Codecov Report

Merging #2051 (86ea91a) into main (6fc5570) will increase coverage by 0.00%.
The diff coverage is 100.00%.

❗ Current head 86ea91a differs from pull request most recent head 7d075f8. Consider uploading reports for the commit 7d075f8 to get more accurate results

@@           Coverage Diff           @@
##             main    #2051   +/-   ##
=======================================
  Coverage   98.56%   98.56%           
=======================================
  Files         114      115    +1     
  Lines        5017     5032   +15     
=======================================
+ Hits         4945     4960   +15     
  Misses         72       72           

There are also data.table::uniqueN() and dplyr::n_distinct() for the similar length(unique(x)). WDYT about optionally linting this here?

I would put that as a separate linter.

The closer analogue is nested_ifelse_linter(), whose primary recommended alternatives are data.table::fcase() and dplyr::case_when().

But in that case, there are ways to avoid the nested ifelse() using base functions, so the user doesn't strictly need any new packages to fix the issue.

For length(unique(.)), there's no improvement in {base} (AFAIK).

Doesn't mean we shouldn't offer such a linter (since it'll be optional for users to turn on anyway), just by way of explaining why we didn't write that so far (it's certainly crossed my mind).