Fix bin unit tests

[cb1kenobi]

Fixes #83.

[kriszyp]

Awesome! Thank you!

[cap10morgan]

I'm now realizing I probably should have done this as a part of my PR review, but I'm having a hard time getting a couple of these to pass locally:

1. harper/unitTests/bin/copyDB.test.js

   Line 96 in ea7d5bb

   assert(100 - (stat_after.size / stat_before_compact.size) * 100 < 10);

   This fails, and I don't really understand what it's asserting now. Subtracting from 100 would give you the percentage of compaction, right? The original code without that was (statAfter.size / statBeforeCompact.size) * 100 which would be the percentage size of the "after" compared to the "before." So now wouldn't we want the assertion to be >= 90? At any rate, I'm now getting ~87% compaction. Is that indicative of a problem, or should we make this test a little more forgiving?

2. harper/unitTests/bin/copyDB.test.js

   Lines 110 to 113 in ea7d5bb

   	assert(
   	100 - (stat_after.size / stat_before_compact.size) * 100 < 10,
   	'after size ' + stat_after.size + ' should be' + ' much less than before size ' + stat_before_compact.size
   	);

   This fails for (I think) similar reasons. And the test failure says: after size 278528 should be much less than before size 2211840

Help me @cb1kenobi, you're my only hope!

[cb1kenobi]

@cap10morgan This test was failing for me too in both repos. I figured it was a bad test. In theory the after size is less than the before size which would result in a ratio < 1.0. I assumed 10 was the percentage compaction, so you are spot on.

Personally, I would assert the after < before and call it a day. I don't know how consistent the compaction is and any compaction should prove compaction worked, right?

[cap10morgan]

@cap10morgan This test was failing for me too in both repos. I figured it was a bad test. In theory the after size is less than the before size which would result in a ratio < 1.0. I assumed 10 was the percentage compaction, so you are spot on.

Personally, I would assert the after < before and call it a day. I don't know how consistent the compaction is and any compaction should prove compaction worked, right?

Yeah, that all makes sense and I agree the simpler the better. I might do that in a future PR. I want to make the minimum changes to get tests passing and migrated into this repo first. You can see where I'm starting in #117.

[cb1kenobi]

@cap10morgan I'm still a little puzzled why the tests passed on my Mac. Maybe the cold weather?