<!doctype html public "-//w3c//dtd html 4.0 transitional//en"> <html> <head> <meta http-equiv="Content-Type" content="text/html; charset=iso-8859-1"> <meta name="Author" content="jh"> <meta name="GENERATOR" content="Mozilla/4.73 [en] (Win98; I) [Netscape]"> <title>Eliza Chat</title> </head> <body background="paper.jpg"> <center><b><font size=+2>ELIZA Chat</font></b>.</center> <p><br> <br> <p>There are several "operators" available: <br>  pattern matching: <br>        |       or <br>        &       and <br>        ~       and-not <br>  special pattern matching: <br>        ^       match start <br>        =       exact match <br>  pass 1 scan operators: <br>        #       remark <br>        "       output to stderr <br>        '       output to stdout <br>  response operators: <br>        1-9     weights are the numerals. <br>        %       include the rest of the response. <br>  <br>Some special notes: <br>        #,',",weights, and the first ( must be in column one. <p>Syntax of a simple pattern match: <p>( A | B & C ~ ( D | E | F)) <p>The above would mean if A or B and C and-not D or E or F was in the <br>sentence then pick one of the responses. <p>The responses take on the form: <p>9this is most likely <br>4somewhat likely <br>1least likely <p>However, when a response is triggered, its weight is decreased by one. <br>  <p>Since the above makes little sense, an example will illustrate the use of <br>Chat databases better. <p>#!/usr/bin/chat <br>#change the above line if the chat executable is in a different place. <br>#an example of a "Chat" database <br>"this will be printed to stderr <br>'this will be printed to stdout <p>(apple & ( orange | lemon ) ~ pine cones) <br>9You must have a citrus fruit and an apple. <br>3you do not have a pine cone but you have an apple. <br>2what was % suppose to mean? <p>(=your sentence matches this one exactly) <br>2you are unimaginative. <br>9you lack imagination. <br>  </body> </html>