PyKQML

PyKQML is an implementation of KQML messaging in Python.

Installation

PyKQML can be installed as

pip install pykqml

Note that releases of PyKQML up to 0.5 work in Python 2 only, whereas releases above and including 1.0 work in Python 3 only.

To install for Python 2, use:

pip install pykqml==0.5

To install for Python 3 (or force an upgrade to a compatible version), use

pip install "pykqml>0.5"

Usage

PyKQML implements the following KQML classes, which allow constructing and manipulating KQML messages programmatically:

KQMLToken
KQMLString
KQMLQuotation
KQMLList
KQMLPerformative
KQMLReader
KQMLDispatcher
KQMLModule

You can import KQML classes as, for instance,

from kqml import KQMLList

You can create a new KQML messaging agent in the context of the TRIPS system as

from kqml import KQMLModule

class MyAgent(KQMLModule):
    name = "MyAgent" # This is the name of the agent to register with

    def __init__(self, **kwargs):
        # Call the parent class' constructor which sends a registration
        # message, setting the agent's name to be recognized by the
        # Facilitator.
        super(MyAgent, self).__init__(name=self.name, **kwargs)

        # Subscribe to REQUESTs of interest. The list will change
        # depending on the role of the agent
        for req in ('what-next', 'commit', 'evaluate'):
            self.subscribe_request(req)

        # Subscribe to TELLs of interest if needed. This list will change
        # depending on the role of the agent
        for tell in ('log-speechact', ):
            self.subscribe_tell(tell)

        # Now signal that the agent is ready to receive messages
        self.ready()

        # Finally, start the listener for incoming messages
        self.start()


    def receive_request(self, msg, content):
        # First, figure out what kind of request this is
        task = content.head().upper()
        # Here you typically decide what to do based on the
        # type of request.

        # Construct reply message's content
        reply_content = KQMLList()
        # Set whatever needs to be set in the reply content

        # Finally, wrap the content in a message and reply
        reply_msg = KQMLPerformative('reply')
        reply_msg.set('content', reply_content)
        self.reply(msg, reply_msg)

Testing

You can run all tests by running nosetests in the top level folder.