Update of /var/cvs/src/org/mmbase/module/core
In directory james.mmbase.org:/tmp/cvs-serv23605
Modified Files:
MMObjectNode.java
Log Message:
I think the comment out of useInputStream was related to MMB-1808. It caused
some test-cases to fail though (which set binary values per byte[]), this is
solved now by changes in BinaryDataType.
See also: http://cvs.mmbase.org/viewcvs/src/org/mmbase/module/core
See also: http://www.mmbase.org/jira/browse/MMB-1808
Index: MMObjectNode.java
===================================================================
RCS file: /var/cvs/src/org/mmbase/module/core/MMObjectNode.java,v
retrieving revision 1.232
retrieving revision 1.233
diff -u -b -r1.232 -r1.233
--- MMObjectNode.java 16 Apr 2009 13:48:09 -0000 1.232
+++ MMObjectNode.java 18 Apr 2009 07:20:20 -0000 1.233
@@ -38,7 +38,7 @@
* @author Eduard Witteveen
* @author Michiel Meeuwissen
* @author Ernst Bunders
- * @version $Id: MMObjectNode.java,v 1.232 2009/04/16 13:48:09 michiel Exp $
+ * @version $Id: MMObjectNode.java,v 1.233 2009/04/18 07:20:20 michiel Exp $
*/
public class MMObjectNode implements org.mmbase.util.SizeMeasurable,
java.io.Serializable, org.mmbase.util.PublicCloneable<MMObjectNode> { //
Comparable<MMObjectNode> {
@@ -781,6 +781,7 @@
}
/*
+ MMB-1808
if (value instanceof InputStream) {
value = useInputStream(fieldName, (InputStream) value);
}
@@ -918,7 +919,8 @@
// cannot return it directly, it would kill the inputstream, and
perhaps it cannot be saved in db anymore then.
// Sad, we have a buffer always now.
// XXX think of something that the buffer is only needed if
actually used a second time
- // help-file, i think
+ // May be we don't have to do this if the
InputStream#markSupported()
+ // We could #reset() then if needed
return new ByteArrayInputStream(useInputStream(fieldName,
(InputStream) value));
}
_______________________________________________
Cvs mailing list
Cvs@lists.mmbase.org
http://lists.mmbase.org/mailman/listinfo/cvs
