Save register type for TERMIDs and snippets.

diff --git a/include/idzebra/snippet.h b/include/idzebra/snippet.h
index e778e3c..d4cfb1d 100644 (file)
--- a/include/idzebra/snippet.h
+++ b/include/idzebra/snippet.h
@@ -1,4 +1,4 @@
-/* $Id: snippet.h,v 1.1 2005-06-07 11:36:38 adam Exp $
+/* $Id: snippet.h,v 1.2 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -29,6 +29,7 @@ YAZ_BEGIN_CDECL
 
 struct zebra_snippet_word {
     zint seqno;
+    int reg_type;
     int ord;
     char *term;
     int match;
@@ -41,9 +42,11 @@ typedef struct zebra_snippet_word zebra_snippet_word;
 zebra_snippets *zebra_snippets_create();
 void zebra_snippets_destroy(zebra_snippets *l);
 void zebra_snippets_append(zebra_snippets *l,
-                          zint seqno, int ord, const char *term);
+                          zint seqno, int reg_type,
+                          int ord, const char *term);
 void zebra_snippets_append_match(zebra_snippets *l,
-                                zint seqno, int ord, const char *term,
+                                zint seqno, int reg_type,
+                                int ord, const char *term,
                                 int match);
 zebra_snippet_word *zebra_snippets_list(zebra_snippets *l);
 void zebra_snippets_log(zebra_snippets *l, int log_level);

diff --git a/include/rset.h b/include/rset.h
index b37becc..bf8f87b 100644 (file)
--- a/include/rset.h
+++ b/include/rset.h
@@ -1,4 +1,4 @@
-/* $Id: rset.h,v 1.55 2005-06-07 07:41:04 adam Exp $
+/* $Id: rset.h,v 1.56 2005-06-07 14:53:38 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -58,6 +58,7 @@ struct rset_term {
                       This info is used to return encoded term back for
                       search-result-1 .
                   */
+    int reg_type;  /** register type */
     RSET rset;     /** the rset corresponding to this term */
     void *rankpriv;/** private stuff for the ranking algorithm */
     struct ord_list *ol;
@@ -65,7 +66,8 @@ struct rset_term {
 
 typedef struct rset_term *TERMID; 
 TERMID rset_term_create (const char *name, int length, const char *flags,
-                        int type, NMEM nmem, struct ord_list *ol);
+                        int type, NMEM nmem, struct ord_list *ol,
+                        int reg_type);
 
 /** rsfd is a "file descriptor" for reading from a rset */
 struct rsfd {  /* the stuff common to all rsfd's. */

diff --git a/index/extract.c b/index/extract.c
index ca6d012..acd63b2 100644 (file)
--- a/index/extract.c
+++ b/index/extract.c
@@ -1,4 +1,4 @@
-/* $Id: extract.c,v 1.184 2005-06-07 11:36:38 adam Exp $
+/* $Id: extract.c,v 1.185 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -1498,7 +1498,7 @@ ZEBRA_RES zebra_snippets_rec_keys(ZebraHandle zh, struct recKeys *reckeys,
        seqno = (int) key.mem[key.len-1];
 
        zebra_term_untrans_iconv(zh, nmem, src[0], &dst_term, src+1);
-       zebra_snippets_append(snippets, seqno, key.mem[0], dst_term);
+       zebra_snippets_append(snippets, seqno, src[0], key.mem[0], dst_term);
         while (*src++)
            ;
         off = src - reckeys->buf;

diff --git a/index/index.h b/index/index.h
index 6d656ac..95371fa 100644 (file)
--- a/index/index.h
+++ b/index/index.h
@@ -1,4 +1,4 @@
-/* $Id: index.h,v 1.141 2005-06-07 11:36:38 adam Exp $
+/* $Id: index.h,v 1.142 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -368,7 +368,7 @@ RSET rset_trunc(ZebraHandle zh, ISAM_P *isam_p, int no,
                const char *term, int length_term, const char *flags,
                int preserve_position, int term_type, NMEM rset_nmem,
                struct rset_key_control *kctrl, int scope,
-               struct ord_list *ol);
+               struct ord_list *ol, int reg_type);
 
 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 63b27e8..b416592 100644 (file)
--- a/index/trunc.c
+++ b/index/trunc.c
@@ -1,4 +1,4 @@
-/* $Id: trunc.c,v 1.59 2005-06-06 21:31:08 adam Exp $
+/* $Id: trunc.c,v 1.60 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -400,13 +400,14 @@ RSET rset_trunc(ZebraHandle zi, ISAM_P *isam_p, int no,
                const char *term, int length, const char *flags,
                int preserve_position, int term_type, NMEM rset_nmem,
                struct rset_key_control *kctrl, int scope,
-               struct ord_list *ol)
+               struct ord_list *ol, int reg_type)
 {
     TERMID termid;
     RSET result;
     int trunc_chunk;
     
-    termid = rset_term_create(term, length, flags, term_type, rset_nmem, ol);
+    termid = rset_term_create(term, length, flags, term_type, rset_nmem, ol,
+                             reg_type);
     if (no < 1)
        return rsnull_create(rset_nmem, kctrl, termid);
     

diff --git a/index/zrpn.c b/index/zrpn.c
index 4b39270..1b74b9a 100644 (file)
--- a/index/zrpn.c
+++ b/index/zrpn.c
@@ -1,4 +1,4 @@
-/* $Id: zrpn.c,v 1.196 2005-06-07 07:41:05 adam Exp $
+/* $Id: zrpn.c,v 1.197 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -1018,7 +1018,7 @@ static ZEBRA_RES term_trunc(ZebraHandle zh,
                       grep_info->isam_p_indx, term_dst,
                       strlen(term_dst), rank_type, 1 /* preserve pos */,
                       zapt->term->which, rset_nmem,
-                      kc, kc->scope, ol);
+                      kc, kc->scope, ol, reg_type);
     if (!*rset)
        return ZEBRA_FAIL;
     return ZEBRA_OK;
@@ -1965,7 +1965,7 @@ static ZEBRA_RES rpn_search_APT_numeric(ZebraHandle zh,
                       strlen(term_dst), rank_type,
                       0 /* preserve position */,
                       zapt->term->which, rset_nmem, 
-                      kc, kc->scope, 0);
+                      kc, kc->scope, 0, reg_type);
        if (!result_sets[num_result_sets])
            break;
        num_result_sets++;
@@ -2159,7 +2159,7 @@ static RSET xpath_trunc(ZebraHandle zh, NMEM stream,
     rset = rset_trunc(zh, grep_info.isam_p_buf,
                      grep_info.isam_p_indx, term, strlen(term),
                      flags, 1, term_type,rset_nmem,
-                     kc, kc->scope, 0);
+                     kc, kc->scope, 0, reg_type);
     grep_info_delete(&grep_info);
     return rset;
 }
@@ -2963,7 +2963,7 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
            rset = rset_trunc(zh, &scan_info_array[j0].list[ptr[j0]].isam_p, 1,
                              glist[lo].term, strlen(glist[lo].term),
                              NULL, 0, zapt->term->which, rset_nmem, 
-                             kc, kc->scope, 0);
+                             kc, kc->scope, 0, reg_id);
        }
        ptr[j0]++; /* move index for this set .. */
        /* get result set for remaining scan terms */
@@ -2984,7 +2984,7 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
                            glist[lo].term,
                            strlen(glist[lo].term), NULL, 0,
                            zapt->term->which,rset_nmem,
-                           kc, kc->scope, 0);
+                           kc, kc->scope, 0, reg_id);
                    rset = rsmulti_or_create(rset_nmem, kc,
                                             kc->scope, 0 /* termid */,
                                             2, rsets);
@@ -3054,7 +3054,7 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
            (zh, &scan_info_array[j0].list[before-1-ptr[j0]].isam_p, 1,
             glist[lo].term, strlen(glist[lo].term),
             NULL, 0, zapt->term->which, rset_nmem,
-            kc, kc->scope, 0);
+            kc, kc->scope, 0, reg_id);
        
        ptr[j0]++;
        
@@ -3073,7 +3073,7 @@ ZEBRA_RES rpn_scan(ZebraHandle zh, ODR stream, Z_AttributesPlusTerm *zapt,
                    glist[lo].term,
                    strlen(glist[lo].term), NULL, 0,
                    zapt->term->which, rset_nmem,
-                   kc, kc->scope, 0);
+                   kc, kc->scope, 0, reg_id);
                rset = rsmulti_or_create(rset_nmem, kc,
                                         kc->scope, 0 /* termid */, 2, rsets);
                

diff --git a/index/zsets.c b/index/zsets.c
index b4650a1..b20041f 100644 (file)
--- a/index/zsets.c
+++ b/index/zsets.c
@@ -1,4 +1,4 @@
-/* $Id: zsets.c,v 1.87 2005-06-07 11:36:38 adam Exp $
+/* $Id: zsets.c,v 1.88 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -1073,11 +1073,10 @@ ZEBRA_RES zebra_snippets_hit_vector(ZebraHandle zh, const char *setname,
            if (termid)
            {
                struct ord_list *ol;
-               key_logdump_txt(YLOG_LOG, &key, termid->name);
                for (ol = termid->ol; ol; ol = ol->next)
                {
-                   yaz_log(YLOG_LOG, "   ord=%d", ol->ord);
                    zebra_snippets_append(snippets, key.mem[key.len-1],
+                                         termid->reg_type,
                                          ol->ord, termid->name);
                }
            }

diff --git a/rset/rsbetween.c b/rset/rsbetween.c
index e2411fd..d988a6b 100644 (file)
--- a/rset/rsbetween.c
+++ b/rset/rsbetween.c
@@ -1,4 +1,4 @@
-/* $Id: rsbetween.c,v 1.40 2005-06-06 21:31:08 adam Exp $
+/* $Id: rsbetween.c,v 1.41 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -96,7 +96,7 @@ static void checkterm(RSET rs, char *tag, NMEM nmem)
 {
     if (!rs->term)
     {
-        rs->term = rset_term_create(tag, strlen(tag), "", 0, nmem, 0);
+        rs->term = rset_term_create(tag, strlen(tag), "", 0, nmem, 0, 0);
         rs->term->rset = rs;
     }
 }

diff --git a/rset/rset.c b/rset/rset.c
index c7821d3..52f3033 100644 (file)
--- a/rset/rset.c
+++ b/rset/rset.c
@@ -1,4 +1,4 @@
-/* $Id: rset.c,v 1.49 2005-06-07 07:41:05 adam Exp $
+/* $Id: rset.c,v 1.50 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -307,9 +307,11 @@ struct ord_list *ord_list_dup(NMEM nmem, struct ord_list *list)
    \param type Term Type, Z_Term_general, Z_Term_characterString,..
    \param nmem memory for term.
    \param ol ord list
+   \param reg_type register type
 */
 TERMID rset_term_create(const char *name, int length, const char *flags,
-                       int type, NMEM nmem, struct ord_list *ol)
+                       int type, NMEM nmem, struct ord_list *ol,
+                       int reg_type)
 
 {
     TERMID t;
@@ -331,6 +333,7 @@ TERMID rset_term_create(const char *name, int length, const char *flags,
     else
         t->flags = nmem_strdup(nmem, flags);
     t->type = type;
+    t->reg_type = reg_type;
     t->rankpriv = 0;
     t->rset = 0;
     t->ol = ord_list_dup(nmem, ol);

diff --git a/util/snippet.c b/util/snippet.c
index bca55ef..a3045a3 100644 (file)
--- a/util/snippet.c
+++ b/util/snippet.c
@@ -1,4 +1,4 @@
-/* $Id: snippet.c,v 1.2 2005-06-07 13:10:52 adam Exp $
+/* $Id: snippet.c,v 1.3 2005-06-07 14:53:39 adam Exp $
    Copyright (C) 1995-2005
    Index Data ApS
 
@@ -47,14 +47,14 @@ void zebra_snippets_destroy(zebra_snippets *l)
 }
 
 void zebra_snippets_append(zebra_snippets *l,
-                          zint seqno, int ord, const char *term)
+                          zint seqno, int reg_type, int ord, const char *term)
 {
-    zebra_snippets_append_match(l, seqno, ord, term, 0);
+    zebra_snippets_append_match(l, seqno, reg_type, ord, term, 0);
 }
 
 void zebra_snippets_append_match(zebra_snippets *l,
-                                zint seqno, int ord, const char *term,
-                                int match)
+                                zint seqno, int reg_type, 
+                                int ord, const char *term, int match)
 {
     struct zebra_snippet_word *w = nmem_malloc(l->nmem, sizeof(*w));
 
@@ -66,6 +66,7 @@ void zebra_snippets_append_match(zebra_snippets *l,
     l->tail = w;
 
     w->seqno = seqno;
+    w->reg_type = reg_type;
     w->ord = ord;
     w->term = nmem_strdup(l->nmem, term);
     w->match = match;
@@ -80,8 +81,9 @@ void zebra_snippets_log(zebra_snippets *l, int log_level)
 {
     zebra_snippet_word *w;
     for (w = l->front; w; w = w->next)
-       yaz_log(log_level, "term=%s%s seqno=" ZINT_FORMAT " ord=%d",
-               w->term, (w->match ? "*" : ""), w->seqno, w->ord);
+       yaz_log(log_level, "term=%s%s seqno=" ZINT_FORMAT " reg_type=%c "
+               "ord=%d",
+               w->term, (w->match ? "*" : ""), w->seqno, w->reg_type, w->ord);
 }
 
 zebra_snippets *zebra_snippets_window(zebra_snippets *doc, zebra_snippets *hit,
@@ -96,12 +98,17 @@ zebra_snippets *zebra_snippets_window(zebra_snippets *doc, zebra_snippets *hit,
     while(1)
     {
        int window_start;
+       int reg_type;
        zebra_snippet_word *hit_w, *doc_w;
        int min_ord = 0; /* not set yet */
        for (hit_w = zebra_snippets_list(hit); hit_w; hit_w = hit_w->next)
            if (hit_w->ord > ord &&
-               (min_ord == 0 || hit_w->ord < min_ord))
+               (min_ord == 0 || 
+                (hit_w->ord < min_ord && hit_w->reg_type == reg_type)))
+           {
                min_ord = hit_w->ord;
+               reg_type = hit_w->reg_type;
+           }
        if (min_ord == 0)
            break;
        ord = min_ord;
@@ -119,7 +126,7 @@ zebra_snippets *zebra_snippets_window(zebra_snippets *doc, zebra_snippets *hit,
                int seq_no_last = 0;
                while (look_w && look_w->seqno < hit_w->seqno + window_size)
                {
-                   if (look_w->ord == ord)
+                   if (look_w->ord == ord && look_w->reg_type == reg_type)
                    {
                        seq_no_last = look_w->seqno;
                        number_this++;
@@ -134,29 +141,31 @@ zebra_snippets *zebra_snippets_window(zebra_snippets *doc, zebra_snippets *hit,
                }
            }
        }
-       yaz_log(YLOG_LOG, "ord=%d", ord);
-       yaz_log(YLOG_LOG, "first_seq_no_best_window=%d", first_seq_no_best_window);
-       yaz_log(YLOG_LOG, "last_seq_no_best_window=%d", last_seq_no_best_window);
-       yaz_log(YLOG_LOG, "number_best_window=%d", number_best_window);
+       yaz_log(YLOG_DEBUG, "ord=%d", ord);
+       yaz_log(YLOG_DEBUG, "first_seq_no_best_window=%d", first_seq_no_best_window);
+       yaz_log(YLOG_DEBUG, "last_seq_no_best_window=%d", last_seq_no_best_window);
+       yaz_log(YLOG_DEBUG, "number_best_window=%d", number_best_window);
 
        window_start = (first_seq_no_best_window + last_seq_no_best_window -
                        window_size) / 2;
        for (doc_w = zebra_snippets_list(doc); doc_w; doc_w = doc_w->next)
-           if (doc_w->ord == ord 
+           if (doc_w->ord == ord && doc_w->reg_type == reg_type
                && doc_w->seqno >= window_start
                && doc_w->seqno < window_start + window_size)
            {
                int match = 0;
                for (hit_w = zebra_snippets_list(hit); hit_w; hit_w = hit_w->next)
                {
-                   if (hit_w->ord == ord && hit_w->seqno == doc_w->seqno)
+                   if (hit_w->ord == ord && hit_w->reg_type == reg_type &&
+                       hit_w->seqno == doc_w->seqno)
                        
                    {
                        match = 1;
                        break;
                    }
                }
-               zebra_snippets_append_match(result, doc_w->seqno, ord,
+               zebra_snippets_append_match(result, doc_w->seqno,
+                                           doc_w->reg_type, ord,
                                            doc_w->term, match);
            }
     }