feat: implement conflicts() for defining mutually exclusive arguments

[bcoe]

This pull request sees @madcampos' hard work over the finish line 🎉

I've made a few minor changes to make the conflicts() api almost identical to its cousin implies().

[bcoe]

@madcampos this is slated to go out with the next release of yargs, I would love the extra help testing:

npm cache clear; npm i yargs@next

Thanks for your contribution!

[lrlna]

should we give an example of how this might work, even like so .conflicts({ 'c': 'a', 's': 't'}), or nah?

[lrlna]

oh woops! i am sorry! i didn't refresh yesterday's page to see if you merged this; ignore me :D

[madcampos]

Will do as soon as possible. 😉

[JeroenVdb]

@madcampos I was exited to use this new feature but it doesn't seem to work with dot-notations? Is this intentionally left out or should it be possible to implement?

var args = require('yargs')
  .option('some.foo', {
	  type: 'string'
  })
  .option('some.bar', {
	  type: 'string'
  })
  .conflicts('some.bar', 'some.foo')
  .argv;

[madcampos]

Sorry, didn't have this case in mind...
I'll review the code and see if it's hard or not to do.

[pensive-platypus]

Is it possible to implement it in a way where one key could conflict with many keys.

.conflicts('foo', ['bar', 'baz'])
//or
.conflicts({'foo': ['bar', 'baz']})

Great work on this btw. :D

[madcampos]

Sorry for the long due feedback…
To implement it as properties of an object the complexity would grow as the object properties would need to be mapped so conflicts knows which are the conflicting keys.

I could not fined a easy and performant object mapper/flattener so it became a problem.