Download kommunalvalgsdata fra kmdvalg.dk.

KV2013

library(XML)
# main page
url <- "http://kmdvalg.dk/main"

# read links
url.data <- readLines(url)
doc <- htmlTreeParse(url.data,  useInternalNodes = TRUE)

#grab urls
urls <- unlist(xpathApply(doc, "//div[@class='LetterGroup']//a[@href]", 
                          xmlGetAttr, "href"))

# function
get.data <- function(url){
  # read url
  cat("url:", url)
  url.data <- readLines(url)
  tables <- htmlTreeParse(url.data,  useInternalNodes = TRUE, encoding = "UTF-8")
  
  #grab data
  st.nu <- xpathSApply(tables, "//td[@class='StemmerNu']", xmlValue)
  st.sidst <- xpathSApply(tables, "//td[@class='StemmerSidst']", xmlValue)
  navn <- xpathSApply(tables, "//td[@class='kandidat']", xmlValue)
  kom <- xpathSApply(tables, "//title", xmlValue)
  
  # clean
  st.nu <- st.nu[c(rep(TRUE,1),FALSE)]
  st.sidst <- st.sidst[c(rep(TRUE,1),FALSE)]
  navn <- navn[c(rep(FALSE,1),TRUE)]
  # clean some more 
  st.nu <- gsub("\\.","", st.nu)
  st.sidst <- gsub("\\.","", st.sidst)
  st.sidst <- gsub("\\(|\\)", "", st.sidst)
  kom <- gsub("Kommunalvalg ","", kom)
  kom <- rep(kom, length(st.nu))
  return(cbind(navn,st.nu,st.sidst,kom))
}

data <- list()

for (i in seq_along(urls)){
  # extracting information
  data[[i]] <- get.data(urls[i])
  # waiting one second between hits
  Sys.sleep(1)
  cat(" done!\n")
}

data <- data.frame(do.call(rbind, data), stringsAsFactors=F)
colnames(data) <- c("kandidatlister","stemmetal","ændring","kommune")