Commands

How it works

Commands are processed by regex. Almost all commands (can) have derivatives.

Everything happens in the processing.js file.

Commands are stored in an array (q) of objects.

A command object is created as follows :

pattern : regular expression of the matching command
before : (optional) function to execute before answering command
answer : (optional) string to output as answer
after : (optional) function to execute after answering command

Of course, at least one of the three optional properties is necessary to make something happen.

When it receives a command, the script will check regex matches in order that they are written in the code.

Random answers

In the same file, there is a phrases array that contains Samaritan well-known answers.

The script will randomy pick one of these sententences matching the /^random$/ regular expression after 3.5 seconds of "calculating response".

Calculating response

You can call the "calculating response" UI feature using the output.calculating function and pass the following object in parameter :

state : boolean determining if "calculating response" mode is enabled or disabled
duration : (optional) number of seconds while this mode is enabled
callback : (optional) function which will be executed when calculating mode stops automatically or manually.

Fallback

I wanted to implement a fallback solution to make Samaritan answer to everything. I firstly tought about Mitsuku because it's the best, but it doesn't provide an API. So I think Cleverbot would be great enough, that's why there's an empty cleverbotobject in the index.js file.