Nothing
R/loadEncodingTableVN.R
In vietnameseConverter: Convert Vietnamese Encodings
Defines functions
loadEncodingTableVN
Documented in
loadEncodingTableVN
#' Load conversion table for Vietnamese characters
#' @param version Version of this table (1 or 2) - 2 is currentlty being used
#' @return data.frame
#' @details
#' The table was adapted from \url{http://vietunicode.sourceforge.net/charset/}
#' @export
#'
loadEncodingTableVN <- function(version)
{
if(version == 1){
out <- structure(list(Unicode_Hex = c("U+00C0",
"U+00C1", "U+00C2", "U+00C3", "U+00C8", "U+00C9", "U+00CA", "U+00CC",
"U+00CD", "U+00D2", "U+00D3", "U+00D4", "U+00D5", "U+00D9", "U+00DA",
"U+00DD", "U+00E0", "U+00E1", "U+00E2", "U+00E3", "U+00E8", "U+00E9",
"U+00EA", "U+00EC", "U+00ED", "U+00F2", "U+00F3", "U+00F4", "U+00F5",
"U+00F9", "U+00FA", "U+00FD", "U+0102", "U+0103", "U+0110", "U+0111",
"U+0128", "U+0129", "U+0168", "U+0169", "U+01A0", "U+01A1", "U+01AF",
"U+01B0", "U+1EA0", "U+1EA1", "U+1EA2", "U+1EA3", "U+1EA4", "U+1EA5",
"U+1EA6", "U+1EA7", "U+1EA8", "U+1EA9", "U+1EAA", "U+1EAB", "U+1EAC",
"U+1EAD", "U+1EAE", "U+1EAF", "U+1EB0", "U+1EB1", "U+1EB2", "U+1EB3",
"U+1EB4", "U+1EB5", "U+1EB6", "U+1EB7", "U+1EB8", "U+1EB9", "U+1EBA",
"U+1EBB", "U+1EBC", "U+1EBD", "U+1EBE", "U+1EBF", "U+1EC0", "U+1EC1",
"U+1EC2", "U+1EC3", "U+1EC4", "U+1EC5", "U+1EC6", "U+1EC7", "U+1EC8",
"U+1EC9", "U+1ECA", "U+1ECB", "U+1ECC", "U+1ECD", "U+1ECE", "U+1ECF",
"U+1ED0", "U+1ED1", "U+1ED2", "U+1ED3", "U+1ED4", "U+1ED5", "U+1ED6",
"U+1ED7", "U+1ED8", "U+1ED9", "U+1EDA", "U+1EDB", "U+1EDC", "U+1EDD",
"U+1EDE", "U+1EDF", "U+1EE0", "U+1EE1", "U+1EE2", "U+1EE3", "U+1EE4",
"U+1EE5", "U+1EE6", "U+1EE7", "U+1EE8", "U+1EE9", "U+1EEA", "U+1EEB",
"U+1EEC", "U+1EED", "U+1EEE", "U+1EEF", "U+1EF0", "U+1EF1", "U+1EF2",
"U+1EF3", "U+1EF4", "U+1EF5", "U+1EF6", "U+1EF7", "U+1EF8", "U+1EF9"
), VNI_Hex = c("41 D8", "41 D9", "41 C2", "41 D5", "45 D8", "45 D9",
"45 C2", "CC", "CD", "4F D8", "4F D9", "4F C2", "4F D5", "55 D8",
"55 D9", "59 D9", "61 F8", "61 F9", "61 E2", "61 F5", "65 F8",
"65 F9", "65 E2", "EC", "ED", "6F F8", "6F F9", "6F E2", "6F F5",
"75 F8", "75 F9", "79 F9", "41 CA", "61 EA", "D1", "F1", "D3",
"F3", "55 D5", "75 F5", "D4", "F4", "D6", "F6", "41 CF", "61 EF",
"41 DB", "61 FB", "41 C1", "61 E1", "41 C0", "61 E0", "41 C5",
"61 E5", "41 C3", "61 E3", "41 C4", "61 E4", "41 C9", "61 E9",
"41 C8", "61 E8", "41 DA", "61 FA", "41 DC", "61 FC", "41 CB",
"61 EB", "45 CF", "65 EF", "45 DB", "65 FB", "45 D5", "65 F5",
"45 C1", "65 E1", "45 C0", "65 E0", "45 C5", "65 E5", "45 C3",
"65 E3", "45 C4", "65 E4", "C6", "E6", "D2", "F2", "4F CF", "6F EF",
"4F DB", "6F FB", "4F C1", "6F E1", "4F C0", "6F E0", "4F C5",
"6F E5", "4F C3", "6F E3", "4F C4", "6F E4", "D4 D9", "F4 F9",
"D4 D8", "F4 F8", "D4 DB", "F4 FB", "D4 D5", "F4 F5", "D4 CF",
"F4 EF", "55 CF", "75 EF", "55 DB", "75 FB", "D6 D9", "F6 F9",
"D6 D8", "F6 F8", "D6 DB", "F6 FB", "D6 D5", "F6 F5", "D6 CF",
"F6 EF", "59 D8", "79 F8", "CE", "EE", "59 DB", "79 FB", "59 D5",
"79 F5"), VPS_Hex = c("80", "C1", "C2", "82", "D7", "C9", "CA",
"B5", "B4", "BC", "B9", "D4", "BE", "A8", "DA", "DD", "E0", "E1",
"E2", "E3", "E8", "E9", "EA", "EC", "ED", "F2", "F3", "F4", "F5",
"F9", "FA", "9A", "88", "E6", "F1", "C7", "B8", "EF", "AC", "DB",
"F7", "D6", "D0", "DC", "2", "E5", "81", "E4", "83", "C3", "84",
"C0", "85", "C4", "1C", "C5", "3", "C6", "8D", "A1", "8E", "A2",
"8F", "A3", "F0", "A4", "4", "A5", "5", "CB", "DE", "C8", "FE",
"EB", "90", "89", "93", "8A", "94", "8B", "95", "CD", "6", "8C",
"B7", "CC", "10", "CE", "11", "86", "BD", "D5", "96", "D3", "97",
"D2", "98", "B0", "99", "87", "12", "B6", "9D", "A7", "9E", "A9",
"9F", "AA", "A6", "AB", "13", "AE", "14", "F8", "D1", "FB", "AD",
"D9", "AF", "D8", "B1", "BA", "1D", "BB", "15", "BF", "B2", "FF",
"19", "9C", "FD", "9B", "B3", "CF"), VISCII_Hex = c("C0", "C1",
"C2", "C3", "C8", "C9", "CA", "CC", "CD", "D2", "D3", "D4", "A0",
"D9", "DA", "DD", "E0", "E1", "E2", "E3", "E8", "E9", "EA", "EC",
"ED", "F2", "F3", "F4", "F5", "F9", "FA", "FD", "C5", "E5", "D0",
"F0", "CE", "EE", "9D", "FB", "B4", "BD", "BF", "DF", "80", "D5",
"C4", "E4", "84", "A4", "85", "A5", "86", "A6", "6", "E7", "87",
"A7", "81", "A1", "82", "A2", "2", "C6", "5", "C7", "83", "A3",
"89", "A9", "CB", "EB", "88", "A8", "8A", "AA", "8B", "AB", "8C",
"AC", "8D", "AD", "8E", "AE", "9B", "EF", "98", "B8", "9A", "F7",
"99", "F6", "8F", "AF", "90", "B0", "91", "B1", "92", "B2", "93",
"B5", "95", "BE", "96", "B6", "97", "B7", "B3", "DE", "94", "FE",
"9E", "F8", "9C", "FC", "BA", "D1", "BB", "D7", "BC", "D8", "FF",
"E6", "B9", "F1", "9F", "CF", "1E", "DC", "14", "D6", "19", "DB"
), TCVN3_Hex = c("41 B5", "41 B8", "A2", "41 B7", "45 CC", "45 D0",
"A3", "49 D7", "49 DD", "4F DF", "4F E3", "A4", "4F E2", "55 EF",
"55 F3", "59 FD", "B5", "B8", "A9", "B7", "CC", "D0", "AA", "D7",
"DD", "DF", "E3", "AB", "E2", "EF", "F3", "FD", "A1", "A8", "A7",
"AE", "49 DC", "DC", "55 F2", "F2", "A5", "AC", "A6", "AD", "41 B9",
"B9", "41 B6", "B6", "A2 CA", "CA", "A2 C7", "C7", "A2 C8", "C8",
"A2 C9", "C9", "A2 CB", "CB", "A1 BE", "BE", "A1 BB", "BB", "A1 BC",
"BC", "A1 BD", "BD", "A1 C6", "C6", "45 D1", "D1", "45 CE", "CE",
"45 CF", "CF", "A3 D5", "D5", "A3 D2", "D2", "A3 D3", "D3", "A3 D4",
"D4", "A3 D6", "D6", "49 D8", "D8", "49 DE", "DE", "4F E4", "E4",
"4F E1", "E1", "A4 E8", "E8", "A4 E5", "E5", "A4 E6", "E6", "A4 E7",
"E7", "A4 E9", "E9", "A5 ED", "ED", "A5 EA", "EA", "A5 EB", "EB",
"A5 EC", "EC", "A5 EE", "EE", "55 F4", "F4", "55 F1", "F1", "A6 F8",
"F8", "A6 F5", "F5", "A6 F6", "F6", "A6 F7", "F7", "A6 F9", "F9",
"59 FA", "FA", "59 FE", "FE", "59 FB", "FB", "59 FC", "FC"),
VIQR = c("A`", "A'", "A^", "A~", "E`", "E'", "E^", "I`",
"I'", "O`", "O'", "O^", "O~", "U`", "U'", "Y'", "a`", "a'",
"a^", "a~", "e`", "e'", "e^", "i`", "i'", "o`", "o'", "o^",
"o~", "u`", "u'", "y'", "A(", "a(", "DD", "dd", "I~", "i~",
"U~", "u~", "O+", "o+", "U+", "u+", "A.", "a.", "A?", "a?",
"A^'", "a^'", "A^`", "a^`", "A^?", "a^?", "A^~", "a^~", "A^.",
"a^.", "A('", "a('", "A(`", "a(`", "A(?", "a(?", "A(~", "a(~",
"A(.", "a(.", "E.", "e.", "E?", "e?", "E~", "e~", "E^'",
"e^'", "E^`", "e^`", "E^?", "e^?", "E^~", "e^~", "E^.", "e^.",
"I?", "i?", "I.", "i.", "O.", "o.", "O?", "o?", "O^'", "o^'",
"O^`", "o^`", "O^?", "o^?", "O^~", "o^~", "O^.", "o^.", "O+'",
"o+'", "O+`", "o+`", "O+?", "o+?", "O+~", "o+~", "O+.", "o+.",
"U.", "u.", "U?", "u?", "U+'", "u+'", "U+`", "u+`", "U+?",
"u+?", "U+~", "u+~", "U+.", "u+.", "Y`", "y`", "Y.", "y.",
"Y?", "y?", "Y~", "y~"), English_Name = c("LATIN CAPITAL LETTER A WITH GRAVE",
"LATIN CAPITAL LETTER A WITH ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER A WITH TILDE", "LATIN CAPITAL LETTER E WITH GRAVE",
"LATIN CAPITAL LETTER E WITH ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER I WITH GRAVE", "LATIN CAPITAL LETTER I WITH ACUTE",
"LATIN CAPITAL LETTER O WITH GRAVE", "LATIN CAPITAL LETTER O WITH ACUTE",
"LATIN CAPITAL LETTER O WITH CIRCUMFLEX", "LATIN CAPITAL LETTER O WITH TILDE",
"LATIN CAPITAL LETTER U WITH GRAVE", "LATIN CAPITAL LETTER U WITH ACUTE",
"LATIN CAPITAL LETTER Y WITH ACUTE", "LATIN SMALL LETTER A WITH GRAVE",
"LATIN SMALL LETTER A WITH ACUTE", "LATIN SMALL LETTER A WITH CIRCUMFLEX",
"LATIN SMALL LETTER A WITH TILDE", "LATIN SMALL LETTER E WITH GRAVE",
"LATIN SMALL LETTER E WITH ACUTE", "LATIN SMALL LETTER E WITH CIRCUMFLEX",
"LATIN SMALL LETTER I WITH GRAVE", "LATIN SMALL LETTER I WITH ACUTE",
"LATIN SMALL LETTER O WITH GRAVE", "LATIN SMALL LETTER O WITH ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX", "LATIN SMALL LETTER O WITH TILDE",
"LATIN SMALL LETTER U WITH GRAVE", "LATIN SMALL LETTER U WITH ACUTE",
"LATIN SMALL LETTER Y WITH ACUTE", "LATIN CAPITAL LETTER A WITH BREVE",
"LATIN SMALL LETTER A WITH BREVE", "LATIN CAPITAL LETTER D WITH STROKE",
"LATIN SMALL LETTER D WITH STROKE", "LATIN CAPITAL LETTER I WITH TILDE",
"LATIN SMALL LETTER I WITH TILDE", "LATIN CAPITAL LETTER U WITH TILDE",
"LATIN SMALL LETTER U WITH TILDE", "LATIN CAPITAL LETTER O WITH HORN",
"LATIN SMALL LETTER O WITH HORN", "LATIN CAPITAL LETTER U WITH HORN",
"LATIN SMALL LETTER U WITH HORN", "LATIN CAPITAL LETTER A WITH DOT BELOW",
"LATIN SMALL LETTER A WITH DOT BELOW", "LATIN CAPITAL LETTER A WITH HOOK ABOVE",
"LATIN SMALL LETTER A WITH HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER A WITH BREVE AND ACUTE",
"LATIN SMALL LETTER A WITH BREVE AND ACUTE", "LATIN CAPITAL LETTER A WITH BREVE AND GRAVE",
"LATIN SMALL LETTER A WITH BREVE AND GRAVE", "LATIN CAPITAL LETTER A WITH BREVE AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH BREVE AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH BREVE AND TILDE",
"LATIN SMALL LETTER A WITH BREVE AND TILDE", "LATIN CAPITAL LETTER A WITH BREVE AND DOT BELOW",
"LATIN SMALL LETTER A WITH BREVE AND DOT BELOW", "LATIN CAPITAL LETTER E WITH DOT BELOW",
"LATIN SMALL LETTER E WITH DOT BELOW", "LATIN CAPITAL LETTER E WITH HOOK ABOVE",
"LATIN SMALL LETTER E WITH HOOK ABOVE", "LATIN CAPITAL LETTER E WITH TILDE",
"LATIN SMALL LETTER E WITH TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER I WITH HOOK ABOVE",
"LATIN SMALL LETTER I WITH HOOK ABOVE", "LATIN CAPITAL LETTER I WITH DOT BELOW",
"LATIN SMALL LETTER I WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH DOT BELOW",
"LATIN SMALL LETTER O WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH HOOK ABOVE",
"LATIN SMALL LETTER O WITH HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER O WITH HORN AND ACUTE",
"LATIN SMALL LETTER O WITH HORN AND ACUTE", "LATIN CAPITAL LETTER O WITH HORN AND GRAVE",
"LATIN SMALL LETTER O WITH HORN AND GRAVE", "LATIN CAPITAL LETTER O WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH HORN AND TILDE",
"LATIN SMALL LETTER O WITH HORN AND TILDE", "LATIN CAPITAL LETTER O WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER O WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER U WITH DOT BELOW",
"LATIN SMALL LETTER U WITH DOT BELOW", "LATIN CAPITAL LETTER U WITH HOOK ABOVE",
"LATIN SMALL LETTER U WITH HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND ACUTE",
"LATIN SMALL LETTER U WITH HORN AND ACUTE", "LATIN CAPITAL LETTER U WITH HORN AND GRAVE",
"LATIN SMALL LETTER U WITH HORN AND GRAVE", "LATIN CAPITAL LETTER U WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER U WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND TILDE",
"LATIN SMALL LETTER U WITH HORN AND TILDE", "LATIN CAPITAL LETTER U WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER U WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER Y WITH GRAVE",
"LATIN SMALL LETTER Y WITH GRAVE", "LATIN CAPITAL LETTER Y WITH DOT BELOW",
"LATIN SMALL LETTER Y WITH DOT BELOW", "LATIN CAPITAL LETTER Y WITH HOOK ABOVE",
"LATIN SMALL LETTER Y WITH HOOK ABOVE", "LATIN CAPITAL LETTER Y WITH TILDE",
"LATIN SMALL LETTER Y WITH TILDE"), ascii = c("A", "A", "A",
"A", "E", "E", "E", "I", "I", "O", "O", "O", "O", "U", "U",
"Y", "a", "a", "a", "a", "e", "e", "e", "i", "i", "o", "o",
"o", "o", "u", "u", "y", "A", "a", "D", "d", "I", "i", "U",
"u", "O", "o", "U", "u", "A", "a", "A", "a", "A", "a", "A",
"a", "A", "a", "A", "a", "A", "a", "A", "a", "A", "a", "A",
"a", "A", "a", "A", "a", "E", "e", "E", "e", "E", "e", "E",
"e", "E", "e", "E", "e", "E", "e", "E", "e", "I", "i", "I",
"i", "O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O",
"o", "O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O",
"o", "U", "u", "U", "u", "U", "u", "U", "u", "U", "u", "U",
"u", "U", "u", "Y", "y", "Y", "y", "Y", "y", "Y", "y")), row.names = c(NA,
-134L), class = "data.frame") #out[,1] <- as_utf8(out[,1])
out$Unicode <- intToUtf8(gsub("U+", "0x", out$Unicode_Hex, fixed = T), multiple = TRUE)
out$VNI <- sapply(strsplit(out$VNI_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
out$VPS <- sapply(strsplit(out$VPS_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
out$VISCII <- sapply(strsplit(out$VISCII_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
out$TCVN3 <- sapply(strsplit(out$TCVN3_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
}
if(version == 2){
out <- structure(list(Unicode_Hex = c("U+00C0", "U+00C1", "U+00C2",
"U+00C3", "U+00C8", "U+00C9", "U+00CA", "U+00CC", "U+00CD", "U+00D2",
"U+00D3", "U+00D4", "U+00D5", "U+00D9", "U+00DA", "U+00DD", "U+00E0",
"U+00E1", "U+00E2", "U+00E3", "U+00E8", "U+00E9", "U+00EA", "U+00EC",
"U+00ED", "U+00F2", "U+00F3", "U+00F4", "U+00F5", "U+00F9", "U+00FA",
"U+00FD", "U+0102", "U+0103", "U+0110", "U+0111", "U+0128", "U+0129",
"U+0168", "U+0169", "U+01A0", "U+01A1", "U+01AF", "U+01B0", "U+1EA0",
"U+1EA1", "U+1EA2", "U+1EA3", "U+1EA4", "U+1EA5", "U+1EA6", "U+1EA7",
"U+1EA8", "U+1EA9", "U+1EAA", "U+1EAB", "U+1EAC", "U+1EAD", "U+1EAE",
"U+1EAF", "U+1EB0", "U+1EB1", "U+1EB2", "U+1EB3", "U+1EB4", "U+1EB5",
"U+1EB6", "U+1EB7", "U+1EB8", "U+1EB9", "U+1EBA", "U+1EBB", "U+1EBC",
"U+1EBD", "U+1EBE", "U+1EBF", "U+1EC0", "U+1EC1", "U+1EC2", "U+1EC3",
"U+1EC4", "U+1EC5", "U+1EC6", "U+1EC7", "U+1EC8", "U+1EC9", "U+1ECA",
"U+1ECB", "U+1ECC", "U+1ECD", "U+1ECE", "U+1ECF", "U+1ED0", "U+1ED1",
"U+1ED2", "U+1ED3", "U+1ED4", "U+1ED5", "U+1ED6", "U+1ED7", "U+1ED8",
"U+1ED9", "U+1EDA", "U+1EDB", "U+1EDC", "U+1EDD", "U+1EDE", "U+1EDF",
"U+1EE0", "U+1EE1", "U+1EE2", "U+1EE3", "U+1EE4", "U+1EE5", "U+1EE6",
"U+1EE7", "U+1EE8", "U+1EE9", "U+1EEA", "U+1EEB", "U+1EEC", "U+1EED",
"U+1EEE", "U+1EEF", "U+1EF0", "U+1EF1", "U+1EF2", "U+1EF3", "U+1EF4",
"U+1EF5", "U+1EF6", "U+1EF7", "U+1EF8", "U+1EF9"), ASCII = c("A",
"A", "A", "A", "E", "E", "E", "I", "I", "O", "O", "O", "O", "U",
"U", "Y", "a", "a", "a", "a", "e", "e", "e", "i", "i", "o", "o",
"o", "o", "u", "u", "y", "A", "a", "D", "d", "I", "i", "U", "u",
"O", "o", "U", "u", "A", "a", "A", "a", "A", "a", "A", "a", "A",
"a", "A", "a", "A", "a", "A", "a", "A", "a", "A", "a", "A", "a",
"A", "a", "E", "e", "E", "e", "E", "e", "E", "e", "E", "e", "E",
"e", "E", "e", "E", "e", "I", "i", "I", "i", "O", "o", "O", "o",
"O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O",
"o", "O", "o", "O", "o", "O", "o", "U", "u", "U", "u", "U", "u",
"U", "u", "U", "u", "U", "u", "U", "u", "Y", "y", "Y", "y", "Y",
"y", "Y", "y"), English_Name = c("LATIN CAPITAL LETTER A WITH GRAVE",
"LATIN CAPITAL LETTER A WITH ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER A WITH TILDE", "LATIN CAPITAL LETTER E WITH GRAVE",
"LATIN CAPITAL LETTER E WITH ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER I WITH GRAVE", "LATIN CAPITAL LETTER I WITH ACUTE",
"LATIN CAPITAL LETTER O WITH GRAVE", "LATIN CAPITAL LETTER O WITH ACUTE",
"LATIN CAPITAL LETTER O WITH CIRCUMFLEX", "LATIN CAPITAL LETTER O WITH TILDE",
"LATIN CAPITAL LETTER U WITH GRAVE", "LATIN CAPITAL LETTER U WITH ACUTE",
"LATIN CAPITAL LETTER Y WITH ACUTE", "LATIN SMALL LETTER A WITH GRAVE",
"LATIN SMALL LETTER A WITH ACUTE", "LATIN SMALL LETTER A WITH CIRCUMFLEX",
"LATIN SMALL LETTER A WITH TILDE", "LATIN SMALL LETTER E WITH GRAVE",
"LATIN SMALL LETTER E WITH ACUTE", "LATIN SMALL LETTER E WITH CIRCUMFLEX",
"LATIN SMALL LETTER I WITH GRAVE", "LATIN SMALL LETTER I WITH ACUTE",
"LATIN SMALL LETTER O WITH GRAVE", "LATIN SMALL LETTER O WITH ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX", "LATIN SMALL LETTER O WITH TILDE",
"LATIN SMALL LETTER U WITH GRAVE", "LATIN SMALL LETTER U WITH ACUTE",
"LATIN SMALL LETTER Y WITH ACUTE", "LATIN CAPITAL LETTER A WITH BREVE",
"LATIN SMALL LETTER A WITH BREVE", "LATIN CAPITAL LETTER D WITH STROKE",
"LATIN SMALL LETTER D WITH STROKE", "LATIN CAPITAL LETTER I WITH TILDE",
"LATIN SMALL LETTER I WITH TILDE", "LATIN CAPITAL LETTER U WITH TILDE",
"LATIN SMALL LETTER U WITH TILDE", "LATIN CAPITAL LETTER O WITH HORN",
"LATIN SMALL LETTER O WITH HORN", "LATIN CAPITAL LETTER U WITH HORN",
"LATIN SMALL LETTER U WITH HORN", "LATIN CAPITAL LETTER A WITH DOT BELOW",
"LATIN SMALL LETTER A WITH DOT BELOW", "LATIN CAPITAL LETTER A WITH HOOK ABOVE",
"LATIN SMALL LETTER A WITH HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER A WITH BREVE AND ACUTE",
"LATIN SMALL LETTER A WITH BREVE AND ACUTE", "LATIN CAPITAL LETTER A WITH BREVE AND GRAVE",
"LATIN SMALL LETTER A WITH BREVE AND GRAVE", "LATIN CAPITAL LETTER A WITH BREVE AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH BREVE AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH BREVE AND TILDE",
"LATIN SMALL LETTER A WITH BREVE AND TILDE", "LATIN CAPITAL LETTER A WITH BREVE AND DOT BELOW",
"LATIN SMALL LETTER A WITH BREVE AND DOT BELOW", "LATIN CAPITAL LETTER E WITH DOT BELOW",
"LATIN SMALL LETTER E WITH DOT BELOW", "LATIN CAPITAL LETTER E WITH HOOK ABOVE",
"LATIN SMALL LETTER E WITH HOOK ABOVE", "LATIN CAPITAL LETTER E WITH TILDE",
"LATIN SMALL LETTER E WITH TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER I WITH HOOK ABOVE",
"LATIN SMALL LETTER I WITH HOOK ABOVE", "LATIN CAPITAL LETTER I WITH DOT BELOW",
"LATIN SMALL LETTER I WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH DOT BELOW",
"LATIN SMALL LETTER O WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH HOOK ABOVE",
"LATIN SMALL LETTER O WITH HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER O WITH HORN AND ACUTE",
"LATIN SMALL LETTER O WITH HORN AND ACUTE", "LATIN CAPITAL LETTER O WITH HORN AND GRAVE",
"LATIN SMALL LETTER O WITH HORN AND GRAVE", "LATIN CAPITAL LETTER O WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH HORN AND TILDE",
"LATIN SMALL LETTER O WITH HORN AND TILDE", "LATIN CAPITAL LETTER O WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER O WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER U WITH DOT BELOW",
"LATIN SMALL LETTER U WITH DOT BELOW", "LATIN CAPITAL LETTER U WITH HOOK ABOVE",
"LATIN SMALL LETTER U WITH HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND ACUTE",
"LATIN SMALL LETTER U WITH HORN AND ACUTE", "LATIN CAPITAL LETTER U WITH HORN AND GRAVE",
"LATIN SMALL LETTER U WITH HORN AND GRAVE", "LATIN CAPITAL LETTER U WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER U WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND TILDE",
"LATIN SMALL LETTER U WITH HORN AND TILDE", "LATIN CAPITAL LETTER U WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER U WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER Y WITH GRAVE",
"LATIN SMALL LETTER Y WITH GRAVE", "LATIN CAPITAL LETTER Y WITH DOT BELOW",
"LATIN SMALL LETTER Y WITH DOT BELOW", "LATIN CAPITAL LETTER Y WITH HOOK ABOVE",
"LATIN SMALL LETTER Y WITH HOOK ABOVE", "LATIN CAPITAL LETTER Y WITH TILDE",
"LATIN SMALL LETTER Y WITH TILDE"), Unicode_int = c(192L, 193L,
194L, 195L, 200L, 201L, 202L, 204L, 205L, 210L, 211L, 212L, 213L,
217L, 218L, 221L, 224L, 225L, 226L, 227L, 232L, 233L, 234L, 236L,
237L, 242L, 243L, 244L, 245L, 249L, 250L, 253L, 258L, 259L, 208L,
273L, 296L, 297L, 360L, 361L, 416L, 417L, 431L, 432L, 7840L,
7841L, 7842L, 7843L, 7844L, 7845L, 7846L, 7847L, 7848L, 7849L,
7850L, 7851L, 7852L, 7853L, 7854L, 7855L, 7856L, 7857L, 7858L,
7859L, 7860L, 7861L, 7862L, 7863L, 7864L, 7865L, 7866L, 7867L,
7868L, 7869L, 7870L, 7871L, 7872L, 7873L, 7874L, 7875L, 7876L,
7877L, 7878L, 7879L, 7880L, 7881L, 7882L, 7883L, 7884L, 7885L,
7886L, 7887L, 7888L, 7889L, 7890L, 7891L, 7892L, 7893L, 7894L,
7895L, 7896L, 7897L, 7898L, 7899L, 7900L, 7901L, 7902L, 7903L,
7904L, 7905L, 7906L, 7907L, 7908L, 7909L, 7910L, 7911L, 7912L,
7913L, 7914L, 7915L, 7916L, 7917L, 7918L, 7919L, 7920L, 7921L,
7922L, 7923L, 7924L, 7925L, 7926L, 7927L, 7928L, 7929L), TCVN3_int = list(
c(65L, 181L), c(65L, 184L), 162L, c(65L, 183L), c(69L, 204L
), c(69L, 208L), 163L, c(73L, 215L), c(73L, 221L), c(79L,
223L), c(79L, 227L), 164L, c(79L, 226L), c(85L, 239L), c(85L,
243L), c(89L, 253L), 181L, 184L, 169L, 183L, 204L, 208L,
170L, 215L, 221L, 223L, 227L, 171L, 226L, 239L, 243L, 253L,
161L, 168L, 167L, 174L, c(73L, 220L), 220L, c(85L, 242L),
242L, 165L, 172L, 166L, 173L, c(65L, 185L), 185L, c(65L,
182L), 182L, c(162L, 202L), 202L, c(162L, 199L), 199L, c(162L,
200L), 200L, c(162L, 201L), 201L, c(162L, 203L), 203L, c(161L,
190L), 190L, c(161L, 187L), 187L, c(161L, 188L), 188L, c(161L,
189L), 189L, c(161L, 198L), 198L, c(69L, 209L), 209L, c(69L,
206L), 206L, c(69L, 207L), 207L, c(163L, 213L), 213L, c(163L,
210L), 210L, c(163L, 211L), 211L, c(163L, 212L), 212L, c(163L,
214L), 214L, c(73L, 216L), 216L, c(73L, 222L), 222L, c(79L,
228L), 228L, c(79L, 225L), 225L, c(164L, 232L), 232L, c(164L,
229L), 229L, c(164L, 230L), 230L, c(164L, 231L), 231L, c(164L,
233L), 233L, c(165L, 237L), 237L, c(165L, 234L), 234L, c(165L,
235L), 235L, c(165L, 236L), 236L, c(165L, 238L), 238L, c(85L,
244L), 244L, c(85L, 241L), 241L, c(166L, 248L), 248L, c(166L,
245L), 245L, c(166L, 246L), 246L, c(166L, 247L), 247L, c(166L,
249L), 249L, c(89L, 250L), 250L, c(89L, 254L), 254L, c(89L,
251L), 251L, c(89L, 252L), 252L), VISCII_int = c(192L, 193L,
194L, 195L, 200L, 201L, 202L, 204L, 205L, 210L, 211L, 212L, NA,
217L, 218L, 221L, 224L, 225L, 226L, 227L, 232L, 233L, 234L, 236L,
237L, 242L, 243L, 244L, 245L, 249L, 250L, 253L, 197L, 229L, 208L,
240L, 206L, 238L, 157L, 251L, 180L, 189L, 191L, 223L, 8364L,
213L, 196L, 228L, 8222L, 164L, 8230L, 165L, 8224L, 166L, 6L,
231L, 8225L, 167L, 129L, 161L, 8218L, 162L, 2L, 198L, 5L, 199L,
402L, 163L, 8240L, 169L, 203L, 235L, 710L, 168L, 352L, 170L,
8249L, 171L, 338L, 172L, 141L, 173L, 381L, 174L, 8250L, 239L,
732L, 184L, 353L, 247L, 8482L, 246L, 143L, 175L, 144L, 176L,
8216L, 177L, 8217L, 178L, 8220L, 181L, 8226L, 190L, 8211L, 182L,
8212L, 183L, 179L, 222L, 8221L, 254L, 382L, 248L, 339L, 252L,
186L, 209L, 187L, 215L, 188L, 216L, 255L, 230L, 185L, 241L, 376L,
207L, 30L, 220L, 20L, 214L, 25L, 219L), VNI_int = list(c(65L,
216L), c(65L, 217L), c(65L, 194L), c(65L, 213L), c(69L, 216L),
c(69L, 217L), c(69L, 194L), 204L, 205L, c(79L, 216L), c(79L,
217L), c(79L, 194L), c(79L, 213L), c(85L, 216L), c(85L, 217L
), c(89L, 217L), c(97L, 248L), c(97L, 249L), c(97L, 226L),
c(97L, 245L), c(101L, 248L), c(101L, 249L), c(101L, 226L),
236L, 237L, c(111L, 248L), c(111L, 249L), c(111L, 226L),
c(111L, 245L), c(117L, 248L), c(117L, 249L), c(121L, 249L
), c(65L, 202L), c(97L, 234L), 209L, 241L, 211L, 243L, c(85L,
213L), c(117L, 245L), 212L, 244L, 214L, 246L, c(65L, 207L
), c(97L, 239L), c(65L, 219L), c(97L, 251L), c(65L, 193L),
c(97L, 225L), c(65L, 192L), c(97L, 224L), c(65L, 197L), c(97L,
229L), c(65L, 195L), c(97L, 227L), c(65L, 196L), c(97L, 228L
), c(65L, 201L), c(97L, 233L), c(65L, 200L), c(97L, 232L),
c(65L, 218L), c(97L, 250L), c(65L, 220L), c(97L, 252L), c(65L,
203L), c(97L, 235L), c(69L, 207L), c(101L, 239L), c(69L,
219L), c(101L, 251L), c(69L, 213L), c(101L, 245L), c(69L,
193L), c(101L, 225L), c(69L, 192L), c(101L, 224L), c(69L,
197L), c(101L, 229L), c(69L, 195L), c(101L, 227L), c(69L,
196L), c(101L, 228L), 198L, 230L, 210L, 242L, c(79L, 207L
), c(111L, 239L), c(79L, 219L), c(111L, 251L), c(79L, 193L
), c(111L, 225L), c(79L, 192L), c(111L, 224L), c(79L, 197L
), c(111L, 229L), c(79L, 195L), c(111L, 227L), c(79L, 196L
), c(111L, 228L), c(212L, 217L), c(244L, 249L), c(212L, 216L
), c(244L, 248L), c(212L, 219L), c(244L, 251L), 212:213,
244:245, c(212L, 207L), c(244L, 239L), c(85L, 207L), c(117L,
239L), c(85L, 219L), c(117L, 251L), c(214L, 217L), c(246L,
249L), c(214L, 216L), c(246L, 248L), c(214L, 219L), c(246L,
251L), 214:213, 246:245, c(214L, 207L), c(246L, 239L), c(89L,
216L), c(121L, 248L), 206L, 238L, c(89L, 219L), c(121L, 251L
), c(89L, 213L), c(121L, 245L)), VPS_int = c(8364L, 193L,
194L, 8218L, 215L, 201L, 202L, 181L, 180L, 188L, 185L, 212L,
190L, 168L, 218L, 221L, 224L, 225L, 226L, 227L, 232L, 233L, 234L,
236L, 237L, 242L, 243L, 244L, 245L, 249L, 250L, 353L, 710L, 230L,
241L, 199L, 184L, 239L, 172L, 219L, 247L, 214L, 208L, 220L, 2L,
229L, 129L, 228L, 402L, 195L, 8222L, 192L, 8230L, 196L, 28L,
197L, 3L, 198L, 141L, 161L, 381L, 162L, 143L, 163L, 240L, 164L,
4L, 165L, 5L, 203L, 222L, 200L, 254L, 235L, 144L, 8240L, 8220L,
352L, 8221L, 8249L, 8226L, 205L, 6L, 338L, 183L, 204L, 16L, 206L,
17L, 8224L, 189L, 213L, 8211L, 211L, 8212L, 210L, 732L, 176L,
8482L, 8225L, 18L, 182L, 157L, 167L, 382L, 169L, 376L, 170L,
166L, 171L, 19L, 174L, 20L, 248L, 209L, 251L, 173L, 217L, 175L,
216L, 177L, 186L, 29L, 187L, 21L, 191L, 178L, 255L, 25L, 339L,
253L, 8250L, 179L, 207L), VNU_int = c(129L, 8364L, 8222L, 195L,
200L, 201L, 8240L, 204L, 205L, 210L, 338L, 141L, 213L, 217L,
732L, 382L, 161L, 376L, 181L, 172L, 191L, 190L, 197L, 214L, 207L,
221L, 220L, 225L, 223L, 239L, 238L, 251L, 402L, 175L, 8225L,
189L, 296L, 216L, 360L, 241L, 381L, 231L, 339L, 245L, 7840L,
173L, 8218L, 168L, 8230L, 182L, 7846L, 183L, 7848L, 184L, 6L,
185L, 7852L, 186L, 7854L, 176L, 7856L, 177L, 7858L, 178L, 7860L,
179L, 7862L, 180L, 7864L, 194L, 7866L, 192L, 7868L, 193L, 7870L,
198L, 7872L, 203L, 7874L, 204L, 7876L, 205L, 7878L, 206L, 7880L,
215L, 7882L, 217L, 7884L, 224L, 7886L, 222L, 7888L, 226L, 7890L,
227L, 7892L, 228L, 7894L, 229L, 7896L, 230L, 7898L, 232L, 7900L,
233L, 143L, 234L, 7904L, 235L, 7906L, 236L, 7908L, 242L, 8482L,
240L, 157L, 246L, 7914L, 247L, 7916L, 248L, 7918L, 249L, 7920L,
250L, 7922L, 252L, 7924L, 255L, 7926L, 253L, 7928L, 254L)), row.names = c(NA,
-134L), class = "data.frame")
out$Unicode <- sapply(out$Unicode_int, intToUtf8)
out$TCVN3 <- sapply(out$TCVN3_int, intToUtf8)
out$VISCII <- sapply(out$VISCII_int, intToUtf8)
out$VNI <- sapply(out$VNI_int, intToUtf8)
out$VPS <- sapply(out$VPS_int, intToUtf8)
out$VNU <- sapply(out$VNU_int, intToUtf8)
}
return(out)
}
Try the vietnameseConverter package in your browser
Any scripts or data that you put into this service are public.
vietnameseConverter documentation built on Oct. 21, 2021, 1:06 a.m.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Defines functions loadEncodingTableVN
Documented in loadEncodingTableVN
#' Load conversion table for Vietnamese characters
#' @param version Version of this table (1 or 2) - 2 is currentlty being used
#' @return data.frame
#' @details
#' The table was adapted from \url{http://vietunicode.sourceforge.net/charset/}
#' @export
#'
loadEncodingTableVN <- function(version)
{
if(version == 1){
out <- structure(list(Unicode_Hex = c("U+00C0",
"U+00C1", "U+00C2", "U+00C3", "U+00C8", "U+00C9", "U+00CA", "U+00CC",
"U+00CD", "U+00D2", "U+00D3", "U+00D4", "U+00D5", "U+00D9", "U+00DA",
"U+00DD", "U+00E0", "U+00E1", "U+00E2", "U+00E3", "U+00E8", "U+00E9",
"U+00EA", "U+00EC", "U+00ED", "U+00F2", "U+00F3", "U+00F4", "U+00F5",
"U+00F9", "U+00FA", "U+00FD", "U+0102", "U+0103", "U+0110", "U+0111",
"U+0128", "U+0129", "U+0168", "U+0169", "U+01A0", "U+01A1", "U+01AF",
"U+01B0", "U+1EA0", "U+1EA1", "U+1EA2", "U+1EA3", "U+1EA4", "U+1EA5",
"U+1EA6", "U+1EA7", "U+1EA8", "U+1EA9", "U+1EAA", "U+1EAB", "U+1EAC",
"U+1EAD", "U+1EAE", "U+1EAF", "U+1EB0", "U+1EB1", "U+1EB2", "U+1EB3",
"U+1EB4", "U+1EB5", "U+1EB6", "U+1EB7", "U+1EB8", "U+1EB9", "U+1EBA",
"U+1EBB", "U+1EBC", "U+1EBD", "U+1EBE", "U+1EBF", "U+1EC0", "U+1EC1",
"U+1EC2", "U+1EC3", "U+1EC4", "U+1EC5", "U+1EC6", "U+1EC7", "U+1EC8",
"U+1EC9", "U+1ECA", "U+1ECB", "U+1ECC", "U+1ECD", "U+1ECE", "U+1ECF",
"U+1ED0", "U+1ED1", "U+1ED2", "U+1ED3", "U+1ED4", "U+1ED5", "U+1ED6",
"U+1ED7", "U+1ED8", "U+1ED9", "U+1EDA", "U+1EDB", "U+1EDC", "U+1EDD",
"U+1EDE", "U+1EDF", "U+1EE0", "U+1EE1", "U+1EE2", "U+1EE3", "U+1EE4",
"U+1EE5", "U+1EE6", "U+1EE7", "U+1EE8", "U+1EE9", "U+1EEA", "U+1EEB",
"U+1EEC", "U+1EED", "U+1EEE", "U+1EEF", "U+1EF0", "U+1EF1", "U+1EF2",
"U+1EF3", "U+1EF4", "U+1EF5", "U+1EF6", "U+1EF7", "U+1EF8", "U+1EF9"
), VNI_Hex = c("41 D8", "41 D9", "41 C2", "41 D5", "45 D8", "45 D9",
"45 C2", "CC", "CD", "4F D8", "4F D9", "4F C2", "4F D5", "55 D8",
"55 D9", "59 D9", "61 F8", "61 F9", "61 E2", "61 F5", "65 F8",
"65 F9", "65 E2", "EC", "ED", "6F F8", "6F F9", "6F E2", "6F F5",
"75 F8", "75 F9", "79 F9", "41 CA", "61 EA", "D1", "F1", "D3",
"F3", "55 D5", "75 F5", "D4", "F4", "D6", "F6", "41 CF", "61 EF",
"41 DB", "61 FB", "41 C1", "61 E1", "41 C0", "61 E0", "41 C5",
"61 E5", "41 C3", "61 E3", "41 C4", "61 E4", "41 C9", "61 E9",
"41 C8", "61 E8", "41 DA", "61 FA", "41 DC", "61 FC", "41 CB",
"61 EB", "45 CF", "65 EF", "45 DB", "65 FB", "45 D5", "65 F5",
"45 C1", "65 E1", "45 C0", "65 E0", "45 C5", "65 E5", "45 C3",
"65 E3", "45 C4", "65 E4", "C6", "E6", "D2", "F2", "4F CF", "6F EF",
"4F DB", "6F FB", "4F C1", "6F E1", "4F C0", "6F E0", "4F C5",
"6F E5", "4F C3", "6F E3", "4F C4", "6F E4", "D4 D9", "F4 F9",
"D4 D8", "F4 F8", "D4 DB", "F4 FB", "D4 D5", "F4 F5", "D4 CF",
"F4 EF", "55 CF", "75 EF", "55 DB", "75 FB", "D6 D9", "F6 F9",
"D6 D8", "F6 F8", "D6 DB", "F6 FB", "D6 D5", "F6 F5", "D6 CF",
"F6 EF", "59 D8", "79 F8", "CE", "EE", "59 DB", "79 FB", "59 D5",
"79 F5"), VPS_Hex = c("80", "C1", "C2", "82", "D7", "C9", "CA",
"B5", "B4", "BC", "B9", "D4", "BE", "A8", "DA", "DD", "E0", "E1",
"E2", "E3", "E8", "E9", "EA", "EC", "ED", "F2", "F3", "F4", "F5",
"F9", "FA", "9A", "88", "E6", "F1", "C7", "B8", "EF", "AC", "DB",
"F7", "D6", "D0", "DC", "2", "E5", "81", "E4", "83", "C3", "84",
"C0", "85", "C4", "1C", "C5", "3", "C6", "8D", "A1", "8E", "A2",
"8F", "A3", "F0", "A4", "4", "A5", "5", "CB", "DE", "C8", "FE",
"EB", "90", "89", "93", "8A", "94", "8B", "95", "CD", "6", "8C",
"B7", "CC", "10", "CE", "11", "86", "BD", "D5", "96", "D3", "97",
"D2", "98", "B0", "99", "87", "12", "B6", "9D", "A7", "9E", "A9",
"9F", "AA", "A6", "AB", "13", "AE", "14", "F8", "D1", "FB", "AD",
"D9", "AF", "D8", "B1", "BA", "1D", "BB", "15", "BF", "B2", "FF",
"19", "9C", "FD", "9B", "B3", "CF"), VISCII_Hex = c("C0", "C1",
"C2", "C3", "C8", "C9", "CA", "CC", "CD", "D2", "D3", "D4", "A0",
"D9", "DA", "DD", "E0", "E1", "E2", "E3", "E8", "E9", "EA", "EC",
"ED", "F2", "F3", "F4", "F5", "F9", "FA", "FD", "C5", "E5", "D0",
"F0", "CE", "EE", "9D", "FB", "B4", "BD", "BF", "DF", "80", "D5",
"C4", "E4", "84", "A4", "85", "A5", "86", "A6", "6", "E7", "87",
"A7", "81", "A1", "82", "A2", "2", "C6", "5", "C7", "83", "A3",
"89", "A9", "CB", "EB", "88", "A8", "8A", "AA", "8B", "AB", "8C",
"AC", "8D", "AD", "8E", "AE", "9B", "EF", "98", "B8", "9A", "F7",
"99", "F6", "8F", "AF", "90", "B0", "91", "B1", "92", "B2", "93",
"B5", "95", "BE", "96", "B6", "97", "B7", "B3", "DE", "94", "FE",
"9E", "F8", "9C", "FC", "BA", "D1", "BB", "D7", "BC", "D8", "FF",
"E6", "B9", "F1", "9F", "CF", "1E", "DC", "14", "D6", "19", "DB"
), TCVN3_Hex = c("41 B5", "41 B8", "A2", "41 B7", "45 CC", "45 D0",
"A3", "49 D7", "49 DD", "4F DF", "4F E3", "A4", "4F E2", "55 EF",
"55 F3", "59 FD", "B5", "B8", "A9", "B7", "CC", "D0", "AA", "D7",
"DD", "DF", "E3", "AB", "E2", "EF", "F3", "FD", "A1", "A8", "A7",
"AE", "49 DC", "DC", "55 F2", "F2", "A5", "AC", "A6", "AD", "41 B9",
"B9", "41 B6", "B6", "A2 CA", "CA", "A2 C7", "C7", "A2 C8", "C8",
"A2 C9", "C9", "A2 CB", "CB", "A1 BE", "BE", "A1 BB", "BB", "A1 BC",
"BC", "A1 BD", "BD", "A1 C6", "C6", "45 D1", "D1", "45 CE", "CE",
"45 CF", "CF", "A3 D5", "D5", "A3 D2", "D2", "A3 D3", "D3", "A3 D4",
"D4", "A3 D6", "D6", "49 D8", "D8", "49 DE", "DE", "4F E4", "E4",
"4F E1", "E1", "A4 E8", "E8", "A4 E5", "E5", "A4 E6", "E6", "A4 E7",
"E7", "A4 E9", "E9", "A5 ED", "ED", "A5 EA", "EA", "A5 EB", "EB",
"A5 EC", "EC", "A5 EE", "EE", "55 F4", "F4", "55 F1", "F1", "A6 F8",
"F8", "A6 F5", "F5", "A6 F6", "F6", "A6 F7", "F7", "A6 F9", "F9",
"59 FA", "FA", "59 FE", "FE", "59 FB", "FB", "59 FC", "FC"),
VIQR = c("A`", "A'", "A^", "A~", "E`", "E'", "E^", "I`",
"I'", "O`", "O'", "O^", "O~", "U`", "U'", "Y'", "a`", "a'",
"a^", "a~", "e`", "e'", "e^", "i`", "i'", "o`", "o'", "o^",
"o~", "u`", "u'", "y'", "A(", "a(", "DD", "dd", "I~", "i~",
"U~", "u~", "O+", "o+", "U+", "u+", "A.", "a.", "A?", "a?",
"A^'", "a^'", "A^`", "a^`", "A^?", "a^?", "A^~", "a^~", "A^.",
"a^.", "A('", "a('", "A(`", "a(`", "A(?", "a(?", "A(~", "a(~",
"A(.", "a(.", "E.", "e.", "E?", "e?", "E~", "e~", "E^'",
"e^'", "E^`", "e^`", "E^?", "e^?", "E^~", "e^~", "E^.", "e^.",
"I?", "i?", "I.", "i.", "O.", "o.", "O?", "o?", "O^'", "o^'",
"O^`", "o^`", "O^?", "o^?", "O^~", "o^~", "O^.", "o^.", "O+'",
"o+'", "O+`", "o+`", "O+?", "o+?", "O+~", "o+~", "O+.", "o+.",
"U.", "u.", "U?", "u?", "U+'", "u+'", "U+`", "u+`", "U+?",
"u+?", "U+~", "u+~", "U+.", "u+.", "Y`", "y`", "Y.", "y.",
"Y?", "y?", "Y~", "y~"), English_Name = c("LATIN CAPITAL LETTER A WITH GRAVE",
"LATIN CAPITAL LETTER A WITH ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER A WITH TILDE", "LATIN CAPITAL LETTER E WITH GRAVE",
"LATIN CAPITAL LETTER E WITH ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER I WITH GRAVE", "LATIN CAPITAL LETTER I WITH ACUTE",
"LATIN CAPITAL LETTER O WITH GRAVE", "LATIN CAPITAL LETTER O WITH ACUTE",
"LATIN CAPITAL LETTER O WITH CIRCUMFLEX", "LATIN CAPITAL LETTER O WITH TILDE",
"LATIN CAPITAL LETTER U WITH GRAVE", "LATIN CAPITAL LETTER U WITH ACUTE",
"LATIN CAPITAL LETTER Y WITH ACUTE", "LATIN SMALL LETTER A WITH GRAVE",
"LATIN SMALL LETTER A WITH ACUTE", "LATIN SMALL LETTER A WITH CIRCUMFLEX",
"LATIN SMALL LETTER A WITH TILDE", "LATIN SMALL LETTER E WITH GRAVE",
"LATIN SMALL LETTER E WITH ACUTE", "LATIN SMALL LETTER E WITH CIRCUMFLEX",
"LATIN SMALL LETTER I WITH GRAVE", "LATIN SMALL LETTER I WITH ACUTE",
"LATIN SMALL LETTER O WITH GRAVE", "LATIN SMALL LETTER O WITH ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX", "LATIN SMALL LETTER O WITH TILDE",
"LATIN SMALL LETTER U WITH GRAVE", "LATIN SMALL LETTER U WITH ACUTE",
"LATIN SMALL LETTER Y WITH ACUTE", "LATIN CAPITAL LETTER A WITH BREVE",
"LATIN SMALL LETTER A WITH BREVE", "LATIN CAPITAL LETTER D WITH STROKE",
"LATIN SMALL LETTER D WITH STROKE", "LATIN CAPITAL LETTER I WITH TILDE",
"LATIN SMALL LETTER I WITH TILDE", "LATIN CAPITAL LETTER U WITH TILDE",
"LATIN SMALL LETTER U WITH TILDE", "LATIN CAPITAL LETTER O WITH HORN",
"LATIN SMALL LETTER O WITH HORN", "LATIN CAPITAL LETTER U WITH HORN",
"LATIN SMALL LETTER U WITH HORN", "LATIN CAPITAL LETTER A WITH DOT BELOW",
"LATIN SMALL LETTER A WITH DOT BELOW", "LATIN CAPITAL LETTER A WITH HOOK ABOVE",
"LATIN SMALL LETTER A WITH HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER A WITH BREVE AND ACUTE",
"LATIN SMALL LETTER A WITH BREVE AND ACUTE", "LATIN CAPITAL LETTER A WITH BREVE AND GRAVE",
"LATIN SMALL LETTER A WITH BREVE AND GRAVE", "LATIN CAPITAL LETTER A WITH BREVE AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH BREVE AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH BREVE AND TILDE",
"LATIN SMALL LETTER A WITH BREVE AND TILDE", "LATIN CAPITAL LETTER A WITH BREVE AND DOT BELOW",
"LATIN SMALL LETTER A WITH BREVE AND DOT BELOW", "LATIN CAPITAL LETTER E WITH DOT BELOW",
"LATIN SMALL LETTER E WITH DOT BELOW", "LATIN CAPITAL LETTER E WITH HOOK ABOVE",
"LATIN SMALL LETTER E WITH HOOK ABOVE", "LATIN CAPITAL LETTER E WITH TILDE",
"LATIN SMALL LETTER E WITH TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER I WITH HOOK ABOVE",
"LATIN SMALL LETTER I WITH HOOK ABOVE", "LATIN CAPITAL LETTER I WITH DOT BELOW",
"LATIN SMALL LETTER I WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH DOT BELOW",
"LATIN SMALL LETTER O WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH HOOK ABOVE",
"LATIN SMALL LETTER O WITH HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER O WITH HORN AND ACUTE",
"LATIN SMALL LETTER O WITH HORN AND ACUTE", "LATIN CAPITAL LETTER O WITH HORN AND GRAVE",
"LATIN SMALL LETTER O WITH HORN AND GRAVE", "LATIN CAPITAL LETTER O WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH HORN AND TILDE",
"LATIN SMALL LETTER O WITH HORN AND TILDE", "LATIN CAPITAL LETTER O WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER O WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER U WITH DOT BELOW",
"LATIN SMALL LETTER U WITH DOT BELOW", "LATIN CAPITAL LETTER U WITH HOOK ABOVE",
"LATIN SMALL LETTER U WITH HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND ACUTE",
"LATIN SMALL LETTER U WITH HORN AND ACUTE", "LATIN CAPITAL LETTER U WITH HORN AND GRAVE",
"LATIN SMALL LETTER U WITH HORN AND GRAVE", "LATIN CAPITAL LETTER U WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER U WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND TILDE",
"LATIN SMALL LETTER U WITH HORN AND TILDE", "LATIN CAPITAL LETTER U WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER U WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER Y WITH GRAVE",
"LATIN SMALL LETTER Y WITH GRAVE", "LATIN CAPITAL LETTER Y WITH DOT BELOW",
"LATIN SMALL LETTER Y WITH DOT BELOW", "LATIN CAPITAL LETTER Y WITH HOOK ABOVE",
"LATIN SMALL LETTER Y WITH HOOK ABOVE", "LATIN CAPITAL LETTER Y WITH TILDE",
"LATIN SMALL LETTER Y WITH TILDE"), ascii = c("A", "A", "A",
"A", "E", "E", "E", "I", "I", "O", "O", "O", "O", "U", "U",
"Y", "a", "a", "a", "a", "e", "e", "e", "i", "i", "o", "o",
"o", "o", "u", "u", "y", "A", "a", "D", "d", "I", "i", "U",
"u", "O", "o", "U", "u", "A", "a", "A", "a", "A", "a", "A",
"a", "A", "a", "A", "a", "A", "a", "A", "a", "A", "a", "A",
"a", "A", "a", "A", "a", "E", "e", "E", "e", "E", "e", "E",
"e", "E", "e", "E", "e", "E", "e", "E", "e", "I", "i", "I",
"i", "O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O",
"o", "O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O",
"o", "U", "u", "U", "u", "U", "u", "U", "u", "U", "u", "U",
"u", "U", "u", "Y", "y", "Y", "y", "Y", "y", "Y", "y")), row.names = c(NA,
-134L), class = "data.frame") #out[,1] <- as_utf8(out[,1])
out$Unicode <- intToUtf8(gsub("U+", "0x", out$Unicode_Hex, fixed = T), multiple = TRUE)
out$VNI <- sapply(strsplit(out$VNI_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
out$VPS <- sapply(strsplit(out$VPS_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
out$VISCII <- sapply(strsplit(out$VISCII_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
out$TCVN3 <- sapply(strsplit(out$TCVN3_Hex, " "), FUN = function(x) {
intToUtf8(paste0("0x", x))})
}
if(version == 2){
out <- structure(list(Unicode_Hex = c("U+00C0", "U+00C1", "U+00C2",
"U+00C3", "U+00C8", "U+00C9", "U+00CA", "U+00CC", "U+00CD", "U+00D2",
"U+00D3", "U+00D4", "U+00D5", "U+00D9", "U+00DA", "U+00DD", "U+00E0",
"U+00E1", "U+00E2", "U+00E3", "U+00E8", "U+00E9", "U+00EA", "U+00EC",
"U+00ED", "U+00F2", "U+00F3", "U+00F4", "U+00F5", "U+00F9", "U+00FA",
"U+00FD", "U+0102", "U+0103", "U+0110", "U+0111", "U+0128", "U+0129",
"U+0168", "U+0169", "U+01A0", "U+01A1", "U+01AF", "U+01B0", "U+1EA0",
"U+1EA1", "U+1EA2", "U+1EA3", "U+1EA4", "U+1EA5", "U+1EA6", "U+1EA7",
"U+1EA8", "U+1EA9", "U+1EAA", "U+1EAB", "U+1EAC", "U+1EAD", "U+1EAE",
"U+1EAF", "U+1EB0", "U+1EB1", "U+1EB2", "U+1EB3", "U+1EB4", "U+1EB5",
"U+1EB6", "U+1EB7", "U+1EB8", "U+1EB9", "U+1EBA", "U+1EBB", "U+1EBC",
"U+1EBD", "U+1EBE", "U+1EBF", "U+1EC0", "U+1EC1", "U+1EC2", "U+1EC3",
"U+1EC4", "U+1EC5", "U+1EC6", "U+1EC7", "U+1EC8", "U+1EC9", "U+1ECA",
"U+1ECB", "U+1ECC", "U+1ECD", "U+1ECE", "U+1ECF", "U+1ED0", "U+1ED1",
"U+1ED2", "U+1ED3", "U+1ED4", "U+1ED5", "U+1ED6", "U+1ED7", "U+1ED8",
"U+1ED9", "U+1EDA", "U+1EDB", "U+1EDC", "U+1EDD", "U+1EDE", "U+1EDF",
"U+1EE0", "U+1EE1", "U+1EE2", "U+1EE3", "U+1EE4", "U+1EE5", "U+1EE6",
"U+1EE7", "U+1EE8", "U+1EE9", "U+1EEA", "U+1EEB", "U+1EEC", "U+1EED",
"U+1EEE", "U+1EEF", "U+1EF0", "U+1EF1", "U+1EF2", "U+1EF3", "U+1EF4",
"U+1EF5", "U+1EF6", "U+1EF7", "U+1EF8", "U+1EF9"), ASCII = c("A",
"A", "A", "A", "E", "E", "E", "I", "I", "O", "O", "O", "O", "U",
"U", "Y", "a", "a", "a", "a", "e", "e", "e", "i", "i", "o", "o",
"o", "o", "u", "u", "y", "A", "a", "D", "d", "I", "i", "U", "u",
"O", "o", "U", "u", "A", "a", "A", "a", "A", "a", "A", "a", "A",
"a", "A", "a", "A", "a", "A", "a", "A", "a", "A", "a", "A", "a",
"A", "a", "E", "e", "E", "e", "E", "e", "E", "e", "E", "e", "E",
"e", "E", "e", "E", "e", "I", "i", "I", "i", "O", "o", "O", "o",
"O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O", "o", "O",
"o", "O", "o", "O", "o", "O", "o", "U", "u", "U", "u", "U", "u",
"U", "u", "U", "u", "U", "u", "U", "u", "Y", "y", "Y", "y", "Y",
"y", "Y", "y"), English_Name = c("LATIN CAPITAL LETTER A WITH GRAVE",
"LATIN CAPITAL LETTER A WITH ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER A WITH TILDE", "LATIN CAPITAL LETTER E WITH GRAVE",
"LATIN CAPITAL LETTER E WITH ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX",
"LATIN CAPITAL LETTER I WITH GRAVE", "LATIN CAPITAL LETTER I WITH ACUTE",
"LATIN CAPITAL LETTER O WITH GRAVE", "LATIN CAPITAL LETTER O WITH ACUTE",
"LATIN CAPITAL LETTER O WITH CIRCUMFLEX", "LATIN CAPITAL LETTER O WITH TILDE",
"LATIN CAPITAL LETTER U WITH GRAVE", "LATIN CAPITAL LETTER U WITH ACUTE",
"LATIN CAPITAL LETTER Y WITH ACUTE", "LATIN SMALL LETTER A WITH GRAVE",
"LATIN SMALL LETTER A WITH ACUTE", "LATIN SMALL LETTER A WITH CIRCUMFLEX",
"LATIN SMALL LETTER A WITH TILDE", "LATIN SMALL LETTER E WITH GRAVE",
"LATIN SMALL LETTER E WITH ACUTE", "LATIN SMALL LETTER E WITH CIRCUMFLEX",
"LATIN SMALL LETTER I WITH GRAVE", "LATIN SMALL LETTER I WITH ACUTE",
"LATIN SMALL LETTER O WITH GRAVE", "LATIN SMALL LETTER O WITH ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX", "LATIN SMALL LETTER O WITH TILDE",
"LATIN SMALL LETTER U WITH GRAVE", "LATIN SMALL LETTER U WITH ACUTE",
"LATIN SMALL LETTER Y WITH ACUTE", "LATIN CAPITAL LETTER A WITH BREVE",
"LATIN SMALL LETTER A WITH BREVE", "LATIN CAPITAL LETTER D WITH STROKE",
"LATIN SMALL LETTER D WITH STROKE", "LATIN CAPITAL LETTER I WITH TILDE",
"LATIN SMALL LETTER I WITH TILDE", "LATIN CAPITAL LETTER U WITH TILDE",
"LATIN SMALL LETTER U WITH TILDE", "LATIN CAPITAL LETTER O WITH HORN",
"LATIN SMALL LETTER O WITH HORN", "LATIN CAPITAL LETTER U WITH HORN",
"LATIN SMALL LETTER U WITH HORN", "LATIN CAPITAL LETTER A WITH DOT BELOW",
"LATIN SMALL LETTER A WITH DOT BELOW", "LATIN CAPITAL LETTER A WITH HOOK ABOVE",
"LATIN SMALL LETTER A WITH HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER A WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER A WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER A WITH BREVE AND ACUTE",
"LATIN SMALL LETTER A WITH BREVE AND ACUTE", "LATIN CAPITAL LETTER A WITH BREVE AND GRAVE",
"LATIN SMALL LETTER A WITH BREVE AND GRAVE", "LATIN CAPITAL LETTER A WITH BREVE AND HOOK ABOVE",
"LATIN SMALL LETTER A WITH BREVE AND HOOK ABOVE", "LATIN CAPITAL LETTER A WITH BREVE AND TILDE",
"LATIN SMALL LETTER A WITH BREVE AND TILDE", "LATIN CAPITAL LETTER A WITH BREVE AND DOT BELOW",
"LATIN SMALL LETTER A WITH BREVE AND DOT BELOW", "LATIN CAPITAL LETTER E WITH DOT BELOW",
"LATIN SMALL LETTER E WITH DOT BELOW", "LATIN CAPITAL LETTER E WITH HOOK ABOVE",
"LATIN SMALL LETTER E WITH HOOK ABOVE", "LATIN CAPITAL LETTER E WITH TILDE",
"LATIN SMALL LETTER E WITH TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER E WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER E WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER I WITH HOOK ABOVE",
"LATIN SMALL LETTER I WITH HOOK ABOVE", "LATIN CAPITAL LETTER I WITH DOT BELOW",
"LATIN SMALL LETTER I WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH DOT BELOW",
"LATIN SMALL LETTER O WITH DOT BELOW", "LATIN CAPITAL LETTER O WITH HOOK ABOVE",
"LATIN SMALL LETTER O WITH HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND ACUTE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND ACUTE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND GRAVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND GRAVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND TILDE",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND TILDE", "LATIN CAPITAL LETTER O WITH CIRCUMFLEX AND DOT BELOW",
"LATIN SMALL LETTER O WITH CIRCUMFLEX AND DOT BELOW", "LATIN CAPITAL LETTER O WITH HORN AND ACUTE",
"LATIN SMALL LETTER O WITH HORN AND ACUTE", "LATIN CAPITAL LETTER O WITH HORN AND GRAVE",
"LATIN SMALL LETTER O WITH HORN AND GRAVE", "LATIN CAPITAL LETTER O WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER O WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER O WITH HORN AND TILDE",
"LATIN SMALL LETTER O WITH HORN AND TILDE", "LATIN CAPITAL LETTER O WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER O WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER U WITH DOT BELOW",
"LATIN SMALL LETTER U WITH DOT BELOW", "LATIN CAPITAL LETTER U WITH HOOK ABOVE",
"LATIN SMALL LETTER U WITH HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND ACUTE",
"LATIN SMALL LETTER U WITH HORN AND ACUTE", "LATIN CAPITAL LETTER U WITH HORN AND GRAVE",
"LATIN SMALL LETTER U WITH HORN AND GRAVE", "LATIN CAPITAL LETTER U WITH HORN AND HOOK ABOVE",
"LATIN SMALL LETTER U WITH HORN AND HOOK ABOVE", "LATIN CAPITAL LETTER U WITH HORN AND TILDE",
"LATIN SMALL LETTER U WITH HORN AND TILDE", "LATIN CAPITAL LETTER U WITH HORN AND DOT BELOW",
"LATIN SMALL LETTER U WITH HORN AND DOT BELOW", "LATIN CAPITAL LETTER Y WITH GRAVE",
"LATIN SMALL LETTER Y WITH GRAVE", "LATIN CAPITAL LETTER Y WITH DOT BELOW",
"LATIN SMALL LETTER Y WITH DOT BELOW", "LATIN CAPITAL LETTER Y WITH HOOK ABOVE",
"LATIN SMALL LETTER Y WITH HOOK ABOVE", "LATIN CAPITAL LETTER Y WITH TILDE",
"LATIN SMALL LETTER Y WITH TILDE"), Unicode_int = c(192L, 193L,
194L, 195L, 200L, 201L, 202L, 204L, 205L, 210L, 211L, 212L, 213L,
217L, 218L, 221L, 224L, 225L, 226L, 227L, 232L, 233L, 234L, 236L,
237L, 242L, 243L, 244L, 245L, 249L, 250L, 253L, 258L, 259L, 208L,
273L, 296L, 297L, 360L, 361L, 416L, 417L, 431L, 432L, 7840L,
7841L, 7842L, 7843L, 7844L, 7845L, 7846L, 7847L, 7848L, 7849L,
7850L, 7851L, 7852L, 7853L, 7854L, 7855L, 7856L, 7857L, 7858L,
7859L, 7860L, 7861L, 7862L, 7863L, 7864L, 7865L, 7866L, 7867L,
7868L, 7869L, 7870L, 7871L, 7872L, 7873L, 7874L, 7875L, 7876L,
7877L, 7878L, 7879L, 7880L, 7881L, 7882L, 7883L, 7884L, 7885L,
7886L, 7887L, 7888L, 7889L, 7890L, 7891L, 7892L, 7893L, 7894L,
7895L, 7896L, 7897L, 7898L, 7899L, 7900L, 7901L, 7902L, 7903L,
7904L, 7905L, 7906L, 7907L, 7908L, 7909L, 7910L, 7911L, 7912L,
7913L, 7914L, 7915L, 7916L, 7917L, 7918L, 7919L, 7920L, 7921L,
7922L, 7923L, 7924L, 7925L, 7926L, 7927L, 7928L, 7929L), TCVN3_int = list(
c(65L, 181L), c(65L, 184L), 162L, c(65L, 183L), c(69L, 204L
), c(69L, 208L), 163L, c(73L, 215L), c(73L, 221L), c(79L,
223L), c(79L, 227L), 164L, c(79L, 226L), c(85L, 239L), c(85L,
243L), c(89L, 253L), 181L, 184L, 169L, 183L, 204L, 208L,
170L, 215L, 221L, 223L, 227L, 171L, 226L, 239L, 243L, 253L,
161L, 168L, 167L, 174L, c(73L, 220L), 220L, c(85L, 242L),
242L, 165L, 172L, 166L, 173L, c(65L, 185L), 185L, c(65L,
182L), 182L, c(162L, 202L), 202L, c(162L, 199L), 199L, c(162L,
200L), 200L, c(162L, 201L), 201L, c(162L, 203L), 203L, c(161L,
190L), 190L, c(161L, 187L), 187L, c(161L, 188L), 188L, c(161L,
189L), 189L, c(161L, 198L), 198L, c(69L, 209L), 209L, c(69L,
206L), 206L, c(69L, 207L), 207L, c(163L, 213L), 213L, c(163L,
210L), 210L, c(163L, 211L), 211L, c(163L, 212L), 212L, c(163L,
214L), 214L, c(73L, 216L), 216L, c(73L, 222L), 222L, c(79L,
228L), 228L, c(79L, 225L), 225L, c(164L, 232L), 232L, c(164L,
229L), 229L, c(164L, 230L), 230L, c(164L, 231L), 231L, c(164L,
233L), 233L, c(165L, 237L), 237L, c(165L, 234L), 234L, c(165L,
235L), 235L, c(165L, 236L), 236L, c(165L, 238L), 238L, c(85L,
244L), 244L, c(85L, 241L), 241L, c(166L, 248L), 248L, c(166L,
245L), 245L, c(166L, 246L), 246L, c(166L, 247L), 247L, c(166L,
249L), 249L, c(89L, 250L), 250L, c(89L, 254L), 254L, c(89L,
251L), 251L, c(89L, 252L), 252L), VISCII_int = c(192L, 193L,
194L, 195L, 200L, 201L, 202L, 204L, 205L, 210L, 211L, 212L, NA,
217L, 218L, 221L, 224L, 225L, 226L, 227L, 232L, 233L, 234L, 236L,
237L, 242L, 243L, 244L, 245L, 249L, 250L, 253L, 197L, 229L, 208L,
240L, 206L, 238L, 157L, 251L, 180L, 189L, 191L, 223L, 8364L,
213L, 196L, 228L, 8222L, 164L, 8230L, 165L, 8224L, 166L, 6L,
231L, 8225L, 167L, 129L, 161L, 8218L, 162L, 2L, 198L, 5L, 199L,
402L, 163L, 8240L, 169L, 203L, 235L, 710L, 168L, 352L, 170L,
8249L, 171L, 338L, 172L, 141L, 173L, 381L, 174L, 8250L, 239L,
732L, 184L, 353L, 247L, 8482L, 246L, 143L, 175L, 144L, 176L,
8216L, 177L, 8217L, 178L, 8220L, 181L, 8226L, 190L, 8211L, 182L,
8212L, 183L, 179L, 222L, 8221L, 254L, 382L, 248L, 339L, 252L,
186L, 209L, 187L, 215L, 188L, 216L, 255L, 230L, 185L, 241L, 376L,
207L, 30L, 220L, 20L, 214L, 25L, 219L), VNI_int = list(c(65L,
216L), c(65L, 217L), c(65L, 194L), c(65L, 213L), c(69L, 216L),
c(69L, 217L), c(69L, 194L), 204L, 205L, c(79L, 216L), c(79L,
217L), c(79L, 194L), c(79L, 213L), c(85L, 216L), c(85L, 217L
), c(89L, 217L), c(97L, 248L), c(97L, 249L), c(97L, 226L),
c(97L, 245L), c(101L, 248L), c(101L, 249L), c(101L, 226L),
236L, 237L, c(111L, 248L), c(111L, 249L), c(111L, 226L),
c(111L, 245L), c(117L, 248L), c(117L, 249L), c(121L, 249L
), c(65L, 202L), c(97L, 234L), 209L, 241L, 211L, 243L, c(85L,
213L), c(117L, 245L), 212L, 244L, 214L, 246L, c(65L, 207L
), c(97L, 239L), c(65L, 219L), c(97L, 251L), c(65L, 193L),
c(97L, 225L), c(65L, 192L), c(97L, 224L), c(65L, 197L), c(97L,
229L), c(65L, 195L), c(97L, 227L), c(65L, 196L), c(97L, 228L
), c(65L, 201L), c(97L, 233L), c(65L, 200L), c(97L, 232L),
c(65L, 218L), c(97L, 250L), c(65L, 220L), c(97L, 252L), c(65L,
203L), c(97L, 235L), c(69L, 207L), c(101L, 239L), c(69L,
219L), c(101L, 251L), c(69L, 213L), c(101L, 245L), c(69L,
193L), c(101L, 225L), c(69L, 192L), c(101L, 224L), c(69L,
197L), c(101L, 229L), c(69L, 195L), c(101L, 227L), c(69L,
196L), c(101L, 228L), 198L, 230L, 210L, 242L, c(79L, 207L
), c(111L, 239L), c(79L, 219L), c(111L, 251L), c(79L, 193L
), c(111L, 225L), c(79L, 192L), c(111L, 224L), c(79L, 197L
), c(111L, 229L), c(79L, 195L), c(111L, 227L), c(79L, 196L
), c(111L, 228L), c(212L, 217L), c(244L, 249L), c(212L, 216L
), c(244L, 248L), c(212L, 219L), c(244L, 251L), 212:213,
244:245, c(212L, 207L), c(244L, 239L), c(85L, 207L), c(117L,
239L), c(85L, 219L), c(117L, 251L), c(214L, 217L), c(246L,
249L), c(214L, 216L), c(246L, 248L), c(214L, 219L), c(246L,
251L), 214:213, 246:245, c(214L, 207L), c(246L, 239L), c(89L,
216L), c(121L, 248L), 206L, 238L, c(89L, 219L), c(121L, 251L
), c(89L, 213L), c(121L, 245L)), VPS_int = c(8364L, 193L,
194L, 8218L, 215L, 201L, 202L, 181L, 180L, 188L, 185L, 212L,
190L, 168L, 218L, 221L, 224L, 225L, 226L, 227L, 232L, 233L, 234L,
236L, 237L, 242L, 243L, 244L, 245L, 249L, 250L, 353L, 710L, 230L,
241L, 199L, 184L, 239L, 172L, 219L, 247L, 214L, 208L, 220L, 2L,
229L, 129L, 228L, 402L, 195L, 8222L, 192L, 8230L, 196L, 28L,
197L, 3L, 198L, 141L, 161L, 381L, 162L, 143L, 163L, 240L, 164L,
4L, 165L, 5L, 203L, 222L, 200L, 254L, 235L, 144L, 8240L, 8220L,
352L, 8221L, 8249L, 8226L, 205L, 6L, 338L, 183L, 204L, 16L, 206L,
17L, 8224L, 189L, 213L, 8211L, 211L, 8212L, 210L, 732L, 176L,
8482L, 8225L, 18L, 182L, 157L, 167L, 382L, 169L, 376L, 170L,
166L, 171L, 19L, 174L, 20L, 248L, 209L, 251L, 173L, 217L, 175L,
216L, 177L, 186L, 29L, 187L, 21L, 191L, 178L, 255L, 25L, 339L,
253L, 8250L, 179L, 207L), VNU_int = c(129L, 8364L, 8222L, 195L,
200L, 201L, 8240L, 204L, 205L, 210L, 338L, 141L, 213L, 217L,
732L, 382L, 161L, 376L, 181L, 172L, 191L, 190L, 197L, 214L, 207L,
221L, 220L, 225L, 223L, 239L, 238L, 251L, 402L, 175L, 8225L,
189L, 296L, 216L, 360L, 241L, 381L, 231L, 339L, 245L, 7840L,
173L, 8218L, 168L, 8230L, 182L, 7846L, 183L, 7848L, 184L, 6L,
185L, 7852L, 186L, 7854L, 176L, 7856L, 177L, 7858L, 178L, 7860L,
179L, 7862L, 180L, 7864L, 194L, 7866L, 192L, 7868L, 193L, 7870L,
198L, 7872L, 203L, 7874L, 204L, 7876L, 205L, 7878L, 206L, 7880L,
215L, 7882L, 217L, 7884L, 224L, 7886L, 222L, 7888L, 226L, 7890L,
227L, 7892L, 228L, 7894L, 229L, 7896L, 230L, 7898L, 232L, 7900L,
233L, 143L, 234L, 7904L, 235L, 7906L, 236L, 7908L, 242L, 8482L,
240L, 157L, 246L, 7914L, 247L, 7916L, 248L, 7918L, 249L, 7920L,
250L, 7922L, 252L, 7924L, 255L, 7926L, 253L, 7928L, 254L)), row.names = c(NA,
-134L), class = "data.frame")
out$Unicode <- sapply(out$Unicode_int, intToUtf8)
out$TCVN3 <- sapply(out$TCVN3_int, intToUtf8)
out$VISCII <- sapply(out$VISCII_int, intToUtf8)
out$VNI <- sapply(out$VNI_int, intToUtf8)
out$VPS <- sapply(out$VPS_int, intToUtf8)
out$VNU <- sapply(out$VNU_int, intToUtf8)
}
return(out)
}
Try the vietnameseConverter package in your browser
Any scripts or data that you put into this service are public.
R Package Documentation
Browse R Packages
We want your feedback!
Note that we can't provide technical support on individual packages. You should contact the package authors for that.
Embedding an R snippet on your website
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.