expand format tests to ensure double formatting is sane

[a-frantz]

Formatting the output of format a second time should produce identical results to the first format. Formatting is of course deterministic, so this isn't about testing the same input twice, but rather that supplying the output of format as an input to a subsequent format doesn't make any changes.

[Scott8440]

I can take this one if not already being worked on

[a-frantz]

Awesome! It's all yours :)