Add rudimentary tests for ioreg

[bgamari]

No description provided.

[bharrisau]

Preferring //! over /*!. Not sure if that is a Zinc style thing, but /* */ is falling out of favour anyway.

[farcaller]

Yup, that's also what the rust style suggests.

[bgamari]

@bharrisau @farcaller any thoughts on the above? As I said, my weak opinion is that we should just keep things simple (yet still quite readable, IMHO). However, if others feel strongly about using Shiny/Hamcrest I'm willing to go that route.

Either way, I would like to start clearing out my patch queue and move on to other things.

[farcaller]

It may be reasonable to convert hamcrest's expect to expect!, that will take line!() into account. Overall I prefer hamcrest tests as they do add to condition readability as compared to plain assertions (or, uh, I've written to much test code in rspec).

I'm not enforcing both libs, but my two points are:

• there should be one way we do tests across zinc;
• shiny/hamcrest really worked out nice in os tests, and I'm cleaning up PT tests to use shiny as well and so far I like the result.

I'll approve this one PR no matter what you do use for tests, but let's get into ML to discuss if we're going to use shiny/hamcrest everywhere until I started doing that on my own. And if we end up using the libs, the code will be refactored.

PS: expect some latency in my responses, out on vacation ☀️

[bgamari]

r?

[bharrisau]

I'm still away, so I've only given it a quick scan - but it looks good to me.

Looks like Travis is failing on an ioreg related issue. Can you double check that then r=me.