X-Git-Url: http://git.indexdata.com/?p=idzebra-moved-to-github.git;a=blobdiff_plain;f=recctrl%2Frecgrs.c;h=c41906fcac47312827e1d40e42329b5fd140d7bf;hp=b832911adf338ed90708317d03f93ccbadbc7771;hb=896c0427df9d8eff5de6a1735dcd992e067df844;hpb=a749ce4ff4739420474e7d40f400321a86f67684

diff --git a/recctrl/recgrs.c b/recctrl/recgrs.c
index b832911..c41906f 100644
--- a/recctrl/recgrs.c
+++ b/recctrl/recgrs.c
@@ -1,9 +1,26 @@
-/*
- * Copyright (C) 1994-2002, Index Data
- * All rights reserved.
- *
- * $Id: recgrs.c,v 1.54 2002-07-05 16:07:02 adam Exp $
- */
+/* $Id: recgrs.c,v 1.59 2002-08-02 19:26:56 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 <stdio.h>
 #include <assert.h>
@@ -105,6 +122,16 @@ 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 +191,67 @@ 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];
+                    
+                    sprintf (attr_tag_path_full, "@%s/%.*s",
+                             xp->name, flen, 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);
+                    
+                    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;
     }
 }
 
@@ -396,10 +481,13 @@ static int grs_extract_sub(struct grs_handlers *h, struct recExtractCtrl *p,
         if ((oid_ent_to_oid (&oe, oidtmp)))
             (*p->schemaAdd)(p, oidtmp);
     }
+
+    /* ensure our data1 tree is UTF-8 */
+    data1_iconv (p->dh, mem, n, "UTF-8", data1_get_encoding(p->dh, n));
+
 #if 0
     data1_pr_tree (p->dh, n, stdout);
 #endif
-    data1_iconv (p->dh, mem, n, "ISO-8859-1", "UTF-8");
 
     (*p->init)(p, &wrd);
     if (dumpkeys(n, p, 0, &wrd) < 0)
@@ -558,6 +646,9 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p)
         nmem_destroy (mem);
 	return 0;
     }
+    /* ensure our data1 tree is UTF-8 */
+    data1_iconv (p->dh, mem, node, "UTF-8", data1_get_encoding(p->dh, node));
+
 #if 0
     data1_pr_tree (p->dh, node, stdout);
 #endif
@@ -711,17 +802,16 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p)
 #if 0
     data1_pr_tree (p->dh, node, stdout);
 #endif
-#if 1
-    data1_iconv (p->dh, mem, node, "ISO-8859-1", "UTF-8");
-#endif
     logf (LOG_DEBUG, "grs_retrieve: transfer syntax mapping");
     switch (p->output_format = (p->input_format != VAL_NONE ?
 				p->input_format : VAL_SUTRS))
     {
-	
     case VAL_TEXT_XML:
         add_idzebra_info (p, top, mem);
 
+        if (p->encoding)
+            data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");
+
 	if (!(p->rec_buf = data1_nodetoidsgml(p->dh, node, selected,
 					      &p->rec_len)))
 	    p->diagnostic = 238;
@@ -755,6 +845,8 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p)
 	    p->rec_len = (size_t) (-1);
 	break;
     case VAL_SUTRS:
+        if (p->encoding)
+            data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");
 	if (!(p->rec_buf = data1_nodetobuf(p->dh, node, selected,
 					   &p->rec_len)))
 	    p->diagnostic = 238;
@@ -791,6 +883,8 @@ static int grs_retrieve(void *clientData, struct recRetrieveCtrl *p)
 	    p->diagnostic = 238;
 	    break;
 	}
+        if (p->encoding)
+            data1_iconv (p->dh, mem, node, p->encoding, "UTF-8");
 	if (!(p->rec_buf = data1_nodetomarc(p->dh, marctab, node,
 					selected, &p->rec_len)))
 	    p->diagnostic = 238;