improvements for high-bit text literals (from Gisle Aas)

p4raw-id: //depot/perl@5192

5 files changed, 7 insertions, 122 deletions

diff --git a/t/pragma/warn/doop b/t/pragma/warn/doop
index c16e24f919..5803b44581 100644
--- a/t/pragma/warn/doop
+++ b/t/pragma/warn/doop
@@ -1,29 +1,6 @@
-  doop.c	AOK
-
-  \x%s will produce malformed UTF-8 character; use \x{%s} for that
-
-
-__END__
 # doop.c
 use utf8 ;
 $_ = "\x80  \xff" ;
 chop ;
 EXPECT
 ########
-# doop.c
-BEGIN {
-    if (ord("\t") == 5) {
-        print "SKIPPED\n# Character codes differ on ebcdic machines.";
-        exit 0;
-    }
-}
-use warnings 'utf8'  ;
-use utf8 ;
-$_ = "\x80  \xff" ;
-chop ;
-no warnings 'utf8'  ;
-$_ = "\x80  \xff" ;
-chop ;
-EXPECT
-\x80 will produce malformed UTF-8 character; use \x{80} for that at - line 10.
-\xff will produce malformed UTF-8 character; use \x{ff} for that at - line 10.
diff --git a/t/pragma/warn/pp b/t/pragma/warn/pp
index b392029767..8f42ba64ec 100644
--- a/t/pragma/warn/pp
+++ b/t/pragma/warn/pp
@@ -28,11 +28,6 @@
   Constant subroutine %s undefined			<<<TODO
   Constant subroutine (anonymous) undefined		<<<TODO
 
-  Mandatory Warnings
-  ------------------
-  Malformed UTF-8 character (not tested: difficult to produce with
-                             perl now)
-
 __END__
 # pp.c
 use warnings 'substr' ;
@@ -113,20 +108,3 @@ $_ = "\x80  \xff" ;
 reverse ;
 EXPECT
 ########
-# pp.c
-BEGIN {
-    if (ord("\t") == 5) {
-        print "SKIPPED\n# Character codes differ on ebcdic machines.";
-        exit 0;
-    }
-}
-use warnings 'utf8'  ;
-use utf8 ;
-$_ = "\x80  \xff" ;
-reverse ;
-no warnings 'utf8'  ;
-$_ = "\x80  \xff" ;
-reverse ;
-EXPECT
-\x80 will produce malformed UTF-8 character; use \x{80} for that at - line 10.
-\xff will produce malformed UTF-8 character; use \x{ff} for that at - line 10.
diff --git a/t/pragma/warn/sv b/t/pragma/warn/sv
index 9a2428e0cb..758137f2e8 100644
--- a/t/pragma/warn/sv
+++ b/t/pragma/warn/sv
@@ -269,25 +269,6 @@ EXPECT
 Undefined value assigned to typeglob at - line 3.
 ########
 # sv.c
-BEGIN {
-    if (ord("\t") == 5) {
-        print "SKIPPED\n# ebcdic \\x characters differ.";
-        exit 0;
-    }
-}
-use utf8 ;
-$^W =0 ;
-{
-  use warnings 'utf8' ;
-  my $a = rindex "a\xff bc ", "bc" ;
-  no warnings 'utf8' ;
-  $a = rindex "a\xff bc ", "bc" ;
-}
-my $a = rindex "a\xff bc ", "bc" ;
-EXPECT
-\xff will produce malformed UTF-8 character; use \x{ff} for that at - line 12.
-########
-# sv.c
 use warnings 'y2k';
 use Config;
 BEGIN {
diff --git a/t/pragma/warn/toke b/t/pragma/warn/toke
index 271ef6365c..cfdea78d3c 100644
--- a/t/pragma/warn/toke
+++ b/t/pragma/warn/toke
@@ -89,10 +89,6 @@ toke.c	AOK
 	sub time {} 
 	my $a = time()
 
-    \x%.*s will produce malformed UTF-8 character; use \x{%.*s} for that
-        use utf8 ; 
-	$_ = "\xffe"
-
     Unrecognized escape \\%c passed through
         $a = "\m" ;
 
@@ -447,21 +443,6 @@ EXPECT
 
 ########
 # toke.c
-BEGIN {
-    if (ord("\t") == 5) {
-        print "SKIPPED\n# Ebcdic platforms have different \\x constructs.";
-        exit 0;
-    }
-}
-use warnings 'utf8' ;
-use utf8 ;
-$_ = " \xffe " ;
-no warnings 'utf8' ;
-$_ = " \xffe " ;
-EXPECT
-\xff will produce malformed UTF-8 character; use \x{ff} for that at - line 10.
-########
-# toke.c
 my $a = rand + 4 ;
 EXPECT
 Warning: Use of "rand" without parens is ambiguous at - line 2.
diff --git a/t/pragma/warn/utf8 b/t/pragma/warn/utf8
index cb1f202b8d..6a2fe5446c 100644
--- a/t/pragma/warn/utf8
+++ b/t/pragma/warn/utf8
@@ -14,48 +14,16 @@
      <<<<<< Add a test when somethig actually calls utf16_to_utf8
 
 __END__
-# utf8.c [utf8_to_uv]
+# utf8.c [utf8_to_uv] -W
 use utf8 ;
-my $a = ord "\x80" ;
-EXPECT
-########
-# utf8.c [utf8_to_uv]
-BEGIN {
-    if (ord("\t") == 5) {
-        print "SKIPPED\n# Ebcdic platforms have different \\x constructs.";
-        exit 0;
-    }
-}
-use utf8 ;
-my $a = ord "\x80" ;
+my $a = "snøstorm" ;
 {
-    use warnings 'utf8' ;
-    my $a = ord "\x80" ;
     no warnings 'utf8' ;
-    my $a = ord "\x80" ;
-}
-EXPECT
-\x80 will produce malformed UTF-8 character; use \x{80} for that at - line 12.
-########
-# utf8.c [utf8_to_uv]
-use utf8 ;
-my $a = ord "\xf080" ;
-EXPECT
-########
-# utf8.c [utf8_to_uv]
-BEGIN {
-    if (ord("\t") == 5) {
-        print "SKIPPED\n# Ebcdic platforms have different \\x constructs.";
-        exit 0;
-    }
-}
-use utf8 ;
-my $a = ord "\xf080" ;
-{
+    my $a = "snøstorm";
     use warnings 'utf8' ;
-    my $a = ord "\xf080" ;
-    no warnings 'utf8' ;
-    my $a = ord "\xf080" ;
+    my $a = "snøstorm";
 }
 EXPECT
-\xf0 will produce malformed UTF-8 character; use \x{f0} for that at - line 12.
+Malformed UTF-8 character at - line 3.
+Malformed UTF-8 character at - line 8.
+########