Hello.
I've recently updated Lingua::KO::Hangul::Util [1],
and added functions decomposing complex Jamo to simple Jamo
and composing simple Jamo to complex Jamo:
they are named decomposeJamo() and composeJamo(), resp.
(This is just putting new functions in old package.)
In comparison with a tweak against Unicode::Normalize [2],
Jamo composition works differently.
Lingua::KO::Hangul::Util::composeJamo() composes
a sequence of simple jamo having the same Hangul syllable type [3],
i.e., L{n} to L, V{n} to V, and T{n} to T.
This implementation follows K.Kim's proposal. [4]
U+1181 and U+118C are also composed properly.
Composition of U+1169,U+1167,U+1175:
<jungseong O, jungseong YEO, jungseong I>
via LKHU => U+1181.
<jungseong O-YE>
via U::N => U+1169,U+1168.
<jungseong O, jungseong YE>
In contrast, tweaked Unicode::Normalize prefers L+V composition
(L + V => LV) to V+V composition, since the algorithm by UAX#15
composes characters from left to right (in LVV, LV is leftmore than VV).
If an old (non-modern) complex jamo is included,
the result of composition is different as following:
Composition of U+1100,U+1161,U+1169:
<choseong KIYEOK, jungseong A, jungseong O>
via LKHU => U+1100,U+1176.
<choseong KIYEOK, jungseong A-O>
via U::N => U+AC00,U+1169.
<syllable GA, jungseong O>
A similar situation occurs in the case of L+V+T+T
where L+V constitutes a modern syllable:
Composition of U+1100,U+1161,U+11A8,U+11AF:
<choseong KIYEOK, jungseong A, jongseong KIYEOK, jongseong RIEUL>
via LKHU => U+1100,U+1161,U+11C3.
<choseong KIYEOK, jungseong A, jongseong KIYEOK-RIEUL>
via U::N => U+AC01,U+11AF.
<syllable GAG, jongseong RIEUL>
[1] http://search.cpan.org/author/SADAHIRO/Lingua-KO-Hangul-Util-0.20/
[2] http://www.xray.mpe.mpg.de/mailing-lists/perl-unicode/2003-04/msg00028.html
[3] http://www.unicode.org/Public/UNIDATA/HangulSyllableType.txt
[4] http://std.dkuug.dk/JTC1/SC22/WG20/docs/N954.PDF (full);
http://std.dkuug.dk/JTC1/SC22/WG20/docs/N953.PDF (summary)
regards,
SADAHIRO Tomoyuki
