Olympic Arts

Android read file from URL

admin — Thu, 24 Feb 2011 15:13:35 +0000

How to read a file from the internet

While coding y next app for Android, I found out that I needed to read files from URL locations and saven them on my Android device.
I searched for the first sample and put that code to work in my app. That was a quick disappointment. It dit not work and it gave me the message:

java.io.IOException: Server returned HTTP response code: 401 for URL
DefaultRequestDirector(5824): Authentication error: Unable to respond to any of these challenges: {}

I did not know what to make of this, since my file on the net is not protected by authentication. A lot of searching on the web did not reveal any solution to my problem (only solutions where you had to authenticate).

After long hours of searching I found a workaround for my problem. Just send in some dummy authentication and it will work.

HTTPClient / UrlConnection example

Here is my sample of my workaround

URI _url = new URI(location);
HttpGet httpget = new HttpGet(_url);
DefaultHttpClient httpclient = new DefaultHttpClient();
httpclient.getCredentialsProvider().setCredentials(AuthScope.ANY, new UsernamePasswordCredentials("admin","admin"));
String _generatedFilename = generateFilename(_filename, savedconfiglist);
HttpResponse response = httpclient.execute(httpget);
HttpEntity entity = response.getEntity();
if (response.getStatusLine().getStatusCode()==200) {
saveArmyConfig(_generatedFilename, entity.getContent());
} else {
Log.d("TESTING", "error receiving content");
}
} catch (IOException e) {
e.printStackTrace();
} catch (URISyntaxException e) {
e.printStackTrace();
}

Kind regards,
Marc

Android XML parsing workaround

admin — Thu, 16 Dec 2010 14:09:06 +0000

The problem

I came across a problem while using XML in my android application. When trying to parse XML in android 2.2 and above, I read the qName variable and use that to do my magic. But in sdk version 7 (2.1) this poses a serious problem. A bug in the Android code, causes the qName to be empty and the localname to be filled with your XML element value.

The solution

I have coded a Utility that works around this minor issue. The code is as follows:

public static String qLocal(String qName, String localname) {
String _retValue = "";
if (Build.VERSION.SDK_INT<=7) {
_retValue = localname;
} else {
_retValue = qName;
}
return _retValue;
}

Call the function above in your startElement(…) and endElement(…) functions and assign the return value to a local variable and u are good to go and cross version compatible again.
I Hope this helps you avoiding this annoying issue with XML parsing on Android 2.1.

Kind regards,

Marc

Android Form Field Validation

admin — Mon, 04 Oct 2010 14:23:28 +0000

Validation ??

After coding a while for the Android platform, I was getting annoyed that there was no default validation support (i.e that I could find) for the validation of EditText, CheckBox etc fields.
Therefore I have written a small set of interfaces/Classes that gives you and me the possibility to do the validation you need.

The basics

To be able to use the validation you need to implement de Validator Interface and extend the AbstractValidator class. In the abstract class a couple of handy methods are already implemented, but may be overridden for your specific usage.
An example of an implemented validator:

package nl.android.validator;

import android.widget.CheckBox;

/**
* A Checkbox required field
* implementation of the validator.
* With this validator a checkbox can be tested
* for the required option.
*
* @author http://nl.linkedin.com/in/marcdekwant
*
*/
public class CheckBoxRequiredValidator extends AbstractValidator {

private CheckBox _source;

/** CONSTRUCTORS */

public CheckBoxRequiredValidator(CheckBox source) {
super(true);
_source = source;
}

public CheckBoxRequiredValidator(CheckBox source, String requiredMessage) {
super(true);
_source = source;
_requiredMessage = requiredMessage;
}

@Override
public ValidationResult validate() {
ValidationResult _v = super.validate();
if (_v==null) {
if (_source.isChecked()) {
_v = new ValidationResult(true, "");
} else {
_v = new ValidationResult(false, _requiredMessage);
}
}
return _v;
}

@Override
public Object getSource() {
return _source;
}

}

Example usage in a activity

.....
private List<validator> validators = new ArrayList<validator>();
.....
private void someMethod() {
validators.add(new CheckBoxRequiredValidator(myCheckbox, "This checkbox is required"));
}
.....
private void myAcceptMethod() {
List<validationResult> _validationResults = AbstractValidator.validateAll(validators);
if (_validationResults.size()==0) {
// Validation is succesfull
}
}

Source code

Android Validation zip

Kind regards,

Marc

JEE Webservice basic auth security check with Glassfish

admin — Thu, 08 Jul 2010 11:26:32 +0000

No @RolesAllowed available for Webservices

The JEE specification is growing into a nice and easy to use specification. But some features, like Webservices and RolesAllowed, are not good enough yet. To be more precise. @RolesAllowed is not supported for plain Webservices (not doubling as EJB’s) until version 3.0 in the Servlet Specification.

After much reading on this subject, I have come to the conclusion that a pragmatic solution is required. My solution is just to inspect the principals based upon the logged in user and check myself if the user is allowed to continue. But you can also define the webservice as EJB.

My setup

Glassfish 3
JEE
Basic authentication configured in JEE project and glassfish server by JDBC realm means

My @RolesAllowed / isUserInRole solution

Off course an annotation can be constructed for the example below, but that I leave to other to implement.

Check the principal yourself

public static boolean accessAllowed(String role) {
boolean retValue = false;
Subject s;
try {
s = (Subject) PolicyContext.getContext("javax.security.auth.Subject.container");
Set<principal> principals = s.getPrincipals();
for (Principal p : principals) {
if (p.getName().equals(role)) {
// Access allowed
retValue = true;
}
}
} catch (PolicyContextException e) {
// Something is wrong. We do nothing by return access NOT allowed
}
return retValue;
}

Define the webservice as EJB

Another solution would be to define the webservice as an EJB also. I have not chosen this approach, but it is a matter of taste.

I hope the JEE specification will add/implement @RolesAllowed much needed feature for basic authenticated webservices, as in EJB’s, as fast as they can.

Kind regards,

Marc

Flex: Dynamic binding using a mapping file

admin — Wed, 20 Jan 2010 08:08:01 +0000

Introduction

Binding can be an time consuming task in Flex. If you have a domain model and the DTO’s for your components, you need an easy way to bind the two together. Here you have two options. The first is to write alot of BindingUtils statements in either models. And finally you can write a mapping file in xml, create a custom binder class which uses the mapping file and bind the two models together.

I have chosen the last option, because this gives me some more flexibility handling the bindings.

Requirements

Flex Builder 3
Tutorial code (See below for the files: the xml, xsd and binder.as)

Actual coding

To make this idear happen, I needed several things. First, I needed my two models (the real domain model and the DTO for the view components) which needed to be bound together. Prerequisite for these models is that they have a single point of entry to be able to naviagate to all objects within the model.
Secondly, a mapping file. I have created a easy to use mapping file in which I configure the binding rules.
And finally, the class that does the actual binding based upon the mapping file and both models.

Code samples

The is not much more to say about this, so here I will show you samples of all code.

Binding file

version="1.0" encoding="utf-8"?>
xmlns="http://www.example.org/binding" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xsi:schemaLocation="http://www.example.org/binding assets/binding.xsd ">

id="myList" bidirectional="true">
topLevelObject="myDomainRoot" object="myDomainRoot.subDomainObj" field="myList" isCollection="true"/>
>
topLevelObject="myDTORoot" object="customer.address" field="myList" />
topLevelObject="myDTORoot" object="company" field="anotherList" />
>
>

>

XSD file

version="1.0" encoding="UTF-8"?>
id="mapping" xmlns="http://www.w3.org/2001/XMLSchema" targetNamespace="http://www.example.org/binding" xmlns:tns="http://www.example.org/binding" elementFormDefault="qualified">
name="mapping">
>
minOccurs="1" maxOccurs="unbounded">
name="binding">
>
>
name="sourceBinding" minOccurs="1" maxOccurs="1">
>
name="topLevelObject" type="string" />
name="object" type="string" />
name="field" type="string" />
name="isCollection" type="string" />
>
>
name="targets" minOccurs="1" maxOccurs="1">
>
>
name="targetBinding" minOccurs="1" maxOccurs="unbounded">
>
name="topLevelObject" type="string" />
name="object" type="string" />
name="field" type="string" />
>
>
>
>
>
>
name="id" type="string" />
name="bidirectional" type="string" />
>
>
>
>
>
>

Mapping explenation
A binding element represents a complete bi-rectional binding
A sourceBinding element represents the source end object and property of the required binding. TopLevelObject indicates the root of the model. The object (for example rootObject.subObject) represents the navigation to the object that needs to facilitate the binding. If the topLevelObject and the first object in the property Object are the same, this is ignored in the navigation. It just means that the model has no nested objects, but only properties.
A targetBinding element represents the target to be bound. Same rules as the source binding apply.

Binding class

import flash.utils.ByteArray;

import mx.binding.utils.BindingUtils;
import mx.logging.ILogger;
import mx.logging.Log;
import mx.logging.LogEventLevel;
import mx.logging.targets.TraceTarget;

/**
* This class provides the binding capability between
* 2 models. The binding is read from a XML file and
* used in the binding process.
*
* @author http://nl.linkedin.com/in/marcdekwant
*/
public class Binder {

[Embed( source='/assets/bindings.xml', mimeType="application/octet-stream" )]
private var _rawFile:Class

// The from file read xml representation of the bindingfile
private var _bindings:XML;
// The original domain model
private var _model:*;
// The target model (DTO's related to the UI components)
private var _dto:*;

private var logger:ILogger = Log.getLogger( "Binder" );

/**
* Constructor
*
* @param dto the target domain model
* @param model the source domain model
*/
public function Binder( dto:*, model:* ) {
this._dto = dto;
this._model = model;
var file:ByteArray = new _rawFile();
var str:String = file.readUTFBytes( file.length );
_bindings = new XML( str );
initLogging();
createBindings();
}

/**
* Initialize the logging
* TODO refactor to controller or application level
*/
private function initLogging():void {
var logTarget:TraceTarget = new TraceTarget();
logTarget.level = LogEventLevel.ALL;
logTarget.filters = [ "*", "nl.dialoog.binding.*" ];
logTarget.includeCategory = true;
logTarget.includeDate = true;
logTarget.includeLevel = true;
logTarget.includeTime = true;
Log.addTarget( logTarget );
}

/**
* This method retrieves a property from a
* given object. THis method gives the binding
* class the capability to chain objects.
*
* @param _objTarget The object
* @param _innerObj the property of the targetObject
* @return the innerObj from the targetobj
*/
private function resolveNestedObject( _objTarget:*, _innerObj:String ):* {
return _objTarget[ _innerObj ];
}

/**
* This method resolves the property from an object chain to
* be bound to another property from an object chain.
*
* @param _obj The baseObject (domain model root, or DTO root)
* @param _xmlObj the target to be found in the root
* @return the target object to be bound.
*/
private function resolveSourceObject( _obj:*, _xmlObj:XML ):* {
var _objArray:Array = _xmlObj.@object.split( "." );
for ( var i:int = 0; i < _objArray.length; i++ ) {
if ( _xmlObj.@topLevelObject != _xmlObj.@object ) {
_obj = resolveNestedObject( _obj, _objArray[ i ]);
}
}
return _obj;
}

/**
* Actual method that hold the logic in creating bindings
* based upon the binding xml file.
*/
private function createBindings():void {
logger.debug( "Start create bindings..." );
var _objSource:* = null;
var _objTarget:* = null;

// Set the default namespace from the binding.xml, in order to process the
// mapping file. "" relates to the xmlns="" attribute in the root element
default xml namespace = _bindings.namespace("");

// traverse over all elements
for each ( var xmlBinding:XML in _bindings.binding ) {
logger.debug( "Create binding for " + xmlBinding.@id + " bi-directional:" + xmlBinding.@bidirectional );
var sourceBinding:XML = xmlBinding.sourceBinding[ 0 ];
_objSource = resolveSourceObject( this._model, sourceBinding );
logger.debug( "Collection binding..." );
// traverse over all elements
for each ( var xmlTargetBinding:XML in xmlBinding.targets.targetBinding ) {
logger.debug( "Construction of bindings from " + sourceBinding.@object + "." + sourceBinding.@field + " to " + xmlTargetBinding.@object + "." + xmlTargetBinding.@field );
_objTarget = resolveSourceObject( this._dto, xmlTargetBinding );
// A collection requires a slightly different binding then properties
// because the ArrayCollection contains binding mechanism in its own class.
if ( sourceBinding.@isCollection == "true" ) {
logger.debug( "Collection binding..." );
_objTarget[ xmlTargetBinding.@field ][ "source" ] = _objSource[ sourceBinding.@field ][ "source" ];
} else {
logger.debug( "Property binding..." );
var _sourceStr:String = sourceBinding.@field;
var _targetStr:String = xmlTargetBinding.@field;
BindingUtils.bindProperty( _objSource, _sourceStr, _objTarget, _targetStr );
BindingUtils.bindProperty( _objTarget, _targetStr, _objSource, _sourceStr );
}
}
}
logger.debug("Bindings created...");
}

}
}

Sample Application file

version="1.0" encoding="utf-8"?>
xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical"
creationComplete="onCreationComplete();">

>

import mx.binding.utils.BindingUtils;

[Bindable]
private var myDomainRoot:MyDomainRoot;
[Bindable]
private var myDTORoot:MyDTORoot;

private function onCreationComplete():void {
myDomainRoot= new MyDomainRoot();
myDTORoot = new MyDTORoot();
var _binder:Binder = new Binder(myDTORoot,myDomainRoot);
}

]]>
>

>

Kind regards,

Marc de Kwant

Flex 3 Custom validation of grouped input fields

admin — Tue, 01 Dec 2009 17:12:27 +0000

Introduction

Flex 3 has nicely build-in validation mechanism, but sometimes you wish they had not embedded this so deep into the components, that customization becomes a laborish task. I have also read many articles on customizations of this mechanism, but I have not found a working solution to my specific wish. What I want is validation of single input fields, multiple input and arbitrary placement of errors
besides the tooltip method.

Requirements

Flex Builder 3
Tutorial code (See step 7 for both files)

Steps

Now I am going to describe the steps I took to create this validator.

Step 1

I have created a Flex (web) project in my workspace. In this project I have 1 mxml file (my application file). I have named this ValidationTutorial.mxml.

version="1.0" encoding="utf-8"?>
xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical" xmlns:local="*">
>

In this mxml I will put 3 input field (day, month, year) in a hbox. Furthermore I will place a hidden Label
below the hbox. These are my basic components that I will be using to demonstrate the multiple input validator.

version="1.0" encoding="utf-8"?>
xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical" xmlns:local="*">
>
id="day" />
id="month" />
id="year" />
>
id="errorMsgbirthdayValidationModel" visible="false" includeInLayout="false"/>
>

Step 2

Code our custom validator. Now we have to create the validator. file -> new -> actionscript class and call it BirthdayValidator (make it extend Validator). override the protected function doValidation, put a call to super.doValidation(value) in it and we have our foundation for the validator in place. Your actionscript class should look as follows

package {

public class BirthdayValidator extends Validator {

public function BirthdayValidator() {
super();
}

override protected function doValidation(value:Object):Array {
// Init the results array with the basic validation
var results:Array = super.doValidation(value);
return results;
}

}
}

Step 3

This is a relatively small step, but contains the magic that will enable me to validate and control the output regarding 3 input fields. Make a mental note of the valError element, which I will use later on.

id="birthdayValidationModel">
>
>
>{day.text}>
>{month.text}>
>{year.text}>
>>
>
>
>

Step 4

We define the validator and link all components together. this means we are going to tell the validator that the source for validation is our model and the results will be displayed in the label. If you now run the application, nothing will happen when tabbing across the fields. This is because we have not defined the triggers for the validator and the actual implementation of the validator. We will put this on the last input with focusOut. The implementation for the validator is for tutorial purpose simple.

id="birthdayVal"
required="true"
source="{birthdayValidationModel}"
property="birthday"
listener="{errorMsgbirthdayValidationModel}"
trigger="{year}"
triggerEvent="focusOut"/>

We are now close to a working example.

Step 5

Now we define the invalid handler and the valid handler. These methods will contain code to set/unset validation texts. Configure them into the validator and edit the label so that it responds to the errors in it. When you have done this, you will see validation results below the inputs when you enter no values and focusOut on the year input (last input field).

id="birthdayVal"
required="true"
source="{birthdayValidationModel}"
property="birthday"
listener="{errorMsgbirthdayValidationModel}"
trigger="{year}"
triggerEvent="focusOut"
invalid="onInvalidHandler(event)"
valid="onValidHandler(event)"/>

And the methods….

/**
* Validationresult event handler. This function
* places the validation results into the appropriate
* label for error text placement beneath an input
* text.
*
* @param event Results of the input box(es) validation
*/
public static function onInvalidHandler( event:ValidationResultEvent ):void {
var _error:String = "";
var validator:Validator = event.target as Validator;
for each ( var result:ValidationResult in event.results ) {
_error+=result.errorCode;
}
// All validation results will be displayed in a label component
var errorLabel:Label = validator.listener as Label;
errorLabel.text = _error;
event.stopImmediatePropagation();
}

/**
* Validation result handler. This function removes all
* error texts from the error label for the current input
* component.
*
* @param event Results of the input box(es) validation
*/
public static function onValidHandler( event:ValidationResultEvent ):void {
var _error:String = "";
var validator:Validator = event.target as Validator;
// All validation results will be displayed in a label component
var errorLabel:Label = validator.listener as Label;
errorLabel.text = _error;
event.stopImmediatePropagation();
}

Step 6

Add styling. Just some nice sugar, but styling is everything. Add this styling to your ValidationTutorial.mxml And add the following method to the script to use this styling.
Add to the 3 input field the following property

styleName="{ValidatorUtil.showErrorBorder((birthdayValidationModel.birthday.valError as XML), this.day)}"

Step 7

Now comes the 4th element in the mx:model into focus. When you run the application you will see deep blue borders, but they do not change to red borders were there is an error detected by the validation. For this to work we need to programaticly fill the error element to reflect the inputs that are invalid.
This we will do in the validator and a helper method in our ValidationTutorial.mxml

The BirthdayValidator

package {
import mx.validators.ValidationResult;
import mx.validators.Validator;

public class BirthdayValidator extends Validator {

private static const ERROR_START_ELEMENT_TAG:String = "";
private static const ERROR_END_ELEMENT_TAG:String = "";
private static const ERROR_DAY_TAG:String = "day";
private static const ERROR_MONTH_TAG:String = "month";
private static const ERROR_YEAR_TAG:String = "year";

public function BirthdayValidator() {
super();
}

override protected function doValidation(value:Object):Array {
var day:String = value.day;
var month:String = value.month;
var year:String = value.year;
var errorString:String = ERROR_START_ELEMENT_TAG;

// Init the results array with the basic validation
var results:Array = super.doValidation(value);
if (day == "" || day == null) {
// Texts and codes need to be refactered out of this util
// into a text file
results.push(new ValidationResult(true,
"day", "noDay", " No day"));
errorString += ERROR_DAY_TAG;
}
// Check month field.
if (month == "" || month == null) {
// Texts and codes need to be refactered out of this util
// into a text file
results.push(new ValidationResult(true,
"month", "noMonth", " No month"));
errorString += ERROR_MONTH_TAG;
}
// Check month field.
if (year == "" || year == null) {
// Texts and codes need to be refactered out of this util
// into a text file
results.push(new ValidationResult(true,
"year", "noYear", " No year"));
errorString += ERROR_YEAR_TAG;
}
errorString += ERROR_END_ELEMENT_TAG;
value.valError = new XML(errorString);
return results;
}

}
}

The ValidatorTutorial.mxml

version="1.0" encoding="utf-8"?>
xmlns:mx="http://www.adobe.com/2006/mxml" layout="vertical" xmlns:local="*">

>

.errorBorderStyle {

border-color: red;
border-style: solid;
border-thickness: 2px;
border-thickness-bottom: 2px;
border-thickness-left: 2px;
border-thickness-right: 2px;
border-thickness-top: 2px;

}

.defaultBorderStyle {

border-color: blue;
border-style: solid;
border-thickness: 2px;
border-thickness-bottom: 2px;
border-thickness-left: 2px;
border-thickness-right: 2px;
border-thickness-top: 2px;

}

>

>

import mx.core.UIComponent;
import mx.validators.ValidationResult;
import mx.validators.Validator;
import mx.events.ValidationResultEvent;

/**
* Validationresult event handler. This function
* places the validation results into the appropriate
* label for error text placement beneath an input
* text.
*
* @param event Results of the input box(es) validation
*/
public static function onInvalidHandler( event:ValidationResultEvent ):void {
var _error:String = "";
var validator:Validator = event.target as Validator;
for each ( var result:ValidationResult in event.results ) {
_error+=result.errorCode;
}
// All validation results will be displayed in a label component
var errorLabel:Label = validator.listener as Label;
errorLabel.text = _error;
event.stopImmediatePropagation();
}

/**
* Validation result handler. This function removes all
* error texts from the error label for the current input
* component.
*
* @param event Results of the input box(es) validation
*/
public static function onValidHandler( event:ValidationResultEvent ):void {
var _error:String = "";
var validator:Validator = event.target as Validator;
// All validation results will be displayed in a label component
var errorLabel:Label = validator.listener as Label;
errorLabel.text = _error;
event.stopImmediatePropagation();
}

/**
* Function reserved for complex objects. In this we mean mx:model
* type objects for multi input validations. (dates where day,month,year
* are seperated input controls).
*
* @param errors A xmllist of elements containing the controls that need
* or do not need error styling.
* @param input the control we like to style.
* @return the new styling information.
*/
public static function showErrorBorder( errors:XML, input:UIComponent ):String {
var styleName:String = "defaultBorderStyle";
// Check for not null, otherwise the flashplayer
// will silently ignore the nullpointer and break off
// any further styling for the target control
if ( errors != null && errors.error != null ) {
for each ( var error:XML in errors.error ) {
if ( error == input.id ) {
styleName = "errorBorderStyle";
break;
}
}
}
return styleName;
}

]]>
>

id="birthdayValidationModel">
>
>
>{day.text}>
>{month.text}>
>{year.text}>
>>
>
>
>

>

id="day" styleName="{showErrorBorder((birthdayValidationModel.birthday.valError as XML), this.day)}"/>
id="month" styleName="{showErrorBorder((birthdayValidationModel.birthday.valError as XML), this.month)}"/>
id="year" styleName="{showErrorBorder((birthdayValidationModel.birthday.valError as XML), this.year)}"/>

>

id="errorMsgbirthdayValidationModel" visible="{errorMsgbirthdayValidationModel.text!=''}" includeInLayout="{errorMsgbirthdayValidationModel.text!=''}"/>

id="birthdayVal"
required="true"
source="{birthdayValidationModel}"
property="birthday"
listener="{errorMsgbirthdayValidationModel}"
trigger="{year}"
triggerEvent="focusOut"
invalid="onInvalidHandler(event)"
valid="onValidHandler(event)"/>

>

Final note

And now the magic is complete. You have build a multiple input validator, with error message customization and placement. You can add many enhancement, that I will leave up to your imagination.

I hope you enjoy coding Flex/ActionScript as I do.

Kind regards,

Marc de Kwant

How To: Load a spring application from a jar file

admin — Mon, 24 Aug 2009 07:18:59 +0000

Consider the following situation:

Loading 3 application context files in a Swing application. 2 of the context files are in jar-files.
All context files depend on the component-scan option. The Application context gets loaded from my Swing app.
The method I use, is the import resource option in the context file. Read on how to configure this in your application context.

My DAO class

package nl.foo.bar.dao;

import org.springframework.stereotype.Repository;

import nl.foo.bar.domain.GOClass;

@Repository("goClassDAO")
public class GOClassDAOJPA extends BaseDAOJPA<goclass> implements GOClassDAO {
}

ApplicationContext-dao.xml

base-package="nl.foo.bar.dao">

type="annotation"

expression="org.springframework.stereotype.Repository"/>
>

ApplicationContext-controller.xml (this gets loaded from my swing application)

resource="classpath:ApplicationContext-dao.xml" />

base-package="nl.foo.bar.gui.controller">

type="annotation"

expression="org.springframework.stereotype.Controller"/>
>

Code snippet from my swing application

String[] locations = {"ApplicationContext-controller.xml"};
ClassPathXmlApplicationContext ctx = new ClassPathXmlApplicationContext(locations);
ApplicationController appController = (ApplicationController)ctx.getBean("applicationController");

I have tried various other options in the past (like adding “classpath:*/ApplicationContext-dao.xml” and variations to the locations) too no success.

The way to go is the import-tag in my opinion. Works like a charm.

Kind regards,
Marc de Kwant

Webservices with Spring and Castor

admin — Mon, 17 Aug 2009 09:33:19 +0000

Today, webservices are used primarily to expose some services of an application to the outside world. There are a lot of tools available to consume webservices. Tools like Mule, Axis and Spring.
This article describes a method of exposing Java services as webservices. The technique’s / tools used to accomplice this are; Java, Spring and Castor.
There are many other tools available, like Axis, to expose java services as webservices. The reason I used Spring and Castor is the simplicity in creating webservices. Spring takes care of all boilerplate coding and Castor’s mapping capability takes care of the Object <-> XML mapping without the need to generate any java objects the way Axis does. This way you have clean java code, supported with configuration in Spring and a clean Castor mapping to expose your service as a webservice.

Knowledge

I presume that you posses the following knowledge

Java web component developer knowledge level
Know how to use Eclipse and the svn and/or Maven plugin
Basic understanding of Spring
Basic understanding of Maven (but you do not need knowledge besides being able to install the tooling)
Basic understanding of webservices
Basic understanding of Tomcat

Requirements

To be able to run the provided example code you need the following applications and tools.

For a complete overview of the required libraries, you can view the pom.xml in the example code. There is NO need o download the tooling yourself, this will be done by maven.

The example

This example exposes a webservice that returns a Artist object from a (HSQL) database, given an id. The example will not explain how to create a CRUD application using DAO’s and services etc. This should be knowledge you already have. I have provided an example project, that contains the necessary CRUD code for this webservice project to use.

Project configuration

The next thing you need to do is to download the example projects from google code. These are the Top2000Persistence and Top2000Webservices project.

First you need to checkout the persistence project. The webservice project depends on it. The persistence project is a smaple project using Spring, JPA and Hibernate.

The way I do this is the following:

Startup your Eclipse
File -> new -> Other -> Maven -> Checkout maven projects……
select svn from the dropdown
Paste the url http://top2000persistence.googlecode.com/svn/trunk/
click on Finish

The project wil be downloaded into your Eclipse and configured as a maven/java project. Below an image showing the steps I have described here:

New Maven Project Dialog

Now you can build the project either with Eclipse, Maven plugin in Eclipse or cmdline Maven command. For example, if you right-click the project -> run as -> maven test, this should be part of your output

[INFO] Scanning for projects...
[INFO] ------------------------------------------------------------------------
[INFO] Building Top2000Persistence
[INFO]
[INFO] Id: Top2000:Top2000Persistence:jar:1.0-SNAPSHOT
[INFO] task-segment: [test]
[INFO] ------------------------------------------------------------------------
[INFO] [resources:resources]
[INFO] Using default encoding to copy filtered resources.
[INFO] [compiler:compile]
[INFO] Nothing to compile - all classes are up to date
[INFO] [resources:testResources]
[INFO] Using default encoding to copy filtered resources.
[INFO] [compiler:testCompile]
[INFO] Nothing to compile - all classes are up to date
[INFO] [surefire:test]
.....
.....
.....

Results :

Tests run: 4, Failures: 0, Errors: 0, Skipped: 1

[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESSFUL
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 9 seconds
[INFO] Finished at: Mon Jun 29 11:28:38 CEST 2009
[INFO] Final Memory: 2M/9M
[INFO] ------------------------------------------------------------------------

If you experience other results that the above, you might need to manaully download libraries and install then into your local maven repository. In the section resources you will find the resource for the maven documentation and the maven install plugin documentation.

The same you must do for the Top2000Webservices project.
The way I do this is the following:

File -> new -> Other -> Maven -> Checkout maven projects……
select svn from the dropdown
Paste the url http://top2000webservice.googlecode.com/svn/trunk/
click on Finish

The project wil be downloaded into your Eclipse and configured as a maven/web project/java. The dependency Top2000Persistence will be included by the maven process. You should have 2 project which look something like this:

New Maven Web project

You can now do right-click project -> run as -> maven test, the output should be as follows:

[INFO] Scanning for projects...
[INFO] ------------------------------------------------------------------------
[INFO] Building Top2000Webservices
[INFO]
[INFO] Id: Top2000:Top2000Webservices:war:1.0
[INFO] task-segment: [test]
[INFO] ------------------------------------------------------------------------
[INFO] [resources:resources]
[INFO] Using default encoding to copy filtered resources.

......
......
......

[INFO] ------------------------------------------------------------------------
[INFO] BUILD SUCCESSFUL
[INFO] ------------------------------------------------------------------------
[INFO] Total time: 6 seconds
[INFO] Finished at: Mon Jun 29 12:18:37 CEST 2009
[INFO] Final Memory: 3M/12M
[INFO] ------------------------------------------------------------------------

Now we have 2 projects in our workspace, where the web project has a dependency on the persistence project. Maven test confirms that the projects compile and run the available test without problems.

The next thing we need to do is configure the HSQL database location for our webservices project. The persistence project uses the HSQL database in memory type for testing purposes. For usage in other purposes you need to create a file called local.config and put a set of properties into it. Below a sample of the required properties:

ds_driverclassname org.hsqldb.jdbcDriver
ds_url /test <-- You need to provide this
ds_username sa
ds_password
vendor_dbplatform org.hibernate.dialect.HSQLDialect
showsql false
generateddl false

Make sure the local.config file is in your webapp folder. (I have provided a sample in there when you download the project. Just replace my url with your own).

Next thing we need to do is add a server config to our eclipse workspace. I am not going to explain in detail how to do this, since I expect you already know how. But in short terms:

Download and install tomcat 6.x
In server view right-click –> new server –> select tomcat 6 –> point it to the home dit of tomcat
Add the top2000webservice project
Finish

Now you can start the server and the Top2000webservice application should start without any problems.

Resources

Here are usefull resources that can help you in running and working with the example project.

Easy JavaScript client-side form validation

admin — Wed, 03 Jun 2009 07:46:00 +0000

When you are developing a website or an application, there will come a moment that you have to start using Javascript to validate your HTML forms. Here you have 4 options. Option one is not to validate. Although tempting, it is not the best option to use. Option two is to depend on your application framework and hope they have implemented client-side validation into it besides the server-side validation. Option three is to custom code some basic validation scripting yourself.
This is a great way to learn more about Javascript, but also the figurative pain in the butt. Option four and final option is to grab a existing Javascript validation framework from the Net and use it. I would recommend option four which brings me to the actual goal of my post

Many Javascript validation frameworks that are available have one thing in common. You MUST code Javascript to use it. And this is just the thing you are trying to avoid. Spry is a nice example here. Very powerful framework, but you have to know about Javascript to use it. Would it not be nice that u can use a Javascript validation framework without having to know Javascript. I know some designers would love it, since they do not have to depend on other developers for the Javascript. On the other side. I as a developer can spend my time on other issues besides validation coding. That mean that I will have more time if I provide my designer with a non intrusive way to implement client-side Javascript validation.

Criteria for Form validation

As you can imagine I have done some coding and a basic framework has been created. I have build a client-side Javascript form validation framework that adheres to the following criteria.

No Javascript coding is necessary for basic validation
Error messages should be displayed where you want then to
Language support
Easy extensibility for custom validation code

Again it is not a all encompassing framework, but it will give you the necessary tooling to make it suit your purpose.

Requirements for the Javascript Form Validation framework

Let us start with what we have to add to our HTML page and form in order for this validation framework to work. Only two thing are needed to activate the framework in your HTML form.

Add the Javascript file to your HTML page
Add validation=”true” to your form-tag

Add the Javascript file

Add the validation attribute in the form-tag

DONE. This is all you have to code to make the Javascript validation active. But it does not validate anything yet.

Form validation options

Adding validation is just as easy as activating it. You have to add attributes in the input-tags to use validation. Before I show you how to do this I will sum up the validation options you have.

required check
Zipcode NL check
E-mail check
ISO date check

This is not much, but as I said I have build a basic framework which can easily be extended and augmented if necessary.
Besides these validations you can also use

Masking of Text input fields
Add custom Javascript checks

Enough said. Here is some example code in which I use the options above

Required check

Date check

As you can see, adding client-side form validation is a no brainer with this Framework. There are some other things you might want to know about. First, error messaging. In order to use error messaging I have provided two means to display them. The first one is global errors. This means all errors will be collected and displayed in one location of your choosing. The other option is local errors. Error will be displayed near the input tag.

How does javascript form validation work?

Well you have to add another attribute into the form-tag and you must add either a div-tag or span-tag(s) which act as container for the messages. For example. You want to display all errors in one place. This is how you have to do it:

Put the messages attribute into your formtag

messages="global"

and Put a message container in your code like this:

Now your messages will all be displayed at the location of the div. If you want to display the messages locally you must do this:

messages="local"

and provide a span-tag for each input-tag you have validation on:

Now the errors corresponding to the naam input field will be displayed in the span naam_error. As you can see the span-id must contain the name of the input field. The framework will use this to find and disply the errors in the correct span.

Custom javascript form validation

I am fully aware of the fact that I have not provided a all encompassing validation library. Thus I have provided the means of extension. You can add your own javascript method, which will be used in the validation framework. The only two things you have to do are:

Add a attribute to your input-tag for the custom validation

And build your own Javascript method
The method should have the signature below. When your custom validation fails you must return the error message to be displayed. Off course you can use a predefined error available in the validation library.

function exampleCustomCheck(message) {<br />
if (document.simpleform.naamCustom.value!='') return null;<br />
return message;<br />
}

Masking in form validation

The last item I want to show of is Masking. I love masking. It gives you a very powerful way of forcing the user to adhere to your desired input. I have implemented 2 types of masking. First there are a set of predefined masking option available. These are:

date_iso
date_us
time
phone
ssn
visa

Also I have provided the means for custom masking. This can be used in 3 ways. You can use the Char A, 9 and X to define what a user can enter in the field. A = alphanumerical, 9 = Numerical and X = all characters. Some examples:

mask="9999 AA"

This means that you can only enter 4 digits, a space will be forced and 2 letters.

mask="XX-XX-99"

This means that you can enter 2 characters, a – is forced, 2 characters, a – is forced and 2 digits.

I leave it up to you for the endless combinations that you can make with this.

This concludes my client-side Javascript validation framework. I hope you will use it to your satisfaction and keep me posted of wishes.

Client-side Javascript validation Framework

Regards.

Marc de Kwant