Refactor getBrowserNameByOS

[Xercoy]

getBrowserNameByOS doesn't have unit tests, so I decided to add one.

Right now, it's basically a subtest that runs based on the value of runtime.GOOS since the way getBrowserNameByOS is written is only meant to handle cases where we're dealing with a darwin based OS. Not only does the unit test do that, but it tests every possible denormalized string value.

I was thinking that since the function only handles cases for a darwin OS, we can ensure that any other value for runtime.GOOS will result in a return value of a null string. While it does adhere to the current behavior, I feel like it could be overkill.

The code for it also seemed like it could have been simplified a bit for a few reasons:

• variable name 'k' for the sole string argument didn't seem like it provided any context as to what the argument was
• doubly nested switch, case blocks for basic logic just seemed like it could be simplified.

To ensure that my changes to the function work as intended, I created the test, ran the entire test suite and ensured it passed, modified getBrowserNameByOS, then ran tests a final time to ensure its behavior remained unchanged.

[Bunchhieng]

Looks good. Feel free to modify the checkOS method to support Windows or Linux in this PR or I can add a follow-up issue to address it. Thanks

[Bunchhieng]

This section is mainly for the dependencies I use on the project. Your name will show up in the release. 👍

[Xercoy]

Ah, definitely. I'll add a credit to stretchr maybe?

[Bunchhieng]

I was thinking if we can refactor this to support Windows and Linux.

[Xercoy]

That sounds good - would the denormalized values be anything different?

[Bunchhieng]

I add this issue. there is a list of all the OS name. the browser string may be vary between each OS.
#21

[Bunchhieng]

Would you be able to fix the merge conflict? @Xercoy

[Xercoy]

@Bunchhieng I'll work on it and ping you when it's done. Apologies for the delay.

[Xercoy]

closing in favor of #50