coord-munch.r

#' For munching, only grobs are lines and polygons: everything else is 
#' transfomed into those special cases by the geom.
#' @examples
#' nz <- data.frame(map("nz", plot=FALSE)[c("x","y")])
#' munch_data(nz, segment_length = 0.1)
munch_data <- function(data, segment_length) {
  data <- add_group(data)
  n <- nrow(data)
  aes_df <- data[setdiff(names(data), c("x", "y"))]

  # Distance between points - NA indicates a break / terminal points
  dist <- with(data, sqrt((x[-n] - x[-1]) ^ 2 + (y[-n] - y[-1]) ^ 2))
  dist[data$group[-1] != data$group[-n]] <- NA
  
  # How many extra new  to create per old segment
  extra <- floor(dist / segment_length)
 
  # The hard part: figure out how to do the interpolation in a vectorised 

  # start and end, and number of pieces between them
  # start + seq(0, 1, length = n + 1)[-n] * end 
  
  data.frame(
    x = x, y = y,
    aes_df[rep(seq_len(nrow(aes_df)), by = expand)]
  )
}