When you already did something with B2B11g, then you probably already know how to connect to the B2B UI, and have found out that it's not available by default.

By default the B2B UI is not targetted to a managed server. It can be done very easy in the weblogic console (http://adminserver-host:port/console). Go to Deployments and look for b2bui, when found click on it.

Then check the boxes of the  b2bui components and click on "Change Targets" (this button will activate after checking the boxes).
Then in the next screen click yes:
 And after a while the changes are applied:
After this the b2b console can be addressed via: http://managedserver-host:port/b2bconsole.

Recently I renamed some docTypes in B2B11g because of a typo. Last week I wanted to export the repository to put it into subversion and to be able to propagate the implementation to the test environment.

But when exporting I ran into the following exception:

MDS-00001: exception in Metadata Services layer MDS-00511: failure to create document /soa/b2b/Custom/DocVersion_V1.2/MyAcknowledgementDocType /MyAcknowledgement /MyAcknowledgement.xsd 
\DocVersion_V1.2\MyAcknowledgementDocType \MyAcknowledgement \MyAcknowledgement_1p2.xsd 
C:\Windows\TEMP\Export817535\soa\b2b\Custom\DocVersion_V1.2\MyAcknowledgementDocType \MyAcknowledgement \MyAcknowledgement_1p2.xsd (The system cannot find the path specified)

What I actually did in the renaming was to create a new doctype with docdefinition, along the incorrect ones, with copy/pasting the oldnames and changing the typo (...Acknowlegdement... into ...Acknowledgement...). But apparently I got an extra trailingspace in my new name.  It appeared in two docTypes.

I removed the deployments of the particular agreements, removed the document registrations at the Trading Partners (both host and remote) and then removed the DocType under the Administration/Documents menu.

After that I could create the documentTypes again (now payed extra attention to possible trailing spaces), recreated the document registration at the partners and the agreements. After that I could export the repository succesfully again.

Bottom line: you apparently can have trailing spaces in the names, but this will lead into export errors...

In the past I wrote an article about high availability architecture for Oracle Integration B2B10g.
Currently I'm working on an AS2 implementation with Oracle B2B11g. B2B11g is now part of SOASuite11g. Although it can be seen as a gateway product, much like OSB, since it is part of SOASuite I would not recommend to place B2B in de DMZ (Demilitarized Zone).

So what you need is a proxy and reverse proxy in the DMZ to forward messages from B2B to your trading partner (forward proxy) and route messages from your trading partner to B2B (reverse proxy).

The idea in our setup was actually to use Microsoft IIS for this. The reverse proxy was succesfull implementated, where it also functions as a SSL-decoder. But the IIS-expert had difficulties to turn the forward-proxy functioning on. Now that should not be to hard, but it need to have proxy, and proxy-connect (for proxying SSL-requests) installed. Apparently those are seperate installs for IIS. You may conclude: me know nothing about IIS. Also, the IIS-expert was not around very much and we needed a proxy in place pretty much immediatly.

So I installed Apache for that. I installed a Apache 2.0.x. Not the latest  release, I know, because I hoped to be able to re-use my settings from my earlier implementation. However, those turned out to be from 1.3 (forgot about that).

Anyway I had to change my settings a little.

First you need to turn some modules (mod_proxy, mod_proxy_http and mod_proxy_connect) on in the http.conf

#2012-10-31, M. van den Akker, Darwin-IT Professionals
# mod_proxy en mod_proxy_http needed for proxy
LoadModule proxy_module modules/mod_proxy.so
LoadModule proxy_connect_module modules/mod_proxy_connect.so
LoadModule proxy_http_module modules/mod_proxy_http.so

The proxy-connect module is needed to be able to proxy for SSL connections.
Since I like to seperate my customsettings from the other ones I created a separate config file for the proxy-settings. This I include at the end of the http.conf:

#2012-10-31, M. van den Akker, Darwin-IT Professionals
#Include proxy config
Include conf/proxy.conf

Then of course the proxysettings in the proxy.conf file:

#2012-10-31, M. van den Akker, Darwin-IT Professionals

Listen 1234
NameVirtualHost proxy-web-b2b:1234
ProxyRequests On
ProxyVia On
AllowCONNECT 443 8443 
<Proxy *>
  #Order deny,allow
  #Deny from all
  Allow from all
</Proxy> 
<VirtualHost proxy-web-b2b:1234>
  ServerAdmin webmaster@proxy-web-b2b
  ErrorLog logs/error_proxyfwd1234.log
  CustomLog logs/access_proxyfwd1234.log common
</VirtualHost>

The AllowCONNECT setting denotes the ports that are allowed for SSL-Connect forwards.

Since we use this temporarly only for forward proxying, I don't have the Apache 2.0 settings for reverse proxying at hand. Since B2B11g is the initiator of the https-connection it needs to have the SSL certificates of remote tradingpartners, besides the AS2 encryption/signing certificates in the keystore. So as is done in our case it does not make much sense to have the reverse-proxy (IIS in our implementation) do the un-SSL-ing (decoding). Then you'll need to maintain those certificates in two keystores.


For the forward proxy the weblogic server of B2B needs to be started with the following settings:

• -Dhttp.proxySet=true
• -Dhttp.proxyHost=proxy-web-b2b
• -Dhttp.proxyPort=1234

In the B2B configuration on the delivery channels the checkbox "Use Proxy" needs to be checked.

That's about it.

Did you encounter it as well? You're working wih a SOA-project within JDeveloper 11g and suddenly you see that there are componenttypes in your project (within the application navigator) that do not belong to your project. Componenttypes of  mostly BPEL components that are from other projects, even in other applications (.jws).

I encountered this in both Jdeveloper 11g PS4 (11.1.1.5 ) and PS5 (11.1.1.6). But probably because I work with those now more intensively. Probably this occurs in earlier versions too.

Now, I found that it is probably a bug in JDeveloper because sometimes I work within a component, say a BPEL process, and go in the application navigator to another project to look something up. To see how I did it in that project or how a particular xsd looks like. Then I work further, but the Application Navigator is open in that other project. JDeveloper apparently looses track of the actual source folder and writes the changes in my component in to that other project. I assume it is when you add a partnerlink, eg. an adapter, where the component type has to be changed with references.

Until now I just deleted those "wandering" component types. And found no problems or inconsistencies in the original/correct componenttypes within their projects.
But I hope this bug is solved.
Maybe someone that reads this and can file a bug within Oracle could do so? I would be greatfull. For the rest be carefull...

A few years ago I was very proud that I managed to call java from BPEL through WSIF. I think that the method pretty much should work in 11g. But my method back than used the deprecated tool Schema Compile. So I would need to figure out how to do that with JAXB or something. But now in the later patchsets of Weblogic/SOASuite there is a far more nicer method: Spring. Now, my experience with Spring starts since yesterday... Spring is a method to embed simple java objects (Plain Old Java Objects: POJO's) into an application server context. In SOASuite it is implemented in a real cool way, so that you can embed them in a real declarative way. If you google your way around you can find several examples on Spring and SOASuite. Documentation you can find here. And Edwin Biemond created an article about it. But unfortunately, it did not become clear to me how I had to implement my setup.

In my previous article I explained how to call Oracle Reports from java. I want to call that from BPEL or Mediator. Earlier I thought about the Socket adapter or the HTTP service-reference. But those options failed because the examples I found expect XML as an input (REST), where I needed to call an URL with parameters. Or they prevented me to call a url on port 80 (on which our reportservers listen...).
In that article you read that I had a service, and the input is a bean and the response is a bean. The reason of that was that I needed to pass multiple attributes as parameter and receive multiple attributes as a response. In reallife you might have a service with a bean structure as a request and response that reflect a complete hierarchical XML structure. The examples I found just talked about classes with elementary datatypes as  parameters and results. So how to set up your spring context to be able to do your call? At first I thought I needed just one context for the service. But it turns out that you have to create a context for every bean that plays a part in the interface (basically your service and every bean in the call parameter and response class structure).

To be able to expose the java classes to a WSDL (that is generated for you when you wire the service to BPEL) you have to generate Interfaces on your classes. So building further on my previous story, that would be on the classes ReportRequest, ReportResponse and CallReportsService.
This can be generated for you as follows. Open the ReportRequest class and right click in the source, choose Refactor and then Extract Interface:

In my previous article I put a method in my request bean to construct the Reports Server url (getReportsURL()). This method I don't want to have exposed in my WSDL. So I did not select that in the list. You can do the same for the other two classes.

Now it's time to declare the Spring-bean-structure. Open the composite editor and right-click:
When you choose Insert SpringContext, you might get a "Spring Not Available" error:
Go to Help->Check for Updates and search for the Spring & Weblogic SCA Addon:
Add it and restart JDeveloper.

Now you can add a SpringContext to your composite:
 Edit the Spring Context (CallReportsServer.xml). Then in the component palette you can look for the Weblogic SCA tab, and drag a service into the composite:

Name it CallReportsService, and as a target the CallReportsServiceBean. As a type you can browse to your classes, choose here the interface ICallReportsService.
Now you can add a bean to the context. The tag for this can be found on the "Spring 2.5 Core" page. But you can also type it in of course:
 Then you have to define it. Convenient is the Inspector palette:

 The Id of the bean should be the same as the target of the service: CallReportsServiceBean.
The class can be browsed the same way as the type in the Service. Choose here the implementation class "CallReportsService".

Now we can do the same for the other two beans. For the two input/output beans (ReportRequest and ReportResponse) I choose the bean implementation itself as the type. In my first setup I encountered that at Runtime SOASuite tries to call the public no-parameter-constructor of the type. Since an interface does not have a constructor, it failed. So I replaced that with the implementation bean.

If you have the three beans you can tie them together:
This gets you the following context definitions.
CallReportsServer.xml:

<?xml version="1.0" encoding="windows-1252" ?>
<beans xmlns="http://www.springframework.org/schema/beans"
       xmlns:util="http://www.springframework.org/schema/util"
       xmlns:jee="http://www.springframework.org/schema/jee"
       xmlns:lang="http://www.springframework.org/schema/lang"
       xmlns:aop="http://www.springframework.org/schema/aop"
       xmlns:tx="http://www.springframework.org/schema/tx"
       xmlns:sca="http://xmlns.oracle.com/weblogic/weblogic-sca"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.5.xsd http://www.springframework.org/schema/util http://www.springframework.org/schema/util/spring-util-2.5.xsd http://www.springframework.org/schema/aop http://www.springframework.org/schema/aop/spring-aop-2.5.xsd http://www.springframework.org/schema/jee http://www.springframework.org/schema/jee/spring-jee-2.5.xsd http://www.springframework.org/schema/lang http://www.springframework.org/schema/lang/spring-lang-2.5.xsd http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-2.5.xsd http://www.springframework.org/schema/tool http://www.springframework.org/schema/tool/spring-tool-2.5.xsd http://xmlns.oracle.com/weblogic/weblogic-sca META-INF/weblogic-sca.xsd">
  <!--Spring Bean definitions go here-->
  <sca:service name="CallReportsService" target="CallReportsServiceBean"
               type="nl.darwinit.soa.callreportsservice.ICallReportsService"/>
  <bean id="CallReportsServiceBean"
        class="nl.darwinit.soa.callreportsservice.CallReportsService"/>
  <sca:reference type="nl.darwinit.soa.callreportsservice.ReportRequest"
                 name="ReportRequest.ReportRequest"/>
  <sca:reference type="nl.darwinit.soa.callreportsservice.ReportResponse"
                 name="ReportResponse.ReportResponse"/>
</beans>

ReportRequest.xml:

<?xml version="1.0" encoding="windows-1252" ?>
<beans xmlns="http://www.springframework.org/schema/beans"
       xmlns:util="http://www.springframework.org/schema/util"
       xmlns:jee="http://www.springframework.org/schema/jee"
       xmlns:lang="http://www.springframework.org/schema/lang"
       xmlns:aop="http://www.springframework.org/schema/aop"
       xmlns:tx="http://www.springframework.org/schema/tx"
       xmlns:sca="http://xmlns.oracle.com/weblogic/weblogic-sca"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.5.xsd http://www.springframework.org/schema/util http://www.springframework.org/schema/util/spring-util-2.5.xsd http://www.springframework.org/schema/aop http://www.springframework.org/schema/aop/spring-aop-2.5.xsd http://www.springframework.org/schema/jee http://www.springframework.org/schema/jee/spring-jee-2.5.xsd http://www.springframework.org/schema/lang http://www.springframework.org/schema/lang/spring-lang-2.5.xsd http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-2.5.xsd http://www.springframework.org/schema/tool http://www.springframework.org/schema/tool/spring-tool-2.5.xsd http://xmlns.oracle.com/weblogic/weblogic-sca META-INF/weblogic-sca.xsd">
  <!--Spring Bean definitions go here-->
  <sca:service name="ReportRequest" target="ReportRequestBean"
               type="nl.darwinit.soa.callreportsservice.ReportRequest"/>
  <bean id="ReportRequestBean"
        class="nl.darwinit.soa.callreportsservice.ReportRequest"/>
</beans>

ReportResponse.xml:

<?xml version="1.0" encoding="windows-1252" ?>
<beans xmlns="http://www.springframework.org/schema/beans"
       xmlns:util="http://www.springframework.org/schema/util"
       xmlns:jee="http://www.springframework.org/schema/jee"
       xmlns:lang="http://www.springframework.org/schema/lang"
       xmlns:aop="http://www.springframework.org/schema/aop"
       xmlns:tx="http://www.springframework.org/schema/tx"
       xmlns:sca="http://xmlns.oracle.com/weblogic/weblogic-sca"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.5.xsd http://www.springframework.org/schema/util http://www.springframework.org/schema/util/spring-util-2.5.xsd http://www.springframework.org/schema/aop http://www.springframework.org/schema/aop/spring-aop-2.5.xsd http://www.springframework.org/schema/jee http://www.springframework.org/schema/jee/spring-jee-2.5.xsd http://www.springframework.org/schema/lang http://www.springframework.org/schema/lang/spring-lang-2.5.xsd http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-2.5.xsd http://www.springframework.org/schema/tool http://www.springframework.org/schema/tool/spring-tool-2.5.xsd http://xmlns.oracle.com/weblogic/weblogic-sca META-INF/weblogic-sca.xsd">
  <!--Spring Bean definitions go here-->
  <sca:service name="ReportResponse" target="ReportResponseBean"
               type="nl.darwinit.soa.callreportsservice.ReportResponse"/>
  <bean id="ReportResponseBean"
        class="nl.darwinit.soa.callreportsservice.ReportResponse"/>
</beans>

Well, actually that is about it. Now lets create a BPEL process.


I changed the xsd to:

<?xml version="1.0" encoding="UTF-8"?> 
<schema attributeFormDefault="unqualified"
 elementFormDefault="qualified"
 targetNamespace="http://xmlns.darwin-it.nl/CallReportsService/CallReportsService/CallReportsService"
 xmlns="http://www.w3.org/2001/XMLSchema" xmlns:xsd="http://www.w3.org/2001/XMLSchema" >
 <element name="process">
  <complexType>
   <sequence>
    <xsd:element name="aanvraagID" type="xsd:string" minOccurs="0"/>
    <xsd:element name="destination" type="xsd:string" minOccurs="0"/>
    <xsd:element name="httpTimeOut" type="xsd:int" minOccurs="0"/>
    <xsd:element name="jobName" type="xsd:string" minOccurs="0"/>
    <xsd:element name="previewJN" type="xsd:string" minOccurs="0"/>
    <xsd:element name="reportName" type="xsd:string" minOccurs="0"/>
    <xsd:element name="reportsBaseUrl" type="xsd:string" minOccurs="0"/>
    <xsd:element name="userId" type="xsd:string" minOccurs="0"/>
   </sequence>
  </complexType>
 </element>
 <element name="processResponse">
  <complexType>
   <sequence>
    <xsd:element name="errorCode" type="xsd:string" minOccurs="0"/>
    <xsd:element name="errorMessage" type="xsd:string" minOccurs="0"/>
    <xsd:element name="jobId" type="xsd:string" minOccurs="0"/>
    <xsd:element name="jobStatus" type="xsd:string" minOccurs="0"/>
    <xsd:element name="jobStatusCode" type="xsd:string" minOccurs="0"/>
    <xsd:element name="responseXml" type="xsd:string" minOccurs="0"/>
   </sequence>
  </complexType>
 </element>
</schema>

Now we can tie the CallReportServer component to the BPEL Component:

 After the compilation, the wiring has to be redone, which will result in the message that the wsdl is created:
Now you can implement the BPEL Process.
Since I choose the ReportRequest implementation bean as a type in the ReportRequest Spring Context, it turns out that it generated an element for the reportURL getter. You could remove that from the wsdl, but since it is not exposed as a service, you can also leave it that way. Just don't map it to any input.

Now, you would think you're done. But there is one little caveat in the deployment of the spring contexts. We tied the beans neatly together. But the Spring Context of the CallReportServer.xml expects in the references Interfaces as a type. So change those to the interfaces of the beans:

<?xml version="1.0" encoding="windows-1252" ?>

<beans xmlns="http://www.springframework.org/schema/beans"
       xmlns:util="http://www.springframework.org/schema/util"
       xmlns:jee="http://www.springframework.org/schema/jee"
       xmlns:lang="http://www.springframework.org/schema/lang"
       xmlns:aop="http://www.springframework.org/schema/aop"
       xmlns:tx="http://www.springframework.org/schema/tx"
       xmlns:sca="http://xmlns.oracle.com/weblogic/weblogic-sca"
       xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
       xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-2.5.xsd http://www.springframework.org/schema/util http://www.springframework.org/schema/util/spring-util-2.5.xsd http://www.springframework.org/schema/aop http://www.springframework.org/schema/aop/spring-aop-2.5.xsd http://www.springframework.org/schema/jee http://www.springframework.org/schema/jee/spring-jee-2.5.xsd http://www.springframework.org/schema/lang http://www.springframework.org/schema/lang/spring-lang-2.5.xsd http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-2.5.xsd http://www.springframework.org/schema/tool http://www.springframework.org/schema/tool/spring-tool-2.5.xsd http://xmlns.oracle.com/weblogic/weblogic-sca META-INF/weblogic-sca.xsd">
  <!--Spring Bean definitions go here-->
  <sca:service name="CallReportsService" target="CallReportsServiceBean"
               type="nl.darwinit.soa.callreportsservice.ICallReportsService"/>
  <bean id="CallReportsServiceBean"
        class="nl.darwinit.soa.callreportsservice.CallReportsService"/>
  <sca:reference type="nl.darwinit.soa.callreportsservice.IReportRequest"
                 name="ReportRequest.ReportRequest"/>
  <sca:reference type="nl.darwinit.soa.callreportsservice.IReportResponse"
                 name="ReportResponse.ReportResponse"/>
</beans>

(nl.darwinit.soa.callreportsservice.ReportRequest changed to nl.darwinit.soa.callreportsservice.IReportRequest, same for IReportResponse)

Now you can deploy and test.

As you can see, it's a very declarative way to create a webservice on a java class.

I wrapped up my complete project here.

In my current project we're replacing Oracle Workflow with SOASuite/BPEL 11g. One of the problems turn out to be Oracle Reports in batch. There is a procedure in the database that queries for all requests for which an Oracle Report has to be run. For each of them it calls another procedure that constructs the URL to the Reports server and invokes it. The response is searched for a status, for an OK status.

Since potentially there can be hundreds or thousands of letters to be printed for requests, this procedure can run for a very long hours. Up to maybe three hours or so. From Oracle Workflow this is no problem at all, since it runs in the database in a background job itself. Normally you have one or more background engines running in dbms_jobs. The OWF processes can be rebuild in BPEL as easily. The batch procedure is then called through the database adapter. But since it can run for hours (potentially) the database-session needs to be open all the time. The inherent synchronous database adapter needs a time out of hours. And even then you might run into a time-out. So for BPEL this is not a good architecture.

There are several options to solve it. We could make the pl/sql service asynchronous, by creating a request and reply queue in Advanced Queueing. Then subscribe and register a pl/sql function that dequeues the message, calls the batch procedure and queues the reply message to the reply queue, using a correlation id. BPEL enqueues on the request queue a message to request for a batch run. And the wait for the correlated reply message. A good explanation on how to set that up is here on Ask Tom.

But I thought it would be nicer to call the Reports right from the middle tier. Then I would need to query the Requests for which to print in one database call, and for each one call the Oracle Report one by one. (Did you know that BPEL 2.0 had a very nice For-loop construction?)

So I needed to be able to call Reports from Java.

My Basic example I found here, is:

import java.net.*;
import java.io.*;

public class URLConnectionReader {
    public static void main(String[] args) throws Exception {
        URL yahoo = new URL("http://www.yahoo.com/");
        URLConnection yc = yahoo.openConnection();
        BufferedReader in = new BufferedReader(
                                new InputStreamReader(
                                yc.getInputStream()));
        String inputLine;

        while ((inputLine = in.readLine()) != null)
            System.out.println(inputLine);
        in.close();
    }
}

My first technical working example was:

   CallReportsService callReportsService = new CallReportsService();
        StringBuffer reportsUrl = new StringBuffer("http://reportsserver.darwin-it.local/reports/rwservlet?outputimageformat=gif&statusformat=xml&server=rep_dev_darwin&");
        reportsUrl.append("report=");
        String reportName = "DWN1234R";
        reportsUrl.append(reportName);
        reportsUrl.append("&P_PREVIEW=");
        String preview = "J";
        reportsUrl.append(preview);
        reportsUrl.append("&P_AVR_ID=");
        String aanvraagID = "7249834";
        reportsUrl.append(aanvraagID);
        reportsUrl.append("&JOBNAME=");
       
       
        String jobName = "DWN1234R_PREVIEW";
        reportsUrl.append(jobName);
                                         
                
        URL reportSvrURL = new URL(reportsUrl.toString());
       
        URLConnection reportSvrConnection = reportSvrURL.openConnection();
        reportSvrConnection.setConnectTimeout(60000);
        BufferedReader in = new BufferedReader(
                                new InputStreamReader(
                                reportSvrConnection.getInputStream()));
        String inputLine;

        while ((inputLine = in.readLine()) != null)
            System.out.println(inputLine);
        in.close();
    }"

Now eventually I want to have this running in a Spring context (had to figure that out too). So I want a Request and a Response bean. Since the response of the Reports server is in XML format, I want to have a ResponseParser. And then of course the service itsself.

The main runReport method of the CallReportsService is as follows:

    /**
     * Run a Report on the Report Server
     * @param repReq
     * @return ReportResponse
     */
    public ReportResponse runReport(ReportRequest repReq) {
        StringBuffer strBuf = new StringBuffer();
        ReportRequest repRequest = repReq;//(ReportRequest)iRepReq;
        
        ReportResponse repResp = null;
        URL reportSvrURL;
        try {
            reportSvrURL = new URL(repRequest.getReportsUrl());

            URLConnection reportSvrConnection = reportSvrURL.openConnection();
            reportSvrConnection.setConnectTimeout(repRequest.getHttpTimeOut());
            BufferedReader in =
                new BufferedReader(new InputStreamReader(reportSvrConnection.getInputStream()));
            String inputLine;
            do {
                inputLine = in.readLine();
                if (inputLine != null) {
                    strBuf.append(inputLine);
                    strBuf.append('\n');
                }
            } while (inputLine != null);
            in.close();
            ReportResponseParser repParser =
                new ReportResponseParser(strBuf.toString());
            repResp = repParser.getReportResponse();
        } catch (MalformedURLException e) {
            if (repResp == null)
                repResp = new ReportResponse();
            repResp.setErrorCode(EC_MALFORMEDURL);
            repResp.setErrorMessage("MalformedURLException: " +
                                    e.getLocalizedMessage());
        } catch (IOException e) {
            if (repResp == null)
                repResp = new ReportResponse();
            repResp.setErrorCode(EC_IO);
            repResp.setErrorMessage("IOException: " + e.getLocalizedMessage());
        }
        return repResp;
    }

It uses a request bean as an input. The request bean contains the method to build up a request url:

    /**
     * Build a URL-String from the paratamers in the request
     */
    public String getReportsUrl() {
        StringBuffer reportsUrl = new StringBuffer(getReportsBaseUrl());
        reportsUrl.append("report=");
        reportsUrl.append(getReportName());
        reportsUrl.append("&userid=");
        reportsUrl.append(getUserId());
        reportsUrl.append("&destination=");
        reportsUrl.append(getDestination());
        reportsUrl.append("&P_PREVIEW=");
        reportsUrl.append(getPreviewJN());
        reportsUrl.append("&P_AVR_ID=");
        reportsUrl.append(getAanvraagID());
        reportsUrl.append("&JOBNAME=");
        reportsUrl.append(getJobName());
        return reportsUrl.toString();
    }

The attribute "reportsBaseUrl" of the bean should contain the base URL to the reports server: "http://reportsserver.darwin-it.local/reports/rwservlet?outputimageformat=gif&statusformat=xml&server=rep_dev_darwin&"

To parse the XML content I created a few helper methods in the ReportResponseParser class. The first one is the actual parser:

        /**
     * Parse the Response XML String
     */
    private void parseRespXmlStr() {
        try {
            InputSource inputSource = getRespXMLAsInputSource();
            if (inputSource != null) {
                DocumentBuilder docBuilder;
                docBuilder = newDocBuilder();
                Document doc;
                doc = docBuilder.parse(inputSource);
                setDoc(doc);
            }
        } catch (ParserConfigurationException e) {
            setErrorMsg("ParserConfigurationException: " +
                        e.getLocalizedMessage());
            setErrorCode(EC_PARSING);
        } catch (SAXException e) {
            setErrorMsg("SAXException: " + e.getLocalizedMessage());
            setErrorCode(EC_PARSING);
        } catch (IOException e) {
            setErrorMsg("IOException: " + e.getLocalizedMessage());
            setErrorCode(EC_PARSING);
        }
    }

It needs an InputStream. So the String object that is the result from the ReportsServer is to be read into an InputSource:

   /**
     * Create an InputSource/InputStream for the response XML String
     * @return
     */
    private InputSource getRespXMLAsInputSource() {
        String respXMLStr = getResponseXMLStr();
        InputSource inputSource = null;
        if (respXMLStr != null) {
            inputSource = new InputSource();
            inputSource.setCharacterStream(new StringReader(respXMLStr));
        }
        return inputSource;
    }

Possibly, since I catch all the lines in the response into a StringBuffer, I could do it from there as well. But these methods I had "on the shelf". Then I need a DocumentBuilder that does the parsing.

    /**
     * Create a new DocumentBuilder
     * @return
     * @throws ParserConfigurationException
     */
    private DocumentBuilder newDocBuilder() throws ParserConfigurationException {
        DocumentBuilderFactory domFactory =
            DocumentBuilderFactory.newInstance();
        domFactory.setNamespaceAware(true);
        DocumentBuilder docBuilder = domFactory.newDocumentBuilder();
        return docBuilder;
    }

To evaluate an xpath expression on the XML Document I need the next method:

    /**
     * Evaluate xpath expression
     *
     * @param xpathExpr the xpath expression
     * @param returnType the return type that is expected.
     * http://www.ibm.com/developerworks/library/x-javaxpathapi.html:
     * XPathConstants.NODESET => node-set maps to an org.w3c.dom.NodeList
     * XPathConstants.BOOLEAN => boolean maps to a java.lang.Boolean
     * XPathConstants.NUMBER => number maps to a java.lang.Double
     * XPathConstants.STRING => string maps to a java.lang.String
     * XPathConstants.NODE
     *
     * @throws XPathExpressionException
     */
    public Object evaluate(String xpathExpr,
                           QName returnType) throws XPathExpressionException {
        XPathFactory factory = XPathFactory.newInstance();
        XPath xpath = factory.newXPath();
        /*XMLNSResolver nsRes = getNsRes();
           if (nsRes != null) {
               xpath.setNamespaceContext(nsRes);
           }*/
        XPathExpression expr = xpath.compile(xpathExpr);
        Document doc = getDoc();
        Object resultObj = expr.evaluate(doc, returnType);
        return resultObj;
    }

Now, the result is an Object. And we expect a NodeList:

    /**
     * Select Nodes using Xpath
     *
     * @param xpath
     * @return NodeList
     * @throws XPathExpressionException
     */
    public NodeList selectNodes(String xpath) throws XPathExpressionException {
        NodeList nl = (NodeList)evaluate(xpath, XPathConstants.NODESET);
        return nl;
    }

The queries I want to perform on the response are basically single strings. So the next method just gets the simple text content of the queried node:

    /**
     * Get the text content of a node.
     * @param xpath
     * @return
     */
    public String getXpathTextContent(String xpath) {
        String result = null;
        try {
            NodeList nl = selectNodes(xpath);
            if (nl.getLength() > 0) {
                result = nl.item(0).getNodeValue();
            }
        } catch (XPathExpressionException e) {
            setErrorMsg("XPathExpressionException: " +
                        e.getLocalizedMessage());
            errorCode = EC_XPATH;
            setErrorCode(errorCode);
        }
        return result;
    }

Then you can querie all the relevant elements as follows:

    /**
     * Get the Error Code from the Reponse
     * @return
     */
    public String getRespErrorCode() {
        String errorCode = getXpathTextContent("//serverQueues/error/@code");
        return errorCode;
    }

    /**
     * Get the Error Message from the Reponse
     * @return
     */
    public String getRespErrorMsg() {
        String errorMsg = getXpathTextContent("//serverQueues/error/@message");
        return errorMsg;
    }
    /**
     * Get the job Id from the Reponse
     * @return
     */
    public String getRespJobId() {
        String errorMsg = getXpathTextContent("//serverQueues/job/@id");
        return errorMsg;
    }
    /**
     * Get the job status from the Reponse
     * @return
     */
    public String getRespJobStatus() {
        String errorMsg = getXpathTextContent("//serverQueues/job/status");
        return errorMsg;
    }  
    /**
     * Get the job status Codefrom the Reponse
     * @return
     */
    public String getRespJobStatusCode() {
        String errorMsg = getXpathTextContent("//serverQueues/job/status/@code");
        return errorMsg;
    }
    /**
     * Parse the responseXML into  a ReportResponse bean
     * @return
     */
    ReportResponse getReportResponse() {
        ReportResponse repResp = new ReportResponse();
        repResp.setErrorCode(this.getRespErrorCode());
        repResp.setErrorMessage(this.getRespErrorMsg());
        repResp.setResponseXml(this.getResponseXMLStr());
        repResp.setJobId(this.getRespJobId());
        repResp.setJobStatus(this.getRespJobStatus());
        repResp.setJobStatusCode(this.getRespJobStatusCode());
        return repResp;
    }

In the end it is pretty simple. Most of the code is to be able to call the service with a nice Request bean and  get a response bean back. You could of course get the fields you want using string-manipulation methods. But I found the xml-parsing a more neat method. If you need more info from the response it is easy to query them. Tip: use the xpath-search (from the Search menu) tool from JDeveloper to test your xpath queries.
A tipical Reports response xml is:

<?xml version = '1.0' encoding = 'UTF-8' standalone = 'yes'?>
<serverQueues>
   <job id="5159" queueType="past">
      <name>DWN1234R_PREVIEW</name>
      <type>report</type>
      <status code="4">Rapport  is voltooid.</status>
      <owner>RWUser</owner>
      <server>rep_dev_darwin</server>
      <destination>
         <desType>unknown</desType>
         <desFormat>dflt</desFormat>
      </destination>
      <timingInfo>
         <queued>11-sep-2012 12:45:14</queued>
         <started>11-sep-2012 12:45:14</started>
         <finished>11-sep-2012 12:45:14</finished>
      </timingInfo>
   </job>
</serverQueues>

I put my classes in a SOA project. Because the next step is to create a spring context for them. You can download the project here.

One of my current customers is in the transition of migrating from sending/receiving EDI messages to XML over AS2. The choice of B2B-product is Oracle SoaSuite-Integration B2B11g.

Nice, because in the past I was intensively involved in implementing ebMS with B2B10g. I think the first implementation on Oracle Integration B2B 10g with ebMS in the Netherlands. Unfortunately I was not able to get message encryption and signing working back then. Mainly because of lack of time and lack of knowledge on different certificate formats. Now I had the change to re-try quite the same thing in B2B 11g.

AS2 and ebMS are regarding implementation basically the same: both are transport protocols over HTTP(s). I had to implement SSL and Message Encription and Signing.

My colleague had already found a good starting point: the blog of Anuj Dwivedi on Enabling SSL on Oracle B2B 11g. It's a nice step by step "how-to" to implement transport security.

Unfortunately I missed 2 aspects in the story: how to import a private certificate in a ".p12" file and how to convert a certificate in DER format to PEM. My customer has a certificate in pkcs12 (a .p12 file) format and that one needs to be imported. Then the public certicate is not in PEM format, and that needs to be to get it imported in the JKS keystore using the keytool as described in the blog.

Import private certificate in the keystore

This how-to I found here. This is done by first creating an empty key store:

keytool -genkey -alias darwinKey -keystore ~/Keystores/b2bKeyStore.jks -
keypass welcome1 -storepass welcome1

This creates a JKS-keystore with a key with alias darwinKey. It asks several questions like:

What is your first and last name?
[Unknown]: Martien van den Akker
What is the name of your organizational unit?
[Unknown]: Professionals
What is the name of your organization?
[Unknown]: Darwin-IT
What is the name of your City or Locality?
[Unknown]: Amersfoort
What is the name of your State or Province?
[Unknown]: Utrecht
What is the two-letter country code for this unit?
[Unknown]: NL
Is CN=Martien van den Akker, OU=Professionals, O=Darwin-IT, L=Amersfoort,
ST=Utrecht, C=NL correct?
[no]: yes

Since we want an empty keystore in which we import a private key, we need to delete it:

keytool -delete -alias darwinKey -keystore ~/Keystores/b2bKeyStore.jks -
storepass welcome1

Now the keystore is empty we can import the pkcs12 certificate:

keytool -v -importkeystore -srckeystore ~/Certificaten/
darwinCustomerCertificate.p12 -srcstoretype PKCS12 -destkeystore ~/
Keystores/b2bKeyStore.jks -deststoretype JKS -storepass welcome1

Output:

Enter source keystore password:
Entry for alias darwincustomer successfully imported.
Import command completed: 1 entries successfully imported, 0 entries
failed or cancelled
[Storing /home/oracle/Keystores/b2bKeyStore.jks]

Now I used as a keystore password "welcome1" in the examples above. Later I found that to enable B2B to read the certificate and actually use it, the password of the keystore had to be the same as the passphrase of the private key. So it is best when using the examples above for your actual implementation to use that password right away. But luckily the password can also be changed:

keytool -storepasswd -keystore ~/Keystores/b2bKeyStore.jks

Output:

Enter keystore password:
New keystore password:
Re-enter new keystore password:

Convert DER certificate to PEM

When you get a public certificate, you probably get it in a kind of binary format. You can also export it from your private key in your keystore with the keytool utility:

keytool -exportcert -alias darwinCustomerCert -file ~/Certificaten/
darwinCustomerPublicCert.cer -keystore ~/Keystores/b2bKeyStore.jks -storepass
welcome1

Output:

Certificate stored in file </home/oracle/Certificaten/darwinCustomerPublicCert.cer> 

The exported certificate is stored in DER-format. To create a PEM file out of it, OpenSSL can be used. The command for this, I found here and here.

openssl x509 -inform der -in darwinCustomerPublicCert.cer -outform pem -out
darwinCustomerPublicCert.pem

This will deliver a readable certificate that is importable in the keystore.