X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=util%2Fzebramap.c;h=71c715aee635c793278d2f456a9107f5656b93fd;hb=98e21f950e6d0babe88a6a1ccb03b90f0b3a6b1a;hp=e7878997fc3e13636d5406843f991c55313ed42a;hpb=2730140edaab8fc1df36e4b82fb1a4fcc523de05;p=idzebra-moved-to-github.git diff --git a/util/zebramap.c b/util/zebramap.c index e787899..71c715a 100644 --- a/util/zebramap.c +++ b/util/zebramap.c @@ -4,7 +4,43 @@ * Sebastian Hammer, Adam Dickmeiss * * $Log: zebramap.c,v $ - * Revision 1.12 1999-02-12 13:29:25 adam + * Revision 1.24 2002-04-04 20:50:37 adam + * Multi register works with record paths and data1 profile path + * + * Revision 1.23 2001/11/15 08:41:24 adam + * Fix for weight (bug introduced by previous commit). + * + * Revision 1.22 2001/11/14 22:06:27 adam + * Rank-weight may be controlled via query. + * + * Revision 1.21 2001/01/22 10:42:56 adam + * Added numerical sort. + * + * Revision 1.20 2000/03/02 14:35:19 adam + * Added structure year and date. + * + * Revision 1.19 1999/11/30 13:48:04 adam + * Improved installation. Updated for inclusion of YAZ header files. + * + * 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 @@ -51,13 +87,22 @@ #include #include -#include +#include #include #include #define ZEBRA_MAP_TYPE_SORT 1 #define ZEBRA_MAP_TYPE_INDEX 2 +#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; int completeness; @@ -74,15 +119,18 @@ struct zebra_map { chrmaptab maptab; const char *maptab_name; struct zebra_map *next; + struct zm_token *replace_tokens; }; struct zebra_maps { 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; }; void zebra_maps_close (ZebraMaps zms) @@ -94,6 +142,8 @@ void zebra_maps_close (ZebraMaps zms) chrmaptab_destroy (zm->maptab); zm = zm->next; } + wrbuf_free (zms->wrbuf_1, 1); + wrbuf_free (zms->wrbuf_2, 1); nmem_destroy (zms->nmem); xfree (zms); } @@ -107,7 +157,7 @@ static void zebra_map_read (ZebraMaps zms, const char *name) int lineno = 0; struct zebra_map **zm = 0, *zp; - if (!(f = yaz_path_fopen(zms->tabpath, name, "r"))) + if (!(f = yaz_path_fopen_base(zms->tabpath, name, "r", zms->tabroot))) { logf(LOG_WARN|LOG_ERRNO, "%s", name); return ; @@ -120,13 +170,14 @@ static void zebra_map_read (ZebraMaps zms, const char *name) zm = &zms->map_list; else zm = &(*zm)->next; - *zm = nmem_malloc (zms->nmem, sizeof(**zm)); + *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) { @@ -134,7 +185,7 @@ static void zebra_map_read (ZebraMaps zms, const char *name) zm = &zms->map_list; else zm = &(*zm)->next; - *zm = nmem_malloc (zms->nmem, sizeof(**zm)); + *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; @@ -142,6 +193,7 @@ static void zebra_map_read (ZebraMaps zms, const char *name) (*zm)->maptab = NULL; (*zm)->completeness = 0; (*zm)->positioned = 0; + (*zm)->replace_tokens = 0; } else if (zm && !yaz_matchstr (argv[0], "charmap") && argc == 2) { @@ -160,6 +212,58 @@ static void zebra_map_read (ZebraMaps zms, const char *name) if ((*zm)->type == ZEBRA_MAP_TYPE_SORT) (*zm)->u.sort.entry_size = atoi (argv[1]); } + else if (zm && !yaz_matchstr (argv[0], "replace") && argc >= 2) + { + 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]); +#endif + token->token_from = 0; + if (argc >= 2) + { + char *cp = argv[1]; + int *dp = token->token_from = (int *) + nmem_malloc (zms->nmem, (1+strlen(cp))*sizeof(int)); + while (*cp) + if (*cp == '$') + { + *dp++ = ' '; + cp++; + } + else if (*cp == '.') + { + *dp++ = ZEBRA_REPLACE_ANY; + cp++; + } + else + { + *dp++ = zebra_prim(&cp); +#if 0 + logf (LOG_LOG, " char %2X %c", dp[-1], dp[-1]); +#endif + } + *dp = '\0'; + } + if (argc >= 3) + { + char *cp = argv[2]; + char *dp = token->token_to = + nmem_malloc (zms->nmem, strlen(cp)+1); + while (*cp) + if (*cp == '$') + { + *dp++ = ' '; + cp++; + } + else + *dp++ = zebra_prim(&cp); + *dp = '\0'; + } + else + token->token_to = 0; + } } if (zm) (*zm)->next = NULL; @@ -171,18 +275,22 @@ static void zebra_map_read (ZebraMaps zms, const char *name) static void zms_map_handle (void *p, const char *name, const char *value) { - ZebraMaps zms = p; + ZebraMaps zms = (ZebraMaps) p; zebra_map_read (zms, value); } -ZebraMaps zebra_maps_open (Res res) +ZebraMaps zebra_maps_open (Res res, const char *base) { - ZebraMaps zms = xmalloc (sizeof(*zms)); + ZebraMaps zms = (ZebraMaps) xmalloc (sizeof(*zms)); int i; zms->nmem = nmem_create (); - zms->tabpath = nmem_strdup (zms->nmem, res_get (res, "profilePath")); + zms->tabpath = nmem_strdup (zms->nmem, + res_get_def (res, "profilePath", ".")); + zms->tabroot = 0; + if (base) + zms->tabroot = nmem_strdup (zms->nmem, base); zms->map_list = NULL; zms->temp_map_str[0] = '\0'; @@ -191,12 +299,16 @@ ZebraMaps zebra_maps_open (Res res) zms->temp_map_ptr[0] = zms->temp_map_str; zms->temp_map_ptr[1] = NULL; - zms->lookup_array = + 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(); + return zms; } @@ -210,11 +322,11 @@ chrmaptab zebra_charmap_get (ZebraMaps zms, unsigned reg_id) struct zebra_map *zm = zebra_map_get (zms, reg_id); if (!zm) { - zm = nmem_malloc (zms->nmem, sizeof(*zm)); + zm = (struct zebra_map *) nmem_malloc (zms->nmem, sizeof(*zm)); logf (LOG_WARN, "Unknown register type: %c", reg_id); zm->reg_id = reg_id; - zm->maptab_name = NULL; + zm->maptab_name = nmem_strdup (zms->nmem, "@"); zm->maptab = NULL; zm->type = ZEBRA_MAP_TYPE_INDEX; zm->completeness = 0; @@ -228,7 +340,8 @@ chrmaptab zebra_charmap_get (ZebraMaps zms, unsigned reg_id) if (!zm->maptab_name || !yaz_matchstr (zm->maptab_name, "@")) return NULL; if (!(zm->maptab = chrmaptab_create (zms->tabpath, - zm->maptab_name, 0))) + zm->maptab_name, 0, + zms->tabroot))) logf(LOG_WARN, "Failed to read character table %s", zm->maptab_name); else @@ -252,19 +365,47 @@ const char **zebra_maps_input (ZebraMaps zms, unsigned reg_id, return zms->temp_map_ptr; } +#if 0 +int zebra_maps_input_tokens (ZebraMaps zms, unsigned reg_id, + const char *input_str, int input_len, + WRBUF wrbuf) +{ + 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; + + out = chr_map_input (maptab, str, len); + while (len[1] > 0) + { + while (out && *out && **out == *CHR_SPACE) + out = chr_map_input (maptab, str, len); + } +} +#endif + const char *zebra_maps_output(ZebraMaps zms, unsigned reg_id, const char **from) { - chrmaptab maptab; - unsigned char i = (unsigned char) **from; - static char buf[2] = {0,0}; - - maptab = zebra_charmap_get (zms, reg_id); - if (maptab) - return chr_map_output (maptab, from, 1); - (*from)++; - buf[0] = i; - return buf; + chrmaptab maptab = zebra_charmap_get (zms, reg_id); + if (!maptab) + return 0; + return chr_map_output (maptab, from, 1); } @@ -372,36 +513,48 @@ int zebra_maps_is_sort (ZebraMaps zms, unsigned reg_id) return 0; } -int zebra_maps_sort (ZebraMaps zms, Z_SortAttributes *sortAttributes) +int zebra_maps_sort (ZebraMaps zms, Z_SortAttributes *sortAttributes, + int *numerical) { AttrType use; + 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, + unsigned *reg_id, char **search_type, char *rank_type, int *complete_flag, int *sort_flag) { AttrType completeness; AttrType structure; AttrType relation; AttrType sort_relation; + AttrType weight; int completeness_value; int structure_value; int relation_value; int sort_relation_value; + int weight_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); 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); + weight_value = attr_find (&weight, NULL); if (completeness_value == 2 || completeness_value == 3) *complete_flag = 1; @@ -411,10 +564,13 @@ int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt, *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,%d", weight_value); + } if (*complete_flag) *reg_id = 'p'; else @@ -433,7 +589,6 @@ int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt, case -1: case 1: /* phrase */ case 2: /* word */ - case 3: /* key */ case 108: /* string */ *search_type = "phrase"; break; @@ -449,8 +604,126 @@ int zebra_maps_attr (ZebraMaps zms, Z_AttributesPlusTerm *zapt, *reg_id = 'u'; *search_type = "phrase"; break; + case 3: /* key */ + *reg_id = '0'; + *search_type = "phrase"; + break; + case 4: /* year */ + *reg_id = 'y'; + *search_type = "phrase"; + break; + case 5: /* date */ + *reg_id = 'd'; + *search_type = "phrase"; + 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, + 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 || !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 (;;) + { + 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; + } + 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 i = -1; + int no_replaces = 0; + struct zebra_map *zm = zebra_map_get (zms, reg_id); + + wrbuf_rewind(wrbuf); + for (i = -1; i <= input_len; ) + { + struct zm_token *token; + char replace_string[128]; + int replace_out; + int replace_in = 0; + + 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; + } + } +#if 0 + logf (LOG_LOG, "out:%.*s:", wrbuf_len(wrbuf), wrbuf_buf(wrbuf)); +#endif + return no_replaces; +}