Refactor tests to use snapshots and abort_bad_argument()

[olivroy]

Here are a couple extra suggestions when trying to improve error messages.

I highly recommend using snapshot tests, as you can quickly verify that your output changed how you want it.

It is better to add a call argument to abort_bad_argument() to make testing easier

expect_snapshot(error = TRUE) is a lifesaver https://testthat.r-lib.org/articles/snapshotting.html

Also discovered that abort() / cli_abort() have a dedicated argument to take care of extra info: footer, so renamed it to this (and if custom, don't show footer)

custom requires must to be optional.

Edit: made a mistake in the first place that I easily caught with snapshot tests! e8d8868 It is easier to experiment this way

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 100.00%. Comparing base (66a84ec) to head (e8d8868).

Additional details and impacted files

@@            Coverage Diff            @@
##              main       #50   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            6         6           
  Lines          287       286    -1     
=========================================
- Hits           287       286    -1     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[wjakethompson]

This great--thank you!!