R/snake-case.R

In OuhscBbmc/OuhscMunge: Data Manipulation Operations

#' @name snake_case
#' @export
#'
#' @title Convert variable names to snake_case
#'
#' @description This function attempts to convert variables to snake_case, even if it's already in snake_case.
#' The important regex lines were posted by Stack Overflow user [epost](http://stackoverflow.com/users/129879/epost)
#' in ["Elegant Python function to convert CamelCase to snake_case?"](http://stackoverflow.com/a/1176023/1082435).
#'
#' @param x A vector of names to convert.
#'
#' @return A vector of converted names.
#'
#' @note This series of regexes has an advantages over the current
#' implementations of [`lettercase::str_snake_case()`](https://github.com/decisionpatterns/lettercase)
#' and [`snakecase::to_snake_case()`](https://tazinho.github.io/snakecase/reference/caseconverter.html).
#' The former converts "PatientDOB" to "patientdob" and the latter converts
#' "patient.dob" to "patient_._dob".  I'll keep an eye on these packages
#' (*i.e.*, [lettercase #1](https://github.com/decisionpatterns/lettercase/issues/1) for 'camelCase'
#' and [snakecase #101](https://github.com/Tazinho/snakecase/issues/101)).  I'd prefer
#' to use one of them, instead of maintaining the functions.
#'
#' @author Will Beasley
#'
#' @examples
#' snake_case(colnames(datasets::OrchardSprays))
#' snake_case(colnames(datasets::iris))
#' snake_case(c("PatientID", "PatientDOB", "DOB", "name.last", "name.first"))

snake_case <- function(x) {
  # Third & fourth lines use http://stackoverflow.com/questions/1175208/elegant-python-function-to-convert-camelcase-to-snake-case
  s <- gsub("'"                         , ""        , x, perl = TRUE)     # Remove apostrophes to preserve contractions.
  s <- gsub("\\W"                       , "_"       , s, perl = TRUE)     # Replace any nondigit/noncharacter/nonunderscore with an underscore.
  s <- gsub("(.)([A-Z][a-z]+)"          , "\\1_\\2" , s, perl = TRUE)     # Separate w/ underscores based on capitalization
  s <- tolower(gsub("([a-z0-9])([A-Z])" , "\\1_\\2" , s, perl = TRUE))
  s <- gsub("_+"                        , "_"       , s, perl = TRUE)     # Replace consecutive underscores with single.
  s <- gsub("^_|_$"                     , ""        , s, perl = TRUE)     # remove leading and trailing underscores
  # s <- gsub("\\."                     , "_"       , s, perl = TRUE)     # Replace dots with underscores.
  return(s)
}