Converts Japanese Hiragana/Katakana to Romaji for keyboard compatibility.

This transform enables passwords with Japanese words to be typed on any keyboard layout while maintaining memorability for Japanese speakers.

Use Case

• Dictionary: Japanese words (memorability in native language)
• Transform: Romaji conversion (ASCII output for compatibility)
• Result: Memorable for Japanese speakers, compatible with all systems

Examples

# Load Japanese dictionary
ExkPasswd.Dictionary.load_custom(:japanese, ["さくら", "やま", "うみ", "そら", "かぜ"])

config = ExkPasswd.Config.new!(
  dictionary: :japanese,
  word_length: 2..4,
  word_length_bounds: 1..10,
  separator: "-",
  meta: %{
    transforms: [%ExkPasswd.Transform.Romaji{}]
  }
)

ExkPasswd.generate(config)
#=> "45-sakura-yama-umi-89"
# Memorable: Japanese speaker remembers "さくら やま うみ"
# Compatible: Works on any keyboard, any system

# Katakana loanwords are also supported
ExkPasswd.Dictionary.load_custom(:katakana, ["ファイル", "コーヒー", "パーティー"])
# Converts to: fairu, koohii, paatii

Supported Scripts

• Hiragana (あいうえお...) - All standard characters
• Katakana (アイウエオ...) - All standard characters plus extended sounds
• Extended Katakana - Foreign sound combinations (ファ, ウィ, ヴァ, ティ, etc.)

Romanization Style

This implementation follows Modified Hepburn with Wāpuro input conventions, which matches how Japanese speakers type on QWERTY keyboards using IME (Input Method Editor).

Standard Hepburn Features

• Sokuon (っ/ッ): Doubles the next consonant
  • がっこう → gakkou
  • まっちゃ → matcha (special case: っち → tch)
• Palatalized sounds (ゃ/ゅ/ょ): Combines with previous consonant
  • きょう → kyou, しゃ → sha, ちゃ → cha
• ん before labials: Romanized as "m" before p, b, m
  • さんぽ → sampo, しんぶん → shimbun

Extended Features (Katakana Loanwords)

• Long vowel marker (ー): Duplicates the previous vowel for ASCII compatibility
  • コーヒー → koohii (not kōhī with macrons)
  • ラーメン → raamen, ビール → biiru
• ヴ (vu) sound: Used for "v" in foreign words
  • ヴァイオリン → vaiorin (violin)
  • ヴ → vu, ヴァ → va, ヴィ → vi, ヴェ → ve, ヴォ → vo
• Small vowel combinations: Used for foreign sounds
  • フ + small vowels: ファ → fa, フィ → fi, フェ → fe, フォ → fo
  • ウ + small vowels: ウィ → wi, ウェ → we, ウォ → wo
  • テ/デ + small イ: ティ → ti, ディ → di
  • ト/ド + small ウ: トゥ → tu, ドゥ → du
  • シ/チ/ジ + small エ: シェ → she, チェ → che, ジェ → je

Real-World Examples

ExkPasswd.Transform.apply(%Romaji{}, "ファイル", nil)     #=> "fairu" (file)
ExkPasswd.Transform.apply(%Romaji{}, "パーティー", nil)   #=> "paatii" (party)
ExkPasswd.Transform.apply(%Romaji{}, "ウィンドウ", nil)   #=> "windou" (window)
ExkPasswd.Transform.apply(%Romaji{}, "チェック", nil)     #=> "chekku" (check)
ExkPasswd.Transform.apply(%Romaji{}, "コンピューター", nil) #=> "kompyuutaa" (computer)

Limitations

Kanji Not Supported

This transform cannot convert Kanji (漢字) to romaji without additional morphological analysis or dictionary lookup. Kanji characters will be passed through unchanged.

To use Japanese words with Kanji:

1. Pre-convert to Hiragana/Katakana using your system IME or online tools
2. Use kana-only dictionaries
3. Use the contains_kanji?/1 function to detect Kanji in input

Particle Handling

The transform does not distinguish grammatical particles (は as "wa", を as "o") from their regular usage, as this requires sentence-level context. For password generation from word lists, this distinction is not necessary.

Kanji Detection

Use contains_kanji?/1 to check if text contains Kanji characters:

ExkPasswd.Transform.Romaji.contains_kanji?("桜")      #=> true
ExkPasswd.Transform.Romaji.contains_kanji?("さくら")  #=> false
ExkPasswd.Transform.Romaji.contains_kanji?("日本語")  #=> true