Add ":any" as a hook scope alias for ":all"

[fj]

As per a discussion on the mailing list, David suggested it might be useful to alias ":all" to ":any" for before/after hooks. Here is a patch that does that.

[dchelimsky]

Here's the thread: http://groups.google.com/group/rspec/browse_thread/thread/26ea96043f46b3b6

Rick DeNatale pointed out that :any would just be confusing in an after hook: after(:any), which leads me to think we should probably not do this as it would just add confusion. I'm going to leave this open for a bit to see if anyone else wants to voice an opinion.

[fj]

Not to shoot my own patch in the foot, but my personal opinion is to have only one way to do it. I think whatever ambiguity there may be in before(:all) isn't adequately compensated by the additional confusion of having before(:any), which sounds like it would do something subtly different.

[markiz]

Why is :any aliased to :all instead of :each? FWIW, I think the difference is clear now and won't be so clear with introduction of this alias.

[fj]

@markiz: Because that's what the discussion was about. It was triggered by this reasoning: "before(:each) runs before each spec runs, while before(:all) runs before any spec runs". My vote is to leave it the way it is, but I wrote the patch to show what it might look like if we did have a better reason for aliases in the future.

[dchelimsky]

On the rspec-users mailing list, Evgeniy Dolzhenko suggested:

before(:suite)
before(:group)
before(:example)

If we're going to change the existing names at all, I'm a big fan of this. The names work very well together, with both before and after. The default would be :example. And these don't really conflict/confuse with the existing language - it's just an alternative that we would support and promote as the default over time.

Thoughts?

[fj]

What's the difference between :suite and :group?

[dchelimsky]

before(:suite) and after(:suite) already exist, and are the first and last hooks to run, outside of any example groups.

[fj]

I have made a new pull request which supersedes this one to use Evgeniy Dolzhenko's suggestions. You can see it here.

In the meantime, I will close this pull request.