You can not select more than 25 topics
Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
74 lines
1.6 KiB
74 lines
1.6 KiB
3 months ago
|
#!/usr/bin/perl
|
||
|
|
#
|
||
|
|
# Generate Expansion Table of Compatibility Ligatures
|
||
|
|
# (c) 2003 Martin Mares <mj@atrey.karlin.mff.cuni.cz>
|
||
|
|
#
|
||
|
|
|
||
|
|
use strict;
|
||
|
|
use warnings;
|
||
|
|
|
||
|
|
print STDERR "Reading ligature list\n";
|
||
|
|
open(L, "misc/u-ligatures") || die "lig file open";
|
||
|
|
my %ligs = ();
|
||
|
|
while (<L>) {
|
||
|
|
chomp;
|
||
|
|
$ligs{$_} = 1;
|
||
|
|
}
|
||
|
|
close L;
|
||
|
|
|
||
|
|
print STDERR "Reading decompositions\n";
|
||
|
|
open(I, "unidata/UnicodeData.txt") || die "Unable to open UniCode data file";
|
||
|
|
my %decs = ();
|
||
|
|
while (<I>) {
|
||
|
|
chomp;
|
||
|
|
(/^$/ || /^#/) && next;
|
||
|
|
my ($code,$name,$cat,$comb,$bidir,$decomp,$d0,$d1,$n0,$mirr,$cmt1,$cmt2,$upper,$lower,$title) = split /;/;
|
||
|
|
$code =~ /^....$/ || next;
|
||
|
|
if (my ($d) = ($decomp =~ /^<compat> (.*)/)) {
|
||
|
|
$decs{$code} = $d;
|
||
|
|
}
|
||
|
|
}
|
||
|
|
close I;
|
||
|
|
|
||
|
|
sub expand($) {
|
||
|
|
my ($c) = @_;
|
||
|
|
if (defined $decs{$c}) {
|
||
|
|
return join (" ", map { expand($_) } split(/\s+/, $decs{$c}));
|
||
|
|
} else {
|
||
|
|
return $c;
|
||
|
|
}
|
||
|
|
}
|
||
|
|
|
||
|
|
print STDERR "Searching for a perfect hash function\n";
|
||
|
|
my $n = keys %ligs;
|
||
|
|
my $div = $n-1;
|
||
|
|
DIV: while (++$div) {
|
||
|
|
#print STDERR "Trying $div... ";
|
||
|
|
my @c = ();
|
||
|
|
foreach my $l (keys %ligs) {
|
||
|
|
my $i = (hex $l) % $div;
|
||
|
|
if (defined $c[$i]) {
|
||
|
|
#print STDERR "collision\n";
|
||
|
|
next DIV;
|
||
|
|
}
|
||
|
|
$c[$i] = 1;
|
||
|
|
}
|
||
|
|
#print STDERR "FOUND\n";
|
||
|
|
last;
|
||
|
|
}
|
||
|
|
|
||
|
|
print STDERR "Filling hash table with $div entries for $n ligatures\n";
|
||
|
|
my @ht = map { "NULL" } 1..$div;
|
||
|
|
foreach my $l (keys %ligs) {
|
||
|
|
my $i = (hex $l) % $div;
|
||
|
|
my $w = join(", ", map { "0x$_" } split(/ /, expand($l)));
|
||
|
|
$ht[$i] = "/* $l */ (const u16 []) { $w, 0 }";
|
||
|
|
}
|
||
|
|
|
||
|
|
print "#define LIG_HASH_SIZE $div\n\n";
|
||
|
|
print "static const u16 *_U_lig_hash[] = {\n";
|
||
|
|
for (my $i=0; $i<$div; $i++) {
|
||
|
|
print "\t", $ht[$i], ",\n";
|
||
|
|
}
|
||
|
|
print "};\n";
|