X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Frecgrs.c;h=49fdf91d7af50a0c18225d1e0297e00e908119aa;hb=856d020355a71f37c0b55564f97d52072646b7c8;hp=568774f10502eb0f4867df52ce2099f498dbcded;hpb=bdf6fff005b3a0821fd6673bd441f79ffc97e81b;p=idzebra-moved-to-github.git diff --git a/recctrl/recgrs.c b/recctrl/recgrs.c index 568774f..49fdf91 100644 --- a/recctrl/recgrs.c +++ b/recctrl/recgrs.c @@ -1,4 +1,4 @@ -/* $Id: recgrs.c,v 1.78 2003-04-25 08:57:36 adam Exp $ +/* $Id: recgrs.c,v 1.85 2003-10-07 09:18:21 adam Exp $ Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002,2003 Index Data Aps @@ -62,7 +62,8 @@ static int read_grs_type (struct grs_handlers *h, strcpy (p->type, cp+1); for (gh = h->handlers; gh; gh = gh->next) { - if (!memcmp (type, gh->type->type, cp-type)) + if (!memcmp (type, gh->type->type, cp-type) && + gh->type->type[cp-type] == '\0') { if (!gh->initFlag) { @@ -99,6 +100,7 @@ static void *grs_init(RecType recType) grs_add_handler (h, recTypeGrs_tcl); #endif grs_add_handler (h, recTypeGrs_marc); + grs_add_handler (h, recTypeGrs_marcxml); #if HAVE_EXPAT_H grs_add_handler (h, recTypeGrs_xml); #endif @@ -135,26 +137,26 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) if (p->which == XPATH_PREDICATE_RELATION) { if (p->u.relation.name[0]) { if (*p->u.relation.name != '@') { - logf(LOG_WARN, + yaz_log(LOG_WARN, " Only attributes (@) are supported in xelm xpath predicates"); - logf(LOG_WARN, "predicate %s ignored", p->u.relation.name); + yaz_log(LOG_WARN, "predicate %s ignored", p->u.relation.name); return (1); } attname = p->u.relation.name + 1; res = 0; /* looking for the attribute with a specified name */ for (attr = n->u.tag.attributes; attr; attr = attr->next) { - logf(LOG_DEBUG," - attribute %s <-> %s", attname, attr->name ); + yaz_log(LOG_DEBUG," - attribute %s <-> %s", attname, attr->name ); if (!strcmp(attr->name, attname)) { if (p->u.relation.op[0]) { if (*p->u.relation.op != '=') { - logf(LOG_WARN, + yaz_log(LOG_WARN, "Only '=' relation is supported (%s)",p->u.relation.op); - logf(LOG_WARN, "predicate %s ignored", p->u.relation.name); + yaz_log(LOG_WARN, "predicate %s ignored", p->u.relation.name); res = 1; break; } else { - logf(LOG_DEBUG," - value %s <-> %s", + yaz_log(LOG_DEBUG," - value %s <-> %s", p->u.relation.value, attr->value ); if (!strcmp(attr->value, p->u.relation.value)) { res = 1; break; @@ -166,6 +168,7 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) } } } + yaz_log(LOG_DEBUG, "return %d", res); return res; } else { return 1; @@ -180,7 +183,7 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) return (d1_check_xpath_predicate(n, p->u.boolean.left) || d1_check_xpath_predicate(n, p->u.boolean.right)); } else { - logf(LOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op); + yaz_log(LOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op); return 1; } } @@ -219,7 +222,7 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) int ok = 0; sprintf (pexpr, "%s\n", tagpath); - logf(LOG_DEBUG,"Checking tagpath %s",tagpath); + yaz_log(LOG_DEBUG,"Checking tagpath %s",tagpath); while (xpe) { struct DFA_state **dfaar = xpe->dfa->states; @@ -230,8 +233,6 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) unsigned char c; int start_line = 1; - logf(LOG_DEBUG," - xpath %s",xpe->xpath_expr); - c = *pexpr++; t = s->trans; i = s->tran_no; if ((c >= t->ch[0] && c <= t->ch[1]) || (!t->ch[0])) { p = pexpr; @@ -248,6 +249,11 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) p++; } while (i >= 0); } + if (ok) + yaz_log(LOG_DEBUG," xpath match %s",xpe->xpath_expr); + else + yaz_log(LOG_DEBUG," xpath no match %s",xpe->xpath_expr); + pexpr--; if (ok) { #ifdef ENHANCED_XELM @@ -263,11 +269,11 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) backwards trough xpath location steps ... */ for (i=xpe->xpath_len - 1; i>0; i--) { - logf(LOG_DEBUG,"Checking step %d: %s on tag %s", + yaz_log(LOG_DEBUG,"Checking step %d: %s on tag %s", i,xp[i].part,nn->u.tag.tag); if (!d1_check_xpath_predicate(nn, xp[i].predicate)) { - logf(LOG_DEBUG," Predicates didn't match"); + yaz_log(LOG_DEBUG," Predicates didn't match"); ok = 0; break; } @@ -287,7 +293,7 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) xfree(pexpr); if (ok) { - logf(LOG_DEBUG,"Got it"); + yaz_log(LOG_DEBUG,"Got it"); return xpe->termlists; } else { return NULL; @@ -349,6 +355,7 @@ static void index_xpath (data1_node *n, struct recExtractCtrl *p, data1_node *nn; int termlist_only = 1; + yaz_log(LOG_DEBUG, "index_xpath level=%d use=%d", level, use); if ((!n->root->u.root.absyn) || (n->root->u.root.absyn->enable_xpath_indexing)) { termlist_only = 0; @@ -361,18 +368,16 @@ static void index_xpath (data1_node *n, struct recExtractCtrl *p, wrd->length = n->u.data.len; if (p->flagShowRecords) { - printf("%*s data=", (level + 1) * 4, ""); + printf("%*s XData:\"", (level + 1) * 4, ""); for (i = 0; ilength && i < 8; i++) fputc (wrd->string[i], stdout); - printf("\n"); + printf("\"\n"); } else { data1_termlist *tl; int xpdone = 0; flen = 0; - // logf (LOG_LOG,"Index value, %s",wrd->string); - /* we have to fetch the whole path to the data tag */ for (nn = n; nn; nn = nn->parent) { if (nn->which == DATA1N_tag) { @@ -388,7 +393,7 @@ static void index_xpath (data1_node *n, struct recExtractCtrl *p, tag_path_full[flen] = 0; /* If we have a matching termlist... */ - if ((tl = xpath_termlist_by_tagpath(tag_path_full, n))) { + if (n->root->u.root.absyn && (tl = xpath_termlist_by_tagpath(tag_path_full, n))) { for (; tl; tl = tl->next) { wrd->reg_type = *tl->structure; /* this is the ! case, so structure is for the xpath index */ @@ -397,8 +402,8 @@ static void index_xpath (data1_node *n, struct recExtractCtrl *p, wrd->attrUse = use; (*p->tokenAdd)(wrd); xpdone = 1; - /* this is just the old fashioned attribute based index */ } else { + /* this is just the old fashioned attribute based index */ wrd->attrSet = (int) (tl->att->parent->reference); wrd->attrUse = tl->att->locals->local; (*p->tokenAdd)(wrd); @@ -452,136 +457,134 @@ static void index_xpath (data1_node *n, struct recExtractCtrl *p, data1_xattr *xp; data1_termlist *tl; int do_xpindex; - + tag_path_full[flen] = 0; - + /* Add tag start/end xpath index, only when there is a ! in the apropriate xelm - directive, or default xpath indexing is enabled */ + directive, or default xpath indexing is enabled */ if (!(do_xpindex = 1 - termlist_only)) { - if ((tl = xpath_termlist_by_tagpath(tag_path_full, n))) { - for (; tl; tl = tl->next) { if (!tl->att) {do_xpindex = 1;} } - } + if ((tl = xpath_termlist_by_tagpath(tag_path_full, n))) { + for (; tl; tl = tl->next) { if (!tl->att) {do_xpindex = 1;} } + } } if (do_xpindex) { - (*p->tokenAdd)(wrd); /* index element pag (AKA tag path) */ + (*p->tokenAdd)(wrd); /* index element pag (AKA tag path) */ } - + if (use == 1) /* only for the starting tag... */ { - #define MAX_ATTR_COUNT 50 - - data1_termlist *tll[MAX_ATTR_COUNT]; - - int i = 0; - - /* get termlists for attributes, and find out, if we have to do xpath indexing */ - for (xp = n->u.tag.attributes; xp; xp = xp->next) { - i++; - } - - i = 0; - for (xp = n->u.tag.attributes; xp; xp = xp->next) { - char comb[512]; - int do_xpindex = 1 - termlist_only; - data1_termlist *tl; - char attr_tag_path_full[1024]; - int int_len = flen; - - /* this could be cached as well */ - sprintf (attr_tag_path_full, "@%s/%.*s", - xp->name, int_len, tag_path_full); - - tll[i] = xpath_termlist_by_tagpath(attr_tag_path_full,n); - - /* if there is a ! in the xelm termlist, or default indexing is on, - proceed with xpath idx */ - if ((tl = tll[i])) - { - for (; tl; tl = tl->next) - { - if (!tl->att) - do_xpindex = 1; - } - } - - if (do_xpindex) { - - /* attribute (no value) */ - wrd->reg_type = '0'; - wrd->attrUse = 3; - wrd->string = xp->name; - wrd->length = strlen(xp->name); - - wrd->seqno--; - (*p->tokenAdd)(wrd); - - if (xp->value && - strlen(xp->name) + strlen(xp->value) < sizeof(comb)-2) { - - /* attribute value exact */ - strcpy (comb, xp->name); - strcat (comb, "="); - strcat (comb, xp->value); - - wrd->attrUse = 3; - wrd->reg_type = '0'; - wrd->string = comb; - wrd->length = strlen(comb); - wrd->seqno--; - - (*p->tokenAdd)(wrd); - } - } - i++; - } - - i = 0; - for (xp = n->u.tag.attributes; xp; xp = xp->next) { - data1_termlist *tl; - char attr_tag_path_full[1024]; - int int_len = flen; - int xpdone = 0; - - sprintf (attr_tag_path_full, "@%s/%.*s", - xp->name, int_len, tag_path_full); - - if ((tl = tll[i])) - { - /* If there is a termlist given (=xelm directive) */ - for (; tl; tl = tl->next) - { - if (!tl->att) { - /* add xpath index for the attribute */ - index_xpath_attr (attr_tag_path_full, xp->name, - xp->value, tl->structure, - p, wrd); - xpdone = 1; - } else { - /* add attribute based index for the attribute */ - if (xp->value) { - wrd->attrSet = (int) - (tl->att->parent->reference); - wrd->attrUse = tl->att->locals->local; - wrd->reg_type = *tl->structure; - wrd->string = xp->value; - wrd->length = strlen(xp->value); - (*p->tokenAdd)(wrd); - } - } - } - } - /* if there was no termlist for the given path, - or the termlist didn't have a ! element, index - the attribute as "w" */ - if ((!xpdone) && (!termlist_only)) - { - index_xpath_attr (attr_tag_path_full, xp->name, - xp->value, "w", p, wrd); - } - i++; - } - } + data1_termlist *tll[MAX_ATTR_COUNT]; + + int i = 0; + + /* get termlists for attributes, and find out, if we have to do xpath indexing */ + for (xp = n->u.tag.attributes; xp; xp = xp->next) { + i++; + } + + i = 0; + for (xp = n->u.tag.attributes; xp; xp = xp->next) { + char comb[512]; + int do_xpindex = 1 - termlist_only; + data1_termlist *tl; + char attr_tag_path_full[1024]; + int int_len = flen; + + /* this could be cached as well */ + sprintf (attr_tag_path_full, "@%s/%.*s", + xp->name, int_len, tag_path_full); + + tll[i] = xpath_termlist_by_tagpath(attr_tag_path_full,n); + + /* if there is a ! in the xelm termlist, or default indexing is on, + proceed with xpath idx */ + if ((tl = tll[i])) + { + for (; tl; tl = tl->next) + { + if (!tl->att) + do_xpindex = 1; + } + } + + if (do_xpindex) { + + /* attribute (no value) */ + wrd->reg_type = '0'; + wrd->attrUse = 3; + wrd->string = xp->name; + wrd->length = strlen(xp->name); + + wrd->seqno--; + (*p->tokenAdd)(wrd); + + if (xp->value && + strlen(xp->name) + strlen(xp->value) < sizeof(comb)-2) { + + /* attribute value exact */ + strcpy (comb, xp->name); + strcat (comb, "="); + strcat (comb, xp->value); + + wrd->attrUse = 3; + wrd->reg_type = '0'; + wrd->string = comb; + wrd->length = strlen(comb); + wrd->seqno--; + + (*p->tokenAdd)(wrd); + } + } + i++; + } + + i = 0; + for (xp = n->u.tag.attributes; xp; xp = xp->next) { + data1_termlist *tl; + char attr_tag_path_full[1024]; + int int_len = flen; + int xpdone = 0; + + sprintf (attr_tag_path_full, "@%s/%.*s", + xp->name, int_len, tag_path_full); + + if ((tl = tll[i])) + { + /* If there is a termlist given (=xelm directive) */ + for (; tl; tl = tl->next) + { + if (!tl->att) { + /* add xpath index for the attribute */ + index_xpath_attr (attr_tag_path_full, xp->name, + xp->value, tl->structure, + p, wrd); + xpdone = 1; + } else { + /* add attribute based index for the attribute */ + if (xp->value) { + wrd->attrSet = (int) + (tl->att->parent->reference); + wrd->attrUse = tl->att->locals->local; + wrd->reg_type = *tl->structure; + wrd->string = xp->value; + wrd->length = strlen(xp->value); + (*p->tokenAdd)(wrd); + } + } + } + } + /* if there was no termlist for the given path, + or the termlist didn't have a ! element, index + the attribute as "w" */ + if ((!xpdone) && (!termlist_only)) + { + index_xpath_attr (attr_tag_path_full, xp->name, + xp->value, "w", p, wrd); + } + i++; + } + } } } } @@ -646,7 +649,7 @@ static void index_termlist (data1_node *par, data1_node *n, tlist->att->parent->name, tlist->att->name, tlist->att->value, tlist->source); - printf (" data=\""); + printf (" XData:\""); for (i = 0; ilength && i < 40; i++) fputc (wrd->string[i], stdout); fputc ('"', stdout); @@ -711,8 +714,8 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level, { index_termlist (n, n, p, level, wrd); /* index start tag */ - assert (n->root->u.root.absyn); - index_xpath (n, p, level, wrd, 1); + if (n->root->u.root.absyn) + index_xpath (n, p, level, wrd, 1); } if (n->child) @@ -729,8 +732,8 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level, printf("%*s", level * 4, ""); printf("Data: "); if (n->u.data.len > 256) - printf("'%.240s ... %.6s'\n", n->u.data.data, - n->u.data.data + n->u.data.len-6); + printf("'%.170s ... %.70s'\n", n->u.data.data, + n->u.data.data + n->u.data.len-70); else if (n->u.data.len > 0) printf("'%.*s'\n", n->u.data.len, n->u.data.data); else @@ -811,6 +814,7 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p, if ((oid_ent_to_oid (&oe, oidtmp))) (*p->schemaAdd)(p, oidtmp); } + data1_concat_text(p->dh, mem, n); /* ensure our data1 tree is UTF-8 */ data1_iconv (p->dh, mem, n, "UTF-8", data1_get_encoding(p->dh, n)); @@ -857,10 +861,10 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c) if (!(eset = data1_getesetbyname(dh, n->u.root.absyn, c->u.simple->u.generic))) { - logf(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic); + yaz_log(LOG_LOG, "Unknown esetname '%s'", c->u.simple->u.generic); return 25; /* invalid esetname */ } - logf(LOG_DEBUG, "Esetname '%s' in simple compspec", + yaz_log(LOG_DEBUG, "Esetname '%s' in simple compspec", c->u.simple->u.generic); espec = eset->spec; break; @@ -877,23 +881,23 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c) data1_getesetbyname(dh, n->u.root.absyn, p->u.elementSetName))) { - logf(LOG_LOG, "Unknown esetname '%s'", + yaz_log(LOG_LOG, "Unknown esetname '%s'", p->u.elementSetName); return 25; /* invalid esetname */ } - logf(LOG_DEBUG, "Esetname '%s' in complex compspec", + yaz_log(LOG_DEBUG, "Esetname '%s' in complex compspec", p->u.elementSetName); espec = eset->spec; break; case Z_ElementSpec_externalSpec: if (p->u.externalSpec->which == Z_External_espec1) { - logf(LOG_DEBUG, "Got Espec-1"); + yaz_log(LOG_DEBUG, "Got Espec-1"); espec = p->u.externalSpec-> u.espec1; } else { - logf(LOG_LOG, "Unknown external espec."); + yaz_log(LOG_LOG, "Unknown external espec."); return 25; /* bad. what is proper diagnostic? */ } break; @@ -905,12 +909,12 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c) } if (espec) { - logf (LOG_DEBUG, "Element: Espec-1 match"); + yaz_log(LOG_DEBUG, "Element: Espec-1 match"); return data1_doespec1(dh, n, espec); } else { - logf (LOG_DEBUG, "Element: all match"); + yaz_log(LOG_DEBUG, "Element: all match"); return -1; } } @@ -987,7 +991,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) gri.mem = mem; gri.dh = p->dh; - logf (LOG_DEBUG, "grs_retrieve"); + yaz_log(LOG_DEBUG, "grs_retrieve"); if (read_grs_type (h, &gri, p->subType, &node)) { p->diagnostic = 14; @@ -1000,6 +1004,8 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) nmem_destroy (mem); return 0; } + data1_concat_text(p->dh, mem, node); + /* ensure our data1 tree is UTF-8 */ data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); @@ -1008,7 +1014,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) #endif top = data1_get_root_tag (p->dh, node); - logf (LOG_DEBUG, "grs_retrieve: size"); + yaz_log(LOG_DEBUG, "grs_retrieve: size"); tagname = data1_systag_lookup(node->u.root.absyn, "size", "size"); if (tagname && (dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem))) @@ -1023,7 +1029,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) if (tagname && p->score >= 0 && (dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem))) { - logf (LOG_DEBUG, "grs_retrieve: %s", tagname); + yaz_log(LOG_DEBUG, "grs_retrieve: %s", tagname); dnew->u.data.what = DATA1I_num; dnew->u.data.data = dnew->lbuf; sprintf(dnew->u.data.data, "%d", p->score); @@ -1035,7 +1041,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) if (tagname && p->localno > 0 && (dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem))) { - logf (LOG_DEBUG, "grs_retrieve: %s", tagname); + yaz_log(LOG_DEBUG, "grs_retrieve: %s", tagname); dnew->u.data.what = DATA1I_text; dnew->u.data.data = dnew->lbuf; @@ -1070,7 +1076,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) */ if (requested_schema != VAL_NONE) { - logf (LOG_DEBUG, "grs_retrieve: schema mapping"); + yaz_log(LOG_DEBUG, "grs_retrieve: schema mapping"); for (map = node->u.root.absyn->maptabs; map; map = map->next) { if (map->target_absyn_ref == requested_schema) @@ -1098,7 +1104,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) * the overlap of schema and formatting which is inherent in the MARC * family) */ - yaz_log (LOG_DEBUG, "grs_retrieve: syntax mapping"); + yaz_log(LOG_DEBUG, "grs_retrieve: syntax mapping"); if (node->u.root.absyn) for (map = node->u.root.absyn->maptabs; map; map = map->next) { @@ -1114,7 +1120,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) break; } } - yaz_log (LOG_DEBUG, "grs_retrieve: schemaIdentifier"); + yaz_log(LOG_DEBUG, "grs_retrieve: schemaIdentifier"); if (node->u.root.absyn && node->u.root.absyn->reference != VAL_NONE && p->input_format == VAL_GRS1) @@ -1152,7 +1158,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) } } - logf (LOG_DEBUG, "grs_retrieve: element spec"); + yaz_log(LOG_DEBUG, "grs_retrieve: element spec"); if (p->comp && (res = process_comp(p->dh, node, p->comp)) > 0) { p->diagnostic = res; @@ -1168,7 +1174,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) #if 0 data1_pr_tree (p->dh, node, stdout); #endif - logf (LOG_DEBUG, "grs_retrieve: transfer syntax mapping"); + yaz_log(LOG_DEBUG, "grs_retrieve: transfer syntax mapping"); switch (p->output_format = (p->input_format != VAL_NONE ? p->input_format : VAL_SUTRS)) {