An R function I wrote in the process of translating a list kept in a .docx file to a proper R dataframe. After moving it from .docx to .txt via pandoc, I needed to turn section headers into levels of a grouping factor

fill_down.R

#' convert positional information to two columns
#' 
#' Sometimes text is organized by position. This function
#' turns positional group labels (e.g headers ) into the levels of a grouping variable
#' @param x character vector containing group labels followed by group members
#' @param pattern regular expression that identifies the group labels
fill_down <- function(x, pattern){
  ## find matches of the pattern 
  x <- as.character(x)
  value_matches <- grepl(pattern = pattern, x = x)
  ## get their positions
  loc <- which(value_matches)
  ## the first of these should be labelled
  stopifnot(min(loc) == 1)
  
  start <- loc
  end <- c(loc[-1], length(x) + 1)
  
  # measure the intervals between labels
  intervals <- end - start - 1
  
  rps <- Map(f = rep_len, x[loc], intervals)
  
  ## combine replicated values in a single vector"
  grps <- do.call(c, rps)
  
  ## get the values between labels
  xvals <- x[!value_matches]
  
  stopifnot(length(grps) == length(xvals))
  
  dplyr::data_frame(grps, xvals)
}

# imagine a list of letters divided into two sections: "A" and "B":

test <- c("A", "b", "c", "e", "B", "g", "h", "i")
fill_down(test, "A|B")

# 
# grps xvals
# 1    A     b
# 2    A     c
# 3    A     e
# 4    B     g
# 5    B     h
# 6    B     i