Use snapshot tests

[hfrick]

This PR moves tests around errors and warnings to our current standard:

• expect_error(regexp = NA) is now expect_no_error() (and not expect_no_condition() due to a <dplyr_regroup> condition like in expect_error(regexp = NA) -> expect_no_condition() parsnip#1188 (comment))
• expect_error() otherwise is now expect_snapshot(error = TRUE)
• expect_warning() is now expect_snapshot()
• expect_snapshot_error() and expect_snapshot_warning() are now expect_snapshot() (with and without error = TRUE, respectively) to add the context to the snapshot

I didn't see any expect_message() or expect_condition().

I'll open separate issues for anything in the snapshots that needs fixing after we merge this because then I can link to the snapshot more easily.

[hfrick]

Any suggestions to solve #539 welcome! I think I'll start with making tests self-contained because looks like it's about when purrr is loaded and I think I recall a comment about code in tests/testthat/ scripts outside of test_that() calls potentially having side-effects like this.

[simonpcouch]

Awesome—will defer to your judgment on which snaps deserve attention. :) Just a couple Find + Replace Wonk suggestions.

This PR helped me deal with my feelings about:

  expect_snapshot(error = TRUE, {
    int_pctl(bt_small, id)
  })

vs.

  expect_snapshot(error = TRUE, int_pctl(bt_small, id))

vs.

expect_snapshot(
  error = TRUE,
  int_pctl(bt_small, id)
)

So thank you for that😝

[hfrick]

Thanks for catching the indentation bits! 🙌

[github-actions]

This pull request has been automatically locked. If you believe you have found a related problem, please file a new issue (with a reprex: https://reprex.tidyverse.org) and link to this issue.