X-Git-Url: http://git.indexdata.com/?a=blobdiff_plain;f=index%2Frecgrs.c;h=3137760ddd9648d4bb898d03df9d78b18bf9bd29;hb=27a6d4f7f7425896345ab5a2bfdf35a96c97416e;hp=7adf1609992aca539627a2fd4b1ba103e2f172e4;hpb=7415d28c149c1bab51fe93aeaccdd14085b69bd9;p=idzebra-moved-to-github.git diff --git a/index/recgrs.c b/index/recgrs.c index 7adf160..3137760 100644 --- a/index/recgrs.c +++ b/index/recgrs.c @@ -1,5 +1,5 @@ -/* $Id: recgrs.c,v 1.1 2006-07-03 14:27:09 adam Exp $ - Copyright (C) 1995-2006 +/* $Id: recgrs.c,v 1.16 2007-03-07 21:08:36 adam Exp $ + Copyright (C) 1995-2007 Index Data ApS This file is part of the Zebra server. @@ -15,9 +15,9 @@ 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. +along with this program; if not, write to the Free Software +Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA + */ #include @@ -27,6 +27,7 @@ Free Software Foundation, 59 Temple Place - Suite 330, Boston, MA #include #include +#include #include #include @@ -245,7 +246,7 @@ 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)); @@ -295,8 +296,6 @@ 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 != '=') { @@ -305,8 +304,6 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) 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; } @@ -317,7 +314,6 @@ int d1_check_xpath_predicate(data1_node *n, struct xpath_predicate *p) } } } - yaz_log(YLOG_DEBUG, "return %d", res); return res; } else { return 1; @@ -392,20 +388,37 @@ pop, 2003-01-17 data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) { data1_absyn *abs = n->root->u.root.absyn; - data1_xpelement *xpe = abs->xp_elements; + + data1_xpelement *xpe = 0; data1_node *nn; #ifdef ENHANCED_XELM struct xpath_location_step *xp; #endif char *pexpr = xmalloc(strlen(tagpath)+5); - int ok = 0; sprintf (pexpr, "/%s\n", tagpath); - for (; xpe; xpe = xpe->next) + + for (xpe = abs->xp_elements; xpe; xpe = xpe->next) + xpe->match_state = -1; /* don't know if it matches yet */ + + for (xpe = abs->xp_elements; xpe; xpe = xpe->next) { int i; - ok = dfa_match_first(xpe->dfa->states, pexpr); - + int ok = xpe->match_state; + if (ok == -1) + { /* don't know whether there is a match yet */ + data1_xpelement *xpe1; + + assert(xpe->dfa); + ok = dfa_match_first(xpe->dfa->states, pexpr); + +#if OPTIMIZE_MELM + /* mark this and following ones with same regexp */ + for (xpe1 = xpe; xpe1; xpe1 = xpe1->match_next) + xpe1->match_state = ok; +#endif + } + assert (ok == 0 || ok == 1); if (ok) { #ifdef ENHANCED_XELM /* we have to check the perdicates up to the root node */ @@ -419,12 +432,8 @@ 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--) { - 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; } @@ -440,7 +449,7 @@ data1_termlist *xpath_termlist_by_tagpath(char *tagpath, data1_node *n) xfree(pexpr); - if (ok) { + if (xpe) { yaz_log(YLOG_DEBUG, "Got it"); return xpe->termlists; } else { @@ -528,18 +537,20 @@ 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); @@ -547,6 +558,7 @@ static void index_xpath(struct source_parser *sp, data1_node *n, 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 */ @@ -555,50 +567,37 @@ static void index_xpath(struct source_parser *sp, data1_node *n, 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, 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'; @@ -624,27 +623,9 @@ static void index_xpath(struct source_parser *sp, data1_node *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... */ { @@ -652,55 +633,36 @@ static void index_xpath(struct source_parser *sp, data1_node *n, 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]; /* this could be cached as well */ 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_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); - + wrd->index_name = ZEBRA_XPATH_ATTR_NAME; wrd->index_type = '0'; wrd->term_buf = comb; @@ -709,7 +671,7 @@ static void index_xpath(struct source_parser *sp, data1_node *n, (*p->tokenAdd)(wrd); } - } + } i++; } @@ -721,7 +683,6 @@ static void index_xpath(struct source_parser *sp, data1_node *n, sprintf (attr_tag_path_full, "@%s/%s", xp->name, tag_path_full); - if ((tl = tll[i])) { /* If there is a termlist given (=xelm directive) */ @@ -750,7 +711,7 @@ 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 the attribute as "w" */ - if ((!xpdone) && (!termlist_only)) + if (!xpdone && !termlist_only) { index_xpath_attr (attr_tag_path_full, xp->name, xp->value, "w", p, wrd); @@ -937,6 +898,8 @@ int grs_extract_tree(struct recExtractCtrl *p, data1_node *n) } (*p->init)(p, &wrd); + /* data1_pr_tree(p->dh, n, stdout); */ + return dumpkeys(n, p, &wrd); } @@ -950,12 +913,7 @@ static int grs_extract_sub(void *clientData, struct recExtractCtrl *p, int oidtmp[OID_SIZE]; RecWord wrd; - gri.readf = p->readf; - gri.seekf = p->seekf; - gri.tellf = p->tellf; - gri.endf = p->endf; - gri.fh = p->fh; - gri.offset = p->offset; + gri.stream = p->stream; gri.mem = mem; gri.dh = p->dh; gri.clientData = clientData; @@ -980,6 +938,9 @@ static int grs_extract_sub(void *clientData, struct recExtractCtrl *p, /* ensure our data1 tree is UTF-8 */ data1_iconv (p->dh, mem, n, "UTF-8", data1_get_encoding(p->dh, n)); + + data1_remove_idzebra_subtree (p->dh, n); + #if 0 data1_pr_tree (p->dh, n, stdout); #endif @@ -987,10 +948,8 @@ static int grs_extract_sub(void *clientData, struct recExtractCtrl *p, (*p->init)(p, &wrd); if (dumpkeys(n, p, &wrd) < 0) { - data1_free_tree(p->dh, n); return RECCTRL_EXTRACT_ERROR_GENERIC; } - data1_free_tree(p->dh, n); return RECCTRL_EXTRACT_OK; } @@ -1146,12 +1105,7 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, int dummy; mem = nmem_create(); - gri.readf = p->readf; - gri.seekf = p->seekf; - gri.tellf = p->tellf; - gri.endf = NULL; - gri.fh = p->fh; - gri.offset = 0; + gri.stream = p->stream; gri.mem = mem; gri.dh = p->dh; gri.clientData = clientData; @@ -1160,12 +1114,14 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, node = (*grs_read)(&gri); if (!node) { - p->diagnostic = 14; + 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); + #if 0 data1_pr_tree (p->dh, node, stdout); #endif @@ -1234,7 +1190,7 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, onode = node; if (!(node = data1_map_record(p->dh, onode, map, mem))) { - p->diagnostic = 14; + p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; nmem_destroy (mem); return 0; } @@ -1263,7 +1219,7 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, onode = node; if (!(node = data1_map_record(p->dh, onode, map, mem))) { - p->diagnostic = 14; + p->diagnostic = YAZ_BIB1_SYSTEM_ERROR_IN_PRESENTING_RECORDS; nmem_destroy (mem); return 0; } @@ -1313,9 +1269,6 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, p->odr)) > 0) { p->diagnostic = res; - if (onode) - data1_free_tree(p->dh, onode); - data1_free_tree(p->dh, node); nmem_destroy(mem); return 0; } @@ -1432,10 +1385,6 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p, p->rec_buf = new_buf; } } - if (node) - data1_free_tree(p->dh, node); - if (onode) - data1_free_tree(p->dh, onode); nmem_destroy(mem); return 0; }