wuggy/Dactyloidae
wuggy/Dactyloidae
1
0
Fork 0
mirror of https://repo.dactyloidae.xyz/Dactyloidae/UXP.git synced 2026-09-20 15:27:32 +09:00
Code Issues Projects Releases Packages Wiki Activity

Issue #326 - Part 4: Update Unicode data to release 10.0

Browse source 
genUnicodePropertyData.pl's output:
 unknown ICU script MASARAM_GONDI at ../tools/genUnicodePropertyData.pl line 442, <FH> line 2641.
 unknown ICU script SOYOMBO at ../tools/genUnicodePropertyData.pl line 442, <FH> line 2663.
 unknown ICU script ZANABAZAR_SQUARE at ../tools/genUnicodePropertyData.pl line 442, <FH> line 2679.
 Data for CharProp2 = 14864
 Data for FullWidth = 2176
 Data for FullWidthInverse = 2688
 Total data = 19728
This commit is contained in:
Job Bautista 2022-06-23 13:28:04 +08:00 • committed by roytam1
commit 54489c3029
5 changed files with 75 additions and 115 deletions
Download patch file Download diff file Expand all files Collapse all files

46
intl/unicharutil/tools/genUnicodePropertyData.pl
View file

@ -32,12 +32,11 @@
#
# We also require the file
# http://www.unicode.org/Public/security/latest/IdentifierStatus.txt
# http://www.unicode.org/Public/security/latest/IdentifierType.txt
# This file should be in a sub-directory "security" immediately below the
# directory containing the other Unicode data files.
#
# We also require the latest data file for UTR50, currently revision-16:
# http://www.unicode.org/Public/vertical/revision-16/VerticalOrientation-16.txt
# We also require the latest data file for UTR50, currently revision-17:
# http://www.unicode.org/Public/vertical/revision-17/VerticalOrientation-17.txt
# This file should be in a sub-directory "vertical" immediately below the
# directory containing the other Unicode data files.
#
@ -168,9 +167,7 @@ my %idType = (
# These match the IdentifierType enum in nsUnicodeProperties.h.
my %mappedIdType = (
"Restricted" => 0,
"Allowed" => 1,
"Aspirational" => 2 # for Aspirational characters that are not excluded
# by another attribute.
"Allowed" => 1
);
my %bidicategoryCode = (
@ -441,7 +438,11 @@ while (<FH>) {
while (<FH>) {
if (m/([0-9A-F]{4,6})(?:\.\.([0-9A-F]{4,6}))*\s+;\s+([^ ]+)/) {
my $script = uc($3);
warn "unknown ICU script $script" unless exists $scriptCode{$script};
unless (exists $scriptCode{$script}) {
warn "unknown ICU script $script";
$scriptCode{$script} = $scriptCode{"UNKNOWN"};
}
$script = $scriptCode{$script};
my $script = $scriptCode{$script};
my $start = hex "0x$1";
my $end = (defined $2) ? hex "0x$2" : $start;
@ -623,33 +624,6 @@ while (<FH>) {
}
close FH;
# read IdentifierType.txt, to find Aspirational characters
open FH, "< $UNICODE/security/IdentifierType.txt" or die "can't open UCD file IdentifierType.txt\n";
push @versionInfo, "";
while (<FH>) {
chomp;
s/\xef\xbb\xbf//;
push @versionInfo, $_;
last if /Date:/;
}
while (<FH>) {
if (m/([0-9A-F]{4,6})(?:\.\.([0-9A-F]{4,6}))*\s+;\s+([^#]+)/) {
my $idtype = $3;
foreach (split(/ /, $idtype)) {
warn "unknown Identifier Type $_" unless exists $idType{$_};
}
my $start = hex "0x$1";
my $end = (defined $2) ? hex "0x$2" : $start;
if ($idtype =~ /Aspirational/ and (not $idtype =~ /Exclusion|Not_XID|Not_NFKC/)) {
for (my $i = $start; $i <= $end; ++$i) {
$idtype[$i] = $mappedIdType{'Aspirational'};
}
}
}
}
close FH;
open FH, "< $UNICODE/Unihan_Variants.txt" or die "can't open UCD file Unihan_Variants.txt (from Unihan.zip)\n";
push @versionInfo, "";
while (<FH>) {
@ -686,8 +660,8 @@ while (<FH>) {
}
close FH;
# read VerticalOrientation-16.txt
open FH, "< $UNICODE/vertical/VerticalOrientation-16.txt" or die "can't open UTR50 data file VerticalOrientation-16.txt\n";
# read VerticalOrientation-17.txt
open FH, "< $UNICODE/vertical/VerticalOrientation-17.txt" or die "can't open UTR50 data file VerticalOrientation-17.txt\n";
push @versionInfo, "";
while (<FH>) {
chomp;