book-store: shortened test descriptions

[rpottsoh]

Descriptions have been shortened so as to make the json nicer for automatic test generators, so test case functions do not end up being too long.

[kotp]

#470 mentioned shortening the names a few times... Looks like this was known, so happy to see it being addressed.

[rpottsoh]

@kotp it was discussed in #470 but at the time it wasn't as high of a priority as just coming up with meaningful descriptions. At the time I was not fully aware of how the canonical data was being used, I wasn't aware of automatic test generation at the time.

[Insti]

The old array formatting without all those line breaks was nicer.
Having the expected values consistently as floats was better.
It's helpful to order the tests in a progressive fashion. (Empty basket should be the first test)

The shorter descriptions are a step in the right direction, but they still don't communicate why that is a useful thing to be testing.

"Two each of first 3 books and 1 copy each of rest"

Why are we not testing "1 each of the first 3 books and 2 copies of books four and five"?

[Insti]

Try and remember to include a link to the problem description when discussing canonical-data files:
https://github.com/exercism/x-common/blob/master/exercises/book-store/description.md

[robkeim]

I know it's a matter of style, but personally I find this use of vertical space harder to read than the previous version where all of the books in a basket were on the same line.

From a functional standpoint it looks good, and I like the shortened descriptions.

[ErikSchierboom]

I'm with @robkeim here. The vertical grouping is significantly less readable to me. The descriptions themselves look good to me.

[kotp]

Much better

[ErikSchierboom]

Agreed. This looks great!