X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Frecgrs.c;h=597488814a7d1111713ca8664146e286b37edb6e;hb=3f1ccb1a5d47e99a07c4236f149f199dfebd21d0;hp=23025671c6ec83e66e9cc13422ab08ef0d328e2c;hpb=08dfaf179e9c3bf0ec66a3a34fab11d5cd5d1fa8;p=idzebra-moved-to-github.git diff --git a/recctrl/recgrs.c b/recctrl/recgrs.c index 2302567..5974888 100644 --- a/recctrl/recgrs.c +++ b/recctrl/recgrs.c @@ -3,7 +3,25 @@ * All rights reserved. * * $Log: recgrs.c,v $ - * Revision 1.40 2001-03-29 21:31:31 adam + * Revision 1.46 2002-04-13 18:16:43 adam + * More XPATH work; common sequence numbers for extract keys + * + * Revision 1.45 2002/04/12 14:40:42 adam + * Work on XPATH + * + * Revision 1.44 2002/04/11 20:09:47 adam + * work on string tag indexing + * + * Revision 1.43 2002/03/21 23:06:36 adam + * Source 'tag' in abs-file + * + * Revision 1.42 2002/02/20 17:30:01 adam + * Work on new API. Locking system re-implemented + * + * Revision 1.41 2001/05/22 21:01:47 adam + * Removed print of data1 tree on stdout so that inetd works again. + * + * Revision 1.40 2001/03/29 21:31:31 adam * Fixed "record begin" for Tcl filter. * * Revision 1.39 2000/12/05 19:09:15 adam @@ -312,8 +330,81 @@ static void grs_destroy(void *clientData) free (h); } -static void index_tag (data1_node *par, data1_node *n, - struct recExtractCtrl *p, int level, RecWord *wrd) +static void index_xpath (data1_node *n, struct recExtractCtrl *p, + int level, RecWord *wrd, int use) +{ + int i; + char tag_path_full[1024]; + size_t flen = 0; + data1_node *nn; + + switch (n->which) + { + case DATA1N_data: + wrd->reg_type = 'w'; + wrd->string = n->u.data.data; + wrd->length = n->u.data.len; + wrd->attrSet = VAL_IDXPATH, + wrd->attrUse = use; + if (p->flagShowRecords) + { + printf("%*s data=", (level + 1) * 4, ""); + for (i = 0; ilength && i < 8; i++) + fputc (wrd->string[i], stdout); + printf("\n"); + } + else + { + (*p->tokenAdd)(wrd); + } + break; + case DATA1N_tag: + for (nn = n; nn; nn = nn->parent) + { + if (n->which == DATA1N_tag) + { + size_t tlen = strlen(nn->u.tag.tag); + if (tlen + flen > (sizeof(tag_path_full)-2)) + return; + memcpy (tag_path_full + flen, nn->u.tag.tag, tlen); + flen += tlen; + tag_path_full[flen++] = '/'; + } + else if (n->which == DATA1N_root) + { + size_t tlen = strlen(nn->u.root.type); + if (tlen + flen > (sizeof(tag_path_full)-2)) + return; + memcpy (tag_path_full + flen, nn->u.root.type, tlen); + flen += tlen; + tag_path_full[flen++] = '/'; + break; + } + } + wrd->reg_type = '0'; + wrd->string = tag_path_full; + wrd->length = flen; + wrd->attrSet = VAL_IDXPATH, + wrd->attrUse = use; + if (p->flagShowRecords) + { + printf("%*s tag=", (level + 1) * 4, ""); + for (i = 0; ilength && i < 40; i++) + fputc (wrd->string[i], stdout); + if (i == 40) + printf (" .."); + printf("\n"); + } + else + { + (*p->tokenAdd)(wrd); + } + break; + } +} + +static void index_termlist (data1_node *par, data1_node *n, + struct recExtractCtrl *p, int level, RecWord *wrd) { data1_termlist *tlist = 0; data1_datatype dtype = DATA1K_string; @@ -342,6 +433,11 @@ static void index_tag (data1_node *par, data1_node *n, wrd->string = n->u.data.data; wrd->length = n->u.data.len; } + else if (!strcmp (tlist->source, "tag") && n->which == DATA1N_tag) + { + wrd->string = n->u.tag.tag; + wrd->length = strlen(n->u.tag.tag); + } else if (sscanf (tlist->source, "attr(%511[^)])", xattr) == 1 && n->which == DATA1N_tag) { @@ -384,10 +480,9 @@ static void index_tag (data1_node *par, data1_node *n, } } -static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) +static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level, + RecWord *wrd) { - RecWord wrd; - (*p->init)(p, &wrd); /* set defaults */ for (; n; n = n->next) { if (p->flagShowRecords) /* display element description to user */ @@ -395,7 +490,7 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) if (n->which == DATA1N_root) { printf("%*s", level * 4, ""); - printf("Record type: '%s'\n", n->u.root.absyn->name); + printf("Record type: '%s'\n", n->u.root.type); } else if (n->which == DATA1N_tag) { @@ -427,15 +522,19 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) } } - if (n->child) - if (dumpkeys(n->child, p, level + 1) < 0) - return -1; - if (n->which == DATA1N_tag) { - index_tag (n, n, p, level, &wrd); + index_termlist (n, n, p, level, wrd); + /* index start tag */ + if (!n->root->u.root.absyn) + index_xpath (n, p, level, wrd, 1); } + if (n->child) + if (dumpkeys(n->child, p, level + 1, wrd) < 0) + return -1; + + if (n->which == DATA1N_data) { data1_node *par = get_parent_tag(p->dh, n); @@ -454,8 +553,20 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) } if (par) - index_tag (par, n, p, level, &wrd); + index_termlist (par, n, p, level, wrd); + if (!n->root->u.root.absyn) + index_xpath (n, p, level, wrd, 1016); + } + + if (n->which == DATA1N_tag) + { + /* index end tag */ + if (!n->root->u.root.absyn) + index_xpath (n, p, level, wrd, 2); + } + + if (p->flagShowRecords && n->which == DATA1N_root) { printf("%*s-------------\n\n", level * 4, ""); @@ -468,15 +579,19 @@ int grs_extract_tree(struct recExtractCtrl *p, data1_node *n) { oident oe; int oidtmp[OID_SIZE]; + RecWord wrd; oe.proto = PROTO_Z3950; oe.oclass = CLASS_SCHEMA; - oe.value = n->u.root.absyn->reference; - - if ((oid_ent_to_oid (&oe, oidtmp))) - (*p->schemaAdd)(p, oidtmp); - - return dumpkeys(n, p, 0); + if (n->u.root.absyn) + { + oe.value = n->u.root.absyn->reference; + + if ((oid_ent_to_oid (&oe, oidtmp))) + (*p->schemaAdd)(p, oidtmp); + } + (*p->init)(p, &wrd); + return dumpkeys(n, p, 0, &wrd); } static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p, @@ -486,6 +601,7 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p, struct grs_read_info gri; oident oe; int oidtmp[OID_SIZE]; + RecWord wrd; gri.readf = p->readf; gri.seekf = p->seekf; @@ -502,13 +618,21 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p, return RECCTRL_EXTRACT_EOF; oe.proto = PROTO_Z3950; oe.oclass = CLASS_SCHEMA; +#if 0 if (!n->u.root.absyn) return RECCTRL_EXTRACT_ERROR; - oe.value = n->u.root.absyn->reference; - if ((oid_ent_to_oid (&oe, oidtmp))) - (*p->schemaAdd)(p, oidtmp); - - if (dumpkeys(n, p, 0) < 0) +#endif + if (n->u.root.absyn) + { + oe.value = n->u.root.absyn->reference; + if ((oid_ent_to_oid (&oe, oidtmp))) + (*p->schemaAdd)(p, oidtmp); + } +#if 0 + data1_pr_tree (p->dh, n, stdout); +#endif + (*p->init)(p, &wrd); + if (dumpkeys(n, p, 0, &wrd) < 0) { data1_free_tree(p->dh, n); return RECCTRL_EXTRACT_ERROR; @@ -716,20 +840,21 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) * family) */ logf (LOG_DEBUG, "grs_retrieve: syntax mapping"); - for (map = node->u.root.absyn->maptabs; map; map = map->next) - { - if (map->target_absyn_ref == p->input_format) - { - onode = node; - if (!(node = data1_map_record(p->dh, onode, map, mem))) - { - p->diagnostic = 14; - nmem_destroy (mem); - return 0; - } - break; - } - } + if (node->u.root.absyn) + for (map = node->u.root.absyn->maptabs; map; map = map->next) + { + if (map->target_absyn_ref == p->input_format) + { + onode = node; + if (!(node = data1_map_record(p->dh, onode, map, mem))) + { + p->diagnostic = 14; + nmem_destroy (mem); + return 0; + } + break; + } + } logf (LOG_DEBUG, "grs_retrieve: schemaIdentifier"); if (node->u.root.absyn && node->u.root.absyn->reference != VAL_NONE && @@ -783,7 +908,9 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) else if (p->comp && !res) selected = 1; +#if 0 data1_pr_tree (p->dh, node, stdout); +#endif logf (LOG_DEBUG, "grs_retrieve: transfer syntax mapping"); switch (p->output_format = (p->input_format != VAL_NONE ? p->input_format : VAL_SUTRS))