More XPATH work; common sequence numbers for extract keys

diff --git a/include/recctrl.h b/include/recctrl.h
index 1b3cce8..485b746 100644 (file)
--- a/include/recctrl.h
+++ b/include/recctrl.h
@@ -4,7 +4,10 @@
  * Sebastian Hammer, Adam Dickmeiss
  *
  * $Log: recctrl.h,v $
- * Revision 1.34  2000-03-20 19:08:36  adam
+ * Revision 1.35  2002-04-13 18:16:42  adam
+ * More XPATH work; common sequence numbers for extract keys
+ *
+ * Revision 1.34  2000/03/20 19:08:36  adam
  * Added remote record import using Z39.50 extended services and Segment
  * Requests.
  *
@@ -142,7 +145,7 @@ typedef struct {
     unsigned reg_type;
     char *string;
     int  length;
-    int  *seqnos;
+    int  seqno;
     ZebraMaps zebra_maps;
     struct recExtractCtrl *extractCtrl;
 } RecWord;

diff --git a/index/extract.c b/index/extract.c
index 040aff0..09e9009 100644 (file)
--- a/index/extract.c
+++ b/index/extract.c
@@ -3,7 +3,7 @@
  * All rights reserved.
  * Sebastian Hammer, Adam Dickmeiss
  *
- * $Id: extract.c,v 1.115 2002-04-04 20:50:36 adam Exp $
+ * $Id: extract.c,v 1.116 2002-04-13 18:16:43 adam Exp $
  */
 #include <stdio.h>
 #include <assert.h>
@@ -44,7 +44,7 @@ static void logRecord (int showFlag)
 static void extract_init (struct recExtractCtrl *p, RecWord *w)
 {
     w->zebra_maps = p->zebra_maps;
-    w->seqnos = p->seqno;
+    w->seqno = 1;
     w->attrSet = VAL_BIB1;
     w->attrUse = 1016;
     w->reg_type = 'w';
@@ -1359,11 +1359,11 @@ void extract_add_index_string (RecWord *p, const char *string,
     unsigned short attrUse;
     int lead = 0;
     int diff = 0;
-    int *pseqno = &p->seqnos[p->reg_type];
+    int *pseqno = &p->seqno;
     ZebraHandle zh = p->extractCtrl->handle;
     ZebraExplainInfo zei = zh->reg->zei;
     struct recKeys *keys = &zh->reg->keys;
-
+    
     if (keys->buf_used+1024 > keys->buf_max)
     {
         char *b;
@@ -1434,8 +1434,6 @@ void extract_add_index_string (RecWord *p, const char *string,
         dst += sizeof(*pseqno);
     }
     keys->buf_used = dst - keys->buf;
-    if (*pseqno)
-       (*pseqno)++;
 }
 
 static void extract_add_sort_string (RecWord *p, const char *string,
@@ -1510,8 +1508,8 @@ static void extract_add_incomplete_field (RecWord *p)
        if (!i)
            return;
        extract_add_string (p, buf, i);
+        p->seqno++;
     }
-    (p->seqnos[p->reg_type])++; /* to separate this from next one  */
 }
 
 static void extract_add_complete_field (RecWord *p)

diff --git a/index/zrpn.c b/index/zrpn.c
index bb8dc7a..78f6180 100644 (file)
--- a/index/zrpn.c
+++ b/index/zrpn.c
@@ -3,7 +3,7 @@
  * All rights reserved.
  * Sebastian Hammer, Adam Dickmeiss
  *
- * $Id: zrpn.c,v 1.114 2002-04-12 14:55:22 adam Exp $
+ * $Id: zrpn.c,v 1.115 2002-04-13 18:16:43 adam Exp $
  */
 #include <stdio.h>
 #include <assert.h>
@@ -85,17 +85,17 @@ static int attr_find_ex (AttrType *src, oid_value *attributeSetP,
             case Z_AttributeValue_complex:
                 if (src->minor >= element->value.complex->num_list)
                    break;
+                if (element->attributeSet && attributeSetP)
+                {
+                    oident *attrset;
+                    
+                    attrset = oid_getentbyoid (element->attributeSet);
+                    *attributeSetP = attrset->value;
+                }
                 if (element->value.complex->list[src->minor]->which ==  
                     Z_StringOrNumeric_numeric)
                {
                    ++(src->minor);
-                   if (element->attributeSet && attributeSetP)
-                   {
-                       oident *attrset;
-                       
-                       attrset = oid_getentbyoid (element->attributeSet);
-                       *attributeSetP = attrset->value;
-                   }
                    return
                        *element->value.complex->list[src->minor-1]->u.numeric;
                }
@@ -856,9 +856,9 @@ static int string_term (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
     truncation_value = attr_find (&truncation, NULL);
     logf (LOG_DEBUG, "truncation value %d", truncation_value);
 
-    if (use_value == -1)
+    if (use_value == -1)    /* no attribute - assumy "any" */
         use_value = 1016;
-    if (use_value == -2)
+    if (use_value == -2)    /* string attribute - assumy "any" */
         use_value = 1016;
 
     for (base_no = 0; base_no < num_bases; base_no++)
@@ -1931,8 +1931,8 @@ static RSET rpn_search_xpath (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
     oid_value curAttributeSet = attributeSet;
     char term_dict[2048];
     int base_no;
+    int reg_type = '0';
     struct grep_info grep_info;
-    struct rpn_char_map_info rcmi;
 
     yaz_log (LOG_LOG, "rpn_search_xpath 1");
     attr_init (&use, zapt, 1);
@@ -1949,14 +1949,14 @@ static RSET rpn_search_xpath (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
         return rset;
     }
 
-    rpn_char_map_prepare (zh->reg, '0', &rcmi);
-
-    if (grep_info_prepare (zh, zapt, &grep_info, '0', stream))
+    dict_grep_cmap (zh->reg->dict, 0, 0);
+    if (grep_info_prepare (zh, zapt, &grep_info, reg_type, stream))
        return 0;
 
     yaz_log (LOG_LOG, "rpn_search_xpath 2");
     for (base_no = 0; base_no < num_bases; base_no++)
     {
+        const char *termp = use_string;
         rset_between_parms parms;
         RSET rset_start_tag, rset_end_tag;
         int ord, ord_len, i, r, max_pos;
@@ -1986,8 +1986,11 @@ static RSET rpn_search_xpath (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
         }
         term_dict[prefix_len++] = ')';
         term_dict[prefix_len++] = 1;
-        term_dict[prefix_len++] = '0';
+        term_dict[prefix_len++] = reg_type;
+
+        termp = use_string;
         strcpy (term_dict+prefix_len, use_string);
+        
         grep_info.isam_p_indx = 0;
         yaz_log (LOG_LOG, "rpn_search_xpath 3 %s", term_dict+prefix_len);
         r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
@@ -2016,8 +2019,12 @@ static RSET rpn_search_xpath (ZebraHandle zh, Z_AttributesPlusTerm *zapt,
         }
         term_dict[prefix_len++] = ')';
         term_dict[prefix_len++] = 1;
-        term_dict[prefix_len++] = '0';
+        term_dict[prefix_len++] = reg_type;
+
+        termp = use_string;
+
         strcpy (term_dict+prefix_len, use_string);
+
         grep_info.isam_p_indx = 0;
         r = dict_lookup_grep (zh->reg->dict, term_dict, 0,
                               &grep_info, &max_pos, 0, grep_handle);

diff --git a/recctrl/recgrs.c b/recctrl/recgrs.c
index deb7046..5974888 100644 (file)
--- a/recctrl/recgrs.c
+++ b/recctrl/recgrs.c
@@ -3,7 +3,10 @@
  * All rights reserved.
  *
  * $Log: recgrs.c,v $
- * Revision 1.45  2002-04-12 14:40:42  adam
+ * Revision 1.46  2002-04-13 18:16:43  adam
+ * More XPATH work; common sequence numbers for extract keys
+ *
+ * Revision 1.45  2002/04/12 14:40:42  adam
  * Work on XPATH
  *
  * Revision 1.44  2002/04/11 20:09:47  adam
@@ -477,10 +480,9 @@ static void index_termlist (data1_node *par, data1_node *n,
     }
 }
 
-static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level)
+static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level,
+                    RecWord *wrd)
 {
-    RecWord wrd;
-    (*p->init)(p, &wrd);      /* set defaults */
     for (; n; n = n->next)
     {
        if (p->flagShowRecords) /* display element description to user */
@@ -522,14 +524,14 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level)
 
        if (n->which == DATA1N_tag)
        {
-            index_termlist (n, n, p, level, &wrd);
+            index_termlist (n, n, p, level, wrd);
             /* index start tag */
             if (!n->root->u.root.absyn)
-                index_xpath (n, p, level, &wrd, 1);
+                index_xpath (n, p, level, wrd, 1);
        }
 
        if (n->child)
-           if (dumpkeys(n->child, p, level + 1) < 0)
+           if (dumpkeys(n->child, p, level + 1, wrd) < 0)
                return -1;
 
 
@@ -551,9 +553,9 @@ static int dumpkeys(data1_node *n, struct recExtractCtrl *p, int level)
            }
 
            if (par)
-               index_termlist (par, n, p, level, &wrd);
+               index_termlist (par, n, p, level, wrd);
             if (!n->root->u.root.absyn)
-                index_xpath (n, p, level, &wrd, 1016);
+                index_xpath (n, p, level, wrd, 1016);
 
        }
 
@@ -561,7 +563,7 @@ 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);
+                index_xpath (n, p, level, wrd, 2);
        }
 
 
@@ -577,6 +579,7 @@ 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;
@@ -587,7 +590,8 @@ int grs_extract_tree(struct recExtractCtrl *p, data1_node *n)
         if ((oid_ent_to_oid (&oe, oidtmp)))
             (*p->schemaAdd)(p, oidtmp);
     }
-    return dumpkeys(n, p, 0);
+    (*p->init)(p, &wrd);
+    return dumpkeys(n, p, 0, &wrd);
 }
 
 static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p,
@@ -597,6 +601,7 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p,
     struct grs_read_info gri;
     oident oe;
     int oidtmp[OID_SIZE];
+    RecWord wrd;
 
     gri.readf = p->readf;
     gri.seekf = p->seekf;
@@ -626,7 +631,8 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p,
 #if 0
     data1_pr_tree (p->dh, n, stdout);
 #endif
-    if (dumpkeys(n, p, 0) < 0)
+    (*p->init)(p, &wrd);
+    if (dumpkeys(n, p, 0, &wrd) < 0)
     {
        data1_free_tree(p->dh, n);
        return RECCTRL_EXTRACT_ERROR;

diff --git a/tab/default.idx b/tab/default.idx
index 2632158..cec40e5 100644 (file)
--- a/tab/default.idx
+++ b/tab/default.idx
@@ -1,5 +1,5 @@
 # Zebra indexes as referred to from the *.abs-files.
-#  $Id: default.idx,v 1.8 2000-03-02 14:35:19 adam Exp $
+#  $Id: default.idx,v 1.9 2002-04-13 18:16:43 adam Exp $
 #
 # Traditional word index
 index w
@@ -25,6 +25,7 @@ charmap numeric.chr
 # Null map index (no mapping at all)
 index 0
 completeness 0
+position 1
 charmap @
 
 # Sort register (no mapping at all)