icu4c/source/data/translit/Latn_Kana.txt

# © 2016 and later: Unicode, Inc. and others.
# License & terms of use: http://www.unicode.org/copyright.html
# Generated using tools/cldr/cldr-to-icu/build-icu-data.xml
#
# File: Latn_Kana.txt
# Generated from CLDR
#

# note: a global filter is more efficient, but MUST include all source chars
#:: [\u0000-\u007E 、。 \u3099-゜ ァ-ー ｡-ﾟ [:Latin:][:Katakana:] [:nonspacing mark:]] ;
# MINIMAL FILTER GENERATED FOR: Latin-Katakana
### WARNING -- must add width filter, both here and below!!! ###
:: [[ᄀ-ᄒᄚᄡ\u1160-ᅵᆪᆬ-ᆭᆰ-ᆵ←-↓│■○\u3000-。「-」\u3099-\u309Aァ-ロワヲ-ヴヷヺ-ー！-～￠-￦][',.A-Za-z~À-ÖØ-öø-ďĒ-ĥĨ-İĴ-ķĹ-ľŃ-ňŌ-őŔ-ťŨ-žƠ-ơƯ-ưǍ-ǜǞ-ǣǦ-ǭǰǴ-ǵǸ-țȞ-ȟȦ-ȳ\u0304Ӣ-ӣӮ-ӯḀ-ẙẠ-ỹᾱᾹῑῙῡῩK-Å]] ;
:: [:Latin:] fullwidth-halfwidth ();
:: NFD (NFC);
:: Lower ();    # whenever transliterating from cased to uncased script, include this
# :: NFD () ;   # this would catch the odd cases where a lowercase is not in NFD, but none are important for Japanese
# Uses modified Hepburn. Small changes to make  unambiguous.
# | Kunrei-shiki: Hepburn/MHepburn
# | ------------------------------
# | si: shi
# | si ~ya: sha
# | si ~yu: shu
# | si ~yo: sho
# | zi: ji
# | zi ~ya: ja
# | zi ~yu: ju
# | zi ~yo: jo
# | ti: chi
# | ti ~ya: cha
# | ti ~yu: chu
# | ti ~yu: cho
# | tu: tsu
# | di: ji/dji
# | du: zu/dzu
# | hu: fu
# | For foreign words:
# | -----------------
# | se ~i si
# | si ~e she
# |
# | ze ~i zi
# | zi ~e je
# |
# | te ~i ti
# | ti ~e che
# | te ~u tu
# |
# | de ~i di
# | de ~u du
# | de ~i di
# |
# | he ~u: hu
# | hu ~a fa
# | hu ~i fi
# | hu ~e he
# | hu ~o ho
# Most small forms are generated, but if necessary
# explicit small forms are given with ~a, ~ya, etc.
#------------------------------------------------------
# Variables
$vowel = [aeiou] ;
$consonant = [bcdfghjklmnpqrstvwxyz] ;
$macron = \u0304 ;
# Variables used for doubled-consonants with tsu
$kana = [ぁ-ゔ] ;
$voice = [\u3099゛];
$semivoice = [\u309A゜];
$k_start = [カキクケコかきくけこ] ;
$s_start = [サシスセソさしすせそ] ;
$j_start = [シし] $voice ;
$t_start = [タチツテトたちつてと] ;
$n_start = [ナニヌネノンなにぬねの] ;
$h_start = [ハヒヘホはひへほ] ;
$f_start = [フふ] ;
$m_start = [マミムメモまみむめも] ;
$y_start = [ヤユヨやゆよ] ;
$r_start = [ラリルレロらりるれろ] ;
$w_start = [ワヰヱヲわゐゑを] ;
$v_start = [ワヰヱヲ]\u3099 ;
$voweled_basekana = [ァ-オカキクケコサシスセソタチッツテトナ-ノハヒフヘホマ-ヲヵヶ] ;
# if ン is followed by $n_quoter, then it needs an
# apostrophe after its romaji form to disambiguate it.
# e.g., ン ア ! =  ナ, so represent as "n'a", not "na".
$n_quoter  =  [ア イ ウ エ オ ナ ニ ヌ ネ ノ ヤ ユ ヨ ン] ;
$small_y = [ャィュェョ] ;
$iteration = ゝ ;
#------------------------------------------------------
# katakana rules
# Punctuation
'.' ↔ 。;
',' ↔ 、;
# ' ' } [a-z] → ; # delete spaces before latin
# ' ' ← [^' '゠-ヿ] {} ['゠-ヿ] ; #insert spaces before hiragana
# Iteration Mark
# Copy previous letter § marks
# TODO
# | $1 $1 ← ($kana [[:M:]$voice$semivoice]?) $iteration
# Specials for katakana -- not shared with hiragana
va ↔ ワ\u3099 ;
vi ↔ ヰ\u3099 ;
ve ↔ ヱ\u3099 ;
vo ↔ ヲ\u3099 ;
'~ka' ↔ ヵ ;
'~ke' ↔ ヶ ;
# ~~~ begin shared rules ~~~
#special
ya ← '~'ャ;
yi ← '~'ィ ;
yu ← '~'ュ;
ye ← '~'ェ;
yo ← '~'ョ;
#normal
a ↔ ア ;
b | '~' ← ヒ \u3099} $small_y ;
by } $vowel → ヒ\u3099 | '~y' ;
ba ↔ ハ\u3099 ;
bi ↔ ヒ\u3099 ;
bu ↔ フ\u3099 ;
be ↔ ヘ\u3099 ;
bo ↔ ホ\u3099 ;
c } i → | s ;
c } e → | s ;
da ↔ タ\u3099 ;
di ↔ テ\u3099ィ ;
du ↔ テ\u3099ゥ ;
de ↔ テ\u3099 ;
do ↔ ト\u3099 ;
dzu ↔ ツ\u3099 ;
dja ← チ\u3099ャ ;
dji'~i' ← チ\u3099ィ ; # liu
dju ← チ\u3099ュ ;
dje ← チ\u3099ェ ;
djo ← チ\u3099ョ ;
dji ↔ チ\u3099 ;
dj  } $vowel → チ\u3099 | '~y' ;
# TODO: QUESTION: use ĵĴżŻ instead of dj, dz
cha ← チャ ;
chi'~i' ← チィ ; # liu
chu ← チュ ;
che ← チェ ;
cho ← チョ ;
chi ↔ チ ;
ch } $vowel → チ | '~y' ;
e ↔ エ ;
g | '~' ← キ\u3099} $small_y ;
gy  } $vowel → キ\u3099 | '~y' ;
ga ↔ カ\u3099 ;
gi ↔ キ\u3099 ;
gu ↔ ク\u3099 ;
ge ↔ ケ\u3099 ;
go ↔ コ\u3099 ;
i ↔ イ ;
# j  } $vowel → シ\u3099 | '~y' ;
ja ↔ シ\u3099ャ ;
ji'~i' ← シ\u3099ィ ; # liu
ju ↔ シ\u3099ュ ;
je ↔ シ\u3099ェ ;
jo ↔ シ\u3099ョ ;
ji ↔ シ\u3099 ;
k | '~' ← キ} $small_y ;
ky  } $vowel → キ | '~y' ;
ka ↔ カ ;
ki ↔ キ ;
ku ↔ ク ;
ke ↔ ケ ;
ko ↔ コ ;
m | '~' ← ミ} $small_y ;
my  } $vowel → ミ | '~y' ;
ma ↔ マ ;
mi ↔ ミ ;
mu ↔ ム ;
me ↔ メ ;
mo ↔ モ ;
m } [pbfv] → ン ;
n | '~' ← ニ } $small_y ;
ny  } $vowel → ニ | '~y' ;
na ↔ ナ ;
ni ↔ ニ ;
nu ↔ ヌ ;
ne ↔ ネ ;
no ↔ ノ ;
o ↔ オ ;
p | '~' ← ヒ\u309A } $small_y ;
py  } $vowel → ヒ\u309A | '~y' ;
pa ↔ ハ\u309A ;
pi ↔ ヒ\u309A ;
pu ↔ フ\u309A ;
pe ↔ ヘ\u309A ;
po ↔ ホ\u309A ;
h | '~' ← ヒ } $small_y ;
hy  } $vowel → ヒ | '~y' ;
ha ↔ ハ ;
hi ↔ ヒ ;
hu ↔ ヘゥ ;
he ↔ ヘ ;
ho ↔ ホ ;
# f | '~' ← フ } $small_y ;
# f } $vowel → フ | '~' ;
fa ↔ ファ ;
fi ↔ フィ ;
fe ↔ フェ ;
fo ↔ フォ ;
fu ↔ フ ;
r | '~' ← リ } $small_y ;
ry  } $vowel → リ | '~y' ;
ra ↔ ラ ;
ri ↔ リ ;
ru ↔ ル ;
re ↔ レ ;
ro ↔ ロ ;
za ↔ サ\u3099 ;
zi ↔ セ\u3099ィ ;
zu ↔ ス\u3099 ;
ze ↔ セ\u3099 ;
zo ↔ ソ\u3099 ;
sa ↔ サ ;
si ↔ セィ ;
su ↔ ス ;
se ↔ セ ;
so ↔ ソ ;
sha ← シャ ;
shi'~i' ← シィ ; # liu
shu ← シュ ;
she ← シェ ;
sho ← ショ ;
shi ↔ シ ;
sh } $vowel → シ | '~y' ;
ta ↔ タ ;
ti ↔ ティ ;
tu ↔ テゥ ;
te ↔ テ ;
to ↔ ト ;
tsu ↔ ツ ;
# v  } $vowel → ウ\u3099 | '~' ;
#'v~a' ← ウ\u3099ァ ; # liu
#'v~i' ← ウ\u3099ィ ; # liu
#'v~e' ← ウ\u3099ェ ; # liu
#'v~o' ← ウ\u3099ォ ; # liu
vu ↔ ウ\u3099 ;
u ↔ ウ ;
# w  } $vowel → ウ | '~' ;
wa ↔ ワ ;
wi ↔ ヰ ;
wu → ウ ;
we ↔ ヱ ;
wo ↔ ヲ ;
ya ↔ ヤ ;
yi → イ ;
yu ↔ ユ ;
ye → エ ;
yo ↔ ヨ ;
# double consonants
#specials
s } sh → ッ ;
t } ch → ッ ;
#voiced
j } j ↔ ッ } $j_start ;
b } b ↔ ッ } [$h_start$f_start] $voice;
d } d ↔ ッ } $t_start $voice;
g } g ↔ ッ } $k_start $voice;
p } p ↔ ッ } [$h_start$f_start] $semivoice;
# v } v ↔ ッ } [ワヰウヱヲう]  $voice ;
z } z ↔ ッ } $s_start $voice;
v } v ↔ ッ } $v_start;
# normal
k } k ↔ ッ } $k_start ;
m } m ↔ ッ } $m_start ;
n } n ↔ ッ } $n_start ;
h } h ↔ ッ } $h_start ;
f } f ↔ ッ } $f_start ;
r } r ↔ ッ } $r_start ;
t } t ↔ ッ } $t_start ;
s } s ↔ ッ } $s_start ;
w } w  ↔ ッ } $w_start;
y } y ↔ ッ } $y_start;
# completeness
x } x → ッ ;
c } k → ッ ;
c } c → ッ ;
c } q → ッ ;
l } l → ッ ;
q } q → ッ ;
# y } y → ッ ;
# w } w → ッ ;
# prolonged vowel mark. this indicates a doubling of
# the preceding vowel sound
#a ← a { ー ; # liu
#e ← e { ー ; # liu
#i ← i { ー ; # liu
#o ← o { ー ; # liu
#u ← u { ー ; # liu
$macron ↔ ー ;
# small forms
'~a' ↔ ァ ;
'~i' ↔ ィ ;
'~u' ↔ ゥ ;
'~e' ↔ ェ ;
'~o' ↔ ォ ;
'~tsu' ↔ ッ ;
'~wa' ↔ ヮ ;
'~ya' ↔ ャ ;
'~yi' → ィ ;
'~yu' ↔ ュ ;
'~ye' → ェ ;
'~yo' ↔ ョ ;
# iteration marks
# TODO: make more accurate
j $1 ← sh (y* $vowel) {ヽ$voice ;
dj $1 ← ch (y* $vowel) {ヽ$voice ;
dz $1 ← ts (y* $vowel) {ヽ$voice ;
g $1 ← k (y* $vowel) {ヽ$voice ;
z $1 ← s (y* $vowel) {ヽ$voice ;
d $1 ← t (y* $vowel) {ヽ$voice ;
h $1 ← b (y* $vowel) {ヽ$voice ;
v $1 ← w (y* $vowel) {ヽ$voice ;
sh $1 ← sh (y* $vowel) {ヽ$voice ;
j $1 ← j (y* $vowel) {ヽ$voice ;
ch $1 ← ch (y* $vowel) {ヽ$voice ;
dj $1 ← dj(y* $vowel) {ヽ$voice ;
ts $1 ← ts (y* $vowel) {ヽ$voice ;
dz $1 ← dz (y* $vowel) {ヽ$voice ;
$1 ← ($consonant y* $vowel) {ヽ$voice? ;
$1 ← (.) {ヽ $voice? ; # otherwise repeat last character
← ヽ $voice? ; # delete if no characters found
# h- rule: lengthens vowel if not followed by a vowel.
# At the point this is applied, latin [cons]?vowel sequences
# have been converted to katakana in NFD form.
$voweled_basekana [\u3099 \u309A]? { h → ー ;
# one-way latin- → kana rules. these do not occur in
# well-formed romaji representing actual japanese text.
# their purpose is to make all romaji map to kana of
# some sort.
# the following are not really necessary, but produce
# slightly more natural results.
cy → セィ ;
dy → テ\u3099ィ ;
hy → ヒ ;
sy → セィ ;
ty → ティ ;
zy → セ\u3099ィ ;
h → ヘ ;
# isolated consonants listed here so as not to mask
# longer rules above.
ch → チ;
sh → シ ;
dz → ツ\u3099 ;
dj → チ\u3099;
b → フ\u3099 ;
d → テ\u3099 ;
g → ク\u3099 ;
k → ク ;
m → ム ;
n'' ← ン } $n_quoter ;
n ↔ ン ;
p → フ\u309A ;
r → ル ;
s → ス ;
t → テ ;
y → イ ;
z → ス\u3099 ;
v → ウ\u3099 ;
f → フ;
j  → シ\u3099;
w → ウ;
ß → | ss ;
æ → | e ;
ð → | d ;
ø → | u ;
þ → | th ;
# simple substitutions using backup
c → | k ;
l → | r ;
q → | k ;
x → | ks ;
# ~~~ END shared rules ~~~
#------------------------------------------------------
# Final cleanup
'~' → ; # delete stray tildes between letters
[:Katakana:] { '' } [:Latin:] → ; # delete stray quotes between letters
# [ʾ[:Nonspacing Mark:]-[\u3099-゜]] → ; # delete any non-spacing marks that we didn't use
:: NFC (NFD) ;
:: ([[:Katakana:][\u309B\u309C\u30A0\u30FC\uFF70\uFF9E\uFF9F]] halfwidth-fullwidth);
# note: a global filter is more efficient, but MUST include all source chars!!
#:: ([\u0000-\u007E 、。 \u3099-゜ ァ-ー ｡-ﾟ [:Latin:][:Katakana:] [:nonspacing mark:]]);
# MINIMAL FILTER GENERATED FOR: Latin-Katakana BACKWARD
:: ( [[\ -~¢-£¥-¦¬\u0304₩｡-ﾾￂ-ￇￊ-ￏￒ-ￗￚ-ￜ￨-￮][~、-。がぎぐげござじずぜぞだぢづでどば-ぱび-ぴぶ-ぷべ-ぺぼ-ぽゔ\u3099-゛ゞァ-ヺー-ヾ][\u309B\u309C\u30A0\u30FC\uFF70\uFF9E\uFF9F]] ) ;
# eof