const TABLE_A: &[(char, &[&str])] = &[
('\u{00C0}', &["A"]), ('\u{00C1}', &["A"]), ('\u{00C2}', &["A"]), ('\u{00C3}', &["A"]), ('\u{00C4}', &["AE", "A"]), ('\u{00C5}', &["AA", "A"]), ('\u{00C6}', &["AE"]), ('\u{00C7}', &["C"]), ('\u{00C8}', &["E"]), ('\u{00C9}', &["E"]), ('\u{00CA}', &["E"]), ('\u{00CB}', &["E"]), ('\u{00CC}', &["I"]), ('\u{00CD}', &["I"]), ('\u{00CE}', &["I"]), ('\u{00CF}', &["I"]), ('\u{00D0}', &["D"]), ('\u{00D1}', &["N", "NXX"]), ('\u{00D2}', &["O"]), ('\u{00D3}', &["O"]), ('\u{00D4}', &["O"]), ('\u{00D5}', &["O"]), ('\u{00D6}', &["OE", "O"]), ('\u{00D8}', &["OE"]), ('\u{00D9}', &["U"]), ('\u{00DA}', &["U"]), ('\u{00DB}', &["U"]), ('\u{00DC}', &["UE", "UXX", "U"]), ('\u{00DD}', &["Y"]), ('\u{00DE}', &["TH"]), ('\u{0100}', &["A"]), ('\u{0102}', &["A"]), ('\u{0104}', &["A"]), ('\u{0106}', &["C"]), ('\u{0108}', &["C"]), ('\u{010A}', &["C"]), ('\u{010C}', &["C"]), ('\u{010E}', &["D"]), ('\u{0110}', &["D"]), ('\u{0112}', &["E"]), ('\u{0114}', &["E"]), ('\u{0116}', &["E"]), ('\u{0118}', &["E"]), ('\u{011A}', &["E"]), ('\u{011C}', &["G"]), ('\u{011E}', &["G"]), ('\u{0120}', &["G"]), ('\u{0122}', &["G"]), ('\u{0124}', &["H"]), ('\u{0126}', &["H"]), ('\u{0128}', &["I"]), ('\u{012A}', &["I"]), ('\u{012C}', &["I"]), ('\u{012E}', &["I"]), ('\u{0130}', &["I"]), ('\u{0131}', &["I"]), ('\u{0132}', &["IJ"]), ('\u{0134}', &["J"]), ('\u{0136}', &["K"]), ('\u{0139}', &["L"]), ('\u{013B}', &["L"]), ('\u{013D}', &["L"]), ('\u{013F}', &["L"]), ('\u{0141}', &["L"]), ('\u{0143}', &["N"]), ('\u{0145}', &["N"]), ('\u{0147}', &["N"]), ('\u{014A}', &["N"]), ('\u{014C}', &["O"]), ('\u{014E}', &["O"]), ('\u{0150}', &["O"]), ('\u{0152}', &["OE"]), ('\u{0154}', &["R"]), ('\u{0156}', &["R"]), ('\u{0158}', &["R"]), ('\u{015A}', &["S"]), ('\u{015C}', &["S"]), ('\u{015E}', &["S"]), ('\u{0160}', &["S"]), ('\u{0162}', &["T"]), ('\u{0164}', &["T"]), ('\u{0166}', &["T"]), ('\u{0168}', &["U"]), ('\u{016A}', &["U"]), ('\u{016C}', &["U"]), ('\u{016E}', &["U"]), ('\u{0170}', &["U"]), ('\u{0172}', &["U"]), ('\u{0174}', &["W"]), ('\u{0176}', &["Y"]), ('\u{0178}', &["Y"]), ('\u{0179}', &["Z"]), ('\u{017B}', &["Z"]), ('\u{017D}', &["Z"]), ('\u{1E9E}', &["SS"]), ];
#[non_exhaustive]
#[derive(Debug, Clone, Copy, PartialEq, Eq, Default, Hash)]
pub enum TransliterationStyle {
#[default]
Expanded,
Simple,
XxSuffix,
}
pub fn transliterations(c: char) -> &'static [&'static str] {
match TABLE_A.binary_search_by_key(&c, |&(cp, _)| cp) {
Ok(idx) => TABLE_A[idx].1,
Err(_) => &[],
}
}
pub fn transliterate_char(c: char, style: TransliterationStyle) -> Option<&'static str> {
let variants = transliterations(c);
if variants.is_empty() {
return None;
}
let selected = match (c, style) {
('\u{00C4}', TransliterationStyle::Simple) => "A",
('\u{00C4}', _) => "AE",
('\u{00C5}', TransliterationStyle::Simple) => "A",
('\u{00C5}', _) => "AA",
('\u{00D1}', TransliterationStyle::XxSuffix) => "NXX",
('\u{00D1}', _) => "N",
('\u{00D6}', TransliterationStyle::Simple) => "O",
('\u{00D6}', _) => "OE",
('\u{00DC}', TransliterationStyle::Simple) => "U",
('\u{00DC}', TransliterationStyle::XxSuffix) => "UXX",
('\u{00DC}', _) => "UE",
_ => variants[0],
};
Some(selected)
}
pub fn transliterate(s: &str, style: TransliterationStyle) -> String {
let mut out = String::with_capacity(s.len());
for c in s.chars() {
for u in c.to_uppercase() {
match transliterate_char(u, style) {
Some(t) => out.push_str(t),
None => out.push(u),
}
}
}
out
}
#[non_exhaustive]
#[derive(Debug, Clone, Copy, PartialEq, Eq, Default, Hash)]
pub enum CyrillicLanguage {
#[default]
Russian,
Belarusian,
Bulgarian,
Serbian,
Ukrainian,
Macedonian,
}
const TABLE_B: &[(char, &str)] = &[
('\u{0401}', "E"), ('\u{0402}', "D"), ('\u{0404}', "IE"), ('\u{0405}', "DZ"), ('\u{0406}', "I"), ('\u{0407}', "I"), ('\u{0408}', "J"), ('\u{0409}', "LJ"), ('\u{040A}', "NJ"), ('\u{040C}', "K"), ('\u{040E}', "U"), ('\u{040F}', "DZ"), ('\u{0410}', "A"), ('\u{0411}', "B"), ('\u{0412}', "V"), ('\u{0413}', "G"), ('\u{0414}', "D"), ('\u{0415}', "E"), ('\u{0416}', "ZH"), ('\u{0417}', "Z"), ('\u{0418}', "I"), ('\u{0419}', "I"), ('\u{041A}', "K"), ('\u{041B}', "L"), ('\u{041C}', "M"), ('\u{041D}', "N"), ('\u{041E}', "O"), ('\u{041F}', "P"), ('\u{0420}', "R"), ('\u{0421}', "S"), ('\u{0422}', "T"), ('\u{0423}', "U"), ('\u{0424}', "F"), ('\u{0425}', "KH"), ('\u{0426}', "TS"), ('\u{0427}', "CH"), ('\u{0428}', "SH"), ('\u{0429}', "SHCH"), ('\u{042A}', "IE"), ('\u{042B}', "Y"), ('\u{042D}', "E"), ('\u{042E}', "IU"), ('\u{042F}', "IA"), ('\u{046A}', "U"), ('\u{0474}', "Y"), ('\u{0490}', "G"), ('\u{0492}', "G"), ('\u{04BA}', "C"), ];
pub fn transliterate_cyrillic_char(
c: char,
lang: CyrillicLanguage,
is_first: bool,
) -> Option<&'static str> {
let base = match TABLE_B.binary_search_by_key(&c, |&(cp, _)| cp) {
Ok(idx) => TABLE_B[idx].1,
Err(_) => return None,
};
use CyrillicLanguage::*;
let selected = match c {
'\u{0401}' if lang == Belarusian => "IO",
'\u{040C}' if lang == Macedonian => "KJ",
'\u{040F}' if lang == Macedonian => "DJ",
'\u{0413}' if matches!(lang, Belarusian | Serbian | Ukrainian) => "H",
'\u{0416}' if lang == Serbian => "Z",
'\u{0418}' if lang == Ukrainian => "Y",
'\u{0425}' if matches!(lang, Serbian | Macedonian) => "H",
'\u{0426}' if matches!(lang, Serbian | Macedonian) => "C",
'\u{0427}' if lang == Serbian => "C",
'\u{0428}' if lang == Serbian => "S",
'\u{0429}' if lang == Bulgarian => "SHT",
'\u{0492}' if lang == Macedonian => "GJ",
'\u{0404}' if lang == Ukrainian && is_first => "YE",
'\u{0407}' if lang == Ukrainian && is_first => "YI",
'\u{0419}' if lang == Ukrainian && is_first => "Y",
'\u{042E}' if lang == Ukrainian && is_first => "YU",
'\u{042F}' if lang == Ukrainian && is_first => "YA",
_ => base,
};
Some(selected)
}
pub fn transliterate_cyrillic(s: &str, lang: CyrillicLanguage) -> String {
let mut out = String::with_capacity(s.len());
let mut is_first = true;
for c in s.chars() {
for u in c.to_uppercase() {
match transliterate_cyrillic_char(u, lang, is_first) {
Some(t) => out.push_str(t),
None => out.push(u),
}
is_first = false;
}
}
out
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn table_has_95_entries() {
assert_eq!(TABLE_A.len(), 95);
}
#[test]
fn table_strictly_ascending_no_duplicates() {
for pair in TABLE_A.windows(2) {
assert!(
pair[0].0 < pair[1].0,
"not strictly ascending at {:?}, {:?}",
pair[0].0,
pair[1].0
);
}
}
#[test]
fn every_variant_nonempty_and_uppercase_alpha() {
for &(cp, variants) in TABLE_A {
assert!(!variants.is_empty(), "{cp:?} has no variants");
for v in variants {
assert!(!v.is_empty(), "{cp:?} has an empty variant");
assert!(
v.chars().all(|c| c.is_ascii_uppercase()),
"{cp:?} variant {v:?} is not [A-Z]+"
);
}
}
}
#[test]
fn no_key_is_ascii() {
for &(cp, _) in TABLE_A {
assert!(!cp.is_ascii(), "{cp:?} should not be ASCII");
}
}
#[test]
fn expanded_is_always_variants_zero() {
for &(cp, variants) in TABLE_A {
assert_eq!(
transliterate_char(cp, TransliterationStyle::Expanded),
Some(variants[0]),
"Expanded mismatch at {cp:?}"
);
}
}
#[test]
fn multi_valued_set_is_exactly_five() {
let multi: Vec<char> = TABLE_A
.iter()
.filter(|&&(_, v)| v.len() > 1)
.map(|&(cp, _)| cp)
.collect();
assert_eq!(
multi,
vec!['\u{00C4}', '\u{00C5}', '\u{00D1}', '\u{00D6}', '\u{00DC}']
);
}
#[test]
fn lowercase_round_trip() {
for &(cp, _) in TABLE_A {
if cp == '\u{0130}' {
continue;
}
let upper_str = cp.to_string();
let lower_str: String = cp.to_lowercase().collect();
assert_eq!(
transliterate(&lower_str, TransliterationStyle::Expanded),
transliterate(&upper_str, TransliterationStyle::Expanded),
"lowercase of {cp:?} ({lower_str:?}) does not round-trip"
);
}
}
#[test]
fn sharp_s_uppercases_to_ss() {
assert_eq!(transliterate("ß", TransliterationStyle::Expanded), "SS");
assert_eq!(transliterate("ß", TransliterationStyle::Simple), "SS");
assert_eq!(transliterate("ß", TransliterationStyle::XxSuffix), "SS");
assert_eq!(
transliterate_char('\u{1E9E}', TransliterationStyle::Expanded),
Some("SS")
);
}
#[test]
fn golden_vector_teresa_canon() {
assert_eq!(
transliterate("Térèsa", TransliterationStyle::XxSuffix),
"TERESA"
);
assert_eq!(
transliterate("CAÑON", TransliterationStyle::XxSuffix),
"CANXXON"
);
}
#[test]
fn not_in_table_returns_empty_and_none() {
assert_eq!(transliterations('A'), &[] as &[&str]);
assert_eq!(
transliterate_char('A', TransliterationStyle::Expanded),
None
);
assert_eq!(transliterate("A1!", TransliterationStyle::Expanded), "A1!");
}
const LANGS: &[CyrillicLanguage] = &[
CyrillicLanguage::Russian,
CyrillicLanguage::Belarusian,
CyrillicLanguage::Bulgarian,
CyrillicLanguage::Serbian,
CyrillicLanguage::Ukrainian,
CyrillicLanguage::Macedonian,
];
#[test]
fn table_b_has_48_entries() {
assert_eq!(TABLE_B.len(), 48);
}
#[test]
fn table_b_strictly_ascending_no_duplicates() {
for pair in TABLE_B.windows(2) {
assert!(
pair[0].0 < pair[1].0,
"not strictly ascending at {:?}, {:?}",
pair[0].0,
pair[1].0
);
}
}
#[test]
fn table_b_keys_are_cyrillic_block() {
for &(cp, _) in TABLE_B {
assert!(!cp.is_ascii(), "{cp:?} should not be ASCII");
let n = cp as u32;
assert!(
(0x0400..=0x04FF).contains(&n),
"{cp:?} (U+{n:04X}) is outside the Cyrillic block"
);
}
}
#[test]
fn table_b_base_output_is_nonempty_upper_ascii() {
for &(cp, base) in TABLE_B {
assert!(!base.is_empty(), "{cp:?} has an empty base transliteration");
assert!(
base.chars().all(|c| c.is_ascii_uppercase()),
"{cp:?} base {base:?} is not [A-Z]+"
);
}
}
#[test]
fn every_conditional_output_is_upper_ascii() {
for &(cp, _) in TABLE_B {
for &lang in LANGS {
for is_first in [false, true] {
let out = transliterate_cyrillic_char(cp, lang, is_first)
.expect("every TABLE_B key resolves for every language");
assert!(
!out.is_empty() && out.chars().all(|c| c.is_ascii_uppercase()),
"{cp:?} / {lang:?} / first={is_first} → {out:?} is not [A-Z]+"
);
}
}
}
}
#[test]
fn conditional_set_is_exactly_the_seventeen_documented_rows() {
let mut conditional: Vec<char> = TABLE_B
.iter()
.filter(|&&(cp, base)| {
LANGS.iter().any(|&lang| {
[false, true]
.iter()
.any(|&f| transliterate_cyrillic_char(cp, lang, f) != Some(base))
})
})
.map(|&(cp, _)| cp)
.collect();
conditional.sort();
assert_eq!(
conditional,
vec![
'\u{0401}', '\u{0404}', '\u{0407}', '\u{040C}', '\u{040F}', '\u{0413}', '\u{0416}', '\u{0418}', '\u{0419}', '\u{0425}', '\u{0426}', '\u{0427}', '\u{0428}', '\u{0429}', '\u{042E}', '\u{042F}', '\u{0492}', ]
);
assert_eq!(conditional.len(), 17);
}
#[test]
fn cyrillic_lowercase_round_trips() {
for &(cp, _) in TABLE_B {
let upper = cp.to_string();
let lower: String = cp.to_lowercase().collect();
for &lang in LANGS {
assert_eq!(
transliterate_cyrillic(&lower, lang),
transliterate_cyrillic(&upper, lang),
"lowercase of {cp:?} does not round-trip for {lang:?}"
);
}
}
}
#[test]
fn cyrillic_non_table_chars_pass_through() {
assert_eq!(
transliterate_cyrillic("ABC-1", CyrillicLanguage::Russian),
"ABC-1"
);
assert_eq!(
transliterate_cyrillic("\u{0470}", CyrillicLanguage::Russian),
"\u{0470}"
);
assert_eq!(
transliterate_cyrillic_char('A', CyrillicLanguage::Russian, false),
None
);
}
#[test]
fn cyrillic_base_column_vectors() {
let cases: &[(&str, &str)] = &[
("Иванов", "IVANOV"),
("Смирнова", "SMIRNOVA"),
("Жуков", "ZHUKOV"), ("Щедрин", "SHCHEDRIN"), ("Хабаров", "KHABAROV"), ("Цветаева", "TSVETAEVA"),
];
for (native, expected) in cases {
assert_eq!(
transliterate_cyrillic(native, CyrillicLanguage::Russian),
*expected,
"{native}"
);
}
}
#[test]
fn cyrillic_language_conditional_vectors() {
assert_eq!(
transliterate_cyrillic("Живков", CyrillicLanguage::Serbian),
"ZIVKOV"
);
assert_eq!(
transliterate_cyrillic("Чачак", CyrillicLanguage::Serbian),
"CACAK"
);
assert_eq!(
transliterate_cyrillic("Щерев", CyrillicLanguage::Bulgarian),
"SHTEREV"
);
assert_eq!(
transliterate_cyrillic("Христов", CyrillicLanguage::Bulgarian),
"KHRISTOV"
);
assert_eq!(
transliterate_cyrillic("Ќоста", CyrillicLanguage::Macedonian),
"KJOSTA"
);
assert_eq!(
transliterate_cyrillic("Гончарова", CyrillicLanguage::Belarusian),
"HONCHAROVA"
);
}
#[test]
fn cyrillic_ukrainian_positional_vectors() {
assert_eq!(
transliterate_cyrillic("Євген", CyrillicLanguage::Ukrainian),
"YEVHEN" );
assert_eq!(
transliterate_cyrillic("Юрій", CyrillicLanguage::Ukrainian),
"YURII" );
assert_eq!(
transliterate_cyrillic("Гаєвий", CyrillicLanguage::Ukrainian),
"HAIEVYI" );
assert_eq!(
transliterate_cyrillic("Мирний", CyrillicLanguage::Ukrainian),
"MYRNYI"
);
assert_eq!(
transliterate_cyrillic("Мирний", CyrillicLanguage::Russian),
"MIRNII"
);
}
}