Refactor documentation and test nomenclature

[olivroy]

Sorry.. this got out of hand again. Please let me know if you agree with this. If not, I can send smaller separate PRs of the relevant bits!

TL;DR: The goal is basically to simplify the x() function to x() where applicable as recommended in the documentation tidyverse style guide. (only the function bullet)

This PR started as a follow-up to #1683 as I wanted to make clarifications and improved some things I did there after taking a fresh look.
cbb1dca and 018b232 do this. I finally discovered the way to avoid spurious (snapshot git diffs while working on Windows: Set Posix line ending in Rproj (13b2666). Used https://www.tidyverse.org/blog/2020/03/roxygen2-7-1-0/ as ref too.

I'd like to get your views on how I edited the shiny related functions docs 64e80b3. (I removed details that felt redundant? Let me know if you agree. Otherwise, I will rework it!

Changes

• Renamed is_non_empty_string() to is_non_empty_chr() for clarity (accepts length > 1) e0c4a10 (string is usually for character + length = 1)

• renamed tests to use the convention ab0ed09

• Moved some test helper functions into own test helper files to improve testing workflow (i.e. functions are loaded with load_all() af9cdd6

• Removed duplicate definition of test helper in tests/testthat/helper.R bcbb80d

• Removed the check for xml2 installation since it is import now (I may do a follow-up where I remove check_suggests() with appropriate skip_if_not_installed() bcbb80d

• Did minor test refactoring where I saw it useful bcbb80d

• simplified some dplyr expressions in examples d2aa113, bc70cdb

I verified that they are all correct with this code

v1 <-  towny |>
     dplyr::select(name, csd_type, census_div, population_2021) |>
     dplyr::group_by(census_div) |>
     dplyr::summarize(
       population = sum(population_2021),
       .groups = "drop_last"
     ) |>
     dplyr::arrange(population) |>
     dplyr::slice_head(n = 5) |>
     dplyr::mutate(ranking = dplyr::row_number()) |>
     dplyr::select(ranking, dplyr::everything()) 
v2 <- towny |>
  dplyr::select(name, csd_type, census_div, population_2021) |>
  dplyr::group_by(census_div) |>
  dplyr::summarize(
    population = sum(population_2021),
    .groups = "drop_last"
  ) |>
  dplyr::slice_min(population, n = 5) |>
  dplyr::mutate(ranking = dplyr::row_number(), .before = 0) 

waldo::compare(
  v1, v2
)

Other considerations:

When I noticed it, avoid self-reference, and 2 cross-refs to the same function in the same paragrah.
The diff is large because I reflowed roxygen comments to be about 80 characters long.

[rich-iannone]

This is monumental work! I am super impressed so far (and it's okay that this touches so many files, it has to).

[olivroy]

Thanks for your interest! I pushed another chunk of changes to go around rough edges.

[rich-iannone]

LGTM!