X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=util%2Fzebramap.c;h=3fe7f06d4210f9826221df8c4841c9fb280a7e6c;hp=6418560a587e90457ab04ccd41bc38eacfaa4a47;hb=e199777080c6fa0963d51b7df1763fd5286ca9a4;hpb=6ba9698e88c0283e40fa5980a1a6b551fff2d597 diff --git a/util/zebramap.c b/util/zebramap.c index 6418560..3fe7f06 100644 --- a/util/zebramap.c +++ b/util/zebramap.c @@ -1,5 +1,5 @@ -/* $Id: zebramap.c,v 1.49 2006-05-19 13:49:38 adam Exp $ - Copyright (C) 1995-2006 +/* $Id: zebramap.c,v 1.61 2007-10-31 16:56:15 adam Exp $ + Copyright (C) 1995-2007 Index Data ApS This file is part of the Zebra server. @@ -28,44 +28,45 @@ #include #include -#include +#include #define ZEBRA_MAP_TYPE_SORT 1 #define ZEBRA_MAP_TYPE_INDEX 2 +#define ZEBRA_MAP_TYPE_STATICRANK 3 #define ZEBRA_REPLACE_ANY 300 struct zebra_map { - unsigned reg_id; + const char *id; int completeness; int positioned; + int alwaysmatches; + int first_in_field; int type; union { struct { - int dummy; - } index; - struct { int entry_size; } sort; } u; chrmaptab maptab; const char *maptab_name; + zebra_maps_t zebra_maps; struct zebra_map *next; }; -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; int no_maps; + 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) @@ -74,19 +75,44 @@ void zebra_maps_close(ZebraMaps zms) chrmaptab_destroy(zm->maptab); zm = zm->next; } - wrbuf_free(zms->wrbuf_1, 1); + wrbuf_destroy(zms->wrbuf_1); nmem_destroy(zms->nmem); xfree(zms); } -ZEBRA_RES zebra_maps_read_file(ZebraMaps zms, const char *fname) +zebra_map_t zebra_add_map(zebra_maps_t zms, const char *index_type, + int map_type) +{ + 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->maptab = 0; + zm->type = map_type; + zm->completeness = 0; + zm->positioned = 0; + zm->alwaysmatches = 0; + zm->first_in_field = 0; + + zm->next = 0; + *zms->last_map = zm; + zms->last_map = &zm->next; + + zms->no_maps++; + + return zm; +} + +ZEBRA_RES zebra_maps_read_file(zebra_maps_t zms, const char *fname) { FILE *f; char line[512]; char *argv[10]; int argc; int lineno = 0; - struct zebra_map **zm = 0, *zp; + int failures = 0; + zebra_map_t zm = 0; if (!(f = yaz_fopen(zms->tabpath, fname, "r", zms->tabroot))) { @@ -95,78 +121,116 @@ ZEBRA_RES zebra_maps_read_file(ZebraMaps zms, const char *fname) } while ((argc = readconf_line(f, &lineno, line, 512, argv, 10))) { - if (!yaz_matchstr(argv[0], "index") && argc == 2) + if (argc == 1) + { + yaz_log(YLOG_WARN, "%s:%d: Missing arguments for '%s'", + fname, lineno, argv[0]); + failures++; + break; + } + if (argc > 2) + { + yaz_log(YLOG_WARN, "%s:%d: Too many arguments for '%s'", + fname, lineno, argv[0]); + failures++; + break; + } + 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")) { - 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; - zms->no_maps++; + zm = zebra_add_map(zms, argv[1], ZEBRA_MAP_TYPE_SORT); + zm->u.sort.entry_size = 80; } - else if (!yaz_matchstr(argv[0], "sort") && argc == 2) + else if (!yaz_matchstr(argv[0], "staticrank")) { - 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; - zms->no_maps++; + zm = zebra_add_map(zms, argv[1], ZEBRA_MAP_TYPE_STATICRANK); + zm->completeness = 1; } - else if (zm && !yaz_matchstr(argv[0], "charmap") && argc == 2) + else if (!zm) + { + yaz_log(YLOG_WARN, "%s:%d: Missing sort/index before '%s'", + fname, lineno, argv[0]); + failures++; + } + 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); + failures++; + } + } + else if (!yaz_matchstr(argv[0], "completeness") && argc == 2) { - (*zm)->maptab_name = nmem_strdup(zms->nmem, argv[1]); + zm->completeness = atoi(argv[1]); } - else if (zm && !yaz_matchstr(argv[0], "completeness") && argc == 2) + else if (!yaz_matchstr(argv[0], "position") && argc == 2) { - (*zm)->completeness = atoi(argv[1]); + zm->positioned = atoi(argv[1]); } - else if (zm && !yaz_matchstr(argv[0], "position") && argc == 2) + else if (!yaz_matchstr(argv[0], "alwaysmatches") && argc == 2) { - (*zm)->positioned = atoi(argv[1]); + if (zm->type != ZEBRA_MAP_TYPE_STATICRANK) + zm->alwaysmatches = atoi(argv[1]); + else + { + yaz_log(YLOG_WARN|YLOG_FATAL, "%s:%d: alwaysmatches for " + "staticrank is invalid", fname, lineno); + failures++; + } } - else if (zm && !yaz_matchstr(argv[0], "entrysize") && 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 + { + yaz_log(YLOG_WARN, + "%s:%d: entrysize only valid in sort section", + fname, lineno); + failures++; + } + + } + else { - if ((*zm)->type == ZEBRA_MAP_TYPE_SORT) - (*zm)->u.sort.entry_size = atoi(argv[1]); + yaz_log(YLOG_WARN, "%s:%d: Unrecognized directive '%s'", + fname, lineno, argv[0]); + failures++; } } - if (zm) - (*zm)->next = NULL; yaz_fclose(f); - for (zp = zms->map_list; zp; zp = zp->next) - zms->lookup_array[zp->reg_id] = zp; - + if (failures) + return ZEBRA_FAIL; return ZEBRA_OK; } -ZebraMaps zebra_maps_open(Res res, const char *base_path, +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->no_maps = 0; - zms->tabpath = nmem_strdup(zms->nmem, profile_path); + 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 = NULL; + zms->map_list = 0; + zms->last_map = &zms->map_list; zms->temp_map_str[0] = '\0'; zms->temp_map_str[1] = '\0'; @@ -174,49 +238,51 @@ ZebraMaps zebra_maps_open(Res res, const char *base_path, 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); zms->wrbuf_1 = wrbuf_alloc(); - for (i = 0; i<256; i++) - zms->lookup_array[i] = 0; return zms; } -struct zebra_map *zebra_map_get(ZebraMaps zms, unsigned reg_id) +zebra_map_t zebra_map_get(zebra_maps_t zms, const char *id) { - assert(reg_id >= 0 && reg_id <= 255); - return zms->lookup_array[reg_id]; + zebra_map_t zm; + for (zm = zms->map_list; zm; zm = zm->next) + if (!strcmp(zm->id, id)) + break; + return zm; } -chrmaptab zebra_charmap_get(ZebraMaps zms, unsigned reg_id) +zebra_map_t zebra_map_get_or_add(zebra_maps_t zms, const char *id) { - struct zebra_map *zm = zebra_map_get(zms, reg_id); + struct zebra_map *zm = zebra_map_get(zms, id); if (!zm) { - zm = (struct zebra_map *) nmem_malloc(zms->nmem, sizeof(*zm)); + zm = zebra_add_map(zms, id, ZEBRA_MAP_TYPE_INDEX); - /* no reason to warn if no maps are installed at ALL */ - if (zms->no_maps) - yaz_log(YLOG_WARN, "Unknown register type: %c", reg_id); + /* no reason to warn if no maps are installed at ALL + Note that zebra_add_maps increments no_maps .. + */ + if (zms->no_maps > 1) + yaz_log(YLOG_WARN, "Unknown register type: %s", id); + else + zms->no_maps = 0; - zm->reg_id = reg_id; zm->maptab_name = nmem_strdup(zms->nmem, "@"); - zm->maptab = NULL; - zm->type = ZEBRA_MAP_TYPE_INDEX; 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, "@")) return NULL; - if (!(zm->maptab = chrmaptab_create(zms->tabpath, + if (!(zm->maptab = chrmaptab_create(zm->zebra_maps->tabpath, zm->maptab_name, - zms->tabroot))) + zm->zebra_maps->tabroot))) yaz_log(YLOG_WARN, "Failed to read character table %s", zm->maptab_name); else @@ -225,28 +291,26 @@ chrmaptab zebra_charmap_get(ZebraMaps zms, unsigned reg_id) return zm->maptab; } -const char **zebra_maps_input(ZebraMaps zms, unsigned reg_id, +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, 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; } -const char **zebra_maps_search(ZebraMaps zms, unsigned reg_id, +const char **zebra_maps_search(zebra_map_t zm, const char **from, int len, int *q_map_match) { chrmaptab maptab; *q_map_match = 0; - maptab = zebra_charmap_get(zms, reg_id); + maptab = zebra_charmap_get(zm); if (maptab) { const char **map; @@ -260,50 +324,74 @@ const char **zebra_maps_search(ZebraMaps zms, unsigned reg_id, if (map) return map; } - 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; } -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); } - /* ------------------------------------ */ -int zebra_maps_is_complete(ZebraMaps zms, unsigned reg_id) +int zebra_maps_is_complete(zebra_map_t zm) { - struct zebra_map *zm = zebra_map_get(zms, reg_id); if (zm) return zm->completeness; return 0; } -int zebra_maps_is_positioned(ZebraMaps zms, unsigned reg_id) +int zebra_maps_is_positioned(zebra_map_t zm) { - struct zebra_map *zm = zebra_map_get(zms, reg_id); if (zm) return zm->positioned; return 0; } + +int zebra_maps_is_index(zebra_map_t zm) +{ + if (zm) + return zm->type == ZEBRA_MAP_TYPE_INDEX; + return 0; +} + +int zebra_maps_is_staticrank(zebra_map_t zm) +{ + if (zm) + return zm->type == ZEBRA_MAP_TYPE_STATICRANK; + return 0; +} -int zebra_maps_is_sort(ZebraMaps zms, unsigned reg_id) +int zebra_maps_is_sort(zebra_map_t zm) { - struct zebra_map *zm = zebra_map_get(zms, reg_id); if (zm) return zm->type == ZEBRA_MAP_TYPE_SORT; return 0; } -int zebra_maps_sort(ZebraMaps zms, Z_SortAttributes *sortAttributes, +int zebra_maps_is_alwaysmatches(zebra_map_t zm) +{ + if (zm) + return zm->alwaysmatches; + return 0; +} + +int zebra_maps_is_first_in_field(zebra_map_t zm) +{ + if (zm) + return zm->first_in_field; + return 0; +} + +int zebra_maps_sort(zebra_maps_t zms, Z_SortAttributes *sortAttributes, int *numerical) { AttrType use; @@ -319,8 +407,8 @@ int zebra_maps_sort(ZebraMaps zms, Z_SortAttributes *sortAttributes, return attr_find(&use, NULL); } -int zebra_maps_attr(ZebraMaps zms, Z_AttributesPlusTerm *zapt, - unsigned *reg_id, char **search_type, char *rank_type, +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; @@ -331,6 +419,7 @@ int zebra_maps_attr(ZebraMaps zms, Z_AttributesPlusTerm *zapt, AttrType use; int completeness_value; int structure_value; + const char *structure_str = 0; int relation_value; int sort_relation_value; int weight_value; @@ -344,7 +433,7 @@ int zebra_maps_attr(ZebraMaps zms, Z_AttributesPlusTerm *zapt, attr_init_APT(&use, zapt, 1); completeness_value = attr_find(&completeness, NULL); - structure_value = attr_find(&structure, 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); @@ -354,7 +443,7 @@ int zebra_maps_attr(ZebraMaps zms, Z_AttributesPlusTerm *zapt, *complete_flag = 1; else *complete_flag = 0; - *reg_id = 0; + *index_type = 0; *sort_flag =(sort_relation_value > 0) ? 1 : 0; *search_type = "phrase"; @@ -365,16 +454,10 @@ int zebra_maps_attr(ZebraMaps zms, Z_AttributesPlusTerm *zapt, weight_value = 34; sprintf(rank_type, "rank,w=%d,u=%d", weight_value, use_value); } - if (relation_value == 103) - { - *search_type = "always"; - *reg_id = 'w'; - return 0; - } if (*complete_flag) - *reg_id = 'p'; + *index_type = "p"; else - *reg_id = 'w'; + *index_type = "w"; switch (structure_value) { case 6: /* word list */ @@ -394,40 +477,46 @@ int zebra_maps_attr(ZebraMaps zms, Z_AttributesPlusTerm *zapt, 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 */ - *reg_id = 'y'; + *index_type = "y"; *search_type = "phrase"; break; case 5: /* date */ - *reg_id = 'd'; + *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; } -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) { - wrbuf_rewind(zms->wrbuf_1); - wrbuf_write(zms->wrbuf_1, input_str, input_len); - return zms->wrbuf_1; + wrbuf_rewind(zm->zebra_maps->wrbuf_1); + wrbuf_write(zm->zebra_maps->wrbuf_1, input_str, input_len); + return zm->zebra_maps->wrbuf_1; } /*