bug #504 fix

The GRS module now ignores any previously existing GRS subtree of the form
  <idzebra xmlns="http://www.indexdata.dk/zebra/">
    ...
  </idzebra>
both under indexing and retrieval of records. This is to facilitate remote
update of GRS records, which have been augmented by zebra-internal meta data
info, such that this part is not appended twice.
   The local storage is still including the GRS <idzebra ..> subtree, but
it is not indexed any more, and stripped before retrieval, i.e. for all
practical purposes it is not existent.

diff --git a/data1/Makefile.am b/data1/Makefile.am
index acff185..3aba070 100644 (file)
--- a/data1/Makefile.am
+++ b/data1/Makefile.am
@@ -1,10 +1,11 @@
-## $Id: Makefile.am,v 1.7 2006-07-05 15:03:45 adam Exp $
+## $Id: Makefile.am,v 1.8 2006-07-06 12:42:21 marc Exp $
 
 noinst_LTLIBRARIES=libidzebra-data1.la
 
 libidzebra_data1_la_SOURCES = d1_handle.c d1_read.c d1_attset.c d1_tagset.c \
   d1_absyn.c d1_grs.c d1_sutrs.c d1_varset.c d1_espec.c d1_doespec.c d1_map.c \
-  d1_marc.c d1_write.c d1_expout.c d1_sumout.c d1_soif.c d1_prtree.c d1_if.c
+  d1_marc.c d1_write.c d1_expout.c d1_sumout.c d1_soif.c d1_prtree.c d1_if.c \
+  d1_utils.c
 
 AM_CPPFLAGS=-I$(top_srcdir)/include $(YAZINC)
 

diff --git a/data1/d1_utils.c b/data1/d1_utils.c
new file mode 100644 (file)
index 0000000..396d1db
--- /dev/null
+++ b/data1/d1_utils.c
@@ -0,0 +1,125 @@
+/* $Id: d1_utils.c,v 1.1 2006-07-06 12:42:21 marc Exp $
+   Copyright (C) 1995-2006
+   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>
+#include <stdlib.h>
+
+#include <yaz/log.h>
+#include <idzebra/data1.h>
+
+
+void data1_remove_node (data1_handle dh, data1_node *n)
+{
+    fprintf (stdout, "REMOVE tag %s \n", n->u.tag.tag);
+    /* n is first childen */
+    if(n->parent->child == n){
+        n->parent->child = n->next;
+
+        /* n is the only child */
+        if(! n->next){
+            n->parent->last_child = 0;                
+        }
+    } 
+    /* n is one of the following childrens */
+    else {
+        data1_node * before;
+        
+        /* need to find sibling before me */
+        before = n->parent->child;
+        while (before->next != n)
+            before = before->next;
+        
+        before->next = n->next;
+        
+        /* n is last child of many */
+        if ( n->parent->last_child == n){
+            n->parent->last_child = before;
+        }
+    }
+    /* break pointers to root, parent and following siblings */
+    n->parent = 0;
+    n->root = 0;
+    n->next = 0;
+}
+
+void data1_remove_idzebra_subtree (data1_handle dh, data1_node *n)
+{
+    switch (n->which)
+        {
+            /*
+        case DATA1N_root:
+            break;
+            */
+        case DATA1N_tag:
+            if (!strcmp(n->u.tag.tag, "idzebra")){
+                if (n->u.tag.attributes){
+                    data1_xattr *xattr = n->u.tag.attributes;
+                    
+                    for (; xattr; xattr = xattr->next){
+                        if (!strcmp(xattr->name, "xmlns") 
+                            & !strcmp(xattr->value, 
+                                      "http://www.indexdata.dk/zebra/"))
+                            data1_remove_node (dh, n);
+                    }
+                }
+            }
+                
+            break;
+            /*
+        case DATA1N_data:
+            break;
+        case DATA1N_comment:
+            break;
+        case DATA1N_preprocess:
+            break;
+        case DATA1N_variant:
+        break;
+            */
+        default:
+            break;
+        }
+    if (n->child)
+        data1_remove_idzebra_subtree (dh, n->child);
+    if (n->next)
+        data1_remove_idzebra_subtree (dh, n->next);
+    /*
+    else
+        {
+            if (n->parent && n->parent->last_child != n)
+                fprintf(out, "%*sWARNING: last_child=%p != %p\n", level, "",
+                        n->parent->last_child, n);
+        }
+    */
+    
+}
+
+
+
+/*
+ * Local variables:
+ * c-basic-offset: 4
+ * indent-tabs-mode: nil
+ * End:
+ * vim: shiftwidth=4 tabstop=8 expandtab
+ */
+

diff --git a/include/idzebra/data1.h b/include/idzebra/data1.h
index b09f3a1..51f138d 100644 (file)
--- a/include/idzebra/data1.h
+++ b/include/idzebra/data1.h
@@ -1,4 +1,4 @@
-/* $Id: data1.h,v 1.16 2006-07-06 11:40:24 adam Exp $
+/* $Id: data1.h,v 1.17 2006-07-06 12:42:22 marc Exp $
    Copyright (C) 1995-2006
    Index Data ApS
 
@@ -371,6 +371,9 @@ YAZ_EXPORT data1_node *data1_read_nodex (data1_handle dh, NMEM m,
 YAZ_EXPORT data1_node *data1_read_record(data1_handle dh, 
                                         int (*rf)(void *, char *, size_t),
                                         void *fh, NMEM m);
+
+YAZ_EXPORT void data1_remove_node (data1_handle dh, data1_node *n);
+YAZ_EXPORT void data1_remove_idzebra_subtree (data1_handle dh, data1_node *n);
 YAZ_EXPORT data1_tag *data1_gettagbynum(data1_handle dh,
                                        data1_tagset *s,
                                        int type, int value);

diff --git a/index/recgrs.c b/index/recgrs.c
index cd11eda..8a31f3c 100644 (file)
--- a/index/recgrs.c
+++ b/index/recgrs.c
@@ -1,4 +1,4 @@
-/* $Id: recgrs.c,v 1.2 2006-07-06 11:40:25 adam Exp $
+/* $Id: recgrs.c,v 1.3 2006-07-06 12:42:22 marc Exp $
    Copyright (C) 1995-2006
    Index Data ApS
 
@@ -937,6 +937,8 @@ 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);
 }
 
@@ -980,6 +982,9 @@ 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
@@ -1164,6 +1169,8 @@ int zebra_grs_retrieve(void *clientData, struct recRetrieveCtrl *p,
     }
     data1_concat_text(p->dh, mem, node);
 
+    data1_remove_idzebra_subtree (p->dh, node);
+
 #if 0
     data1_pr_tree (p->dh, node, stdout);
 #endif

diff --git a/test/filters/Makefile.am b/test/filters/Makefile.am
index 8dd2236..fd7aa60 100644 (file)
--- a/test/filters/Makefile.am
+++ b/test/filters/Makefile.am
@@ -1,13 +1,14 @@
-# $Id: Makefile.am,v 1.3 2006-07-03 14:27:21 adam Exp $
+# $Id: Makefile.am,v 1.4 2006-07-06 12:42:23 marc Exp $
 
-check_PROGRAMS = text grs.xml grs.marc
+check_PROGRAMS = text grs.xml grs.xml.idzebra grs.marc
 
 TESTS = $(check_PROGRAMS)
 
-EXTRA_DIST = zebra.cfg record.xml record.mrc
+EXTRA_DIST = zebra.cfg record.xml record-idzebra.xml record.mrc
 
 text_SOURCES = text.c
 grs_xml_SOURCES = grs.xml.c
+grs_xml_idzebra_SOURCES = grs.xml.idzebra.c
 grs_marc_SOURCES = grs.marc.c
 
 AM_CPPFLAGS = -I$(top_srcdir)/include $(YAZINC)