Support Windows paths

[rifraf]

This change allows me to type rubocop c:\d\e in Windows instead of c:/d/e.
In particular, if I use sublime_rubocop, this allows me to check projects and folders

[coveralls]

Coverage decreased (-0.41%) when pulling 6984e81 on rifraf:master into 0c09856 on bbatsov:master.

[bbatsov]

The supplied patch is not working correctly (see the failing tests).

[bbatsov]

Btw, @jonas054 @yujinakayama any thoughts on this? I'm wondering what's the best way to handle this, but I haven't dealt with Windows paths in ages.

[jonas054]

I'm not sure why we have a problem in the first place. Is it because we're assuming that the file separator is / in our code? Either way, if this patch, with an added if File::ALT_SEPARATOR, solves the problem on Windows, then I think all is well. I don't have a Windows environment set up to test anything, though.

[rifraf]

Yes - it is simply that (as stated in the readme) paths are / separated when entered on the command-line. A Windows user would expect to be using .
It can be argued that it is correct according to the documentation, but it took me a while to realise it. Also sublime_rubocop gets it wrong - it uses \ separators.
Thanks for the attention

[yujinakayama]

I have never dealt with Windows paths too.

My initial concerns were:

1. Is this a common way to handle Windows paths in CLI tools written in Ruby?
2. Doesn't this have a harmful effect on non-Windows platforms?

Then I've investigated a bit:

1. RSpec also uses this way. fixes #396 rspec/rspec-core#605
2. Ruby always uses File::SEPARATOR (/) as an internal path separator regardless of platform. File::ALT_SEPARATOR would have a value if the platform's file system uses another path separator (ref in ja). On Un*x, it's nil. So I think it won't harm with if File::ALT_SEPARATOR condition.

Thus, I think it's ok if we add if File::ALT_SEPARATOR.

[bbatsov]

@yujinakayama Excellent research.

@rifraf Please, update your PR according to @yujinakayama's comments and update the changelog as well.

[coveralls]

Coverage increased (+0.05%) when pulling 6028dd6 on rifraf:master into 0c09856 on bbatsov:master.

[coveralls]

Coverage decreased (-0.01%) when pulling a9fc439 on rifraf:master into ca2a23c on bbatsov:master.

[bbatsov]

Add your nickname to the list at the bottom of the file as well and squash the commits together.

[rifraf]

Sorry - I'm at the limits of my knowledge here. I wasn't really expecting all these magic tests to run (and complain) while I thought I was working quietly on a fork where no-one was watching...

I don't know how to squash the commits - can you help?
Also am I already updating the original PR or do I ship a new one.

[bbatsov]

No problem. I've squashed and merged your changes manually.

[coveralls]

Coverage decreased (-0.01%) when pulling 963262c on rifraf:master into ca2a23c on bbatsov:master.

[coveralls]

Coverage decreased (-0.01%) when pulling aaccba8 on rifraf:master into ca2a23c on bbatsov:master.

[coveralls]

Coverage decreased (-0.03%) when pulling 5b19885 on rifraf:master into ca2a23c on bbatsov:master.

[coveralls]

Coverage decreased (-0.03%) when pulling dfd6504 on rifraf:master into ca2a23c on bbatsov:master.