The data in the Pcd GL is organized hierarchically. A JNDI based API allows to build and traverse "folder" hierachies, and create, modify, read and delete objects. The data are physically stored in a database.
If you need to learn more about JNDI, a good starting point for you might be SUN's JNDI tutorial. Many details on JNDI are found in the JNDI API documentation.
Applications that use the Pcd GL fall into two groups: consumers and object providers.
An object provider provides a Java implementation of its own semantic object (that's trivial), and conversion functionality, so-called state and object factories, that allow to convert instances of this semantic object into the low-level data model of the Pcd, and vice versa.
A consumer uses the JNDI API with or without the Pcd specific extensions to lookup and work with the objects that are stored in the Pcd. Consumers do typically not work with the low-level data-model of the Pcd, but with the respective semantic objects of the object providers.
The API comprises four packages:
com.sapportals.portal.pcd.gl |
Pcd-specific extensions of the package javax.naming.directory |
com.sapportals.portal.pcd.gl.admin |
Miscellaneous administration interfaces |
com.sapportals.portal.pcd.gl.notif |
Using this implementation of the observer pattern, the semantic objects can be notified of any (direct or indirect) change to their underlying persistence object |
com.sapportals.portal.pcd.gl.spi |
Classes that are valid return values in the StateFactory-call. In bind or rebind calls, the object providers use these classes to indicate that a non-standard object should be created in the persistency |
The Pcd GL extends the functionality of the standard JNDI API in following ways:
See the package description of the package com.sapportals.portal.pcd.gl for more information.