mirror of
https://github.com/postgres/postgres.git
synced 2025-11-10 17:42:29 +03:00
Pre-beta mechanical code beautification.
Run pgindent, pgperltidy, and reformat-dat-files. This set of diffs is a bit larger than typical. We've updated to pg_bsd_indent 2.1.2, which properly indents variable declarations that have multi-line initialization expressions (the continuation lines are now indented one tab stop). We've also updated to perltidy version 20230309 and changed some of its settings, which reduces its desire to add whitespace to lines to make assignments etc. line up. Going forward, that should make for fewer random-seeming changes to existing code. Discussion: https://postgr.es/m/20230428092545.qfb3y5wcu4cm75ur@alvherre.pgsql
This commit is contained in:
Tom Lane
@@ -40,7 +40,7 @@ my $cp950txt = &read_source("CP950.TXT");
|
||||
foreach my $i (@$cp950txt)
|
||||
{
|
||||
my $code = $i->{code};
|
||||
my $ucs = $i->{ucs};
|
||||
my $ucs = $i->{ucs};
|
||||
|
||||
# Pick only the ETEN extended characters in the range 0xf9d6 - 0xf9dc
|
||||
# from CP950.TXT
|
||||
@@ -51,12 +51,12 @@ foreach my $i (@$cp950txt)
|
||||
{
|
||||
push @$all,
|
||||
{
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
comment => $i->{comment},
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
comment => $i->{comment},
|
||||
direction => BOTH,
|
||||
f => $i->{f},
|
||||
l => $i->{l}
|
||||
f => $i->{f},
|
||||
l => $i->{l}
|
||||
};
|
||||
}
|
||||
}
|
||||
@@ -64,7 +64,7 @@ foreach my $i (@$cp950txt)
|
||||
foreach my $i (@$all)
|
||||
{
|
||||
my $code = $i->{code};
|
||||
my $ucs = $i->{ucs};
|
||||
my $ucs = $i->{ucs};
|
||||
|
||||
# BIG5.TXT maps several BIG5 characters to U+FFFD. The UTF-8 to BIG5 mapping can
|
||||
# contain only one of them. XXX: Doesn't really make sense to include any of them,
|
||||
|
||||
@@ -33,7 +33,7 @@ while (<$in>)
|
||||
next if (!m/<a u="([0-9A-F]+)" b="([0-9A-F ]+)"/);
|
||||
my ($u, $c) = ($1, $2);
|
||||
$c =~ s/ //g;
|
||||
my $ucs = hex($u);
|
||||
my $ucs = hex($u);
|
||||
my $code = hex($c);
|
||||
|
||||
# The GB-18030 character set, which we use as the source, contains
|
||||
@@ -73,11 +73,11 @@ while (<$in>)
|
||||
|
||||
push @mapping,
|
||||
{
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
direction => BOTH,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
close($in);
|
||||
|
||||
@@ -37,13 +37,13 @@ while (my $line = <$in>)
|
||||
|
||||
push @all,
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => $ucs1,
|
||||
direction => BOTH,
|
||||
ucs => $ucs1,
|
||||
ucs_second => $ucs2,
|
||||
code => $code,
|
||||
comment => $rest,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
code => $code,
|
||||
comment => $rest,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
elsif ($line =~ /^0x(\w+)\s*U\+(\w+)\s*#\s*(\S.*)?\s*$/)
|
||||
@@ -51,7 +51,7 @@ while (my $line = <$in>)
|
||||
|
||||
# non-combined characters
|
||||
my ($c, $u, $rest) = ($1, $2, "U+" . $2 . $3);
|
||||
my $ucs = hex($u);
|
||||
my $ucs = hex($u);
|
||||
my $code = hex($c);
|
||||
|
||||
next if ($code < 0x80 && $ucs < 0x80);
|
||||
@@ -59,11 +59,11 @@ while (my $line = <$in>)
|
||||
push @all,
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
comment => $rest,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
comment => $rest,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
}
|
||||
|
||||
@@ -120,521 +120,521 @@ foreach my $i (grep defined $_->{sjis}, @mapping)
|
||||
push @mapping, (
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x4efc,
|
||||
code => 0x8ff4af,
|
||||
comment => '# CJK(4EFC)'
|
||||
ucs => 0x4efc,
|
||||
code => 0x8ff4af,
|
||||
comment => '# CJK(4EFC)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x50f4,
|
||||
code => 0x8ff4b0,
|
||||
comment => '# CJK(50F4)'
|
||||
ucs => 0x50f4,
|
||||
code => 0x8ff4b0,
|
||||
comment => '# CJK(50F4)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x51EC,
|
||||
code => 0x8ff4b1,
|
||||
comment => '# CJK(51EC)'
|
||||
ucs => 0x51EC,
|
||||
code => 0x8ff4b1,
|
||||
comment => '# CJK(51EC)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5307,
|
||||
code => 0x8ff4b2,
|
||||
comment => '# CJK(5307)'
|
||||
ucs => 0x5307,
|
||||
code => 0x8ff4b2,
|
||||
comment => '# CJK(5307)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5324,
|
||||
code => 0x8ff4b3,
|
||||
comment => '# CJK(5324)'
|
||||
ucs => 0x5324,
|
||||
code => 0x8ff4b3,
|
||||
comment => '# CJK(5324)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x548A,
|
||||
code => 0x8ff4b5,
|
||||
comment => '# CJK(548A)'
|
||||
ucs => 0x548A,
|
||||
code => 0x8ff4b5,
|
||||
comment => '# CJK(548A)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5759,
|
||||
code => 0x8ff4b6,
|
||||
comment => '# CJK(5759)'
|
||||
ucs => 0x5759,
|
||||
code => 0x8ff4b6,
|
||||
comment => '# CJK(5759)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x589E,
|
||||
code => 0x8ff4b9,
|
||||
comment => '# CJK(589E)'
|
||||
ucs => 0x589E,
|
||||
code => 0x8ff4b9,
|
||||
comment => '# CJK(589E)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5BEC,
|
||||
code => 0x8ff4ba,
|
||||
comment => '# CJK(5BEC)'
|
||||
ucs => 0x5BEC,
|
||||
code => 0x8ff4ba,
|
||||
comment => '# CJK(5BEC)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5CF5,
|
||||
code => 0x8ff4bb,
|
||||
comment => '# CJK(5CF5)'
|
||||
ucs => 0x5CF5,
|
||||
code => 0x8ff4bb,
|
||||
comment => '# CJK(5CF5)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5D53,
|
||||
code => 0x8ff4bc,
|
||||
comment => '# CJK(5D53)'
|
||||
ucs => 0x5D53,
|
||||
code => 0x8ff4bc,
|
||||
comment => '# CJK(5D53)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x5FB7,
|
||||
code => 0x8ff4be,
|
||||
comment => '# CJK(5FB7)'
|
||||
ucs => 0x5FB7,
|
||||
code => 0x8ff4be,
|
||||
comment => '# CJK(5FB7)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6085,
|
||||
code => 0x8ff4bf,
|
||||
comment => '# CJK(6085)'
|
||||
ucs => 0x6085,
|
||||
code => 0x8ff4bf,
|
||||
comment => '# CJK(6085)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6120,
|
||||
code => 0x8ff4c0,
|
||||
comment => '# CJK(6120)'
|
||||
ucs => 0x6120,
|
||||
code => 0x8ff4c0,
|
||||
comment => '# CJK(6120)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x654E,
|
||||
code => 0x8ff4c1,
|
||||
comment => '# CJK(654E)'
|
||||
ucs => 0x654E,
|
||||
code => 0x8ff4c1,
|
||||
comment => '# CJK(654E)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x663B,
|
||||
code => 0x8ff4c2,
|
||||
comment => '# CJK(663B)'
|
||||
ucs => 0x663B,
|
||||
code => 0x8ff4c2,
|
||||
comment => '# CJK(663B)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6665,
|
||||
code => 0x8ff4c3,
|
||||
comment => '# CJK(6665)'
|
||||
ucs => 0x6665,
|
||||
code => 0x8ff4c3,
|
||||
comment => '# CJK(6665)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6801,
|
||||
code => 0x8ff4c6,
|
||||
comment => '# CJK(6801)'
|
||||
ucs => 0x6801,
|
||||
code => 0x8ff4c6,
|
||||
comment => '# CJK(6801)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6A6B,
|
||||
code => 0x8ff4c9,
|
||||
comment => '# CJK(6A6B)'
|
||||
ucs => 0x6A6B,
|
||||
code => 0x8ff4c9,
|
||||
comment => '# CJK(6A6B)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6AE2,
|
||||
code => 0x8ff4ca,
|
||||
comment => '# CJK(6AE2)'
|
||||
ucs => 0x6AE2,
|
||||
code => 0x8ff4ca,
|
||||
comment => '# CJK(6AE2)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6DF2,
|
||||
code => 0x8ff4cc,
|
||||
comment => '# CJK(6DF2)'
|
||||
ucs => 0x6DF2,
|
||||
code => 0x8ff4cc,
|
||||
comment => '# CJK(6DF2)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x6DF8,
|
||||
code => 0x8ff4cb,
|
||||
comment => '# CJK(6DF8)'
|
||||
ucs => 0x6DF8,
|
||||
code => 0x8ff4cb,
|
||||
comment => '# CJK(6DF8)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7028,
|
||||
code => 0x8ff4cd,
|
||||
comment => '# CJK(7028)'
|
||||
ucs => 0x7028,
|
||||
code => 0x8ff4cd,
|
||||
comment => '# CJK(7028)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x70BB,
|
||||
code => 0x8ff4ae,
|
||||
comment => '# CJK(70BB)'
|
||||
ucs => 0x70BB,
|
||||
code => 0x8ff4ae,
|
||||
comment => '# CJK(70BB)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7501,
|
||||
code => 0x8ff4d0,
|
||||
comment => '# CJK(7501)'
|
||||
ucs => 0x7501,
|
||||
code => 0x8ff4d0,
|
||||
comment => '# CJK(7501)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7682,
|
||||
code => 0x8ff4d1,
|
||||
comment => '# CJK(7682)'
|
||||
ucs => 0x7682,
|
||||
code => 0x8ff4d1,
|
||||
comment => '# CJK(7682)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x769E,
|
||||
code => 0x8ff4d2,
|
||||
comment => '# CJK(769E)'
|
||||
ucs => 0x769E,
|
||||
code => 0x8ff4d2,
|
||||
comment => '# CJK(769E)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7930,
|
||||
code => 0x8ff4d4,
|
||||
comment => '# CJK(7930)'
|
||||
ucs => 0x7930,
|
||||
code => 0x8ff4d4,
|
||||
comment => '# CJK(7930)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7AE7,
|
||||
code => 0x8ff4d9,
|
||||
comment => '# CJK(7AE7)'
|
||||
ucs => 0x7AE7,
|
||||
code => 0x8ff4d9,
|
||||
comment => '# CJK(7AE7)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7DA0,
|
||||
code => 0x8ff4dc,
|
||||
comment => '# CJK(7DA0)'
|
||||
ucs => 0x7DA0,
|
||||
code => 0x8ff4dc,
|
||||
comment => '# CJK(7DA0)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x7DD6,
|
||||
code => 0x8ff4dd,
|
||||
comment => '# CJK(7DD6)'
|
||||
ucs => 0x7DD6,
|
||||
code => 0x8ff4dd,
|
||||
comment => '# CJK(7DD6)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x8362,
|
||||
code => 0x8ff4df,
|
||||
comment => '# CJK(8362)'
|
||||
ucs => 0x8362,
|
||||
code => 0x8ff4df,
|
||||
comment => '# CJK(8362)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x85B0,
|
||||
code => 0x8ff4e1,
|
||||
comment => '# CJK(85B0)'
|
||||
ucs => 0x85B0,
|
||||
code => 0x8ff4e1,
|
||||
comment => '# CJK(85B0)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x8807,
|
||||
code => 0x8ff4e4,
|
||||
comment => '# CJK(8807)'
|
||||
ucs => 0x8807,
|
||||
code => 0x8ff4e4,
|
||||
comment => '# CJK(8807)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x8B7F,
|
||||
code => 0x8ff4e6,
|
||||
comment => '# CJK(8B7F)'
|
||||
ucs => 0x8B7F,
|
||||
code => 0x8ff4e6,
|
||||
comment => '# CJK(8B7F)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x8CF4,
|
||||
code => 0x8ff4e7,
|
||||
comment => '# CJK(8CF4)'
|
||||
ucs => 0x8CF4,
|
||||
code => 0x8ff4e7,
|
||||
comment => '# CJK(8CF4)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x8D76,
|
||||
code => 0x8ff4e8,
|
||||
comment => '# CJK(8D76)'
|
||||
ucs => 0x8D76,
|
||||
code => 0x8ff4e8,
|
||||
comment => '# CJK(8D76)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x90DE,
|
||||
code => 0x8ff4ec,
|
||||
comment => '# CJK(90DE)'
|
||||
ucs => 0x90DE,
|
||||
code => 0x8ff4ec,
|
||||
comment => '# CJK(90DE)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x9115,
|
||||
code => 0x8ff4ee,
|
||||
comment => '# CJK(9115)'
|
||||
ucs => 0x9115,
|
||||
code => 0x8ff4ee,
|
||||
comment => '# CJK(9115)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x9592,
|
||||
code => 0x8ff4f1,
|
||||
comment => '# CJK(9592)'
|
||||
ucs => 0x9592,
|
||||
code => 0x8ff4f1,
|
||||
comment => '# CJK(9592)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x973B,
|
||||
code => 0x8ff4f4,
|
||||
comment => '# CJK(973B)'
|
||||
ucs => 0x973B,
|
||||
code => 0x8ff4f4,
|
||||
comment => '# CJK(973B)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x974D,
|
||||
code => 0x8ff4f5,
|
||||
comment => '# CJK(974D)'
|
||||
ucs => 0x974D,
|
||||
code => 0x8ff4f5,
|
||||
comment => '# CJK(974D)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x9751,
|
||||
code => 0x8ff4f6,
|
||||
comment => '# CJK(9751)'
|
||||
ucs => 0x9751,
|
||||
code => 0x8ff4f6,
|
||||
comment => '# CJK(9751)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x999E,
|
||||
code => 0x8ff4fa,
|
||||
comment => '# CJK(999E)'
|
||||
ucs => 0x999E,
|
||||
code => 0x8ff4fa,
|
||||
comment => '# CJK(999E)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x9AD9,
|
||||
code => 0x8ff4fb,
|
||||
comment => '# CJK(9AD9)'
|
||||
ucs => 0x9AD9,
|
||||
code => 0x8ff4fb,
|
||||
comment => '# CJK(9AD9)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x9B72,
|
||||
code => 0x8ff4fc,
|
||||
comment => '# CJK(9B72)'
|
||||
ucs => 0x9B72,
|
||||
code => 0x8ff4fc,
|
||||
comment => '# CJK(9B72)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x9ED1,
|
||||
code => 0x8ff4fe,
|
||||
comment => '# CJK(9ED1)'
|
||||
ucs => 0x9ED1,
|
||||
code => 0x8ff4fe,
|
||||
comment => '# CJK(9ED1)'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xF929,
|
||||
code => 0x8ff4c5,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-F929'
|
||||
ucs => 0xF929,
|
||||
code => 0x8ff4c5,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-F929'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xF9DC,
|
||||
code => 0x8ff4f2,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-F9DC'
|
||||
ucs => 0xF9DC,
|
||||
code => 0x8ff4f2,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-F9DC'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA0E,
|
||||
code => 0x8ff4b4,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA0E'
|
||||
ucs => 0xFA0E,
|
||||
code => 0x8ff4b4,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA0E'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA0F,
|
||||
code => 0x8ff4b7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA0F'
|
||||
ucs => 0xFA0F,
|
||||
code => 0x8ff4b7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA0F'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA10,
|
||||
code => 0x8ff4b8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA10'
|
||||
ucs => 0xFA10,
|
||||
code => 0x8ff4b8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA10'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA11,
|
||||
code => 0x8ff4bd,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA11'
|
||||
ucs => 0xFA11,
|
||||
code => 0x8ff4bd,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA11'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA12,
|
||||
code => 0x8ff4c4,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA12'
|
||||
ucs => 0xFA12,
|
||||
code => 0x8ff4c4,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA12'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA13,
|
||||
code => 0x8ff4c7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA13'
|
||||
ucs => 0xFA13,
|
||||
code => 0x8ff4c7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA13'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA14,
|
||||
code => 0x8ff4c8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA14'
|
||||
ucs => 0xFA14,
|
||||
code => 0x8ff4c8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA14'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA15,
|
||||
code => 0x8ff4ce,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA15'
|
||||
ucs => 0xFA15,
|
||||
code => 0x8ff4ce,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA15'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA16,
|
||||
code => 0x8ff4cf,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA16'
|
||||
ucs => 0xFA16,
|
||||
code => 0x8ff4cf,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA16'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA17,
|
||||
code => 0x8ff4d3,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA17'
|
||||
ucs => 0xFA17,
|
||||
code => 0x8ff4d3,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA17'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA18,
|
||||
code => 0x8ff4d5,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA18'
|
||||
ucs => 0xFA18,
|
||||
code => 0x8ff4d5,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA18'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA19,
|
||||
code => 0x8ff4d6,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA19'
|
||||
ucs => 0xFA19,
|
||||
code => 0x8ff4d6,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA19'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA1A,
|
||||
code => 0x8ff4d7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1A'
|
||||
ucs => 0xFA1A,
|
||||
code => 0x8ff4d7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1A'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA1B,
|
||||
code => 0x8ff4d8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1B'
|
||||
ucs => 0xFA1B,
|
||||
code => 0x8ff4d8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1B'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA1C,
|
||||
code => 0x8ff4da,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1C'
|
||||
ucs => 0xFA1C,
|
||||
code => 0x8ff4da,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1C'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA1D,
|
||||
code => 0x8ff4db,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1D'
|
||||
ucs => 0xFA1D,
|
||||
code => 0x8ff4db,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1D'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA1E,
|
||||
code => 0x8ff4de,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1E'
|
||||
ucs => 0xFA1E,
|
||||
code => 0x8ff4de,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1E'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA1F,
|
||||
code => 0x8ff4e0,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1F'
|
||||
ucs => 0xFA1F,
|
||||
code => 0x8ff4e0,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA1F'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA20,
|
||||
code => 0x8ff4e2,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA20'
|
||||
ucs => 0xFA20,
|
||||
code => 0x8ff4e2,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA20'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA21,
|
||||
code => 0x8ff4e3,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA21'
|
||||
ucs => 0xFA21,
|
||||
code => 0x8ff4e3,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA21'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA22,
|
||||
code => 0x8ff4e5,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA22'
|
||||
ucs => 0xFA22,
|
||||
code => 0x8ff4e5,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA22'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA23,
|
||||
code => 0x8ff4e9,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA23'
|
||||
ucs => 0xFA23,
|
||||
code => 0x8ff4e9,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA23'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA24,
|
||||
code => 0x8ff4ea,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA24'
|
||||
ucs => 0xFA24,
|
||||
code => 0x8ff4ea,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA24'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA25,
|
||||
code => 0x8ff4eb,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA25'
|
||||
ucs => 0xFA25,
|
||||
code => 0x8ff4eb,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA25'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA26,
|
||||
code => 0x8ff4ed,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA26'
|
||||
ucs => 0xFA26,
|
||||
code => 0x8ff4ed,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA26'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA27,
|
||||
code => 0x8ff4ef,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA27'
|
||||
ucs => 0xFA27,
|
||||
code => 0x8ff4ef,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA27'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA28,
|
||||
code => 0x8ff4f0,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA28'
|
||||
ucs => 0xFA28,
|
||||
code => 0x8ff4f0,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA28'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA29,
|
||||
code => 0x8ff4f3,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA29'
|
||||
ucs => 0xFA29,
|
||||
code => 0x8ff4f3,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA29'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA2A,
|
||||
code => 0x8ff4f7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2A'
|
||||
ucs => 0xFA2A,
|
||||
code => 0x8ff4f7,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2A'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA2B,
|
||||
code => 0x8ff4f8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2B'
|
||||
ucs => 0xFA2B,
|
||||
code => 0x8ff4f8,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2B'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA2C,
|
||||
code => 0x8ff4f9,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2C'
|
||||
ucs => 0xFA2C,
|
||||
code => 0x8ff4f9,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2C'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFA2D,
|
||||
code => 0x8ff4fd,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2D'
|
||||
ucs => 0xFA2D,
|
||||
code => 0x8ff4fd,
|
||||
comment => '# CJK COMPATIBILITY IDEOGRAPH-FA2D'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFF07,
|
||||
code => 0x8ff4a9,
|
||||
comment => '# FULLWIDTH APOSTROPHE'
|
||||
ucs => 0xFF07,
|
||||
code => 0x8ff4a9,
|
||||
comment => '# FULLWIDTH APOSTROPHE'
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0xFFE4,
|
||||
code => 0x8fa2c3,
|
||||
comment => '# FULLWIDTH BROKEN BAR'
|
||||
ucs => 0xFFE4,
|
||||
code => 0x8fa2c3,
|
||||
comment => '# FULLWIDTH BROKEN BAR'
|
||||
},
|
||||
|
||||
# additional conversions for EUC_JP -> UTF-8 conversion
|
||||
{
|
||||
direction => TO_UNICODE,
|
||||
ucs => 0x2116,
|
||||
code => 0x8ff4ac,
|
||||
comment => '# NUMERO SIGN'
|
||||
ucs => 0x2116,
|
||||
code => 0x8ff4ac,
|
||||
comment => '# NUMERO SIGN'
|
||||
},
|
||||
{
|
||||
direction => TO_UNICODE,
|
||||
ucs => 0x2121,
|
||||
code => 0x8ff4ad,
|
||||
comment => '# TELEPHONE SIGN'
|
||||
ucs => 0x2121,
|
||||
code => 0x8ff4ad,
|
||||
comment => '# TELEPHONE SIGN'
|
||||
},
|
||||
{
|
||||
direction => TO_UNICODE,
|
||||
ucs => 0x3231,
|
||||
code => 0x8ff4ab,
|
||||
comment => '# PARENTHESIZED IDEOGRAPH STOCK'
|
||||
ucs => 0x3231,
|
||||
code => 0x8ff4ab,
|
||||
comment => '# PARENTHESIZED IDEOGRAPH STOCK'
|
||||
});
|
||||
|
||||
print_conversion_tables($this_script, "EUC_JP", \@mapping);
|
||||
|
||||
@@ -36,27 +36,27 @@ foreach my $i (@$mapping)
|
||||
push @$mapping,
|
||||
( {
|
||||
direction => BOTH,
|
||||
ucs => 0x20AC,
|
||||
code => 0xa2e6,
|
||||
comment => '# EURO SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x20AC,
|
||||
code => 0xa2e6,
|
||||
comment => '# EURO SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x00AE,
|
||||
code => 0xa2e7,
|
||||
comment => '# REGISTERED SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x00AE,
|
||||
code => 0xa2e7,
|
||||
comment => '# REGISTERED SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x327E,
|
||||
code => 0xa2e8,
|
||||
comment => '# CIRCLED HANGUL IEUNG U',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x327E,
|
||||
code => 0xa2e8,
|
||||
comment => '# CIRCLED HANGUL IEUNG U',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
});
|
||||
|
||||
print_conversion_tables($this_script, "EUC_KR", $mapping);
|
||||
|
||||
@@ -30,8 +30,8 @@ my @extras;
|
||||
|
||||
foreach my $i (@$mapping)
|
||||
{
|
||||
my $ucs = $i->{ucs};
|
||||
my $code = $i->{code};
|
||||
my $ucs = $i->{ucs};
|
||||
my $code = $i->{code};
|
||||
my $origcode = $i->{code};
|
||||
|
||||
my $plane = ($code & 0x1f0000) >> 16;
|
||||
@@ -56,12 +56,12 @@ foreach my $i (@$mapping)
|
||||
{
|
||||
push @extras,
|
||||
{
|
||||
ucs => $i->{ucs},
|
||||
code => ($i->{code} + 0x8ea10000),
|
||||
rest => $i->{rest},
|
||||
ucs => $i->{ucs},
|
||||
code => ($i->{code} + 0x8ea10000),
|
||||
rest => $i->{rest},
|
||||
direction => TO_UNICODE,
|
||||
f => $i->{f},
|
||||
l => $i->{l}
|
||||
f => $i->{f},
|
||||
l => $i->{l}
|
||||
};
|
||||
}
|
||||
}
|
||||
|
||||
@@ -33,17 +33,17 @@ while (<$in>)
|
||||
next if (!m/<a u="([0-9A-F]+)" b="([0-9A-F ]+)"/);
|
||||
my ($u, $c) = ($1, $2);
|
||||
$c =~ s/ //g;
|
||||
my $ucs = hex($u);
|
||||
my $ucs = hex($u);
|
||||
my $code = hex($c);
|
||||
if ($code >= 0x80 && $ucs >= 0x0080)
|
||||
{
|
||||
push @mapping,
|
||||
{
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
direction => BOTH,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
}
|
||||
|
||||
@@ -30,27 +30,27 @@ my $mapping = &read_source("JOHAB.TXT");
|
||||
push @$mapping,
|
||||
( {
|
||||
direction => BOTH,
|
||||
ucs => 0x20AC,
|
||||
code => 0xd9e6,
|
||||
comment => '# EURO SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x20AC,
|
||||
code => 0xd9e6,
|
||||
comment => '# EURO SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x00AE,
|
||||
code => 0xd9e7,
|
||||
comment => '# REGISTERED SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x00AE,
|
||||
code => 0xd9e7,
|
||||
comment => '# REGISTERED SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => BOTH,
|
||||
ucs => 0x327E,
|
||||
code => 0xd9e8,
|
||||
comment => '# CIRCLED HANGUL IEUNG U',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x327E,
|
||||
code => 0xd9e8,
|
||||
comment => '# CIRCLED HANGUL IEUNG U',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
});
|
||||
|
||||
print_conversion_tables($this_script, "JOHAB", $mapping);
|
||||
|
||||
@@ -37,13 +37,13 @@ while (my $line = <$in>)
|
||||
|
||||
push @mapping,
|
||||
{
|
||||
code => $code,
|
||||
ucs => $ucs1,
|
||||
code => $code,
|
||||
ucs => $ucs1,
|
||||
ucs_second => $ucs2,
|
||||
comment => $rest,
|
||||
direction => BOTH,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
comment => $rest,
|
||||
direction => BOTH,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
elsif ($line =~ /^0x(\w+)\s*U\+(\w+)\s*#\s*(\S.*)?\s*$/)
|
||||
@@ -51,7 +51,7 @@ while (my $line = <$in>)
|
||||
|
||||
# non-combined characters
|
||||
my ($c, $u, $rest) = ($1, $2, "U+" . $2 . $3);
|
||||
my $ucs = hex($u);
|
||||
my $ucs = hex($u);
|
||||
my $code = hex($c);
|
||||
my $direction;
|
||||
|
||||
@@ -74,12 +74,12 @@ while (my $line = <$in>)
|
||||
|
||||
push @mapping,
|
||||
{
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
comment => $rest,
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
comment => $rest,
|
||||
direction => $direction,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
}
|
||||
|
||||
@@ -22,13 +22,13 @@ my $mapping = read_source("CP932.TXT");
|
||||
# Drop these SJIS codes from the source for UTF8=>SJIS conversion
|
||||
my @reject_sjis = (
|
||||
0xed40 .. 0xeefc, 0x8754 .. 0x875d, 0x878a, 0x8782,
|
||||
0x8784, 0xfa5b, 0xfa54, 0x8790 .. 0x8792,
|
||||
0x8784, 0xfa5b, 0xfa54, 0x8790 .. 0x8792,
|
||||
0x8795 .. 0x8797, 0x879a .. 0x879c);
|
||||
|
||||
foreach my $i (@$mapping)
|
||||
{
|
||||
my $code = $i->{code};
|
||||
my $ucs = $i->{ucs};
|
||||
my $ucs = $i->{ucs};
|
||||
|
||||
if (grep { $code == $_ } @reject_sjis)
|
||||
{
|
||||
@@ -40,67 +40,67 @@ foreach my $i (@$mapping)
|
||||
push @$mapping,
|
||||
( {
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x00a2,
|
||||
code => 0x8191,
|
||||
comment => '# CENT SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x00a2,
|
||||
code => 0x8191,
|
||||
comment => '# CENT SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x00a3,
|
||||
code => 0x8192,
|
||||
comment => '# POUND SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x00a3,
|
||||
code => 0x8192,
|
||||
comment => '# POUND SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x00a5,
|
||||
code => 0x5c,
|
||||
comment => '# YEN SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x00a5,
|
||||
code => 0x5c,
|
||||
comment => '# YEN SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x00ac,
|
||||
code => 0x81ca,
|
||||
comment => '# NOT SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x00ac,
|
||||
code => 0x81ca,
|
||||
comment => '# NOT SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x2016,
|
||||
code => 0x8161,
|
||||
comment => '# DOUBLE VERTICAL LINE',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x2016,
|
||||
code => 0x8161,
|
||||
comment => '# DOUBLE VERTICAL LINE',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x203e,
|
||||
code => 0x7e,
|
||||
comment => '# OVERLINE',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x203e,
|
||||
code => 0x7e,
|
||||
comment => '# OVERLINE',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x2212,
|
||||
code => 0x817c,
|
||||
comment => '# MINUS SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x2212,
|
||||
code => 0x817c,
|
||||
comment => '# MINUS SIGN',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
},
|
||||
{
|
||||
direction => FROM_UNICODE,
|
||||
ucs => 0x301c,
|
||||
code => 0x8160,
|
||||
comment => '# WAVE DASH',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
ucs => 0x301c,
|
||||
code => 0x8160,
|
||||
comment => '# WAVE DASH',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
});
|
||||
|
||||
print_conversion_tables($this_script, "SJIS", $mapping);
|
||||
|
||||
@@ -33,7 +33,7 @@ while (<$in>)
|
||||
next if (!m/<a u="([0-9A-F]+)" b="([0-9A-F ]+)"/);
|
||||
my ($u, $c) = ($1, $2);
|
||||
$c =~ s/ //g;
|
||||
my $ucs = hex($u);
|
||||
my $ucs = hex($u);
|
||||
my $code = hex($c);
|
||||
|
||||
next if ($code == 0x0080 || $code == 0x00FF);
|
||||
@@ -42,11 +42,11 @@ while (<$in>)
|
||||
{
|
||||
push @mapping,
|
||||
{
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
ucs => $ucs,
|
||||
code => $code,
|
||||
direction => BOTH,
|
||||
f => $in_file,
|
||||
l => $.
|
||||
f => $in_file,
|
||||
l => $.
|
||||
};
|
||||
}
|
||||
}
|
||||
@@ -56,11 +56,11 @@ close($in);
|
||||
push @mapping,
|
||||
{
|
||||
direction => BOTH,
|
||||
code => 0xa2e8,
|
||||
ucs => 0x327e,
|
||||
comment => 'CIRCLED HANGUL IEUNG U',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
code => 0xa2e8,
|
||||
ucs => 0x327e,
|
||||
comment => 'CIRCLED HANGUL IEUNG U',
|
||||
f => $this_script,
|
||||
l => __LINE__
|
||||
};
|
||||
|
||||
print_conversion_tables($this_script, "UHC", \@mapping);
|
||||
|
||||
@@ -23,33 +23,33 @@ use convutils;
|
||||
my $this_script = 'src/backend/utils/mb/Unicode/UCS_to_most.pl';
|
||||
|
||||
my %filename = (
|
||||
'WIN866' => 'CP866.TXT',
|
||||
'WIN874' => 'CP874.TXT',
|
||||
'WIN1250' => 'CP1250.TXT',
|
||||
'WIN1251' => 'CP1251.TXT',
|
||||
'WIN1252' => 'CP1252.TXT',
|
||||
'WIN1253' => 'CP1253.TXT',
|
||||
'WIN1254' => 'CP1254.TXT',
|
||||
'WIN1255' => 'CP1255.TXT',
|
||||
'WIN1256' => 'CP1256.TXT',
|
||||
'WIN1257' => 'CP1257.TXT',
|
||||
'WIN1258' => 'CP1258.TXT',
|
||||
'ISO8859_2' => '8859-2.TXT',
|
||||
'ISO8859_3' => '8859-3.TXT',
|
||||
'ISO8859_4' => '8859-4.TXT',
|
||||
'ISO8859_5' => '8859-5.TXT',
|
||||
'ISO8859_6' => '8859-6.TXT',
|
||||
'ISO8859_7' => '8859-7.TXT',
|
||||
'ISO8859_8' => '8859-8.TXT',
|
||||
'ISO8859_9' => '8859-9.TXT',
|
||||
'WIN866' => 'CP866.TXT',
|
||||
'WIN874' => 'CP874.TXT',
|
||||
'WIN1250' => 'CP1250.TXT',
|
||||
'WIN1251' => 'CP1251.TXT',
|
||||
'WIN1252' => 'CP1252.TXT',
|
||||
'WIN1253' => 'CP1253.TXT',
|
||||
'WIN1254' => 'CP1254.TXT',
|
||||
'WIN1255' => 'CP1255.TXT',
|
||||
'WIN1256' => 'CP1256.TXT',
|
||||
'WIN1257' => 'CP1257.TXT',
|
||||
'WIN1258' => 'CP1258.TXT',
|
||||
'ISO8859_2' => '8859-2.TXT',
|
||||
'ISO8859_3' => '8859-3.TXT',
|
||||
'ISO8859_4' => '8859-4.TXT',
|
||||
'ISO8859_5' => '8859-5.TXT',
|
||||
'ISO8859_6' => '8859-6.TXT',
|
||||
'ISO8859_7' => '8859-7.TXT',
|
||||
'ISO8859_8' => '8859-8.TXT',
|
||||
'ISO8859_9' => '8859-9.TXT',
|
||||
'ISO8859_10' => '8859-10.TXT',
|
||||
'ISO8859_13' => '8859-13.TXT',
|
||||
'ISO8859_14' => '8859-14.TXT',
|
||||
'ISO8859_15' => '8859-15.TXT',
|
||||
'ISO8859_16' => '8859-16.TXT',
|
||||
'KOI8R' => 'KOI8-R.TXT',
|
||||
'KOI8U' => 'KOI8-U.TXT',
|
||||
'GBK' => 'CP936.TXT');
|
||||
'KOI8R' => 'KOI8-R.TXT',
|
||||
'KOI8U' => 'KOI8-U.TXT',
|
||||
'GBK' => 'CP936.TXT');
|
||||
|
||||
# make maps for all encodings if not specified
|
||||
my @charsets = (scalar(@ARGV) > 0) ? @ARGV : sort keys(%filename);
|
||||
|
||||
@@ -16,10 +16,10 @@ our @EXPORT =
|
||||
|
||||
# Constants used in the 'direction' field of the character maps
|
||||
use constant {
|
||||
NONE => 0,
|
||||
TO_UNICODE => 1,
|
||||
NONE => 0,
|
||||
TO_UNICODE => 1,
|
||||
FROM_UNICODE => 2,
|
||||
BOTH => 3
|
||||
BOTH => 3
|
||||
};
|
||||
|
||||
#######################################################################
|
||||
@@ -53,12 +53,12 @@ sub read_source
|
||||
exit;
|
||||
}
|
||||
my $out = {
|
||||
code => hex($1),
|
||||
ucs => hex($2),
|
||||
comment => $4,
|
||||
code => hex($1),
|
||||
ucs => hex($2),
|
||||
comment => $4,
|
||||
direction => BOTH,
|
||||
f => $fname,
|
||||
l => $.
|
||||
f => $fname,
|
||||
l => $.
|
||||
};
|
||||
|
||||
# Ignore pure ASCII mappings. PostgreSQL character conversion code
|
||||
@@ -124,14 +124,14 @@ sub print_conversion_tables_direction
|
||||
my $tblname;
|
||||
if ($direction == TO_UNICODE)
|
||||
{
|
||||
$fname = lc("${csname}_to_utf8.map");
|
||||
$fname = lc("${csname}_to_utf8.map");
|
||||
$tblname = lc("${csname}_to_unicode_tree");
|
||||
|
||||
print "- Writing ${csname}=>UTF8 conversion table: $fname\n";
|
||||
}
|
||||
else
|
||||
{
|
||||
$fname = lc("utf8_to_${csname}.map");
|
||||
$fname = lc("utf8_to_${csname}.map");
|
||||
$tblname = lc("${csname}_from_unicode_tree");
|
||||
|
||||
print "- Writing UTF8=>${csname} conversion table: $fname\n";
|
||||
@@ -378,10 +378,10 @@ sub print_radix_table
|
||||
|
||||
unshift @segments,
|
||||
{
|
||||
header => "Dummy map, for invalid values",
|
||||
header => "Dummy map, for invalid values",
|
||||
min_idx => 0,
|
||||
max_idx => $widest_range,
|
||||
label => "dummy map"
|
||||
label => "dummy map"
|
||||
};
|
||||
|
||||
###
|
||||
@@ -397,7 +397,7 @@ sub print_radix_table
|
||||
###
|
||||
for (my $j = 0; $j < $#segments - 1; $j++)
|
||||
{
|
||||
my $seg = $segments[$j];
|
||||
my $seg = $segments[$j];
|
||||
my $nextseg = $segments[ $j + 1 ];
|
||||
|
||||
# Count the number of zero values at the end of this segment.
|
||||
@@ -527,17 +527,17 @@ sub print_radix_table
|
||||
if ($max_val <= 0xffff)
|
||||
{
|
||||
$vals_per_line = 8;
|
||||
$colwidth = 4;
|
||||
$colwidth = 4;
|
||||
}
|
||||
elsif ($max_val <= 0xffffff)
|
||||
{
|
||||
$vals_per_line = 4;
|
||||
$colwidth = 6;
|
||||
$colwidth = 6;
|
||||
}
|
||||
else
|
||||
{
|
||||
$vals_per_line = 4;
|
||||
$colwidth = 8;
|
||||
$colwidth = 8;
|
||||
}
|
||||
|
||||
###
|
||||
@@ -607,8 +607,10 @@ sub print_radix_table
|
||||
# Print the next line's worth of values.
|
||||
# XXX pad to begin at a nice boundary
|
||||
printf $out " /* %02x */ ", $i;
|
||||
for (my $j = 0;
|
||||
$j < $vals_per_line && $i <= $seg->{max_idx}; $j++)
|
||||
for (
|
||||
my $j = 0;
|
||||
$j < $vals_per_line && $i <= $seg->{max_idx};
|
||||
$j++)
|
||||
{
|
||||
# missing values represent zero.
|
||||
my $val = $seg->{values}->{$i} || 0;
|
||||
@@ -671,10 +673,10 @@ sub build_segments_recurse
|
||||
push @segments,
|
||||
{
|
||||
header => $header . ", leaf: ${path}xx",
|
||||
label => $label,
|
||||
level => $level,
|
||||
depth => $depth,
|
||||
path => $path,
|
||||
label => $label,
|
||||
level => $level,
|
||||
depth => $depth,
|
||||
path => $path,
|
||||
values => $map
|
||||
};
|
||||
}
|
||||
@@ -696,10 +698,10 @@ sub build_segments_recurse
|
||||
push @segments,
|
||||
{
|
||||
header => $header . ", byte #$level: ${path}xx",
|
||||
label => $label,
|
||||
level => $level,
|
||||
depth => $depth,
|
||||
path => $path,
|
||||
label => $label,
|
||||
level => $level,
|
||||
depth => $depth,
|
||||
path => $path,
|
||||
values => \%children
|
||||
};
|
||||
}
|
||||
@@ -789,12 +791,12 @@ sub make_charmap_combined
|
||||
if (defined $c->{ucs_second})
|
||||
{
|
||||
my $entry = {
|
||||
utf8 => ucs2utf($c->{ucs}),
|
||||
utf8 => ucs2utf($c->{ucs}),
|
||||
utf8_second => ucs2utf($c->{ucs_second}),
|
||||
code => $c->{code},
|
||||
comment => $c->{comment},
|
||||
f => $c->{f},
|
||||
l => $c->{l}
|
||||
code => $c->{code},
|
||||
comment => $c->{comment},
|
||||
f => $c->{f},
|
||||
l => $c->{l}
|
||||
};
|
||||
push @combined, $entry;
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user