Pradeep's Blog

Data Mapping Using Xsl

2006-12-12T16:43:00.000+11:00

In B2B communications resolution of data between trading partners is one of the most important issue. Let’s consider a scenario where your trading partner sends you a Purchase Order and this PO has got only got a shipping address id. Once you get this PO, you must resolve the shipping address id to an actual destination so that your delivery guys can deliver it. Except the few Big ERP systems most ERP systems do not have to capability to resolve this id to a particular address. Hence the data has to be resolved by a pre-process before the PO enters your ERP System.

Thought there are lots of ways to map information between partners, let me show you how it can be done using an XML mapping file. Let’s also considered that the incoming file is either’s XML or has been transformed to XML. I’ll now show you how an address id can be mapped to the complete address.

Lets consider that the Xml file that we get have the address id as shown

Now instead of the ident field inside the partyaddress something like this would be very useful

In order for us to map the ID to the address we need a mapping file. The actual mapping file could be something like the one shown below; this mapping file will be used in the xsl file to populate the address instead of the address ids.
Now the Xsl part, this is where the actual mapping is done. In the Xsl file create a variable that reads the mapping file like

Now to get the city mapped we could map it as shown above. Here the @id in the mapping file is compared with the ident field in the main xml file. If they are equal then we get the city name from the mapping file.

Google Bomb

2006-10-31T16:45:00.000+11:00

Before me telling what Google Bomb is try searching for miserable failure or failure and liar in Google. Got the answer? Those two are perfect examples of Google bombing. As you all know Google Search has a set of rules by which it ranks pages. So once you understand how these rules work, you can build pages that get better ranking for a particular keyword. The most important criterion for page ranking by Google is the number of pages that link to your page with that keyword. Now that you know this, you can increase the page rank of your webpage for a certain keyword, by increasing the number of links to your page with that keyword. The more the number of pages that link to you with the specific keyword the better the ranking.
The process of influencing Google’s ranking by flooding the pages on the net, with links to a particular page (with a certain keyword) is known as Google Bomb. Most of these Bombs are dropped for commercial reasons.

Deploying AJAX application created using Google’s Web Toolkit

2006-09-17T19:18:00.000+10:00

AJAX without java script! Few months back and I wouldn’t have believed this would be possible. With Google’s Web Toolkit you could code your AJAX applications in Java, and it converts it into Java script for you. Any application created using this Toolkit has three parts to it. First the client side code, this is the code that gets converted into Java script (and runs on the browser) , then the server side code this is the part of the application that runs in your web server (like Tomcat), then the public files, these can be any images or other files that are required by your application.

Since the client side code is translated into Java script, your client side code can only have features that a normal web browser can support (So features like Threads are not supported). In the server side code you can do any thing that Java supports.

The Google’s web toolkit supports two modes for running your AJAX applications. The first one is the hosted mode; this mode is very useful while developing the application. In this mode Google’s simulates the output of your application. The other mode is the web mode. In this mode the Java code is compiled into Java script code and you can run your application from a web server.

The documentation provided by Google for deploying the application in web mode is very limited. I recently deployed my application in Tomcat 5.5 and these are the steps that I followed,

1. Use the application.compile.cmd file to create all the client side files. These files would be placed in the www directory.
2. Get all the server side files compiled.

Consider that we are going to deploy in the ROOT directory in a tomcat.

1. Copy all the client side files (the content of the www directory) into the ROOT directory in tomcat.
2. Then copy all the classes (server side) files to ROOT/WEB-INF/Classes directory.
3. If there are any dependent jar files copy them to ROOT/WEB-INF/lib directory. [NOTE: The packages javax.servelt.*; have to be removed from both the jar files that are provided by Google].
4. Then map the server side servlets (the classes that talk to the client) via the web.xml file in the ROOT/WEB-INF directory.

The url pattern in the servlet mapping should be same in the client code, the application.gwt.xml and in the web.xml files.

If even after following these steps you have problems running you application in web mode, see your web server's logs for more information.

Ajax

2006-08-18T09:33:00.000+10:00

Over the past few months there has been a buzz about the term AJAX. AJAX stands for Asynchronous Javascript And XML.

AJAX as a technology came into the spotlight mainly after Google (GMail) started using it. But AJAX is not a completely new technology, it's old technology with some new tricks to improve the interactiveness of the webpages.

One good news for coders is that you probably know most of the technologies behind building AJAX applications. Some of the technologies involved are HTML, Javascript, DOM, and XML. AJAX appliactions are much more responsive because they are able to send requests to the server without actually reloading the whole page. This is achieved by using the XMLHttpRequest object. The XMLHttpRequest Object enables us to bypass the normal request mechanism followed by the web pages (where the whole page actually reloads) . Instead this object along with Javascript makes it possible to send information to the server without actually refreshing the page, and thus improving interactiveness.

Even Microsoft has taken note of the huge popularity of AJAX and (as always) has planned to have "it's own" AJAX, which is to be integrated with ASP.Net soon.

Search Engine Optimization Tools

2006-07-17T14:26:00.000+10:00

As a search engine optimiser (SEO) you definitely require a set of tools that display your ranking in various search engines, generate keyword, analyse backlinks, etc. Listed below are the various free tools that are available on the net for SEO.

Ranking in various search engines:

http://www.googlerankings.com/

In order to use this tool you need to register with Google and get an API key. This tool enables users to check the rank of your web site for a certain key word with Google, Yahoo and Msn.

I was able to get a high ranking of 16 for my site for the key words “software complexity”.

Keyword Suggestions/Density Checker:

Though most search engines do not consider the keywords anymore for better ranking, its better to have appropriate keywords in your sites. These tools can help you in deciding keywords.

http://googlerankings.com/ultimate_seo_tool.php
https://adwords.google.com/select/KeywordToolExternal

Though these tool are good it would have been better if it could narrow down the keywords rather that giving a big list for us to choose from.

Other tools:

These are other third party tools that are available for free on these sites
http://www.webconfs.com/
http://www.seochat.com/seo-tools/
http://www.seocompany.ca/tool/seo-tools.html

Though these sites provide free tools using Google’s API, I consider most of the functionality provided by these tools to be redundant.

Google’s Alert:

One other good tool that might help webmaster to get more backlinks to their sites is to register with Google alert. With this alerting facility you can easily build more backlinks. Click here to register http://www.google.com/alerts

Passing parameters to XSL from Java

2006-07-03T10:07:00.000+10:00

Generating a HTML from XML can be done using the javax.xml.transform class in the Java’s API. The transformation can be done using the code below;

TransformerFactory tFactory = TransformerFactory.newInstance();
Transformer transformer = tFactory.newTransformer
(new javax.xml.transform.stream.StreamSource(xslFileName));
transformer.transform(new javax.xml.transform.stream.StreamSource
(xmlFileName),new javax.xml.transform.stream.StreamResult
( new FileOutputStream(htmlFileName)));

Note: Remember to provide the correct XSL, XML and HTML file names and the XSL and the XML file should be valid (i.e. all tags should be closed properly)

Now, suppose you want to customise the HTML output depending on certain criteria in your program, how do you do that? This can be done by sending information to the XSL file as parameters. So, the next big question, how can parameters be passed into the XSL file from a java programme?
The javax.xml.transform class provides you with a method called setParameter(String name, Object value). An example of this is given below;

Timestamp ts = new Timestamp(System.currentTimeMillis());
transformer.setParameter("timestamp", ts);

Here the timestamp is passed to the XSL filename. The XSL file can get this value by using a global param field as shown below.
<xsl:param name="timestamp"/>

and used like this
<xsl:value-of select="$timestamp"/>

to print the time that is obtained from the java class.

Note: The parameter name in the XSL should match the one in the java class.

CyVis v0.9.0 Released!

2006-07-02T21:28:00.000+10:00

Are you a Java developer? Do you want a tool that visualises the complexity of your software? We have just released the first public version of CyVis.
CyVis is a software metrics collection, analysis and visualisation tool for software built using Java. It enables developers to identify complex part of their software very easily.
CyVis collects data from java class or jar files hence there is no need for the actual source code. Once the metrics are collected, the statistical information is shown in a way that enables you to immediately detect which part of your software is complicated and which is not.
CyVis can either be integrated with your Ant script or used in command interface to generate reports. It also has a GUI Interface which we consider to be of great help to its users.
Try this cute little tool once and you will love it. Visit http://cyvis.sourceforge.net for more information.

Writing Custom Ant Task

2006-06-05T10:35:00.000+10:00

I wanted to create an Ant task for a project (CyVis - a free software metrics collection and analysis tool for java based software) that I am working on. Writing our own Ant task is actually very simple. This blog explains how to write a very simple (something like a hello world) task that can be integrated into ant.

First write a class that extends org.apache.tools.ant.Task. Then for each attribute, write a setter method like the HelloTask class shown below.

Now compile the class and place it in the classpath or in the same dir as the Ant home directory in order to use the task. Then you can call you task as shown in the script below. Note that before you can use your task, you should use taskdef to define it.

I had placed the HelloTask class file in a jar file and placed the jar file in the Ant’s home directory. You can either do this or define a classpath to say where it’s placed. Running this Ant script would print out some thing like this:

For more complex examples see the Ant’s documentation.

Customizing JTree

2006-05-29T10:48:00.000+10:00

The default look and feel that the JTree provides might not be reasonable for certain applications. For example, the tree structure in eclipse has its own custom icons to represent different things (packages, java files, text files, etc). Let’s have a look at how to customise the JTree, so that we can get to a level of customization as shown below.

There are actually two ways of doing the customization. The first method is straight-forward and this is all that you have to do:

1. Create the tree structure.
2. Create an object of DefaultTableCellRenderer and set the icons as shown below:

ImageIcon customLeafIcon = new ImageIcon(“class.gif”);
DefaultTreeCellRenderer renderer = new DefaultTreeCellRenderer(); renderer.setLeafIcon(customLeafIcon);
tree.setCellRenderer(renderer);

The tree would look something like this now.

Similarly you can change lots of other things. For more details have a look at the API. There are some drawbacks to this method such as, all leaf node icons being the same. So for those of you who would like to tweak a bit more, method two would serve as the better alternative.

The second method is not as simple as the first one. This is what you have to do.

1. Create the tree structure.
2. Create a class that extends the DefaultTableCellRenderer class.
3. Override the method getTreeCellRendererComponent
4. Do all the customization in this method.
5. Create an object of the class that we just created, and set it as the new tree cell renderer.

TreeRenderer renderer = new TreeRenderer();

tree.setCellRenderer(renderer);

The code for the class that extends DefaultTableCellRenderer would look something like this:

And the resulting tree would look something like this.

Session Façade:

2006-05-22T09:46:00.000+10:00

Session façade is one of the design patterns used in enterprise applications. Before I explain how it’s actually implemented lets have a look at the various components that make up the enterprise system.
1. Client: Either a browser or a custom software.
2. Session beans: This is where all the business logic takes place in the enterprise system.
3. Entity Bean: It represents the database in object form.
4. And lastly the database.

Now let’s see how session façade can be implemented for a banking application. Let’s consider that the client can access his account via a browser or ATM. If the client wants to transfer money, from his account to another account, there are few things that have to be done by the bank; check his authorization and his balance.

In session façade the client software would never talk directly to the data base (i.e. the entity bean representing the database) instead it would talk to the session bean where, all the business logic like checking his credit balance and his password would be done. It’s the session bean that talks to the entity bean to get the required data to validate and perform the transfer.

The diagram below shows what I have explained.

Creating Custom Charts in Java

2006-05-15T10:34:00.000+10:00

I am working on an application that required me to draw custom charts. This is what I was required to do: Draw a chart that consists of smaller graphs (which should each have tool tips) and the chart itself should be scrollable. This blog explains how this can be done and the problems I am facing.

This is what I had as a plan before I started coding,

1. For the chart, it has to extend JComponent, so that we can add the graphs to it and to make it scrollable depending on the number of graphs added.
2. The graphs too should extend JComponent so that we can have tool tips on them.

So the code came up something like this

The most important thing in this is to overload the getPreferredSize() method in the graph Component. If this is not done then the graph are given a default size of (0, 0) and the graphs don’t appear. So it’s important to calculate the size of the graph and return it in that method.
The main problem that I am facing is, in deciding the layout manager and the size of the graphs (which is being worked on).

Progress bar for your application:

2006-05-08T10:38:00.000+10:00

Majority of applications need progress bars to show the progress of time consuming operations. Most people have common questions like: Do I need Threads? How to integrate progress bar into my application? This blog answers these common questions.

As for the first question, yes, you need Threads. Without Threads you cannot multitask, meaning you cannot update the progress bar while your time consuming task is running.

The next question is how to integrate progress bar using Threads? Let’s assume that you have a time consuming task in class A. In order to use thread make this class extend Thread. Put the time consuming task in the run method (This enables the task to run in a separate thread by calling the start () method with an object of class A). In this class have an integer variable that you increment according to the progress of your task.

Most applications make the progress bar to popup in a new window. So have it in another class that extends JDialog or JFrame. Add a timer to this class with an action listener added to it. In the action performed method get the integer from class A and set the progress bar’s value to it.

Now that we have both the classes, you can start the task & the progress bar by calling the start () function in class A, followed immediately by a function that starts the timer in the progress bar class.

Finding Class Dependencies in Java

2006-05-01T07:17:00.000+10:00

A requirement for our project, CyVis, was to find the classes that a particular class is dependent on. Finding the class dependencies for a particular class can be done as follows

1. Get the type of all the variables declared in the class. This includes the global variables and local variables in all the methods.
2. Get the return type for all the methods.
3. The type of exception thrown in all the methods.
4. The class that the particular class extends.
5. The interface that the particular class implements.

These five would give us a list of classes, that the particular class is dependent upon. Given below is a code snippet that shows, the things you have to look for in order to find the dependent classes. How you find that is up to you. I am using ASM to get this information directly from the class files.

The import statements do not necessarily give you the actual list of dependent classes. There may be some imports that are actually not needed. For the above class the dependent classes can be found from the fields that I have circled.

Using Derby from Command Line

2006-05-01T07:09:00.000+10:00

Is there a way to add tables into Derby from command line? The answer is yes. Let’s have a look at how it can be done. Derby provides us tools (java programs), that enable us to create, delete, select and update tables in the database from the command line. Given below is a batch file that enables us to access the database.

***********Batch File *****************
@echo off
rem -- Run Derby ij --
set LIBPATH=C:\Sun\AppServer\derby\lib
java -classpath

"%LIBPATH%\derbytools.jar;%LIBPATH%\derby.jar;%LIBPATH%\derbyclient.jar" org.apache.derby.tools.ij %1.
**********End of Batch File ************

This batch file actually sets that classpath and runs a java program with the argument that we pass along. This argument is a file containing the SQL command that we would like to run. Given below is a text file containing a SQL create statement.

*********Text File ********************
connect 'jdbc:derby://localhost/testdb;create=true' user 'app' password 'app';
CREATE TABLE test_table (
col1 VARCHAR(24) NOT NULL,
col2 VARCHAR(24),
col3 VARCHAR(24),
CONSTRAINT pk_test_table PRIMARY KEY (col1)
);
disconnect;
********* End Of Text File **************

The first and last line of the text file enable us to connects and disconnect with Derby. Any SQL statements that you want to run can be written in the middle. Considering that you have saved the batch file as runDerby.bat & the text file as sql.txt. You can run the SQL statement in the text file on Derby as follows.

Prompt> runDerby sql.txt

For more information look here.

Advantages of Unit Testing:

2006-04-24T10:43:00.000+10:00

My project requires me to perform unit testing. Initially we did not follow test driven development, instead the tests were written only after we had some sort of a stable architecture. Writing these test made me realize the advantages of test driven development & unit testing. Here is what I think are the advantages:

You get a very good understanding of the requirements.
Code quality improves, because of you breaking the functionality into much smaller units for unit testing.
Improves your confidence in the application you are building.
I think it enables you to find the problems early.
You are confident to change existing code.

How to integrate Emma with Ant?

2006-04-24T10:28:00.001+10:00

I had to integrate Emma with Ant for my project. I will explain the basic steps involved in doing it. Let’s consider that your directory structure is like this-
Base dir
==Source- Containing the source files.
==Classes – Containing all the class files.
==Test – Containing all the JUNit files.
==Lib – Containing all the jar files needed.

The first task is to set the properties & the class paths. This for our directory structure can be done as:

The second task in your ant code is to compile all the file source and test files.

Now we have the class files in the classes directory.The third and the last task is to perform test coverage using emmajava. But before that, emma task should be defined as shown below:

Now we are all set to use emmajava. It can be done as follows:

You have to substitute the “MyTestSuite”, with a class that starts your testing. Also include & exclude packages according to your needs. The above code generates both html and txt reports (which are placed in the coverage-report folder that was newly created).

The complete documentation of emmajava can be seen at emma’s site .

Do you have a big taxing for loop

2006-04-10T09:23:00.000+10:00

Does you program have very big for loops that it has to run through always? Are these loops making your program slower? Here is a method to over come it.

Lets just consider that you program requires to go through a for loop a million times, this means that there are a million conditions checked (for the loop to go on).

for(long i=0;i<=list.size();i++)

{

System.out.println(list.get(i).name);

}
This is very taxing on any application. So here what you can do

try
{
   long i=0;
   while(true)
   {
       System.out.println(list.get(i).name);
     i++;
   }
}
catch(ArrayIndexOutOfBounds e)
{
//don’t handle the exception
}


Isn’t it simple?

Good logging design in java

2006-04-10T08:54:00.000+10:00

I had to use java’s logging API in my project. Though lots of the information (API & examples) regarding logging can be found online, most of these examples were very simple. These examples were there to explain only to show how certain things can be done using the logging API. None of these sites really explained how logging could be done for larger projects. I looked into other open source projects like hibernate, to look at how they had done their logging before I could start mine. So with that knowledge of how hibernate has done its logging and from my own experience, here are few tips on logging.

- Its always good to have a separate class that generates the logger object. So that any changes to logging mechanism requires you to make changes in just one place.
- Its also better to have your own Log Formatter class, so that you can create customise logging outputs.
- You may have more than one Logformatter class if you are logging to more than one handler.
- I would also recommend you to have just one handler for one logger (I had more that one handler and it complicates things).

System properties in Java

2006-04-03T08:49:00.000+10:00

Want to find out the system's default temp directory, users home directory, user’s name, the path, file or the line separator. These things become very useful if you consider to your application to be platform independent. Finding out these things in java has always been easy.
The System class maintains a set of properties, key/value pairs, which define properties of the current working environment. When the JRE first start these system properties are initialised to contain information about the runtime environment. The complete list of system properties can be found here:

Reading System Properties:

The System class has two methods that you can use to read the system properties: getProperty and getProperties. The example below shows how you can print the user’s home directory.

String home = System.getProperty("java.home");

System.out.println(home);

Writing System Properties:

Though java allows us to overwrite the system properties, it should be done with caution. Any how here’s how it can be done.

java.home =”c:\temp “;

This would make change the user’s home directory to the path specifies in the string.

References:

System Properties

Reading System Properties

Program to list all system Properties

Protectinf System Properties

Logging Packages’s in Java:

2006-04-03T08:42:00.000+10:00

Advantages of logging:

Logging can generate detailed information about the operation of an application. Unlike normal debugging statements they can be easily turned off when not needed and are easy to maintain & increase productivity. Unlike debugging statements logging frameworks allows us to generate output to different destinations like console, files, sockets etc. But the best of all these functionality would be the ability to turn them off and on (either completely or partially) according to our needs.

Logging packages for Java:

My project required me to use a logging framework. So I had to decide on the framework that we would be using. Of the various logging packages, log4j and Java’s own logging API are notable. Both of them have very similar functionality. Logj4 has been around for a long time and is used widely, whereas the java’s logging package was added with jre1.4 in 2003.

Java’s logging package has all the functions that log4j has. In fact it is nearly a copy of log4j with a bit more added to it. The difference is in the areas of appender/handler implementations, formatter/layout implementations, configuration flexibility, and in the levels of logging.

My choice:

Even though log4j is thoroughly tested and very popular, my choice was to go with Java’s own API because it’s already bundled with JRE and moreover has all the functionality that I would need.

How to use Hibernate

2006-03-27T04:16:00.000+11:00

As mentioned in my earlier blog hibernate is used to persist data stored in the objects into a data base. For this to happen there are three things that are needed. One a java data classes (whose object’s properties we want to persist).Two an xml file mapping this class to the respective table, and also the properties of the class into appropriate classes. This file tells Hibernate how & which data to store. The third thing that we need is an xml file that configures Hibernate. The purpose of this file is to tell Hibernate what database it has to use, mapping file information, and other defaults.

Once these three are ready, we can use another class to populate the data, then hibernate can be used to in synchronise the object’s properties into their mapped columns in the database.

Need for an ORM package like Hibernate?

2006-03-26T14:49:00.000+11:00

Relational database management systems are proven data management technology and are almost always a requirement in any Java project. While using objects in our application, we sometimes reach the point where we want them to persist using a RDBMS. We would typically open a JDBC connection, create an SQL statement and copy all our objects property values over.

Since Java does not have SQL type, we are forced to use String concatenation which errors prone and time consuming. This may be easy for objects with small set of properties, but for an object with many properties just preparing the SQL statement would itself be complicated.

That’s not the only problem, what about things like associations, foreign key constraints? And situations like adding a new property to an object half way through the development, trying to use the application with other database; do we go and change all associated SQL statements in the code to suit our new needs? Remember that the problem applies to both loading properties from the database into object’s properties and Vis-versa.

So what can Hibernate do for us? Hibernate basically intends to takes most of that burden of our shoulder. By using Hibernate we only have to define the way we map our classes to tables once - which property maps to which column, which class to which table, etc. After this, we should be able to do things like this:

session.save(MyObj);//statement 1

myObject = session.load(MyObject.class, objectId);//statement 2

The first statement saves the objects properties into the appropriate column in the database table & the second statement the opposite as per our mapping.

Hibernate will automatically generate all the SQL needed to store the object. It relieves us of manual JDBC result set handling, objects conversion, and keeps our application portable to all SQL databases, leaving us to concentrate on the business logic alone. Hence Hibernate helps in rapid product development, minimises maintenance & development costs.

References:

• Road to Hibernate Tutorial

• Tutorials at hibernate.org

• Hibernate in action

Good Ways of Handling Exception in Java

2006-03-20T07:14:00.000+11:00

Not that I know everything in exception handling, but here's what i think are good ways of handling exceptions.

Never use a catch block that suppresses the exception (for example by doing nothing). This makes the program to continue on as if nothing ever happened.
Always try to catch the specific exception, unless you are sure that it makes no difference. For example when invoking a method that throws more that one type of exception, never just catch them using Exception.
Handle exceptions where actions can be taken to deal with the exceptions caused. For example if a block of code throws an exception, handle it there only ifs possible for you to recover from the exception, or else throw the exception to a level where the system can recover.
Having said to throw the exceptions to an upper level where it can be comfortably handled, it dosent mean that you throw an fileNotFound Exception to a place where the applications logical level.
When deciding on checked or unchecked exceptions exceptions for your APIs, its good to throw an checked exception if its possible for the client to recover from the exception, or else it can be should unchecked exception.
Allways clean up afer exception using finally.

How to integrate Junit into Ant?

2006-03-20T06:48:00.000+11:00

Ant provides a flexible JUnit task for compling and executing JUnit tests. I hope that you all know what how test classes are written in java. Consider that we have a class named metricEngine and a class named metricEngineTest to test it.

The first thing that has to be done is that both the classes needs to be compiled. This can be done in Ant as follows (considering that both the files are in basedir itself).

<target description="Compiling Files." depends="init" name="compile.source">
<echo>Compiling Files
<javac destdir="${destdir}" srcdir="${basedir}" target="1.5" source="1.5">
<classpath refid="classpath">
<include name="**">
</javac>
</target>

The classpath includes the list of libary files needed to compile these files. They can be set as follows (consider that the basedir/lib contains all the required library (jar) files.

<path id="classpath">
<fileset dir="${basedir}/lib">
<include name="*.jar">
</fileset>
</path>

Now that both the files have been compiled we can test them, this can be done as follows

<target description="Runs the test suite" depends="compile.test" name="test">
<printsummary="yes" fork="yes" haltonfailure="no" showoutput="yes">
<formatter type="plain" usefile="false">
<classpath refid="classpath">
<lt todir="${basedir}/test-result.txt" fork="yes">
<fileset dir="${destdir}">
<include name="**/metricEngineTest.class">
</fileset>
</batchtest>
</junit>
</target>

MOre information on the JUnit tag can be found with ANT Documentation.

Data Recovery, My Experience !

2006-03-16T12:05:00.000+11:00

My portable hard disk crashed (my computer was able to recognise it but not able to open it) today. I guess some important file that deals with the file format & things like that got corrupted. There was a lot of data in the disk (some were not important & others I had backup). But I did not have a back up for a collection of e-books (around 6GB).

I used software called recovermyfiles. It was pretty good, once I checked the files that I wanted it to recover it started recovering files. But just to scan 1 percent of the disk it took about 15 minutes & the files that were recovered were all named as recovered filesomenumber.extension. I seemed that too much time would have been needed to just sort the recovered data.

So, after just fifteen minutes of my hard disk crashing I am reformatting it (about 80% over now). And what about the ebook collection, well definitely I was not going to read all of them & I know how to get the books that I may need. Good thing it happened, i now dont have so much unwanted stuff...