Technology & Science Interest

Why I haven't posted in a while

2011-11-24T10:09:00.001-05:00

Some have asked why I have not posted any new software tips in quite a while. 2 reasons.

First. I've been promoted to oversee all the engineers and so that keeps me pretty busy.

Secondly, because I am working on two masters degrees at the same time.

So, I will be happy to answer questions when I have time but I don't plan on posting any new software tips for quite sometime unless I find something so amazing I just have too.

Compiling OpenSSL on Mac OS X Snow Leopard

2010-12-12T07:52:00.000-05:00

Snow Leopard comes with a dated version of OpenSSL 0.9.8l and for various reasons you may want to install the latest version. Here is how I was able to install the latest version (1.0.0c as of the writing of this blog) on Snow Leopard.

NOTE: This Blog assumes you have some familiarity with the terminal, using the command line, and opening and closing applications. If you are not comfortable with any of these things then I do not suggest you try this on a production machine.

Before beginning check your current version of OpenSSL by opening a terminal and typing:

openssl version

You should get back the following if you have not updated OpenSSL since installing Snow Leopard:

OpenSSL 0.9.8l 5 Nov 2009

Assuming you don't already have 1.0.0c or later already installed, you have XCode installed from Apple, and you have administrative rights to the machine you are installing on, then you are ready to begin.

STEP 1: Download the Source Code from the OpenSSL website (http://www.openssl.org/source/).

STEP 2: Open a Terminal and navigate to the directory where you download the source code to. In my case the source code was download to ~/Downloads so I typed on the command line:

cd ~/Downloads

STEP 3: Configure the build by typing

./configure darwin64-x86_64-cc

STEP 4: Type

make

STEP 5 (Optional but fun): Test that everything was created correctly by typing

make test

STEP 6: Install the newly built OpenSSL application by typing:

sudo make install
(Type password when prompted)

STEP 7: Navigate to the directory where the old openssl is installed by typing:

cd /usr/bin

STEP 8: Rename the old openssl file to another name so that you can go back to it if you need to:

sudo mv openssl opensslOld

STEP 9: Create a link to the new openssl file

sudo ln -s /usr/local/ssl/bin/openssl openssl

STEP 10: Test that everything worked by typing:

openssl version

If it all worked you should now see

OpenSSL 1.0.0c 2 Dec 2010

Devoxx Java Conference

2010-10-26T19:30:00.001-04:00

My team and I are off to Belgium in two weeks for the Devoxx Conference. First time I have gone to Devoxx and Belgium! Looking forward to it, even though I despise flying.

Using Apache Velocity with Java Properties File

2010-10-08T09:46:00.003-04:00

The Apache Velocity project is an amazingly useful tool which allows you to merge data with templates. Here is a quick example of how it works.

First suppose you had a "template" text file that looks like the following

$awesomeProgrammer needs to purchase a new $device.

Go ahead and save this in a file called "test.tpl." Then suppose you had a "context" or "properties" file that read like this:

#This is a comment

awesomeProgrammer=Brent Coffey
device=Mac Pro

Save this file as "test.properties." Now you want to programmatically mash together the properties file with the the template file so that the output looks like this:

Brent Coffey needs to purchase a new Mac Pro.

This is what the Apache Velocity Engine will do for you with very little code here is how it works.

Assumptions: First this example assumes you have already have a working knowledge of Java, Eclipse, jars, and classpaths. Therefore create a new project in Eclipse and add the velocity-dep-1.4.jar to the classpath. You can download the jar from the Apache Velocity web sit by clicking here.

Second note that the example I'm giving is a down and dirty example, and I would not recommend using this in production because I'm not going to handle all the try catch statements or possible errors that could arise nor is this an example of "good" OO programming. So take it for what it's worth; a plain-ole super simple example of how to use velocity to merge a properties file with template.

I wrote this code to be really self explanatory. I may come back later and explain more if it's needed or requested. Let me know.

Here is the code:

To run this example ensure that the two files you created above (test.tpl, test.properties) are in the root of the executing directory. Pass those two files to the application as command line arguments in the specified order (template first followed by properties file) and if everything is located in the correct place you should see the string: "Brent Coffey needs to purchase a new Mac Pro." printed out to the console.

Man Pages as PDF

2010-09-30T05:24:00.001-04:00

Sometimes I want to read a man pages in a seperate window from my terminal I am working in. Instead of opening another terminal I read this handy blog on how to open a man page as a pdf. (Why I didn't think of this myself I don't know as I do simular type things for html and jpeg files all the time...oh well). The solution suggested in the blog is to "simply" type:

man -t XXXX | open -f -a /Applications/Preview.app

Where XXXX is the thing you want the man page for such as ls, ipconfig, snmpget, etc. I thought that was quite a handy tip but a bear to type. After all it's pretty simple to just type: man ls. Whereas having to type the other monstrosity of a command makes for an interesting blog but frankly less than useful utility.

So, for fear of stating the obvious, I took the blog one step further and created a script to do the work for me and to greatly simply the task. Here is the simple script, which I named manpdf:

#!/bin/bash
man -t $1 | open -f -a /Applications/Preview.app

NOTE:There are tons of ways you could make this even more useful (post those in the comments if you wish):

Save the script, as root, this script in your path, in my case I saved it to /usr/bin. Then make it executable by typing:

sudo chmod uga+x manpdf

Now you can simply type: manpdf XXXX, to open a man page for a given command.

GeoTagging Photos and then What to do with them

2010-08-16T14:01:00.002-04:00

How to GeoTag Photos and What to do once they are GeoTagged

My Hardware:

1. Canon EOS Digital T2i Camera
2. Delorome PN-40 GPS
3. MacBook Pro
4. LowePro FastPack 350

My Software:
1. Mac OS X
2. GPS Babel (Free)
3. PhotoLinker ($50 recommended) or GPSPhotoLinker(Free)
4. Aperture 3

Web Sites to use GeoTagged Photos (Listed in the order I would recommend using them)
1. Panoramio.com (Easiest to use if you want your photos in Google Earth, only for scenic/landscape photography!)
2. EveryTrail.com (If you went on a trip and you are logging your journey. Can place pictures and video of your trip along with GPS Track)
3. HistoryPin (For Historical Photos of a place)

Start of Photo Shoot Step 1:
1. Ensure your Camera has the correct timeset, battery is fully charged for GPS and Camera.
2. Turn on your GPS
3. On your GPS, turn tracks on at the start of your photoshoot or trip.
4. Once satellite lock is aquired on GPS begin journey toss gps in side mesh pouch of LowePro backpack and forget about until you are done.
5. Hike and shoot photos at will.

After Photo Shoot Step 2:
1. Stop GPS Track and Save.
2. Turn off GPS.
2. Turn off Camera.

Back at your Computer Step 3:
GPS Babel

1. Copy Pictures to new folder on your computer. (Do not import to Apeture or Lightroom or anyother photo management tool yet!)
2. Open GPS Babel
3. Connect GPS to Computer
4. Turn GPS On (and ensure GPS is ready to send data to computer (differs by model)).
5. Select GPS From top pull down menu in GPS Babel
6. Ensure ONLY Tracks is checked
7. Select GPS XML from dropdown menu
8. Click on File Name and choose to save your GPS XML file in the folder you created in step 1 above.
9. Hit Apply
10. Close GPS Babel.

Photolinker
1. Load Track (GPS XML file created above). Click Tracks->Load Tracks from File...
2. Load all pictures saved above. Click Photos->Load Photos from File
4. Click on GeoTagging Console button in the lower right side.
5. Set Timezone to the timezone that the Camera is using for the pictures!!
6. Click on one picture.
7. Select all loaded photos (Ctrl+A)
8. Click Geotag X Photos button in lower right side of GeoTagging Console
9. Click Write Changed Tags
10. Exit Photolinker

Aperture

1. Open Aperture.
2. Click Import
3. Select folder where new photos were stored
4. Import Pictures as you would normally
5. Make any adjustments to your photos as you normally would using Apertures tools.
6. Once photos are imported, click on the Places button (Top Right Side) to see your photos all geotagged.
7. Go back to Split View by clicking Split View Button (Top Right Side).
8. Click File->Export->Versions...
9. Create a new folder and name it "contrib"
10. Give the photos a generic name
11. Click ok.
12. Wait for all photos to be exported (may take a while).

NOTE: You now have JPEG versions of all of you photos on your harddrive. These can now be used to upload to web sites. Aperture to the best of my knowledge does not have an export for EveryTrail, Panoramio, or HistoryPin so this is why the manual export was necessary though annoying.

Using your Photos to Benifit the World Step 4:

Panoramio.com
1. Open Browser
2. Go to www.panoramio.com
3. Create an Account
4. Upload photos in your "contrib" folder by clicking Upload link at top of page.
5. Follow Directions on site.
6. Your photos, if approved should be avaible in on google earth in a few days, though it may take up to a month. My experience has been days though.

EveryTrail.com
1. Go To www.everytrail.com
2. Create an Account
3. Upload GPS XML file you created
4. Write a few sentences about your trip.
5. Upload Photos from the "contrib" folder.
6. Follow additional directions and photos should be placed on track correctly.

HistoryPin

You get the idea...same type thing... follow web sites directions

Final Step Step 5:

1. Tell all your friends about your photos on Panoramio and EveryTrail using the twitter and facebook links on Panoramio and EveryTrail!
2. Post your pictures on your blog! ;-)

I’m on

bcoffe
1 trip
see my profile

Recent trips by bcoffe

Check out EveryTrail mobile travel guides

JUnit Parameterized Test

2010-08-06T10:25:00.001-04:00

Sometimes you want to test a method with various inputs to perhaps check to ensure the method works with all kinds of different data. I had just this issue the other day and knowing that JUnit 4 supports "Parameterized" Test I decided to use that methodology. Here is how it works.

Problem

Imagine you have a method in a class that makes a word plural by adding an "s" or an "es" to the end of the word (yes I know, silly example, but just go with me on this!).

Example if you pass in the String "dog" it should return "dogs." If you pass in the word "fish" it returns "fishes." (Because this is a simplistic example and this blog is about parameterized test, not how to make a word plural our method will only work for these two words.)

Method to Test

JUnit Test

Comments in the code tell what is being done

Thats It. When you run your JUnit test in the Eclipse IDE you should see this result:

Turning a White Sky, Blue

2010-07-18T07:28:00.001-04:00

I work a lot with Photoshop for various software engineering projects, but every now and then I have a family project that requires the use of my Photoshop skills. The other day my oldest daughter and I went on a brief hike. Along the way I stopped to take some pictures with my new Canon Rebel T2i and Ultra-Wide-Angle lens. I got some great pictures but lots of them showed the sky as white instead of blue. A very common photograph problem. Here is one of the pictures before I photoshopped it:

As you can see the sky is mostly white. On the web there are various techniques people blog for fixing this, but most of them are tedious as they require selecting the entire sky, including all of the hints of sky inbetween the leaves! Yuck.

So here is how I fix the problem in less than 1 Minute without having to tediously select every hint of sky.

1. Create a new layer by clicking the new layer icon at the bottom of your layers window

3. Change the Overlay mode to Darkest Color (This is the magic step!)

2. Select the new layer by clicking once on it.

3. Click on the Gradient Tool

4. Ensure that Gradient is set to Linear Gradient and then create a gradient that goes from the desired dark blue color to a light color blue by clicking on the gradient controls at the top. Click the drop down color bar to set the colors of the gradient and click the icon next to it to change to Linear Gradient

5. I choose the colors #269df0 and #dce2ec as the dark and light colors, respectively, inside of the Gradient editor.

6. Holding down the left mouse button, on your new layer, from about where you want the lightest part of the sky drag up and a little past the top of your picture and let go. You may need to do this a couple of times to get the colors right where you want them. Here are my end results. (For more of a real sky look download some free cloud brushes and use them to put a few white clouds in)

7. For even a better effect. Do the water as well! You do it the exact same way, creating a new layer, etc. but drag the gradient from middle to bottom (instead of middle to top as you did on the sky).

Final result with water:

Highlight in Dock Stack Snow Leopard

2010-06-14T17:46:00.000-04:00

I really love my mac. I love it so much I can't figure out how I, at one time, developed on a windows computer. I can't imagine going back. What I love about the mac is that just about anytime I find something I don't like about the way it works I can usually just find a way to change it.

I didn't like the fact that on the dock stack when I placed my mouse over something it didn't hightlight. Googled and found the answer here. Here are the steps:

Open Terminal
type: defaults write com.apple.dock mouse-over-hilite-stack -boolean YES
type: killall dock

Thats it! Now your dock stack will highlight as you mouse over an item.

Execute a formula stored as a String

2010-06-03T13:00:00.004-04:00

I recently had a situation where I needed to calculate a formula that came from an outside source as input into my program. In this case the formula was from a record in the database. Each formula was different.
Coming from a database, the formula was passed into my application as a String so it looked something like this "(x + y) /z". Since the formula was something sent to the application during runtime I needed to parse the String and turn it into an actual formula, replacing the variables (i.e. x, y, and z) with actual values that came from another place. In this case other servers on a network.

Then to make things more complicated some of the formulas where not your typical math formula but where more like this "(x + Math.Round(y)) / z" or something to that effect

So what to do? I needed a robust solution that could handle parsing a String and executing it as a valid java code during runtime. This is when I discovered the really cool open source product called Janino.

Here is how it works.

1. You pass in the Parameter Names
2. You pass in the Parameter Types
3. You pass in the result type (in my case all mine would be doubles)
4. You pass in the formula
5. Call evaluate

This example is taken almost right from there web site. You can also do even more powerful stuff with it. But this is just to wet your appetite.

Definitely check this out. Better than writing the code yourself. Janino is available from the Maven central repository.

AspectJ Softening (Getting Rid of Annoying Try Catch Statements)

2010-06-03T09:07:00.004-04:00

I love Java. I really do. But I absolutely hate the "try catch" construct. Granted sometimes I can recover from a thrown exception by putting meaningful work in the catch statement. But more often than not I find myself simply putting a log statement in the catch section of the code. So the code looks like

Now I end up "crosscutting" my code all over the place with log statements and the code is harder to read because the meaningful work is obscured by boilerplate try catch code. What is an engineer to do?

Well fortunately the people at AspectJ have a way to fix this with a construct. Here is what it looks like:

The key words "declare soft" are followed by the type of exception you want to catch in this case "Exception."Then they are followed by the "pointcut" in this case

This will now catch the exception thrown from "someMethodThatFails()." Which is exactly what I want! Now one last step to log the error I write a really good log statement:

Now I can back to my original code and remove the try catch block and am left with this one line of code!

Beautiful easy to read, and now all my log statements are not crosscutting the code but are elegantly all contained inside of one Aspect which I named ExceptionHandler.aj

I defiantly recommend doing this everywhere you have a try catch and no real work is being done in the catch.

One last trick. Suppose a constructor throws an exception, as is the case if you are using the Janino ExpressionEvaluater constructor? How do you catch that? Well this gave me fits the key is to use the "new" as if it where the method. Here is what I mean:

See I use the keywords "declare soft" followed by the exception I want to catch ("ScanException"), followed by a valid constructor pointcut "within(org.example.myClass) && execution(*.new(..));"

NOTE I had significant problems when I tried putting something before the "new" besides a splat "*"

Hope this helps get you started.

Change to new BASH Shell (4.1) for Mac OS X

2010-05-08T16:01:00.008-04:00

UPDATE: THIS STILL WORKS WITH BASH 4.2
Suppose you want to do some uber awesome shell scripting. So you need to get the latest version of the Bash shell installed, this way you can use all the new features. Well have no fear, this is super simple. Here is how to update the default bash shell (3.2) to bash 4.1 on Mac OS X Snow Leopard 10.6.3.

(Be sure you have admin/root rights to your box and that you have Apple XCode installed. If you do then you are ready to get started. If you don't have apple XCode installed it can be downloaded for free from Apple's Developer Website but you must be registered to get it.)

NOTE: This blog was a major help, but it skipped one step that might mess up a newbie so I'm explicitly documenting it here.

Step one download the latest Bash code by clicking here.
Open up your terminal and change to the directory where you saved the download file. In my case I typed in:cd Downloads
Uncompress the tar file by typing: tar zxvf bash-4.1.tar
Change to newly uncompressed directory : cd bash-4.1
Now configure, compile, and install code by typing: ./configure && make && sudo make install
Make the new shell your default shell by typing (NOTE: replace {user_name} with your actual user name. You can get your actual user name if you do not know it by typing 'whoami' before typing this command): chsh -s /usr/local/bin/bash {user_name}
Now you need to register your shell by typing: sudo bash -c "echo /usr/local/bin/bash >> /private/etc/shells"
Shutdown your terminal and reopen it.
Check to make sure it all worked by typing: help If it worked it should say the following: GNU bash, version 4.1.0(1)-release (i386-apple-darwin10.3.0))
Change to /bin by typing: cd /bin
Move bash to bash-old: sudo mv bash bash-old
Create a symlink to the new bash command: sudo ln -s /usr/local/bin/bash bash

All Done!

Google Earth Distance of Track

2010-04-12T19:10:00.003-04:00

Well I hate to say it but sometimes I miss the obvious...or not so obvious. Today I went for a Bike ride with my Delorme PN-40 attached to my bike. Again I wanted to know the distance of my ride using google earth (obviously the GPS tells me but I want it in google earth!). So following my last blog I was getting ready to upload the data to the aforementioned web site, when I discovered how to do this directly in Google Earth!

Drum role.... Click the imported track over in the left hand side! Once you do that a dialog appears and tells you the distance. Ugh...... Why that had to be so hard to figure out I don't know.

You can even click on individual waypoints along your track and the same dialog will appear but tell you your speed and time at the given waypoint!

Importing a Track from a DeLorme PN-40 to Google Earth Using a Mac

2010-03-20T16:01:00.003-04:00

I bought a new Fuji Sarges 1.0 bike yesterday and I have to say I love it! It's comfortable and perfect for running around the neighborhood with the kids. I only live about 5 miles from my office so I may also try riding it to work a few times this summer once I get used to the gears and figure out where to park it at the office. But like all good Engineers I wasted no time in thinking about how I could hook up my GPS to it to figure out how fast I am going, track my rides, and figure out how far I went on a ride. Here is how i figured I did it.

First off note that you can find your distance just by turning on your gps enable a new track when you start your ride and then save your newly created "track" at the end of your ride. When you save it will tell you the total distance. But that's no fun! I want to see the track in Google Earth on my mac!

Step One: Get the Track off of the GPS

Download the free GPS Babel tool for the mac from www.gpsbabel.org. Be sure to scroll to the bottom of the Dowmload page and choose the Mac Beta version. I had issues with the release version.
Install GPS Babel as you would any Mac software.
Connect your DeLerome PN-40 to your mac via the provided USB Cable
Turn on your GPS
Open GPS Babel
When your GPS prompts, Choose Data Transfer mode by clicking enter
Import Tracks with GPS Babel and Save them as a KML file. Hit Apply

NOTE: Here is what your GPS Babel screen may look like:

Step Two: Import to Google Earth

Open Google Earth
Click File->Open
Select the File you saved

Your Path and track points should now show up on Google Earth.

But wait I want to know the distance of my path? I could find no way to do this in google earth itself! Unbelievable! I would think I could right click on the path and click a label called "Distance" but alas, there is no such option. Well lots of other people had this same issue and so here is what you can do:

Step Three: Third Party To the rescue

Open Safari
Go to http://www.emaltd.net/google/gec/utilities/
Go Back to Google Earth
On the left hand side, find your path
Right Click
Click Copy
Paste the path in the text box on the web site provided.
Check British Units if you want the results in miles
Click Calculate

Here are my results:

Remember to click the "View in Google Earth" button on the right.

This list is only the list of distances.

Total: 1.3765 mile 1 of 1: Bike around neighborhood (Path) Total: 1.3765 mile

Using AspectJ for the Observable Design Pattern

2010-03-19T09:36:00.007-04:00

In a nutshell the Observable design pattern says that certain entities (i.e. Observers) want to be notified when an attribute of a class changes. The class that is changing we call the the "Subject" or the "Observable." Often the "subject" is a pojo with just a bunch of getters and setters. A common example would be a pojo representing a record in a database table. Suppose you have a table in a database called "Animal." Each record in that table represents an animal such that the table has the following columns:

Id, Name, Feet, Says

and the following type of data:

1, Cat, 4, meow
2, Dog, 4, woof

The records in the database then can be represented by a simple class made up of getters and setters like the following:

package org.example.model;

public class Animal {

private int id;

private String name;

private String says;

private int feet;

public int getId() {

return id;

}

public void setId(int id) {

this.id = id;

}

public String getName() {

return name;

}

public void setName(String name) {

this.name = name;

}

public String getSays() {

return says;

}

public void setSays(String says) {

this.says = says;

}

public int getFeet() {

return feet;

}

public void setFeet(int feet) {

this.feet = feet;

}

Now suppose some class, an Observer, wants to be notified when a value for a given animal changes. Such that when:

animal.setName("Iguana");

the Observer is notified of the change. The best way to do this is using AOP. Here is how.

First create a new interface called subject. This interface will need to be implemented by all the classes that you want to observe. Don't worry though you don't have to change all your classes. We are going to use "static" crosscutting to alter the structure of your objects without actually touching your object. Here is my subject interface:

package org.example.designpattern.observable;

public interface Subject {

public boolean addObserver (Observer observer);

public boolean removeObserver (Observer observer);

public void notifyObservers ();

}

Next we are going to implement the Subject interface with a new class I called Observable. Here is my observable class:

Now the fun stuff. Somehow we have to get our pojo (in this case our pojo is Animal) to implement "Subject" but we don't actually want to change our pojo class. So how do we do this? Well here is how "static" cross cutting works. Create a new aspect and use the @DeclareMixin annotation to tell the weaver which class this aspect is supposed to crosscut. Here is the code:

package org.example.aspects;

import org.aspectj.lang.annotation.DeclareMixin;

import org.example.designpattern.observable.Subject;

import org.example.designpattern.observable.Observable;

public aspect ObservableDesignPattern {

@DeclareMixin("org.example.model.*")

public Subject makeObservable(Object observableObject) {

return new Observable(observableObject);

}

Notice that this code says everything that is in the org.example.model hierarchy is to implement the Subject interface and the actual code that does the implementation will be found in the Observable class we created above. This way no code has to change in "Animal" (i.e. our Subject pojo) at all!

Next we need to do one last thing. Remember we said that the Observer should be notified when something changes. Well in order to do this we need to use "dynamic" crosscutting to call the notifyObservers method after every setter is called in the Animal class (i.e. remember that is our Subject we are watching for changes in).

So we expand our aspect to look like this now:

public aspect ObservableDesignPattern {

@DeclareMixin("org.example.model.*")

public Subject makeObservable(Object observableObject) {

return new Observable(observableObject);

}

pointcut notifyObservers (Subject subject, Object newValue) : execution (public * org.example.model.*.set*(..)) && target (subject) && args (newValue);

void around (Subject subject, Object newValue) : notifyObservers(subject, newValue) {

proceed(subject, newValue);

subject.notifyObservers();

}

Now we are almost done but there is a bug! This will call notifyObservers every single time someone calls "set" regardless if they passed in a new value or not. But remember the Observer only want to be notified when something actually changes. So we expand our aspect one last time to look like the following. Notice we are using reflection to compare the before value with the new value and are only calling notify if they are different. Here is the code:

public aspect ObservableDesignPattern {

@DeclareMixin("org.example.model.*")

public Subject makeObservable(Object observableObject) {

return new Observable(observableObject);

}

@SuppressWarnings("unchecked")

private Class[] argTypes = new Class[] {};

private Object[] args = new Object[] {};

pointcut notifyObservers(Subject subject, Object newValue) : execution (public * org.example.model.*.set*(..)) && target (subject) && args (newValue);

void around(Subject subject, Object newValue) : notifyObservers(subject, newValue) {

try {

String getMethod = "g"

+ thisJoinPoint.getSignature().getName().substring(1);

Object value = subject.getClass().getMethod(getMethod, argTypes)

.invoke(subject, args);

boolean clean = (newValue == null) ? value == null : newValue

.equals(value);

if (!clean) {

proceed(subject, newValue);

subject.notifyObservers();

}

} catch (Exception ex) {

throw new IllegalArgumentException(ex.getMessage());

}

That's it. To test the code we write a JUnit test like the following. If everything is setup correctly all test should pass..

package org.example.designpattern.observable;

import static org.junit.Assert.assertEquals;

import org.example.model.Animal;

import org.junit.Before;

import org.junit.Test;

public class ObservableTest implements Observer {

Animal animal;

boolean subjectChangeCalled;

@Before

public void setUp() throws Exception {

animal = new Animal();

subjectChangeCalled = false;

}

@Test

public void testAddObserver() {

boolean result = false;

result = ((Subject) animal).addObserver(this);

assertEquals(true, result);

}

@Test

public void testRemoveObserver() {

boolean result = false;

((Subject) animal).addObserver(this);

result = ((Subject) animal).removeObserver(this);

assertEquals(true, result);

}

@Test

public void testNotifyObserversDoesGetCalled() {

((Subject) animal).addObserver(this);

animal.setName("Iguana");

assertEquals(true, subjectChangeCalled);

}

@Test

public void testNotifyObserversDoesntGetCalled() {

((Subject) animal).addObserver(this);

animal.setName("Iguana");

subjectChangeCalled = false;

animal.setName("Iguana");

assertEquals(false, subjectChangeCalled);

}

public void subjectChanged(Object observerableObject) {

subjectChangeCalled = true;

}

Spring Hibernate Integration HOW-TO using MySQL

2010-02-24T16:02:00.006-05:00

This blog assumes you are using Eclipse 3.5 and already know how to setup and use a basic Maven Project. If not check out my blog on creating a Maven, Spring, JUnit, AspectJ Project. I will be using Spring 3.0.1 and Hibernate 3.3.2.

This blog also assumes you have MySQL up and running and know the basics of how to create a database, create a table, populate a table using SQL, have permissions to the database and table, and know how to drop a table. If not you will first need to read the MySQL documentation from Oracle.

This blog is meant to walk someone through setting up a basic "HelloWorld" type Spring Hibernate integration project. Little explanation as to the "why," and the "what" will be given. Rather this blog answers the question "how." After gaining a basic understanding of the concepts of Spring and Hibernate, sometimes it's especially helpful to just see "how" without all the computer science lingo.

In this blog I am going to walk through the bare-bones steps needed to create a basic Spring Hibernate integrated project. This demo will retrieve, store, and delete animals with their associated attributes from a MySQL database. Hopefully this blog will help someone understand how to actually configure a basic Spring Hibernate project.

STEP 1: Create A Database Table

Login to MySQL
Create a new database called test, if you don't already have this database setup
Create a new table in your newly created database called "Animal" using the provided SQL

create table Animal (id int,name varchar(25), says varchar(10), feet int(3), index(id));

STEP 2: Populate Database

At the MySQL Prompt execute the following commands:

INSERT INTO Animal (id, name, says, feet) VALUES (1, 'cow', 'moo', 4);
INSERT INTO Animal (id, name, says, feet) VALUES (2, 'chicken', 'cluck', 2);
INSERT INTO Animal (id, name, says, feet) VALUES (3, 'horse', 'nay', 4);

STEP 3: Setup Your Working Project

Create a new maven project
Add the following dependancies with the specified versions numbers to your POM.xml:

log4J : log4j 1.2.15
org.springframework : spring-context : [3,)
org.springframework : spring-jdbc : [3,)
org.springframework : spring-orm : 3.0.1.RELEASE
org.hibernate : hibernate-core : 3.3.2.GA
javax.persistence : ebb : 3.0-public_review
mysql : mysql-connector-java : [5,)
org.slf4j : slf4j-simple : 1.5.8
javassist : javassist : 3.8.0.GA

Save your Project
Right click on the POM and select Run As Maven Build

If prompted type "compile" as your goal and then click Apply and Run.

You may get an error from the javax.persistence dependency. This is because the Maven repository does not contain the javax.persistence JAR file. So you will need to find this JAR yourself and place it in the correct directory yourself. Here is how to dot this:

Close Eclipse
Download the file at http://prdownloads.sourceforge.net/hibernate/hibernate-annotations-3.1beta5.tar.gz?download
Untar the file
Find the ejb3-persistence.jar
Rename to ejb-3.0-public_review.jar
Copy this renamed jar to ~/m2/repository/javax/persistence/ejb/3.0-public_review/

Open Eclipse and perform steps 3 & 4 again. The Maven build error should now be gone.

STEP 4: Create the Projects Basic Structure

Right click on the folder "src/main/java"
Click New->Package
Name your package org.example
Click Finish
Right click on the folder named "src/main/resources"
Click New->Folder
Name the folder "beans"
Click Finish
Right click on the folder named "src/main/resources"
Click New->Folder
Name the folder "hibernate"
Click Finish

STEP 5: Create POJO Representation of Database Table

Right Click on the package you named "org.example"
Click New->Class
Name the class the same as your table, in this case, "Animal."
Write your Animal class to correspond to your database. It should look like this:

package org.example;

public class Animal {

private int id;

private String name;

private String says;

private int feet;

public int getId() {

return id;

}

public void setId(int id) {

this.id = id;

}

public String getName() {

return name;

}

public void setName(String name) {

this.name = name;

}

public String getSays() {

return says;

}

public void setSays(String says) {

this.says = says;

}

public int getFeet() {

return feet;

}

public void setFeet(int feet) {

this.feet = feet;

}

STEP 6: Create Hibernate Mapping Between POJO and Database Table

Right click on the folder you created called "hibernate"
Click New->Other…
Expand the Hibernate DropDown
Select Hibernate XML Mapping file (hbm.xml)
Name the File Animal.hbm.xml
Click Next
Click Browse
Type Animal
Select Animal in the "org.example" Package
Click Ok
Click Finish
Open the newly created Animal.hbm.xml file
Map the Table Columns to the Animal Object Attributes by making the Animal.hbm.xml file look like the following:

xml version="1.0"?>

DOCTYPE hibernate-mapping PUBLIC

"-//Hibernate/Hibernate Mapping DTD 3.0//EN"

"http://hibernate.sourceforge.net/hibernate-mapping-3.0.dtd">

<hibernate-mapping package="org.example">

class>

hibernate-mapping>

STEP 7: Create Repository Interface

The Repository interface is where you should write all of your code that deals with connecting to the database and returning the data in the form of objects back to the main business logic of your application

Right click on the "org.example" package
Click New->Interface
Name the interface "AnimalRepository"
Click Finish
Add all the mandatory methods to your repository that your business logic will want to call. Here is the interface code:

package org.example;

import java.util.List;

public interface AnimalRepository {

public Animal getAnimalById(int id);

public List getAnimalsNamed(String name);

public List getAnimalsWithFeet(int feet);

public void saveAnimal(Animal animal);

public void killAnimal(Animal animal);

}

STEP 8: Implement the AnimalRepository Interface

Right click on the "org.example" package
Click New->Class
Name the Class "AnimalBarn"
Click Add… (Next to Interfaces)
Choose "AnimalRepository"
Click Ok
Click Finish
Code the AnimalBarn Implementation. Here is what it looks like:

package org.example;

import java.util.List;

import org.hibernate.Session;

import org.hibernate.SessionFactory;

import org.springframework.stereotype.Repository;

import org.springframework.transaction.annotation.Transactional;

@Repository

public class AnimalBarn implements AnimalRepository {

private SessionFactory sessionFactory;

public AnimalBarn(SessionFactory sessionFactory) {

this.sessionFactory = sessionFactory;

}

public void setSessionFactory(SessionFactory sessionFactory) {

this.sessionFactory = sessionFactory;

}

@Transactional(readOnly=true)

public Animal getAnimalById(int id) {

Session session = sessionFactory.getCurrentSession();

return (Animal) session.get(Animal.class, id);

}

@Transactional(readOnly=true)

public List getAnimalsNamed(String name) {

Session session = sessionFactory.getCurrentSession();

String hql = "from Animal where name='" + name +"'";

return session.createQuery(hql).list();

}

@Transactional(readOnly=false)

public void saveAnimal(Animal animal) {

Session session = sessionFactory.getCurrentSession();

session.saveOrUpdate(animal);

}

@Transactional(readOnly=false)

public void killAnimal(Animal animal) {

Session session = sessionFactory.getCurrentSession();

session.delete(animal);

}

@Transactional(readOnly=true)

public List getAnimalsWithFeet(int feet) {

Session session = sessionFactory.getCurrentSession();

String hql = "from Animal where feet=" + feet;

return session.createQuery(hql).list();

}

STEP 9: Create Spring Configuration File

Right click on the beans folder
Click New->Spring Bean Configuration File
Name the file application-config.xml
Click Next
Select beans - http://www.springframework.org/shema/beans
Select tx - http://www.springframework.org/schema/tx
Click Finish
Setup beans in XML. Here is what the completed file looks like:

xml version="1.0" encoding="UTF-8"?>

<beans xmlns="http://www.springframework.org/schema/beans"

xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

xmlns:tx="http://www.springframework.org/schema/tx"

xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-3.0.xsd

http://www.springframework.org/schema/tx http://www.springframework.org/schema/tx/spring-tx-3.0.xsd">

<constructor-arg ref="sessionFactory"/>

bean>

<list>

<value>resources/hibernate/Animal.hbm.xmlvalue>

list>

property>

bean>

bean>

bean>

<tx:annotation-driven transaction-manager="transacationManager"/>

beans>

STEP 10: Implement the Business Logic

Right click on the "org.example" package
Click New->Class
Name the class "BarnYard"
Select the check box asking you to create a "public static void main(String[] args) method stub
Click Finish
Code the main method. Here is the code I used:

package org.example;

import java.util.List;

import org.apache.log4j.BasicConfigurator;

import org.springframework.context.ApplicationContext;

import org.springframework.context.support.ClassPathXmlApplicationContext;

public class BarnYard {

/**

* @param args

public static void main(String[] args) {

//Log4J Logger so you can see all that is going on in the frameworks...

//BasicConfigurator.configure();

//Get the Spring Bean AnimalRepository interface which is implemented by animalBarn

ApplicationContext context = new ClassPathXmlApplicationContext("resources/beans/application-config.xml");

AnimalRepository animalRepository = (AnimalRepository)context.getBean("animalRepository");

//Showing how to get an Object by it's primary key.

Animal animal = animalRepository.getAnimalById(1);

System.out.println("The " + animal.getName() + " says " + animal.getSays() + " and has " + animal.getFeet() + " feet.");

//Showing how to query for an object based on an attribute

int feet = 4;

List animals = (List) animalRepository.getAnimalsWithFeet(feet);

int i = 0;

System.out.println("The Following Animal have " + feet + " feet:");

for (Animal anotherAnimal: animals) {

System.out.println(++i + ". " + anotherAnimal.getName());

}

//Showing how to query for an object based on an attribute

feet = 2;

animals = (List) animalRepository.getAnimalsWithFeet(feet);

System.out.println("The Following Animal have " + feet + " feet:");

i = 0;

for (Animal anotherAnimal: animals) {

System.out.println(++i +". " + anotherAnimal.getName());

}

//Showing how to save an Object to the Database...

Animal newAnimal = new Animal();

newAnimal.setId(5);

newAnimal.setName("piggy");

newAnimal.setSays("oink");

newAnimal.setFeet(4);

animalRepository.saveAnimal(newAnimal);

Animal pig = animalRepository.getAnimalById(5);

System.out.println("The " + pig.getName() + " says " + pig.getSays() + " and has " + pig.getFeet() + " feet.");

//Showing that updating an Object persist to the database...

newAnimal.setName("pig");

animalRepository.saveAnimal(newAnimal);

pig = animalRepository.getAnimalById(5);

System.out.println("The " + pig.getName() + " says " + pig.getSays() + " and has " + pig.getFeet() + " feet.");

//Showing how to delete an Object from the Database

animalRepository.killAnimal(newAnimal);

}

STEP 11: Setup Run Configuration

Right Click on BarnYard.java
Click on Run As…->Run Configurations...
Right Click on Java Application
Click New
Name the Project BarnYard
Click on the Classpath tab
Click on User Enties
Click Advanced…
Click Add Folders
Click Ok
Select src main
Select Ok
Click Apply

STEP 12: Run Application

Click Run

If everything complied correctly you should see the following (To turn on all the log messages uncomment the line BarnYard which calls BasicConfigurator.configure();):

log4j:WARN No appenders could be found for logger (org.springframework.context.support.ClassPathXmlApplicationContext).

log4j:WARN Please initialize the log4j system properly.

8 [main] INFO org.hibernate.cfg.Environment - Hibernate 3.3.2.GA

9 [main] INFO org.hibernate.cfg.Environment - hibernate.properties not found

11 [main] INFO org.hibernate.cfg.Environment - Bytecode provider name : javassist

14 [main] INFO org.hibernate.cfg.Environment - using JDK 1.4 java.sql.Timestamp handling

145 [main] INFO org.hibernate.cfg.HbmBinder - Mapping class: org.example.Animal -> Animal

200 [main] INFO org.hibernate.connection.ConnectionProviderFactory - Initializing connection provider: org.springframework.orm.hibernate3.LocalDataSourceConnectionProvider

404 [main] INFO org.hibernate.cfg.SettingsFactory - RDBMS: MySQL, version: 5.1.39

404 [main] INFO org.hibernate.cfg.SettingsFactory - JDBC driver: MySQL-AB JDBC Driver, version: mysql-connector-java-5.1.12 ( Revision: ${bzr.revision-id} )

420 [main] INFO org.hibernate.dialect.Dialect - Using dialect: org.hibernate.dialect.MySQLDialect

423 [main] INFO org.hibernate.transaction.TransactionFactoryFactory - Transaction strategy: org.springframework.orm.hibernate3.SpringTransactionFactory

424 [main] INFO org.hibernate.transaction.TransactionManagerLookupFactory - No TransactionManagerLookup configured (in JTA environment, use of read-write or transactional second-level cache is not recommended)

424 [main] INFO org.hibernate.cfg.SettingsFactory - Automatic flush during beforeCompletion(): disabled

424 [main] INFO org.hibernate.cfg.SettingsFactory - Automatic session close at end of transaction: disabled

424 [main] INFO org.hibernate.cfg.SettingsFactory - JDBC batch size: 15

424 [main] INFO org.hibernate.cfg.SettingsFactory - JDBC batch updates for versioned data: disabled

425 [main] INFO org.hibernate.cfg.SettingsFactory - Scrollable result sets: enabled

425 [main] INFO org.hibernate.cfg.SettingsFactory - JDBC3 getGeneratedKeys(): enabled

425 [main] INFO org.hibernate.cfg.SettingsFactory - Connection release mode: auto

425 [main] INFO org.hibernate.cfg.SettingsFactory - Maximum outer join fetch depth: 2

425 [main] INFO org.hibernate.cfg.SettingsFactory - Default batch fetch size: 1

425 [main] INFO org.hibernate.cfg.SettingsFactory - Generate SQL with comments: disabled

426 [main] INFO org.hibernate.cfg.SettingsFactory - Order SQL updates by primary key: disabled

426 [main] INFO org.hibernate.cfg.SettingsFactory - Order SQL inserts for batching: disabled

426 [main] INFO org.hibernate.cfg.SettingsFactory - Query translator: org.hibernate.hql.ast.ASTQueryTranslatorFactory

427 [main] INFO org.hibernate.hql.ast.ASTQueryTranslatorFactory - Using ASTQueryTranslatorFactory

427 [main] INFO org.hibernate.cfg.SettingsFactory - Query language substitutions: {}

427 [main] INFO org.hibernate.cfg.SettingsFactory - JPA-QL strict compliance: disabled

427 [main] INFO org.hibernate.cfg.SettingsFactory - Second-level cache: enabled

427 [main] INFO org.hibernate.cfg.SettingsFactory - Query cache: disabled

427 [main] INFO org.hibernate.cfg.SettingsFactory - Cache region factory : org.hibernate.cache.impl.NoCachingRegionFactory

427 [main] INFO org.hibernate.cfg.SettingsFactory - Optimize cache for minimal puts: disabled

427 [main] INFO org.hibernate.cfg.SettingsFactory - Structured second-level cache entries: disabled

430 [main] INFO org.hibernate.cfg.SettingsFactory - Statistics: disabled

430 [main] INFO org.hibernate.cfg.SettingsFactory - Deleted entity synthetic identifier rollback: disabled

430 [main] INFO org.hibernate.cfg.SettingsFactory - Default entity-mode: pojo

430 [main] INFO org.hibernate.cfg.SettingsFactory - Named query checking : enabled

454 [main] INFO org.hibernate.impl.SessionFactoryImpl - building session factory

585 [main] INFO org.hibernate.impl.SessionFactoryObjectFactory - Not binding factory to JNDI, no JNDI name configured

The cow says moow and has 4 feet.

The Following Animal have 4 feet:

1. cow

2. horse

The Following Animal have 2 feet:

1. chicken

The piggy says oink and has 4 feet.

The pig says oink and has 4 feet

All Done!

Using JDBC the Spring 3 Way

2010-02-18T19:32:00.002-05:00

Normally when you go to query a database there are lots of boilerplate things you need to do to make a connection, query the database, process the result set, and finally close the connection. Well Spring makes things easy. I'm going to show you how to do all these things in Spring. (NOTE: I'm using MySQL as my database)

Step one add the Spring Jars to your application (NOTE: I'm using the Maven group Id's and artifact Id's, if you are not using Maven you will download the associated jars and put them in your classpath ):

mysql : mysql-connector-java
org.springframework : spring-jdbc
org.springframework : spring-context

Second Setup your application-config.xml:

xml version="1.0" encoding="UTF-8"?>

<beans xmlns="http://www.springframework.org/schema/beans"

xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"

xmlns:jdbc="http://www.springframework.org/schema/jdbc"

xsi:schemaLocation="http://www.springframework.org/schema/beans http://www.springframework.org/schema/beans/spring-beans-3.0.xsd

http://www.springframework.org/schema/jdbc http://www.springframework.org/schema/jdbc/spring-jdbc-3.0.xsd">

bean>

<constructor-arg ref="mysqlDataSource"/>

bean>

beans>

Notice I have setup 2 beans mysqlDataSource and jdbcTemplate, respectively. I have passed the mysqlDataSource to the jdbcTemplate as a constructor-arg. Also note that in mysqlDataSource you will need to change the username, password, and URL to point to your database with the correct login credentials.

Next I create a table in the database:

CREATE TABLE animal (name VARCHAR(255), says VARCHAR (10), feet int(3));

Create a pojo that will hold a row of the table. In my case since I have table called animal, each row of the table corresponds to a particular animal. Therefore, I create a java class called animal that looks like this:

package org.example;

public class Animal {

private String name;

private String says;

private int feet;

public String getName() {

return name;

}

public void setName(String name) {

this.name = name;

}

public String getSays() {

return says;

}

public void setSays(String says) {

this.says = says;

}

public int getFeet() {

return feet;

}

public void setFeet(int feet) {

this.feet = feet;

}

Next write a class that will implement springs RowMapper interface: This is import because this is the class that will map a single result set to an Animal object. This is also where Spring magic starts to happen so pay attention to what is going on.

package org.example;

import java.sql.ResultSet;

import java.sql.SQLException;

import org.springframework.jdbc.core.RowMapper;

public class AnimalMapper implements RowMapper {

public Animal mapRow(ResultSet rs, int i) throws SQLException {

Animal animal = new Animal();

animal.setName(rs.getString("name"));

animal.setSays(rs.getString("says"));

animal.setFeet(rs.getInt("feet"));

return animal;

}

Now I write my calling code: Notice I will get the SimpleJdbcTemplate bean from the application-config.xml which I setup at the top of this blog. I will then use this bean created by Spring to call "update" to setup the database. Finally I will retrieve the data I place in the database using the "query" method. Notice how the query method takes as an argument a SQL String, and a RowMapper Object. It will then return a List of whatever types of objects your RowMapper was created to return. Notice how I never called "next" on the resultSet or had a while loop to iterate over the result set to create the list. I also never closed the connection, because, again, Spring does this all for me!

package org.example;

import java.util.List;

import org.springframework.context.ApplicationContext;
import org.springframework.context.support.ClassPathXmlApplicationContext;
import org.springframework.jdbc.core.simple.SimpleJdbcTemplate;

public class HelloAnimals {

public static void main(String[] args) {
BasicConfigurator.configure();
ApplicationContext context = new ClassPathXmlApplicationContext("./application-config.xml");
SimpleJdbcTemplate jdbcTemplate = (SimpleJdbcTemplate) context.getBean("jdbcTemplate");
jdbcTemplate.update("DELETE FROM animal;");
jdbcTemplate.update("INSERT INTO animal (name, says, feet) VALUES ('cow', 'moow', 4);");
jdbcTemplate.update("INSERT INTO animal (name, says, feet) VALUES ('chicken', 'cluck', 2);");
jdbcTemplate.update("INSERT INTO animal (name, says, feet) VALUES ('dog', 'woof', 4);");
List animals = jdbcTemplate.query("SELECT * FROM animal",new AnimalMapper());

for (Animal animal : animals) {
System.out.println("Animals name is " + animal.getName() +
" and says " + animal.getSays() +
" and has " + animal.getFeet() + " feet");

}
}
}

My Results:

Animals name is cow and says moow and has 4 feet
Animals name is chicken and says cluck and has 2 feet
Animals name is dog and says woof and has 4 feet

Mona Lisa Evolution Part II

2010-02-17T20:48:00.001-05:00

Earlier I posted a short article on evolving the "Mona Lisa" using genetic algorithms. Today I stumbled on another blog that refers does the same thing. The results are simply amazing! Check out the pictures on this blog.

Setter Injection & Custom Property Editors

2010-02-17T12:56:00.003-05:00

Below is an example of a typical use of setter injection. As you can see, some property of a java class is going to be set to the string "Hello World."

bean>

Out of the box Spring supports automagically setting java class properties to the following types besides just String:

1. Number
2. Boolean
3. Date
4. Resource
5. Properties
6. Byte
7. Locale

But sometimes you may want to use setter injection to set a property of a java class to a non-supported property type, like "java.util.regex.Pattern." Consider this class for example:

package org.example;

import java.util.regex.Pattern;

public class HelloPattern implements Hello {

Pattern pattern;

public Pattern getPattern() {

return pattern;

}

public void setPattern(Pattern pattern) {

this.pattern = pattern;

}

public void printMeAPattern() {

System.out.println("I got a pattern like this: " + pattern.toString());

}

So how would you use setter injection in spring to set "pattern"?

Well Spring allows you to extend something called the "PropertyEditorSupport" class and override two of it's methods; setAsText(String) and getAsText(). Once you extend this class, you will use your newly created class to support the setting of properties. Here is how it works.

First extend PropertyEditorSupport. Here is mine:

package org.example;

import java.beans.PropertyEditorSupport;

import java.util.regex.Pattern;

public class PatternEditorSupport extends PropertyEditorSupport {

public void setAsText(String regex) {

if (regex != null) {

setValue(Pattern.compile(regex));

} else {

setValue(null);

}

public String getAsText() {

String rtnString = "";

Pattern pattern = (Pattern)getValue();

if (pattern != null) {

rtnString = pattern.pattern();

}

return rtnString;

}

Notice a few of things.

1. You must override the setAsText(String) and the getAsText() methods as I stated above.
2. You must call setValue(String) in the setAsText(String) method
3. You must call getValue() and cast it to the correct type in the getAsText() method.
4. You must return an empty value from getAsText() in case of null.
5. You must send null to setValue(String) if for some reason the String is null.

After having written your PropertyEditorSupport extended class you will then add the following to your bean xml (usually application-config.xml):

<map>

map>

property>

bean>

Finally create your bean in your xml file just like normal and the Spring framework will automagically call the correct CustomEditorConfigurer for the property in question. In this case I'm going to set a Pattern for a zip code...

<property name="pattern" value="[0-9]{5}">property>

bean>

Spring knows to call the correct CustomEditorConfigurer because Spring knows that the property "pattern" in set in the class org.example.HelloPattern with a setter requiring type java.util.regex.Pattern to be passed in. So it will look in the CustomEditors map and see if there is a Custom editor to support that type. Sure enough, in this case there is. So Spring calls the PatternEditorSupport class to handle the setAsText(String) and getAsText() functions. Our property "pattern" has now been properly intialized to a Pattern.

Important to remember: In Spring, everything starts as a String (because it's XML) and the Spring framework tries to convert this starting String to something else before calling the set method on the property; albeit a number, a date, a custom something, etc. If none of those conversions work Spring will only then pass it as a String to the property.

Using properties files in Spring

2010-02-16T17:48:00.001-05:00

When one wants to set a value in the Spring Framework using "Setter Injection" typically the application-context.xml has the following type of XML in it:

bean>

The problem with this is that if you want to change the "value" then you have to wade through an xml file to find the value you want to change. This could easily cause a bug such as a missing " or an accidently deleted > or /. Not to mention it is just hard to read a bunch of XML when you are just looking to set a single value. Any little typo in the XML that will cause a runtime exception that may be difficult to find and fix. A way to keep the "values" out the xml file is to use a traditional properties file and refer to it from the application-context.xml file. Here is what this solution would look like:

bean>

<context:property-placeholder location="./resources/values.properties"/>

And the values.properties file would read like this:

values.question=Hello World using properties!

That's it.

JUnit 4 in Eclispe 3.5

2010-02-12T20:39:00.002-05:00

Here is a simple example on how to use JUnit 4 inside of Eclipse 3.5. First to setup an Eclipse 3.5 project to use JUnit you can follow my blog on how to setup an Eclipse project using Maven, JUnit, Spring, etc...

Step 1 Create a Class to Test:

Create a new package by right clicking on the src/main/java folder. I called my package org.example.testing
Create a new Java Class inside of your package by right clicking and selecting class. I called my class "Cow."

Here is my "Cow" class:

public class Cow {

private String says;

private int feet;

private int stomachs;

public Cow () {

this ("moow", 4, 4);

}

public Cow (String says, int feet, int stomachs) {

this.says = says;

this.feet = feet;

this.stomachs = stomachs;

}

public String getSays() {

return says;

}

public void setSays(String says) {

this.says = says;

}

public int getFeet() {

return feet;

}

public void setFeet(int feet) {

this.feet = feet;

}

public int getStomachs() {

return stomachs;

}

public void setStomachs(int stomachs) {

this.stomachs = stomachs;

}

public void throwSomeExceptions () {

int test = 5/0;

}

Step 2 Create a Test Case

Create create the same package you created above in your src/test/java folder.
Create a JUnit Case, by right clicking on your newly created package and then clicking Other... and finding the JUnit Case.

You will be given a dialog box that looks like this:

Be sure "New JUnit 4 test" is selected.

Give your test case a name. I called mine CowTest
Select setUp()
For the "Class under test:" text box, select "Browse..." and Select the Cow class you created earlier.

You may be given a warning at the bottom that says, "JUnit 4 requires a Java 5 project." If so click the Configure link and change the compliance level to 1.5 or higher.

Click Next and select the Cow methods to test. In this case I choose to test the following methods:

getSays();
getFeet();
getStomachs();
throwSomeException();

Click Finish.

Fill in the functions in your test case until it looks like this:

public class CowTest {

private Cow cow;

@Before

public void setUp() throws Exception {

cow = new Cow ("moow", 4, 3);

}

@Test

public void testGetSays() {

assertEquals (cow.getSays(), "moow");

}

@Test

public void testGetFeet() {

assertEquals (cow.getFeet(), 4);

}

@Test

public void testGetStomachs() {

assertEquals (cow.getStomachs(), 4);

}

@Test (expected = Exception.class)

public void testThrowSomeExceptions() {

cow.throwSomeExceptions();

}

Step 3 Run the JUnit Test

To run the JUnit test right click on your test case and select Run as JUnit test

The test should now run and you should see get this GUI displayed inside of Eclipse telling you which test ran, passed, and failed (In this case 4 Test where run, 3 Passed, 1 Failed):

Go back to your JUnit Test case and change your setUp() code to:

public void setUp() throws Exception {

cow = new Cow ("moow", 4, 4);

}

Rerun the JUnit test, just as you did before and you should see this GUI displayed:

Congratulations you have just successfully used JUnit 4 in Eclipse 3.5! As you add more test cases to the package, you can run them all at once by right clicking on the Package and selecting Run As JUnit test. It will now run all of the test cases in the package.

Evolutionary Computing with Java

2010-02-12T16:54:00.001-05:00

One the most interesting areas in computer science is the area of Evolutionary Computing. I found this open source Java framework called the "Watchmaker Framework." This framework is a group of algorithms based on biological evolution. The authors show an example of their framework in action by evolving the Mona Lisa from a maximum of 50 polygons over a series of generations. But like evolution in the biological world, running the application may take a while! ;-) However, it's very impressive even after a couple of minutes of running to say the least!

Here are my results after about 1.5 Minutes and a population size of 5

Check out the watchmaker framework here.

Syntax Highlighting in Web Pages

2010-02-12T13:13:00.004-05:00

Since my blogs are mostly about technology and software engineering, I wanted to have some method of putting code in my blogs and it to have good syntax highlighting.

Without syntax highlighting my blogs look like this:

public void bland () {
System.out.println("This stinks!");
}

with Syntax highlighting turned on though they look like this:

I found a number of solutions on the web, but this blog had the simplest and best method I found.

Follow the direction on the linked blog and your done. Source code is actually written by SyntaxHighlighter.

Simpler though is to simply to cut and paste cut directly from Eclipse to the Blogger New Post Compose page. No css or HTML to worry about at all! Here is what text pasted from Eclipse looks like:

package org.example.testing;

public class Cow {

private String says;

private int feet;

private int stomachs;

public Cow () {

this ("moow", 4, 4);

}

public Cow (String says, int feet, int stomachs) {

this.says = says;

this.feet = feet;

this.stomachs = stomachs;

}

public String getSays() {

return says;

}

public void setSays(String says) {

this.says = says;

}

public int getFeet() {

return feet;

}

public void setFeet(int feet) {

this.feet = feet;

}

public int getStomachs() {

return stomachs;

}

public void setStomachs(int stomachs) {

this.stomachs = stomachs;

}

public void throwSomeExceptions () {

int test = 5/0;

}

Mac Blog Software

2010-02-11T22:06:00.003-05:00

I have been looking for good Blogging Software for the Mac. After much searching I have settled on Blogo.

I first tried MarsEdit.

Pro's: I liked the simplicity, liked how to add Categories to the blog post.

Con's: To simple. Pictures are put in as img links. Not as images. It also didn't automatically resize my images to fit with in the contours of my Blogger Blog. It was also not intuitive as to how to make a thumbnail image a link to the full image.

Second I tried Ecto

Pro's: Images show up as images. More of a WYSWYG editor than MarsEdit.

Con's: Couldn't figure out how to upload photos to picasoweb.

Third I tried MacJournal

Pro's: Color and lots of features
Con's: Cryptic, just couldn't do simple thing easily.

Finally I tried Blogo

Pro': Simple, Easy to use, automatically sized images to fit my blog, easy to link images to full size image. Overall I found this product to be the simplest and to provide the best integration with Blogger and a general pleasure to use so long as you are doing really simple stuff. The image functionality and the good rich text support is why I choose this software.

Con's: Picture is just a place holder in editor, adding Categories to blogs are awkward. Anything like code pasting or advance stuff is very difficult.

In conclusion while I'm using Blogo, I wouldn't say any of the products are will knock your socks off. There certainly is a market here for some one to write the Ultimate Blogging software.

How To Setup an Eclipse Project to use Maven, AspectJ, Log4J, Spring, and JUnit

2010-02-11T14:45:00.002-05:00

I see lots of blogs on how to do any one of these things, but I want to do them all in the same project. So how do I do this? First it's easy, but a little tricky so follow closely. This blog assumes you have setup your eclipse environment with all of these plugins already. If not check out my blog telling you how to configure the "Ideal Eclipse 3.5" (at least it's ideal to me) ;-)

(To see the full images just right click and say open Image in New Window)

1. Click File->New->Other...

2. Click Maven Project

3. Walk through setup wizard as shown in these screen shots. (There are plenty of blogs that can explain what each option means, but for now just follow along)

4. Click Finish and then Double click on the file called pom.xml Your screen should look like this.

5. Click on the Dependencies tab in the lower left hand side of the middle window

6. Click Add...

7. Then click the little folder with the Jar in it over on the upper right side.

8. Type in Junit and select it from the list

9. Continue clicking Add and then the folder with a Jar icon until you have added Junit, Log4J, spring-context, and aspectjrt. Your Dependency list should look like this.

10. Click on the tab labeled Reporting

11. Like you did for dependencies, click add, and then the icon of the folder with a jar until you have all the following plugins added: surefire-report-maven-plugin, maven-javadoc-plugin, jdepend-maven-plugin, dashboard-maven-plugin. When completed it should look like this:

12. For fun click the tab labeled Dependency Graph. It should now look like this:

13. Now Right click on your project folder and click Configure->Convert to AspectJ Project (Don't worry you won't lose your Maven Project Settings)

14. You will notice now a new folder has been added in your Package Explorer called AspectJ Runtime Library

And that's it. You now have configured a New Project which can use Maven, AspectJ, JUnit, Log4J, and Spring. Just be sure to keep your JUnit test in your src/test/java folder and you should be good to go. Hope this was helpful.

Installing Hibernate Tools in Eclipse 3.5

2010-02-11T12:05:00.001-05:00

How do I get Hibernate Tools to work in Eclipse 3.5?"
(Personally I use Eclipse 3.5 for Cocoa 32 Bit Addition)

Well here are the steps (assuming you already have Eclipse 3.5 installed)

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "Hibernate" (without the quotes)
Type in this location: http://download.jboss.org/jbosstools/updates/stable/ (This is Hibernate Update site)
Expand the JBoss Tools Box
Click Hibernate Tools
Click Finish and follow the directions accepting the license agreements

Ideal Eclipse 3.5

2010-02-11T11:27:00.002-05:00

I'm a Software Manager and I have a team of developers who mostly code in Java and Flex but occasionally in C++. Occasionally new people join the team and they need to setup their work environment so that they have all the same tools as everyone else. We had been keeping a document that told step by step how to setup their Eclipse IDE. But as things changed, new versions of plugins get released, Eclipse changed, or we discover a new useful plugin, the document has to be continually revised and someone has to take the time to update it. Then it occurred to me one day why not just create an Eclipse distribution that is already setup with all the plugins and configurations we use? Anyone else doing this? Well here is how I created our own personal Eclipse Distribution. Any ideas for making this better would be great!

I noticed that when you download Eclipse the download is really just one big folder zipped up. This gave me the idea that if I download the Eclipse we as a team want to use and simply start installing the plugins we want, when we are done we should be able to simply zip the Eclipse folder back up and put it out on our server. When we decide to use a new version, instead of updating documentation, we just setup one new Eclipse the way we want and put it on the server and let everyone download it. No reading, no tedious writing, no downloading and no configuring!

So here is how I have setup our Eclipse Distribution:

Download Eclipse IDE for Java EE Developers 3.5 Cocoa 32Bit Version from the Eclipse.org website and unzip/untar as normal
Click Help->Check For Updates and get all the latest
Click Help->Install New Software...
Select the Galileo Update Site from the Drop Down Menu labeled "Work With"
Select Collaboration (Why not? Might be useful one day) ;-)
Expand Programming Languages and select Eclipse C/C++ Development Tools
Click Next and Finish
Click Yes to restart Eclipse when prompted.
Click SVN Kit 1.3.0
Click Finish
Click Next when prompted ensuring both boxes
Click Next
Accept the license agreement
Click Finish
Click Ok when you get a Security Warning.
Click Yes to restart
Click Help->Install New Software...
Click on the Eclipse Web Tools Platform from the Drop Down Menu labeled "Work With"
Expand the Web Tools Platform SDK
Select Eclipse XML Editors and Tools SDK
Select JavaScript Developer Tools SDK
Select XSL Tools SDK
Select Web Page Editor
Click Next and then Finish and Yes to restart when prompted just as before
Select the Mylyn Update Site from the Drop Down Menu labeled "Work With"
Select All boxes.
Click Next and then Finish and Yes to restart when prompted just as before (It will tell you somethings have been modified...that's fine...just keep going)
Follow my Blog on installing AspectJ in Eclipse
Follow my Blog on installing Maven in Eclipse (In our case you can unselect the Subclipse package since we are using Subversive to talk to Subversion) (NOTE: This step is not necessary if you are going to install the Spring Tool Suite in the next step. Just skip this step)
Follow my Blog on installing Spring IDE in Eclipse
Follow my Blog on installing Hibernate tools in Eclipse
Now find your eclipse folder on your hard drive.
On the Mac you right click and then click Compress

Now you have your own Custom Eclipse IDE to distribute to your team.

Installing Spring in Eclipse 3.5

2010-02-11T11:07:00.002-05:00

How do I get Spring to work in Eclipse 3.5?" (Personally I use Eclipse 3.5 for Cocoa 32 Bit Addition) Well here are the steps (assuming you already have Eclipse 3.5 installed)

In order to install Spring and the Spring Tool Suite first you have to have Mylyn installed as well as Mylyn Extra's.

Install Mylyn

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "Mylyn" (without the quotes)
Type in this location: http://download.eclipse.org/tools/mylyn/update/e3.4/ (This is Mylyn Update site)
Select all the boxes.
Click Finish and follow the directions accepting the license agreements

Install Mylyn Extra's

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "Mylyn Extra's" (without the quotes)
Type in this location: http://download.eclipse.org/tools/mylyn/update/extras (This is Mylyn Extra's Update site)
Select all the boxes.
Click Finish and follow the directions accepting the license agreements

Install Spring

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "Spring" (without the quotes)
Type in this location: http://www.springsource.org/update/e3.4 (This is Spring Update site)
Select all the boxes.
Click Finish and follow the directions accepting the license agreements

Installing m2Eclipse (Maven) in Eclipse

2010-02-11T10:44:00.002-05:00

How do I get Maven to work in Eclipse 3.5?"
(Personally I use Eclipse 3.5 for Cocoa 32 Bit Addition)

(NOTE: If you are using the Spring Tool Suite then it already includes m2Eclipse so no need to follow these steps.)

Well here are the steps (assuming you already have Eclipse 3.5 installed)

m2Eclipse Installation

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "m2Eclipse (Maven)" (without the quotes)
Type in this location: http://m2eclipse.sonatype.org/sites/m2e (This is m2Eclispe Update site)
Select all the check boxes, if you have Subclipse, AspectJ, and Mylyn installed. Otherwise uncheck those particular boxes. If you just select everything it will tell you which ones won't be installed because you are missing a dependency. So don't worry about it if you don't know which ones I'm referring to. Just happily continue.
Click Finish and follow the directions accepting the license agreements

m2Eclipe Extras Installation

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "m2Eclipse (Maven)" (without the quotes)
Type in this location: http://m2eclipse.sonatype.org/sites/m2e-extras (This is m2Eclispe Extras Update site)
Select all the check boxes, if you have Subclipse, AspectJ, and Mylyn installed. Otherwise uncheck those particular boxes. If you just select everything it will tell you which ones won't be installed because you are missing a dependency. So don't worry about it if you don't know which ones I'm referring to. Just happily continue.
Click Finish and follow the directions accepting the license agreements

What is UTM? (Universal Transverse Mercator)

2010-02-10T16:29:00.001-05:00

(First I assume you know what Latitude and Longitude is if you are reading this.)

UTM was developed as a way of determining coordinates on a paper map or really any flat 2D surface (such as a computer screen!). Here is how it works in a nutshell:

UTM is written like this 18S 3741214233566 (Like all the numbers? ;-)) So here is how to read this "garble-de-goop" (That's a technical term!)

UTM divides the world up from 80 degrees South latitude to 84 degrees North Latitude.

These areas are labeled from bottom to top & left to right. Exactly how you read a typical X & Y graph. But instead of using just numbers UTM uses a mix of numbers and letters.

The Y axis starts with the letter F and goes through the Letter X

The X axis starts with the number 1 and goes through the number 60

Now there are good mathematical reasons for all of this. If you are interested you can read a book on the subject. A simple book on navigation will explain this well. I personally have enjoyed reading Outdoor Navigation With GPS: Hiking, Geocaching, Canoeing, Kayaking, Fishing, Outdoor Photography, Backpacking, Mountain Biking by Stephen Hinch
. He explains it in more detail in his book. But for our purposes I'm going to keep this really simple.

So if you have ever played the board game battleship you should understand that if I said 18 S that would refer to the grid area somewhere around Maryland, Virginia, and Delaware.

Now the next 13 digits are the "eastings and the northings." This may not always be 13 digits, but typically is. When there are an odd number of digits the easting digits are always one less than the northing digits. So in our example the easting measurement is 374121 (6 Digits) and the northing measurement is 4233566 (7 Digits).

Now you find the map for the 18 S (which can be found at usgs.gov) and you simply go over on the X Axis 374121 and on the Y axis up to 423566. Remember what I said about the Y Axis having letters and the X axis going from 0 to 60? Well now once you have found the right map, the letters and numbers I told you about are irrelevant. We now switch back to good old numbers on both the X and Y Axis. Just read them like you would any X & Y graph now.

In other words the first number and letter tell you on a global scale what general area map you need to find. Once you have found the corresponding map, you then basically just use the next string of numbers to find the point on the map. Simple right?

Click Here for more detailed explanation here along with pictures and sample maps from the US Government..

Create/Convert to an AspectJ Project in Eclipse 3.5

2010-02-10T15:46:00.001-05:00

To create an AspectJ project:

Click on File->New->Other...
Select AspectJ Project.

To add AspectJ code to an existing Java Project:

Right click on the project folder in the Package Explorer window
Click Configure->Convert to AspectJ Project.

Your project will now allow you to create Aspects and of course use the AspectJ weaver. Don't worry your current Java code should still compile and work fine. The only thing you have done is told Eclipse to use the AspectJ weaver when compiling this project.

That's it! Keep in mind that ApsectJ has it's own unique compiler separate from javac. This AspectJ compiler is called a "weaver" and this weaver is now installed on your computer and you are able to create AspectJ projects that will use this new weaver. Don't worry your java code should still work fine.

But if you have an Ant build script that is calling the "javac" compiler you will need to change it to call the "ajc" compiler or your aspects will not compile.

How to Install AspectJ in Eclipse 3.5

2010-02-10T15:45:00.003-05:00

If you read my last post you may be thinking, "Ok AOP sounds interesting but how do I get AspectJ to workin Eclipse 3.5?"
(Personally I use Eclipse 3.5 for Cocoa 32 Bit Addition)

Well here are the steps (assuming you already have Eclipse 3.5 installed)

In Eclipse, click on Help -> Install New Software...
Click Add...
Type in a name, I typed, "AspectJ" (without the quotes)
Type in this location: http://download.eclipse.org/tools/ajdt/35/update (This is AspectJ Update site for Eclipse 3.5 for 3.6 use http://download.eclipse.org/tools/ajdt/36/update)
Select all the check boxes
Click Finish and follow the directions accepting the license agreements

An Intro to Aspect Oriented Programming

2010-02-10T13:55:00.006-05:00

What in the world is Aspect Oriented Programming (AOP)? Well if you don't know your not alone. In fact I had no idea until about two weeks ago, and no one on my software team had a clue either. But I kept seeing the word "AOP" pop up on blogs so I decided to investigate.

Well I hate big words and long explanations so in a nutshell AOP is a programming methodology whereby the primary aspects (i.e. requirements!) are separated from the secondary aspects. So let me explain. Say you are writing ATM software. It's primary job/aspect/requirement is to move money from Account A into your hands. Any other requirements that are used to support that main requirement of moving money (such as logging, security, etc) are probably secondary aspects/jobs/requirements and can/should be written using aspects in an aspect oriented language such as AspectJ. Why you ask?

Well these secondary requirements, like logging, tend to "cross cut" many of your traditional objects and make for ugly hard to follow code example:

Requirements:

1. Move Money
2. Log all Changes in Account Balance

In this case "logging" is a secondary aspect/job/requirement of the code. So here is how AOP would solve this problem?

Notice that the moveMoney function now only deals with "moving money" while the logAccountChange "advice" (think of the term "advice" as a method inside a class) deals with the logging and only the logging.

That is what AOP is about in a nutshell; separating out primary aspects/requirements from secondary aspects/requirements. It is not a replacement to OOP it is complementary to it.

Using Pixel Bender

2009-03-20T13:47:00.001-04:00

I'm going to demonstrate how to write a short Pixel Bender script which will change the color of a graphic from Green to Yellow or Red.

The Pixel Bender application is a simple development tool that comes with Adobe Flash CS4. You can also download it from Adobe Labs if you don't have it.

(I'm using an Intel Macbook Pro 2.53GHz with 4GB of RAM to write this code, this will also work on a Microsoft Windows PC)

Part 1:

1. Start Pixel Bender. I use Spotlight to start all my applications. (Command + Spacebar on the Mac and start typing Pixel Bender ToolKit)

2. When Pixel Bender starts you will see the window below. Look in the lower right hand corner and click the "Create a new filter" button.

3. When you click the "Create a new filter" button, Pixel Bender will put the basic stub code in the editor pane (Lower Left hand side).

NOTE: Don't let the code intimidate you. This will be easy. Don't worry about what all the code means at the moment. We will get to it in later blogs. We are just going to cover the bare minimum here.

4. Now Load a sample image so that when we are done we can see the results of our code. To do this click on File > Load Image 1...

5. Find the Image you want to use to test the code. This image doesn't have to be the image you want to use in the final product. But In this case we need to find a green button image in order for the script to work as intended.

I created a simple green button in Adobe Illustrator and saved it as a jpeg (Pixel Bender can use either jpeg or png images to test the code). Here is the image I used (yes it's a little cheesy but it's mine, get's the job done, and it's free to use.)

5. Now look down at the source code. Note that there are two variables in the code, (src, and dst) and also note that there is one function (evaluatePixel()).

src is the variable that stores your "source" image (your original image). In this case src has already been set to our green button. Pixel Bender did this for you. You don't need to do anything to make this happen.
dst is the variable that stores your output image, your "destination" image.
The evaluatePixel() function is the function that is called by default when the script runs. (If you are familiar with C++ you can think of it as kind of like the "main" function).

6. In order to turn a green pixel to Yellow or Red pixel you have to understand that basically all pixels are made up of 3 color channels (Red, Green, and Blue).

A red pixel has the red channel turned up, and the green and blue channels turned down. Think of it like volume on a radio. To best hear the Left side speaker turn the left side speaker up and turn the right side speaker down. The stronger you want the sound out of the left speaker the more you can turn up the left speaker and turn down the right speaker.
Thus a green pixel has the Green channel turned up and the two other channels turned down.
A yellow pixel has the red and green channels equal and the blue channel down.
A white pixel has all the channels turned all the way up.
A grey pixel has all the channels equal but not turned fully up or fully down.
A black pixel has all the channels turned down.

So to make our pixel bender script turn our green image to a red image we are going to take the red channel of the dst variable and set it equal to the green channel of the src variable and we are going to set the dst variable green channel to the src variable red channel. To do this we are going to insert the following code inside the evaluatePixel() function below the line that says "dst = sampleNearest(src, outCoord());"

float greenChannel = dst.g;
float redChannel = dst.r;

dst.r = greenChannel;
dst.g = redChannel;

7, If you followed all these steps your Pixel Bender window should now look like this:

8. Almost done. Double check your code for typos. When your sure it matches mine then click the run button in the lower right hand side.

Your image should turn red. (Don't worry this is a non-destructive operation. You original image is not actually changed only what is displayed on the screen! Your screen should now look like this.)

9. Now lets add the code that allows us to choose whether we want to turn the image red or yellow. To do this we first have to create two "parameters." These parameters will let an outside program or person change the image color. We can't use a boolean type to do this because if we do it won't work later when we want to put it into Adobe's Flash CS4. So we need to create two integer parameters. To do this we add the following code above the evaluatePixel() function, but below the variable declarations. You can name the parameters whatever you want but I choose "alarm" and "warning" because ultimately I want to use this to show whether a computer on a network is working properly or not.

parameter int alarm
<
minValue : 0;
maxValue : 1;
defaultValue : 0;
description : "Change To Red";
>;

parameter int warning
<
minValue : 0;
maxValue : 1;
defaultValue : 0;
description : "Change to Yellow";
>;

10. Now we need to add the code to the evaluatePixel() function that will evaluate the pixels based on the parameters. If the alarm parameter is set to 1 (i.e. ON) then we want a red image. On the other hand if the warning parameter is set to 1 then we want the image to turn yellow. To do this we change the evaluatePixel() function to look like the following:

dst = sampleNearest(src,outCoord());
float greenChannel = dst.g;
float redChannel = dst.r;

if (alarm == 1)
{
dst.r = greenChannel;
dst.g = redChannel;
}
else if (warning == 1)
{
dst.r = greenChannel;
}

11. All Done! If you followed all these steps hit the Run button again and now your script should be all done and working.

Test your script by moving the sliders on the upper right hand side from 0 to 1. Be sure to save your pixel bender script. I called mine SystemState.pbk

(NOTE: If you set both your sliders to 1 the image will be Red. You can change that if you want. I personally choose not to)