Refactor, eslint, Skip, dry-run and prefix features

[Globegitter]

Since I recently learned quite a bit about tests I thought I would use this knowledge to refactor enpeem and improve the test situation a bit.

This primarily fixes #3 but also adds https://github.com/eslint/eslint, wraps now everything into Promises and simplified a few things.

Let me know what you think @mikermcneil .

[Globegitter]

Managed to miss a few things. So not ready to merge yet. Will fix up today.

[Globegitter]

This should be in a good state now and almost ready to merge. I added some of the new syntax to the README and now have also all new command additions unit and/or accpetance/integration tested.

Another thing to note, this now only supports node >=0.12.0 and io.js >= 1.0.0 since it is using native Promises.

Also for tests I am using babeljs to get the much nice ES6/7 syntax. This is done via:

require('babel/register')({
  experimental: true,
  loose: true
});

in the tests/runner.js file. I am not doing that for the actual app-code, since it costs quite a bit of performance. So there is actually one problem with the tests, since they are loaded via babel that then requires the whole app code using babel, which means when I accidentally use ES6/7 syntax that would not work on pure node.js/io.js the tests will not catch that. See Travis node 0.10 tests. They pass even though this would not work on actual node 0.10.

Now writing that all down tells me I should find a solution that babel ignores certain folders. So probably best to wait for that before merging.

[Globegitter]

Now waiting on babel/babel#1134 and can think of implementing eslint/eslint#2217 / http://eslint.org/docs/user-guide/configuring#configuration-cascading-and-hierarchy

[Globegitter]

@mikermcneil Let me know what you think about these changes - I am happy to explain to you all the changes in more details. If so I would just do the last clean-ups and it would be ready to merge.