All 15 entries tagged Flex

I am investigating the possibility of moving the query layer of a web application off the SQL Server that hosts the data (so as to allow me to more easily move the database to another platform). Last week I mastered BlazeDS for Java remoting. That was easy. So I decided to add Spring to the mix, following this excellent article from Adobe Labs. I’ve never used Spring before, but understand how it simplifies such applications.

It didn’t take long to understand the example from Adobe, and adapt it to query the SQLServer database that I have used.

And when I pressed the ‘query’ button on my Flex app, I was astonished. 10,000 rows appeared immediately as an ArrayCollection. Whereas my Flex/JSON/Java app would request 100 rows at a time, with a new query each time the datagrid page was changed, I can now load every single record and immediately work with it in the browser.

Next trick? Hibernate.

Firstly, I must acknowledge this very useful explanation from Ashier's blog. 

A BlazeDS project is simply a standard web application, with a few additional features. In this case I have created a Tomcat application in Eclipse. You can see the various components in this image of the Eclipse resource navigator:  



The additional elements within the web app are:

1. The flex directory in WEB-INF, containing the BlazeDS configuration files for this application, with the required service configurations (for example, making Java classes available to Flash).
2. Various jar library files in the lib directory. 
3. The flash application (swf and html files) generated from the Flex 3.0 project.
4. Some configuration settings added to the web.xml file.

To get it all set up, I followed these steps:

1. Create the Tomcat project in Eclipse.

2. Add the jar files.

3. Add the flex directory with template config files.

4. Modify the web.xml file so that BlazeDS requests are handled.

We need to add the listener...

<listener>
<listener-class>flex.messaging.HttpFlexSession</listener-class>
</listener>

And a message broker servlet...

<servlet>
<servlet-name>MessageBrokerServlet</servlet-name>
<display-name>MessageBrokerServlet</display-name>
<servlet-class>flex.messaging.MessageBrokerServlet</servlet-class>
<init-param>
<param-name>services.configuration.file</param-name>
<param-value>/WEB-INF/flex/services-config.xml</param-value>
</init-param>
<load-on-startup>1</load-on-startup>
</servlet>

<servlet-mapping>
<servlet-name>MessageBrokerServlet</servlet-name>
<url-pattern>/messagebroker/*</url-pattern>
</servlet-mapping>
<welcome-file-list>
<welcome-file>default.html</welcome-file>
<welcome-file>default.htm</welcome-file>
</welcome-file-list>

5. Create a new Flex 3.0 project.

Use settings like this:

Note that I have set it to use LiveCycle Data Services (now known as BlazeDS) and a J2EE application server. In the next screen, the location and url of the web app must be specified. Note how I have set the output folder of the Flex application to be the root of the web application, so that when I run the Flex app it is built and deployed directly into the web app.

6. Write and build a class in the Eclipse web app project.

I created a class called Hello in the package hello. To start with it contains two simple methods. One method is called sayHello, and returns a string. The other method is multiiply(int i, int j). This method returns the value of i * j.

7. Make that class and its methods available to the Flex application via BlazeDS.

The following 'destination' needs to be set up in the remoting-config.xml: 

<destination id="Hello">
<properties>
<source>hello.Hello</source>
</properties>
</destination>

The source maps to the Hello class in the hello package.

I also pared-down the services-config.xml file so that only the remoting service is being used (as suggested by Ashier):

<services>
<service-include file-path="remoting-config.xml" />
</services>

8. Add a RemoteObject to my MXML application file in Flex 3.0.

This points to the class that I have created and registered with BlazeDS. See how the source points to the Hello class in the hello package. Also see how the RemoteObject has its own id. We use that to referece it and call its methods.

<mx:RemoteObject id="blazeService" fault="faultHandler(event)" source="hello.Hello" destination="Hello">
<mx:method name="sayHello" result="resultHandler(event)" />
<mx:method name="multiply" result="resultHandler(event)" />
</mx:RemoteObject>

Notice how I have registered the two methods, and specified a resultHandler method (in this case the same handler, as it just displays the string that is returned in a text box).

9. Call one of the methods, and handle the result.

I have a button that, when pressed, calls a method that calls the multiply method on the RemoteObject...

blazeService.multiply(i, j);

This method executes asynchronously on the server, in the class Hello. The values of i and j are passed to the Java method. The result is receieved bythe specified result handler method in the action script:

private function resultHandler(evt:ResultEvent):void {
result_text.text = evt.message.body.toString();
}

In this case, I simply get the single string that is returned, and display it in a text box.

Alternatively, I could get a series of arguments, an ArrayCollection, an Object. I can also create a value object class in the Flex app, mirroring a class in the Java. This class may then be passed back and forth in the method call (i'll document this in another article).

A taxonomy is a hierarchical structure of categorisations, with a progression to more detailed categories towards the leaves. The taxonomy defines the set of categories that may be applied to resources. Taxonomies are common in academia. Sometimes they are formal and explicit mechanisms. Often they are informal aids to research and academic communications.

Many web publishing systems allow content to be “tagged” with keywords. For example, a Sitebuilder page may be given one or more tags. Similarly a blog entry is classified using tags. Social bookmarking systems like del.icio.us use keyword tagging. Bibliographical databases also commonly use taxonomies of keywords. Most of these systems can be searched using keywords. Many of these search interfaces can be accessed programmatically.

Most keyword tagging and searching systems, such as that used in Sitebuilder, are not taxonomical. There is no taxonomical structure to which an author or researcher can refer when tagging resources or searching for resources.

I have identified a set of real use cases in which a taxonomical approach to keyword searching would be beneficial. These cases also imply that resources are tagged according to the keywords available in the taxonomy.

Here is a description of how the tool could work:

Searching

The search interface contains a tree structure representing the taxonomy of keywords.
The tree structure can be explored, drilling down through its branches.
Individual keywords, or whole branches of keywords can be selected.
There is also a text box allowing the user to type in keywords (with autosuggest giving options as they type).
As keywords are selected, they are listed in a text area showing all currently selected search terms.
Once all of the required keywords have been chosen, a search is done, returning a list of all matching resources.
The search could be ordered in several different ways, for example with resources that have a higher match coming at the top of the list.

Here is an illustration of how the interface might work (it doesn’t yet do the search):

Configuring the taxonomy

The taxonomy could be stored as an xml file on the web (for example in the same location as the search tool).
The xml file could be hand coded, or the search tool could provide an interface for creating a taxonomy file and editing its tags.
The taxonomy file to be used could be specified by the author of the web page on which it appears.
The end user could be allowed to choose a taxonomy file, it could be possible to search for taxonomy files in Sitebuilder.

Configuring the sources

The search tool will need to know about the web application in which resources are stored (for example Sitebuilder).
It will need a method for searching each web application (how to build the search url).
One or more sources could be specified by the author of the page on which the tool is deployed – it could be set to search Sitebuilder, Blogs, etc.
A version of the tool could be made available allowing the end user to choose sources.

Saving and sharing searches

Searches could be saved onto the end users local machine (Shared Object in Flash, filesystem in AIR).
Searches could also be shared with other users.

Using the taxonomy to tag resources

The search tool could be used to assist in tagging resources using a selected taxonomy.
For example, a user selects a series of keywords, and the tool displays the text to paste into the keyword tag field of the Sitebuilder page properties.

Tasks

1. find out how to save an xml file from Flash into Sitebuilder.

The application behaves just as before, but the plumbing behind it has been rationalised and improved using Cairngorm. The project now looks like this in the Navigator:

The migration was quite straightforwards. The design of the original app was quite similar to that of a Cairngorm app. However, the framework offers a few neat tricks:

1. A simple and easy to understand event based controller. Views can initiate events, which are then passed onto the required command by the controller. Once the event is called, the originating view is forgotten about.
2. The AppModelLocator singleton, which gives access to centrally stored value objects. Elements in views are bound directly to value objects and properties in the model. If a command updates the model, then changes are made to the bound elements.
3. In this app, tab navigators are used to contain the various views. The TabNavigator.selectedIndex property of each tab navigator is bound to a property in the model. That means that a command can change the currently displayed tab by changing the relevant property in the model.
4. Chaining of commands – a command can call another command. For example, when viewing a single publication record, the GetSinglePublication command calls further commands depending upon the type of publication that is being viewed.

Most importantly, my code is now much easier to understand and to develop further. As Cairngorm is a widely understood and documented framework, it will be easily understood by other developers.

It’s possible. If we do it, then this would be the course material. I reckon (optimistically) that it could be covered in half a day….

1. What is Flex capable of?
   1.1 What is a Rich Internet Application
   1.2 Examples
       1.2.1 Databases (and Hibernate)
       1.2.2 XML and RSS
       1.2.3 Charting
       1.2.4 Video and audio
       1.2.5 The Adobe Internet Runtime

2. The development process
   2.1 Introduction to Eclipse and FlexBuilder
   2.2 Debugging and inspecting
  2.3 The Cairngorm framework

3. Designing an application as a set of user interface views
   3.1 Creating a view as a custom component
   3.2 Indexing views in a ViewStack (and other navigation controls)
   3.3 Using Flex components to build user interface views
       3.3.1 Using containers to control layout (canvas, vbox, hbox, panel)
       3.3.2 Using Flex interface controls (label, text, textarea, button, datagrid, list, tree)
       3.3.3 Binding controls to data

4. Encapsulating data as models and value objects
   4.1 Creating custom classes
   4.2 Using the AppModelLocator
   4.3 Binding view controls to data in the AppModelLocator

5. Making things happen
   5.1 Creating event handlers in ActionScript within views
   5.2 Creating custom events to call actions and to carry data
   5.3 Creating command classes to process events and to update data and views
   5.4 Mapping events to classes in AppController

6. Connecting with external data services
   6.1 Loading xml and other data using HTTPService
   6.2 Coping with security sand box restrictions
   6.3 Creating and using an asynchronous service delegate
   6.4 Accessing and using the results of a service delegate

7. Publishing your Flex application using SWFObject

Interested?

Explanation

Here is a screenshot of the Navigator window for the project:



The Cairngorm source code is included in the "com" folder (although it could be compiled and included as an external library). The "uk.ac.cairngormtemplate" folder contains my application. This has six subfolders, each containing classes responsible for different parts of the MVC architecture: view, control, command, business (actually meaning data and comms services), model, vo (value object). As with all Flex applications, there is a single container mxml file, called Main.mxml (an mxml file is equivalent to a html file, with code, styles and visual components). I'll start my explanation of this app and Cairngorm with this top-level container.

The Main.mxml file contains four key items:

1. a Services component - this contains all of the services used for connecting to external data sources etc;
2. the AppController - this maps events onto commands, so that when a ui dispatches an event (a request for something to happen), the correct code does the required job;
3. the AppModelLocator - this stores and indexes all of the data models used in the application, all of the data is stored in one place!
4. a ViewStack listing all of the alternative user interface views that the app can display.

Here's the code for the ViewStack:

<mx:ViewStack width="100%" paddingBottom="10" paddingTop="10" resizeToContent="true" selectedIndex="{model.currentView}">
   <view:InitialView />
   <view:AlternativeView />
</mx:ViewStack>   

There are two views in this application, an initial view and an alternative. Each view is defined in its own mxml file in the "uk.ac.cairngormtemplate.view" folder. Notice how the selectedIndex property of the ViewStack is bound to the .currentView property contained in the modelLocator (initially set at 0, which means that the first view in the stack is displayed). This means that the currently selected view may be changed by changing the model.currentView property (more on that below).

So now let's drill down into the InitialView component. We can then see the details of what is presented to the user, and what happens when the user responds to that interface.

InitialView is itself an mxml component based upon the standard VBox component. It contains some script and some UI components. Firstly, the script contains a bindable reference to the AppModelLocator singleton. This means that we can bind UI components to data in the AppModelLocator. There is in fact a text component that is bound to a value object (a class that wraps a set of value properties):

<mx:Text text="{model.messageVO.message}" />

Initially this displays the message "Hello everyone".

Below this text component there is a button "translate". Clicking this button calls an actionscript method in the script on the page. The method is called translate(). It could at that point do some validation, or request further input. However, in this case all that we want it to do is fire of a request for the text "Hello everyone" to be translated.

The least impressive Flex app ever

This is where Cairngorm starts:

private function translate():void
{
   var cgEvent : DataChangeEvent = new DataChangeEvent("Dumelang");
   CairngormEventDispatcher.getInstance().dispatchEvent( cgEvent );               
}

The method creates an instance of the DataChangeEvent, passing it the new message "Dumelang". The DataChangeEvent is defined in the "uk.ac.cairngormtemplate.control" folder. It is a cairngorm event extended to carry our message. We can create events that can carry whatever data we need to be passed to a command. (For example, a user types their username and password into a login box, this is then wrapped in a LoginEvent, and the event dispatched. A LoginCommand receieves the event and attempts a login using a LoginDelegate and login service. If succesful, the command changes the current view to one appropriate to a signed in user).

The next step is to dispatch that event and its message to the Cairngorm controller. This should then get passed on to the required "command" object (the command object contains the code for executing the requested command). But how does the controller know which command object to use? Events are mapped on to commands in the control.AppController class:

public class AppController extends FrontController
{

   public static const DATACHANGE_EVENT : String = "DATACHANGE_EVENT";

   public function AppController()
   {
       addCommand( AppController.DATACHANGE_EVENT, DataChangeCommand );
   }
}

Our request (to translate "Hello everyone") is then passed to a DataChangeCommand object. This is where we do the work that results in the translation. In this simple example, the .messageVO.message of the modelLocator is simply changed to read "Dumelang". However, we could instead refer to an English-Steswana dictionary to retrieve the translation. This might even require the use of an external data service. The DataChangeCommand would then delegate the work of sending this external request to a Delegate class in the "business" folder. The delegate, on retrieving the required translation, calls back to the DataChangeCommand to complete the requried command, or alternatively calls a seperate command.

So, with the text control on the current view bound to the .messageVO.message property in the modelLocator, and that property now altered (by DataChangeCommand) to "Dumelang" that is precisely what the end user sees on their screen.

But what if we also need to change the current view, for example to the AlternativeView control in the ViewStack? Easy. The command simply alters the .currentView property of the modelLocator. As the selectedIndex property of the ViewStack is bound to .currentView, it immediately changes.

In this example, at the end of DataChangeCommand, a second command is requested using a class that I have called ViewChangeEvent.

var cgEvent2 : ViewChangeEvent = new ViewChangeEvent();
CairngormEventDispatcher.getInstance().dispatchEvent( cgEvent2 );

This class calls ViewChangeCommand, which simply alternates between the InitialView and the DefaultView.

Look - it speaks Setswana! and now the view has changed to one with a white background! Hoorah!

Conclusion

This may be a simple example, but I can already see how using the Cairngorm framework will simplify the task of building very complex applications, resulting in better, more rational and sustainable code.

1. Think about how you might want to be able to search/aggregate/organise your content, how other people might want to search it and see it organised, and how you want other people to think about your content (especially if you are trying to establish a schematic structure in their minds).
2. Be systematic, especially with punctuation, spaces, spellings.
3. Maintain a list of your tags. A concept map is a good mechanism for maintaining this list. You could use MindManager to create the list.
4. Cooperate – use the same tags that other people use, develop tagsonomies with other people (formally or informally).
5. Use a combination of very specific and more general tags (for example: e-learning, elab, quizbuilder) – think about your tags as being arranged in a tree structure (specific at the leaves general at the branches).
6. Combine different tags that identify different aspects of the tagged content (for example, use a tag to identify what kind of content it is (essay, review), what it is about (philosophy, Kant).
7. Tagging your work with a unique identifier associated with yourself allows you to aggregate your work from wherever it appears (e.g. robert_o-toole in Sitebuilder gets you this result )

I’m gonna do this….

1) I’m building an RSS tag searching tool to replace the slighty rickety javascript tool that I am using for Renaissance Studies and some other departments.

2) The search tool can load its tagxonomy from an xml file stored on the web.

3) I’m going to add a feature that allows different tagxonomies to be loaded as required. If the xml files are stored in Sitebuilder and tagged as “tagxonomy” then the search tool could even auto-discover tagxonomies for the user to choose from.

4) A tagxonomy file could also be tagged with the unique identifier of a person, group or department (using the web groups codes for these), thus associating the file with the profile.

5) I could allow for the tagxonomy to be edited in the search tool (adding new tags or tag collections for example).

6) The tagxonomy could then be uploaded onto Sitebuilder, if the user has permissions, it could replace an existing tagxonomy, if not they can create their own new tagxonomy.

7) It will also be possible to select a series of tags from the tree, and have an string automatically generated from that. The string can be pasted into any content tagging application. Importantly, the string should include tags that identify the tagxonomy from which the tags where taken.

Best of all, i’ll try to build it as an Adobe AIR desktop application (not sure how Single Sign On will work with that).

I reckon that will take a few weeks to build (as I’m in the middle another big project).

This example will load the tags defined for the Renaissance Studies project:

Does that sound like fun?