In CarlBoneri/roxydoc2: Helps document R packages properly.

knitr::opts_chunk$set(echo = TRUE)

roxydoc2 is intended as a set of useful global tools for R package authors to use when developing packages. It provides a series of wrapper and utility functions aimed at supplementing the devtools and roxygen2 packages.

Implementations

1) Set global author variables for package docs 2) Allow for flexible code creation, at the console or in a script with the focus of documentation being the last thought 3) Properly format all latext and markdown output strings 4) Detailed data and output descriptions from actual code outputs 5) Addition of any copyright notices, or boilerplate to each script 6) Organization of code even when the author fights it... 7) Work at the console as much as you'd like, with script compiling running in the background

Setting the global variables for the author

o <- lapply(list.files("R", pattern = "docr_", full.names=TRUE), source)
docr.defaults(author = "Joe Shmow", 
              email = "[e-mail]", 
              company = "Joe's cRab shack", 
              website = "https://joeknowsr.com", 
              github = "JoeShmow")

This will then store the data to the .GlobalEnv as

knitr::kable(docr.get_defaults())

So now when docr.author is called:

docr.author()

Documenting datasets

The primary purpose of R is generally to create, or manipulate raw data. Often it is time consuming and tedious to properly format what the dataset looks like, or contains. These functions aim to ease the burdon.

dataframes

raw <- capture.output(docr.describe_df(df_in = mtcars, include_slice = TRUE))
cat(raw, sep = "\n")

lists

raw2 <- capture.output(docr.list_named(as.list(sapply(as.list(mtcars),`[[`,1))))
cat(raw2, sep = "\n")

And to add boilerplate to all files:

cat(docr.meta_private(),sep = "\n")

And writing code at the console and prepping for doc/scripts:

docr.clean_fun <- function(fun_name = NULL, trim_blanks = FALSE,
                           rm_comments = FALSE, do_cat = TRUE){
  deparse(get(fun_name), control = "all") %>% (function(x){
    x[[1]] <- sprintf("%s <- %s",fun_name, x[[1]])

    if(trim_blanks){
      x <- x[nchar(x) > 0]
    }

    if(rm_comments){
      x <- grep("^[[:space:]]{1,}\\#", x, invert = TRUE, value = TRUE)
    }
    if(do_cat){
      cat(x, sep = "\n")
    }else {
      x
    }
  })
}
this.fun <- function(x = 1, y = 2,...){
    x * y
}
Fun <- docr.clean_fun('this.fun',do_cat = FALSE)
Args <- docr.param('this.fun')
Auth <- capture.output(docr.author())
Meta <- docr.meta_private()
cat(c(Meta,sprintf("#' %s",docr.fun('this.fun')),
      Args[[1]], sprintf("#' %s\n", Args[2:length(Args)]), Auth, Fun),sep = "\n")

CarlBoneri/roxydoc2 documentation built on Nov. 6, 2019, 8:01 a.m.