Explicitly check that our $_; is special, and forced to main::

[p5sagit/p5-mst-13.2.git] / t / op / lc.t

diff --git a/t/op/lc.t b/t/op/lc.t
index 243b0c7..6b7625b 100644 (file)
--- a/t/op/lc.t
+++ b/t/op/lc.t
@@ -6,7 +6,12 @@ BEGIN {
     require './test.pl';
 }
 
-plan tests => 59;
+plan tests => 93;
+
+is(lc(undef),     "", "lc(undef) is ''");
+is(lcfirst(undef), "", "lcfirst(undef) is ''");
+is(uc(undef),     "", "uc(undef) is ''");
+is(ucfirst(undef), "", "ucfirst(undef) is ''");
 
 $a = "HELLO.* world";
 $b = "hello.* WORLD";
@@ -111,8 +116,14 @@ is("\u\x{587}" , "\x{535}\x{582}", "ligature titlecase");
 is("\U\x{587}" , "\x{535}\x{552}", "ligature uppercase");
 
 # mktables had problems where many-to-one case mappings didn't work right.
-# The lib/unifold.t should give the fourth folding, "casefolding", a good
-# workout.
+# The lib/uni/fold.t should give the fourth folding, "casefolding", a good
+# workout (one cannot directly get that from Perl). 
+# \x{01C4} is LATIN CAPITAL LETTER DZ WITH CARON
+# \x{01C5} is LATIN CAPITAL LETTER D WITH SMALL LETTER Z WITH CARON
+# \x{01C6} is LATIN SMALL LETTER DZ WITH CARON
+# \x{03A3} is GREEK CAPITAL LETTER SIGMA
+# \x{03C2} is GREEK SMALL LETTER FINAL SIGMA
+# \x{03C3} is GREEK SMALL LETTER SIGMA
 
 is(lc("\x{1C4}") , "\x{1C6}",      "U+01C4 lc is U+01C6");
 is(lc("\x{1C5}") , "\x{1C6}",      "U+01C5 lc is U+01C6, too");
@@ -130,17 +141,17 @@ $b = "\x{3a3}FOO.BAR"; # \x{3a3} == GREEK CAPITAL LETTER SIGMA.
 ($c = $b) =~ s/(\w+)/lc($1)/ge;
 is($c , $a, "Using s///e to change case.");
 
-($c = $a) =~ s/(\w+)/uc($1)/ge;
+($c = $a) =~ s/(\p{IsWord}+)/uc($1)/ge;
 is($c , $b, "Using s///e to change case.");
 
-($c = $b) =~ s/(\w+)/lcfirst($1)/ge;
+($c = $b) =~ s/(\p{IsWord}+)/lcfirst($1)/ge;
 is($c , "\x{3c3}FOO.bAR", "Using s///e to change case.");
 
-($c = $a) =~ s/(\w+)/ucfirst($1)/ge;
+($c = $a) =~ s/(\p{IsWord}+)/ucfirst($1)/ge;
 is($c , "\x{3a3}foo.Bar", "Using s///e to change case.");
 
 # #18931: perl5.8.0 bug in \U..\E processing
-# Test case from Nick Clark.
+# Test case from Nicholas Clark.
 for my $a (0,1) {
     $_ = 'abcdefgh';
     $_ .= chr 256;
@@ -155,6 +166,57 @@ for my $a (0,1) {
        chop $a;
        $a =~ s/^(\s*)(\w*)/$1\u$2/;
        is($a, v10, "[perl #18857]");
-       $test++;
     } 
 }
+
+
+# [perl #38619] Bug in lc and uc (interaction between UTF-8, substr, and lc/uc)
+
+for ("a\x{100}", "xyz\x{100}") {
+    is(substr(uc($_), 0), uc($_), "[perl #38619] uc");
+}
+for ("A\x{100}", "XYZ\x{100}") {
+    is(substr(lc($_), 0), lc($_), "[perl #38619] lc");
+}
+for ("a\x{100}", "ßyz\x{100}") { # ß to Ss (different length)
+    is(substr(ucfirst($_), 0), ucfirst($_), "[perl #38619] ucfirst");
+}
+
+# Related to [perl #38619]
+# the original report concerns PERL_MAGIC_utf8.
+# these cases concern PERL_MAGIC_regex_global.
+
+for (map { $_ } "a\x{100}", "abc\x{100}", "\x{100}") {
+    chop; # get ("a", "abc", "") in utf8
+    my $return =  uc($_) =~ /\G(.?)/g;
+    my $result = $return ? $1 : "not";
+    my $expect = (uc($_) =~ /(.?)/g)[0];
+    is($return, 1,       "[perl #38619]");
+    is($result, $expect, "[perl #38619]");
+}
+
+for (map { $_ } "A\x{100}", "ABC\x{100}", "\x{100}") {
+    chop; # get ("A", "ABC", "") in utf8
+    my $return =  lc($_) =~ /\G(.?)/g;
+    my $result = $return ? $1 : "not";
+    my $expect = (lc($_) =~ /(.?)/g)[0];
+    is($return, 1,       "[perl #38619]");
+    is($result, $expect, "[perl #38619]");
+}
+
+for (1, 4, 9, 16, 25) {
+    is(uc "\x{03B0}" x $_, "\x{3a5}\x{308}\x{301}" x $_,
+       'uc U+03B0 grows threefold');
+
+    is(lc "\x{0130}" x $_, "i\x{307}" x $_, 'lc U+0130 grows');
+}
+
+# bug #43207
+my $temp = "Hello";
+for ("$temp") {
+    lc $_;
+    is($_, "Hello");
+}
+
+# new in Unicode 5.1.0
+is(lc("\x{1E9E}"), "\x{df}", "lc(LATIN CAPITAL LETTER SHARP S)");