refactor step_regex() and step_count()

[EmilHvitfeldt]

Related to #1156.

This PR does a small refactor to step_regex() and step_count() to bring them up to the same standard as #1156.

Further, the steps new use check_name(), and will thus error in result is set to a name that is in the data set. Previously it would silently overwrite.

[simonpcouch]

This is great!

A paper trail for future passersby: I was missing a bit of context while trying to understand how many of these lines transition eval_select() output object$input to names(object$input). I initially had trouble wrapping my head around how that was fully safe (e.g. depending on object$input being c(Species = "Species")), and this seemed to suggest to me not:

recipes/R/selections.R

Lines 204 to 208 in 59486ec

	# Return names not positions, as these names are
	# used for both the training and test set and their positions
	# may have changed. If renaming is allowed, add the new names.
	out <- names(data)[sel]
	names <- names(sel)

This is fine for both step_count() and step_regex(), though, because both leave allow_rename = FALSE as default in recipes_eval_select() and don't surface that argument to the user.🐛

[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.