X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=index%2Frecgrs.c;h=386e5412666ace4ac0d74b44ee9a894ddd4d06d9;hp=4d621efddbfad14eac5cd30373a21f17584d0b06;hb=c3ff843e467932c6027a8b3b2ebda7b44612447e;hpb=f5fb1bd0a0df8a28471285cdf622897f4c2391b3 diff --git a/index/recgrs.c b/index/recgrs.c index 4d621ef..386e541 100644 --- a/index/recgrs.c +++ b/index/recgrs.c @@ -1,8 +1,5 @@ -/* $Id: recgrs.c,v 1.7 2006-09-29 10:02:47 adam Exp $ - Copyright (C) 1995-2006 - Index Data ApS - -This file is part of the Zebra server. +/* This file is part of the Zebra server. + Copyright (C) Index Data Zebra is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free @@ -20,13 +17,17 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */ +#if HAVE_CONFIG_H +#include +#endif #include #include #include #include #include -#include +#include +#include #include #include @@ -69,30 +70,30 @@ static int sp_range(struct source_parser *sp, data1_node *n, RecWord *wrd) { int start, len; RecWord tmp_w; - + /* ( */ sp_lex(sp); if (sp->lookahead != '(') return 0; sp_lex(sp); /* skip ( */ - + /* 1st arg: string */ if (!sp_expr(sp, n, wrd)) return 0; - + if (sp->lookahead != ',') - return 0; + return 0; sp_lex(sp); /* skip , */ - + /* 2nd arg: start */ if (!sp_expr(sp, n, &tmp_w)) return 0; start = atoi_n(tmp_w.term_buf, tmp_w.term_len); - + if (sp->lookahead == ',') { sp_lex(sp); /* skip , */ - + /* 3rd arg: length */ if (!sp_expr(sp, n, &tmp_w)) return 0; @@ -100,18 +101,24 @@ static int sp_range(struct source_parser *sp, data1_node *n, RecWord *wrd) } else len = wrd->term_len; - + /* ) */ if (sp->lookahead != ')') - return 0; + return 0; sp_lex(sp); - - if (wrd->term_buf && wrd->term_len) + + if (wrd->term_buf) { - wrd->term_buf += start; - wrd->term_len -= start; - if (wrd->term_len > len) - wrd->term_len = len; + if (start >= wrd->term_len) + wrd->term_len = 0; + else + { + wrd->term_len -= start; + wrd->term_buf += start; + + if (wrd->term_len > len) + wrd->term_len = len; + } } return 1; } @@ -131,7 +138,7 @@ static int sp_first(struct source_parser *sp, data1_node *n, RecWord *wrd) RecWord search_w; int i; sp_lex(sp); /* skip , */ - + if (!sp_expr(sp, n, &search_w)) return 0; for (i = 0; iterm_len; i++) @@ -175,7 +182,7 @@ static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd) else if (sp->len == 3 && !memcmp(sp->tok, "tag", sp->len)) { if (n->which == DATA1N_tag) - { + { wrd->term_buf = n->u.tag.tag; wrd->term_len = strlen(n->u.tag.tag); } @@ -191,13 +198,13 @@ static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd) if (!sp_expr(sp, n, &tmp_w)) return 0; - + wrd->term_buf = ""; wrd->term_len = 0; if (n->which == DATA1N_tag) { data1_xattr *p = n->u.tag.attributes; - while (p && strlen(p->name) != tmp_w.term_len && + while (p && strlen(p->name) != tmp_w.term_len && memcmp (p->name, tmp_w.term_buf, tmp_w.term_len)) p = p->next; if (p) @@ -236,7 +243,7 @@ static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd) wrd->term_buf = b; sp_lex(sp); } - else + else { wrd->term_buf = ""; wrd->term_len = 0; @@ -245,7 +252,7 @@ static int sp_expr(struct source_parser *sp, data1_node *n, RecWord *wrd) return 1; } -static struct source_parser *source_parser_create() +static struct source_parser *source_parser_create(void) { struct source_parser *sp = xmalloc(sizeof(*sp)); @@ -260,8 +267,8 @@ static void source_parser_destroy(struct source_parser *sp) nmem_destroy(sp->nmem); xfree(sp); } - -static int sp_parse(struct source_parser *sp, + +static int sp_parse(struct source_parser *sp, data1_node *n, RecWord *wrd, const char *src) { sp->len = 0; @@ -279,14 +286,14 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) int res = 1; char *attname; data1_xattr *attr; - + if (!p) { return 1; } else { if (p->which == XPATH_PREDICATE_RELATION) { if (p->u.relation.name[0]) { if (*p->u.relation.name != '@') { - yaz_log(YLOG_WARN, + yaz_log(YLOG_WARN, " Only attributes (@) are supported in xelm xpath predicates"); yaz_log(YLOG_WARN, "predicate %s ignored", p->u.relation.name); return 1; @@ -295,21 +302,17 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) res = 0; /* looking for the attribute with a specified name */ for (attr = n->u.tag.attributes; attr; attr = attr->next) { - yaz_log(YLOG_DEBUG," - attribute %s <-> %s", attname, attr->name ); - if (!strcmp(attr->name, attname)) { if (p->u.relation.op[0]) { if (*p->u.relation.op != '=') { - yaz_log(YLOG_WARN, + yaz_log(YLOG_WARN, "Only '=' relation is supported (%s)",p->u.relation.op); yaz_log(YLOG_WARN, "predicate %s ignored", p->u.relation.name); res = 1; break; } else { - yaz_log(YLOG_DEBUG," - value %s <-> %s", - p->u.relation.value, attr->value ); if (!strcmp(attr->value, p->u.relation.value)) { res = 1; break; - } + } } } else { /* attribute exists, no value specified */ @@ -317,20 +320,19 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) } } } - yaz_log(YLOG_DEBUG, "return %d", res); return res; } else { return 1; } - } + } else if (p->which == XPATH_PREDICATE_BOOLEAN) { if (!strcmp(p->u.boolean.op,"and")) { - return d1_check_xpath_predicate(n, p->u.boolean.left) - && d1_check_xpath_predicate(n, p->u.boolean.right); + return d1_check_xpath_predicate(n, p->u.boolean.left) + && d1_check_xpath_predicate(n, p->u.boolean.right); } else if (!strcmp(p->u.boolean.op,"or")) { - return (d1_check_xpath_predicate(n, p->u.boolean.left) - || d1_check_xpath_predicate(n, p->u.boolean.right)); + return (d1_check_xpath_predicate(n, p->u.boolean.left) + || d1_check_xpath_predicate(n, p->u.boolean.right)); } else { yaz_log(YLOG_WARN, "Unknown boolean relation %s, ignored",p->u.boolean.op); return 1; @@ -348,7 +350,7 @@ static int dfa_match_first(struct DFA_state **dfaar, const char *text) int i; const char *p = text; unsigned char c; - + for (c = *p++, t = s->trans, i = s->tran_no; --i >= 0; t++) { if (c >= t->ch[0] && c <= t->ch[1]) @@ -373,7 +375,7 @@ static int dfa_match_first(struct DFA_state **dfaar, const char *text) } /* *ostrich* - + New function, looking for xpath "element" definitions in abs, by tagpath, using a kind of ugly regxp search.The DFA was built while parsing abs, so here we just go trough them and try to match @@ -395,12 +397,12 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) data1_xpelement *xpe = 0; data1_node *nn; -#ifdef ENHANCED_XELM +#ifdef ENHANCED_XELM struct xpath_location_step *xp; #endif char *pexpr = xmalloc(strlen(tagpath)+5); - - sprintf (pexpr, "/%s\n", tagpath); + + sprintf(pexpr, "/%s\n", tagpath); for (xpe = abs->xp_elements; xpe; xpe = xpe->next) xpe->match_state = -1; /* don't know if it matches yet */ @@ -422,30 +424,26 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) xpe1->match_state = ok; #endif } - assert (ok == 0 || ok == 1); + assert(ok == 0 || ok == 1); if (ok) { -#ifdef ENHANCED_XELM +#ifdef ENHANCED_XELM /* we have to check the perdicates up to the root node */ xp = xpe->xpath; - + /* find the first tag up in the node structure */ for (nn = n; nn && nn->which != DATA1N_tag; nn = nn->parent) ; - + /* go from inside out in the node structure, while going backwards trough xpath location steps ... */ for (i = xpe->xpath_len - 1; i>0; i--) { - yaz_log(YLOG_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)) { - yaz_log(YLOG_DEBUG, " Predicates didn't match"); ok = 0; break; } - + if (nn->which == DATA1N_tag) nn = nn->parent; } @@ -453,12 +451,11 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) if (ok) break; } - } - + } + xfree(pexpr); - + if (xpe) { - yaz_log(YLOG_DEBUG, "Got it"); return xpe->termlists; } else { return NULL; @@ -477,32 +474,32 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) *ostrich* Now, if there is a matching xelm described in abs, for the - indexed element or the attribute, then the data is handled according + indexed element or the attribute, then the data is handled according to those definitions... modified by pop, 2002-12-13 */ /* add xpath index for an attribute */ -static void index_xpath_attr (char *tag_path, char *name, char *value, +static void index_xpath_attr(char *tag_path, char *name, char *value, char *structure, struct recExtractCtrl *p, RecWord *wrd) { wrd->index_name = ZEBRA_XPATH_ELM_BEGIN; - wrd->index_type = '0'; + wrd->index_type = "0"; wrd->term_buf = tag_path; wrd->term_len = strlen(tag_path); (*p->tokenAdd)(wrd); - + if (value) { wrd->index_name = ZEBRA_XPATH_ATTR_CDATA; - wrd->index_type = 'w'; + wrd->index_type = "w"; wrd->term_buf = value; wrd->term_len = strlen(value); (*p->tokenAdd)(wrd); } wrd->index_name = ZEBRA_XPATH_ELM_END; - wrd->index_type = '0'; + wrd->index_type = "0"; wrd->term_buf = tag_path; wrd->term_len = strlen(tag_path); (*p->tokenAdd)(wrd); @@ -522,7 +519,7 @@ static void mk_tag_path_full(char *tag_path_full, size_t max, data1_node *n) size_t tlen = strlen(nn->u.tag.tag); if (tlen + flen > (max - 2)) break; - memcpy (tag_path_full + flen, nn->u.tag.tag, tlen); + memcpy(tag_path_full + flen, nn->u.tag.tag, tlen); flen += tlen; tag_path_full[flen++] = '/'; } @@ -532,7 +529,7 @@ static void mk_tag_path_full(char *tag_path_full, size_t max, data1_node *n) } tag_path_full[flen] = 0; } - + static void index_xpath(struct source_parser *sp, data1_node *n, struct recExtractCtrl *p, @@ -545,87 +542,77 @@ static void index_xpath(struct source_parser *sp, data1_node *n, char tag_path_full[1024]; int termlist_only = 1; data1_termlist *tl; - int xpdone = 0; - if ((!n->root->u.root.absyn) || - (n->root->u.root.absyn->xpath_indexing == DATA1_XPATH_INDEXING_ENABLE)) { + + if (!n->root->u.root.absyn + || + n->root->u.root.absyn->xpath_indexing == DATA1_XPATH_INDEXING_ENABLE) + { termlist_only = 0; } + switch (n->which) { case DATA1N_data: wrd->term_buf = n->u.data.data; wrd->term_len = n->u.data.len; - xpdone = 0; mk_tag_path_full(tag_path_full, sizeof(tag_path_full), n); - + /* If we have a matching termlist... */ - if (n->root->u.root.absyn && + if (n->root->u.root.absyn && (tl = xpath_termlist_by_tagpath(tag_path_full, n))) { + zint max_seqno = 0; for (; tl; tl = tl->next) { /* need to copy recword because it may be changed */ RecWord wrd_tl; - wrd->index_type = *tl->structure; - memcpy (&wrd_tl, wrd, sizeof(*wrd)); + wrd->index_type = tl->structure; + memcpy(&wrd_tl, wrd, sizeof(*wrd)); if (tl->source) sp_parse(sp, n, &wrd_tl, tl->source); - if (!tl->index_name) - { - /* this is the ! case, so structure is for the xpath index */ - wrd_tl.index_name = xpath_index; - if (p->flagShowRecords) - { - int i; - printf("%*sXPath index", (level + 1) * 4, ""); - printf (" XData:\""); - for (i = 0; i 40) - printf (" ..."); - fputc ('\n', stdout); - } - else - (*p->tokenAdd)(&wrd_tl); - xpdone = 1; - } else { - /* this is just the old fashioned attribute based index */ - wrd_tl.index_name = tl->index_name; - if (p->flagShowRecords) - { - int i; - printf("%*sIdx: [%s]", (level + 1) * 4, "", - tl->structure); - printf("%s %s", tl->index_name, tl->source); - printf (" XData:\""); - for (i = 0; i 40) - printf (" ..."); - fputc ('\n', stdout); - } - else - (*p->tokenAdd)(&wrd_tl); - } + + /* this is just the old fashioned attribute based index */ + wrd_tl.index_name = tl->index_name; + if (p->flagShowRecords) + { + int i; + printf("%*sIdx: [%s]", (level + 1) * 4, "", + tl->structure); + printf("%s %s", tl->index_name, tl->source); + printf(" XData:\""); + for (i = 0; i 40) + printf(" ..."); + fputc('\n', stdout); + } + else + { + (*p->tokenAdd)(&wrd_tl); + } + if (wrd_tl.seqno > max_seqno) + max_seqno = wrd_tl.seqno; } + if (max_seqno) + wrd->seqno = max_seqno; + } - /* xpath indexing is done, if there was no termlist given, + /* xpath indexing is done, if there was no termlist given, or no ! in the termlist, and default indexing is enabled... */ - if (!p->flagShowRecords && !xpdone && !termlist_only) + if (!p->flagShowRecords && !termlist_only) { wrd->index_name = xpath_index; - wrd->index_type = 'w'; + wrd->index_type = "w"; (*p->tokenAdd)(wrd); } break; case DATA1N_tag: mk_tag_path_full(tag_path_full, sizeof(tag_path_full), n); - wrd->index_type = '0'; + wrd->index_type = "0"; wrd->term_buf = tag_path_full; wrd->term_len = strlen(tag_path_full); wrd->index_name = xpath_index; @@ -633,112 +620,74 @@ static void index_xpath(struct source_parser *sp, data1_node *n, { printf("%*s tag=", (level + 1) * 4, ""); for (i = 0; iterm_len && i < 40; i++) - fputc (wrd->term_buf[i], stdout); + fputc(wrd->term_buf[i], stdout); if (i == 40) - printf (" .."); + printf(" .."); printf("\n"); } else { data1_xattr *xp; - data1_termlist *tl; - int do_xpindex; - - /* Add tag start/end xpath index, only when there is a ! in - the apropriate xelm 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->index_name) - do_xpindex = 1; - } - } - } - if (do_xpindex) { + + if (!termlist_only) (*p->tokenAdd)(wrd); /* index element pag (AKA tag path) */ - } - + if (xpath_is_start == 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]; - + char attr_tag_path_full[1024]; + /* this could be cached as well */ - sprintf (attr_tag_path_full, "@%s/%s", + sprintf(attr_tag_path_full, "@%s/%s", xp->name, 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])) + + if (!termlist_only) { - for (; tl; tl = tl->next) - { - if (!tl->index_name) - do_xpindex = 1; - } - } - - if (do_xpindex) { - /* attribute (no value) */ - wrd->index_type = '0'; - wrd->index_name = ZEBRA_XPATH_ATTR_NAME; + wrd->index_type = "0"; + wrd->index_name = ZEBRA_XPATH_ATTR_NAME; wrd->term_buf = xp->name; wrd->term_len = strlen(xp->name); - + wrd->seqno--; (*p->tokenAdd)(wrd); - - if (xp->value && - strlen(xp->name) + strlen(xp->value) < sizeof(comb)-2) { - + + 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); + strcpy(comb, xp->name); + strcat(comb, "="); + strcat(comb, xp->value); wrd->index_name = ZEBRA_XPATH_ATTR_NAME; - wrd->index_type = '0'; + wrd->index_type = "0"; wrd->term_buf = comb; wrd->term_len = 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 xpdone = 0; - - sprintf (attr_tag_path_full, "@%s/%s", + + sprintf(attr_tag_path_full, "@%s/%s", xp->name, tag_path_full); - if ((tl = tll[i])) { /* If there is a termlist given (=xelm directive) */ @@ -747,16 +696,16 @@ static void index_xpath(struct source_parser *sp, data1_node *n, if (!tl->index_name) { /* add xpath index for the attribute */ - index_xpath_attr (attr_tag_path_full, xp->name, + index_xpath_attr(attr_tag_path_full, xp->name, xp->value, tl->structure, p, wrd); xpdone = 1; } else { /* index attribute value (only path/@attr) */ - if (xp->value) + if (xp->value) { wrd->index_name = tl->index_name; - wrd->index_type = *tl->structure; + wrd->index_type = tl->structure; wrd->term_buf = xp->value; wrd->term_len = strlen(xp->value); (*p->tokenAdd)(wrd); @@ -764,12 +713,12 @@ static void index_xpath(struct source_parser *sp, data1_node *n, } } } - /* if there was no termlist for the given path, - or the termlist didn't have a ! element, index + /* 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)) + if (!xpdone && !termlist_only) { - index_xpath_attr (attr_tag_path_full, xp->name, + index_xpath_attr(attr_tag_path_full, xp->name, xp->value, "w", p, wrd); } i++; @@ -779,27 +728,22 @@ static void index_xpath(struct source_parser *sp, data1_node *n, } } -static void index_termlist (struct source_parser *sp, data1_node *par, +static void index_termlist(struct source_parser *sp, data1_node *par, data1_node *n, struct recExtractCtrl *p, int level, RecWord *wrd) { data1_termlist *tlist = 0; - data1_datatype dtype = DATA1K_string; - /* * cycle up towards the root until we find a tag with an att.. * this has the effect of indexing locally defined tags with * the attribute of their ancestor in the record. */ - + while (!par->u.tag.element) if (!par->parent || !(par=get_parent_tag(p->dh, par->parent))) break; if (!par || !(tlist = par->u.tag.element->termlists)) return; - if (par->u.tag.element->tag) - dtype = par->u.tag.element->tag->kind; - for (; tlist; tlist = tlist->next) { /* consider source */ @@ -815,17 +759,17 @@ static void index_termlist (struct source_parser *sp, data1_node *par, printf("%*sIdx: [%s]", (level + 1) * 4, "", tlist->structure); printf("%s %s", tlist->index_name, tlist->source); - printf (" XData:\""); + printf(" XData:\""); for (i = 0; iterm_len && i < 40; i++) - fputc (wrd->term_buf[i], stdout); - fputc ('"', stdout); + fputc(wrd->term_buf[i], stdout); + fputc('"', stdout); if (wrd->term_len > 40) - printf (" ..."); - fputc ('\n', stdout); + printf(" ..."); + fputc('\n', stdout); } else { - wrd->index_type = *tlist->structure; + wrd->index_type = tlist->structure; wrd->index_name = tlist->index_name; (*p->tokenAdd)(wrd); } @@ -881,7 +825,7 @@ static int dumpkeys_r(struct source_parser *sp, index_termlist(sp, n, n, p, level, wrd); /* index start tag */ if (n->root->u.root.absyn) - index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_ELM_BEGIN, + index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_ELM_BEGIN, 1 /* is start */); } @@ -910,14 +854,14 @@ static int dumpkeys_r(struct source_parser *sp, if (par) index_termlist(sp, par, n, p, level, wrd); - index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_CDATA, + index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_CDATA, 0 /* is start */); } if (n->which == DATA1N_tag) { /* index end tag */ - index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_ELM_END, + index_xpath(sp, n, p, level, wrd, ZEBRA_XPATH_ELM_END, 0 /* is start */); } @@ -939,22 +883,13 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, RecWord *wrd) 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; - if (n->u.root.absyn) - { - oe.value = n->u.root.absyn->reference; - - if ((oid_ent_to_oid (&oe, oidtmp))) - (*p->schemaAdd)(p, oidtmp); - } + if (n->u.root.absyn && n->u.root.absyn->oid) + (*p->schemaAdd)(p, n->u.root.absyn->oid); (*p->init)(p, &wrd); - /* data1_pr_tree(p->dh, n, stdout); */ + /* data1_pr_tree(p->dh, n, stdout); */ return dumpkeys(n, p, &wrd); } @@ -965,8 +900,6 @@ static int grs_extract_sub(void *clientData, struct recExtractCtrl *p, { data1_node *n; struct grs_read_info gri; - oident oe; - int oidtmp[OID_SIZE]; RecWord wrd; gri.stream = p->stream; @@ -977,28 +910,18 @@ static int grs_extract_sub(void *clientData, struct recExtractCtrl *p, n = (*grs_read)(&gri); if (!n) return RECCTRL_EXTRACT_EOF; - oe.proto = PROTO_Z3950; - oe.oclass = CLASS_SCHEMA; -#if 0 - if (!n->u.root.absyn) - return RECCTRL_EXTRACT_ERROR; -#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 (n->u.root.absyn && n->u.root.absyn->oid) + (*p->schemaAdd)(p, n->u.root.absyn->oid); 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)); + data1_iconv(p->dh, mem, n, "UTF-8", data1_get_encoding(p->dh, n)); - data1_remove_idzebra_subtree (p->dh, n); + data1_remove_idzebra_subtree(p->dh, n); #if 0 - data1_pr_tree (p->dh, n, stdout); + data1_pr_tree(p->dh, n, stdout); #endif (*p->init)(p, &wrd); @@ -1013,7 +936,7 @@ int zebra_grs_extract(void *clientData, struct recExtractCtrl *p, data1_node *(*grs_read)(struct grs_read_info *)) { int ret; - NMEM mem = nmem_create (); + NMEM mem = nmem_create(); ret = grs_extract_sub(clientData, p, mem, grs_read); nmem_destroy(mem); return ret; @@ -1098,7 +1021,7 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c, } /* Add Zebra info in separate namespace ... - 359 @@ -1108,7 +1031,7 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c, */ -static void zebra_xml_metadata (struct recRetrieveCtrl *p, data1_node *top, +static void zebra_xml_metadata(struct recRetrieveCtrl *p, data1_node *top, NMEM mem) { const char *idzebra_ns[3]; @@ -1120,29 +1043,29 @@ static void zebra_xml_metadata (struct recRetrieveCtrl *p, data1_node *top, idzebra_ns[1] = "http://www.indexdata.dk/zebra/"; idzebra_ns[2] = 0; - data1_mk_text (p->dh, mem, i2, top); + data1_mk_text(p->dh, mem, i2, top); + + n = data1_mk_tag(p->dh, mem, "idzebra", idzebra_ns, top); - n = data1_mk_tag (p->dh, mem, "idzebra", idzebra_ns, top); + data1_mk_text(p->dh, mem, "\n", top); - data1_mk_text (p->dh, mem, "\n", top); + data1_mk_text(p->dh, mem, i4, n); - data1_mk_text (p->dh, mem, i4, n); - - data1_mk_tag_data_int (p->dh, n, "size", p->recordSize, mem); + data1_mk_tag_data_int(p->dh, n, "size", p->recordSize, mem); if (p->score != -1) { - data1_mk_text (p->dh, mem, i4, n); - data1_mk_tag_data_int (p->dh, n, "score", p->score, mem); + data1_mk_text(p->dh, mem, i4, n); + data1_mk_tag_data_int(p->dh, n, "score", p->score, mem); } - data1_mk_text (p->dh, mem, i4, n); - data1_mk_tag_data_zint (p->dh, n, "localnumber", p->localno, mem); + data1_mk_text(p->dh, mem, i4, n); + data1_mk_tag_data_zint(p->dh, n, "localnumber", p->localno, mem); if (p->fname) { - data1_mk_text (p->dh, mem, i4, n); + data1_mk_text(p->dh, mem, i4, n); data1_mk_tag_data_text(p->dh, n, "filename", p->fname, mem); } - data1_mk_text (p->dh, mem, i2, n); + data1_mk_text(p->dh, mem, i2, n); } int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, @@ -1156,10 +1079,10 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, struct grs_read_info gri; const char *tagname; - int requested_schema = VAL_NONE; + const Odr_oid *requested_schema = 0; data1_marctab *marctab; int dummy; - + mem = nmem_create(); gri.stream = p->stream; gri.mem = mem; @@ -1170,18 +1093,18 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, node = (*grs_read)(&gri); if (!node) { - p->diagnostic = 14; - nmem_destroy (mem); + p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; + nmem_destroy(mem); return 0; } data1_concat_text(p->dh, mem, node); - data1_remove_idzebra_subtree (p->dh, node); + data1_remove_idzebra_subtree(p->dh, node); #if 0 - data1_pr_tree (p->dh, node, stdout); + data1_pr_tree(p->dh, node, stdout); #endif - top = data1_get_root_tag (p->dh, node); + top = data1_get_root_tag(p->dh, node); yaz_log(YLOG_DEBUG, "grs_retrieve: size"); tagname = data1_systag_lookup(node->u.root.absyn, "size", "size"); @@ -1193,7 +1116,7 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, sprintf(dnew->u.data.data, "%d", p->recordSize); dnew->u.data.len = strlen(dnew->u.data.data); } - + tagname = data1_systag_lookup(node->u.root.absyn, "rank", "rank"); if (tagname && p->score >= 0 && (dnew = data1_mk_tag_data_wd(p->dh, top, tagname, mem))) @@ -1213,51 +1136,55 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, yaz_log(YLOG_DEBUG, "grs_retrieve: %s", tagname); dnew->u.data.what = DATA1I_text; dnew->u.data.data = dnew->lbuf; - + sprintf(dnew->u.data.data, ZINT_FORMAT, p->localno); dnew->u.data.len = strlen(dnew->u.data.data); } - if (p->input_format == VAL_TEXT_XML) - zebra_xml_metadata (p, top, mem); + if (!p->input_format) + { /* SUTRS is default input_format */ + p->input_format = yaz_oid_recsyn_sutrs; + } + assert(p->input_format); + + if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_xml)) + zebra_xml_metadata(p, top, mem); #if 0 - data1_pr_tree (p->dh, node, stdout); + data1_pr_tree(p->dh, node, stdout); #endif if (p->comp && p->comp->which == Z_RecordComp_complex && p->comp->u.complex->generic && p->comp->u.complex->generic->which == Z_Schema_oid && p->comp->u.complex->generic->schema.oid) { - oident *oe = oid_getentbyoid (p->comp->u.complex->generic->schema.oid); - if (oe) - requested_schema = oe->value; + requested_schema = p->comp->u.complex->generic->schema.oid; } /* If schema has been specified, map if possible, then check that - * we got the right one + * we got the right one */ - if (requested_schema != VAL_NONE) + if (requested_schema) { yaz_log(YLOG_DEBUG, "grs_retrieve: schema mapping"); for (map = node->u.root.absyn->maptabs; map; map = map->next) { - if (map->target_absyn_ref == requested_schema) + if (!oid_oidcmp(map->oid, requested_schema)) { onode = node; if (!(node = data1_map_record(p->dh, onode, map, mem))) { - p->diagnostic = 14; - nmem_destroy (mem); + p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; + nmem_destroy(mem); return 0; } break; } } - if (node->u.root.absyn && - requested_schema != node->u.root.absyn->reference) + if (node->u.root.absyn + && oid_oidcmp(requested_schema, node->u.root.absyn->oid)) { - p->diagnostic = 238; - nmem_destroy (mem); + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; + nmem_destroy(mem); return 0; } } @@ -1270,53 +1197,31 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, if (node->u.root.absyn) for (map = node->u.root.absyn->maptabs; map; map = map->next) { - if (map->target_absyn_ref == p->input_format) + if (!oid_oidcmp(map->oid, p->input_format)) { onode = node; if (!(node = data1_map_record(p->dh, onode, map, mem))) { - p->diagnostic = 14; - nmem_destroy (mem); + p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; + nmem_destroy(mem); return 0; } break; } } yaz_log(YLOG_DEBUG, "grs_retrieve: schemaIdentifier"); - if (node->u.root.absyn && - node->u.root.absyn->reference != VAL_NONE && - p->input_format == VAL_GRS1) + if (node->u.root.absyn && node->u.root.absyn->oid + && !oid_oidcmp(p->input_format, yaz_oid_recsyn_grs_1)) { - oident oe; - Odr_oid *oid; - int oidtmp[OID_SIZE]; - - oe.proto = PROTO_Z3950; - oe.oclass = CLASS_SCHEMA; - oe.value = node->u.root.absyn->reference; - - if ((oid = oid_ent_to_oid (&oe, oidtmp))) - { - char tmp[128]; - data1_handle dh = p->dh; - char *p = tmp; - int *ii; - - for (ii = oid; *ii >= 0; ii++) - { - if (p != tmp) - *(p++) = '.'; - sprintf(p, "%d", *ii); - p += strlen(p); - } - if ((dnew = data1_mk_tag_data_wd(dh, top, - "schemaIdentifier", mem))) - { - dnew->u.data.what = DATA1I_oid; - dnew->u.data.data = (char *) nmem_malloc(mem, p - tmp); - memcpy(dnew->u.data.data, tmp, p - tmp); - dnew->u.data.len = p - tmp; - } + char oid_str[OID_STR_MAX]; + char *dot_str = oid_oid_to_dotstring(node->u.root.absyn->oid, oid_str); + + if (dot_str && (dnew = data1_mk_tag_data_wd(p->dh, top, + "schemaIdentifier", mem))) + { + dnew->u.data.what = DATA1I_oid; + dnew->u.data.data = (char *) nmem_strdup(mem, dot_str); + dnew->u.data.len = strlen(dot_str); } } @@ -1332,114 +1237,122 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, selected = 1; #if 0 - data1_pr_tree (p->dh, node, stdout); + data1_pr_tree(p->dh, node, stdout); #endif yaz_log(YLOG_DEBUG, "grs_retrieve: transfer syntax mapping"); - switch (p->output_format = (p->input_format != VAL_NONE ? - p->input_format : VAL_SUTRS)) + + p->output_format = p->input_format; + + assert(p->input_format); + if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_xml)) { - case VAL_TEXT_XML: #if 0 - data1_pr_tree (p->dh, node, stdout); + data1_pr_tree(p->dh, node, stdout); #endif /* default output encoding for XML is UTF-8 */ - data1_iconv (p->dh, mem, node, + data1_iconv(p->dh, mem, node, p->encoding ? p->encoding : "UTF-8", data1_get_encoding(p->dh, node)); if (!(p->rec_buf = data1_nodetoidsgml(p->dh, node, selected, &p->rec_len))) - p->diagnostic = 238; + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; else { - char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); - memcpy (new_buf, p->rec_buf, p->rec_len); + char *new_buf = (char*) odr_malloc(p->odr, p->rec_len); + memcpy(new_buf, p->rec_buf, p->rec_len); p->rec_buf = new_buf; } - break; - case VAL_GRS1: - data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); + } + else if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_grs_1)) + { + data1_iconv(p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); dummy = 0; if (!(p->rec_buf = data1_nodetogr(p->dh, node, selected, p->odr, &dummy))) - p->diagnostic = 238; /* not available in requested syntax */ + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; else p->rec_len = -1; - break; - case VAL_EXPLAIN: + } + else if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_explain)) + { /* ensure our data1 tree is UTF-8 */ - data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); - + data1_iconv(p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); + if (!(p->rec_buf = data1_nodetoexplain(p->dh, node, selected, p->odr))) - p->diagnostic = 238; + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; else p->rec_len = -1; - break; - case VAL_SUMMARY: + } + else if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_summary)) + { /* ensure our data1 tree is UTF-8 */ - data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); + data1_iconv(p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node)); if (!(p->rec_buf = data1_nodetosummary(p->dh, node, selected, p->odr))) - p->diagnostic = 238; + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; else p->rec_len = -1; - break; - case VAL_SUTRS: + } + else if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_sutrs)) + { if (p->encoding) - data1_iconv (p->dh, mem, node, p->encoding, + data1_iconv(p->dh, mem, node, p->encoding, data1_get_encoding(p->dh, node)); if (!(p->rec_buf = data1_nodetobuf(p->dh, node, selected, &p->rec_len))) - p->diagnostic = 238; + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; else { - char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); - memcpy (new_buf, p->rec_buf, p->rec_len); + char *new_buf = (char*) odr_malloc(p->odr, p->rec_len); + memcpy(new_buf, p->rec_buf, p->rec_len); p->rec_buf = new_buf; } - break; - case VAL_SOIF: + } + else if (!oid_oidcmp(p->input_format, yaz_oid_recsyn_soif)) + { if (p->encoding) - data1_iconv (p->dh, mem, node, p->encoding, + data1_iconv(p->dh, mem, node, p->encoding, data1_get_encoding(p->dh, node)); if (!(p->rec_buf = data1_nodetosoif(p->dh, node, selected, &p->rec_len))) - p->diagnostic = 238; + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; else { - char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); - memcpy (new_buf, p->rec_buf, p->rec_len); + char *new_buf = (char*) odr_malloc(p->odr, p->rec_len); + memcpy(new_buf, p->rec_buf, p->rec_len); p->rec_buf = new_buf; } - break; - default: + } + else + { if (!node->u.root.absyn) - { - p->diagnostic = 238; - break; - } - for (marctab = node->u.root.absyn->marc; marctab; - marctab = marctab->next) - if (marctab->reference == p->input_format) - break; - if (!marctab) - { - p->diagnostic = 238; - break; - } - if (p->encoding) - data1_iconv (p->dh, mem, node, p->encoding, - data1_get_encoding(p->dh, node)); - if (!(p->rec_buf = data1_nodetomarc(p->dh, marctab, node, - selected, &p->rec_len))) - p->diagnostic = 238; - else - { - char *new_buf = (char*) odr_malloc (p->odr, p->rec_len); - memcpy (new_buf, p->rec_buf, p->rec_len); - p->rec_buf = new_buf; - } + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; + else + { + for (marctab = node->u.root.absyn->marc; marctab; + marctab = marctab->next) + if (marctab->oid && !oid_oidcmp(marctab->oid, p->input_format)) + break; + if (!marctab) + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; + else + { + if (p->encoding) + data1_iconv(p->dh, mem, node, p->encoding, + data1_get_encoding(p->dh, node)); + if (!(p->rec_buf = data1_nodetomarc(p->dh, marctab, node, + selected, &p->rec_len))) + p->diagnostic = YAZ_BIB1_RECORD_NOT_AVAILABLE_IN_REQUESTED_SYNTAX; + else + { + char *new_buf = (char*) odr_malloc(p->odr, p->rec_len); + memcpy(new_buf, p->rec_buf, p->rec_len); + p->rec_buf = new_buf; + } + } + } } nmem_destroy(mem); return 0; @@ -1448,6 +1361,7 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, /* * Local variables: * c-basic-offset: 4 + * c-file-style: "Stroustrup" * indent-tabs-mode: nil * End: * vim: shiftwidth=4 tabstop=8 expandtab