Skip to content
Draft
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
20 changes: 18 additions & 2 deletions src/main/java/org/perlonjava/runtime/regex/UnicodeResolver.java
Original file line number Diff line number Diff line change
Expand Up @@ -1407,8 +1407,20 @@ private static UnicodeSet resolvePerlBlock(String value) {
UnicodeSet result = new UnicodeSet();
for (int valueId = 0; valueId < PerlUnicodeBlockData.valueCount(); valueId++) {
String candidate = PerlUnicodeBlockData.canonicalValue(valueId);
if (valuePattern.matcher(candidate).matches()
|| valuePattern.matcher(loosePropertyName(candidate)).matches()) {
boolean matches = valuePattern.matcher(candidate).matches()
|| valuePattern.matcher(loosePropertyName(candidate)).matches();
int icuValue = unicodePropertyValue(UProperty.BLOCK, candidate);
for (int nameChoice = UProperty.NameChoice.SHORT;
!matches && icuValue >= 0 && nameChoice <= UProperty.NameChoice.LONG;
nameChoice++) {
String officialAlias = UCharacter.getPropertyValueName(
UProperty.BLOCK, icuValue, nameChoice);
matches = officialAlias != null
&& (valuePattern.matcher(officialAlias).matches()
|| valuePattern.matcher(
loosePropertyName(officialAlias)).matches());
}
if (matches) {
result.addAll(PerlUnicodeBlockData.set(valueId));
}
}
Expand Down Expand Up @@ -1476,6 +1488,10 @@ private static UnicodeSet resolvePerlScript(String value, boolean extensions) {

private static String perlBlockWildcardBody(String value) {
String trimmed = value.trim();
if (trimmed.startsWith(":\\A") && trimmed.endsWith("\\z:")
&& trimmed.length() > 6) {
return trimmed.substring(3, trimmed.length() - 3);
}
if (!trimmed.startsWith("#") || !trimmed.endsWith("#")
|| trimmed.length() <= 2) {
return null;
Expand Down
47 changes: 47 additions & 0 deletions src/test/resources/unit/regex/unicode_block_colon_wildcards.t
Original file line number Diff line number Diff line change
@@ -0,0 +1,47 @@
use strict;
use warnings;
use Test::More;

no warnings 'experimental::uniprop_wildcards';

sub compile_property {
my ($property) = @_;
my $pattern = eval 'qr/\A\p{' . $property . '}\z/u';
return ($pattern, $@);
}

my ($basic, $basic_error) = compile_property('Block=:\ABasic_Latin\z:');
ok(defined $basic, 'colon-delimited Block wildcard compiles')
or diag($basic_error);
like('A', $basic, 'colon-delimited Block wildcard matches its block');
unlike(chr(0x03B1), $basic,
'colon-delimited Block wildcard excludes another block');

my ($short, $short_error) = compile_property('Blk=:\AASCII\z:');
ok(defined $short, 'Blk accepts a colon-delimited wildcard alias')
or diag($short_error);
like('A', $short, 'wildcard value accepts an official Block alias');

my ($loose, $loose_error) = compile_property('Block=:\Abasiclatin\z:');
ok(defined $loose, 'Block wildcard compares loose value spellings')
or diag($loose_error);
like('A', $loose, 'loose wildcard value selects Basic_Latin');

my ($union, $union_error) =
compile_property('Block=:\A(?:Basic_Latin|Greek_And_Coptic)\z:');
ok(defined $union, 'Block wildcard accepts value alternation')
or diag($union_error);
like('A', $union, 'Block wildcard alternation includes Basic_Latin');
like(chr(0x0378), $union,
'Block wildcard alternation includes Greek_And_Coptic');

for my $rejected (
['Block=:\ANever_A_Block\z:', 'wildcard matching no Block value'],
['Block=:\A.*\z:', 'star quantifier in Block wildcard'],
['Is_Block=:\ABasic_Latin\z:', 'Is-prefixed Block wildcard'],
) {
my ($pattern, $error) = compile_property($rejected->[0]);
ok(!defined($pattern) && length($error), "$rejected->[1] is rejected");
}

done_testing;