For data-1, do not chop text data in ISO2709 creation . The problem is

that in some cases the chop operation will remove essential content.
However, chop is needed in cases where input is XML/SGML. Therefore,
this operation performed in the data-1 map code instead and chop is
enabled by default. The chop can be disabled with 'nochop' in parameter
in map rule, e.g.
  map title /(3,245)/(3,a) nochop

diff --git a/data1/d1_map.c b/data1/d1_map.c
index 997b9cb..724ceb1 100644 (file)
--- a/data1/d1_map.c
+++ b/data1/d1_map.c
@@ -1,4 +1,4 @@
-/* $Id: d1_map.c,v 1.16 2007-04-16 08:44:31 adam Exp $
+/* $Id: d1_map.c,v 1.17 2007-06-27 22:17:20 adam Exp $
    Copyright (C) 1995-2007
    Index Data ApS
 
@@ -30,6 +30,15 @@ Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
 #include <yaz/tpath.h>
 #include <d1_absyn.h>
 
+struct data1_mapunit
+{
+    int no_data;
+    int no_chop;
+    char *source_element_name;
+    data1_maptag *target_path;
+    struct data1_mapunit *next;
+};
+
 data1_maptab *data1_read_maptab (data1_handle dh, const char *file)
 {
     NMEM mem = data1_nmem_get (dh);
@@ -109,6 +118,10 @@ data1_maptab *data1_read_maptab (data1_handle dh, const char *file)
                (*mapp)->no_data = 1;
            else
                (*mapp)->no_data = 0;
+           if (argc > 3 && !data1_matchstr(argv[3], "nochop"))
+               (*mapp)->no_chop = 1;
+           else
+               (*mapp)->no_chop = 0;
            (*mapp)->source_element_name =
                (char *)nmem_malloc(mem, strlen(argv[1])+1);
            strcpy((*mapp)->source_element_name, argv[1]);
@@ -163,23 +176,6 @@ data1_maptab *data1_read_maptab (data1_handle dh, const char *file)
 }
 
 /*
- * Locate node with given elementname.
- * NOTE: This is stupid - we don't find repeats this way.
- */
-static data1_node *find_node(data1_node *p, char *elementname)
-{
-    data1_node *c, *r;
-
-    for (c = p->child; c; c = c->next)
-       if (c->which == DATA1N_tag && c->u.tag.element &&
-           !data1_matchstr(c->u.tag.element->name, elementname))
-           return c;
-       else if ((r = find_node(c, elementname)))
-           return r;
-    return 0;
-}
-
-/*
  * See if the node n is equivalent to the tag t.
  */
 static int tagmatch(data1_node *n, data1_maptag *t)
@@ -305,6 +301,11 @@ static int map_children(data1_handle dh, data1_node *n, data1_maptab *map,
                             cur->child =
                                 dup_child (dh, c->child,
                                            &cur->last_child, mem, cur);
+                            if (!m->no_chop)
+                            {
+                                data1_concat_text(dh, mem, cur->child);
+                                data1_chop_text(dh, mem, cur->child);
+                            }
                        }
                    }
                }

diff --git a/data1/d1_marc.c b/data1/d1_marc.c
index 05fbb76..a091cc6 100644 (file)
--- a/data1/d1_marc.c
+++ b/data1/d1_marc.c
@@ -1,4 +1,4 @@
-/* $Id: d1_marc.c,v 1.18 2007-04-16 08:44:31 adam Exp $
+/* $Id: d1_marc.c,v 1.19 2007-06-27 22:17:20 adam Exp $
    Copyright (C) 1995-2007
    Index Data ApS
 
@@ -160,35 +160,22 @@ data1_marctab *data1_read_marctab (data1_handle dh, const char *file)
 }
 
 
-/*
- * Locate some data under this node. This routine should handle variants
- * prettily.
- */
-static char *get_data(data1_node *n, int *len, int chop)
+static void get_data2(data1_node *n, int *len, char *dst, size_t max)
 {
-    char *r;
+    *len = 0;
 
     while (n)
     {
         if (n->which == DATA1N_data)
         {
-            int i;
-            *len = n->u.data.len;
-           
-           if (chop)
+            if (dst && *len < max)
             {
-               for (i = 0; i<*len; i++)
-                   if (!d1_isspace(n->u.data.data[i]))
-                       break;
-               while (*len && d1_isspace(n->u.data.data[*len - 1]))
-                   (*len)--;
-               *len = *len - i;
-               if (*len > 0)
-                   return n->u.data.data + i;
-           }
-           else
-               if (*len > 0)
-                   return n->u.data.data;
+                size_t copy_len = max - *len;
+                if (copy_len > n->u.data.len)
+                    copy_len = n->u.data.len;
+                memcpy(dst + *len, n->u.data.data, copy_len);
+            }
+            *len += n->u.data.len;
         }
         if (n->which == DATA1N_tag)
             n = n->child;
@@ -197,9 +184,6 @@ static char *get_data(data1_node *n, int *len, int chop)
        else
             break;     
     }
-    r = "";
-    *len = strlen(r);
-    return r;
 }
 
 static void memint (char *p, int val, int len)
@@ -271,11 +255,7 @@ static int nodetomarc(data1_handle dh,
        else if (!strcmp(field->u.tag.tag, "leader"))
        {
            int dlen = 0;
-           char *dbuf = get_data(subf, &dlen, 0);
-           if (dlen > 24)
-               dlen = 24;
-           if (dbuf && dlen > 0)
-               memcpy (leader, dbuf, dlen);
+            get_data2(subf, &dlen, leader, 24);
            continue;
        }
        else if (!strcmp(field->u.tag.tag, "controlfield"))
@@ -320,7 +300,7 @@ static int nodetomarc(data1_handle dh,
                    continue; /* we skip comments, cdata .. */
                 len += p->identifier_length;
            }
-           get_data(subf, &dlen, control_field ? 0 : 1);
+            get_data2(subf, &dlen, 0, 0);
             len += dlen;
         }
     }
@@ -413,8 +393,6 @@ static int nodetomarc(data1_handle dh,
        }
        for (; subf; subf = subf->next)
         {
-           char *data;
-
             if (!control_field)
             {
                 const char *identifier = "a";
@@ -439,8 +417,7 @@ static int nodetomarc(data1_handle dh,
                 memcpy (op + data_p+1, identifier, p->identifier_length-1);
                 data_p += p->identifier_length;
             }
-           data = get_data(subf, &dlen, control_field ? 0 : 1);
-            memcpy (op + data_p, data, dlen);
+            get_data2(subf, &dlen, op + data_p, 100000);
             data_p += dlen;
         }
         op[data_p++] = ISO2709_FS;

diff --git a/include/idzebra/data1.h b/include/idzebra/data1.h
index 5d2ffdc..ab533a5 100644 (file)
--- a/include/idzebra/data1.h
+++ b/include/idzebra/data1.h
@@ -1,4 +1,4 @@
-/* $Id: data1.h,v 1.24 2007-06-27 22:04:45 adam Exp $
+/* $Id: data1.h,v 1.25 2007-06-27 22:17:20 adam Exp $
    Copyright (C) 1995-2007
    Index Data ApS
 
@@ -100,13 +100,7 @@ typedef struct data1_maptag
     struct data1_maptag *next;
 } data1_maptag;
 
-typedef struct data1_mapunit
-{
-    int no_data;
-    char *source_element_name;
-    data1_maptag *target_path;
-    struct data1_mapunit *next;
-} data1_mapunit;
+typedef struct data1_mapunit data1_mapunit;
 
 typedef struct data1_maptab
 {
@@ -117,6 +111,7 @@ typedef struct data1_maptab
     struct data1_maptab *next;
 } data1_maptab;
 
+
 typedef struct data1_name
 {
     char *name;