Details
-
Type:
Bug
-
Status:
Closed
-
Priority:
Major
-
Resolution: Cannot Reproduce
-
Affects Version/s: None
-
Fix Version/s: 1.1.3
-
Component/s: None
-
Labels:None
-
JDK version and platform:jdk 1.5_03 on Windows
Description
I'm new to XStream (tried to use it with mule) and while doing my first experiment I of course stepped on a mine. Looky:
XStream xstream = new XStream();
Object o = Arrays.asList(new Integer[]
);
String xml = xstream.toXML(o);
log.info("xml: " + xml);
Object newObject = xstream.fromXML(xml);
log.info("new object: " + newObject);
gives:
[05-10 22:27:10] INFO XStreamTest: xml: <java.util.Arrays-ArrayList>
<a class="int-array">
<int>42</int>
</a>
</java.util.Arrays-ArrayList>
Exception in thread "main" java.lang.NullPointerException
at java.lang.String.valueOf(Unknown Source)
at java.util.AbstractCollection.toString(Unknown Source)
at java.lang.String.valueOf(Unknown Source)
at java.lang.StringBuffer.append(Unknown Source)
at h2o.playground.XStreamTest.main(XStreamTest.java:24)
It seems XStream has problems properly recognizing the static inner class in java.util.Arrays which is a special kind of ArrayList that doesn't copy the array it contains.
From reading the feature list I'm not sure if this is a bug in XStream; if I need to configure something in a special way or this is just a special case that does not work, please let me know and disregard this bug.
Thanks,
Holger
Activity
Apparently this was a VM bug in 1.5_03; it doesn't happen any longer in 1.5_04 - everything works as expected now.
Holger Hoffstätte
added a comment - Apparently this was a VM bug in 1.5_03; it doesn't happen any longer in 1.5_04 - everything works as expected now. Hello, I work for Sun on the Java VM and tripped across your report and decided to investigate the crash. I reproduced it in all Client VM configurations from 1.4.2 to mustang so it's a previously unreported bug. I've created bug 6317220 to track this issue in case you're interested.
Tom
Tom Rodriguez
added a comment - Hello, I work for Sun on the Java VM and tripped across your report and decided to investigate the crash. I reproduced it in all Client VM configurations from 1.4.2 to mustang so it's a previously unreported bug. I've created bug 6317220 to track this issue in case you're interested.
Tom Seems to work now also for current 1.4.2 releases. Unit test fixture added to CollectionsTest.
Jörg Schaible
added a comment - Seems to work now also for current 1.4.2 releases. Unit test fixture added to CollectionsTest. 
Some more info. It turns out that the object is re-created properly, but for some reason toString() fails. I suspected that it's related to my using eclipse 3.0.2 in 1.4-mode with jdk 1.5 but that doesn't seem to be the case; it happens when compiled from the command line, too.
Interestingly enough, the following:
List newObject = (List)xstream.fromXML(xml);
log.info("new.size: " + newObject.size());
log.info("new(0):" + newObject.get(0));
gives:
INFO: new.size: 1
Exception in thread "main" java.lang.NullPointerException
at java.lang.String.valueOf(Unknown Source)
at java.lang.StringBuilder.append(Unknown Source)
at h2o.playground.XStreamTest.main(XStreamTest.java:28)
so it seems the problen is the List's internal array. It gets better though:
Object first = newObject.get(0);
log.info("first.class:" + first.getClass());
gives the best fireworks:
INFO: new.size: 1
#
#
#
#
#
#
..which indicates that the internal array is not null and contains something really bogus. I didn't investigate the stack trace further