Make the Target Folder always in-sync from the Source Folder using Beyond Compare

  1. Create a folder compare session with beyond compare like the following:
  2. Save the session with a file name without any special characters (e.g. < or >) or spaces (i.e. necessary not to confuse the scheduler).
  3. Save the following beyond compare script to right-updater.bc file.
    #Write to right-updater.log
    log normal "right-updater.log"
    #Load a folder compare session.
    load "%1"
    #Confirmation will be set to yes by default. 
    option confirm:yes-to-all
    #Make the right equals to the left of the loaded folder from the session.
    sync mirror:left->right
  4. Save the following powershell script to bcrunner.ps1 file along the side right-updater.bc file.
    #expects 2 command line parameters
    param([string]$script, [string]$session)
    #The beyond compare command to be used.
    $cmd="C:\Program Files\Beyond Compare 4\"
    #Holds the session argument for the beyond compare command.
    #Write on screen the actual command to be executed.
    write-output "$cmd /silent @$script $args"
    #Perform the command.
    &$cmd /silent @$script $args
  5. Schedule using windows task scheduler and on the actions tab fill in the following fields like:
    Action Start a program
    Program/script powershell
    Add arguments(optional) ./bcrunner.ps1 right-updater.bc <SAVED_SESSION_FROM_STEP_2>
    Start in (optional) <DIRECTORY_OF_BCRUNNER.PS1_FILE_FROM_STEP_4>

    Note: Fill-in the other tabs of the scheduler accordingly.

Triggering builds remotely with Jenkins using HTTP Request plugin

Use Cases

Lower Jenkins (e.g. v1.x) but requires later Jenkins (e.g. v2.x) functionality (e.g. using git plugin which has some transitive dependencies which requires v2.x).

For some reason , we just wanted a Jenkins server to delegate a task to another Jenkins server.

You've created a script (i.e. python, ruby, groovy, etc) that will trigger a Jenkins task via HTTP protocol. This is not part of this document but the Backend Jenkins Server section is applicable.

Backend Jenkins Server

This is the server that will perform the task.

Setting-up a role that can do remote build

  1. Click Jenkins -> Manage Jenkins -> Manage and Assign Roles entry and select Manage Roles.
  2. In the Global roles, add a new role (e.g. remote_triggers) that is allowed to trigger a task remotely.
  3. Once added enable the following:
Overall Read
Job Build
Job Read
Job Workspace

Example 1: Sample role configuration

If we have remote_triggers as the new role, we have the following:

Assigning a role to a user

  1. Click Jenkins -> Manage Jenkins -> Manage and Assign Roles entry and select Assign Role.
  2. In the Global roles, add a known Jenkins user.
  3. Once added assign the role.

Example 2: Assign a role following Example 1.

If the known user is GwDevOpsDev, we have the following:

Setting-up the Build Triggers

  1. On the Build Triggers section, select the Trigger builds remotely (e.g., from scripts).
  2. Provide an Authentication Token.

Note: You can use just remove the hyphen for this.

This will enable you to trigger the job remotely via HTTP Post using one of the following:

Without Build Parameters


With Build Parameters


Example 3: Address to call a task remotely.

If we have the following:

ITEM_NAME=Statement Date Update to Saturday

The URL to call this job remotely with build parameters must be:

User Facing Jenkins Server

This is the Jenkins server that is normally use by the user to invoke tasks but the task is actually being done by a another Jenkins server.


  • HTTP Request Plugin

Setting-up Credential to be used with the HTTP request.

  1. Click Jenkins -> Credentials entry.
  2. Click the global store seen as follows:
  3. Click Add Credentials.
  4. Set the Kind field to Username with Password.
  5. Set the Scope field to Global if not yet set.
  6. Set the Username field to the user that can to remote triggering configured on the backend Jenkins server (e.g. GwDevOpsDev from Example 2).
  7. Set the Password field to a valid password.
  8. Set the ID field to something unique.
    Note: You can use  for this.
  9. Click the OK button.

Setting-up a New Item to Delegate Task to Backend Jenkins Server

  1. Click New Item.
  2. Provide an Item name.
  3. Select Freestyle project.
  4. Click the OK button.
  5. Click Add build step-> HTTP Request entry.
  6. Fill-in the URL with the one configured in the backend Jenkins server (e.g. from Example 3)
  7. Select POST for the HTTP mode.
  8. Click the Advanced… button.
  9. In the Authenticate field select the set up from Setting-up Credential to be used with the HTTP request section earlier.
  10. Click the Save button.
  11. Now every time you click Build now it will complete immediately since the actual build will be done on the Backend Jenkins Server.

Unit Test with Expando Object


Expando object can make dynamic object, properties and methods if you like. This can be used for stubbing an object to be used for testing.

For example if have the following:

Code 1 - processor function that requires the SSampleStructure structure

//Structure works with objects with similar features but no common inheritance and interface declarations
structure SSampleStructure {
  function someAction() : int

function processor(sample : SSampleStructure) : int {
  return sample.someAction() + 10

The SSampleStructure structure will be compliant to all the classes that has someAction function that returns an int. And the actual compliant implementation of the structure is the following:

Code 2 - Actual implementation that is compatible with SSampleStructure

class SampleCodeWithStaticMethods {

  private construct() {}

  function someAction() : int {
    return (Math.round(Math.random()*100) as int) + 100

We can notice that in Code 2, we cannot have a consistent output for someAction method since it is return arbitrary values. And what the processor function does is just add 10 on it.

What we can do here is to create an Expando object that will also comply with SSampleStructure structure like the following:

Code 3 - Using Expando class

uses gw.lang.reflect.Expando
    // By declaring the Dynamic type for the variable or function argument, you can
// programmatically access any property or method without requiring compile-time confirmation of its existence.
var testObj : Dynamic =

// An expando object is an object whose properties, and potentially methods, are created dynamically on assignment.
    new Expando()

// We can attached function using the block syntax as below.  
testObj.someAction = \ -> 25

Since the testObj of Code 3, also has someAction function that returns an int. This is compliant with SSampleStructure structure. So we can safely run the following test of logic without any problem:

Code 4 - Actual test

function testProcessor() {
  print("Is35: " + (processor(testObj)==35))


As we can see from Code 4, we are using testObj to test the processor method.

Preparing Your Powershell Profile for Editing

  1. Execute the following to check the PROFILE variable:
    Get-Variable | Where-Object {$_.Name -like "PROFILE"}
  2. Test if its actually existing on the disk using the following command:
    Test-Path $PROFILE

    This will return True if the profile is already existing on the disk and ready for editing. If this returns False execute the following command:

    New-Item -Path $PROFILE -ItemType "file" -Force

Install Jabba in Powershell behind proxy

Jabba is a java version manager (see simplifies the installation and switching of/to different JDK.

If your powershell is behind the firewall follow the procedure here.

Install Jabba using the following cmdlet:

Note: If you encountered something like jabba.ps1 cannot be loaded because running scripts is disabled on this system. For
more information, see about_Execution_Policies at see running unsigned script.

Invoke-Expression (
        wget -UseBasicParsing

Running Unsigned Script Temporarily in Powershell

  1. Get the current execution policy using the following command and remember it:
  2. Set the execution policy to unrestricted using the following command:
    Set-ExecutionPolicy -Scope CurrentUser -ExecutionPolicy Unrestricted
  3. Execute the unsigned script.
  4. Return the back execution policy to what it was (i.e. from step 1). Using the following command syntax similar to step 2:

    Note: This is not required if you know that you are only running the script that you trust.

    Set-ExecutionPolicy -Scope CurrentUser -ExecutionPolicy <EXECUTION_POLICY_FROM_STEP1>

Powershell Behind Proxy

Using the default network credentials

(New-Object System.Net.WebClient).Proxy.Credentials = [System.Net.CredentialCache]::DefaultNetworkCredentials

Using a different credentials

Using Get-Credential

(New-Object System.Net.WebClient).Proxy.Credentials = Get-Credential

Using NetworkCredential Object

(New-Object System.Net.WebClient).Proxy.Credentials = New-Object System.Net.NetworkCredential(<username>, <password>)

Note: You can add any one of these to your powershell profile to allow it to have network connectivity by default using proxy.

Calling Web Service without Generating Stubs (JAX-WS API)


Normally calling a SOAP WebAPI requires the developer to access WSDL then based on it generates the stubs linked to a particular server. After generating it, we need to compile those to make them ready to be used. The problem with that one is. Do we use the same server for development, testing and production? If yes, we need to regenerate the stubs for testing and if we are satisfied with the testing. We need to regenerate the stubs for production. Wait? The one tested is not really the one going to production. The simplest way to do is not to generate the stubs but understand how the soap message was structured. If we know the structure of the SOAP message (i.e. SOAP UI can be used here), then we can create it by hand. The following is the sample of assembling a SOAP message for invoking the evaluate method from the StudioService endpoint of PolicyCenter 7 (i.e. One of the Centers of Guidewire) using Java.

Code 1. Using JAXWS API with Java

import java.util.Iterator;

public class GWStudioService {

    public static void main(String[] args) throws SOAPException, IOException {
        String endpointUrl = "http://localhost:8180/pc/ws/gw/internal/webservice/studio/StudioService";
        String targetNamespace = "";

        QName serviceName = new QName(targetNamespace,"evaluateGosu");
        QName portName = new QName(targetNamespace,"StudioServiceSoap12Port");

        /** Create a service and add at least one port to it. **/
        Service service = Service.create(serviceName);
        service.addPort(portName, SOAPBinding.SOAP12HTTP_BINDING, endpointUrl);

        /** Create a Dispatch instance from a service.**/
        Dispatch dispatch = service.createDispatch(portName,
                SOAPMessage.class, Service.Mode.MESSAGE);

        /** Create SOAPMessage request. **/
        // compose a request message
        MessageFactory mf = MessageFactory.newInstance(SOAPConstants.SOAP_1_2_PROTOCOL);

        // Create a message.  This example works with the SOAPPART.
        SOAPMessage request = mf.createMessage();
        SOAPPart part = request.getSOAPPart();

        // Obtain the SOAPEnvelope and header and body elements.
        SOAPEnvelope env = part.getEnvelope();
        SOAPHeader header = env.getHeader();
        SOAPBody body = env.getBody();

        String authNS = "";

        SOAPElement auth = header.addChildElement("authentication", "auth", authNS);
        SOAPElement username = auth.addChildElement("username", "auth");
        SOAPElement password = auth.addChildElement("password", "auth");

        // Construct the message payload.
        SOAPElement operation = body.addChildElement("evaluateGosu", "stud",
        SOAPElement value = operation.addChildElement("code", "stud");


        SOAPMessage response = dispatch.invoke(request);


        SOAPPart inSOAPPart = response.getSOAPPart();
        SOAPEnvelope inEnv = inSOAPPart.getEnvelope();
        SOAPBody inBody =  inEnv.getBody();

        Iterator inGosuResp = inBody.getChildElements(new QName(targetNamespace, "evaluateGosuResponse"));
        Iterator inReturn = QName(targetNamespace, "return"));
        Iterator inEntries= QName(targetNamespace, "Entry"));

        while (inEntries.hasNext()) {
            SOAPElement inEntry =;


Output 1: Sample output of running code 1

<soap12:Envelope xmlns:soap12=""><soap12:Header/><soap12:Body><evaluateGosuResponse xmlns="">






To understand the output of the entry we can use the following online Hex to String converter.