BEGIN {
chdir 't' if -d 't';
- @INC = '../lib';
- require './test.pl';
+ require './test.pl'; require './charset_tools.pl';
+ set_up_inc(qw '../lib ../dist/Math-BigInt/lib');
}
# This is truth in an if statement, and could be a skip message
my $no_signedness = $] > 5.009 ? '' :
"Signed/unsigned pack modifiers not available on this perl";
-plan tests => 13856;
+plan tests => 14713;
use strict;
-use warnings;
+use warnings qw(FATAL all);
use Config;
-my $Is_EBCDIC = (defined $Config{ebcdic} && $Config{ebcdic} eq 'define');
my $Perl = which_perl();
my @valid_errors = (qr/^Invalid type '\w'/);
}
for my $size ( 16, 32, 64 ) {
- if (exists $Config{"u${size}size"} and $Config{"u${size}size"} != ($size >> 3)) {
+ if (defined $Config{"u${size}size"} and ($Config{"u${size}size"}||0) != ($size >> 3)) {
push @valid_errors, qr/^Perl_my_$maybe_not_avail$size\(\) not available/;
}
}
{
my $sum = 129; # ASCII
- $sum = 103 if $Is_EBCDIC;
+ $sum = 103 if $::IS_EBCDIC;
my $x;
is( ($x = unpack("%32B*", "Now is the time for all good blurfl")), $sum );
my $foo;
open(BIN, $Perl) || die "Can't open $Perl: $!\n";
+ binmode BIN;
sysread BIN, $foo, 8192;
close BIN;
# Is this a stupid thing to do on VMS, VOS and other unusual platforms?
skip("-- the IEEE infinity model is unavailable in this configuration.", 1)
- if (($^O eq 'VMS') && !defined($Config{useieee}));
+ if (($^O eq 'VMS') && !defined($Config{useieee}) || !$Config{d_double_has_inf});
skip("-- $^O has serious fp indigestion on w-packed infinities", 1)
if (
- ($^O eq 'mpeix')
- ||
($^O eq 'ultrix')
||
($^O =~ /^svr4/ && -f "/etc/issue" && -f "/etc/.relid") # NCR MP-RAS
if ($^O eq 'vos');
eval { $x = pack 'w', $inf };
- like ($@, qr/^Cannot compress integer/, "Cannot compress integer");
+ like ($@, qr/^Cannot compress Inf/, "Cannot compress infinity");
}
SKIP: {
skip("-- the full range of an IEEE double may not be available in this configuration.", 3)
- if (($^O eq 'VMS') && !defined($Config{useieee}));
+ if (($^O eq 'VMS') && !defined($Config{useieee}) || !$Config{d_double_style_ieee});
skip("-- $^O does not like 2**1023", 3)
if (($^O eq 'ultrix'));
# temps
sub foo { my $a = "a"; return $a . $a++ . $a++ }
{
- use warnings;
+ use warnings qw(NONFATAL all);;
my $warning;
local $SIG{__WARN__} = sub {
$warning = $_[0];
print "# test unpack-pack lengths\n";
-my @templates = qw(c C i I s S l L n N v V f d q Q);
+my @templates = qw(c C W i I s S l L n N v V f d q Q);
foreach my $base (@templates) {
my @tmpl = ($base);
- $base =~ /^[cnv]/i or push @tmpl, "$base>", "$base<";
+ $base =~ /^[cwnv]/i or push @tmpl, "$base>", "$base<";
foreach my $t (@tmpl) {
SKIP: {
my @t = eval { unpack("$t*", pack("$t*", 12, 34)) };
['p', 'Z3', "foo", "fo\0"],
['u', 'Z*', "foo\0bar \0", "foo"],
['u', 'Z8', "foo\0bar \0", "foo"],
-)
+)
{
my ($what, $template, $in, $out) = @$_;
my $got = $what eq 'u' ? (unpack $template, $in) : (pack $template, $in);
}
if ($calc_sum == $calc_sum - 1 && $calc_sum == $max_p1) {
# we're into floating point (either by getting out of the range of
- # UV arithmetic, or because we're doing a floating point checksum)
+ # UV arithmetic, or because we're doing a floating point checksum)
# and our calculation of the checksum has become rounded up to
# max_checksum + 1
$calc_sum = 0;
numbers ('c', -128, -1, 0, 1, 127);
numbers ('C', 0, 1, 127, 128, 255);
+numbers ('W', 0, 1, 127, 128, 255, 256, 0x7ff, 0x800, 0xfffd);
numbers ('s', -32768, -1, 0, 1, 32767);
numbers ('S', 0, 1, 32767, 32768, 65535);
numbers ('i', -2147483648, -1, 0, 1, 2147483647);
numbers_with_total ('q', -1,
-9223372036854775808, -1, 0, 1,9223372036854775807);
# This total is icky, but the true total is 2**65-1, and need a way to generate
-# the epxected checksum on any system including those where NVs can preserve
+# the expected checksum on any system including those where NVs can preserve
# 65 bits. (long double is 128 bits on sparc, so they certainly can)
# or where rounding is down not up on binary conversion (crays)
numbers_with_total ('Q', sub {
skip "cannot pack '$format' on this perl", 5
if is_valid_error($@);
- print "# [$value][$nat][$be][$le][$@]\n";
+ {
+ use warnings qw(NONFATAL utf8);
+ print "# [$value][$nat][$be][$le][$@]\n";
+ }
SKIP: {
skip "cannot compare native byteorder with big-/little-endian", 1
{
# /
- my ($x, $y, $z);
+ my ($x, $y, $z, @a);
eval { ($x) = unpack '/a*','hello' };
like($@, qr!'/' must follow a numeric type!);
undef $x;
eval { $x = unpack '/a*','hello' };
like($@, qr!'/' must follow a numeric type!);
+ # [perl #60204] Unhelpful error message from unpack
+ eval { @a = unpack 'v/a*','h' };
+ is($@, '');
+ is(scalar @a, 0);
+ eval { $x = unpack 'v/a*','h' };
+ is($@, '');
+ is($x, undef);
+
undef $x;
eval { ($z,$x,$y) = unpack 'a3/A C/a* C/Z', "003ok \003yes\004z\000abc" };
is($@, '');
foreach (
['a/a*/a*', '212ab345678901234567','ab3456789012'],
['a/a*/a*', '3012ab345678901234567', 'ab3456789012'],
- ['a/a*/b*', '212ab', $Is_EBCDIC ? '100000010100' : '100001100100'],
- )
+ ['a/a*/b*', '212ab', $::IS_EBCDIC ? '100000010100' : '100001100100'],
+ )
{
my ($pat, $in, $expect) = @$_;
undef $x;
eval { ($x) = unpack $pat, $in };
is($@, '');
- is($x, $expect) ||
+ is($x, $expect) ||
printf "# list unpack ('$pat', '$in') gave %s, expected '$expect'\n",
encode_list ($x);
}
-SKIP: {
- skip("(EBCDIC and) version strings are bad idea", 2) if $Is_EBCDIC;
-
- is("1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000));
- is("1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000));
+{
+ is("1.20.300.4000", sprintf "%vd", pack("U*",utf8::native_to_unicode(1),utf8::native_to_unicode(20),300,4000));
+ is("1.20.300.4000", sprintf "%vd", pack(" U*",utf8::native_to_unicode(1),utf8::native_to_unicode(20),300,4000));
}
-isnt(v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000));
+isnt(v1.20.300.4000, sprintf "%vd", pack("C0U*",utf8::native_to_unicode(1),utf8::native_to_unicode(20),300,4000));
-my $rslt = $Is_EBCDIC ? "156 67" : "199 162";
-is(join(" ", unpack("C*", chr(0x1e2))), $rslt);
+my $rslt = join " ", map { ord } split "", byte_utf8a_to_utf8n("\xc7\xa2");
+# The ASCII UTF-8 of U+1E2 is "\xc7\xa2"
+is(join(" ", unpack("U0 C*", chr(0x1e2))), $rslt);
# does pack U create Unicode?
is(ord(pack('U', 300)), 300);
# is unpack U the reverse of pack U for byte string?
is("@{[unpack('U*', pack('U*', 100, 200))]}", "100 200");
-
-SKIP: {
- skip "Not for EBCDIC", 4 if $Is_EBCDIC;
-
- # does unpack C unravel pack U?
- is("@{[unpack('C*', pack('U*', 100, 200))]}", "100 195 136");
-
+{
# does pack U0C create Unicode?
- is("@{[pack('U0C*', 100, 195, 136)]}", v100.v200);
+ my $cp202 = chr(202);
+ utf8::upgrade $cp202;
+ my @bytes202;
+ { # This is portable across character sets
+ use bytes;
+ @bytes202 = map { ord } split "", $cp202;
+ }
+
+ # This test requires the first number to be invariant; 64 is invariant on
+ # ASCII and EBCDIC.
+ is("@{[pack('U0C*', 64, @bytes202)]}", v64.v202);
# does pack C0U create characters?
- is("@{[pack('C0U*', 100, 200)]}", pack("C*", 100, 195, 136));
+ # The U* is expecting Unicode, so convert to that.
+ is("@{[pack('C0U*', map { utf8::native_to_unicode($_) } 64, 202)]}",
+ pack("C*", 64, @bytes202));
# does unpack U0U on byte data warn?
{
+ use warnings qw(NONFATAL all);;
+
+ my $bad = pack("U0C", 202);
local $SIG{__WARN__} = sub { $@ = "@_" };
- my @null = unpack('U0U', chr(255));
- like($@, qr/^Malformed UTF-8 character /);
+ my @null = unpack('U0U', $bad);
+ like($@, qr/^Malformed UTF-8 character: /);
}
}
['@4', 'N', "\0"x4],
['a*@8a*', 'Camel', 'Dromedary', "Camel\0\0\0Dromedary"],
['a*@4a', 'Perl rules', '!', 'Perl!'],
-)
+)
{
my ($template, @in) = @$_;
my $out = pop @in;
['@3', "ice"],
['@2a2', "water", "te"],
['a*@1a3', "steam", "steam", "tea"],
-)
+)
{
my ($template, $in, @out) = @$_;
my @got = eval {unpack $template, $in};
}
{ # more on grouping (W.Laun)
- use warnings;
- my $warning;
- local $SIG{__WARN__} = sub {
- $warning = $_[0];
- };
# @ absolute within ()-group
my $badc = pack( '(a)*', unpack( '(@1a @0a @2)*', 'abcd' ) );
is( $badc, 'badc' );
my @a = unpack( '(@1c)((@2c)@3c)', $buf );
is( "@a", "@b" );
- # various unpack count/code scenarios
+ # various unpack count/code scenarios
my @Env = ( a => 'AAA', b => 'BBB' );
my $env = pack( 'S(S/A*S/A*)*', @Env/2, @Env );
# 2 4 5 7 10 1213
eval { @pup = unpack( 'S/(S/A* S/A*)', substr( $env, 0, 13 ) ) };
like( $@, qr{length/code after end of string} );
-
+
# postfix repeat count
$env = pack( '(S/A* S/A*)' . @Env/2, @Env );
}
{ # syntax checks (W.Laun)
- use warnings;
+ use warnings qw(NONFATAL all);;
my @warning;
local $SIG{__WARN__} = sub {
push( @warning, $_[0] );
eval { my @inf = unpack( 'c/*a', "\x03AAA\x02BB" ); };
like( $@, qr{'/' does not take a repeat count} );
- # white space where possible
+ # white space where possible
my @Env = ( a => 'AAA', b => 'BBB' );
my $env = pack( ' S ( S / A* S / A* )* ', @Env/2, @Env );
my @pup = unpack( ' S / ( S / A* S / A* ) ', $env );
# comma warning only once
@warning = ();
$x = pack( 'C(C,C)C,C', 65..71 );
- like( scalar @warning, 1 );
+ cmp_ok( scalar(@warning), '==', 1 );
# forbidden code in []
eval { my $x = pack( 'A[@4]', 'XXXX' ); };
# @ repeat default 1
my $s = pack( 'AA@A', 'A', 'B', 'C' );
my @c = unpack( 'AA@A', $s );
- is( $s, 'AC' );
- is( "@c", "A C C" );
+ is( $s, 'AC' );
+ is( "@c", "A C C" );
# no unpack code after /
eval { my @a = unpack( "C/", "\3" ); };
}
{ # Repeat count [SUBEXPR]
- my @codes = qw( x A Z a c C B b H h s v n S i I l V N L p P f F d
+ my @codes = qw( x A Z a c C W B b H h s v n S i I l V N L p P f F d
s! S! i! I! l! L! j J);
my $G;
if (eval { pack 'q', 1 } ) {
@val{@codes} = map { / [Xx] (?{ undef })
| [AZa] (?{ 'something' })
| C (?{ 214 })
+ | W (?{ 8188 })
| c (?{ 114 })
| [Bb] (?{ '101' })
| [Hh] (?{ 'b8' })
| [svnSiIlVNLqQjJ] (?{ 10111 })
- | [FfDd] (?{ 1.36514538e67 })
+ | [FfDd] (?{ 1.36514538e37 })
| [pP] (?{ "try this buffer" })
/x; $^R } @codes;
my @end = (0x12345678, 0x23456781, 0x35465768, 0x15263748);
my $p = eval { pack $junk1, @list2 };
skip "cannot pack '$type' on this perl", 12
if is_valid_error($@);
+ die "pack $junk1 failed: $@" if $@;
my $half = int( (length $p)/2 );
for my $move ('', "X$half", "X!$half", 'x1', 'x!8', "x$half") {
$b =~ s/(?:17000+|16999+)\d+(e-45) /17$1 /gi; # stringification is gamble
is($b, "@a @a");
+ use warnings qw(NONFATAL all);;
my $warning;
local $SIG{__WARN__} = sub {
$warning = $_[0];
is(unpack('c'), 65, "one-arg unpack (change #18751)"); # defaulting to $_
{
- my $a = "X\t01234567\n" x 100;
+ my $a = "X\x0901234567\n" x 100; # \t would not be EBCDIC TAB
my @a = unpack("(a1 c/a)*", $a);
is(scalar @a, 200, "[perl #15288]");
is($a[-1], "01234567\n", "[perl #15288]");
is($a[-2], "X", "[perl #15288]");
}
+{
+ use warnings qw(NONFATAL all);;
+ my $warning;
+ local $SIG{__WARN__} = sub {
+ $warning = $_[0];
+ };
+
+ # This test is looking for the encoding of the bit pattern "\x66\x6f\x6f",
+ # which is ASCII "foo"
+ my $out = pack("u99", native_to_uni("foo") x 99);
+ like($warning, qr/Field too wide in 'u' format in pack at /,
+ "Warn about too wide uuencode");
+ is($out, ("_" . "9F]O" x 21 . "\n") x 4 . "M" . "9F]O" x 15 . "\n",
+ "Use max width in case of too wide uuencode");
+}
+
# checksums
{
# verify that unpack advances correctly wrt a checksum
my (@x) = unpack("b10a", "abcd");
my (@y) = unpack("%b10a", "abcd");
is($x[1], $y[1], "checksum advance ok");
+
+ SKIP: {
+ skip("-- non-IEEE float", 1) if !$Config{d_double_style_ieee};
+ # verify that the checksum is not overflowed with C0
+ is(unpack("C0%128U", "abcd"), unpack("U0%128U", "abcd"), "checksum not overflowed");
+ }
+}
+
+my $U_1FFC_bytes = byte_utf8a_to_utf8n("\341\277\274");
+{
+ # U0 and C0 must be scoped
+ my (@x) = unpack("a(U0)U", "b$U_1FFC_bytes");
+ is($x[0], 'b', 'before scope');
+ is($x[1], 8188, 'after scope');
+
+ is(pack("a(U0)U", "b", 8188), "b$U_1FFC_bytes");
+}
+
+{
+ # counted length prefixes shouldn't change C0/U0 mode
+ # (note the length is actually 0 in this test, as the C/ is replaced by C0
+ # due to the \0 in the string)
+ is(join(',', unpack("aC/UU", "b\0$U_1FFC_bytes")), 'b,8188');
+ is(join(',', unpack("aC/CU", "b\0$U_1FFC_bytes")), 'b,8188');
+
+ # The U expects Unicode, so convert from native
+ my $first_byte = utf8::native_to_unicode(ord substr($U_1FFC_bytes, 0, 1));
+
+ is(join(',', unpack("aU0C/UU", "b\0$U_1FFC_bytes")), "b,$first_byte");
+ is(join(',', unpack("aU0C/CU", "b\0$U_1FFC_bytes")), "b,$first_byte");
+}
+
+{
+ # "Z0" (bug #34062)
+ my (@x) = unpack("C*", pack("CZ0", 1, "b"));
+ is(join(',', @x), '1', 'pack Z0 doesn\'t destroy the character before');
+}
+
+{
+ # Encoding neutrality
+ # String we will pull apart and rebuild in several ways:
+ my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06";
+ my $up = $down;
+ utf8::upgrade($up);
+
+ my %expect =
+ # [expected result,
+ # how many chars it should progress,
+ # (optional) expected result of pack]
+ (a5 => ["\xf8\xf9\xfa\xfb\xfc", 5],
+ A5 => ["\xf8\xf9\xfa\xfb\xfc", 5],
+ Z5 => ["\xf8\xf9\xfa\xfb\xfc", 5, "\xf8\xf9\xfa\xfb\x00\xfd"],
+ b21 => ["000111111001111101011", 3, "\xf8\xf9\x1a\xfb"],
+ B21 => ["111110001111100111111", 3, "\xf8\xf9\xf8\xfb"],
+ H5 => ["f8f9f", 3, "\xf8\xf9\xf0\xfb"],
+ h5 => ["8f9fa", 3, "\xf8\xf9\x0a\xfb"],
+ "s<" => [-1544, 2],
+ "s>" => [-1799, 2],
+ "S<" => [0xf9f8, 2],
+ "S>" => [0xf8f9, 2],
+ "l<" => [-67438088, 4],
+ "l>" => [-117835013, 4],
+ "L>" => [0xf8f9fafb, 4],
+ "L<" => [0xfbfaf9f8, 4],
+ n => [0xf8f9, 2],
+ N => [0xf8f9fafb, 4],
+ v => [63992, 2],
+ V => [0xfbfaf9f8, 4],
+ c => [-8, 1],
+ U0U => [0xf8, 1],
+ w => ["8715569050387726213", 9],
+ q => ["-283686952306184", 8],
+ Q => ["18446460386757245432", 8],
+ );
+
+ for my $string ($down, $up) {
+ for my $format (sort {lc($a) cmp lc($b) || $a cmp $b } keys %expect) {
+ SKIP: {
+ my $expect = $expect{$format};
+ # unpack upgraded and downgraded string
+ my @result = eval { unpack("$format C0 W", $string) };
+ skip "cannot pack/unpack '$format C0 W' on this perl", 5 if
+ $@ && is_valid_error($@);
+ is(@result, 2, "Two results from unpack $format C0 W");
+
+ # pack to downgraded
+ my $new = pack("$format C0 W", @result);
+ is(length($new), $expect->[1]+1,
+ "pack $format C0 W should give $expect->[1]+1 chars");
+ is($new, $expect->[2] || substr($string, 0, length $new),
+ "pack $format C0 W returns expected value");
+
+ # pack to upgraded
+ $new = pack("a0 $format C0 W", chr(256), @result);
+ is(length($new), $expect->[1]+1,
+ "pack a0 $format C0 W should give $expect->[1]+1 chars");
+ is($new, $expect->[2] || substr($string, 0, length $new),
+ "pack a0 $format C0 W returns expected value");
+ }
+ }
+ }
+}
+
+{
+ # Encoding neutrality, numbers
+ my $val = -2.68;
+ for my $format (qw(s S i I l L j J f d F D q Q
+ s! S! i! I! l! L! n! N! v! V!)) {
+ SKIP: {
+ my $down = eval { pack($format, $val) };
+ skip "cannot pack/unpack $format on this perl", 9 if
+ $@ && is_valid_error($@);
+ ok(!utf8::is_utf8($down), "Simple $format pack doesn't get upgraded");
+ my $up = pack("a0 $format", chr(256), $val);
+ ok(utf8::is_utf8($up), "a0 $format with high char leads to upgrade");
+ is($down, $up, "$format generated strings are equal though");
+ my @down_expanded = unpack("$format W", $down . chr(0xce));
+ is(@down_expanded, 2, "Expand to two values");
+ is($down_expanded[1], 0xce,
+ "unpack $format left us at the expected position");
+ my @up_expanded = unpack("$format W", $up . chr(0xce));
+ is(@up_expanded, 2, "Expand to two values");
+ is($up_expanded[1], 0xce,
+ "unpack $format left us at the expected position");
+ is($down_expanded[0], $up_expanded[0], "$format unpack was neutral");
+ is(pack($format, $down_expanded[0]), $down, "Pack $format undoes unpack $format");
+ }
+ }
+}
+
+{
+ # C *is* neutral
+ my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06";
+ my $up = $down;
+ utf8::upgrade($up);
+ my @down = unpack("C*", $down);
+ my @expect_down = (0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, 0x05, 0x06);
+ is("@down", "@expect_down", "byte expand");
+ is(pack("C*", @down), $down, "byte join");
+
+ my @up = unpack("C*", $up);
+ my @expect_up = (0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, 0x05, 0x06);
+ is("@up", "@expect_up", "UTF-8 expand");
+ is(pack("U0C0C*", @up), $up, "UTF-8 join");
+}
+
+{
+ # Harder cases for the neutrality test
+
+ # u format
+ my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06";
+ my $up = $down;
+ utf8::upgrade($up);
+ is(pack("u", $down), pack("u", $up), "u pack is neutral");
+ is(unpack("u", pack("u", $down)), $down, "u unpack to downgraded works");
+ is(unpack("U0C0u", pack("u", $down)), $up, "u unpack to upgraded works");
+
+ # p/P format
+ # This actually only tests something if the address contains a byte >= 0x80
+ my $str = "abc\xa5\x00\xfede";
+ $down = pack("p", $str);
+ is(pack("P", $str), $down);
+ is(pack("U0C0p", $str), $down);
+ is(pack("U0C0P", $str), $down);
+ is(unpack("p", $down), "abc\xa5", "unpack p downgraded");
+ $up = $down;
+ utf8::upgrade($up);
+ is(unpack("p", $up), "abc\xa5", "unpack p upgraded");
+
+ is(unpack("P7", $down), "abc\xa5\x00\xfed", "unpack P downgraded");
+ is(unpack("P7", $up), "abc\xa5\x00\xfed", "unpack P upgraded");
+
+ # x, X and @
+ $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06";
+ $up = $down;
+ utf8::upgrade($up);
+
+ is(unpack('@4W', $down), 0xfc, "\@positioning on downgraded string");
+ is(unpack('@4W', $up), 0xfc, "\@positioning on upgraded string");
+
+ is(unpack('@4x2W', $down), 0xfe, "x moving on downgraded string");
+ is(unpack('@4x2W', $up), 0xfe, "x moving on upgraded string");
+ is(unpack('@4x!4W', $down), 0xfc, "x! moving on downgraded string");
+ is(unpack('@4x!4W', $up), 0xfc, "x! moving on upgraded string");
+ is(unpack('@5x!4W', $down), 0x05, "x! moving on downgraded string");
+ is(unpack('@5x!4W', $up), 0x05, "x! moving on upgraded string");
+
+ is(unpack('@4X2W', $down), 0xfa, "X moving on downgraded string");
+ is(unpack('@4X2W', $up), 0xfa, "X moving on upgraded string");
+ is(unpack('@4X!4W', $down), 0xfc, "X! moving on downgraded string");
+ is(unpack('@4X!4W', $up), 0xfc, "X! moving on upgraded string");
+ is(unpack('@5X!4W', $down), 0xfc, "X! moving on downgraded string");
+ is(unpack('@5X!4W', $up), 0xfc, "X! moving on upgraded string");
+ is(unpack('@5X!8W', $down), 0xf8, "X! moving on downgraded string");
+ is(unpack('@5X!8W', $up), 0xf8, "X! moving on upgraded string");
+
+ is(pack("W2x", 0xfa, 0xe3), "\xfa\xe3\x00", "x on downgraded string");
+ is(pack("W2x!4", 0xfa, 0xe3), "\xfa\xe3\x00\x00",
+ "x! on downgraded string");
+ is(pack("W2x!2", 0xfa, 0xe3), "\xfa\xe3", "x! on downgraded string");
+ is(pack("U0C0W2x", 0xfa, 0xe3), "\xfa\xe3\x00", "x on upgraded string");
+ is(pack("U0C0W2x!4", 0xfa, 0xe3), "\xfa\xe3\x00\x00",
+ "x! on upgraded string");
+ is(pack("U0C0W2x!2", 0xfa, 0xe3), "\xfa\xe3", "x! on upgraded string");
+ is(pack("W2X", 0xfa, 0xe3), "\xfa", "X on downgraded string");
+ is(pack("U0C0W2X", 0xfa, 0xe3), "\xfa", "X on upgraded string");
+ is(pack("W2X!2", 0xfa, 0xe3), "\xfa\xe3", "X! on downgraded string");
+ is(pack("U0C0W2X!2", 0xfa, 0xe3), "\xfa\xe3", "X! on upgraded string");
+ is(pack("W3X!2", 0xfa, 0xe3, 0xa6), "\xfa\xe3", "X! on downgraded string");
+ is(pack("U0C0W3X!2", 0xfa, 0xe3, 0xa6), "\xfa\xe3",
+ "X! on upgraded string");
+
+ # backward eating through a ( moves the group starting point backwards
+ is(pack("a*(Xa)", "abc", "q"), "abq",
+ "eating before strbeg moves it back");
+ is(pack("a*(Xa)", "ab" . chr(512), "q"), "abq",
+ "eating before strbeg moves it back");
+
+ # Check marked_upgrade
+ is(pack('W(W(Wa@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, "a", 0xa4, 0xa5, 0xa6),
+ "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6");
+ $up = "a";
+ utf8::upgrade($up);
+ is(pack('W(W(Wa@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, $up, 0xa4, 0xa5, 0xa6),
+ "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6", "marked upgrade caused by a");
+ is(pack('W(W(WW@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, 256, 0xa4, 0xa5, 0xa6),
+ "\xa1\xa2\xa3\x{100}\x00\xa4\x00\xa5\x00\xa6",
+ "marked upgrade caused by W");
+ is(pack('W(W(WU0aC0@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, "a", 0xa4, 0xa5, 0xa6),
+ "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6", "marked upgrade caused by U0");
+
+ # a, A and Z
+ $down = "\xa4\xa6\xa7";
+ $up = $down;
+ utf8::upgrade($up);
+ utf8::upgrade(my $high = "\xfeb");
+
+ for my $format ("a0", "A0", "Z0", "U0a0C0", "U0A0C0", "U0Z0C0") {
+ is(pack("a* $format a*", "ab", $down, "cd"), "abcd",
+ "$format format on plain string");
+ is(pack("a* $format a*", "ab", $up, "cd"), "abcd",
+ "$format format on upgraded string");
+ is(pack("a* $format a*", $high, $down, "cd"), "\xfebcd",
+ "$format format on plain string");
+ is(pack("a* $format a*", $high, $up, "cd"), "\xfebcd",
+ "$format format on upgraded string");
+ my @down = unpack("a1 $format a*", "\xfeb");
+ is("@down", "\xfe b", "unpack $format");
+ my @up = unpack("a1 $format a*", $high);
+ is("@up", "\xfe b", "unpack $format");
+ }
+ is(pack("a1", $high), "\xfe");
+ is(pack("A1", $high), "\xfe");
+ is(pack("Z1", $high), "\x00");
+ is(pack("a2", $high), "\xfeb");
+ is(pack("A2", $high), "\xfeb");
+ is(pack("Z2", $high), "\xfe\x00");
+ is(pack("a5", $high), "\xfeb\x00\x00\x00");
+ is(pack("A5", $high), "\xfeb ");
+ is(pack("Z5", $high), "\xfeb\x00\x00\x00");
+ is(pack("a*", $high), "\xfeb");
+ is(pack("A*", $high), "\xfeb");
+ is(pack("Z*", $high), "\xfeb\x00");
+
+ utf8::upgrade($high = byte_utf8a_to_utf8n("\xc3\xbe") . "b");
+ is(pack("U0a2", $high), uni_to_native("\xfe"));
+ is(pack("U0A2", $high), uni_to_native("\xfe"));
+ is(pack("U0Z1", $high), uni_to_native("\x00"));
+ is(pack("U0a3", $high), uni_to_native("\xfe") . "b");
+ is(pack("U0A3", $high), uni_to_native("\xfe") . "b");
+ is(pack("U0Z3", $high), uni_to_native("\xfe\x00"));
+ is(pack("U0a6", $high), uni_to_native("\xfe") . "b" . uni_to_native("\x00\x00\x00"));
+ is(pack("U0A6", $high), uni_to_native("\xfe") . "b ");
+ is(pack("U0Z6", $high), uni_to_native("\xfe") . "b" . uni_to_native("\x00\x00\x00"));
+ is(pack("U0a*", $high), uni_to_native("\xfe") . "b");
+ is(pack("U0A*", $high), uni_to_native("\xfe") . "b");
+ is(pack("U0Z*", $high), uni_to_native("\xfe") . "b" . uni_to_native("\x00"));
+}
+{
+ # pack /
+ my @array = 1..14;
+ my @out = unpack("N/S", pack("N/S", @array) . "abcd");
+ is("@out", "@array", "pack N/S works");
+ @out = unpack("N/S*", pack("N/S*", @array) . "abcd");
+ is("@out", "@array", "pack N/S* works");
+ @out = unpack("N/S*", pack("N/S14", @array) . "abcd");
+ is("@out", "@array", "pack N/S14 works");
+ @out = unpack("N/S*", pack("N/S15", @array) . "abcd");
+ is("@out", "@array", "pack N/S15 works");
+ @out = unpack("N/S*", pack("N/S13", @array) . "abcd");
+ is("@out", "@array[0..12]", "pack N/S13 works");
+ @out = unpack("N/S*", pack("N/S0", @array) . "abcd");
+ is("@out", "", "pack N/S0 works");
+ is(pack("Z*/a0", "abc"), "0\0", "pack Z*/a0 makes a short string");
+ is(pack("Z*/Z0", "abc"), "0\0", "pack Z*/Z0 makes a short string");
+ is(pack("Z*/a3", "abc"), "3\0abc", "pack Z*/a3 makes a full string");
+ is(pack("Z*/Z3", "abc"), "3\0ab\0", "pack Z*/Z3 makes a short string");
+ is(pack("Z*/a5", "abc"), "5\0abc\0\0", "pack Z*/a5 makes a long string");
+ is(pack("Z*/Z5", "abc"), "5\0abc\0\0", "pack Z*/Z5 makes a long string");
+ is(pack("Z*/Z"), "1\0\0", "pack Z*/Z makes an extended string");
+ is(pack("Z*/Z", ""), "1\0\0", "pack Z*/Z makes an extended string");
+ is(pack("Z*/a", ""), "0\0", "pack Z*/a makes an extended string");
+}
+{
+ # unpack("A*", $unicode) strips general unicode spaces
+ is(unpack("A*", "ab \n" . uni_to_native("\xa0") . " \0"), "ab \n" . uni_to_native("\xa0"),
+ 'normal A* strip leaves \xa0');
+ is(unpack("U0C0A*", "ab \n" . uni_to_native("\xa0") . " \0"), "ab \n" . uni_to_native("\xa0"),
+ 'normal A* strip leaves \xa0 even if it got upgraded for technical reasons');
+ is(unpack("A*", pack("a*(U0U)a*", "ab \n", 0xa0, " \0")), "ab",
+ 'upgraded strings A* removes \xa0');
+ is(unpack("A*", pack("a*(U0UU)a*", "ab \n", 0xa0, 0x1680, " \0")), "ab",
+ 'upgraded strings A* removes all unicode whitespace');
+ is(unpack("A5", pack("a*(U0U)a*", "ab \n", 0x1680, "def", "ab")), "ab",
+ 'upgraded strings A5 removes all unicode whitespace');
+ is(unpack("A*", pack("U", 0x1680)), "",
+ 'upgraded strings A* with nothing left');
+}
+{
+ # Testing unpack . and .!
+ is(unpack(".", "ABCD"), 0, "offset at start of string is 0");
+ is(unpack(".", ""), 0, "offset at start of empty string is 0");
+ is(unpack("x3.", "ABCDEF"), 3, "simple offset works");
+ is(unpack("x3.", "ABC"), 3, "simple offset at end of string works");
+ is(unpack("x3.0", "ABC"), 0, "self offset is 0");
+ is(unpack("x3(x2.)", "ABCDEF"), 2, "offset is relative to inner group");
+ is(unpack("x3(X2.)", "ABCDEF"), -2,
+ "negative offset relative to inner group");
+ is(unpack("x3(X2.2)", "ABCDEF"), 1, "offset is relative to inner group");
+ is(unpack("x3(x2.0)", "ABCDEF"), 0, "self offset in group is still 0");
+ is(unpack("x3(x2.2)", "ABCDEF"), 5, "offset counts groups");
+ is(unpack("x3(x2.*)", "ABCDEF"), 5, "star offset is relative to start");
+
+ my $high = chr(8188) x 6;
+ is(unpack("x3(x2.)", $high), 2, "utf8 offset is relative to inner group");
+ is(unpack("x3(X2.)", $high), -2,
+ "utf8 negative offset relative to inner group");
+ is(unpack("x3(X2.2)", $high), 1, "utf8 offset counts groups");
+ is(unpack("x3(x2.0)", $high), 0, "utf8 self offset in group is still 0");
+ is(unpack("x3(x2.2)", $high), 5, "utf8 offset counts groups");
+ is(unpack("x3(x2.*)", $high), 5, "utf8 star offset is relative to start");
+
+ is(unpack("U0x3(x2.)", $high), 2,
+ "U0 mode utf8 offset is relative to inner group");
+ is(unpack("U0x3(X2.)", $high), -2,
+ "U0 mode utf8 negative offset relative to inner group");
+ is(unpack("U0x3(X2.2)", $high), 1,
+ "U0 mode utf8 offset counts groups");
+ is(unpack("U0x3(x2.0)", $high), 0,
+ "U0 mode utf8 self offset in group is still 0");
+ is(unpack("U0x3(x2.2)", $high), 5,
+ "U0 mode utf8 offset counts groups");
+ is(unpack("U0x3(x2.*)", $high), 5,
+ "U0 mode utf8 star offset is relative to start");
+
+ is(unpack("x3(x2.!)", $high), 2*3,
+ "utf8 offset is relative to inner group");
+ is(unpack("x3(X2.!)", $high), -2*3,
+ "utf8 negative offset relative to inner group");
+ is(unpack("x3(X2.!2)", $high), 1*3,
+ "utf8 offset counts groups");
+ is(unpack("x3(x2.!0)", $high), 0,
+ "utf8 self offset in group is still 0");
+ is(unpack("x3(x2.!2)", $high), 5*3,
+ "utf8 offset counts groups");
+ is(unpack("x3(x2.!*)", $high), 5*3,
+ "utf8 star offset is relative to start");
+
+ is(unpack("U0x3(x2.!)", $high), 2,
+ "U0 mode utf8 offset is relative to inner group");
+ is(unpack("U0x3(X2.!)", $high), -2,
+ "U0 mode utf8 negative offset relative to inner group");
+ is(unpack("U0x3(X2.!2)", $high), 1,
+ "U0 mode utf8 offset counts groups");
+ is(unpack("U0x3(x2.!0)", $high), 0,
+ "U0 mode utf8 self offset in group is still 0");
+ is(unpack("U0x3(x2.!2)", $high), 5,
+ "U0 mode utf8 offset counts groups");
+ is(unpack("U0x3(x2.!*)", $high), 5,
+ "U0 mode utf8 star offset is relative to start");
+}
+{
+ # Testing pack . and .!
+ is(pack("(a)5 .", 1..5, 3), "123", ". relative to string start, shorten");
+ eval { () = pack("(a)5 .", 1..5, -3) };
+ like($@, qr{'\.' outside of string in pack}, "Proper error message");
+ is(pack("(a)5 .", 1..5, 8), "12345\x00\x00\x00",
+ ". relative to string start, extend");
+ is(pack("(a)5 .", 1..5, 5), "12345", ". relative to string start, keep");
+
+ is(pack("(a)5 .0", 1..5, -3), "12",
+ ". relative to string current, shorten");
+ is(pack("(a)5 .0", 1..5, 2), "12345\x00\x00",
+ ". relative to string current, extend");
+ is(pack("(a)5 .0", 1..5, 0), "12345",
+ ". relative to string current, keep");
+
+ is(pack("(a)5 (.)", 1..5, -3), "12",
+ ". relative to group, shorten");
+ is(pack("(a)5 (.)", 1..5, 2), "12345\x00\x00",
+ ". relative to group, extend");
+ is(pack("(a)5 (.)", 1..5, 0), "12345",
+ ". relative to group, keep");
+
+ is(pack("(a)3 ((a)2 .)", 1..5, -2), "1",
+ ". relative to group, shorten");
+ is(pack("(a)3 ((a)2 .)", 1..5, 2), "12345",
+ ". relative to group, keep");
+ is(pack("(a)3 ((a)2 .)", 1..5, 4), "12345\x00\x00",
+ ". relative to group, extend");
+
+ is(pack("(a)3 ((a)2 .2)", 1..5, 2), "12",
+ ". relative to counted group, shorten");
+ is(pack("(a)3 ((a)2 .2)", 1..5, 7), "12345\x00\x00",
+ ". relative to counted group, extend");
+ is(pack("(a)3 ((a)2 .2)", 1..5, 5), "12345",
+ ". relative to counted group, keep");
+
+ is(pack("(a)3 ((a)2 .*)", 1..5, 2), "12",
+ ". relative to start, shorten");
+ is(pack("(a)3 ((a)2 .*)", 1..5, 7), "12345\x00\x00",
+ ". relative to start, extend");
+ is(pack("(a)3 ((a)2 .*)", 1..5, 5), "12345",
+ ". relative to start, keep");
+
+ is(pack('(a)5 (. @2 a)', 1..5, -3, "a"), "12\x00\x00a",
+ ". based shrink properly updates group starts");
+
+ is(pack("(W)3 ((W)2 .)", 0x301..0x305, -2), "\x{301}",
+ "utf8 . relative to group, shorten");
+ is(pack("(W)3 ((W)2 .)", 0x301..0x305, 2),
+ "\x{301}\x{302}\x{303}\x{304}\x{305}",
+ "utf8 . relative to group, keep");
+ is(pack("(W)3 ((W)2 .)", 0x301..0x305, 4),
+ "\x{301}\x{302}\x{303}\x{304}\x{305}\x00\x00",
+ "utf8 . relative to group, extend");
+
+ is(pack("(W)3 ((W)2 .!)", 0x301..0x305, -2), "\x{301}\x{302}",
+ "utf8 . relative to group, shorten");
+ is(pack("(W)3 ((W)2 .!)", 0x301..0x305, 4),
+ "\x{301}\x{302}\x{303}\x{304}\x{305}",
+ "utf8 . relative to group, keep");
+ is(pack("(W)3 ((W)2 .!)", 0x301..0x305, 6),
+ "\x{301}\x{302}\x{303}\x{304}\x{305}\x00\x00",
+ "utf8 . relative to group, extend");
+
+ is(pack('(W)5 (. @2 a)', 0x301..0x305, -3, "a"),
+ "\x{301}\x{302}\x00\x00a",
+ "utf8 . based shrink properly updates group starts");
+}
+{
+ # Testing @!
+ is(pack('a* @3', "abcde"), "abc", 'Test basic @');
+ is(pack('a* @!3', "abcde"), "abc", 'Test basic @!');
+ is(pack('a* @2', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{301}\x{302}",
+ 'Test basic utf8 @');
+ is(pack('a* @!2', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{301}",
+ 'Test basic utf8 @!');
+
+ is(unpack('@4 a*', "abcde"), "e", 'Test basic @');
+ is(unpack('@!4 a*', "abcde"), "e", 'Test basic @!');
+ is(unpack('@4 a*', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{305}",
+ 'Test basic utf8 @');
+ is(unpack('@!4 a*', "\x{301}\x{302}\x{303}\x{304}\x{305}"),
+ "\x{303}\x{304}\x{305}", 'Test basic utf8 @!');
+}
+{
+ #50256
+ # This test is for the bit pattern "\x61\x62", which is ASCII "ab"
+ my ($v) = split //, unpack ('(B)*', native_to_uni('ab'));
+ is($v, 0); # Doesn't SEGV :-)
+}
+{
+ #73814
+ my $x = runperl( prog => 'print split( /,/, unpack(q(%2H*), q(hello world))), qq(\n)' );
+ is($x, "0\n", "split /a/, unpack('%2H*'...) didn't crash");
+
+ my $y = runperl( prog => 'print split( /,/, unpack(q(%32u*), q(#,3,Q)), qq(\n)), qq(\n)' );
+ is($y, "0\n", "split /a/, unpack('%32u*'...) didn't crash");
+}
+
+#90160
+is(eval { () = unpack "C0 U*", ""; "ok" }, "ok",
+ 'medial U* on empty string');
+
+package o {
+ use overload
+ '""' => sub { ++$o::str; "42" },
+ '0+' => sub { ++$o::num; 42 };
+}
+is pack("c", bless [], "o"), chr(42), 'overloading called';
+is $o::str, undef, 'pack "c" does not call string overloading';
+is $o::num, 1, 'pack "c" does call num overloading';
+
+#[perl #123874]: argument underflow leads to corrupt length
+eval q{ pack "pi/x" };
+ok(1, "argument underflow did not crash");
+
+{
+ # [perl #126325] pack [hH] with a unicode string
+ # the hex encoders would read past the end of the string, using
+ # invalid source bytes
+ my $twenty_nuls = "\0" x 20;
+ # This is the case that failed
+ is(pack("WH40", 0x100, ""), "\x{100}$twenty_nuls",
+ "check pack H zero fills (utf8 target)");
+ my $up_nul = "\0";
+
+ utf8::upgrade($up_nul);
+ # check the other combinations too
+ is(pack("WH40", 0x100, $up_nul), "\x{100}$twenty_nuls",
+ "check pack H zero fills (utf8 target/source)");
+ is(pack("H40", ""), $twenty_nuls,
+ "check pack H zero fills (utf8 none)");
+ is(pack("H40", $up_nul), $twenty_nuls,
+ "check pack H zero fills (utf8 source)");
+}
+
+SKIP:
+{
+ # [perl #129149] the code below would write one past the end of the output
+ # buffer, only detected by ASAN, not by valgrind
+ $Config{ivsize} >= 8
+ or skip "[perl #129149] need 64-bit for this test", 1;
+ fresh_perl_is(<<'EOS', "ok\n", { stderr => 1 }, "pack W overflow");
+print pack("ucW", "0000", 0, 140737488355327) eq "\$,#`P,```\n\0\x{7fffffffffff}"
+ ? "ok\n" : "not ok\n";
+EOS
}
https://perl5.git.perl.org / perl5.git / blobdiff