X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=recctrl%2Frecgrs.c;h=0e44cf2c6a33e300e59b079f1227561768bd6d4f;hb=c8504e6a297055834322bd89b52e8410f549f337;hp=7767bdf96bc0baffd557edbc6bed8b2f21f82c17;hpb=47ea1fc957c7b97bb30a26698f072109cae275e4;p=idzebra-moved-to-github.git diff --git a/recctrl/recgrs.c b/recctrl/recgrs.c index 7767bdf..0e44cf2 100644 --- a/recctrl/recgrs.c +++ b/recctrl/recgrs.c @@ -1,9 +1,24 @@ -/* - * Copyright (C) 1994-2002, Index Data - * All rights reserved. - * - * $Id: recgrs.c,v 1.55 2002-07-25 13:06:44 adam Exp $ - */ +/* $Id: recgrs.c,v 1.69 2002-11-15 21:57:41 adam Exp $ + Copyright (C) 1995,1996,1997,1998,1999,2000,2001,2002 + Index Data Aps + +This file is part of the Zebra server. + +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 +Software Foundation; either version 2, or (at your option) any later +version. + +Zebra is distributed in the hope that it will be useful, but WITHOUT ANY +WARRANTY; without even the implied warranty of MERCHANTABILITY or +FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License +for more details. + +You should have received a copy of the GNU General Public License +along with Zebra; see the file LICENSE.zebra. If not, write to the +Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA +02111-1307, USA. +*/ #include #include @@ -84,9 +99,12 @@ static void *grs_init(RecType recType) grs_add_handler (h, recTypeGrs_tcl); #endif grs_add_handler (h, recTypeGrs_marc); -#if YAZ_HAVE_EXPAT +#if HAVE_EXPAT_H grs_add_handler (h, recTypeGrs_xml); #endif +#if HAVE_PERL + grs_add_handler (h, recTypeGrs_perl); +#endif return h; } @@ -105,6 +123,16 @@ static void grs_destroy(void *clientData) xfree (h); } +/* use + 1 start element (tag) + 2 end element + 3 start attr (and attr-exact) + 4 end attr + + 1016 cdata + 1015 attr data +*/ + static void index_xpath (data1_node *n, struct recExtractCtrl *p, int level, RecWord *wrd, int use) { @@ -164,9 +192,70 @@ static void index_xpath (data1_node *n, struct recExtractCtrl *p, } else { - (*p->tokenAdd)(wrd); + data1_xattr *xp; + (*p->tokenAdd)(wrd); /* index element pag (AKA tag path) */ + if (use == 1) + { + for (xp = n->u.tag.attributes; xp; xp = xp->next) + { + char comb[512]; + /* 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); + } + } + for (xp = n->u.tag.attributes; xp; xp = xp->next) + { + char attr_tag_path_full[1024]; + int int_len = flen; + + sprintf (attr_tag_path_full, "@%s/%.*s", + xp->name, int_len, tag_path_full); + + wrd->reg_type = '0'; + wrd->attrUse = 1; + wrd->string = attr_tag_path_full; + wrd->length = strlen(attr_tag_path_full); + (*p->tokenAdd)(wrd); + + if (xp->value) + { + wrd->attrUse = 1015; + wrd->reg_type = 'w'; + wrd->string = xp->value; + wrd->length = strlen(xp->value); + (*p->tokenAdd)(wrd); + } + + wrd->reg_type = '0'; + wrd->attrUse = 2; + wrd->string = attr_tag_path_full; + wrd->length = strlen(attr_tag_path_full); + (*p->tokenAdd)(wrd); + } + } } - break; } } @@ -293,8 +382,12 @@ 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); + if (!n->root->u.root.absyn) index_xpath (n, p, level, wrd, 1); + else if (n->root->u.root.absyn->enable_xpath_indexing) + index_xpath (n, p, level, wrd, 1); } if (n->child) @@ -310,8 +403,8 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level, { printf("%*s", level * 4, ""); printf("Data: "); - if (n->u.data.len > 32) - printf("'%.24s ... %.6s'\n", n->u.data.data, + if (n->u.data.len > 256) + printf("'%.240s ... %.6s'\n", n->u.data.data, n->u.data.data + n->u.data.len-6); else if (n->u.data.len > 0) printf("'%.*s'\n", n->u.data.len, n->u.data.data); @@ -323,7 +416,8 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level, index_termlist (par, n, p, level, wrd); if (!n->root->u.root.absyn) index_xpath (n, p, level, wrd, 1016); - + else if (n->root->u.root.absyn->enable_xpath_indexing) + index_xpath (n, p, level, wrd, 1016); } if (n->which == DATA1N_tag) @@ -331,9 +425,10 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level, /* index end tag */ if (!n->root->u.root.absyn) index_xpath (n, p, level, wrd, 2); + else if (n->root->u.root.absyn->enable_xpath_indexing) + index_xpath (n, p, level, wrd, 2); } - if (p->flagShowRecords && n->which == DATA1N_root) { printf("%*s-------------\n\n", level * 4, ""); @@ -381,7 +476,7 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p, gri.dh = p->dh; if (read_grs_type (h, &gri, p->subType, &n)) - return RECCTRL_EXTRACT_ERROR; + return RECCTRL_EXTRACT_ERROR_NO_SUCH_FILTER; if (!n) return RECCTRL_EXTRACT_EOF; oe.proto = PROTO_Z3950; @@ -408,7 +503,7 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p, if (dumpkeys(n, p, 0, &wrd) < 0) { data1_free_tree(p->dh, n); - return RECCTRL_EXTRACT_ERROR; + return RECCTRL_EXTRACT_ERROR_GENERIC; } data1_free_tree(p->dh, n); return RECCTRL_EXTRACT_OK; @@ -500,28 +595,52 @@ static int process_comp(data1_handle dh, data1_node *n, Z_RecordComposition *c) } } -static void add_idzebra_info (struct recRetrieveCtrl *p, data1_node *top, - NMEM mem) +/* Add Zebra info in separate namespace ... + + 359 + 447 + records/genera.xml + + +*/ + +static void zebra_xml_metadata (struct recRetrieveCtrl *p, data1_node *top, + NMEM mem) { - const char *idzebra_ns[7]; + const char *idzebra_ns[3]; + const char *i2 = "\n "; + const char *i4 = "\n "; + data1_node *n; - idzebra_ns[0] = "xmlns:idzebra"; + idzebra_ns[0] = "xmlns"; idzebra_ns[1] = "http://www.indexdata.dk/zebra/"; idzebra_ns[2] = 0; - data1_tag_add_attr (p->dh, mem, top, idzebra_ns); + data1_mk_text (p->dh, mem, i2, top); - data1_mk_tag_data_int (p->dh, top, "idzebra:size", p->recordSize, - mem); - if (p->score != -1) - data1_mk_tag_data_int (p->dh, top, "idzebra:score", - p->score, mem); + 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, i4, n); - data1_mk_tag_data_int (p->dh, top, "idzebra:localnumber", p->localno, - 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, "localnumber", p->localno, mem); if (p->fname) - data1_mk_tag_data_text(p->dh, top, "idzebra:filename", - p->fname, mem); + { + 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); } static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) @@ -688,9 +807,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) sprintf(p, "%d", *ii); p += strlen(p); } - *(p++) = '\0'; - - if ((dnew = data1_mk_tag_data_wd(dh, node, + if ((dnew = data1_mk_tag_data_wd(dh, top, "schemaIdentifier", mem))) { dnew->u.data.what = DATA1I_oid; @@ -714,7 +831,7 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) else if (p->comp && !res) selected = 1; -#if 1 +#if 0 data1_pr_tree (p->dh, node, stdout); #endif logf (LOG_DEBUG, "grs_retrieve: transfer syntax mapping"); @@ -722,7 +839,11 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p) p->input_format : VAL_SUTRS)) { case VAL_TEXT_XML: - add_idzebra_info (p, top, mem); + zebra_xml_metadata (p, top, mem); + +#if 0 + data1_pr_tree (p->dh, node, stdout); +#endif if (p->encoding) data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");