1#!./perl -w 2 3BEGIN { 4 chdir 't' if -d 't'; 5 @INC = '../lib'; 6 require './test.pl'; 7} 8 9# This is truth in an if statement, and could be a skip message 10my $no_endianness = $] > 5.009 ? '' : 11 "Endianness pack modifiers not available on this perl"; 12my $no_signedness = $] > 5.009 ? '' : 13 "Signed/unsigned pack modifiers not available on this perl"; 14 15plan tests => 14697; 16 17use strict; 18use warnings qw(FATAL all); 19use Config; 20 21my $Is_EBCDIC = (defined $Config{ebcdic} && $Config{ebcdic} eq 'define'); 22my $Perl = which_perl(); 23my @valid_errors = (qr/^Invalid type '\w'/); 24 25my $ByteOrder = 'unknown'; 26my $maybe_not_avail = '(?:hto[bl]e|[bl]etoh)'; 27if ($no_endianness) { 28 push @valid_errors, qr/^Invalid type '[<>]'/; 29} elsif ($Config{byteorder} =~ /^1234(?:5678)?$/) { 30 $ByteOrder = 'little'; 31 $maybe_not_avail = '(?:htobe|betoh)'; 32} 33elsif ($Config{byteorder} =~ /^(?:8765)?4321$/) { 34 $ByteOrder = 'big'; 35 $maybe_not_avail = '(?:htole|letoh)'; 36} 37else { 38 push @valid_errors, qr/^Can't (?:un)?pack (?:big|little)-endian .*? on this platform/; 39} 40 41if ($no_signedness) { 42 push @valid_errors, qr/^'!' allowed only after types sSiIlLxX in (?:un)?pack/; 43} 44 45for my $size ( 16, 32, 64 ) { 46 if (defined $Config{"u${size}size"} and ($Config{"u${size}size"}||0) != ($size >> 3)) { 47 push @valid_errors, qr/^Perl_my_$maybe_not_avail$size\(\) not available/; 48 } 49} 50 51my $IsTwosComplement = pack('i', -1) eq "\xFF" x $Config{intsize}; 52print "# \$IsTwosComplement = $IsTwosComplement\n"; 53 54sub is_valid_error 55{ 56 my $err = shift; 57 58 for my $e (@valid_errors) { 59 $err =~ $e and return 1; 60 } 61 62 return 0; 63} 64 65sub encode_list { 66 my @result = map {_qq($_)} @_; 67 if (@result == 1) { 68 return @result; 69 } 70 return '(' . join (', ', @result) . ')'; 71} 72 73 74sub list_eq ($$) { 75 my ($l, $r) = @_; 76 return 0 unless @$l == @$r; 77 for my $i (0..$#$l) { 78 if (defined $l->[$i]) { 79 return 0 unless defined ($r->[$i]) && $l->[$i] eq $r->[$i]; 80 } else { 81 return 0 if defined $r->[$i] 82 } 83 } 84 return 1; 85} 86 87############################################################################## 88# 89# Here starteth the tests 90# 91 92{ 93 my $format = "c2 x5 C C x s d i l a6"; 94 # Need the expression in here to force ary[5] to be numeric. This avoids 95 # test2 failing because ary2 goes str->numeric->str and ary doesn't. 96 my @ary = (1,-100,127,128,32767,987.654321098 / 100.0,12345,123456, 97 "abcdef"); 98 my $foo = pack($format,@ary); 99 my @ary2 = unpack($format,$foo); 100 101 is($#ary, $#ary2); 102 103 my $out1=join(':',@ary); 104 my $out2=join(':',@ary2); 105 # Using long double NVs may introduce greater accuracy than wanted. 106 $out1 =~ s/:9\.87654321097999\d*:/:9.87654321098:/; 107 $out2 =~ s/:9\.87654321097999\d*:/:9.87654321098:/; 108 is($out1, $out2); 109 110 like($foo, qr/def/); 111} 112# How about counting bits? 113 114{ 115 my $x; 116 is( ($x = unpack("%32B*", "\001\002\004\010\020\040\100\200\377")), 16 ); 117 118 is( ($x = unpack("%32b69", "\001\002\004\010\020\040\100\200\017")), 12 ); 119 120 is( ($x = unpack("%32B69", "\001\002\004\010\020\040\100\200\017")), 9 ); 121} 122 123{ 124 my $sum = 129; # ASCII 125 $sum = 103 if $Is_EBCDIC; 126 127 my $x; 128 is( ($x = unpack("%32B*", "Now is the time for all good blurfl")), $sum ); 129 130 my $foo; 131 open(BIN, $Perl) || die "Can't open $Perl: $!\n"; 132 binmode BIN; 133 sysread BIN, $foo, 8192; 134 close BIN; 135 136 $sum = unpack("%32b*", $foo); 137 my $longway = unpack("b*", $foo); 138 is( $sum, $longway =~ tr/1/1/ ); 139} 140 141{ 142 my $x; 143 is( ($x = unpack("I",pack("I", 0xFFFFFFFF))), 0xFFFFFFFF ); 144} 145 146{ 147 # check 'w' 148 my @x = (5,130,256,560,32000,3097152,268435455,1073741844, 2**33, 149 '4503599627365785','23728385234614992549757750638446'); 150 my $x = pack('w*', @x); 151 my $y = pack 'H*', '0581028200843081fa0081bd8440ffffff7f8480808014A0808'. 152 '0800087ffffffffffdb19caefe8e1eeeea0c2e1e3e8ede1ee6e'; 153 154 is($x, $y); 155 156 my @y = unpack('w*', $y); 157 my $a; 158 while ($a = pop @x) { 159 my $b = pop @y; 160 is($a, $b); 161 } 162 163 @y = unpack('w2', $x); 164 165 is(scalar(@y), 2); 166 is($y[1], 130); 167 $x = pack('w*', 5000000000); $y = ''; 168 eval { 169 use Math::BigInt; 170 $y = pack('w*', Math::BigInt::->new(5000000000)); 171 }; 172 is($x, $y); 173 174 $x = pack 'w', ~0; 175 $y = pack 'w', (~0).''; 176 is($x, $y); 177 is(unpack ('w',$x), ~0); 178 is(unpack ('w',$y), ~0); 179 180 $x = pack 'w', ~0 - 1; 181 $y = pack 'w', (~0) - 2; 182 183 if (~0 - 1 == (~0) - 2) { 184 is($x, $y, "NV arithmetic"); 185 } else { 186 isnt($x, $y, "IV/NV arithmetic"); 187 } 188 cmp_ok(unpack ('w',$x), '==', ~0 - 1); 189 cmp_ok(unpack ('w',$y), '==', ~0 - 2); 190 191 # These should spot that pack 'w' is using NV, not double, on platforms 192 # where IVs are smaller than doubles, and harmlessly pass elsewhere. 193 # (tests for change 16861) 194 my $x0 = 2**54+3; 195 my $y0 = 2**54-2; 196 197 $x = pack 'w', $x0; 198 $y = pack 'w', $y0; 199 200 if ($x0 == $y0) { 201 is($x, $y, "NV arithmetic"); 202 } else { 203 isnt($x, $y, "IV/NV arithmetic"); 204 } 205 cmp_ok(unpack ('w',$x), '==', $x0); 206 cmp_ok(unpack ('w',$y), '==', $y0); 207} 208 209 210{ 211 print "# test exceptions\n"; 212 my $x; 213 eval { $x = unpack 'w', pack 'C*', 0xff, 0xff}; 214 like($@, qr/^Unterminated compressed integer/); 215 216 eval { $x = unpack 'w', pack 'C*', 0xff, 0xff, 0xff, 0xff}; 217 like($@, qr/^Unterminated compressed integer/); 218 219 eval { $x = unpack 'w', pack 'C*', 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff}; 220 like($@, qr/^Unterminated compressed integer/); 221 222 eval { $x = pack 'w', -1 }; 223 like ($@, qr/^Cannot compress negative numbers/); 224 225 eval { $x = pack 'w', '1'x(1 + length ~0) . 'e0' }; 226 like ($@, qr/^Can only compress unsigned integers/); 227 228 # Check that the warning behaviour on the modifiers !, < and > is as we 229 # expect it for this perl. 230 my $can_endian = $no_endianness ? '' : 'sSiIlLqQjJfFdDpP'; 231 my $can_shriek = 'sSiIlL'; 232 $can_shriek .= 'nNvV' unless $no_signedness; 233 # h and H can't do either, so act as sanity checks in blead 234 foreach my $base (split '', 'hHsSiIlLqQjJfFdDpPnNvV') { 235 foreach my $mod ('', '<', '>', '!', '<!', '>!', '!<', '!>') { 236 SKIP: { 237 # Avoid void context warnings. 238 my $a = eval {pack "$base$mod"}; 239 skip "pack can't $base", 1 if $@ =~ /^Invalid type '\w'/; 240 # Which error you get when 2 would be possible seems to be emergent 241 # behaviour of pack's format parser. 242 243 my $fails_shriek = $mod =~ /!/ && index ($can_shriek, $base) == -1; 244 my $fails_endian = $mod =~ /[<>]/ && index ($can_endian, $base) == -1; 245 my $shriek_first = $mod =~ /^!/; 246 247 if ($no_endianness and ($mod eq '<!' or $mod eq '>!')) { 248 # The ! isn't seem as part of $base. Instead it's seen as a modifier 249 # on > or < 250 $fails_shriek = 1; 251 undef $fails_endian; 252 } elsif ($fails_shriek and $fails_endian) { 253 if ($shriek_first) { 254 undef $fails_endian; 255 } 256 } 257 258 if ($fails_endian) { 259 if ($no_endianness) { 260 # < and > are seen as pattern letters, not modifiers 261 like ($@, qr/^Invalid type '[<>]'/, "pack can't $base$mod"); 262 } else { 263 like ($@, qr/^'[<>]' allowed only after types/, 264 "pack can't $base$mod"); 265 } 266 } elsif ($fails_shriek) { 267 like ($@, qr/^'!' allowed only after types/, 268 "pack can't $base$mod"); 269 } else { 270 is ($@, '', "pack can $base$mod"); 271 } 272 } 273 } 274 } 275 276 SKIP: { 277 skip $no_endianness, 2*3 + 2*8 if $no_endianness; 278 for my $mod (qw( ! < > )) { 279 eval { $x = pack "a$mod", 42 }; 280 like ($@, qr/^'$mod' allowed only after types \S+ in pack/); 281 282 eval { $x = unpack "a$mod", 'x'x8 }; 283 like ($@, qr/^'$mod' allowed only after types \S+ in unpack/); 284 } 285 286 for my $mod (qw( <> >< !<> !>< <!> >!< <>! ><! )) { 287 eval { $x = pack "sI${mod}s", 42, 47, 11 }; 288 like ($@, qr/^Can't use both '<' and '>' after type 'I' in pack/); 289 290 eval { $x = unpack "sI${mod}s", 'x'x16 }; 291 like ($@, qr/^Can't use both '<' and '>' after type 'I' in unpack/); 292 } 293 } 294 295 SKIP: { 296 # Is this a stupid thing to do on VMS, VOS and other unusual platforms? 297 298 skip("-- the IEEE infinity model is unavailable in this configuration.", 1) 299 if (($^O eq 'VMS') && !defined($Config{useieee})); 300 301 skip("-- $^O has serious fp indigestion on w-packed infinities", 1) 302 if ( 303 ($^O eq 'mpeix') 304 || 305 ($^O eq 'ultrix') 306 || 307 ($^O =~ /^svr4/ && -f "/etc/issue" && -f "/etc/.relid") # NCR MP-RAS 308 ); 309 310 my $inf = eval '2**1000000'; 311 312 skip("Couldn't generate infinity - got error '$@'", 1) 313 unless defined $inf and $inf == $inf / 2 and $inf + 1 == $inf; 314 315 local our $TODO; 316 $TODO = "VOS needs a fix for posix-1022 to pass this test." 317 if ($^O eq 'vos'); 318 319 eval { $x = pack 'w', $inf }; 320 like ($@, qr/^Cannot compress integer/, "Cannot compress integer"); 321 } 322 323 SKIP: { 324 325 skip("-- the full range of an IEEE double may not be available in this configuration.", 3) 326 if (($^O eq 'VMS') && !defined($Config{useieee})); 327 328 skip("-- $^O does not like 2**1023", 3) 329 if (($^O eq 'ultrix')); 330 331 # This should be about the biggest thing possible on an IEEE double 332 my $big = eval '2**1023'; 333 334 skip("Couldn't generate 2**1023 - got error '$@'", 3) 335 unless defined $big and $big != $big / 2; 336 337 eval { $x = pack 'w', $big }; 338 is ($@, '', "Should be able to pack 'w', $big # 2**1023"); 339 340 my $y = eval {unpack 'w', $x}; 341 is ($@, '', 342 "Should be able to unpack 'w' the result of pack 'w', $big # 2**1023"); 343 344 # I'm getting about 1e-16 on FreeBSD 345 my $quotient = int (100 * ($y - $big) / $big); 346 ok($quotient < 2 && $quotient > -2, 347 "Round trip pack, unpack 'w' of $big is within 1% ($quotient%)"); 348 } 349 350} 351 352print "# test the 'p' template\n"; 353 354# literals 355is(unpack("p",pack("p","foo")), "foo"); 356SKIP: { 357 skip $no_endianness, 2 if $no_endianness; 358 is(unpack("p<",pack("p<","foo")), "foo"); 359 is(unpack("p>",pack("p>","foo")), "foo"); 360} 361# scalars 362is(unpack("p",pack("p",239)), 239); 363SKIP: { 364 skip $no_endianness, 2 if $no_endianness; 365 is(unpack("p<",pack("p<",239)), 239); 366 is(unpack("p>",pack("p>",239)), 239); 367} 368 369# temps 370sub foo { my $a = "a"; return $a . $a++ . $a++ } 371{ 372 use warnings qw(NONFATAL all);; 373 my $warning; 374 local $SIG{__WARN__} = sub { 375 $warning = $_[0]; 376 }; 377 my $junk = pack("p", &foo); 378 379 like($warning, qr/temporary val/); 380} 381 382# undef should give null pointer 383like(pack("p", undef), qr/^\0+$/); 384SKIP: { 385 skip $no_endianness, 2 if $no_endianness; 386 like(pack("p<", undef), qr/^\0+$/); 387 like(pack("p>", undef), qr/^\0+$/); 388} 389 390# Check for optimizer bug (e.g. Digital Unix GEM cc with -O4 on DU V4.0B gives 391# 4294967295 instead of -1) 392# see #ifdef __osf__ in pp.c pp_unpack 393is((unpack("i",pack("i",-1))), -1); 394 395print "# test the pack lengths of s S i I l L n N v V + modifiers\n"; 396 397my @lengths = ( 398 qw(s 2 S 2 i -4 I -4 l 4 L 4 n 2 N 4 v 2 V 4 n! 2 N! 4 v! 2 V! 4), 399 's!' => $Config{shortsize}, 'S!' => $Config{shortsize}, 400 'i!' => $Config{intsize}, 'I!' => $Config{intsize}, 401 'l!' => $Config{longsize}, 'L!' => $Config{longsize}, 402); 403 404while (my ($base, $expect) = splice @lengths, 0, 2) { 405 my @formats = ($base); 406 $base =~ /^[nv]/i or push @formats, "$base>", "$base<"; 407 for my $format (@formats) { 408 SKIP: { 409 skip $no_endianness, 1 if $no_endianness && $format =~ m/[<>]/; 410 skip $no_signedness, 1 if $no_signedness && $format =~ /[nNvV]!/; 411 my $len = length(pack($format, 0)); 412 if ($expect > 0) { 413 is($expect, $len, "format '$format'"); 414 } else { 415 $expect = -$expect; 416 ok ($len >= $expect, "format '$format'") || 417 print "# format '$format' has length $len, expected >= $expect\n"; 418 } 419 } 420 } 421} 422 423 424print "# test unpack-pack lengths\n"; 425 426my @templates = qw(c C W i I s S l L n N v V f d q Q); 427 428foreach my $base (@templates) { 429 my @tmpl = ($base); 430 $base =~ /^[cwnv]/i or push @tmpl, "$base>", "$base<"; 431 foreach my $t (@tmpl) { 432 SKIP: { 433 my @t = eval { unpack("$t*", pack("$t*", 12, 34)) }; 434 435 skip "cannot pack '$t' on this perl", 4 436 if is_valid_error($@); 437 438 is( $@, '', "Template $t works"); 439 is(scalar @t, 2); 440 441 is($t[0], 12); 442 is($t[1], 34); 443 } 444 } 445} 446 447{ 448 # uuencode/decode 449 450 # Note that first uuencoding known 'text' data and then checking the 451 # binary values of the uuencoded version would not be portable between 452 # character sets. Uuencoding is meant for encoding binary data, not 453 # text data. 454 455 my $in = pack 'C*', 0 .. 255; 456 457 # just to be anal, we do some random tr/`/ / 458 my $uu = <<'EOUU'; 459M` $"`P0%!@<("0H+# T.#Q`1$A,4%187&!D:&QP='A\@(2(C)"4F)R@I*BLL 460M+2XO,#$R,S0U-C<X.3H[/#T^/T!!0D-$149'2$E*2TQ-3D]045)35%565UA9 461M6EM<75Y?8&%B8V1E9F=H:6IK;&UN;W!Q<G-T=79W>'EZ>WQ]?G^`@8*#A(6& 462MAXB)BHN,C8Z/D)&2DY25EI>8F9J;G)V>GZ"AHJ.DI::GJ*FJJZRMKJ^PL;*S 463MM+6VM[BYNKN\O;Z_P,'"P\3%QL?(R<K+S,W.S]#1TM/4U=;7V-G:V]S=WM_@ 464?X>+CY.7FY^CIZNOL[>[O\/'R\_3U]O?X^?K[_/W^_P ` 465EOUU 466 467 $_ = $uu; 468 tr/ /`/; 469 470 is(pack('u', $in), $_); 471 472 is(unpack('u', $uu), $in); 473 474 $in = "\x1f\x8b\x08\x08\x58\xdc\xc4\x35\x02\x03\x4a\x41\x50\x55\x00\xf3\x2a\x2d\x2e\x51\x48\xcc\xcb\x2f\xc9\x48\x2d\x52\x08\x48\x2d\xca\x51\x28\x2d\x4d\xce\x4f\x49\x2d\xe2\x02\x00\x64\x66\x60\x5c\x1a\x00\x00\x00"; 475 $uu = <<'EOUU'; 476M'XL("%C<Q#4"`TI!4%4`\RHM+E%(S,LOR4@M4@A(+<I1*"U-SD])+>("`&1F 477&8%P:```` 478EOUU 479 480 is(unpack('u', $uu), $in); 481 482# This is identical to the above except that backquotes have been 483# changed to spaces 484 485 $uu = <<'EOUU'; 486M'XL("%C<Q#4" TI!4%4 \RHM+E%(S,LOR4@M4@A(+<I1*"U-SD])+>(" &1F 487&8%P: 488EOUU 489 490 # ' # Grr 491 is(unpack('u', $uu), $in); 492 493} 494 495# test the ascii template types (A, a, Z) 496 497foreach ( 498['p', 'A*', "foo\0bar\0 ", "foo\0bar\0 "], 499['p', 'A11', "foo\0bar\0 ", "foo\0bar\0 "], 500['u', 'A*', "foo\0bar \0", "foo\0bar"], 501['u', 'A8', "foo\0bar \0", "foo\0bar"], 502['p', 'a*', "foo\0bar\0 ", "foo\0bar\0 "], 503['p', 'a11', "foo\0bar\0 ", "foo\0bar\0 \0\0"], 504['u', 'a*', "foo\0bar \0", "foo\0bar \0"], 505['u', 'a8', "foo\0bar \0", "foo\0bar "], 506['p', 'Z*', "foo\0bar\0 ", "foo\0bar\0 \0"], 507['p', 'Z11', "foo\0bar\0 ", "foo\0bar\0 \0\0"], 508['p', 'Z3', "foo", "fo\0"], 509['u', 'Z*', "foo\0bar \0", "foo"], 510['u', 'Z8', "foo\0bar \0", "foo"], 511) 512{ 513 my ($what, $template, $in, $out) = @$_; 514 my $got = $what eq 'u' ? (unpack $template, $in) : (pack $template, $in); 515 unless (is($got, $out)) { 516 my $un = $what eq 'u' ? 'un' : ''; 517 print "# ${un}pack ('$template', "._qq($in).') gave '._qq($out). 518 ' not '._qq($got)."\n"; 519 } 520} 521 522print "# packing native shorts/ints/longs\n"; 523 524is(length(pack("s!", 0)), $Config{shortsize}); 525is(length(pack("i!", 0)), $Config{intsize}); 526is(length(pack("l!", 0)), $Config{longsize}); 527ok(length(pack("s!", 0)) <= length(pack("i!", 0))); 528ok(length(pack("i!", 0)) <= length(pack("l!", 0))); 529is(length(pack("i!", 0)), length(pack("i", 0))); 530 531sub numbers { 532 my $base = shift; 533 my @formats = ($base); 534 $base =~ /^[silqjfdp]/i and push @formats, "$base>", "$base<"; 535 for my $format (@formats) { 536 numbers_with_total ($format, undef, @_); 537 } 538} 539 540sub numbers_with_total { 541 my $format = shift; 542 my $total = shift; 543 if (!defined $total) { 544 foreach (@_) { 545 $total += $_; 546 } 547 } 548 print "# numbers test for $format\n"; 549 foreach (@_) { 550 SKIP: { 551 my $out = eval {unpack($format, pack($format, $_))}; 552 skip "cannot pack '$format' on this perl", 2 553 if is_valid_error($@); 554 555 is($@, '', "no error"); 556 is($out, $_, "unpack pack $format $_"); 557 } 558 } 559 560 my $skip_if_longer_than = ~0; # "Infinity" 561 if (~0 - 1 == ~0) { 562 # If we're running with -DNO_PERLPRESERVE_IVUV and NVs don't preserve all 563 # UVs (in which case ~0 is NV, ~0-1 will be the same NV) then we can't 564 # correctly in perl calculate UV totals for long checksums, as pp_unpack 565 # is using UV maths, and we've only got NVs. 566 $skip_if_longer_than = $Config{nv_preserves_uv_bits}; 567 } 568 569 foreach ('', 1, 2, 3, 15, 16, 17, 31, 32, 33, 53, 54, 63, 64, 65) { 570 SKIP: { 571 my $sum = eval {unpack "%$_$format*", pack "$format*", @_}; 572 skip "cannot pack '$format' on this perl", 3 573 if is_valid_error($@); 574 575 is($@, '', "no error"); 576 ok(defined $sum, "sum bits $_, format $format defined"); 577 578 my $len = $_; # Copy, so that we can reassign '' 579 $len = 16 unless length $len; 580 581 SKIP: { 582 skip "cannot test checksums over $skip_if_longer_than bits", 1 583 if $len > $skip_if_longer_than; 584 585 # Our problem with testing this portably is that the checksum code in 586 # pp_unpack is able to cast signed to unsigned, and do modulo 2**n 587 # arithmetic in unsigned ints, which perl has no operators to do. 588 # (use integer; does signed ints, which won't wrap on UTS, which is just 589 # fine with ANSI, but not with most people's assumptions. 590 # This is why we need to supply the totals for 'Q' as there's no way in 591 # perl to calculate them, short of unpack '%0Q' (is that documented?) 592 # ** returns NVs; make sure it's IV. 593 my $max = 1 + 2 * (int (2 ** ($len-1))-1); # The max possible checksum 594 my $max_p1 = $max + 1; 595 my ($max_is_integer, $max_p1_is_integer); 596 $max_p1_is_integer = 1 unless $max_p1 + 1 == $max_p1; 597 $max_is_integer = 1 if $max - 1 < ~0; 598 599 my $calc_sum; 600 if (ref $total) { 601 $calc_sum = &$total($len); 602 } else { 603 $calc_sum = $total; 604 # Shift into range by some multiple of the total 605 my $mult = $max_p1 ? int ($total / $max_p1) : undef; 606 # Need this to make sure that -1 + (~0+1) is ~0 (ie still integer) 607 $calc_sum = $total - $mult; 608 $calc_sum -= $mult * $max; 609 if ($calc_sum < 0) { 610 $calc_sum += 1; 611 $calc_sum += $max; 612 } 613 } 614 if ($calc_sum == $calc_sum - 1 && $calc_sum == $max_p1) { 615 # we're into floating point (either by getting out of the range of 616 # UV arithmetic, or because we're doing a floating point checksum) 617 # and our calculation of the checksum has become rounded up to 618 # max_checksum + 1 619 $calc_sum = 0; 620 } 621 622 if ($calc_sum == $sum) { # HAS to be ==, not eq (so no is()). 623 pass ("unpack '%$_$format' gave $sum"); 624 } else { 625 my $delta = 1.000001; 626 if ($format =~ tr /dDfF// 627 && ($calc_sum <= $sum * $delta && $calc_sum >= $sum / $delta)) { 628 pass ("unpack '%$_$format' gave $sum, expected $calc_sum"); 629 } else { 630 my $text = ref $total ? &$total($len) : $total; 631 fail; 632 print "# For list (" . join (", ", @_) . ") (total $text)" 633 . " packed with $format unpack '%$_$format' gave $sum," 634 . " expected $calc_sum\n"; 635 } 636 } 637 } 638 } 639 } 640} 641 642numbers ('c', -128, -1, 0, 1, 127); 643numbers ('C', 0, 1, 127, 128, 255); 644numbers ('W', 0, 1, 127, 128, 255, 256, 0x7ff, 0x800, 0xfffd); 645numbers ('s', -32768, -1, 0, 1, 32767); 646numbers ('S', 0, 1, 32767, 32768, 65535); 647numbers ('i', -2147483648, -1, 0, 1, 2147483647); 648numbers ('I', 0, 1, 2147483647, 2147483648, 4294967295); 649numbers ('l', -2147483648, -1, 0, 1, 2147483647); 650numbers ('L', 0, 1, 2147483647, 2147483648, 4294967295); 651numbers ('s!', -32768, -1, 0, 1, 32767); 652numbers ('S!', 0, 1, 32767, 32768, 65535); 653numbers ('i!', -2147483648, -1, 0, 1, 2147483647); 654numbers ('I!', 0, 1, 2147483647, 2147483648, 4294967295); 655numbers ('l!', -2147483648, -1, 0, 1, 2147483647); 656numbers ('L!', 0, 1, 2147483647, 2147483648, 4294967295); 657numbers ('n', 0, 1, 32767, 32768, 65535); 658numbers ('v', 0, 1, 32767, 32768, 65535); 659numbers ('N', 0, 1, 2147483647, 2147483648, 4294967295); 660numbers ('V', 0, 1, 2147483647, 2147483648, 4294967295); 661numbers ('n!', -32768, -1, 0, 1, 32767); 662numbers ('v!', -32768, -1, 0, 1, 32767); 663numbers ('N!', -2147483648, -1, 0, 1, 2147483647); 664numbers ('V!', -2147483648, -1, 0, 1, 2147483647); 665# All these should have exact binary representations: 666numbers ('f', -1, 0, 0.5, 42, 2**34); 667numbers ('d', -(2**34), -1, 0, 1, 2**34); 668## These don't, but 'd' is NV. XXX wrong, it's double 669#numbers ('d', -1, 0, 1, 1-exp(-1), -exp(1)); 670 671numbers_with_total ('q', -1, 672 -9223372036854775808, -1, 0, 1,9223372036854775807); 673# This total is icky, but the true total is 2**65-1, and need a way to generate 674# the epxected checksum on any system including those where NVs can preserve 675# 65 bits. (long double is 128 bits on sparc, so they certainly can) 676# or where rounding is down not up on binary conversion (crays) 677numbers_with_total ('Q', sub { 678 my $len = shift; 679 $len = 65 if $len > 65; # unmasked total is 2**65-1 here 680 my $total = 1 + 2 * (int (2**($len - 1)) - 1); 681 return 0 if $total == $total - 1; # Overflowed integers 682 return $total; # NVs still accurate to nearest integer 683 }, 684 0, 1,9223372036854775807, 9223372036854775808, 685 18446744073709551615); 686 687print "# pack nvNV byteorders\n"; 688 689is(pack("n", 0xdead), "\xde\xad"); 690is(pack("v", 0xdead), "\xad\xde"); 691is(pack("N", 0xdeadbeef), "\xde\xad\xbe\xef"); 692is(pack("V", 0xdeadbeef), "\xef\xbe\xad\xde"); 693 694SKIP: { 695 skip $no_signedness, 4 if $no_signedness; 696 is(pack("n!", 0xdead), "\xde\xad"); 697 is(pack("v!", 0xdead), "\xad\xde"); 698 is(pack("N!", 0xdeadbeef), "\xde\xad\xbe\xef"); 699 is(pack("V!", 0xdeadbeef), "\xef\xbe\xad\xde"); 700} 701 702print "# test big-/little-endian conversion\n"; 703 704sub byteorder 705{ 706 my $format = shift; 707 print "# byteorder test for $format\n"; 708 for my $value (@_) { 709 SKIP: { 710 my($nat,$be,$le) = eval { map { pack $format.$_, $value } '', '>', '<' }; 711 skip "cannot pack '$format' on this perl", 5 712 if is_valid_error($@); 713 714 { 715 use warnings qw(NONFATAL utf8); 716 print "# [$value][$nat][$be][$le][$@]\n"; 717 } 718 719 SKIP: { 720 skip "cannot compare native byteorder with big-/little-endian", 1 721 if $ByteOrder eq 'unknown'; 722 723 is($nat, $ByteOrder eq 'big' ? $be : $le); 724 } 725 is($be, reverse($le)); 726 my @x = eval { unpack "$format$format>$format<", $nat.$be.$le }; 727 728 print "# [$value][", join('][', @x), "][$@]\n"; 729 730 is($@, ''); 731 is($x[0], $x[1]); 732 is($x[0], $x[2]); 733 } 734 } 735} 736 737byteorder('s', -32768, -1, 0, 1, 32767); 738byteorder('S', 0, 1, 32767, 32768, 65535); 739byteorder('i', -2147483648, -1, 0, 1, 2147483647); 740byteorder('I', 0, 1, 2147483647, 2147483648, 4294967295); 741byteorder('l', -2147483648, -1, 0, 1, 2147483647); 742byteorder('L', 0, 1, 2147483647, 2147483648, 4294967295); 743byteorder('j', -2147483648, -1, 0, 1, 2147483647); 744byteorder('J', 0, 1, 2147483647, 2147483648, 4294967295); 745byteorder('s!', -32768, -1, 0, 1, 32767); 746byteorder('S!', 0, 1, 32767, 32768, 65535); 747byteorder('i!', -2147483648, -1, 0, 1, 2147483647); 748byteorder('I!', 0, 1, 2147483647, 2147483648, 4294967295); 749byteorder('l!', -2147483648, -1, 0, 1, 2147483647); 750byteorder('L!', 0, 1, 2147483647, 2147483648, 4294967295); 751byteorder('q', -9223372036854775808, -1, 0, 1, 9223372036854775807); 752byteorder('Q', 0, 1, 9223372036854775807, 9223372036854775808, 18446744073709551615); 753byteorder('f', -1, 0, 0.5, 42, 2**34); 754byteorder('F', -1, 0, 0.5, 42, 2**34); 755byteorder('d', -(2**34), -1, 0, 1, 2**34); 756byteorder('D', -(2**34), -1, 0, 1, 2**34); 757 758print "# test negative numbers\n"; 759 760SKIP: { 761 skip "platform is not using two's complement for negative integers", 120 762 unless $IsTwosComplement; 763 764 for my $format (qw(s i l j s! i! l! q)) { 765 SKIP: { 766 my($nat,$be,$le) = eval { map { pack $format.$_, -1 } '', '>', '<' }; 767 skip "cannot pack '$format' on this perl", 15 768 if is_valid_error($@); 769 770 my $len = length $nat; 771 is($_, "\xFF"x$len) for $nat, $be, $le; 772 773 my(@val,@ref); 774 if ($len >= 8) { 775 @val = (-2, -81985529216486896, -9223372036854775808); 776 @ref = ("\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFE", 777 "\xFE\xDC\xBA\x98\x76\x54\x32\x10", 778 "\x80\x00\x00\x00\x00\x00\x00\x00"); 779 } 780 elsif ($len >= 4) { 781 @val = (-2, -19088744, -2147483648); 782 @ref = ("\xFF\xFF\xFF\xFE", 783 "\xFE\xDC\xBA\x98", 784 "\x80\x00\x00\x00"); 785 } 786 else { 787 @val = (-2, -292, -32768); 788 @ref = ("\xFF\xFE", 789 "\xFE\xDC", 790 "\x80\x00"); 791 } 792 for my $x (@ref) { 793 if ($len > length $x) { 794 $x = $x . "\xFF" x ($len - length $x); 795 } 796 } 797 798 for my $i (0 .. $#val) { 799 my($nat,$be,$le) = eval { map { pack $format.$_, $val[$i] } '', '>', '<' }; 800 is($@, ''); 801 802 SKIP: { 803 skip "cannot compare native byteorder with big-/little-endian", 1 804 if $ByteOrder eq 'unknown'; 805 806 is($nat, $ByteOrder eq 'big' ? $be : $le); 807 } 808 809 is($be, $ref[$i]); 810 is($be, reverse($le)); 811 } 812 } 813 } 814} 815 816{ 817 # / 818 819 my ($x, $y, $z); 820 eval { ($x) = unpack '/a*','hello' }; 821 like($@, qr!'/' must follow a numeric type!); 822 undef $x; 823 eval { $x = unpack '/a*','hello' }; 824 like($@, qr!'/' must follow a numeric type!); 825 826 undef $x; 827 eval { ($z,$x,$y) = unpack 'a3/A C/a* C/Z', "003ok \003yes\004z\000abc" }; 828 is($@, ''); 829 is($z, 'ok'); 830 is($x, 'yes'); 831 is($y, 'z'); 832 undef $z; 833 eval { $z = unpack 'a3/A C/a* C/Z', "003ok \003yes\004z\000abc" }; 834 is($@, ''); 835 is($z, 'ok'); 836 837 838 undef $x; 839 eval { ($x) = pack '/a*','hello' }; 840 like($@, qr!Invalid type '/'!); 841 undef $x; 842 eval { $x = pack '/a*','hello' }; 843 like($@, qr!Invalid type '/'!); 844 845 $z = pack 'n/a* N/Z* w/A*','string','hi there ','etc'; 846 my $expect = "\000\006string\0\0\0\012hi there \000\003etc"; 847 is($z, $expect); 848 849 undef $x; 850 $expect = 'hello world'; 851 eval { ($x) = unpack ("w/a", chr (11) . "hello world!")}; 852 is($x, $expect); 853 is($@, ''); 854 855 undef $x; 856 # Doing this in scalar context used to fail. 857 eval { $x = unpack ("w/a", chr (11) . "hello world!")}; 858 is($@, ''); 859 is($x, $expect); 860 861 foreach ( 862 ['a/a*/a*', '212ab345678901234567','ab3456789012'], 863 ['a/a*/a*', '3012ab345678901234567', 'ab3456789012'], 864 ['a/a*/b*', '212ab', $Is_EBCDIC ? '100000010100' : '100001100100'], 865 ) 866 { 867 my ($pat, $in, $expect) = @$_; 868 undef $x; 869 eval { ($x) = unpack $pat, $in }; 870 is($@, ''); 871 is($x, $expect) || 872 printf "# list unpack ('$pat', '$in') gave %s, expected '$expect'\n", 873 encode_list ($x); 874 875 undef $x; 876 eval { $x = unpack $pat, $in }; 877 is($@, ''); 878 is($x, $expect) || 879 printf "# scalar unpack ('$pat', '$in') gave %s, expected '$expect'\n", 880 encode_list ($x); 881 } 882 883 # / with # 884 885 my $pattern = <<'EOU'; 886 a3/A # Count in ASCII 887 C/a* # Count in a C char 888 C/Z # Count in a C char but skip after \0 889EOU 890 891 $x = $y = $z =undef; 892 eval { ($z,$x,$y) = unpack $pattern, "003ok \003yes\004z\000abc" }; 893 is($@, ''); 894 is($z, 'ok'); 895 is($x, 'yes'); 896 is($y, 'z'); 897 undef $x; 898 eval { $z = unpack $pattern, "003ok \003yes\004z\000abc" }; 899 is($@, ''); 900 is($z, 'ok'); 901 902 $pattern = <<'EOP'; 903 n/a* # Count as network short 904 w/A* # Count a BER integer 905EOP 906 $expect = "\000\006string\003etc"; 907 $z = pack $pattern,'string','etc'; 908 is($z, $expect); 909} 910 911 912SKIP: { 913 skip("(EBCDIC and) version strings are bad idea", 2) if $Is_EBCDIC; 914 915 is("1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000)); 916 is("1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000)); 917} 918isnt(v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000)); 919 920my $rslt = $Is_EBCDIC ? "156 67" : "199 162"; 921is(join(" ", unpack("U0 C*", chr(0x1e2))), $rslt); 922 923# does pack U create Unicode? 924is(ord(pack('U', 300)), 300); 925 926# does unpack U deref Unicode? 927is((unpack('U', chr(300)))[0], 300); 928 929# is unpack U the reverse of pack U for Unicode string? 930is("@{[unpack('U*', pack('U*', 100, 200, 300))]}", "100 200 300"); 931 932# is unpack U the reverse of pack U for byte string? 933is("@{[unpack('U*', pack('U*', 100, 200))]}", "100 200"); 934 935 936SKIP: { 937 skip "Not for EBCDIC", 4 if $Is_EBCDIC; 938 939 # does pack U0C create Unicode? 940 is("@{[pack('U0C*', 100, 195, 136)]}", v100.v200); 941 942 # does pack C0U create characters? 943 is("@{[pack('C0U*', 100, 200)]}", pack("C*", 100, 195, 136)); 944 945 # does unpack U0U on byte data warn? 946 { 947 use warnings qw(NONFATAL all);; 948 949 my $bad = pack("U0C", 255); 950 local $SIG{__WARN__} = sub { $@ = "@_" }; 951 my @null = unpack('U0U', $bad); 952 like($@, qr/^Malformed UTF-8 character /); 953 } 954} 955 956{ 957 my $p = pack 'i*', -2147483648, ~0, 0, 1, 2147483647; 958 my (@a); 959 # bug - % had to be at the start of the pattern, no leading whitespace or 960 # comments. %i! didn't work at all. 961 foreach my $pat ('%32i*', ' %32i*', "# Muhahahaha\n%32i*", '%32i* ', 962 '%32i!*', ' %32i!*', "\n#\n#\n\r \t\f%32i!*", '%32i!*#') { 963 @a = unpack $pat, $p; 964 is($a[0], 0xFFFFFFFF) || print "# $pat\n"; 965 @a = scalar unpack $pat, $p; 966 is($a[0], 0xFFFFFFFF) || print "# $pat\n"; 967 } 968 969 970 $p = pack 'I*', 42, 12; 971 # Multiline patterns in scalar context failed. 972 foreach my $pat ('I', <<EOPOEMSNIPPET, 'I#I', 'I # I', 'I # !!!') { 973# On the Ning Nang Nong 974# Where the Cows go Bong! 975# And the Monkeys all say Boo! 976I 977EOPOEMSNIPPET 978 @a = unpack $pat, $p; 979 is(scalar @a, 1); 980 is($a[0], 42); 981 @a = scalar unpack $pat, $p; 982 is(scalar @a, 1); 983 is($a[0], 42); 984 } 985 986 # shorts (of all flavours) didn't calculate checksums > 32 bits with floating 987 # point, so a pathologically long pattern would wrap at 32 bits. 988 my $pat = "\xff\xff"x65538; # Start with it long, to save any copying. 989 foreach (4,3,2,1,0) { 990 my $len = 65534 + $_; 991 is(unpack ("%33n$len", $pat), 65535 * $len); 992 } 993} 994 995 996# pack x X @ 997foreach ( 998 ['x', "N", "\0"], 999 ['x4', "N", "\0"x4], 1000 ['xX', "N", ""], 1001 ['xXa*', "Nick", "Nick"], 1002 ['a5Xa5', "cameL", "llama", "camellama"], 1003 ['@4', 'N', "\0"x4], 1004 ['a*@8a*', 'Camel', 'Dromedary', "Camel\0\0\0Dromedary"], 1005 ['a*@4a', 'Perl rules', '!', 'Perl!'], 1006) 1007{ 1008 my ($template, @in) = @$_; 1009 my $out = pop @in; 1010 my $got = eval {pack $template, @in}; 1011 is($@, ''); 1012 is($out, $got) || 1013 printf "# pack ('$template', %s) gave %s expected %s\n", 1014 encode_list (@in), encode_list ($got), encode_list ($out); 1015} 1016 1017# unpack x X @ 1018foreach ( 1019 ['x', "N"], 1020 ['xX', "N"], 1021 ['xXa*', "Nick", "Nick"], 1022 ['a5Xa5', "camellama", "camel", "llama"], 1023 ['@3', "ice"], 1024 ['@2a2', "water", "te"], 1025 ['a*@1a3', "steam", "steam", "tea"], 1026) 1027{ 1028 my ($template, $in, @out) = @$_; 1029 my @got = eval {unpack $template, $in}; 1030 is($@, ''); 1031 ok (list_eq (\@got, \@out)) || 1032 printf "# list unpack ('$template', %s) gave %s expected %s\n", 1033 _qq($in), encode_list (@got), encode_list (@out); 1034 1035 my $got = eval {unpack $template, $in}; 1036 is($@, ''); 1037 @out ? is( $got, $out[0] ) # 1 or more items; should get first 1038 : ok( !defined $got ) # 0 items; should get undef 1039 or printf "# scalar unpack ('$template', %s) gave %s expected %s\n", 1040 _qq($in), encode_list ($got), encode_list ($out[0]); 1041} 1042 1043{ 1044 my $t = 'Z*Z*'; 1045 my ($u, $v) = qw(foo xyzzy); 1046 my $p = pack($t, $u, $v); 1047 my @u = unpack($t, $p); 1048 is(scalar @u, 2); 1049 is($u[0], $u); 1050 is($u[1], $v); 1051} 1052 1053{ 1054 is((unpack("w/a*", "\x02abc"))[0], "ab"); 1055 1056 # "w/a*" should be seen as one unit 1057 1058 is(scalar unpack("w/a*", "\x02abc"), "ab"); 1059} 1060 1061SKIP: { 1062 print "# group modifiers\n"; 1063 1064 skip $no_endianness, 3 * 2 + 3 * 2 + 1 if $no_endianness; 1065 1066 for my $t (qw{ (s<)< (sl>s)> (s(l(sl)<l)s)< }) { 1067 print "# testing pattern '$t'\n"; 1068 eval { ($_) = unpack($t, 'x'x18); }; 1069 is($@, ''); 1070 eval { $_ = pack($t, (0)x6); }; 1071 is($@, ''); 1072 } 1073 1074 for my $t (qw{ (s<)> (sl>s)< (s(l(sl)<l)s)> }) { 1075 print "# testing pattern '$t'\n"; 1076 eval { ($_) = unpack($t, 'x'x18); }; 1077 like($@, qr/Can't use '[<>]' in a group with different byte-order in unpack/); 1078 eval { $_ = pack($t, (0)x6); }; 1079 like($@, qr/Can't use '[<>]' in a group with different byte-order in pack/); 1080 } 1081 1082 is(pack('L<L>', (0x12345678)x2), 1083 pack('(((L1)1)<)(((L)1)1)>1', (0x12345678)x2)); 1084} 1085 1086{ 1087 sub compress_template { 1088 my $t = shift; 1089 for my $mod (qw( < > )) { 1090 $t =~ s/((?:(?:[SILQJFDP]!?$mod|[^SILQJFDP\W]!?)(?:\d+|\*|\[(?:[^]]+)\])?\/?){2,})/ 1091 my $x = $1; $x =~ s!$mod!!g ? "($x)$mod" : $x /ieg; 1092 } 1093 return $t; 1094 } 1095 1096 my %templates = ( 1097 's<' => [-42], 1098 's<c2x![S]S<' => [-42, -11, 12, 4711], 1099 '(i<j<[s]l<)3' => [-11, -22, -33, 1000000, 1100, 2201, 3302, 1100 -1000000, 32767, -32768, 1, -123456789 ], 1101 '(I!<4(J<2L<)3)5' => [1 .. 65], 1102 'q<Q<' => [-50000000005, 60000000006], 1103 'f<F<d<' => [3.14159, 111.11, 2222.22], 1104 'D<cCD<' => [1e42, -128, 255, 1e-42], 1105 'n/a*' => ['/usr/bin/perl'], 1106 'C/a*S</A*L</Z*I</a*' => [qw(Just another Perl hacker)], 1107 ); 1108 1109 for my $tle (sort keys %templates) { 1110 my @d = @{$templates{$tle}}; 1111 my $tbe = $tle; 1112 $tbe =~ y/</>/; 1113 for my $t ($tbe, $tle) { 1114 my $c = compress_template($t); 1115 print "# '$t' -> '$c'\n"; 1116 SKIP: { 1117 my $p1 = eval { pack $t, @d }; 1118 skip "cannot pack '$t' on this perl", 5 if is_valid_error($@); 1119 my $p2 = eval { pack $c, @d }; 1120 is($@, ''); 1121 is($p1, $p2); 1122 s!(/[aAZ])\*!$1!g for $t, $c; 1123 my @u1 = eval { unpack $t, $p1 }; 1124 is($@, ''); 1125 my @u2 = eval { unpack $c, $p2 }; 1126 is($@, ''); 1127 is(join('!', @u1), join('!', @u2)); 1128 } 1129 } 1130 } 1131} 1132 1133{ 1134 # from Wolfgang Laun: fix in change #13163 1135 1136 my $s = 'ABC' x 10; 1137 my $t = '*'; 1138 my $x = ord($t); 1139 my $buf = pack( 'Z*/A* C', $s, $x ); 1140 my $y; 1141 1142 my $h = $buf; 1143 $h =~ s/[^[:print:]]/./g; 1144 ( $s, $y ) = unpack( "Z*/A* C", $buf ); 1145 is($h, "30.ABCABCABCABCABCABCABCABCABCABC$t"); 1146 is(length $buf, 34); 1147 is($s, "ABCABCABCABCABCABCABCABCABCABC"); 1148 is($y, $x); 1149} 1150 1151{ 1152 # from Wolfgang Laun: fix in change #13288 1153 1154 eval { my $t=unpack("P*", "abc") }; 1155 like($@, qr/'P' must have an explicit size/); 1156} 1157 1158{ # Grouping constructs 1159 my (@a, @b); 1160 @a = unpack '(SL)', pack 'SLSLSL', 67..90; 1161 is("@a", "67 68"); 1162 @a = unpack '(SL)3', pack 'SLSLSL', 67..90; 1163 @b = (67..72); 1164 is("@a", "@b"); 1165 @a = unpack '(SL)3', pack 'SLSLSLSL', 67..90; 1166 is("@a", "@b"); 1167 @a = unpack '(SL)[3]', pack 'SLSLSLSL', 67..90; 1168 is("@a", "@b"); 1169 @a = unpack '(SL)[2] SL', pack 'SLSLSLSL', 67..90; 1170 is("@a", "@b"); 1171 @a = unpack 'A/(SL)', pack 'ASLSLSLSL', 3, 67..90; 1172 is("@a", "@b"); 1173 @a = unpack 'A/(SL)SL', pack 'ASLSLSLSL', 2, 67..90; 1174 is("@a", "@b"); 1175 @a = unpack '(SL)*', pack 'SLSLSLSL', 67..90; 1176 @b = (67..74); 1177 is("@a", "@b"); 1178 @a = unpack '(SL)*SL', pack 'SLSLSLSL', 67..90; 1179 is("@a", "@b"); 1180 eval { @a = unpack '(*SL)', '' }; 1181 like($@, qr/\(\)-group starts with a count/); 1182 eval { @a = unpack '(3SL)', '' }; 1183 like($@, qr/\(\)-group starts with a count/); 1184 eval { @a = unpack '([3]SL)', '' }; 1185 like($@, qr/\(\)-group starts with a count/); 1186 eval { @a = pack '(*SL)' }; 1187 like($@, qr/\(\)-group starts with a count/); 1188 @a = unpack '(SL)3 SL', pack '(SL)4', 67..74; 1189 is("@a", "@b"); 1190 @a = unpack '(SL)3 SL', pack '(SL)[4]', 67..74; 1191 is("@a", "@b"); 1192 @a = unpack '(SL)3 SL', pack '(SL)*', 67..74; 1193 is("@a", "@b"); 1194} 1195 1196{ # more on grouping (W.Laun) 1197 # @ absolute within ()-group 1198 my $badc = pack( '(a)*', unpack( '(@1a @0a @2)*', 'abcd' ) ); 1199 is( $badc, 'badc' ); 1200 my @b = ( 1, 2, 3 ); 1201 my $buf = pack( '(@1c)((@2C)@3c)', @b ); 1202 is( $buf, "\0\1\0\0\2\3" ); 1203 my @a = unpack( '(@1c)((@2c)@3c)', $buf ); 1204 is( "@a", "@b" ); 1205 1206 # various unpack count/code scenarios 1207 my @Env = ( a => 'AAA', b => 'BBB' ); 1208 my $env = pack( 'S(S/A*S/A*)*', @Env/2, @Env ); 1209 1210 # unpack full length - ok 1211 my @pup = unpack( 'S/(S/A* S/A*)', $env ); 1212 is( "@pup", "@Env" ); 1213 1214 # warn when count/code goes beyond end of string 1215 # \0002 \0001 a \0003 AAA \0001 b \0003 BBB 1216 # 2 4 5 7 10 1213 1217 eval { @pup = unpack( 'S/(S/A* S/A*)', substr( $env, 0, 13 ) ) }; 1218 like( $@, qr{length/code after end of string} ); 1219 1220 # postfix repeat count 1221 $env = pack( '(S/A* S/A*)' . @Env/2, @Env ); 1222 1223 # warn when count/code goes beyond end of string 1224 # \0001 a \0003 AAA \0001 b \0003 BBB 1225 # 2 3c 5 8 10 11 13 16 1226 eval { @pup = unpack( '(S/A* S/A*)' . @Env/2, substr( $env, 0, 11 ) ) }; 1227 like( $@, qr{length/code after end of string} ); 1228 1229 # catch stack overflow/segfault 1230 eval { $_ = pack( ('(' x 105) . 'A' . (')' x 105) ); }; 1231 like( $@, qr{Too deeply nested \(\)-groups} ); 1232} 1233 1234{ # syntax checks (W.Laun) 1235 use warnings qw(NONFATAL all);; 1236 my @warning; 1237 local $SIG{__WARN__} = sub { 1238 push( @warning, $_[0] ); 1239 }; 1240 eval { my $s = pack( 'Ax![4c]A', 1..5 ); }; 1241 like( $@, qr{Malformed integer in \[\]} ); 1242 1243 eval { my $buf = pack( '(c/*a*)', 'AAA', 'BB' ); }; 1244 like( $@, qr{'/' does not take a repeat count} ); 1245 1246 eval { my @inf = unpack( 'c/1a', "\x03AAA\x02BB" ); }; 1247 like( $@, qr{'/' does not take a repeat count} ); 1248 1249 eval { my @inf = unpack( 'c/*a', "\x03AAA\x02BB" ); }; 1250 like( $@, qr{'/' does not take a repeat count} ); 1251 1252 # white space where possible 1253 my @Env = ( a => 'AAA', b => 'BBB' ); 1254 my $env = pack( ' S ( S / A* S / A* )* ', @Env/2, @Env ); 1255 my @pup = unpack( ' S / ( S / A* S / A* ) ', $env ); 1256 is( "@pup", "@Env" ); 1257 1258 # white space in 4 wrong places 1259 for my $temp ( 'A ![4]', 'A [4]', 'A *', 'A 4' ){ 1260 eval { my $s = pack( $temp, 'B' ); }; 1261 like( $@, qr{Invalid type } ); 1262 } 1263 1264 # warning for commas 1265 @warning = (); 1266 my $x = pack( 'I,A', 4, 'X' ); 1267 like( $warning[0], qr{Invalid type ','} ); 1268 1269 # comma warning only once 1270 @warning = (); 1271 $x = pack( 'C(C,C)C,C', 65..71 ); 1272 like( scalar @warning, 1 ); 1273 1274 # forbidden code in [] 1275 eval { my $x = pack( 'A[@4]', 'XXXX' ); }; 1276 like( $@, qr{Within \[\]-length '\@' not allowed} ); 1277 1278 # @ repeat default 1 1279 my $s = pack( 'AA@A', 'A', 'B', 'C' ); 1280 my @c = unpack( 'AA@A', $s ); 1281 is( $s, 'AC' ); 1282 is( "@c", "A C C" ); 1283 1284 # no unpack code after / 1285 eval { my @a = unpack( "C/", "\3" ); }; 1286 like( $@, qr{Code missing after '/'} ); 1287 1288 SKIP: { 1289 skip $no_endianness, 6 if $no_endianness; 1290 1291 # modifier warnings 1292 @warning = (); 1293 $x = pack "I>>s!!", 47, 11; 1294 ($x) = unpack "I<<l!>!>", 'x'x20; 1295 is(scalar @warning, 5); 1296 like($warning[0], qr/Duplicate modifier '>' after 'I' in pack/); 1297 like($warning[1], qr/Duplicate modifier '!' after 's' in pack/); 1298 like($warning[2], qr/Duplicate modifier '<' after 'I' in unpack/); 1299 like($warning[3], qr/Duplicate modifier '!' after 'l' in unpack/); 1300 like($warning[4], qr/Duplicate modifier '>' after 'l' in unpack/); 1301 } 1302} 1303 1304{ # Repeat count [SUBEXPR] 1305 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 1306 s! S! i! I! l! L! j J); 1307 my $G; 1308 if (eval { pack 'q', 1 } ) { 1309 push @codes, qw(q Q); 1310 } else { 1311 push @codes, qw(s S); # Keep the count the same 1312 } 1313 if (eval { pack 'D', 1 } ) { 1314 push @codes, 'D'; 1315 } else { 1316 push @codes, 'd'; # Keep the count the same 1317 } 1318 1319 push @codes, map { /^[silqjfdp]/i ? ("$_<", "$_>") : () } @codes; 1320 1321 my %val; 1322 @val{@codes} = map { / [Xx] (?{ undef }) 1323 | [AZa] (?{ 'something' }) 1324 | C (?{ 214 }) 1325 | W (?{ 8188 }) 1326 | c (?{ 114 }) 1327 | [Bb] (?{ '101' }) 1328 | [Hh] (?{ 'b8' }) 1329 | [svnSiIlVNLqQjJ] (?{ 10111 }) 1330 | [FfDd] (?{ 1.36514538e67 }) 1331 | [pP] (?{ "try this buffer" }) 1332 /x; $^R } @codes; 1333 my @end = (0x12345678, 0x23456781, 0x35465768, 0x15263748); 1334 my $end = "N4"; 1335 1336 for my $type (@codes) { 1337 my @list = $val{$type}; 1338 @list = () unless defined $list[0]; 1339 for my $count ('', '3', '[11]') { 1340 my $c = 1; 1341 $c = $1 if $count =~ /(\d+)/; 1342 my @list1 = @list; 1343 @list1 = (@list1) x $c unless $type =~ /[XxAaZBbHhP]/; 1344 for my $groupend ('', ')2', ')[8]') { 1345 my $groupbegin = ($groupend ? '(' : ''); 1346 $c = 1; 1347 $c = $1 if $groupend =~ /(\d+)/; 1348 my @list2 = (@list1) x $c; 1349 1350 SKIP: { 1351 my $junk1 = "$groupbegin $type$count $groupend"; 1352 # print "# junk1=$junk1\n"; 1353 my $p = eval { pack $junk1, @list2 }; 1354 skip "cannot pack '$type' on this perl", 12 1355 if is_valid_error($@); 1356 die "pack $junk1 failed: $@" if $@; 1357 1358 my $half = int( (length $p)/2 ); 1359 for my $move ('', "X$half", "X!$half", 'x1', 'x!8', "x$half") { 1360 my $junk = "$junk1 $move"; 1361 # print "# junk='$junk', list=(@list2)\n"; 1362 $p = pack "$junk $end", @list2, @end; 1363 my @l = unpack "x[$junk] $end", $p; 1364 is(scalar @l, scalar @end); 1365 is("@l", "@end", "skipping x[$junk]"); 1366 } 1367 } 1368 } 1369 } 1370 } 1371} 1372 1373# / is recognized after spaces in scalar context 1374# XXXX no spaces are allowed in pack... In pack only before the slash... 1375is(scalar unpack('A /A Z20', pack 'A/A* Z20', 'bcde', 'xxxxx'), 'bcde'); 1376is(scalar unpack('A /A /A Z20', '3004bcde'), 'bcde'); 1377 1378{ # X! and x! 1379 my $t = 'C[3] x!8 C[2]'; 1380 my @a = (0x73..0x77); 1381 my $p = pack($t, @a); 1382 is($p, "\x73\x74\x75\0\0\0\0\0\x76\x77"); 1383 my @b = unpack $t, $p; 1384 is(scalar @b, scalar @a); 1385 is("@b", "@a", 'x!8'); 1386 $t = 'x[5] C[6] X!8 C[2]'; 1387 @a = (0x73..0x7a); 1388 $p = pack($t, @a); 1389 is($p, "\0\0\0\0\0\x73\x74\x75\x79\x7a"); 1390 @b = unpack $t, $p; 1391 @a = (0x73..0x75, 0x79, 0x7a, 0x79, 0x7a); 1392 is(scalar @b, scalar @a); 1393 is("@b", "@a"); 1394} 1395 1396{ # struct {char c1; double d; char cc[2];} 1397 my $t = 'C x![d] d C[2]'; 1398 my @a = (173, 1.283476517e-45, 42, 215); 1399 my $p = pack $t, @a; 1400 ok( length $p); 1401 my @b = unpack "$t X[$t] $t", $p; # Extract, step back, extract again 1402 is(scalar @b, 2 * scalar @a); 1403 $b = "@b"; 1404 $b =~ s/(?:17000+|16999+)\d+(e-45) /17$1 /gi; # stringification is gamble 1405 is($b, "@a @a"); 1406 1407 use warnings qw(NONFATAL all);; 1408 my $warning; 1409 local $SIG{__WARN__} = sub { 1410 $warning = $_[0]; 1411 }; 1412 @b = unpack "x[C] x[$t] X[$t] X[C] $t", "$p\0"; 1413 1414 is($warning, undef); 1415 is(scalar @b, scalar @a); 1416 $b = "@b"; 1417 $b =~ s/(?:17000+|16999+)\d+(e-45) /17$1 /gi; # stringification is gamble 1418 is($b, "@a"); 1419} 1420 1421is(length(pack("j", 0)), $Config{ivsize}); 1422is(length(pack("J", 0)), $Config{uvsize}); 1423is(length(pack("F", 0)), $Config{nvsize}); 1424 1425numbers ('j', -2147483648, -1, 0, 1, 2147483647); 1426numbers ('J', 0, 1, 2147483647, 2147483648, 4294967295); 1427numbers ('F', -(2**34), -1, 0, 1, 2**34); 1428SKIP: { 1429 my $t = eval { unpack("D*", pack("D", 12.34)) }; 1430 1431 skip "Long doubles not in use", 166 if $@ =~ /Invalid type/; 1432 1433 is(length(pack("D", 0)), $Config{longdblsize}); 1434 numbers ('D', -(2**34), -1, 0, 1, 2**34); 1435} 1436 1437# Maybe this knowledge needs to be "global" for all of pack.t 1438# Or a "can checksum" which would effectively be all the number types" 1439my %cant_checksum = map {$_=> 1} qw(A Z u w); 1440# not a b B h H 1441foreach my $template (qw(A Z c C s S i I l L n N v V q Q j J f d F D u U w)) { 1442 SKIP: { 1443 my $packed = eval {pack "${template}4", 1, 4, 9, 16}; 1444 if ($@) { 1445 die unless $@ =~ /Invalid type '$template'/; 1446 skip ("$template not supported on this perl", 1447 $cant_checksum{$template} ? 4 : 8); 1448 } 1449 my @unpack4 = unpack "${template}4", $packed; 1450 my @unpack = unpack "${template}*", $packed; 1451 my @unpack1 = unpack "${template}", $packed; 1452 my @unpack1s = scalar unpack "${template}", $packed; 1453 my @unpack4s = scalar unpack "${template}4", $packed; 1454 my @unpacks = scalar unpack "${template}*", $packed; 1455 1456 my @tests = ( ["${template}4 vs ${template}*", \@unpack4, \@unpack], 1457 ["scalar ${template} ${template}", \@unpack1s, \@unpack1], 1458 ["scalar ${template}4 vs ${template}", \@unpack4s, \@unpack1], 1459 ["scalar ${template}* vs ${template}", \@unpacks, \@unpack1], 1460 ); 1461 1462 unless ($cant_checksum{$template}) { 1463 my @unpack4_c = unpack "\%${template}4", $packed; 1464 my @unpack_c = unpack "\%${template}*", $packed; 1465 my @unpack1_c = unpack "\%${template}", $packed; 1466 my @unpack1s_c = scalar unpack "\%${template}", $packed; 1467 my @unpack4s_c = scalar unpack "\%${template}4", $packed; 1468 my @unpacks_c = scalar unpack "\%${template}*", $packed; 1469 1470 push @tests, 1471 ( ["% ${template}4 vs ${template}*", \@unpack4_c, \@unpack_c], 1472 ["% scalar ${template} ${template}", \@unpack1s_c, \@unpack1_c], 1473 ["% scalar ${template}4 vs ${template}*", \@unpack4s_c, \@unpack_c], 1474 ["% scalar ${template}* vs ${template}*", \@unpacks_c, \@unpack_c], 1475 ); 1476 } 1477 foreach my $test (@tests) { 1478 ok (list_eq ($test->[1], $test->[2]), $test->[0]) || 1479 printf "# unpack gave %s expected %s\n", 1480 encode_list (@{$test->[1]}), encode_list (@{$test->[2]}); 1481 } 1482 } 1483} 1484 1485ok(pack('u2', 'AA'), "[perl #8026]"); # used to hang and eat RAM in perl 5.7.2 1486 1487$_ = pack('c', 65); # 'A' would not be EBCDIC-friendly 1488is(unpack('c'), 65, "one-arg unpack (change #18751)"); # defaulting to $_ 1489 1490{ 1491 my $a = "X\x0901234567\n" x 100; # \t would not be EBCDIC TAB 1492 my @a = unpack("(a1 c/a)*", $a); 1493 is(scalar @a, 200, "[perl #15288]"); 1494 is($a[-1], "01234567\n", "[perl #15288]"); 1495 is($a[-2], "X", "[perl #15288]"); 1496} 1497 1498{ 1499 use warnings qw(NONFATAL all);; 1500 my $warning; 1501 local $SIG{__WARN__} = sub { 1502 $warning = $_[0]; 1503 }; 1504 my $out = pack("u99", "foo" x 99); 1505 like($warning, qr/Field too wide in 'u' format in pack at /, 1506 "Warn about too wide uuencode"); 1507 is($out, ("_" . "9F]O" x 21 . "\n") x 4 . "M" . "9F]O" x 15 . "\n", 1508 "Use max width in case of too wide uuencode"); 1509} 1510 1511# checksums 1512{ 1513 # verify that unpack advances correctly wrt a checksum 1514 my (@x) = unpack("b10a", "abcd"); 1515 my (@y) = unpack("%b10a", "abcd"); 1516 is($x[1], $y[1], "checksum advance ok"); 1517 1518 # verify that the checksum is not overflowed with C0 1519 if (ord('A') == 193) { 1520 is(unpack("C0%128U", "/bcd"), unpack("U0%128U", "abcd"), "checksum not overflowed"); 1521 } else { 1522 is(unpack("C0%128U", "abcd"), unpack("U0%128U", "abcd"), "checksum not overflowed"); 1523 } 1524} 1525 1526{ 1527 # U0 and C0 must be scoped 1528 my (@x) = unpack("a(U0)U", "b\341\277\274"); 1529 is($x[0], 'b', 'before scope'); 1530 is($x[1], 8188, 'after scope'); 1531 1532 is(pack("a(U0)U", "b", 8188), "b\341\277\274"); 1533} 1534 1535{ 1536 # counted length prefixes shouldn't change C0/U0 mode 1537 # (note the length is actually 0 in this test) 1538 if (ord('A') == 193) { 1539 is(join(',', unpack("aU0C/UU", "b\0\341\277\274")), 'b,0'); 1540 is(join(',', unpack("aU0C/CU", "b\0\341\277\274")), 'b,0'); 1541 } else { 1542 is(join(',', unpack("aC/UU", "b\0\341\277\274")), 'b,8188'); 1543 is(join(',', unpack("aC/CU", "b\0\341\277\274")), 'b,8188'); 1544 is(join(',', unpack("aU0C/UU", "b\0\341\277\274")), 'b,225'); 1545 is(join(',', unpack("aU0C/CU", "b\0\341\277\274")), 'b,225'); 1546 } 1547} 1548 1549{ 1550 # "Z0" (bug #34062) 1551 my (@x) = unpack("C*", pack("CZ0", 1, "b")); 1552 is(join(',', @x), '1', 'pack Z0 doesn\'t destroy the character before'); 1553} 1554 1555{ 1556 # Encoding neutrality 1557 # String we will pull apart and rebuild in several ways: 1558 my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1559 my $up = $down; 1560 utf8::upgrade($up); 1561 1562 my %expect = 1563 # [expected result, 1564 # how many chars it should progress, 1565 # (optional) expected result of pack] 1566 (a5 => ["\xf8\xf9\xfa\xfb\xfc", 5], 1567 A5 => ["\xf8\xf9\xfa\xfb\xfc", 5], 1568 Z5 => ["\xf8\xf9\xfa\xfb\xfc", 5, "\xf8\xf9\xfa\xfb\x00\xfd"], 1569 b21 => ["000111111001111101011", 3, "\xf8\xf9\x1a\xfb"], 1570 B21 => ["111110001111100111111", 3, "\xf8\xf9\xf8\xfb"], 1571 H5 => ["f8f9f", 3, "\xf8\xf9\xf0\xfb"], 1572 h5 => ["8f9fa", 3, "\xf8\xf9\x0a\xfb"], 1573 "s<" => [-1544, 2], 1574 "s>" => [-1799, 2], 1575 "S<" => [0xf9f8, 2], 1576 "S>" => [0xf8f9, 2], 1577 "l<" => [-67438088, 4], 1578 "l>" => [-117835013, 4], 1579 "L>" => [0xf8f9fafb, 4], 1580 "L<" => [0xfbfaf9f8, 4], 1581 n => [0xf8f9, 2], 1582 N => [0xf8f9fafb, 4], 1583 v => [63992, 2], 1584 V => [0xfbfaf9f8, 4], 1585 c => [-8, 1], 1586 U0U => [0xf8, 1], 1587 w => ["8715569050387726213", 9], 1588 q => ["-283686952306184", 8], 1589 Q => ["18446460386757245432", 8], 1590 ); 1591 1592 for my $string ($down, $up) { 1593 for my $format (sort {lc($a) cmp lc($b) || $a cmp $b } keys %expect) { 1594 SKIP: { 1595 my $expect = $expect{$format}; 1596 # unpack upgraded and downgraded string 1597 my @result = eval { unpack("$format C0 W", $string) }; 1598 skip "cannot pack/unpack '$format C0 W' on this perl", 5 if 1599 $@ && is_valid_error($@); 1600 is(@result, 2, "Two results from unpack $format C0 W"); 1601 1602 # pack to downgraded 1603 my $new = pack("$format C0 W", @result); 1604 is(length($new), $expect->[1]+1, 1605 "pack $format C0 W should give $expect->[1]+1 chars"); 1606 is($new, $expect->[2] || substr($string, 0, length $new), 1607 "pack $format C0 W returns expected value"); 1608 1609 # pack to upgraded 1610 $new = pack("a0 $format C0 W", chr(256), @result); 1611 is(length($new), $expect->[1]+1, 1612 "pack a0 $format C0 W should give $expect->[1]+1 chars"); 1613 is($new, $expect->[2] || substr($string, 0, length $new), 1614 "pack a0 $format C0 W returns expected value"); 1615 } 1616 } 1617 } 1618} 1619 1620{ 1621 # Encoding neutrality, numbers 1622 my $val = -2.68; 1623 for my $format (qw(s S i I l L j J f d F D q Q 1624 s! S! i! I! l! L! n! N! v! V!)) { 1625 SKIP: { 1626 my $down = eval { pack($format, $val) }; 1627 skip "cannot pack/unpack $format on this perl", 9 if 1628 $@ && is_valid_error($@); 1629 ok(!utf8::is_utf8($down), "Simple $format pack doesn't get upgraded"); 1630 my $up = pack("a0 $format", chr(256), $val); 1631 ok(utf8::is_utf8($up), "a0 $format with high char leads to upgrade"); 1632 is($down, $up, "$format generated strings are equal though"); 1633 my @down_expanded = unpack("$format W", $down . chr(0xce)); 1634 is(@down_expanded, 2, "Expand to two values"); 1635 is($down_expanded[1], 0xce, 1636 "unpack $format left us at the expected position"); 1637 my @up_expanded = unpack("$format W", $up . chr(0xce)); 1638 is(@up_expanded, 2, "Expand to two values"); 1639 is($up_expanded[1], 0xce, 1640 "unpack $format left us at the expected position"); 1641 is($down_expanded[0], $up_expanded[0], "$format unpack was neutral"); 1642 is(pack($format, $down_expanded[0]), $down, "Pack $format undoes unpack $format"); 1643 } 1644 } 1645} 1646 1647{ 1648 # C *is* neutral 1649 my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1650 my $up = $down; 1651 utf8::upgrade($up); 1652 my @down = unpack("C*", $down); 1653 my @expect_down = (0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, 0x05, 0x06); 1654 is("@down", "@expect_down", "byte expand"); 1655 is(pack("C*", @down), $down, "byte join"); 1656 1657 my @up = unpack("C*", $up); 1658 my @expect_up = (0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, 0x05, 0x06); 1659 is("@up", "@expect_up", "UTF-8 expand"); 1660 is(pack("U0C0C*", @up), $up, "UTF-8 join"); 1661} 1662 1663{ 1664 # Harder cases for the neutrality test 1665 1666 # u format 1667 my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1668 my $up = $down; 1669 utf8::upgrade($up); 1670 is(pack("u", $down), pack("u", $up), "u pack is neutral"); 1671 is(unpack("u", pack("u", $down)), $down, "u unpack to downgraded works"); 1672 is(unpack("U0C0u", pack("u", $down)), $up, "u unpack to upgraded works"); 1673 1674 # p/P format 1675 # This actually only tests something if the address contains a byte >= 0x80 1676 my $str = "abc\xa5\x00\xfede"; 1677 $down = pack("p", $str); 1678 is(pack("P", $str), $down); 1679 is(pack("U0C0p", $str), $down); 1680 is(pack("U0C0P", $str), $down); 1681 is(unpack("p", $down), "abc\xa5", "unpack p downgraded"); 1682 $up = $down; 1683 utf8::upgrade($up); 1684 is(unpack("p", $up), "abc\xa5", "unpack p upgraded"); 1685 1686 is(unpack("P7", $down), "abc\xa5\x00\xfed", "unpack P downgraded"); 1687 is(unpack("P7", $up), "abc\xa5\x00\xfed", "unpack P upgraded"); 1688 1689 # x, X and @ 1690 $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1691 $up = $down; 1692 utf8::upgrade($up); 1693 1694 is(unpack('@4W', $down), 0xfc, "\@positioning on downgraded string"); 1695 is(unpack('@4W', $up), 0xfc, "\@positioning on upgraded string"); 1696 1697 is(unpack('@4x2W', $down), 0xfe, "x moving on downgraded string"); 1698 is(unpack('@4x2W', $up), 0xfe, "x moving on upgraded string"); 1699 is(unpack('@4x!4W', $down), 0xfc, "x! moving on downgraded string"); 1700 is(unpack('@4x!4W', $up), 0xfc, "x! moving on upgraded string"); 1701 is(unpack('@5x!4W', $down), 0x05, "x! moving on downgraded string"); 1702 is(unpack('@5x!4W', $up), 0x05, "x! moving on upgraded string"); 1703 1704 is(unpack('@4X2W', $down), 0xfa, "X moving on downgraded string"); 1705 is(unpack('@4X2W', $up), 0xfa, "X moving on upgraded string"); 1706 is(unpack('@4X!4W', $down), 0xfc, "X! moving on downgraded string"); 1707 is(unpack('@4X!4W', $up), 0xfc, "X! moving on upgraded string"); 1708 is(unpack('@5X!4W', $down), 0xfc, "X! moving on downgraded string"); 1709 is(unpack('@5X!4W', $up), 0xfc, "X! moving on upgraded string"); 1710 is(unpack('@5X!8W', $down), 0xf8, "X! moving on downgraded string"); 1711 is(unpack('@5X!8W', $up), 0xf8, "X! moving on upgraded string"); 1712 1713 is(pack("W2x", 0xfa, 0xe3), "\xfa\xe3\x00", "x on downgraded string"); 1714 is(pack("W2x!4", 0xfa, 0xe3), "\xfa\xe3\x00\x00", 1715 "x! on downgraded string"); 1716 is(pack("W2x!2", 0xfa, 0xe3), "\xfa\xe3", "x! on downgraded string"); 1717 is(pack("U0C0W2x", 0xfa, 0xe3), "\xfa\xe3\x00", "x on upgraded string"); 1718 is(pack("U0C0W2x!4", 0xfa, 0xe3), "\xfa\xe3\x00\x00", 1719 "x! on upgraded string"); 1720 is(pack("U0C0W2x!2", 0xfa, 0xe3), "\xfa\xe3", "x! on upgraded string"); 1721 is(pack("W2X", 0xfa, 0xe3), "\xfa", "X on downgraded string"); 1722 is(pack("U0C0W2X", 0xfa, 0xe3), "\xfa", "X on upgraded string"); 1723 is(pack("W2X!2", 0xfa, 0xe3), "\xfa\xe3", "X! on downgraded string"); 1724 is(pack("U0C0W2X!2", 0xfa, 0xe3), "\xfa\xe3", "X! on upgraded string"); 1725 is(pack("W3X!2", 0xfa, 0xe3, 0xa6), "\xfa\xe3", "X! on downgraded string"); 1726 is(pack("U0C0W3X!2", 0xfa, 0xe3, 0xa6), "\xfa\xe3", 1727 "X! on upgraded string"); 1728 1729 # backward eating through a ( moves the group starting point backwards 1730 is(pack("a*(Xa)", "abc", "q"), "abq", 1731 "eating before strbeg moves it back"); 1732 is(pack("a*(Xa)", "ab" . chr(512), "q"), "abq", 1733 "eating before strbeg moves it back"); 1734 1735 # Check marked_upgrade 1736 is(pack('W(W(Wa@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, "a", 0xa4, 0xa5, 0xa6), 1737 "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6"); 1738 $up = "a"; 1739 utf8::upgrade($up); 1740 is(pack('W(W(Wa@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, $up, 0xa4, 0xa5, 0xa6), 1741 "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6", "marked upgrade caused by a"); 1742 is(pack('W(W(WW@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, 256, 0xa4, 0xa5, 0xa6), 1743 "\xa1\xa2\xa3\x{100}\x00\xa4\x00\xa5\x00\xa6", 1744 "marked upgrade caused by W"); 1745 is(pack('W(W(WU0aC0@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, "a", 0xa4, 0xa5, 0xa6), 1746 "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6", "marked upgrade caused by U0"); 1747 1748 # a, A and Z 1749 $down = "\xa4\xa6\xa7"; 1750 $up = $down; 1751 utf8::upgrade($up); 1752 utf8::upgrade(my $high = "\xfeb"); 1753 1754 for my $format ("a0", "A0", "Z0", "U0a0C0", "U0A0C0", "U0Z0C0") { 1755 is(pack("a* $format a*", "ab", $down, "cd"), "abcd", 1756 "$format format on plain string"); 1757 is(pack("a* $format a*", "ab", $up, "cd"), "abcd", 1758 "$format format on upgraded string"); 1759 is(pack("a* $format a*", $high, $down, "cd"), "\xfebcd", 1760 "$format format on plain string"); 1761 is(pack("a* $format a*", $high, $up, "cd"), "\xfebcd", 1762 "$format format on upgraded string"); 1763 my @down = unpack("a1 $format a*", "\xfeb"); 1764 is("@down", "\xfe b", "unpack $format"); 1765 my @up = unpack("a1 $format a*", $high); 1766 is("@up", "\xfe b", "unpack $format"); 1767 } 1768 is(pack("a1", $high), "\xfe"); 1769 is(pack("A1", $high), "\xfe"); 1770 is(pack("Z1", $high), "\x00"); 1771 is(pack("a2", $high), "\xfeb"); 1772 is(pack("A2", $high), "\xfeb"); 1773 is(pack("Z2", $high), "\xfe\x00"); 1774 is(pack("a5", $high), "\xfeb\x00\x00\x00"); 1775 is(pack("A5", $high), "\xfeb "); 1776 is(pack("Z5", $high), "\xfeb\x00\x00\x00"); 1777 is(pack("a*", $high), "\xfeb"); 1778 is(pack("A*", $high), "\xfeb"); 1779 is(pack("Z*", $high), "\xfeb\x00"); 1780 1781 utf8::upgrade($high = "\xc3\xbeb"); 1782 is(pack("U0a2", $high), "\xfe"); 1783 is(pack("U0A2", $high), "\xfe"); 1784 is(pack("U0Z1", $high), "\x00"); 1785 is(pack("U0a3", $high), "\xfeb"); 1786 is(pack("U0A3", $high), "\xfeb"); 1787 is(pack("U0Z3", $high), "\xfe\x00"); 1788 is(pack("U0a6", $high), "\xfeb\x00\x00\x00"); 1789 is(pack("U0A6", $high), "\xfeb "); 1790 is(pack("U0Z6", $high), "\xfeb\x00\x00\x00"); 1791 is(pack("U0a*", $high), "\xfeb"); 1792 is(pack("U0A*", $high), "\xfeb"); 1793 is(pack("U0Z*", $high), "\xfeb\x00"); 1794} 1795{ 1796 # pack / 1797 my @array = 1..14; 1798 my @out = unpack("N/S", pack("N/S", @array) . "abcd"); 1799 is("@out", "@array", "pack N/S works"); 1800 @out = unpack("N/S*", pack("N/S*", @array) . "abcd"); 1801 is("@out", "@array", "pack N/S* works"); 1802 @out = unpack("N/S*", pack("N/S14", @array) . "abcd"); 1803 is("@out", "@array", "pack N/S14 works"); 1804 @out = unpack("N/S*", pack("N/S15", @array) . "abcd"); 1805 is("@out", "@array", "pack N/S15 works"); 1806 @out = unpack("N/S*", pack("N/S13", @array) . "abcd"); 1807 is("@out", "@array[0..12]", "pack N/S13 works"); 1808 @out = unpack("N/S*", pack("N/S0", @array) . "abcd"); 1809 is("@out", "", "pack N/S0 works"); 1810 is(pack("Z*/a0", "abc"), "0\0", "pack Z*/a0 makes a short string"); 1811 is(pack("Z*/Z0", "abc"), "0\0", "pack Z*/Z0 makes a short string"); 1812 is(pack("Z*/a3", "abc"), "3\0abc", "pack Z*/a3 makes a full string"); 1813 is(pack("Z*/Z3", "abc"), "3\0ab\0", "pack Z*/Z3 makes a short string"); 1814 is(pack("Z*/a5", "abc"), "5\0abc\0\0", "pack Z*/a5 makes a long string"); 1815 is(pack("Z*/Z5", "abc"), "5\0abc\0\0", "pack Z*/Z5 makes a long string"); 1816 is(pack("Z*/Z"), "1\0\0", "pack Z*/Z makes an extended string"); 1817 is(pack("Z*/Z", ""), "1\0\0", "pack Z*/Z makes an extended string"); 1818 is(pack("Z*/a", ""), "0\0", "pack Z*/a makes an extended string"); 1819} 1820{ 1821 # unpack("A*", $unicode) strips general unicode spaces 1822 is(unpack("A*", "ab \n\xa0 \0"), "ab \n\xa0", 1823 'normal A* strip leaves \xa0'); 1824 is(unpack("U0C0A*", "ab \n\xa0 \0"), "ab \n\xa0", 1825 'normal A* strip leaves \xa0 even if it got upgraded for technical reasons'); 1826 is(unpack("A*", pack("a*(U0U)a*", "ab \n", 0xa0, " \0")), "ab", 1827 'upgraded strings A* removes \xa0'); 1828 is(unpack("A*", pack("a*(U0UU)a*", "ab \n", 0xa0, 0x1680, " \0")), "ab", 1829 'upgraded strings A* removes all unicode whitespace'); 1830 is(unpack("A5", pack("a*(U0U)a*", "ab \n", 0x1680, "def", "ab")), "ab", 1831 'upgraded strings A5 removes all unicode whitespace'); 1832 is(unpack("A*", pack("U", 0x1680)), "", 1833 'upgraded strings A* with nothing left'); 1834} 1835{ 1836 # Testing unpack . and .! 1837 is(unpack(".", "ABCD"), 0, "offset at start of string is 0"); 1838 is(unpack(".", ""), 0, "offset at start of empty string is 0"); 1839 is(unpack("x3.", "ABCDEF"), 3, "simple offset works"); 1840 is(unpack("x3.", "ABC"), 3, "simple offset at end of string works"); 1841 is(unpack("x3.0", "ABC"), 0, "self offset is 0"); 1842 is(unpack("x3(x2.)", "ABCDEF"), 2, "offset is relative to inner group"); 1843 is(unpack("x3(X2.)", "ABCDEF"), -2, 1844 "negative offset relative to inner group"); 1845 is(unpack("x3(X2.2)", "ABCDEF"), 1, "offset is relative to inner group"); 1846 is(unpack("x3(x2.0)", "ABCDEF"), 0, "self offset in group is still 0"); 1847 is(unpack("x3(x2.2)", "ABCDEF"), 5, "offset counts groups"); 1848 is(unpack("x3(x2.*)", "ABCDEF"), 5, "star offset is relative to start"); 1849 1850 my $high = chr(8188) x 6; 1851 is(unpack("x3(x2.)", $high), 2, "utf8 offset is relative to inner group"); 1852 is(unpack("x3(X2.)", $high), -2, 1853 "utf8 negative offset relative to inner group"); 1854 is(unpack("x3(X2.2)", $high), 1, "utf8 offset counts groups"); 1855 is(unpack("x3(x2.0)", $high), 0, "utf8 self offset in group is still 0"); 1856 is(unpack("x3(x2.2)", $high), 5, "utf8 offset counts groups"); 1857 is(unpack("x3(x2.*)", $high), 5, "utf8 star offset is relative to start"); 1858 1859 is(unpack("U0x3(x2.)", $high), 2, 1860 "U0 mode utf8 offset is relative to inner group"); 1861 is(unpack("U0x3(X2.)", $high), -2, 1862 "U0 mode utf8 negative offset relative to inner group"); 1863 is(unpack("U0x3(X2.2)", $high), 1, 1864 "U0 mode utf8 offset counts groups"); 1865 is(unpack("U0x3(x2.0)", $high), 0, 1866 "U0 mode utf8 self offset in group is still 0"); 1867 is(unpack("U0x3(x2.2)", $high), 5, 1868 "U0 mode utf8 offset counts groups"); 1869 is(unpack("U0x3(x2.*)", $high), 5, 1870 "U0 mode utf8 star offset is relative to start"); 1871 1872 is(unpack("x3(x2.!)", $high), 2*3, 1873 "utf8 offset is relative to inner group"); 1874 is(unpack("x3(X2.!)", $high), -2*3, 1875 "utf8 negative offset relative to inner group"); 1876 is(unpack("x3(X2.!2)", $high), 1*3, 1877 "utf8 offset counts groups"); 1878 is(unpack("x3(x2.!0)", $high), 0, 1879 "utf8 self offset in group is still 0"); 1880 is(unpack("x3(x2.!2)", $high), 5*3, 1881 "utf8 offset counts groups"); 1882 is(unpack("x3(x2.!*)", $high), 5*3, 1883 "utf8 star offset is relative to start"); 1884 1885 is(unpack("U0x3(x2.!)", $high), 2, 1886 "U0 mode utf8 offset is relative to inner group"); 1887 is(unpack("U0x3(X2.!)", $high), -2, 1888 "U0 mode utf8 negative offset relative to inner group"); 1889 is(unpack("U0x3(X2.!2)", $high), 1, 1890 "U0 mode utf8 offset counts groups"); 1891 is(unpack("U0x3(x2.!0)", $high), 0, 1892 "U0 mode utf8 self offset in group is still 0"); 1893 is(unpack("U0x3(x2.!2)", $high), 5, 1894 "U0 mode utf8 offset counts groups"); 1895 is(unpack("U0x3(x2.!*)", $high), 5, 1896 "U0 mode utf8 star offset is relative to start"); 1897} 1898{ 1899 # Testing pack . and .! 1900 is(pack("(a)5 .", 1..5, 3), "123", ". relative to string start, shorten"); 1901 eval { () = pack("(a)5 .", 1..5, -3) }; 1902 like($@, qr{'\.' outside of string in pack}, "Proper error message"); 1903 is(pack("(a)5 .", 1..5, 8), "12345\x00\x00\x00", 1904 ". relative to string start, extend"); 1905 is(pack("(a)5 .", 1..5, 5), "12345", ". relative to string start, keep"); 1906 1907 is(pack("(a)5 .0", 1..5, -3), "12", 1908 ". relative to string current, shorten"); 1909 is(pack("(a)5 .0", 1..5, 2), "12345\x00\x00", 1910 ". relative to string current, extend"); 1911 is(pack("(a)5 .0", 1..5, 0), "12345", 1912 ". relative to string current, keep"); 1913 1914 is(pack("(a)5 (.)", 1..5, -3), "12", 1915 ". relative to group, shorten"); 1916 is(pack("(a)5 (.)", 1..5, 2), "12345\x00\x00", 1917 ". relative to group, extend"); 1918 is(pack("(a)5 (.)", 1..5, 0), "12345", 1919 ". relative to group, keep"); 1920 1921 is(pack("(a)3 ((a)2 .)", 1..5, -2), "1", 1922 ". relative to group, shorten"); 1923 is(pack("(a)3 ((a)2 .)", 1..5, 2), "12345", 1924 ". relative to group, keep"); 1925 is(pack("(a)3 ((a)2 .)", 1..5, 4), "12345\x00\x00", 1926 ". relative to group, extend"); 1927 1928 is(pack("(a)3 ((a)2 .2)", 1..5, 2), "12", 1929 ". relative to counted group, shorten"); 1930 is(pack("(a)3 ((a)2 .2)", 1..5, 7), "12345\x00\x00", 1931 ". relative to counted group, extend"); 1932 is(pack("(a)3 ((a)2 .2)", 1..5, 5), "12345", 1933 ". relative to counted group, keep"); 1934 1935 is(pack("(a)3 ((a)2 .*)", 1..5, 2), "12", 1936 ". relative to start, shorten"); 1937 is(pack("(a)3 ((a)2 .*)", 1..5, 7), "12345\x00\x00", 1938 ". relative to start, extend"); 1939 is(pack("(a)3 ((a)2 .*)", 1..5, 5), "12345", 1940 ". relative to start, keep"); 1941 1942 is(pack('(a)5 (. @2 a)', 1..5, -3, "a"), "12\x00\x00a", 1943 ". based shrink properly updates group starts"); 1944 1945 is(pack("(W)3 ((W)2 .)", 0x301..0x305, -2), "\x{301}", 1946 "utf8 . relative to group, shorten"); 1947 is(pack("(W)3 ((W)2 .)", 0x301..0x305, 2), 1948 "\x{301}\x{302}\x{303}\x{304}\x{305}", 1949 "utf8 . relative to group, keep"); 1950 is(pack("(W)3 ((W)2 .)", 0x301..0x305, 4), 1951 "\x{301}\x{302}\x{303}\x{304}\x{305}\x00\x00", 1952 "utf8 . relative to group, extend"); 1953 1954 is(pack("(W)3 ((W)2 .!)", 0x301..0x305, -2), "\x{301}\x{302}", 1955 "utf8 . relative to group, shorten"); 1956 is(pack("(W)3 ((W)2 .!)", 0x301..0x305, 4), 1957 "\x{301}\x{302}\x{303}\x{304}\x{305}", 1958 "utf8 . relative to group, keep"); 1959 is(pack("(W)3 ((W)2 .!)", 0x301..0x305, 6), 1960 "\x{301}\x{302}\x{303}\x{304}\x{305}\x00\x00", 1961 "utf8 . relative to group, extend"); 1962 1963 is(pack('(W)5 (. @2 a)', 0x301..0x305, -3, "a"), 1964 "\x{301}\x{302}\x00\x00a", 1965 "utf8 . based shrink properly updates group starts"); 1966} 1967{ 1968 # Testing @! 1969 is(pack('a* @3', "abcde"), "abc", 'Test basic @'); 1970 is(pack('a* @!3', "abcde"), "abc", 'Test basic @!'); 1971 is(pack('a* @2', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{301}\x{302}", 1972 'Test basic utf8 @'); 1973 is(pack('a* @!2', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{301}", 1974 'Test basic utf8 @!'); 1975 1976 is(unpack('@4 a*', "abcde"), "e", 'Test basic @'); 1977 is(unpack('@!4 a*', "abcde"), "e", 'Test basic @!'); 1978 is(unpack('@4 a*', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{305}", 1979 'Test basic utf8 @'); 1980 is(unpack('@!4 a*', "\x{301}\x{302}\x{303}\x{304}\x{305}"), 1981 "\x{303}\x{304}\x{305}", 'Test basic utf8 @!'); 1982} 1983{ 1984 #50256 1985 my ($v) = split //, unpack ('(B)*', 'ab'); 1986 is($v, 0); # Doesn't SEGV :-) 1987} 1988