An ad-hoc command is responsible for executing the provided service and
storing the result of the execution. Each new request will create a new
instance of the command, allowing information related to executions to be
stored in it. For example suppose that a command that retrieves the list of
users on a server is implemented. When the command is executed it gets that
list and the result is stored as a form in the command instance, i.e. the
getForm
method retrieves a form with all the users.
Each command has a node that should be unique within a given JID.
Commands may have zero or more stages. Each stage is usually used for
gathering information required for the command execution. Users are able to
move forward or backward across the different stages. Commands may not be
cancelled while they are being executed. However, users may request the
"cancel" action when submitting a stage response indicating that the command
execution should be aborted. Thus, releasing any collected information.
Commands that require user interaction (i.e. have more than one stage) will
have to provide the data forms the user must complete in each stage and the
allowed actions the user might perform during each stage (e.g. go to the
previous stage or go to the next stage).
All the actions may throw an XMPPException if there is a problem executing
them. The XMPPError
of that exception may have some specific
information about the problem. The possible extensions are:
malformed-action. Extension of a bad-request error.
bad-action. Extension of a bad-request error.
bad-locale. Extension of a bad-request error.
bad-payload. Extension of a bad-request error.
bad-sessionid. Extension of a bad-request error.
session-expired. Extension of a not-allowed error.
See the SpecificErrorCondition
class for detailed description
of each one.
Use the getSpecificErrorConditionFrom
to obtain the specific
information from an XMPPError
.