Sunday, November 22, 2015

The first global day of coderetreat in Thessaloniki

Hi there,

feels like a Sunday afternoon, having had one night with friends from far away, that joined me for the weekend, in order to celebrate my graduation (ok, that's a separate post, so let's leave it for now).

Getting to know the event itself

Traveling back in time, I can remember myself reading a tweet from Patroklos about the 1st global coderetreat event in Thessaloniki. That time, I was in a rush preparing my thesis for October's presentations, so I just put it in my TODOs list. Unfortunately, the fore-mentioned "rush" was an overkill for the last 3 months (I was trying to combine my thesis implementation progress, together with my daily job - did I just say that I will leave the "studies" part out of this post?!), so at that time, my TODOs list maintained greek software management's attitude. Hopefully, some days after, I noticed that there were some cancellations, so, keeping in mind that I had never been in such an event, I'd surely go; I didn't read much about it, I just liked the concept as a whole, so this one would be a new experience (ok, having completed about a year on industry, almost everything that you get exposed on is new) for me.

Reaching the event's place

The event was planned for the 14th of November (Saturday?), so, keeping in my mind that it was a hard week for myself and I, I had to sacrifice Friday's outing for rest. Woke up on time, I reached coho, about 9:20, so I had some time to get my breakfast and get awake.


Patroklos and Nikos were the event organizers. The problem introduced was Conway's Game of Life, which had to be approached, regarding the event's concept, following the rules below:
  • For each session, you have to pair with a different partner.
  • At the end of each session the code has to be deleted (, so, for each new session, a completely new project/code template has to be created).
  • Each pair must only use 1 pc.
  • Each session lasts for 45 minutes; that is, the development process is only 45' minutes.

The Sessions

After the introduction of the game, the rules definition and the event's concept, we were said that there were going to be six different sessions, with the first and the last to have a free character, whereas in general, each session would have an increased difficulty level, compared to the previous one. The difficulty level generally depends on the session's restrictions set initially.

Session 1

A fair introduction to the problem from my partner and some trials to adjust the working environment according to the nearest IDE/language preferences that we had in common. Time seemed to be passing by very quickly on this session and that's why I believe it had no challenging character, just an introductory one to the event.

Session 2

Hmmm...restrictions started: "Solve the problem using TDD's ping-pong approach". One writes the test and the other only tries to get the tests to pass. Obviously, both of them have to refactor when appropriate.

I haven't had prior experience on TDD and I really liked this one; it helps so much for writing cleaner code. 

Session 3

Ok, seems like we 're getting into the meat. What's the next restriction? "Your methods must contain no more than 4 lines of code". Oooh, that one hurt. Or not? After explaining my approach to my partner, it seemed like we had an agreement for this one, but I was still wondering how we could fulfill the requirement. Hopefully, he knew Java 8 and our methods didn't exceed 2 lines.

At this point, let me state the following thought: I have also attended an Oracle's free course for Java 8 lambdas and streams, but this one displayed the difference of using them in production, too. So, I'll surely have to sometime get deeper into it, even for a sample project.

Session 4

"No try/catch, no if/else statements". By default, this one seemed enough challenging, but a big part of the session spent on environment setup and language definitions, as both of us didn't have enough experience with pure PHP. Having had a small amount of time to approach the problem following the restriction set, this one was the hardest one, I would say, as you were obliged to break your code into smaller and smaller functions.

That is, it seemed hard, as we got used to just solving the problem and obviously, the fore-mentioned phrase does not contain keywords like "effectively" or "cleanly".
Many of you, readers/devs, would instantly blame your team leaders or software management team and its roots and I wouldn't completely disagree, if the conversation was about project/delivery management or software architecture, but now, it's all about code; it's not a common phenomenon when a supervisor of yours will tell you how to actually write your code, especially here in Greece.

Nevertheless, we 're still in software and indeed, there is a workaround, for this one, too: for the next time you 're gonna get into some refactoring, if you notice something that would be better changed/written in another way, don't leave it after the upcoming delivery or for tomorrow. Don't try to persuade yourself that you don't currently have enough time for this. Breaking a method to smaller ones, creating a separate package for some common classes or providing some documentation won't last more than 5'.

And by this you 're not only saving time for yourself (that is, there are many times, that even returning to a code block of yours, written 1 month ago, the "wtf" counter starts incrementign surprisingly), but for the rest ones that will read your code in the future.

Imagine what would happen if you applied this simple practice in your daily job and if you also persuaded your colleagues to do so:

  • long-term time saving
  • readable and maintainable code
  • less future refactoring

So, from yesterday, you should start thinking about it, I should start thinking about it, we should start thinking about it, as serious software is a result of good teams and good teams are based not only to team leaders (Figure 1), but on good developers, too.
Figure 1. Boss or Leader?

Session 5

"No mouse". After hearing this restriction, we simply started laughing, but my initial thought was sth like "how on earth would you benefit of writing cleaner code, from this one?". Ok, it may not be one of the direct "clean code" practices, but it's worth following it while writing code (especially on an IDE), for productivity purposes.

I consider myself knowing enough keyboard shortcuts and I feel lucky for this one, as I had the chance to work closely with guys who liked using them for development. so initially, this one didn't seem as a threat, but when it came to switch to browser, things got difficult enough :).

Session 6

The last session was on the same line as the first one, keeping the character of a free session, but we were also encouraged to apply what whe learned during the day.


Ok, it was an exhausting day, in conjuction with an exhausting week, too, but in the end of the day, what only matters, are the feelings for it and my feelings according to the event were great. These feelings were also enchanced from the retrospectives that followed each individual session, where we discussed our approaches and how each one them could be better developed.

First off all, I had the chance to come back to social life (after spending the whole summer for the development of my thesis project, in parallel with my daily job), which means meeting new people who have same interests.

Second of all, I believe that all of us leveled up our experience on time management (depending always on the experience of each one), as the event itself pushes you to make the best out of your communication skills, in order to handle team decision making for each session. I'm also grateful that some of my pairs were js guys, as I lately see myself very keen on reading about JavaScript and front-end, in general.

The good thing is that I saw practices that I have never worked with, like TDD: I was really amazed of the way the ping-pong approach works. In addition, seems like Java 8 is here to reduce lines of code for common problems and increase productivity.

Finishing this part and the post itself, I noticed some technical weaknesses of mine, but I won't blame myself for this; instead I will keep them in mind, as points where more practice is required, 'cause I have a way to go as a developer.


Sunday, July 5, 2015

JAXB Unmarshalling Example

1. Introduction

1.1 Apologies


It's been a loooong time since my last post, so, first of all, let me apologize for this absence

The reasons:
  • The tiny OCAJP badge that you see in the right side of the blog (I will refer to this tin a forthcoming post).
  • I currently have limited time available.

1.2 Example's concept

What we 're gonna see today is how to convert an XML document to a Java Object. This (and the reverse - Java object to XML transformation) usually occurs when a software deals with Web Services.

2. The Example

We 'll here demonstrate how to convert an xml list of objects to a Java object. This can be easier using the JAXB Technology. JAXB stands for Java Architecture for XML Binding. It is used to convert XML to Java objects and vice-versa.

Environment used:
  • JDK 1.7
  • Eclipse Luna

I've the environment that I used, as from JDK versions 1.6 and later, the JAXB dependency is bundled into the JDK, in contrast with previous JDK versions, where you had to at least include the following dependencies onto your classpath: “jaxb-api.jar” and “jaxb-impl.jar”.

That is, we here don't have to include nothing at all.

Here is the project structure of this sample, a simple one as you can see, too:

2.1 The XML file

As I said, we 'll here deal with a smartphones list, so, just for demonstration purposes, two smartphones are here listed:

<?xml version="1.0" encoding="UTF-8"?>
  <model>Galaxy Ace</model>
  <model>Moto G 2014</model>

2.2 The Beans

Each smartphone can be described with a DTO class, so here it is:
package com.toubou91.jaxb.example;

import javax.xml.bind.annotation.XmlAccessType;
import javax.xml.bind.annotation.XmlAccessorType;
import javax.xml.bind.annotation.XmlRootElement;

@XmlRootElement(name = "smartphone")
public class Smartphone {

 private String make;
 private String model;
 private String androidVersion;
 // Getters and setters.
 public String getMake() {
  return make;
 public void setMake(String make) {
  this.make = make;
 public String getModel() {
  return model;
 public void setModel(String model) {
  this.model = model;
 public String getAndroidVersion() {
  return androidVersion;
 public void setAndroidVersion(String androidVersion) {
  this.androidVersion = androidVersion;
 public String toString() {
  return "Smartphone [make: " + getMake() + ", model: " + getModel() + ", android version: "
     + getAndroidVersion() + "]" ;

This is about a bean class containing JAXB annotations, in order to easily handle the properties we want to be traversed from/to XML. According to this When a top level class is annotated with @XmlRootElement maps a class or an enum type to an XML element (in our case,  the <smartphone> tag).

 @XmlAccessorType controls default serialization of fields and properties. That is, it allows us to configure the use of fields or properties to access the data in our domain object (Smartphone object). This is specified as an XmlAccessType (PUBLIC_MEMBER, PROPERTY, FIELD, or NONE) via the  @XmlAccessorType annotation. We use access type FIELD to cause JAXB implementations to create bindings for fields and annotated properties. So in our case all fields (make, model, androidVersion) are marshalled/unmarshalled by JAXB.

The toString() method has to be overrided in order to get a human-readable output format. Otherwise, for each object that will be manipulated, the output will be something like Smartphone@4bbc148 .

We obviously need a second class that holds a list of Smartphone objects:
package com.toubou91.jaxb.example;

import java.util.ArrayList;
import java.util.List;

import javax.xml.bind.annotation.XmlAccessType;
import javax.xml.bind.annotation.XmlAccessorType;
import javax.xml.bind.annotation.XmlElement;
import javax.xml.bind.annotation.XmlRootElement;

@XmlRootElement(name = "smartphones")
public class Smartphones {
 @XmlElement(name = "smartphone", type = Smartphone.class)
 private List<smartphones> smartphones = new ArrayList<smartphones>();
 public Smartphones() {}
 public Smartphones(List<smartphones> smartphones) {
  this.smartphones = smartphones;
 public void setSmartphones(List<smartphones> smartphones) {
  this.smartphones = smartphones;
 public List<smartphones> getSmartphones() {
  return smartphones;

2.3 The Helper class

Let's create a helper class to easily unmarshalla requested XML file.
package com.toubou91.jaxb.example;

import java.util.List;

import javax.xml.bind.JAXBContext;
import javax.xml.bind.JAXBException;
import javax.xml.bind.Unmarshaller;

public class JAXBXMLController {

 public static List<smartphones> unmarshal(File file) throws JAXBException {
  Smartphones smartphones = new Smartphones();
  JAXBContext jaxbContext = JAXBContext.newInstance(Smartphones.class);
   Unmarshaller jaxbUnmarshaller = jaxbContext.createUnmarshaller();
  smartphones = (Smartphones) jaxbUnmarshaller.unmarshal(file);
  return smartphones.getSmartphones();

2.4 The Demo

Finally, let's test what we just created!
package com.toubou91.jaxb.example;

import java.util.List;

import javax.xml.bind.JAXBException;

public class Demo {

	public static void main(String[] args) {

		List<smartphones> smartphones = null;
		try {
			smartphones = JAXBXMLController.unmarshal(new File("src/smartphones.xml"));
		} catch (JAXBException e) {

3. Git repo

You can also find the corresponding source code in this github repo.

Saturday, April 11, 2015

How to change the default installation directory in Windows

1. The problem

I've got a partitioned ultrabook, where OS is installed in the C:\ drive. That is, my C:\ drive has a small amount of available GBs, so, each time I want to install a new software, I have to manually change the setup wizard's default directory ( C:\Program Files\ ) to my D:\ drive.

2. The solution

In order to change the default installation directory for a Windows machine, we have to modify its registry:
  • Start -> Run, type
    and hit OK.

  •  Navigate to HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\Windows\CurrentVersion and change the highlighted variables (right-click -> Modify) with the desired ones:


Sunday, March 22, 2015

How to solve Failed to execute 'send' on 'XMLHttpRequest': Failed to load [filename]

So, the other day I was writing my AngularJS Routing Example for the WCG community and I suddenly got in a kind of trouble: I couldn't seem to get it working!!!

My app's critical point was that I was trying to deploy an app that included ng-view, to a browser.

By default, a browser doesn't allow AJAX requests to files located on your local file system. This happens for security reasons. But, ng-view uses an AJAX request to load templates, so the options are:

  • Configuring your browser in order to allow local files access; (this link explains how to implement this action in Google Chrome)
  • Run your app in a local web server.

The first option is quite straightforward, so I'll stick with the second one. I recommend users that don't have locally a web-server, to go for an open-source solution, like http-server, which can be configured very easily.

I'm a Java lover and I have installed one distribution of Tomcat in all of my working instances, so the following demonstration is about running an Angular app to Tomcat:

  • Copy or move your app's folder to Tomcat's /webapps folder and execute:

  • Navigate to Tomcat's /bin folder and execute the startup script:

  • Verify that Tomcat deployed your app:

  • Access your app from Tomcat, you should have a clear console now:

Tuesday, March 10, 2015

Beautifying Javascript's default alerts with sweetalert

Ok, I was in the middle of my upocoming (scheduled for the 17th of March) AngularJS Form Validation example for the Web Code Geeks (WCG)  community, using Twitter Bootstrap and AngularJS and I was quite excited about that, but still my sample app didn't look so beautiful.

What resulted to this condition, was the fact of the deprecated JavaScript's default alert functions, so after a quick search, I found out that sweetalert matched my needs.

SweetAlert is not a jQuery plugin, but a combination of JS, CSS and some images, which together, produce a very catchy product that every front-end developer would love making use of.

What to include in your project in order to use it

All you have to do, in order to use these fancy alerts in your projects, in to download the package from the fore-mentioned link (or just the lib folder) and reference the JavaScript and CSS files manually:

How to use it

A simple alert:
swal("Hello World!");
An error message - the arguments' order is: title, text, message type:
swal("Oops...", "Something went wrong!", "error");

Using sweetalert in a real project

A good starting point would be this repo of mine, where I'm implementing an AngularJS form validation app, using Bootstrap and sweetalert.

 You will be able to read about the project's analysis (from Angular's perspective) in a week, as this is an example that will be published in WCG, as I fore-mentioned.

For the moment, here are the product's screenshots from the successful and erroneous app's alerts:

Success alert

Error alert

Tuesday, January 13, 2015

Create your own movie database easily with SmartMovieDB!

You can directly jump to the git repository that hosts this project :)

There's a pretty nice source over the internet, regarding the creation of your own movie database.
You can read in detail its usage and functionality over the fore-mentioned link, but, generally, the following schematic explains how it works:

That is, you give as an input a text file that contains the desired movies' names and with the help of the perl scrapper, you get an output of an sql format with all the sql commands that need to be done to your database.

This means of course some extra effort, so why don't we go for an automatic process? What about an sql script runner, which takes as an input an SQL file, connects to the specified database and runs the generated from the scrapper script, queries?  It would be nice if we could stick with Perl, but I didn't had the luxury to waste a lot of time on Perl-MySQL connection, so I chose the easy way, which is Java and fortunately, there is an easy way to do it, again with the support of an open source project (we 'll here need only one part of it, as you noticed, too).

I'm choosing to start with the Java project, which is responsible for reading an sql file and executing the included queries to a given database connection:

  • defines a valid database connection.
  • is responsible for the convertion of sql statements into databse queries.
  • coordinates the game, as the script runner's instance needs a connection argument, which is actually passed by calling the getConnection() static method of
And that's is! We 're good to go with the script runner's part.

What is left now, is to find a way to connect the script runner (that takes an sql file and executes the existing queries into a db) with the perl scrapper.

According to the scrapper's part, I'd firstly like to introduce a small parenthesis: I support open source software, so I'm here using the OMDB Api, instead of the iMDB's API, so, together with some updates that had to be done to the scrapper script, here is the updated version of

#!/usr/bin/perl -w
use strict; 
use XML::Simple;
use Data::Dumper;

my $xml = new XML::Simple;

die "Please make that you the movie title is provided into quotes!\n" if(!@ARGV);
my $movie = shift;
$movie =~ s/\s/+/g;

my $cmd = "curl$movie&y=&plot=short&r=xml";
my $movieData = `$cmd`;
my $data = $xml->XMLin( $movieData );

my $released = escapeSingleQuote($data->{movie}->{released});
my $rating = escapeSingleQuote($data->{movie}->{imdbRating});
my $director = escapeSingleQuote($data->{movie}->{director});
my $genre = escapeSingleQuote($data->{movie}->{genre});
my $writer = escapeSingleQuote($data->{movie}->{writer});
my $runtime = escapeSingleQuote($data->{movie}->{runtime});
my $plot = escapeSingleQuote($data->{movie}->{plot});
my $imdb = escapeSingleQuote($data->{movie}->{imdbID});
my $title = escapeSingleQuote($data->{movie}->{title});
my $votes = escapeSingleQuote($data->{movie}->{imdbVotes});
my $poster = escapeSingleQuote($data->{movie}->{poster});
my $year = escapeSingleQuote($data->{movie}->{year});
my $rated = escapeSingleQuote($data->{movie}->{rated});
my $actors = escapeSingleQuote($data->{movie}->{actors});

my $tstamp = time();

print "INSERT INTO movie_collection VALUES (NULL , '$title', '$year', ";
print "'$rated', '$released', '$genre', '$director', '$writer', '$actors', '$plot', ";
print "'$poster', '$runtime', '$rating', '$votes', '$imdb', '$tstamp');\n";

sub escapeSingleQuote {
 my $str = shift;
 $str =~ s/\'/\\'/g;
 return $str;

Once the sql file creation is done , we 'll make a system call from Perl to run the exported jar file of our Java project. This means that we need to add the following line in the end of the script:   
 my $cmd = "perl \"$_\""; 
system "java -jar absolute/path/to/the/exported/jar/file.jar"
Finally, everything is fired up from command line, so, keeping in mind the exact locations of the required files (movies list, jar and generates sql's file), execute the following command:
perl movielist.txt > sqlInserts.sql

For more details, there's also a git repository that hosts this project :)