dictionary_comprehension.R

require(purrr)
require(dplyr)
require(stringr)
#' Maps over the keys and values of a vector, but expects the function to return
#' a named vector with "key" and "value" entries. These will become the names
#' and values of the final vector. Think of this as like a Python dictionary
#' comprehension: it takes an iterable and produces a dictionary.
#'
#' @param l a vector to map over
#' @param func the function to apply over that vector, which must return a `list(key=?, value=?)`
#' @param map_keys logical. If TRUE, then use `imap` which provides the keys as the second element to the function
#'
#' @examples
#' c(1,2,3) %>% kv_map(~list(key=letters[[.]], value=.))
#' c(1,2,3) %>% kv_map(function(x){ list(key=letters[[x]], value=x) })
#' c(a=1,b=2,c=3) %>% kv_map(~list(key=str_c(.y, '²'), value=.x), map_keys=T)
kv_map = function(l, func, map_keys=F){
  mapper = ifelse(map_keys, purrr::imap, purrr::map)
  mapper(l, func) %>%
    bind_rows() %>%
    pull(var='value', name='key')
}