use warnings;
use 5.010;
-my $IS_EBCDIC = ord ('A') == 193;
+BEGIN {
+ chdir 't' if -d 't';
+ require './test.pl';
+ skip_all_if_miniperl("no dynamic loading on miniperl, no File::Spec (used by charnames)");
+}
sub run_tests;
# It's ok to repeat class names.
#
InLatin1Supplement =>
- $IS_EBCDIC ? ['!\x{7f}', '\x{80}', '!\x{100}']
- : ['!\x{7f}', '\x{80}', '\x{ff}', '!\x{100}'],
+ ['!\N{U+7f}', '\N{U+80}', '\N{U+ff}', '!\x{100}'],
InLatinExtendedA =>
- ['!\x{7f}', '!\x{80}', '!\x{ff}', '\x{100}'],
+ ['!\N{U+7f}', '!\N{U+80}', '!\N{U+ff}', '\x{100}'],
#
# Properties are case-insensitive, and may have whitespace,
# dashes and underscores.
#
- 'in-latin1_SUPPLEMENT' => ['\x{80}',
+ 'in-latin1_SUPPLEMENT' => ['\N{U+80}',
'\N{LATIN SMALL LETTER Y WITH DIAERESIS}'],
' ^ In Latin 1 Supplement '
- => ['!\x{80}', '\N{COFFIN}'],
- 'latin-1 supplement' => ['\x{80}', "0xDF"],
+ => ['!\N{U+80}', '\N{COFFIN}'],
+ 'latin-1 supplement' => ['\N{U+80}', "0xDF"],
);
InNotKana => ['\x{3040}', '!\x{3041}'],
InConsonant => ['d', '!e'],
IsSyriac1 => ['\x{0712}', '!\x{072F}'],
- Syriac1 => ['\x{0712}', '!\x{072F}'],
- '# User-defined character properties my lack \n at the end',
+ IsSyriac1KanaMark => ['\x{309A}', '!\x{3090}'],
+ IsSyriac1KanaMark => ['\x{0730}', '!\x{0712}'],
+ '# User-defined character properties may lack \n at the end',
InGreekSmall => ['\N{GREEK SMALL LETTER PI}',
'\N{GREEK SMALL LETTER FINAL SIGMA}'],
InGreekCapital => ['\N{GREEK CAPITAL LETTER PI}', '!\x{03A2}'],
Dash => ['-'],
ASCII_Hex_Digit => ['!-', 'A'],
- AsciiHexAndDash => ['-', 'A'],
+ IsAsciiHexAndDash => ['-', 'A'],
+);
+
+my @USER_CASELESS_PROPERTIES = (
+ #
+ # User defined properties which differ depending on /i. Second entry is
+ # false regularly, true under /i
+ #
+ 'IsMyUpper' => ["M", "!m" ],
);
#
# Illegal properties
#
-my @ILLEGAL_PROPERTIES = qw [q qrst];
+my @ILLEGAL_PROPERTIES =
+ qw[q qrst f foo isfoo infoo ISfoo INfoo Is::foo In::foo];
my %d;
push @{$d {IsWord}} => map {$class =~ /^[LMN]/ || $_ eq "_"
? $_ : "!$_"} @$chars;
push @{$d {IsSpace}} => map {$class =~ /^Z/ ||
- length ($_) == 1 && ord ($_) >= 0x09
- && ord ($_) <= 0x0D
+ length ($_) == 1 && utf8::native_to_unicode(ord ($_)) >= 0x09
+ && utf8::native_to_unicode(ord ($_)) <= 0x0D
? $_ : "!$_"} @$chars;
}
-delete $d {IsASCII} if $IS_EBCDIC;
-
push @CLASSES => "# Short properties" => %SHORT_PROPERTIES,
"# POSIX like properties" => %d,
"# User defined properties" => @USER_DEFINED_PROPERTIES;
my $count = 0;
for (my $i = 0; $i < @CLASSES; $i += 2) {
$i ++, redo if $CLASSES [$i] =~ /^\h*#\h*(.*)/;
- $count += (length $CLASSES [$i] == 1 ? 4 : 2) * @{$CLASSES [$i + 1]};
+ $count += 2 * (length $CLASSES [$i] == 1 ? 4 : 2) * @{$CLASSES [$i + 1]};
}
-$count += 2 * @ILLEGAL_PROPERTIES;
-$count += 2 * grep {length $_ == 1} @ILLEGAL_PROPERTIES;
+$count += 4 * @ILLEGAL_PROPERTIES;
+$count += 4 * grep {length $_ == 1} @ILLEGAL_PROPERTIES;
+$count += 8 * @USER_CASELESS_PROPERTIES;
-my $tests = 0;
-
-say "1..$count";
+plan(tests => $count);
run_tests unless caller ();
sub match {
- my ($char, $match, $nomatch) = @_;
+ my ($char, $match, $nomatch, $caseless) = @_;
+ $caseless = "" unless defined $caseless;
+ $caseless = 'i' if $caseless;
my ($str, $name);
- given ($char) {
- when (/^\\/) {
- $str = eval qq ["$char"];
- $name = qq ["$char"];
- }
- when (/^0x([0-9A-Fa-f]+)$/) {
- $str = chr hex $1;
- $name = "chr ($char)";
- }
- default {
- $str = $char;
- $name = qq ["$char"];
- }
+ if ($char =~ /^\\/) {
+ $str = eval qq ["$char"];
+ $name = qq ["$char"];
+ }
+ elsif ($char =~ /^0x([0-9A-Fa-f]+)$/) {
+ $str = chr hex $1;
+ $name = "chr ($char)";
+ }
+ else {
+ $str = $char;
+ $name = qq ["$char"];
}
- print "not " unless $str =~ /$match/;
- print "ok ", ++ $tests, " - $name =~ /$match/\n";
- print "not " unless $str !~ /$nomatch/;
- print "ok ", ++ $tests, " - $name !~ /$nomatch/\n";
+ undef $@;
+ my $pat = "qr/$match/$caseless";
+ my $match_pat = eval $pat;
+ is($@, '', "$pat compiled correctly to a regexp: $@");
+ like($str, $match_pat, "$name correctly matched");
+
+ undef $@;
+ $pat = "qr/$nomatch/$caseless";
+ my $nomatch_pat = eval $pat;
+ is($@, '', "$pat compiled correctly to a regexp: $@");
+ unlike($str, $nomatch_pat, "$name correctly did not match");
}
sub run_tests {
match $_, $in_pat, $out_pat for @in;
match $_, $out_pat, $in_pat for @out;
- if (1 == length $class) {
+ if (1 == length $class) { # Repeat without braces if name length 1
my $in_pat = eval qq ['\\p$class'];
my $out_pat = eval qq ['\\P$class'];
foreach my $p (@ILLEGAL_PROPERTIES) {
undef $@;
my $r = eval "'a' =~ /\\p{$p}/; 1";
- print "not " unless !$r && $@ && $@ =~ $pat;
- print "ok ", ++ $tests, " - Unknown Unicode property \\p{$p}\n";
+ is($r, undef, "Unknown Unicode property \\p{$p}");
+ like($@, $pat, "Unknown Unicode property \\p{$p}");
undef $@;
my $s = eval "'a' =~ /\\P{$p}/; 1";
- print "not " unless !$s && $@ && $@ =~ $pat;
- print "ok ", ++ $tests, " - Unknown Unicode property \\P{$p}\n";
+ is($s, undef, "Unknown Unicode property \\p{$p}");
+ like($@, $pat, "Unknown Unicode property \\p{$p}");
if (length $p == 1) {
undef $@;
my $r = eval "'a' =~ /\\p$p/; 1";
- print "not " unless !$r && $@ && $@ =~ $pat;
- print "ok ", ++ $tests, " - Unknown Unicode property \\p$p\n";
+ is($r, undef, "Unknown Unicode property \\p$p");
+ like($@, $pat, "Unknown Unicode property \\p$p");
undef $@;
my $s = eval "'a' =~ /\\P$p/; 1";
- print "not " unless !$s && $@ && $@ =~ $pat;
- print "ok ", ++ $tests, " - Unknown Unicode property \\P$p\n";
+ is($r, undef, "Unknown Unicode property \\P$p");
+ like($@, $pat, "Unknown Unicode property \\P$p");
}
}
+
+ print "# User-defined properties with /i differences\n";
+ foreach my $class (shift @USER_CASELESS_PROPERTIES) {
+ my $chars_ref = shift @USER_CASELESS_PROPERTIES;
+ my @in = grep {!/^!./} @$chars_ref;
+ my @out = map {s/^!(?=.)//; $_} grep { /^!./} @$chars_ref;
+ my $in_pat = eval qq ['\\p{$class}'];
+ my $out_pat = eval qq ['\\P{$class}'];
+
+ # Verify works as regularly for not /i
+ match $_, $in_pat, $out_pat for @in;
+ match $_, $out_pat, $in_pat for @out;
+
+ # Verify that adding /i doesn't change the in set.
+ match $_, $in_pat, $out_pat, 'i' for @in;
+
+ # Verify that adding /i does change the out set to match.
+ match $_, $in_pat, $out_pat, 'i' for @out;
+ }
}
+utf8::IsCn
--
-sub InConsonant {<<'--'} # Not EBCDIC-aware.
-0061 007f
--0061
--0065
--0069
--006f
--0075
---
+sub InConsonant {
-sub IsSyriac1 {<<'--'}
-0712 072C
-0730 074A
---
+ my $return = "+utf8::Lowercase\n&utf8::ASCII\n";
+ $return .= sprintf("-%X\n", ord "a");
+ $return .= sprintf("-%X\n", ord "e");
+ $return .= sprintf("-%X\n", ord "i");
+ $return .= sprintf("-%X\n", ord "o");
+ $return .= sprintf("-%X\n", ord "u");
+ return $return;
+}
-sub Syriac1 {<<'--'}
+sub IsSyriac1 {<<'--'}
0712 072C
0730 074A
--
sub InGreekSmall {return "03B1\t03C9"}
sub InGreekCapital {return "0391\t03A9\n-03A2"}
-sub AsciiHexAndDash {<<'--'}
+sub IsAsciiHexAndDash {<<'--'}
+utf8::ASCII_Hex_Digit
+utf8::Dash
--
+sub IsMyUpper {
+ my $caseless = shift;
+ return "+utf8::"
+ . (($caseless)
+ ? 'Alphabetic'
+ : 'Uppercase')
+ . "\n&utf8::ASCII";
+}
+
+# Verify that can use user-defined properties inside another one
+sub IsSyriac1KanaMark {<<'--'}
++main::IsSyriac1
++main::InKana3
+&utf8::IsMark
+--
+
+# fake user-defined properties; these subs shouldn't be called, because
+# their names don't start with In or Is
+
+sub f { die }
+sub foo { die }
+sub isfoo { die }
+sub infoo { die }
+sub ISfoo { die }
+sub INfoo { die }
+sub Is::foo { die }
+sub In::foo { die }
__END__