JSwat Project How to Write JSwat Commands

How to Write New Commands

All commands must subclass from the JSwatCommand class, defined in the com.bluemarsh.jswat.command package. JSwatCommand provides some basic functionality needed for all JSwat commands. Each JSwat command class is named using the name of the command as the prefix, followed by the "Command" suffix. For example, if I am going to write the 'friday' command, I will name the class "fridayCommand" and place it in the com.bluemarsh.jswat.command package. Then when the user invokes 'friday' at the JSwat command prompt, the CommandManager class will find my friday command and call its "perform" method.

JSwat commands can do just about anything any ordinary Java code can do. But you are probably more concerned with what JSwat commands can do in the context of JSwat. All of the work of a JSwat command is performed in its perform method. This method is given several parameters which enable the command to perform its task. The first parameter is the Session object running the command. The Session provides access to most everything the command could ever want, including threads, context manager, event handler, source code manager, CommandManager, etc. The second argument is a StringTokenizer which provides the command arguments. For instance, if the friday command took two arguments, a thread ID and an object reference, those would be passed in the args parameter. It is up to the friday command to parse the strings and resolve the arguments. The third and final argument to the perform method is the Log that the command may write messages to. This is the same log as the one returned from Session.getStatusLog().

Back