Explore function use in `openxlsx2`

openxlsx2-functions.R

# this script counts the frequency of function use in order to create a function index that make more sense.
# We want to look at what is more user facing.
# for now the data source we use is the examples, vignettes, README etc.
# Maybe some blog posts in the wild may be used eventually.
# TODO go get functions that are used in chaining, and add back `wb_` prefix
# TODO reorder R6 methods so that internal ones are not shown first in documentation
# TODO add_thread, suggest `create_person()` or link to `wb_creators()`, `wb_person()`,
# TODO validate that `wb_dims()`, `wb_color()` are similar to 
# TODO validate vignette content. discuss if the long / complicated way to add styling is relevant in the vignette.
# TODO prefixer https://github.com/dreamRs/prefixer
library(stringr)
library(dplyr)
# exported functions of openxlsx2 
# This has the unforseen advantage of ignoring functions from other packages.
fns <- readLines(fs::path_expand("~/Documents/rrr-forks/openxlsx2/NAMESPACE"))
fns_export <- fns %>% str_subset("export") %>% str_extract("\\((.+)\\)", group = 1)

wb_wrappers <- str_subset(fns_export, "wb_") %>% str_replace('wb_', "\\\\$")

fns_regex <- c(fns_export, wb_wrappers) %>% 
  paste0("\\(", collapse = "|")

files <- fs::dir_ls(path = "~/Documents/rrr-forks/openxlsx2", recurse = T, regexp = ".R$|.Rmd", type = "file") %>% 
  fs::path_filter(regexp = "doc/|inst/", invert = TRUE)

file_content <- files %>% 
  fs::path_rel() %>% 
  purrr::set_names() %>% 
  purrr::map(\(x) readLines(x, encoding = "UTF-8")) %>% 
  purrr::map(\(x) tibble::enframe(x, name = NULL)) %>% 
  dplyr::bind_rows(.id = "file")

file_content2 <- file_content %>% 
  dplyr::mutate(
    file = fs::path(file),
    fn = str_extract_all(value, fns_regex)
  )

file_content3 <- file_content2 %>% 
  dplyr::filter(lengths(fn) != 0) %>% 
  tidyr::unnest(fn) %>% 
  mutate(
    fn = str_remove(fn, "\\($"),
    fn = str_replace(fn, "\\$", "wb_"), # replace back wrapper functions
    fn = str_replace(fn, "colour", "color"), # colour and color are the same
    fn_location = fs::path_dir(file),
    is_in_doc = str_detect(value, "^#'"),
    type = case_when(
      fn_location == "R" & is_in_doc ~ "examples",
      fn_location == "tests/testthat" ~ "tests",
      .default = fn_location
    ),
    value = NULL,
    fn_location = NULL,
    is_in_doc = NULL
  ) %>% 
  relocate(type)

file_content4 <- file_content3 %>% 
  count(fn, type, sort = TRUE) %>% 
  filter(type != "R") %>% 
  mutate(n2 = sum(n), .by = fn) %>% 
  arrange(desc(n2)) %>% 
  select(-n2)
file_content4 %>% 
  filter(type != "tests") %>% 
  summarise(n = sum(n), .by = fn) %>% 
  filter(!str_detect(fn, "xml|create|xlsx"), str_detect(fn, "wb")) %>% 
  arrange(desc(n)) %>% 
  gt::gt(groupname_col = "fn") %>% 
  gt::tab_options(row_group.as_column = TRUE)
# I see that xml_functions seem to be used a lot in internal code.
# Ignoring internal code for the sake o
# the most popular functions are `wb_workbook()`, `wb_color()`

file_content4 %>% 
  gt::gt(groupname_col = "fn") %>% 
  gt::tab_options(row_group.as_column = TRUE)