[CompStrm Wiki] commands

[Bill la F. <laf...@ya...>]

OK, so we have a data proxy class which holds no data
(data is either in the database or cached). But this
data proxy class (really a rolon class) holds the API
both for accessing/updating data and the low-level
business logic, in the case of an application-specific
rolon.

Next step up are commands. There is a table in package
org.agilewiki, commands.txt, which maps command names
to the classes which implement them.

Command objects are singltons and thread safe--they
hold no data. Execution is also atomic for commands
which perform database updates directly--they are
invoked in a transactional context and only one update
command can execute at a time.

Commands which only query or which invoke other things
to perform updates are multi-threaded, but do not run
when an update command is executing.

By default, a command is attached to the base rolon
class, org.agilewiki.ark.Rolon, but you can add a
method to override this--which means the command is
only active when a rolon of the appropriate type is
the current context.

A command has a method which returns a help string.

Most important, command classes have a method which
further restricts when the command is active. For
example, an update command is never allowed when the
user does not have permission to modify the rolon
which is the current context.

Creating a new command takes 2 steps:

1. Write a command class with implements
org.agilewiki.framework.CmdInterface.

2. Add the command name and class to the commands.txt file.


		
__________________________________________________________ 
Yahoo! India Answers: Share what you know. Learn something new. 
http://in.answers.yahoo.com