Mr N

In order to avoid full file-system paths in a Tomcat configuration, you make all the file references point to elements on the classpath. The classpath is an environment variable which is set outside of Tomcat, and is the search path that the class loader uses in order to find classes. When importing another bean definition file you would be far better off using something along the lines of the following, instead of a file:/path/to/file

<import resource="classpath:uk/co/bigsoft/beans.xml"/>

Under Eclipse (3.4 Ganymede (but I’m sure this applies to all versions of Eclipse)) you can link projects together and create project dependencies. Creating a dependency means that when you open a project, it will open the projects it depends on. Also, when you run the project it incorporates the classpath of the dependant projects into its own classpath. If you do this and run Tomcat you will get an error similar to:

class org.springframework.web.context.ContextLoaderListener org.springframework.beans.factory.parsing.BeanDefinitionParsingException: Configuration problem: Failed to import bean definitions from URL location [classpath:uk/co/bigsoft/proj/dino.live.xml] Offending resource: ServletContext resource [/WEB-INF/sitemanager-webapp-config.xml]; nested exception is org.springframework.beans.factory.BeanDefinitionStoreException: IOException parsing XML document from class path resource [uk/co/bigsoft/proj/dino.live.xml]; nested exception is java.io.FileNotFoundException: class path resource [uk/co/bigsoft/proj/dino.live.xml] cannot be opened because it does not exist

This problem occurs because Tomcat is running as a separate application with its own environment - an environment that does not include classpath links to your other Eclipse projects (even though you have linked them together in Eclipse). There follows a step-by-step guide to adding multiple Eclipse projects to Tomcat classpath. This follows on from the Configuring Eclipse to use Tomcat and the Spring Framework article I wrote earlier.

1. After attaching the project to the Tomcat server, double click the Tomcat v6.0 Server at localhost to popup the Tomcat server configuration window.
2. Click General Information->Open launch configuration
3. Select Classpath tab
4. Highlight User Entries (buttons will become enabled)
5. Click Add projects… button
6. Check the boxes of the projects you want to include.
7. Click ok….ok….etc

Been trying to get rid of this error message in my Spring container for a while and I have finally built up enough periphery knowledge to figure out the solution myself, so I thought I’d share it with you - I’m nice like that! The offending message in Eclipse is:

cvc-id.1: There is no ID/IDREF binding for IDREF ‘transactionManager’

After a search I found: http://documentation.softwareag.com/crossvision/ins441/messages/imsg.htm Which provided the very helpful advice:

The document contains an IDREF value which does not exist as an ID value. Modify your document.

In my example it was moaning about the transactionManager but in your case it could be anything. Inside my containers xml configuration I had a bean which uses a reference to a local bean called transactionManager. Local here is the important bit. Local means that it is defined within the scope of this document - not in a linked or imported document, but in this one. In my case it was defined like this:

<bean id="mybeanid" class="uk.co.bigsoft.proj">
  <property name="transactionManager">
    <ref local="transactionManager"/>
  </property>
</bean>

In order to fix the problem, I changed the transactionManager bean reference to not be local:

<bean id="mybeanid" class="uk.co.bigsoft.proj">
  <property
    name="transactionManager"
    ref="transactionManager"/>
</bean>

And the problem went away!

I am officially a SnowHead now. I went on my first Skiing trip with them to Xscape‎, in Milton Keynes which is an indoor real snow dome. SnowHeads is an internet club of skiers who meet up and go skiing together. SnowHeads organises events in the UK and in Europe. So if you like skiing (you don’t even have to be any good), then come along and give it a try. I went with a network architect buddy of mine. On one of the breaks, in the middle, he tells me that the SnowHead people are having trouble with MySQL locking issues inside the SnowHead’s bulletin board system, and if I spoke to their head man he could explain it better. After the event, I went to see the head of SnowHeads to thank him for organising a great evening, and said: “my buddy tells me you are having trouble with your MySQL installation - well I am MySQL-man and could give you a hand". He looks over to my buddy and said: “bloody hell, that is what I call networking". By the sounds of things, Head of SnowHeads and my buddy were discussing the problem together over a beer a couple of days ago, and the next time they met my buddy had produced a fixer as a new member of SnowHeads - Me!

I’ve been doing a lot of work recently with the SpringFramework. It helps with all aspects of building a large scale application by, among other things, providing a generic exception hierarchy which is used by all the component modules. The upshot of this is that I can use Castor or JibX for my XML serialiser without having to change my core code to handle JiBXParseException or a CastorRuntimeException - SpringFramework will abstract away the differences. This is all very handy but when the spring exception is thrown up from the library it contains the original exception (”cause“) which is where all the meaty information lives. Naturally these are specific to the library. During my rounds this one came up when using JibX 1.1.6a, and it took a little while to figure it out:

org.springframework.oxm.jibx.JibxMarshallingFailureException: JiBX marshalling exception: Collection item of type uk.co.bigsoft.proj.event.DeleteEvent has no binding defined; nested exception is org.jibx.runtime.JiBXException: Collection item of type uk.co.bigsoft.proj.event.DeleteEvent has no binding defined at org.springframework.oxm.jibx.JibxUtils.convertJibxException(JibxUtils.java:49) at org.springframework.oxm.jibx.JibxMarshaller.convertJibxException(JibxMarshaller.java:169) at org.springframework.oxm.jibx.JibxMarshaller.marshalWriter(JibxMarshaller.java:193) at org.springframework.oxm.AbstractMarshaller.marshalStreamResult(AbstractMarshaller.java:256) at org.springframework.oxm.AbstractMarshaller.marshal(AbstractMarshaller.java:97) at uk.co.bigsoft.proj.event.TestDeleteEvent.testGet(TestDeleteEvent.java:104) at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method) at sun.reflect.NativeMethodAccessorImpl.invoke(NativeMethodAccessorImpl.java:39) at sun.reflect.DelegatingMethodAccessorImpl.invoke(DelegatingMethodAccessorImpl.java:25) at java.lang.reflect.Method.invoke(Method.java:597) at junit.framework.TestCase.runTest(TestCase.java:168) at junit.framework.TestCase.runBare(TestCase.java:134) at junit.framework.TestResult$1.protect(TestResult.java:110) at junit.framework.TestResult.runProtected(TestResult.java:128) at junit.framework.TestResult.run(TestResult.java:113) at junit.framework.TestCase.run(TestCase.java:124) at junit.framework.TestSuite.runTest(TestSuite.java:232) at junit.framework.TestSuite.run(TestSuite.java:227) at org.junit.internal.runners.JUnit38ClassRunner.run(JUnit38ClassRunner.java:81) at org.eclipse.jdt.internal.junit4.runner.JUnit4TestReference.run(JUnit4TestReference.java:45) at org.eclipse.jdt.internal.junit.runner.TestExecution.run(TestExecution.java:38) at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.runTests(RemoteTestRunner.java:460) at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.runTests(RemoteTestRunner.java:673) at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.run(RemoteTestRunner.java:386) at org.eclipse.jdt.internal.junit.runner.RemoteTestRunner.main(RemoteTestRunner.java:196) Caused by: org.jibx.runtime.JiBXException: Collection item of type uk.co.bigsoft.proj.event.DeleteEvent has no binding defined at uk.co.bigsoft.proj.event.DeleteEvent.JiBX_MungeAdapter.JiBX_jibx_isa_marshal_1_0() at uk.co.bigsoft.proj.event.DeleteEvent.JiBX_jibx_isa_marshal_1_0(EventsContainer.java) at uk.co.bigsoft.proj.event.DeleteEvent.JiBX_jibx_isaEventsContainer_access2.marshal() at uk.co.bigsoft.proj.event.DeleteEvent.marshal(EventsContainer.java) at org.jibx.runtime.impl.MarshallingContext.marshalRoot(MarshallingContext.java:1041) at org.jibx.runtime.impl.MarshallingContext.marshalDocument(MarshallingContext.java:1133) at org.springframework.oxm.jibx.JibxMarshaller.marshalWriter(JibxMarshaller.java:190) … 22 more

…but there is a binding definition for my type X!

<mapping name="eventscontainer"
   class="uk.co.bigsoft.proj.event.EventsContainer">
	
 <collection
   field="events"
   factory="uk.co.bigsoft.proj.event.createEventsContainer">
		
  <structure
    name="isaeventcreate"
    type="uk.co.bigsoft.proj.event.CreateEvent">
   <value name="zone" field="zone"/>
  </structure>
			
 <structure
   name="isaeventmodify"
   type="uk.co.bigsoft.proj.event.ModifyEvent">
  <value name="zone" field="zone"/>
 </structure>

</collection>

In JibX you can serialise collections. The collection may contain different types. In my case they were a set of classes which all implemented the same interface. In JibX you need to define the elements that can appear in a collection, but the collection is by default ordered. When de-serialising the collection it can contain 3 of the first element, 0 of the second, 2 of the next and so on - but they have to be in the order laid out in the binding-mapping. What you can’t do, is have them come in any order. If you want the collection to be able to de-serialise objects in any order, you need to flag it as not ordered. There is a performance hit for this but I think it is worth it. Alternatively, you must make sure that your collection is ordered in the same way as your XML schema definition, which is not always possible. So from the example above add the ordered attribute to the collection elements:

<mapping name="eventscontainer"
   class="uk.co.bigsoft.proj.event.EventsContainer">
	
 <collection
   field="events"
   ordered="false"
   factory="uk.co.bigsoft.proj.event.createEventsContainer">
  :

Finding the differences between files under DOS is a pain, because DOS does not provide any tools to help manipulate files or streams of files.

When I have source trees I need to compare I use the unix power tool diff. Sometimes (because of various editor settings) some of the files will have carriage returns and line feeds (DOS style) and some will have just carriage returns (Unix style). This can make finding differences a bit tricky because every line is different.

There is a set of commands called dos2unix and unix2dos which helps with stripping (or adding respectively) the line feeds, but on some servers it is not installed. Having just done this with 200 files, I thought I'd share the wealth and give it to the world.

for fn in `find . -type f -name "*.java" -print | sort`
do
  echo "$fn"
  sed 's/^M//' $fn > $fn.new
  mv $fn $fn.old
  mv $fn.new $fn
done

Note: that the ^M character is the line feed character. This is obtained by (under bash and most other shells) control+v then press return.

Give things a quick check then delete the originals with:

rm -f `find . -name "*.old"`

Another way to do the same job is to use ed. Create a new file called line-stripper.vi with the following lines:

%s/^M//
wq

Where `^M is the linefeed described above. Then just apply that to every file:

for fn in `ls *.hbm.xml`
do
  cat line-stripper.vi | ed $fn
done

If there are too many files to delete this way then check out my post on rm: Argument list too long.