From: Adam Dickmeiss Date: Fri, 12 Apr 2002 14:40:42 +0000 (+0000) Subject: Work on XPATH X-Git-Tag: ZEBRA.1.3.0~98 X-Git-Url: http://jsfdemo.indexdata.com/cgi-bin?a=commitdiff_plain;h=85728191a4dc93d31e3d1bd0ddc2ae60381ae6ba;p=idzebra-moved-to-github.git Work on XPATH --- diff --git a/dict/Makefile.am b/dict/Makefile.am index 4552c6e..9fd786c 100644 --- a/dict/Makefile.am +++ b/dict/Makefile.am @@ -1,10 +1,9 @@ -## $Id: Makefile.am,v 1.3 2001-10-15 20:57:18 adam Exp $ +## $Id: Makefile.am,v 1.4 2002-04-12 14:40:42 adam Exp $ noinst_LIBRARIES = libdict.a noinst_PROGRAMS = dicttest dictext -INCLUDES = -I$(srcdir)/../include -CFLAGS = @YAZINC@ +INCLUDES = -I$(srcdir)/../include @YAZINC@ LDADD = libdict.a ../bfile/libbfile.a ../dfa/libdfa.a ../util/libutil.a @YAZLIB@ @LIBS@ libdict_a_SOURCES = scan.c dopen.c dclose.c drdwr.c open.c close.c insert.c \ diff --git a/index/Makefile.am b/index/Makefile.am index 559f341..accf045 100644 --- a/index/Makefile.am +++ b/index/Makefile.am @@ -1,4 +1,4 @@ -## $Id: Makefile.am,v 1.11 2002-04-05 08:46:26 adam Exp $ +## $Id: Makefile.am,v 1.12 2002-04-12 14:40:42 adam Exp $ noinst_PROGRAMS = apitest kdump @@ -29,7 +29,7 @@ zebrasrv_SOURCES = zserver.c apitest_SOURCES = apitest.c kdump_SOURCES=kdump.c kcompare.c -CFLAGS = -I$(srcdir)/../include $(YAZINC) $(TCL_INCLUDE) +INCLUDES = -I$(srcdir)/../include $(YAZINC) $(TCL_INCLUDE) LDADD = libzebra.a $(YAZLIB) $(TCL_LIB) diff --git a/index/index.h b/index/index.h index 345cf42..f3ae7f6 100644 --- a/index/index.h +++ b/index/index.h @@ -2,7 +2,7 @@ * Copyright (C) 1995-2002, Index Data * All rights reserved. * Sebastian Hammer, Adam Dickmeiss, Heikki Levanto - * $Id: index.h,v 1.77 2002-04-05 08:46:26 adam Exp $ + * $Id: index.h,v 1.78 2002-04-12 14:40:42 adam Exp $ */ #ifndef INDEX_H @@ -306,7 +306,8 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, int *is_partial); RSET rset_trunc (ZebraHandle zh, ISAMS_P *isam_p, int no, - const char *term, int length_term, const char *flags); + const char *term, int length_term, const char *flags, + int preserve_position); void resultSetAddTerm (ZebraHandle zh, ZebraSet s, int reg_type, const char *db, int set, diff --git a/index/trunc.c b/index/trunc.c index 0f9128e..40c0e50 100644 --- a/index/trunc.c +++ b/index/trunc.c @@ -4,7 +4,10 @@ * Sebastian Hammer, Adam Dickmeiss * * $Log: trunc.c,v $ - * Revision 1.22 2002-04-05 08:46:26 adam + * Revision 1.23 2002-04-12 14:40:42 adam + * Work on XPATH + * + * Revision 1.22 2002/04/05 08:46:26 adam * Zebra with full functionality * * Revision 1.21 2002/04/04 14:14:13 adam @@ -190,8 +193,8 @@ static void heap_close (struct trunc_info *ti) } static RSET rset_trunc_r (ZebraHandle zi, const char *term, int length, - const char *flags, ISAMS_P *isam_p, int from, int to, - int merge_chunk) + const char *flags, ISAMS_P *isam_p, int from, int to, + int merge_chunk, int preserve_position) { RSET result; RSFD result_rsfd; @@ -221,10 +224,12 @@ static RSET rset_trunc_r (ZebraHandle zi, const char *term, int length, { if (i_add <= to - i) rset[rscur] = rset_trunc_r (zi, term, length, flags, - isam_p, i, i+i_add, merge_chunk); + isam_p, i, i+i_add, + merge_chunk, preserve_position); else rset[rscur] = rset_trunc_r (zi, term, length, flags, - isam_p, i, to, merge_chunk); + isam_p, i, to, + merge_chunk, preserve_position); rscur++; } ti = heap_init (rscur, sizeof(struct it_key), key_compare_it); @@ -341,31 +346,32 @@ static RSET rset_trunc_r (ZebraHandle zi, const char *term, int length, int n = ti->indx[ti->ptr[1]]; rset_write (result, result_rsfd, ti->heap[ti->ptr[1]]); -#if 0 -/* section that preserve all keys */ - heap_delete (ti); - if (isc_pp_read (ispt[n], ti->tmpbuf)) - heap_insert (ti, ti->tmpbuf, n); - else - isc_pp_close (ispt[n]); -#else -/* section that preserve all keys with unique sysnos */ - while (1) + if (preserve_position) { - if (!isc_pp_read (ispt[n], ti->tmpbuf)) - { - heap_delete (ti); + heap_delete (ti); + if (isc_pp_read (ispt[n], ti->tmpbuf)) + heap_insert (ti, ti->tmpbuf, n); + else isc_pp_close (ispt[n]); - break; - } - if ((*ti->cmp)(ti->tmpbuf, ti->heap[ti->ptr[1]]) > 1) + } + else + { + while (1) { - heap_delete (ti); - heap_insert (ti, ti->tmpbuf, n); - break; + if (!isc_pp_read (ispt[n], ti->tmpbuf)) + { + heap_delete (ti); + isc_pp_close (ispt[n]); + break; + } + if ((*ti->cmp)(ti->tmpbuf, ti->heap[ti->ptr[1]]) > 1) + { + heap_delete (ti); + heap_insert (ti, ti->tmpbuf, n); + break; + } } } -#endif } heap_close (ti); xfree (ispt); @@ -517,7 +523,8 @@ static int isamd_trunc_cmp (const void *p1, const void *p2) } RSET rset_trunc (ZebraHandle zi, ISAMS_P *isam_p, int no, - const char *term, int length, const char *flags) + const char *term, int length, const char *flags, + int preserve_position) { logf (LOG_DEBUG, "rset_trunc no=%d", no); if (no < 1) @@ -616,6 +623,7 @@ RSET rset_trunc (ZebraHandle zi, ISAMS_P *isam_p, int no, logf (LOG_WARN, "Unknown isam set in rset_trunc"); return rset_create (rset_kind_null, NULL); } - return rset_trunc_r (zi, term, length, flags, isam_p, 0, no, 100); + return rset_trunc_r (zi, term, length, flags, isam_p, 0, no, 100, + preserve_position); } diff --git a/index/zrpn.c b/index/zrpn.c index 2a14aee..06f4259 100644 --- a/index/zrpn.c +++ b/index/zrpn.c @@ -3,7 +3,7 @@ * All rights reserved. * Sebastian Hammer, Adam Dickmeiss * - * $Id: zrpn.c,v 1.112 2002-04-04 14:14:13 adam Exp $ + * $Id: zrpn.c,v 1.113 2002-04-12 14:40:42 adam Exp $ */ #include #include @@ -20,6 +20,7 @@ #include #include #include +#include struct rpn_char_map_info { ZebraMaps zm; @@ -797,12 +798,43 @@ static int string_relation (ZebraHandle zh, Z_AttributesPlusTerm *zapt, } static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, - const char **term_sub, - oid_value attributeSet, NMEM stream, - struct grep_info *grep_info, - int reg_type, int complete_flag, - int num_bases, char **basenames, - char *term_dst) + const char **term_sub, + oid_value attributeSet, NMEM stream, + struct grep_info *grep_info, + int reg_type, int complete_flag, + int num_bases, char **basenames, + char *term_dst); + +static RSET term_trunc (ZebraHandle zh, Z_AttributesPlusTerm *zapt, + const char **term_sub, + oid_value attributeSet, NMEM stream, + struct grep_info *grep_info, + int reg_type, int complete_flag, + int num_bases, char **basenames, + char *term_dst, + const char *rank_type) +{ + int r; + grep_info->isam_p_indx = 0; + r = string_term (zh, zapt, term_sub, attributeSet, stream, grep_info, + reg_type, complete_flag, num_bases, basenames, + term_dst); + if (r < 1) + return 0; + logf (LOG_DEBUG, "term: %s", term_dst); + return rset_trunc (zh, grep_info->isam_p_buf, + grep_info->isam_p_indx, term_dst, + strlen(term_dst), rank_type, 1 /* preserve pos */); +} + + +static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, + const char **term_sub, + oid_value attributeSet, NMEM stream, + struct grep_info *grep_info, + int reg_type, int complete_flag, + int num_bases, char **basenames, + char *term_dst) { char term_dict[2*IT_MAX_WORD+4000]; int j, r, base_no; @@ -810,6 +842,7 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, int truncation_value; AttrType use; int use_value; + const char *use_string = 0; oid_value curAttributeSet = attributeSet; const char *termp; struct rpn_char_map_info rcmi; @@ -817,7 +850,7 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, rpn_char_map_prepare (zh->reg, reg_type, &rcmi); attr_init (&use, zapt, 1); - use_value = attr_find (&use, &curAttributeSet); + use_value = attr_find_ex (&use, &curAttributeSet, &use_string); logf (LOG_DEBUG, "string_term, use value %d", use_value); attr_init (&truncation, zapt, 5); truncation_value = attr_find (&truncation, NULL); @@ -825,78 +858,92 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (use_value == -1) use_value = 1016; + if (use_value == -2) + use_value = 1016; for (base_no = 0; base_no < num_bases; base_no++) { attent attp; + data1_local_attribute id_xpath_attr; data1_local_attribute *local_attr; int max_pos, prefix_len = 0; termp = *term_sub; - if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value))) - { - logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d", - curAttributeSet, use_value, r); - if (r == -1) - { - char val_str[32]; - sprintf (val_str, "%d", use_value); - zh->errCode = 114; - zh->errString = nmem_strdup (stream, val_str); - } - else - { - int oid[OID_SIZE]; - struct oident oident; - - oident.proto = PROTO_Z3950; - oident.oclass = CLASS_ATTSET; - oident.value = curAttributeSet; - oid_ent_to_oid (&oident, oid); - zh->errCode = 121; - zh->errString = nmem_strdup (stream, oident.desc); - } - return -1; - } if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no])) { zh->errCode = 109; /* Database unavailable */ zh->errString = basenames[base_no]; return -1; } + if (curAttributeSet == VAL_IDXPATH) + { + attp.local_attributes = &id_xpath_attr; + attp.attset_ordinal = curAttributeSet; + id_xpath_attr.next = 0; + id_xpath_attr.local = use_value; + } + else + { + if ((r=att_getentbyatt (zh, &attp, curAttributeSet, use_value))) + { + logf (LOG_DEBUG, "att_getentbyatt fail. set=%d use=%d r=%d", + curAttributeSet, use_value, r); + if (r == -1) + { + char val_str[32]; + sprintf (val_str, "%d", use_value); + zh->errCode = 114; + zh->errString = nmem_strdup (stream, val_str); + } + else + { + int oid[OID_SIZE]; + struct oident oident; + + oident.proto = PROTO_Z3950; + oident.oclass = CLASS_ATTSET; + oident.value = curAttributeSet; + oid_ent_to_oid (&oident, oid); + + zh->errCode = 121; + zh->errString = nmem_strdup (stream, oident.desc); + } + return -1; + } + } for (local_attr = attp.local_attributes; local_attr; local_attr = local_attr->next) { int ord; - char ord_buf[32]; - int i, ord_len; - + char ord_buf[32]; + int i, ord_len; + ord = zebraExplain_lookupSU (zh->reg->zei, attp.attset_ordinal, - local_attr->local); + local_attr->local); if (ord < 0) continue; if (prefix_len) term_dict[prefix_len++] = '|'; else term_dict[prefix_len++] = '('; - - ord_len = key_SU_encode (ord, ord_buf); - for (i = 0; ierrCode = 114; - zh->errString = nmem_strdup (stream, val_str); + char val_str[32]; + sprintf (val_str, "%d", use_value); + zh->errCode = 114; + zh->errString = nmem_strdup (stream, val_str); return -1; } - term_dict[prefix_len++] = ')'; + term_dict[prefix_len++] = ')'; term_dict[prefix_len++] = 1; term_dict[prefix_len++] = reg_type; logf (LOG_DEBUG, "reg_type = %d", term_dict[prefix_len-1]); @@ -1334,6 +1381,14 @@ char *normalize_term(ZebraHandle zh, Z_AttributesPlusTerm *zapt, } } +static void grep_info_delete (struct grep_info *grep_info) +{ +#ifdef TERM_COUNT + xfree(grep_info->term_no); +#endif + xfree (grep_info->isam_p_buf); +} + static int grep_info_prepare (ZebraHandle zh, Z_AttributesPlusTerm *zapt, struct grep_info *grep_info, @@ -1392,7 +1447,7 @@ static RSET rpn_search_APT_phrase (ZebraHandle zh, { char term_dst[IT_MAX_WORD+1]; RSET rset[60], result; - int i, r, rset_no = 0; + int i, rset_no = 0; struct grep_info grep_info; char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); const char *termp = termz; @@ -1403,24 +1458,17 @@ static RSET rpn_search_APT_phrase (ZebraHandle zh, while (1) { logf (LOG_DEBUG, "APT_phrase termp=%s", termp); - grep_info.isam_p_indx = 0; - r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info, - reg_type, complete_flag, num_bases, basenames, - term_dst); - if (r < 1) + rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet, + stream, &grep_info, + reg_type, complete_flag, + num_bases, basenames, + term_dst, rank_type); + if (!rset[rset_no]) break; - logf (LOG_DEBUG, "term: %s", term_dst); - rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf, - grep_info.isam_p_indx, term_dst, - strlen(term_dst), rank_type); - assert (rset[rset_no]); if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } -#ifdef TERM_COUNT - xfree(grep_info.term_no); -#endif - xfree (grep_info.isam_p_buf); + grep_info_delete (&grep_info); if (rset_no == 0) { rset_null_parms parms; @@ -1447,7 +1495,7 @@ static RSET rpn_search_APT_or_list (ZebraHandle zh, { char term_dst[IT_MAX_WORD+1]; RSET rset[60], result; - int i, r, rset_no = 0; + int i, rset_no = 0; struct grep_info grep_info; char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); const char *termp = termz; @@ -1457,24 +1505,17 @@ static RSET rpn_search_APT_or_list (ZebraHandle zh, while (1) { logf (LOG_DEBUG, "APT_or_list termp=%s", termp); - grep_info.isam_p_indx = 0; - r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info, - reg_type, complete_flag, num_bases, basenames, - term_dst); - if (r < 1) + rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet, + stream, &grep_info, + reg_type, complete_flag, + num_bases, basenames, + term_dst, rank_type); + if (!rset[rset_no]) break; - logf (LOG_DEBUG, "term: %s", term_dst); - rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf, - grep_info.isam_p_indx, term_dst, - strlen(term_dst), rank_type); - assert (rset[rset_no]); if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } -#ifdef TERM_COUNT - xfree(grep_info.term_no); -#endif - xfree (grep_info.isam_p_buf); + grep_info_delete (&grep_info); if (rset_no == 0) { rset_null_parms parms; @@ -1507,7 +1548,7 @@ static RSET rpn_search_APT_and_list (ZebraHandle zh, { char term_dst[IT_MAX_WORD+1]; RSET rset[60], result; - int i, r, rset_no = 0; + int i, rset_no = 0; struct grep_info grep_info; char *termz = normalize_term(zh, zapt, termz_org, stream, reg_type); const char *termp = termz; @@ -1517,24 +1558,18 @@ static RSET rpn_search_APT_and_list (ZebraHandle zh, while (1) { logf (LOG_DEBUG, "APT_and_list termp=%s", termp); - grep_info.isam_p_indx = 0; - r = string_term (zh, zapt, &termp, attributeSet, stream, &grep_info, - reg_type, complete_flag, num_bases, basenames, - term_dst); - if (r < 1) + rset[rset_no] = term_trunc (zh, zapt, &termp, attributeSet, + stream, &grep_info, + reg_type, complete_flag, + num_bases, basenames, + term_dst, rank_type); + if (!rset[rset_no]) break; - logf (LOG_DEBUG, "term: %s", term_dst); - rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf, - grep_info.isam_p_indx, term_dst, - strlen(term_dst), rank_type); assert (rset[rset_no]); if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } -#ifdef TERM_COUNT - xfree(grep_info.term_no); -#endif - xfree (grep_info.isam_p_buf); + grep_info_delete (&grep_info); if (rset_no == 0) { rset_null_parms parms; @@ -1730,15 +1765,13 @@ static RSET rpn_search_APT_numeric (ZebraHandle zh, logf (LOG_DEBUG, "term: %s", term_dst); rset[rset_no] = rset_trunc (zh, grep_info.isam_p_buf, grep_info.isam_p_indx, term_dst, - strlen(term_dst), rank_type); + strlen(term_dst), rank_type, + 0 /* preserve position */); assert (rset[rset_no]); if (++rset_no >= (int) (sizeof(rset)/sizeof(*rset))) break; } -#ifdef TERM_COUNT - xfree(grep_info.term_no); -#endif - xfree (grep_info.isam_p_buf); + grep_info_delete (&grep_info); if (rset_no == 0) { rset_null_parms parms; @@ -1888,6 +1921,128 @@ static RSET rpn_sort_spec (ZebraHandle zh, Z_AttributesPlusTerm *zapt, return rset_create (rset_kind_null, &parms); } +static RSET rpn_search_xpath (ZebraHandle zh, Z_AttributesPlusTerm *zapt, + oid_value attributeSet, + int num_bases, char **basenames, + NMEM stream, const char *rank_type, RSET rset) +{ + AttrType use; + const char *use_string = 0; + oid_value curAttributeSet = attributeSet; + char term_dict[2048]; + int base_no; + struct grep_info grep_info; + struct rpn_char_map_info rcmi; + + yaz_log (LOG_LOG, "rpn_search_xpath 1"); + attr_init (&use, zapt, 1); + attr_find_ex (&use, &curAttributeSet, &use_string); + + if (curAttributeSet != VAL_IDXPATH) + { + yaz_log (LOG_LOG, "rpn_search_xpath - not 1"); + return rset; + } + if (!use_string) + { + yaz_log (LOG_LOG, "rpn_search_xpath - not 2"); + return rset; + } + + rpn_char_map_prepare (zh->reg, '0', &rcmi); + + if (grep_info_prepare (zh, zapt, &grep_info, '0', stream)) + return 0; + + yaz_log (LOG_LOG, "rpn_search_xpath 2"); + for (base_no = 0; base_no < num_bases; base_no++) + { + rset_between_parms parms; + RSET rset_start_tag, rset_end_tag; + int ord, ord_len, i, r, max_pos; + int prefix_len ; + char ord_buf[32]; + if (zebraExplain_curDatabase (zh->reg->zei, basenames[base_no])) + { + zh->errCode = 109; /* Database unavailable */ + zh->errString = basenames[base_no]; + return rset; + } + + prefix_len = 0; + ord = zebraExplain_lookupSU (zh->reg->zei, curAttributeSet, 1); + if (ord < 0) + continue; + if (prefix_len) + term_dict[prefix_len++] = '|'; + else + term_dict[prefix_len++] = '('; + + ord_len = key_SU_encode (ord, ord_buf); + for (i = 0; ireg->dict, term_dict, 0, + &grep_info, &max_pos, 0, grep_handle); + yaz_log (LOG_LOG, "%s %d positions", use_string, + grep_info.isam_p_indx); + rset_start_tag = + rset_trunc (zh, grep_info.isam_p_buf, + grep_info.isam_p_indx, use_string, strlen(use_string), + rank_type, 1); + + prefix_len = 0; + ord = zebraExplain_lookupSU (zh->reg->zei, curAttributeSet, 2); + if (ord < 0) + continue; + if (prefix_len) + term_dict[prefix_len++] = '|'; + else + term_dict[prefix_len++] = '('; + + ord_len = key_SU_encode (ord, ord_buf); + for (i = 0; ireg->dict, term_dict, 0, + &grep_info, &max_pos, 0, grep_handle); + + yaz_log (LOG_LOG, "%s %d positions", use_string, + grep_info.isam_p_indx); + rset_end_tag = + rset_trunc (zh, grep_info.isam_p_buf, + grep_info.isam_p_indx, use_string, strlen(use_string), + rank_type, 1); + + parms.key_size = sizeof(struct it_key); + parms.cmp = key_compare_it; + parms.rset_l = rset_start_tag; + parms.rset_m = rset; + parms.rset_r = rset_end_tag; + yaz_log (LOG_LOG, "rpn_search_xpath 4"); + rset = rset_create (rset_kind_between, &parms); + } + grep_info_delete (&grep_info); + + return rset; +} + + static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, oid_value attributeSet, NMEM stream, @@ -1900,6 +2055,7 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, int complete_flag; int sort_flag; char termz[IT_MAX_WORD+1]; + RSET rset = 0; zebra_maps_attr (zh->reg->zebra_maps, zapt, ®_id, &search_type, rank_type, &complete_flag, &sort_flag); @@ -1922,35 +2078,37 @@ static RSET rpn_search_APT (ZebraHandle zh, Z_AttributesPlusTerm *zapt, if (!strcmp (search_type, "phrase")) { - return rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream, + rset = rpn_search_APT_phrase (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } else if (!strcmp (search_type, "and-list")) { - return rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream, + rset = rpn_search_APT_and_list (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } else if (!strcmp (search_type, "or-list")) { - return rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream, + rset = rpn_search_APT_or_list (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } else if (!strcmp (search_type, "local")) { - return rpn_search_APT_local (zh, zapt, termz, attributeSet, stream, + rset = rpn_search_APT_local (zh, zapt, termz, attributeSet, stream, rank_type); } else if (!strcmp (search_type, "numeric")) { - return rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream, + rset = rpn_search_APT_numeric (zh, zapt, termz, attributeSet, stream, reg_id, complete_flag, rank_type, num_bases, basenames); } - zh->errCode = 118; - return NULL; + else + zh->errCode = 118; + return rpn_search_xpath (zh, zapt, attributeSet, num_bases, basenames, + stream, rank_type, rset); } static RSET rpn_search_structure (ZebraHandle zh, Z_RPNStructure *zs, @@ -2357,7 +2515,7 @@ void rpn_scan (ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt, &glist[i+before].term, mterm); rset = rset_trunc (zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1, glist[i+before].term, strlen(glist[i+before].term), - NULL); + NULL, 0); ptr[j0]++; for (j = j0+1; jwhich) { case DATA1N_data: wrd->reg_type = 'w'; wrd->string = n->u.data.data; wrd->length = n->u.data.len; - wrd->attrSet = VAL_BIB1; - wrd->attrUse = 1016; + wrd->attrSet = VAL_IDXPATH, + wrd->attrUse = use; if (p->flagShowRecords) { printf("%*s data=", (level + 1) * 4, ""); @@ -351,16 +356,40 @@ static void index_string_tag (data1_node *n, } break; case DATA1N_tag: - wrd->reg_type = 'w'; - wrd->string = n->u.tag.tag; - wrd->length = strlen(n->u.tag.tag); - wrd->attrSet = VAL_BIB1; + 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 < 8; i++) + for (i = 0; ilength && i < 40; i++) fputc (wrd->string[i], stdout); + if (i == 40) + printf (" .."); printf("\n"); } else @@ -496,7 +525,7 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) index_termlist (n, n, p, level, &wrd); /* index start tag */ if (!n->root->u.root.absyn) - index_string_tag (n, p, level, &wrd, 1); + index_xpath (n, p, level, &wrd, 1); } if (n->child) @@ -524,7 +553,7 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) if (par) index_termlist (par, n, p, level, &wrd); if (!n->root->u.root.absyn) - index_string_tag (n, p, level, &wrd, 1016); + index_xpath (n, p, level, &wrd, 1016); } @@ -532,7 +561,7 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level) { /* index end tag */ if (!n->root->u.root.absyn) - index_string_tag (n, p, level, &wrd, 2); + index_xpath (n, p, level, &wrd, 2); } diff --git a/test/api/Makefile.am b/test/api/Makefile.am index ba60a45..5fd0003 100644 --- a/test/api/Makefile.am +++ b/test/api/Makefile.am @@ -6,6 +6,6 @@ EXTRA_DIST=t1.cfg t2.cfg t1_SOURCES = t1.c t2_SOURCES = t2.c -CFLAGS = -I$(top_srcdir)/index -I$(top_srcdir)/include $(YAZINC) +INCLUDES = -I$(top_srcdir)/index -I$(top_srcdir)/include $(YAZINC) LDADD = ../../index/libzebra.a $(YAZLIB) $(TCL_LIB)