jamshell

JAMScript Shell Program
=======================

The JAMScript Shell allows multiple programs to be hooked and executed. The binary is currently named jamshell and would be the name installed to the system path

Starting the Shell (before npm install)
---------------------------------------

Via testing it can be started with:

node index.js

For example: node index.js flow

The Hook names currently defined are: "flow", "jdata", "jdiscovery", "jview", "jamscript". However, these can be modified or removed or more could be added.

Starting the Shell (after npm install)
-----------------------------------------

After installing, running the application from the terminal should be as easy as:

jamshell

For example: jamshell flow

Adding/Modifying Hooks
----------------------

The index.js file in the root directory has a types variable which states the Hooks which are currently supported.
All Hooks extend the Hook class defined in hook.js. The hook class defines a prompt property which will be used as the shell prompt.
The default is App>. This can be modified in the extending class. Each hook can also define the character that triggers the command to be executed. By default the newline character (\n) is used. This can be modified by setting the trigger property in the constructor. For example:

``javascript
var Hook = require('./hook.js');

class MyHook extends Hook{
constructor(opts){
super(opts);
this.trigger = ";" //trigger command when the user presses semi-colon
}
}
`

Adding Commands to Hooks
------------------------

When ever a command is triggered, the execute method of the active Hook receives the command and is expected to run synchronously. The Hook class defines utility methods for parsing commands. These could be used by subclasses. The same build object can used to show Usage help and also parsed to obtain required data from the terminal.
As example of the build object is shown below:

`javascript
class MyHook extends Hook{
...
execute(code){
// command: ds
// description: List the data streams in the data source specified in
// options:
// -f, --from The offset/index to start from
// -t, --to The index to end
// -h, --help See usage help

//If we have the above elements for our command, we can build it as follows:
var build = {
cmd: "ds ",
desc: "List the data streams in the data source specified in ",
opts: [
{cmd: "-f, --from ", desc: "The offset/index to start from"},
{cmd: "-t, --to ", desc: "The index to end"},
{cmd: "-h, --help", desc: "See usage help"}
],
code: code,
number: ['from', 'to']
};
}
...
}
`

In order to parse the build object, we call the method defined in the Hook class as:

`javascript
var parsed = this.parse(build);
`

To generate the Help from that object we can do the following:

`javascript
this.printHelp(build);
`

Ideally, we would want to first check if the user has specified the help flag by doing:

`javascript
var parsed = this.parse(build);
if( parsed.help ){
this.printHelp(build);
return;
}
`

The required options for the build are:

1. cmd: The command pattern as is or ... .
Now bear in mind that the names the command arguments are significant to retrieving the data stored in them.
As an example:

`javascript
var build = {
cmd: "listfiles