How to recover crashed svn repository...

We are using visual svn 2.1.4. It is working on vmware virtual windows 7 on unbuntu linux box. Unfortunately, our electricity infrastructure has a big problem, the electric generator is not activated whenever the electric goes off. Well, this situation is not causing too much problem when our vmware is running on windows box.. But we have changed operating system to linux, it begun to cause problems.. Some files at vmware windows 7 are corrupted..

REPORT of '/svn/ttnetquestuswar/!svn/vcc/default': Could not read chunk size:

We get the above message from svn client, whenever we try to update..

Solution :

1. Stop visual svn server.

2. Backup your repository . You can copy your svn repository to another directory.

3. You should point crashed revision. You can use svnadmin verify command to point the crashed revision.

4. You should dump all revisions until the problematic revision (suppose that it is 1000) with this command ; svnadmin dump -r 1:999 > svndump.txt

5. You should also dump all revisions after the problematic one with this command (suppose the last revision is 2000) ; svnadmin dump --incremental -r 1001:2000 > svndump2.txt

6. Rename your old repository directory..

7. Create a new repository with the same name..

8. Load all changes to the brand new repository with these commands :
svnadmin load < svndump.txt
svnadmin load < svndump2.txt

9. Check your repository with svn verify..

10. Start your visual svn server..

Toad for oracle is not starting..

Well, recently toad is not answering , and I killed its process. But after that, whenever I want to start toad , I get this exception ;

Error reading TtdSQLRecall.HistorySQL: Stream read error

As usual, I googled it, but I could not find anything. As a solution, I have removed SavedSQL.dat file located "user files" directory under toad installation.. It worked for me, I hope it would solve your problem also..

Weblogic 10.3.2 and Hibernate Named Query Problem

If you use Hibernate 3.3.X, 3.4.X with Weblogic 10.3.X , and choose to use named queries , most probably you gonna get below exception ;

org.hibernate.QueryException: ClassNotFoundException: org.hibernate.hql.ast.HqlToken

The problem the versions of antlr.jar , weblogic and hibernate are not compatible. You should put antlr.jar 1.4.X as a first item in weblogic server classpath. You can change startWebLogic.cmd for Windows , startWebLogic.sh for linux file at directory %WEBLOGIC_HOME%/user_projects/domains/%YOUR_DOMAIN%/bin/.

set SAVE_CLASSPATH=%CLASSPATH%

You should change upper line such as for windows environment;

set SAVE_CLASSPATH=%CLASSPATH%
set CLASSPATH=%antlr_path%\antlr.jar;%SAVE_CLASSPATH%

wsimport removes underscores while generating java classes from wsdl

Actually, jaxb is responsible for this behaviour. You should use a binding file to change default behaviour.

Here is the binding file ;


<?xml version='1.0' encoding='UTF-8'?>
<jaxb:bindings
    xmlns:jaxb="http://java.sun.com/xml/ns/jaxb"
    version="2.1">
    <jaxb:globalBindings underscoreBinding="asCharInWord"/>
</jaxb:bindings>

You should  call wsimport such as ;

wsimport -b bindingFile.xml wsdlLocation

How to solve Oracle Cannot access NLS data files or invalid environment specified exception

Recently, I have tried to port a seam application to tomcat from weblogic. I have got below exception...

java.sql.SQLException: ORA-00604: error occurred at recursive SQL level 1
ORA-12705: Cannot access NLS data files or invalid environment specified


After a quite time consuming search, I have found a solution that works for me. My oracle jdbc driver version is 10.2.0.1.0 and name is ojdbc14.jar. Here is the solution ;

 -Duser.language=en  -Duser.region=us

How to display a progressbar during gwt rpc?..

Well, it is a generic non-functional requirement, I have done some research on this item, I have implemented a solution that Thomas Broyer has suggested on gwt group.. This solution has distinct advantage over other solutions, You dont have to change your callback classes, what you have to do is just add a line of code after creation of async gwt-rpc service...

  

    IGwtPersistenceEngineRPCAsync persistenceEngine = GWT.create(IGwtPersistenceEngineRPC.class);
         ((ServiceDefTarget) persistenceEngine).setRpcRequestBuilder(
                          new ProgressRequestBuilder());


    package com.gurselkoca.gwt.client.service;
    import com.allen_sauer.gwt.log.client.Log;
    import com.google.gwt.http.client.Request;
    import com.google.gwt.http.client.RequestBuilder;
    import com.google.gwt.http.client.RequestCallback;
    import com.google.gwt.http.client.Response;
    import com.google.gwt.user.client.rpc.RpcRequestBuilder;
    
    public class ProgressRequestBuilder extends RpcRequestBuilder {
    
     private class RequestCallbackWrapper implements RequestCallback {
    
     private RequestCallback callback;
    
     RequestCallbackWrapper(RequestCallback aCallback) {
     this.callback = aCallback;
     }
    
     @Override
     public void onResponseReceived(Request request, Response response) {
     Log.debug("onResposenReceived is called");
     // put the code to hide your progress bar
     callback.onResponseReceived(request, response);
    
     }
    
     @Override
     public void onError(Request request, Throwable exception) {
     Log.error("onError is called",new Exception(exception));
     // put the code to hide your progress bar
     callback.onError(request, exception);
     }
     }
    
     @Override  
     protected RequestBuilder doCreate(String serviceEntryPoint) {
    
     RequestBuilder rb = super.doCreate(serviceEntryPoint);
     // put the code to show your progress bar
     return rb;  
     }
    
     @Override
     protected void doFinish(RequestBuilder rb) {
     super.doFinish(rb);
     rb.setCallback(new RequestCallbackWrapper(rb.getCallback()));
      
     }
    
    }

How to stop compilation of your gwt project every time you run a GwtTestCase in production (web) mode

When you run a GwtTestCase in production(web) mode , java to javascript compilation is done, whether source files are modified or not. And as already known, java to jaavascript is compilation is very slow. GwtTestCase does not have any option to prevent compilation. I have modified JUnitShell.maybeCompileWebMode method to make compilation depend on VM argument compile. You can download modified JUnitShell from here. You should add -Dcompile=false  to the script responsible running for your GwtTestCase.

How to enable gzip compression for Tomcat?.

You should edit server.xml which is located under conf directory of tomcat installation.
You should add these attributes to connector element.


compression="on" compressableMimeTypes="text/html,text/xml,text/javascript,text/css"

After the modification connector element would be like this ;

<8080" protocol="HTTP/1.1" 

connectionTimeout="20000" redirectPort="8443"

compression="on" compressableMimeTypes="text/html,text/xml,
text/javascript,text/css"/>

SQL Server Datetime and Oracle Timestamp

Recently, I have migrated an oracle database to SQL Server, and I have used SQL Server Datetime for Oracle Timestamp without investigating subtle details of these two types..
Unfortunately, it seems that SQL Server datetime is not suitable for Oracle timestamp, because its precision is not enough to store miliseconds.. Its precision is up to 3.33 miliseconds. SQL Server datetime2 is better choice , its precision is up to 100 nanoseconds.. But you should know that datetime2 is available for SQL Server 2008 and later versions.

How to construct image from byte array for gwt applications

Here is the scenario. You have images stored in database, and you want to use these images in your gwt application. As already known, the gwt image object constructor takes only String , this string should represent the url of the image.. You can  construct the image in the server, and you can save this image to your server's disk, and you can use the url of this image . But , this solution is not elegant, you have to save image to the server's disk..

As a second solution, you can write a servlet whose purpose is to fetch images from database and send them to the clients..  You can map this servlet ImageViewer , with  a  parameter. The request looks like this ;
   /imageViewer?imageId=1234 . You should create your gwt image widget like below;


  Image image = new Image( "/imageViewer?imageId=1234");



 ImageViewer servlet doGet method should be like that ;


public void doGet(HttpServletRequest request, HttpServletResponse response) {
        String imageId =  request.getParameter("imageId");
         byte[] imageData = readImageDataFromDB(imageId);
         response.setContentType("image/png");
         response.getOutputStream().write(imageData,0,imageData.length);
}




As a third solution , you can use the below code to convert image byte array data, to Base64 encoded string.. But you should know that this method does not work for IE7 and the previous versions. IE8 has 32 kb limit for encoded data.


public String getImageData(){
   byte[] imageByteArray = readImageDataFromDB(); //
   String base64 = com.google.gwt.user.server.Base64Utils.toBase64(imageByteArray);
    base64 = "data:image/png;base64,"+base64; 
   return base64;
}

And at gwt layer, you should use the below code ;


 

imageService.getImageData(new AsyncCallback() {

    @Override  
     public void onSuccess(String imageData) {    
        Image image = new Image(imageData);    
         RootPanel.get("image").add(image);  
      }

     @Override 
     public void onFailure(Throwable caught) {
       Window.alert(caught.getMessage()); 
     }

}

JPA 2.0 with GWT

Well, if you are using JPA 2.0 with GWT, and you are using your entity objects as DTO between GWT layer and web layer and you are using annotations for object mapping, you should do the following items ;

• Download jpa2.0 sources . This file does not contain all source files of JPA 2.0 classes, only annotations and some interfaces are included. 
• Put this jar file to your classpath
• Add   <inherits name="javax.Persistence"/>  to  your [module name].gwt.xml file 
• If your gwt version is below 2.1 , then you should add gwt-java-mathh to your project

How to write spring tests without spring transactional management facility

As you know, Spring 3.0 test framework does not allow transaction demercation . You should use beforeTransaction or afterTransaction method annotations to implement test logic after or before the transaction . But it means that you should write one class for every test method. As a result, the number of your test classes increase dramatically. Well, here is the solution , I have used AbstractJUnit4SpringContextTests instead of AbstractTransactionalJUnit4SpringContextTests.

@ContextConfiguration(value = "/applicationContext.xml") 
public class JPASpringTest extends AbstractJUnit4SpringContextTests { 
 @PersistenceContext(unitName="jpadenemelocal") 
 EntityManager entityManager; 
 
 @Autowired 
 protected PlatformTransactionManager transactionManager; 
 
 @Test 
 public void testInsertRolManyToMany() { 
 TransactionStatus status =  transactionManager.getTransaction(null); 
 // your code    
 
 transactionManager.commit(status); 
 } 
 
} 

Maven class loading issues

Recently, I have tried to port one of our project to maven. When I try to use tomcat maven plugin,  I have got the below exception;

java.lang.ClassCastException: tr.com.belsis.common.web.utils.InitServlet cannot be cast to javax.servlet.Servlet at org.apache.catalina.core.StandardWrapper.loadServlet(StandardWrapper.java:1116) at org.apache.catalina.core.StandardWrapper.load(StandardWrapper.java:993)

Multiple javax.servlet.Servlet at class path causes this exception. The solution is to define the scope of servlet dependency as provided as below;

<dependency>
<groupId>javax.servlet</groupId>
<artifactId>servlet-api</artifactId>
<version>2.5</version>
<scope>provided</scope>
</dependency>


Mavenize existing eclipse project

It is very easy, you should just change .project file of your eclipse project..

You should add below xml snippet to under element of .project file.

<buildCommand>
   <name>org.maven.ide.eclipse.maven2Builder</name>
   <arguments>
   </arguments>
  </buildCommand>
   
  
And also you should add below xml snippet under element of .project file.

<nature>org.maven.ide.eclipse.maven2Nature</nature>

edit :

Below snipped should be included .classpath file.

<classpathentry kind="con" path="org.maven.ide.eclipse.MAVEN2_CLASSPATH_CONTAINER">
   <attributes>
    <attribute name="org.eclipse.jst.component.dependency" value="/WEB-INF/lib"/>
   </attributes>
 </classpathentry>





Toplink and Glassfish

If you want to use toplink natively with Glassfish application server , you should make some configuration. First of all, you should change platform class for Glassfish like below, because SunAS9ServerPlatform class constructor expects DatabaseSessionImpl as argument, but it is called with DatabaseSession.

public SunAS9ServerPlatform(DatabaseSession newDatabaseSession) {

       super(newDatabaseSession);

}

Secondly, you should add below ines to your session.xml file.



<server-platform xsi:type="custom-platform">

<server-class>oracle.toplink.platform.server.sunas.SunAS9ServerPlatform</server-class>
<external-transaction-controller-class>oracle.toplink.transaction.sunas.SunAS9TransactionController</external-transaction-controller-class>
</server-platform>
 
 

As a final note, Toplink version is 11.1.1.0.1, Glassfish version is 3.0.1