Fix for the :utf8 read() bug noticed by Matt Sergeant:

"large enough" Unicode characters returned more than one
as their "Unicode size".

p4raw-id: //depot/perl@12981

diff --git a/pp_sys.c b/pp_sys.c
index ea35136..ed70307 100644 (file)
--- a/pp_sys.c
+++ b/pp_sys.c
@@ -1517,6 +1517,9 @@ PP(pp_sysread)
     int fp_utf8;
     Size_t got = 0;
     Size_t wanted;
+    bool charstart = NULL;
+    STRLEN skip;
+    STRLEN charskip;
 
     gv = (GV*)*++MARK;
     if ((PL_op->op_type == OP_READ || PL_op->op_type == OP_SYSREAD)
@@ -1563,6 +1566,9 @@ PP(pp_sysread)
        DIE(aTHX_ "Negative length");
     wanted = length;
 
+    charstart = TRUE;
+    charskip  = 0;
+
 #ifdef HAS_SOCKET
     if (PL_op->op_type == OP_RECV) {
        char namebuf[MAXPATHLEN];
@@ -1683,23 +1689,30 @@ PP(pp_sysread)
        /* Look at utf8 we got back and count the characters */
        char *bend = buffer + count;
        while (buffer < bend) {
-           STRLEN skip = UTF8SKIP(buffer);
-           if (buffer+skip > bend) {
+           if (charstart) {
+               skip = UTF8SKIP(buffer);
+               charskip = 0;
+           }
+           if (buffer - charskip + skip > bend) {
                /* partial character - try for rest of it */
                length = skip - (bend-buffer);
                offset = bend - SvPVX(bufsv);
+               charstart = FALSE;
+               charskip += count;
                goto more_bytes;
            }
            else {
                got++;
                buffer += skip;
+               charstart = TRUE;
+               charskip  = 0;
            }
         }
        /* If we have not 'got' the number of _characters_ we 'wanted' get some more
           provided amount read (count) was what was requested (length)
         */
        if (got < wanted && count == length) {
-           length = (wanted-got);
+           length = wanted - got;
            offset = bend - SvPVX(bufsv);
            goto more_bytes;
        }

diff --git a/t/io/utf8.t b/t/io/utf8.t
index 1a7d27f..2d74cff 100755 (executable)
--- a/t/io/utf8.t
+++ b/t/io/utf8.t
@@ -12,9 +12,7 @@ BEGIN {
 no utf8; # needed for use utf8 not griping about the raw octets
 
 $| = 1;
-my $total_tests = 25;
-if (ord('A') == 193) { $total_tests = 24; } # EBCDIC platforms do not warn on UTF-8
-print "1..$total_tests\n";
+print "1..26\n";
 
 open(F,"+>:utf8",'a');
 print F chr(0x100).'£';
@@ -175,12 +173,38 @@ print "not " unless $x eq $chr;
 print "ok 24\n";
 
 # Now we have a deformed file.
-open F, "<:utf8", "a" or die $!;
-$x = <F>; chomp $x;
-{ local $SIG{__WARN__} = sub { print "ok 25\n"; };
-eval { sprintf "%vd\n", $x; }
+
+if (ord('A') == 193) {
+    print "ok 25 # Skip: EBCDIC\n"; # EBCDIC doesn't complain
+} else {
+    open F, "<:utf8", "a" or die $!;
+    $x = <F>; chomp $x;
+    local $SIG{__WARN__} = sub { print "ok 25\n" };
+    eval { sprintf "%vd\n", $x };
 }
 
 close F;
 unlink('a');
 
+open F, ">a";
+@a = map { chr(1 << ($_ << 2)) } 0..5; # 0x1, 0x10, .., 0x100000
+print F @a;
+close F;
+open F, "<:utf8", "a";
+$a = 0;
+for (@a) {
+    unless (read(F, $b, 1) == 1  &&
+            length($b)     == 1  &&
+            ord($b)        == ord($_) &&
+            tell(F)        == ($a += bytes::length($b))) {
+        print '# ord($_)    == ', ord($_), "\n";
+        print '# ord($b)    == ', ord($b), "\n";
+        print '# length($b) == ', length($b), "\n";
+        print '# tell(F)    == ', tell(F), "\n";
+        print "not ";
+        last;
+    }
+}
+print "ok 26\n";
+
+END { 1 while unlink "a" }