Improved tpl_get, based on tpl_get in the taxize package (http://cran.r-project.org/web/packages/taxize/index.html). Downloads plant species records (with accepted names) indexed by theplantlist.org.

tpl_get2.R

tpl_get2 <- function (dir_, family = NULL) 
{
  require(RCurl)
  require(XML)
  require(plyr)
  temp <-getURL('http://www.theplantlist.org/browse/-/')
  temp <- htmlParse(temp)
  families <- xpathSApply(temp, "//ul[@id='nametree']//a", xmlValue)
  csvlinks <- sprintf('http://www.theplantlist.org%s%s.csv', 
                      xpathSApply(temp, "//ul[@id='nametree']//a", xmlGetAttr, 'href'), 
                      families)
  if (all(!family %in% families)) {
    stop(paste('Requested families not found on TPL.',
               'Use tpl_families() to list plant families indexed by TPL.'),
         call.=FALSE)
  }
  if (any(!family %in% families)) {
    warning(sprintf('Requested families not found on TPL: %s.\n%s', 
                    paste(family[!family %in% families], collapse=', '),
                    'Use tpl_families() to list plant families indexed by TPL.'),
            call.=FALSE)
  }
  if (!is.null(family)) {
    csvlinks <- csvlinks[families %in% family]
    families <- families[families %in% family]
  }
  getcsv <- function(x) {
    download.file(x, destfile=file.path(dir_, basename(x)), quiet=TRUE)
  } 
  message("Downloading csv files to ", dir_, "...")
  dir.create(dir_)
  l_ply(csvlinks, getcsv, .progress = "text")
  message("...el fin")
}