We're happy to present a new version of the wonderful MetaModel component. This version adds a radical new feature: Cross-datastore querying, which means that you can now execute queries that spans multiple datastores (ie. with transparent client-side joining, filtering, grouping etc.). You can check out a simple example of this at kasper's source (blog).

Version 1.2 also adds support for two long-awaited datastores: Microsoft Access databases and dBase databases. Access support is implemented for MetaModel with a core based on the Jackcess project. MetaModel's dBase support is based on a derivate of xBaseJ, courtesy of xBaseJ, American Coders and Joe McVerry.

To look into MetaModel 1.2, here are the crucial resources:

• Downloadables at google code.
• Javadocs available online.
• Maven-support out of the box:

  <dependency>
    <groupId>dk.eobjects.metamodel</groupId>
    <artifactId>MetaModel-full</artifactId>
    <version>1.2</version>
  </dependency>
  

With MetaModel 1.2 we're feature-complete with all of the 1.x features of the MetaModel-roadmap. We hope that you will find it to be as great and useful as we ever intended it to be!

I'm happy to announce the release of MetaModel 1.1.8.

This release is a minor release with updates only relating to MS SQL Server. The changes are, however, profound in this regard. Microsoft SQL Server JDBC drivers are known to be quirky when it comes to metadata exploration and we are happy to say that MetaModel now addresses these issues. So if you're a MS SQL Server you should be sure to get the latest version of MetaModel!

MetaModel is as always available at the following locations:

• Downloadables at google code.
• Javadocs available online.
• Maven-support out of the box:

  <dependency>
   <groupId>dk.eobjects.metamodel</groupId>
   <artifactId>MetaModel-full</artifactId>
   <version>1.1.8</version>
  </dependency>
  

We hope you're all satisified with the improvements of this release and don't hesitate to give us any feedback.

We're pleased to announce the release of MetaModel 1.1.7. The major changes from our latest release is the introduction of two important improvements:

• Microsoft SQL Server is finally supported and integration tests have been added to our portfolio of tests of supported databases. Thank you to Asbjørn Leeth for the major contributions of this feature.
• We've added an option to configure the character encoding for opening CSV files.

With the addition of these two improvements we think that we've added some significant "drops in the ocean" on our way of becoming the most comprehensive and advanced framework for object-oriented querying and datastore-independent schema exploration.

If you use Maven, update your dependencies to the following:

<dependency>
 <groupId>dk.eobjects.metamodel</groupId>
 <artifactId>MetaModel-full</artifactId>
 <version>1.1.7</version>
</dependency>

... or if you don't, head on over to our download site at Google Code and download a copy of the release.

We've released yet another version of MetaModel, namely version 1.1.6.

This release contains very few changes to the 1.1.5 release:

• A convenience method was added to the Query class: select(FunctionType, Column).
• Upgrading the Apache POI version in MetaModel introduced a few bugs that we did not discover in the 1.1.5 milestone. In 1.1.6 we fixed these bugs and unittesting was significantly improved for this part of the code to prevent any new bugs from emerging.

We hope you enjoy this release and excuse for the hectic release schedule - the before mentioned bug fixes where critical and we hope that you appreciate the quick response from the community.

We have just released the newest version of MetaModel, 1.1.5. This release is a minor release which means no API changes, but a few upgrades in terms of performance, flexibility and ease of distribution (full list):

• The most important upgrade have been to CSV performance. We encountered a bug when querying this type of datastore that meant that the whole DataSet was stored in memory while using it. This has undergone quite some refactoring so that it will now stream through memory as expected, thus keeping the door open for very large CSV files.
• A minor change in the column naming scheme have been implemented for the Excel-based DataContext's. This means that if the first row of a spreadsheet contains only blank fields, we will automatically assign the names "[column 1]", "[column 2]" etc. accordingly.
• The downloadable zip or tar.gz file will now contain a "MetaModel-1.1.5-all.jar" file, which is an assembled jar file containing the classes of all MetaModel modules (core, csv, jdbc, excel etc.), which should substantially ease deployment of the framework.

We hope you enjoy the new release of MetaModel and keep up the good work of providing the valuable feedback that drives development of it.

Today we are announcing the first company, Lund&Bendsen, to officially support DataCleaner and MetaModel on a commercial level. These eobjects.org projects are, as you know, independent projects that are run with the community in mind. But as time goes on they grow and for companies to pick them up and start using them in a commercial setting we also welcome third party commercial support to help spread the projects to environments where community-based support is insufficient.

Lund&Bendsen is a Danish company with a strong expertise in Java development and training. Their service offerings include training, customization, integration and enhancement of DataCleaner and MetaModel so if your company is considering applying DataCleaner they might be interested in hiring some professionals to aid them in the process.

Over time more companies are expected to join in on commercial support for the eobjects.org projects. Keep up to date on the DataCleaner support page and don't hesitate to contact us for any inquiries in this regard either.

A new release of MetaModel is ready for download. The new version, 1.1.4, is a bug-fix release with a critical issue for PostgreSQL databases fixed. Other than that no changes from 1.1.3, so it should be a drop-in replacement update.

Enjoy.

• You can download an archived version
• Or get it using maven:

  <dependency>
   <groupId>dk.eobjects.metamodel</groupId>
   <artifactId>MetaModel-full</artifactId>
   <version>1.1.4/version>
  </dependency>
  

As we where recently made aware of, we have once again messed up our maven deployments of MetaModel, sorry! If you're using maven for your Java projects and you just updated your <dependency> tag in your POM files, replacing the version entry "1.1.2" with "1.1.3", I'm sure you ran into a lot of ClassNotFoundException's, because the maven artifacts where in fact empty! We are very sorry about this poor release management situation, but here are a couple of ways that we (you) can fix this:

1. You can add the eobjects maven repository to your POM. The eobjects maven repository contains valid maven artifacts so that's quite an easy fix:

   <repositories>
    <repository>
       <id>eobjects-maven</id>
       <name>Eobjects repository for Maven</name>
       <url>http://datacleaner.sourceforge.net/m2-repo/</url>
     </repository>
   </repositories>
   

2. You can wait a few hours and the central maven repo will have been updated with a couple of new artifacts with the "1.1.3-FINAL" version literal. So your new dependency will look like this:

   <dependency>
      <groupId>dk.eobjects.metamodel</groupId>
      <artifactId>MetaModel-full</artifactId>
      <version>1.1.3-FINAL</version>
   </dependency>
   

We've just released MetaModel version 1.1.3. This is a stabilization release containing some microscopical bugfixes, specifically in regards of Schema serialization. If you're currently using any 1.1.x release of MetaModel, then you should do a drop-in replacement and expect no changes to your code.

As always MetaModel is available from our downloads page and through the maven repositories.

Unless anything urgent comes up this will be the last release of the 1.1 branch of MetaModel. The next focus of MetaModel 1.2 will be to include support for more datastore formats, including dBase and improved XML tag-to-table modelling.

And of course if you have any ideas for development, don't hesitate to let us know!

We have identified a problem with the MetaModel 1.1.1 artifacts in the maven repository, so we will be releasing MetaModel 1.1.2 very shortly. The problem was related to the upload process and caused the jar's in the repository to not contain any .class files! The downloadables from our website did not suffer from this problem, so if you're using those, you're OK.

The new maven artifacts can be downloaded using this dependency tag:

<dependency>
 <groupId>dk.eobjects.metamodel</groupId>
 <artifactId>MetaModel-full</artifactId>
 <version>1.1.2</version>
</dependency>

A single feature have been added in the 1.1.2 release - CSV and XML content is now accessible not only through files but all kinds of input sources, including internet URLs.

Update: After some repository-synchronization waiting time the 1.1.2 release have finally been submitted to the public Maven repositories!

We've just released MetaModel 1.1.1, the successor to the major 1.1 release!

This release is a minor fix release and you should be able to make an easy drop-in replacement of the 1.1 release. Here are the three fixes/improvements that we have been working on for the update:

• Minor bug fixed: The equals() method of SelectClause had a minor bug related to comparing the distinct property.
• Improvement: The Column and Table classes have had a getQualifiedLabel() method added. The qualified label is a dot-separated qualified name such as "MY_SCHEMA.MY_TABLE.MY_COLUMN". The qualified label can be used as a unique identifier for the column but is not necessarily directly transferable to SQL syntax.
• Improvement: Getters and setters have been added to the SelectItem class

As I write this newsitem I'm uploading the new version of MetaModel to the maven repositories! So let me take the time to tell you what's new in this release.

First of all I'd like to say that this is really a release with lots of fundamental changes and we have sacrificed backwards-compatibility at some places, so be sure to check that everything is working exactly as before. That said - those things that we have changed will also cause you compilation problems, so if you do a drop-in replacement and your build fails, then it's because the features have changed. We think this is the easiest way for everybody to deal with changes - it's a lot more obvious that you need to do something if it's really keeping your application from working! The good thing is that the new MetaModel provides a lot of great improvements and new features!

Here's a sum-up of the changes made to MetaModel from version 1.0 to 1.1:

• We've done a major restructuring of the project as to make it more modular and easier to figure out.
• This also means that the way you create DataContext objects have changed. In 1.0 you used the constructor of DataContext. This approach have been replaced by a factory class, which does all the instantiation and initialization stuff for you: DataContextFactory.
• The MetaModel project is now LGPL licensed instead of using the Apache License version 2.0. For more info see MetaModelLicense.
• The built-in query-engine, "Query postprocessor", which is used to serve CSV, Excel and XML content, have gone through numerous improvements to performance and functinality.
• Column types can now be detected, narrowed and transformed using the Query postprocessor engine. This means that you can use the engine to detect and retrieve Integer, Double, Date, Time and Boolean types as well as the old String-based values, even from text-only datastores such as CSV files.
• The JDBC datastores now have a query rewriter component which allows for optimization of queries using native SQL-syntax.
• Query postprocessor now also generates information schemas used to investigate metadata about CSV, Excel and XML files.
• Database compliancy have grown constantly during development and will keep doing so forward on. You can check the supported databases here: MetaModelCompliancy

All in all I think this release marks a high degree of maturity for the MetaModel project and we're very proud to present it to you!

We've made a principal decision at eobjects.org to change the preferred license of our projects from the Apache License 2.0 to the Lesser General Public License (LGPL).

The main difference between the two licenses are that the LGPL requires any modifications to be contributed back to the Open Source community (ie. licensed under a similar license; LGPL or GPL). The eobjects.org projects are gaining the obvious advantages of the LGPL by ensuring that improvements are submitted back to the projects. This also means that we don't risk that anyone sell modified versions of our projects. It is still just as appropriate to use the projects as a part of commercial applications, but any modifications must be contributed back to the community.

• The Apache License 2.0: http://www.apache.org/licenses/LICENSE-2.0
• Lesser General Public License: http://www.gnu.org/licenses/lgpl-3.0.txt

Initially this change in license will affect the two flagship projects of eobjects.org: DataCleaner and MetaModel. This means that the next versions of these projects (DataCleaner 1.5 and MetaModel 1.1 accordingly) will be LGPL licensed. Also, new projects will be LGPL licensed unless special circumstances suggest otherwise.

We're happy to announce that MetaModel 1.0.7 has just been released! The new release should be a drop-in replacement with minor improvements and bugfixes:

• Improved memory handling and fixed a very slight memory leak (#191).
• Added support for RIGHT JOIN when using the embedded query engine (#175).
• XML support have been improved with more precise column types (#176).

You can download the new MetaModel at our google code download site. For all you maven people out there, here's the update to your POM:

<dependency>
 <groupId>dk.eobjects.commons</groupId>
 <artifactId>MetaModel</artifactId>
 <version>1.0.7</version>
</dependency>

Enjoy.