloc (sizeof(*zm)); - zm->reg_type = reg_type; - zm->next = zms->map_list; - zms->map_list = zm; - if (!(zm->maptab = chrmaptab_create (zms->tabpath, name, 0))) - logf(LOG_WARN, "Failed to read character table %s", name); - else - logf(LOG_LOG, "Read table %s", name); return zm->maptab; } const char **zebra_maps_input (ZebraMaps zms, int reg_type, const char **from, int len) { - static char str[2] = {0,0}; - static const char *buf[2] = {0,0}; chrmaptab maptab; - maptab = zebra_map_get (zms, reg_type); + maptab = zebra_charmap_get (zms, reg_type); if (maptab) return chr_map_input(maptab, from, len); - - if (isalnum(**from)) - { - str[0] = isupper(**from) ? tolower(**from) : **from; - buf[0] = str; - } - else - buf[0] = (char*) CHR_SPACE; + + zms->temp_map_str[0] = **from; + (*from)++; - return buf; + return zms->temp_map_ptr; } const char *zebra_maps_output(ZebraMaps zms, int reg_type, const char **from) @@ -123,7 +207,7 @@ const char *zebra_maps_output(ZebraMaps zms, int reg_type, const char **from) unsigned char i = (unsigned char) **from; static char buf[2] = {0,0}; - maptab = zebra_map_get (zms, reg_type); + maptab = zebra_charmap_get (zms, reg_type); if (maptab) return chr_map_output (maptab, from, 1); (*from)++; @@ -138,16 +222,17 @@ typedef struct { int type; int major; int minor; - Z_AttributesPlusTerm *zapt; + Z_AttributeElement **attributeList; + int num_attributes; } AttrType; static int attr_find (AttrType *src, oid_value *attributeSetP) { - while (src->major < src->zapt->num_attributes) + while (src->major < src->num_attributes) { Z_AttributeElement *element; - element = src->zapt->attributeList[src->major]; + element = src->attributeList[src->major]; if (src->type == *element->attributeType) { switch (element->which) @@ -186,10 +271,20 @@ static int attr_find (AttrType *src, oid_value *attributeSetP) return -1; } -static void attr_init (AttrType *src, Z_AttributesPlusTerm *zapt, - int type) +static void attr_init_APT (AttrType *src, Z_AttributesPlusTerm *zapt, int type) +{ + + src->attributeList = zapt->attributeList; + src->num_attributes = zapt->num_attributes; + src->type = type; + src->major = 0; + src->minor = 0; +} + +static void attr_init_AttrList (AttrType *src, Z_AttributeList *list, int type) { - src->zapt = zapt; + src->attributeList = list->attributes; + src->num_attributes = list->num_attributes; src->type = type; src->major = 0; src->minor = 0; @@ -198,12 +293,29 @@ static void attr_init (AttrType *src, Z_AttributesPlusTerm *zapt, /* ------------------------------------ */ int zebra_maps_is_complete (ZebraMaps zms, int reg_type) +{ + struct zebra_map *zm = zebra_map_get (zms, reg_type); + if (zm) + return zm->completeness; + return 0; +} + +int zebra_maps_is_sort (ZebraMaps zms, int reg_type) { - if (reg_type == 'p') - return 1; + struct zebra_map *zm = zebra_map_get (zms, reg_type); + if (zm) + return zm->sort_flag; return 0; } +int zebra_maps_sort (ZebraMaps zms, Z_SortAttributes *sortAttributes) +{ + AttrType use; + attr_init_AttrList (&use, sortAttributes->list, 1); + + return attr_find (&use, NULL); +} + int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt, int *reg_type, char **search_type, int *complete_flag) { @@ -214,12 +326,13 @@ int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt, int structure_value; int relation_value; - attr_init (&structure, zapt, 4); - attr_init (&completeness, zapt, 6); - attr_init (&relation, zapt, 2); + attr_init_APT (&structure, zapt, 4); + attr_init_APT (&completeness, zapt, 6); + attr_init_APT (&relation, zapt, 2); completeness_value = attr_find (&completeness, NULL); structure_value = attr_find (&structure, NULL); + relation_value = attr_find (&relation, NULL); if (completeness_value == 2 || completeness_value == 3) *complete_flag = 1; @@ -233,13 +346,16 @@ int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt, switch (structure_value) { + case 6: /* word list */ + case 105: /* free-form-text */ + case 106: /* document-text */ + *search_type = "ranked"; + *reg_type = 'w'; + break; case -1: case 1: /* phrase */ case 2: /* word */ case 3: /* key */ - case 6: /* word list */ - case 105: /* free-form-text */ - case 106: /* document-text */ case 108: /* string */ if (*complete_flag) *reg_type = 'p';