misconstruedrhinoceros / stackspy Goto Github PK

BETA STATUS -- NOT ALL MODULES WORK

This plugin identifies what technologies comprise a web app's stack.

##Use StackSpy can be required like any other node module, and it returns an object like the example in the design section.

StackSpy can also be used as a command line tool by passing in a url:

  node stackspy.js http://example.com

The main file takes in a url and returns an object of results. Should a non fatal error occur, the object will contain an error flag set to true.

For sanity: All technology names are lowercased regardless of official spelling. The ending 'js' in JavaScript technologies are dropped.

If you don't agree with this convention, we can have an empathetic feedback session where we discuss the merits of this approach, and then change nothing.

ex:

  { client: [
    { name: 'jquery', 
      found: true, 
      version: null, 
      src: 'example.com'
    },
    { name: 'backbone',
      found: true,
      version: null,
      src: 'example.com',
    }],
    server: [
    { name: 'node',
      { certainty: 0.7,
        version: X.X.XX,
        misc: {
          middleware: express,
        }
      }
    }]    
  }  

The client-side technologies are identified by parsing content returned from from HTTP requests.

A request (or series of requests) are made and each request gets passed off to modules which determine if particular technologies are present client side.

Each module should expect a response object straight from an HTTP request made by the request node module.

Each module should return an object that contains the technology tested for, if that technology was found, and if so, the source url and version for that technology.

eg:

  { name: 'backbone',
    found: false,
  } 

  /** OR */
  { name: 'react',
    found: true,
    version: X.X.XX,
    source: cdnjs.com/foo/bar/react.min.js,
  }

###Server The server-side technologies are teased out by piecing together various clues, using the url of the website to interrogate the server.

Each module recieves the url against which it should work.

Each module should return an object with the technology it was checking against, the certainty it found that technology on a 0-1 scale, versioning information, and plugin/middleware information dependent on the technology.

eg:

  { name: 'rails',
    certainty: 0.1,
    version: null
  }

  /** OR */
  
  { name: 'node',
    certainty: 0.7,
    version: X.X.XX,
    misc: {
      middleware: express,
    }
  }

###Certainty The working definition of certainty is (number of clues hitting positive) / (total number of clues run for that tech).

This definition will advance as the project does.