Make CancellationToken optional from tests

[kzu]

Since CancellationToken.None is equivalent to default(CancellationToken)
according to https://github.com/dotnet/corefx/issues/495#issuecomment-72322134,
and since the latter can further be shortened to just default, make use of
both to cleanup a bunch of method calls that all ended up using just
CancellationToken.None.

[sharwell]

This is currently a style matter for this repository, along with the use of ConfigureAwait(false) in tests. It's come up a few times before, e.g. #518, #1440. It's mostly about the desire to eventually have xunit/xunit#345 so user cancellation is a test input that's always available.

[sharwell]

It's possible this style changes in the future, but at this time I'm not ready to apply it. If it changes, the cancellation token would be omitted from signatures altogether as opposed to making it optional.

[kzu]

Hm... the reason for not wanting to make it optional, however, has nothing to do with what I submitted here, since the optional cancellation made optional is not in any test-defining method.

[kzu]

I know this is just a matter of taste, but lines were already long enough that I was scrolling just to find out that it was all just CancellationToken.None and ConfigureAwait(false). Makes the code less readable and (at least in the former case) it's not needed at all. Not sure about the latter in the context of a unit test either.... My guess is it's not needed there either, but I may be wrong...

[kzu]

In any case, getting a PR rejected in 3' after all that regex-fu, is a bit discouraging... Anyway, on to what I was really doing now... ¯\(ツ)/¯

[sharwell]

In any case, getting a PR rejected in 3' after all that regex-fu, is a bit discouraging.

Yeah, I knew that was going to be the case, but I figured it was better to close it early than leave it sitting here to never go anywhere. ☹️

[kzu]

What about your comment that you didn't want it because of the issue with discovering/instantiating/invoking/whatever unit tests when optional parameters where part of the signature? Which is clearly different to what this PR does?

The xunit/xunit#345 issue isn't implemented and even if it were, this PR doesn't prevent a specific test to provide a cancellation if needed/wanted, there's just a default in case you don't care, which is currently the case in all the removed code. So, it just removes all the useless boilerplate all over that's currently not adding any value... So I fail to see how it limits any such future feature.