INSTRUCTIONS FOR SETTING UP, RUNNING and USING A KPML
GENERATION SERVER ...

January 2016
John Bateman, Bremen University

This provides a template for a very simple generation
server that takes SPL expressions as strings across a port and which
writes the generated string on standard output.

1) Startup KPML as usual and load the language resources that the server 
   is to use. Following this, it should be possible to generate with
   the loaded language using the 'say' function as usual. If this is OK,
   then the server can be started as follows.

2) Load the file 
   "... \KPML-LW-PE\KPML4\Processes\server\server-client-lw5\server-eg.lisp"

3) In the Lisp listener, start the server with:

   (cl-user::start-kpml-server)

   This starts a server waiting for SPL inputs as strings on
   the socket 10244. 

4) SPLs can be sent to the socket and the results will be 
   either printed, or returned, depending on the client.
   Very basic example clients for Java and Python are provided.
   For example, following starting the server, the following
   instruction given on the command line:

   .../server/kpml-mini-client.py "(s / sail)"

   should print:

   "Anything sails" 

   if English was the loaded grammar.

