Update kata to Node 12. Added sample tests. Updated the test suite with useful messaging.

There are no random tests -- and never were.

Added useful information.

Replaced tests with chai / mocha and descriptive messages.

Added.

Missing a closing paren on last test case.

Fair enough. That could have been a typo. I've made that change. Sorry about the confusion.

I don't think it's a bad thing within the confines of the problem since it was clearly defined as a limit.

Without that limit it becomes unwieldy and a pain to ensure that it remains accurate if it's in an active code base.

Breaks for 5003.

I can get rid of test case #4.

This kata was based on a really terrible question I answered on StackOverflow. What the asker was looking for did not really make any sense from a functional standpoint, but I thought the solution was easy enough and interesting enough to add.

What if the max is 00000?

Good stuff. This is how I interpreted the task as well.

I thought the point of this one was to accomplish the task with a working closure?