exemple de scraper pour les scrutins publics de l'Assemblée nationale (attention, bug sur certaines listes avec un seul votant, car pas de tag <li> dans le code HTML)

an-votes.r

library(XML)
library(stringr)

# demo page
h = htmlParse("http://www2.assemblee-nationale.fr/scrutins/detail/(legislature)/14/(num)/947")

g = c("SRC", "UMP", "UDI", "ECOLO", "RAD", "GDR")
k = c("socialiste", "mouvement populaire", "démocrates", "écologiste", "radical", "gauche")

for(i in k) {
  
  cat("\nGROUPE :", g[ which(k == i) ], "\n")
  
  for(j in c("Pour", "Contre", "Abstention", "Non-votant")) {
    
    v = paste0("//div[@id='analyse']//p[@class='nomgroupe' and contains(., '", i)
    v = paste0(v, "')]/following-sibling::div[@class='", j, "']/ul/li")
    v = xpathSApply(h, v, xmlValue)

    if(length(v)) {
      cat("\n", j, ":", length(v), ":", paste0(str_trim(v), collapse = ", "), "\n")
    } else {
      cat("\n", j, ": 0 (apparemment)\n")
    }

  }

}

# kthxbye