Improve suggested packages handling + testthat code update

[olivroy]

Summary

This PR aims to take advantage of cli inline markup and rlang utils functions (i.e. rlang::check_installed() as well as using dplyr::tibble() in the code and documentation.
Improving syntax of some error messages and improve speed of tests by using expect_named/length/match/in()

The failing tests on my Windows machine are related to file creation /saving.

Checklist

• I understand and agree to the Code of Conduct.
• I have listed any major changes in the NEWS. (No major change)
• I have added testthat unit tests to tests/testthat for any new functionality. N/A

Edit: sorry for the large PR. At first, I wanted to keep it small, but ended up making changes in all places I thought were relevant for consistency and to avoid many follow-up PRs)

[CLAassistant]

All committers have signed the CLA.

[rich-iannone]

Sorry for not commenting sooner but I do like all the work you've put into this!

I have just one change request at the moment, and addressing that would get this PR closer to merging/completion. Would it be possible to revert all of the tibble::tibble changes back to dplyr::tibble? I know this doesn't make much sense, but I'm hoping to remove the tibble dependency at a later date (I use only a few tibble functions) and I know that dplyr will always re-export the tibble() function.

[olivroy]

@rich-iannone Thank you!

I have added a few more package handling changes + removed tibble::tibble in tests. + modernized a little the test suite.

Edit: I have looked into replacing tibble::enframe() with dplyr::tibble() but it was 10x slower (30microseconds vs 300 microseconds)

Edit 2: I also took a look at the tests. I know it is unrelated to the messages, but it helps to reduce the clutter a little. Bonus, improves speed a little.

library(magrittr)
library(testthat)
#> 
#> Attachement du package : 'testthat'
#> Les objets suivants sont masqués depuis 'package:magrittr':
#> 
#>     equals, is_less_than, not
bench::mark(
  now = letters %>% expect_no_error(),
  before = letters %>% expect_error(regexp = NA),
  check = FALSE
)
#> # A tibble: 2 × 6
#>   expression      min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 now         191.8µs  201.4µs    4585.   246.02KB     10.6
#> 2 before       42.5ms   43.6ms      22.1    4.43MB     12.7
bench::mark(
  now = letters %>% expect_length(26),
  before = letters %>% length() %>% expect_equal(26),
  check = FALSE
)
#> # A tibble: 2 × 6
#>   expression      min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 now           126µs  133.4µs    6709.     56.9KB     10.3
#> 2 before         85ms   86.8ms      11.5   583.4KB     23.1
nms <- names(mtcars)
bench::mark(
  now = mtcars %>% expect_named(nms),
  before = mtcars %>% names() %>% expect_equal(nms),
  check = FALSE
)
#> # A tibble: 2 × 6
#>   expression      min   median `itr/sec` mem_alloc `gc/sec`
#>   <bch:expr> <bch:tm> <bch:tm>     <dbl> <bch:byt>    <dbl>
#> 1 now         140.5µs    149µs   6177.      27.3KB     10.4
#> 2 before       87.6ms    108ms      9.30   427.9KB     13.9

^{Created on 2023-09-23 with reprex v2.0.2}

[rich-iannone]

Thank you for being so diligent with the test changes (and checking the difference in timings). This helps a lot! Let me know when you want me to review and merge; I don’t want to do it too early, as you might have a few more commits to go until you consider it finalized.

[olivroy]

@rich-iannone Ready for your review now.

[rich-iannone]

Thanks again for the huge amount of work done here! I made only a few minor changes but 99.5% of this was perfect.