String handling in R is hard, there are a few packages that help making that task not so hard, but it is still hard. In my opinion string handling in R will never be as good as languages like Python or Perl, it looks like the language hasn't been designed to face that problem. But sincerely speaking, I don't think R realy need that, I can pass without it, I can go fairly well with the packages available, and the most popular is stringr developed by Hadley Wickham. It has several good functions for string handling, but it lacks string interpolation like we do in Python. That's the reason why I developed str_supplant:

str_supplant <- function (string, repl) {
    result <- str_match_all(string, "\\{([^{}]*)\\}")
    if (length(result[[1]]) == 0)
    result <- result[[1]]
    for (i in seq_len(dim(result)[1])) {
        x <- result[i,]
        pattern <- x[1]
        key <- x[2]
        if (!is.null(repl[[key]]))
            string <- gsub(pattern, repl[[key]], string, perl=TRUE)

Here it follows an example:

parms <- list(media='', face='wilson')
str_supplant("{media}/logo.gif, {media}/img/ {face}", parms)
## [1] ", wilson"

The first argument is the string to be interpolated and the second a list with the keys to be used. If a key doesn't match it isn't interpolated.

parms <- list()
str_supplant("My name is {name}", parms)
## [1] "My name is {name}"

That implementation was deeply inspired by Douglas Crockford's supplant implementation. Unfortunately that implementation hasn't been vectorized yet, but I think it is quite useful anyway. I hope it help others like it's been helping me.