The Virtual Enterprise
"There are a lot of dated VisualBasic, COBOL and 4GL-based applications out there running on proprietary platforms. Sun servers, Java and VE provide great vehicles in which to migrate to the Web, as either applications or services in a highly scalable open platform architecture."
Katni Venkat,
Sun Microsystems, Inc

VE supports three ways to interact with a relational database. The first and easiest is VE's ability to automatically generate the database schema out of the application object model, along with the appropriate object-relational mapping. The second way is VE/Designer Database Import tool for importing an existing database schema into a package, and managing the object-relational mapping. Finally, the third way is to embed SQL statements and stored procedures as alternate implementations to corresponding VE processes. Although, the first approach is the easiest and the recommended in new application development, one can mix and match between these three approaches in the same application depending on what is appropriate.

When developing a new application without an existing database, VE can auto generate the database schema along with the object-relational mapping, and all of the SQL statements necessary to conduct the full functionality of the application. This is the recommended approach, as it requires no experience with relational databases or SQL, and doesn't involve any manual activities. Just run the Database Deploy tool, identify the database resource to use, and whether you desire a completely new database instance and schema, or update the last schema to reflect changes to the object model. In the later case, VE will preserve your existing database data.

However, when working with an existing database, it is more suitable to import the database schema into VE objects. In this case, the Database Import tool facilitates the viewing and selection of existing schemas, tables, and stored procedures, and maps them back to objects. The mapping activity can be nil if database tables and columns are mapped one-to-one to objects and attributes, or it can involve additional effort for more complex mapping.

The third approach is to add SQL implementations to VE processes. As described in the Application Logic and Integration pages, VE supports defining alternate implementations to any VE process, including a hard-coded SQL statement or stored procedure. In a case where the application logic has limited need to access an existing database, for either query/retrieval or update, this approach might be the most appropriate. SQL implementation supports defining any SQL statement (including ones with parameters) and stored procedures, or invoking existing stored procedures. In addition, the developer has a choice of mapping the results back to objects, or retaining them as a JDBC ResultSet.
Copyright © 1999-2014  INTELLIUN CORPORATION. All rights reserved.