Added better handling for empty tables

[miversen33]

Addresses #146

Note, I don't really know how to write a test to properly address the change in logic applied here. However I did run make and all tests passed.

That said, its worth considering if this is how you want to address this issue or not :)

[olimorris]

This is a great spot and exactly what is happening. Thank you.

Can you add your test from #146 as well?

[miversen33]

This is a great spot and exactly what is happening. Thank you.

Can you add your test from #146 as well?

Technically the test I provide there fails as the logic in that test is ever so slightly different from what I am actually doing. That said, I will see if I toss something akin to it in here for you :)

[miversen33]

@olimorris I have added a modified version of the test I recommended earlier.

That said, I really don't like it. The logic added is actually in init not dirs_match, but there is no test (from what I saw) for init.allowed_dir so what I added is my best shot at it.

The test is not really clear on what its doing so I added some comments. Tests shouldn't need comments though :(

It works but I would advise you peek at it and make it how you want