jStorm

• jstorm.zip

• Getting Started
• Cook book
• JavaDoc
• FAQ
• To Do list
• White Paper
• SF Project Page

jStorm originates from DBObjects written by Tim Wellhausen and my work with XP doing evolutionary design. DBObjects was rewritten 'Test First' to enable modification and extension through refactoring and jStorm is the result of that effort. In addition to the features of DBObjects we have added simple mechanisms for aggregation and concurrency stamps.

The objective of jStorm is to reduce coupling between the application code and database design, therefore enabling change without exorbitant cost. jStorm strives to issolate these changes to a single place by generating all SQL code from mappings. It is available under the GPL-- Gareth Reeves

More on using and modifying jStorm

I have put together a new release of jStorm that contains no new functionality but some fairly large refactorings. In particular I wanted to remove the code that sets parameters on Prepared Statements and the code that sets data on a object when it is loaded from the database. Both of these used to work with large switch like statements. I added decorators to the parameters and double dispatching to the statement and the object being loaded. If your interested, its in SqlRunner.java

I also added my acceptance tests to the build. They run in jUnit and can be found in the jstorm.acceptance package.

• jStorm only works with MySql right now. We think that the only modification required to run on other databases is in the way that keys are generated (different in most databases ). Look in the 'SqlRunner' class.
• jStorm only does 1 to 1 aggregation.
• jStorm does not do transactions. See concurrency stamps in the cookbook for a poor man's alternative.