トップ 履歴 一覧 Farm ソース 検索 ヘルプ PDF RSS ログイン

myAllTypes.R

*disclaimer
602994

R
R.scripts
myAllTypes.R(370)

# 2018-03-18 copyleft sugiura@nagoya-u.jp

myAllTypes <- function() {
output.file = choose.files()

type.all <- ""

files <- list.files()

for (i in files) {
  type <- ""
  token <- ""


	lines.tmp <- scan(i, what="char", sep="\n")

	data.tmp <- grep("\\*(JPN|NS)...:\t", lines.tmp, value=T)

	body.tmp <- gsub("\\*(JPN|NS)...:\t", "", data.tmp)

	body.tmp <- body.tmp[body.tmp != ""]


	body.lower <- tolower(body.tmp)
	body.token <- unlist(strsplit(body.lower, "\\W"))
	body.token <- body.token[body.token != ""]

	type <- unique(body.token)
	
	type.all <- c(type.all, type)

  }
  type.all <- type.all[type.all != ""]
  type.all <- sort(unique(type.all))
  
  write(type.all, file=output.file)
  
}