1 use Unicode::Normalize qw( NFKD );
2 use Text::Unidecode qw( unidecode );
4 my %g; # present group classes
12 !defined $v || $_ eq $v ? 1 # identical
13 : $v =~ /\A\p{Mn}+\z/ ? 9 # combining accent
14 : $v =~ /\A[\p{Sk}\p{Lm}]+\z/ ? 8 # modifier symbol
15 : $v =~ /\A[\pM\pP]+\z/ ? 7 # mark
16 : NFKD($v) =~ /\Q$_/ ? 2 # decomposed equivalent
17 : unidecode($v) =~ /\Q$_\E+/i ? 4 # transliterated
18 : $v =~ /^\p{Latin}/ ? 5 # latin script
28 g1 => ['unaltered', "same results as without modifier"],
29 g2 => ['accented', "decomposes to the original letter with a combining accent"],
30 g4 => ['similar', "transliterates (mostly) into the unmodified letter"],
31 g5 => ['latin', "a different (accented) latin letter"],
32 g6 => ['symbol', "other character not directly deducible from key"],
33 g7 => ['punctuation', "(punctuation) mark"],
35 g8 => ['mark', "modifier letter or mark (spacing diacritic)"],
36 g9 => ['combining', "diacritical mark to be combined with a following character"],