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 => 14704; 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 'ultrix') 304 || 305 ($^O =~ /^svr4/ && -f "/etc/issue" && -f "/etc/.relid") # NCR MP-RAS 306 ); 307 308 my $inf = eval '2**1000000'; 309 310 skip("Couldn't generate infinity - got error '$@'", 1) 311 unless defined $inf and $inf == $inf / 2 and $inf + 1 == $inf; 312 313 local our $TODO; 314 $TODO = "VOS needs a fix for posix-1022 to pass this test." 315 if ($^O eq 'vos'); 316 317 eval { $x = pack 'w', $inf }; 318 like ($@, qr/^Cannot compress integer/, "Cannot compress integer"); 319 } 320 321 SKIP: { 322 323 skip("-- the full range of an IEEE double may not be available in this configuration.", 3) 324 if (($^O eq 'VMS') && !defined($Config{useieee})); 325 326 skip("-- $^O does not like 2**1023", 3) 327 if (($^O eq 'ultrix')); 328 329 # This should be about the biggest thing possible on an IEEE double 330 my $big = eval '2**1023'; 331 332 skip("Couldn't generate 2**1023 - got error '$@'", 3) 333 unless defined $big and $big != $big / 2; 334 335 eval { $x = pack 'w', $big }; 336 is ($@, '', "Should be able to pack 'w', $big # 2**1023"); 337 338 my $y = eval {unpack 'w', $x}; 339 is ($@, '', 340 "Should be able to unpack 'w' the result of pack 'w', $big # 2**1023"); 341 342 # I'm getting about 1e-16 on FreeBSD 343 my $quotient = int (100 * ($y - $big) / $big); 344 ok($quotient < 2 && $quotient > -2, 345 "Round trip pack, unpack 'w' of $big is within 1% ($quotient%)"); 346 } 347 348} 349 350print "# test the 'p' template\n"; 351 352# literals 353is(unpack("p",pack("p","foo")), "foo"); 354SKIP: { 355 skip $no_endianness, 2 if $no_endianness; 356 is(unpack("p<",pack("p<","foo")), "foo"); 357 is(unpack("p>",pack("p>","foo")), "foo"); 358} 359# scalars 360is(unpack("p",pack("p",239)), 239); 361SKIP: { 362 skip $no_endianness, 2 if $no_endianness; 363 is(unpack("p<",pack("p<",239)), 239); 364 is(unpack("p>",pack("p>",239)), 239); 365} 366 367# temps 368sub foo { my $a = "a"; return $a . $a++ . $a++ } 369{ 370 use warnings qw(NONFATAL all);; 371 my $warning; 372 local $SIG{__WARN__} = sub { 373 $warning = $_[0]; 374 }; 375 my $junk = pack("p", &foo); 376 377 like($warning, qr/temporary val/); 378} 379 380# undef should give null pointer 381like(pack("p", undef), qr/^\0+$/); 382SKIP: { 383 skip $no_endianness, 2 if $no_endianness; 384 like(pack("p<", undef), qr/^\0+$/); 385 like(pack("p>", undef), qr/^\0+$/); 386} 387 388# Check for optimizer bug (e.g. Digital Unix GEM cc with -O4 on DU V4.0B gives 389# 4294967295 instead of -1) 390# see #ifdef __osf__ in pp.c pp_unpack 391is((unpack("i",pack("i",-1))), -1); 392 393print "# test the pack lengths of s S i I l L n N v V + modifiers\n"; 394 395my @lengths = ( 396 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), 397 's!' => $Config{shortsize}, 'S!' => $Config{shortsize}, 398 'i!' => $Config{intsize}, 'I!' => $Config{intsize}, 399 'l!' => $Config{longsize}, 'L!' => $Config{longsize}, 400); 401 402while (my ($base, $expect) = splice @lengths, 0, 2) { 403 my @formats = ($base); 404 $base =~ /^[nv]/i or push @formats, "$base>", "$base<"; 405 for my $format (@formats) { 406 SKIP: { 407 skip $no_endianness, 1 if $no_endianness && $format =~ m/[<>]/; 408 skip $no_signedness, 1 if $no_signedness && $format =~ /[nNvV]!/; 409 my $len = length(pack($format, 0)); 410 if ($expect > 0) { 411 is($expect, $len, "format '$format'"); 412 } else { 413 $expect = -$expect; 414 ok ($len >= $expect, "format '$format'") || 415 print "# format '$format' has length $len, expected >= $expect\n"; 416 } 417 } 418 } 419} 420 421 422print "# test unpack-pack lengths\n"; 423 424my @templates = qw(c C W i I s S l L n N v V f d q Q); 425 426foreach my $base (@templates) { 427 my @tmpl = ($base); 428 $base =~ /^[cwnv]/i or push @tmpl, "$base>", "$base<"; 429 foreach my $t (@tmpl) { 430 SKIP: { 431 my @t = eval { unpack("$t*", pack("$t*", 12, 34)) }; 432 433 skip "cannot pack '$t' on this perl", 4 434 if is_valid_error($@); 435 436 is( $@, '', "Template $t works"); 437 is(scalar @t, 2); 438 439 is($t[0], 12); 440 is($t[1], 34); 441 } 442 } 443} 444 445{ 446 # uuencode/decode 447 448 # Note that first uuencoding known 'text' data and then checking the 449 # binary values of the uuencoded version would not be portable between 450 # character sets. Uuencoding is meant for encoding binary data, not 451 # text data. 452 453 my $in = pack 'C*', 0 .. 255; 454 455 # just to be anal, we do some random tr/`/ / 456 my $uu = <<'EOUU'; 457M` $"`P0%!@<("0H+# T.#Q`1$A,4%187&!D:&QP='A\@(2(C)"4F)R@I*BLL 458M+2XO,#$R,S0U-C<X.3H[/#T^/T!!0D-$149'2$E*2TQ-3D]045)35%565UA9 459M6EM<75Y?8&%B8V1E9F=H:6IK;&UN;W!Q<G-T=79W>'EZ>WQ]?G^`@8*#A(6& 460MAXB)BHN,C8Z/D)&2DY25EI>8F9J;G)V>GZ"AHJ.DI::GJ*FJJZRMKJ^PL;*S 461MM+6VM[BYNKN\O;Z_P,'"P\3%QL?(R<K+S,W.S]#1TM/4U=;7V-G:V]S=WM_@ 462?X>+CY.7FY^CIZNOL[>[O\/'R\_3U]O?X^?K[_/W^_P ` 463EOUU 464 465 $_ = $uu; 466 tr/ /`/; 467 468 is(pack('u', $in), $_); 469 470 is(unpack('u', $uu), $in); 471 472 $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"; 473 $uu = <<'EOUU'; 474M'XL("%C<Q#4"`TI!4%4`\RHM+E%(S,LOR4@M4@A(+<I1*"U-SD])+>("`&1F 475&8%P:```` 476EOUU 477 478 is(unpack('u', $uu), $in); 479 480# This is identical to the above except that backquotes have been 481# changed to spaces 482 483 $uu = <<'EOUU'; 484M'XL("%C<Q#4" TI!4%4 \RHM+E%(S,LOR4@M4@A(+<I1*"U-SD])+>(" &1F 485&8%P: 486EOUU 487 488 # ' # Grr 489 is(unpack('u', $uu), $in); 490 491} 492 493# test the ascii template types (A, a, Z) 494 495foreach ( 496['p', 'A*', "foo\0bar\0 ", "foo\0bar\0 "], 497['p', 'A11', "foo\0bar\0 ", "foo\0bar\0 "], 498['u', 'A*', "foo\0bar \0", "foo\0bar"], 499['u', 'A8', "foo\0bar \0", "foo\0bar"], 500['p', 'a*', "foo\0bar\0 ", "foo\0bar\0 "], 501['p', 'a11', "foo\0bar\0 ", "foo\0bar\0 \0\0"], 502['u', 'a*', "foo\0bar \0", "foo\0bar \0"], 503['u', 'a8', "foo\0bar \0", "foo\0bar "], 504['p', 'Z*', "foo\0bar\0 ", "foo\0bar\0 \0"], 505['p', 'Z11', "foo\0bar\0 ", "foo\0bar\0 \0\0"], 506['p', 'Z3', "foo", "fo\0"], 507['u', 'Z*', "foo\0bar \0", "foo"], 508['u', 'Z8', "foo\0bar \0", "foo"], 509) 510{ 511 my ($what, $template, $in, $out) = @$_; 512 my $got = $what eq 'u' ? (unpack $template, $in) : (pack $template, $in); 513 unless (is($got, $out)) { 514 my $un = $what eq 'u' ? 'un' : ''; 515 print "# ${un}pack ('$template', "._qq($in).') gave '._qq($out). 516 ' not '._qq($got)."\n"; 517 } 518} 519 520print "# packing native shorts/ints/longs\n"; 521 522is(length(pack("s!", 0)), $Config{shortsize}); 523is(length(pack("i!", 0)), $Config{intsize}); 524is(length(pack("l!", 0)), $Config{longsize}); 525ok(length(pack("s!", 0)) <= length(pack("i!", 0))); 526ok(length(pack("i!", 0)) <= length(pack("l!", 0))); 527is(length(pack("i!", 0)), length(pack("i", 0))); 528 529sub numbers { 530 my $base = shift; 531 my @formats = ($base); 532 $base =~ /^[silqjfdp]/i and push @formats, "$base>", "$base<"; 533 for my $format (@formats) { 534 numbers_with_total ($format, undef, @_); 535 } 536} 537 538sub numbers_with_total { 539 my $format = shift; 540 my $total = shift; 541 if (!defined $total) { 542 foreach (@_) { 543 $total += $_; 544 } 545 } 546 print "# numbers test for $format\n"; 547 foreach (@_) { 548 SKIP: { 549 my $out = eval {unpack($format, pack($format, $_))}; 550 skip "cannot pack '$format' on this perl", 2 551 if is_valid_error($@); 552 553 is($@, '', "no error"); 554 is($out, $_, "unpack pack $format $_"); 555 } 556 } 557 558 my $skip_if_longer_than = ~0; # "Infinity" 559 if (~0 - 1 == ~0) { 560 # If we're running with -DNO_PERLPRESERVE_IVUV and NVs don't preserve all 561 # UVs (in which case ~0 is NV, ~0-1 will be the same NV) then we can't 562 # correctly in perl calculate UV totals for long checksums, as pp_unpack 563 # is using UV maths, and we've only got NVs. 564 $skip_if_longer_than = $Config{nv_preserves_uv_bits}; 565 } 566 567 foreach ('', 1, 2, 3, 15, 16, 17, 31, 32, 33, 53, 54, 63, 64, 65) { 568 SKIP: { 569 my $sum = eval {unpack "%$_$format*", pack "$format*", @_}; 570 skip "cannot pack '$format' on this perl", 3 571 if is_valid_error($@); 572 573 is($@, '', "no error"); 574 ok(defined $sum, "sum bits $_, format $format defined"); 575 576 my $len = $_; # Copy, so that we can reassign '' 577 $len = 16 unless length $len; 578 579 SKIP: { 580 skip "cannot test checksums over $skip_if_longer_than bits", 1 581 if $len > $skip_if_longer_than; 582 583 # Our problem with testing this portably is that the checksum code in 584 # pp_unpack is able to cast signed to unsigned, and do modulo 2**n 585 # arithmetic in unsigned ints, which perl has no operators to do. 586 # (use integer; does signed ints, which won't wrap on UTS, which is just 587 # fine with ANSI, but not with most people's assumptions. 588 # This is why we need to supply the totals for 'Q' as there's no way in 589 # perl to calculate them, short of unpack '%0Q' (is that documented?) 590 # ** returns NVs; make sure it's IV. 591 my $max = 1 + 2 * (int (2 ** ($len-1))-1); # The max possible checksum 592 my $max_p1 = $max + 1; 593 my ($max_is_integer, $max_p1_is_integer); 594 $max_p1_is_integer = 1 unless $max_p1 + 1 == $max_p1; 595 $max_is_integer = 1 if $max - 1 < ~0; 596 597 my $calc_sum; 598 if (ref $total) { 599 $calc_sum = &$total($len); 600 } else { 601 $calc_sum = $total; 602 # Shift into range by some multiple of the total 603 my $mult = $max_p1 ? int ($total / $max_p1) : undef; 604 # Need this to make sure that -1 + (~0+1) is ~0 (ie still integer) 605 $calc_sum = $total - $mult; 606 $calc_sum -= $mult * $max; 607 if ($calc_sum < 0) { 608 $calc_sum += 1; 609 $calc_sum += $max; 610 } 611 } 612 if ($calc_sum == $calc_sum - 1 && $calc_sum == $max_p1) { 613 # we're into floating point (either by getting out of the range of 614 # UV arithmetic, or because we're doing a floating point checksum) 615 # and our calculation of the checksum has become rounded up to 616 # max_checksum + 1 617 $calc_sum = 0; 618 } 619 620 if ($calc_sum == $sum) { # HAS to be ==, not eq (so no is()). 621 pass ("unpack '%$_$format' gave $sum"); 622 } else { 623 my $delta = 1.000001; 624 if ($format =~ tr /dDfF// 625 && ($calc_sum <= $sum * $delta && $calc_sum >= $sum / $delta)) { 626 pass ("unpack '%$_$format' gave $sum, expected $calc_sum"); 627 } else { 628 my $text = ref $total ? &$total($len) : $total; 629 fail; 630 print "# For list (" . join (", ", @_) . ") (total $text)" 631 . " packed with $format unpack '%$_$format' gave $sum," 632 . " expected $calc_sum\n"; 633 } 634 } 635 } 636 } 637 } 638} 639 640numbers ('c', -128, -1, 0, 1, 127); 641numbers ('C', 0, 1, 127, 128, 255); 642numbers ('W', 0, 1, 127, 128, 255, 256, 0x7ff, 0x800, 0xfffd); 643numbers ('s', -32768, -1, 0, 1, 32767); 644numbers ('S', 0, 1, 32767, 32768, 65535); 645numbers ('i', -2147483648, -1, 0, 1, 2147483647); 646numbers ('I', 0, 1, 2147483647, 2147483648, 4294967295); 647numbers ('l', -2147483648, -1, 0, 1, 2147483647); 648numbers ('L', 0, 1, 2147483647, 2147483648, 4294967295); 649numbers ('s!', -32768, -1, 0, 1, 32767); 650numbers ('S!', 0, 1, 32767, 32768, 65535); 651numbers ('i!', -2147483648, -1, 0, 1, 2147483647); 652numbers ('I!', 0, 1, 2147483647, 2147483648, 4294967295); 653numbers ('l!', -2147483648, -1, 0, 1, 2147483647); 654numbers ('L!', 0, 1, 2147483647, 2147483648, 4294967295); 655numbers ('n', 0, 1, 32767, 32768, 65535); 656numbers ('v', 0, 1, 32767, 32768, 65535); 657numbers ('N', 0, 1, 2147483647, 2147483648, 4294967295); 658numbers ('V', 0, 1, 2147483647, 2147483648, 4294967295); 659numbers ('n!', -32768, -1, 0, 1, 32767); 660numbers ('v!', -32768, -1, 0, 1, 32767); 661numbers ('N!', -2147483648, -1, 0, 1, 2147483647); 662numbers ('V!', -2147483648, -1, 0, 1, 2147483647); 663# All these should have exact binary representations: 664numbers ('f', -1, 0, 0.5, 42, 2**34); 665numbers ('d', -(2**34), -1, 0, 1, 2**34); 666## These don't, but 'd' is NV. XXX wrong, it's double 667#numbers ('d', -1, 0, 1, 1-exp(-1), -exp(1)); 668 669numbers_with_total ('q', -1, 670 -9223372036854775808, -1, 0, 1,9223372036854775807); 671# This total is icky, but the true total is 2**65-1, and need a way to generate 672# the expected checksum on any system including those where NVs can preserve 673# 65 bits. (long double is 128 bits on sparc, so they certainly can) 674# or where rounding is down not up on binary conversion (crays) 675numbers_with_total ('Q', sub { 676 my $len = shift; 677 $len = 65 if $len > 65; # unmasked total is 2**65-1 here 678 my $total = 1 + 2 * (int (2**($len - 1)) - 1); 679 return 0 if $total == $total - 1; # Overflowed integers 680 return $total; # NVs still accurate to nearest integer 681 }, 682 0, 1,9223372036854775807, 9223372036854775808, 683 18446744073709551615); 684 685print "# pack nvNV byteorders\n"; 686 687is(pack("n", 0xdead), "\xde\xad"); 688is(pack("v", 0xdead), "\xad\xde"); 689is(pack("N", 0xdeadbeef), "\xde\xad\xbe\xef"); 690is(pack("V", 0xdeadbeef), "\xef\xbe\xad\xde"); 691 692SKIP: { 693 skip $no_signedness, 4 if $no_signedness; 694 is(pack("n!", 0xdead), "\xde\xad"); 695 is(pack("v!", 0xdead), "\xad\xde"); 696 is(pack("N!", 0xdeadbeef), "\xde\xad\xbe\xef"); 697 is(pack("V!", 0xdeadbeef), "\xef\xbe\xad\xde"); 698} 699 700print "# test big-/little-endian conversion\n"; 701 702sub byteorder 703{ 704 my $format = shift; 705 print "# byteorder test for $format\n"; 706 for my $value (@_) { 707 SKIP: { 708 my($nat,$be,$le) = eval { map { pack $format.$_, $value } '', '>', '<' }; 709 skip "cannot pack '$format' on this perl", 5 710 if is_valid_error($@); 711 712 { 713 use warnings qw(NONFATAL utf8); 714 print "# [$value][$nat][$be][$le][$@]\n"; 715 } 716 717 SKIP: { 718 skip "cannot compare native byteorder with big-/little-endian", 1 719 if $ByteOrder eq 'unknown'; 720 721 is($nat, $ByteOrder eq 'big' ? $be : $le); 722 } 723 is($be, reverse($le)); 724 my @x = eval { unpack "$format$format>$format<", $nat.$be.$le }; 725 726 print "# [$value][", join('][', @x), "][$@]\n"; 727 728 is($@, ''); 729 is($x[0], $x[1]); 730 is($x[0], $x[2]); 731 } 732 } 733} 734 735byteorder('s', -32768, -1, 0, 1, 32767); 736byteorder('S', 0, 1, 32767, 32768, 65535); 737byteorder('i', -2147483648, -1, 0, 1, 2147483647); 738byteorder('I', 0, 1, 2147483647, 2147483648, 4294967295); 739byteorder('l', -2147483648, -1, 0, 1, 2147483647); 740byteorder('L', 0, 1, 2147483647, 2147483648, 4294967295); 741byteorder('j', -2147483648, -1, 0, 1, 2147483647); 742byteorder('J', 0, 1, 2147483647, 2147483648, 4294967295); 743byteorder('s!', -32768, -1, 0, 1, 32767); 744byteorder('S!', 0, 1, 32767, 32768, 65535); 745byteorder('i!', -2147483648, -1, 0, 1, 2147483647); 746byteorder('I!', 0, 1, 2147483647, 2147483648, 4294967295); 747byteorder('l!', -2147483648, -1, 0, 1, 2147483647); 748byteorder('L!', 0, 1, 2147483647, 2147483648, 4294967295); 749byteorder('q', -9223372036854775808, -1, 0, 1, 9223372036854775807); 750byteorder('Q', 0, 1, 9223372036854775807, 9223372036854775808, 18446744073709551615); 751byteorder('f', -1, 0, 0.5, 42, 2**34); 752byteorder('F', -1, 0, 0.5, 42, 2**34); 753byteorder('d', -(2**34), -1, 0, 1, 2**34); 754byteorder('D', -(2**34), -1, 0, 1, 2**34); 755 756print "# test negative numbers\n"; 757 758SKIP: { 759 skip "platform is not using two's complement for negative integers", 120 760 unless $IsTwosComplement; 761 762 for my $format (qw(s i l j s! i! l! q)) { 763 SKIP: { 764 my($nat,$be,$le) = eval { map { pack $format.$_, -1 } '', '>', '<' }; 765 skip "cannot pack '$format' on this perl", 15 766 if is_valid_error($@); 767 768 my $len = length $nat; 769 is($_, "\xFF"x$len) for $nat, $be, $le; 770 771 my(@val,@ref); 772 if ($len >= 8) { 773 @val = (-2, -81985529216486896, -9223372036854775808); 774 @ref = ("\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFE", 775 "\xFE\xDC\xBA\x98\x76\x54\x32\x10", 776 "\x80\x00\x00\x00\x00\x00\x00\x00"); 777 } 778 elsif ($len >= 4) { 779 @val = (-2, -19088744, -2147483648); 780 @ref = ("\xFF\xFF\xFF\xFE", 781 "\xFE\xDC\xBA\x98", 782 "\x80\x00\x00\x00"); 783 } 784 else { 785 @val = (-2, -292, -32768); 786 @ref = ("\xFF\xFE", 787 "\xFE\xDC", 788 "\x80\x00"); 789 } 790 for my $x (@ref) { 791 if ($len > length $x) { 792 $x = $x . "\xFF" x ($len - length $x); 793 } 794 } 795 796 for my $i (0 .. $#val) { 797 my($nat,$be,$le) = eval { map { pack $format.$_, $val[$i] } '', '>', '<' }; 798 is($@, ''); 799 800 SKIP: { 801 skip "cannot compare native byteorder with big-/little-endian", 1 802 if $ByteOrder eq 'unknown'; 803 804 is($nat, $ByteOrder eq 'big' ? $be : $le); 805 } 806 807 is($be, $ref[$i]); 808 is($be, reverse($le)); 809 } 810 } 811 } 812} 813 814{ 815 # / 816 817 my ($x, $y, $z, @a); 818 eval { ($x) = unpack '/a*','hello' }; 819 like($@, qr!'/' must follow a numeric type!); 820 undef $x; 821 eval { $x = unpack '/a*','hello' }; 822 like($@, qr!'/' must follow a numeric type!); 823 824 # [perl #60204] Unhelpful error message from unpack 825 eval { @a = unpack 'v/a*','h' }; 826 is($@, ''); 827 is(scalar @a, 0); 828 eval { $x = unpack 'v/a*','h' }; 829 is($@, ''); 830 is($x, undef); 831 832 undef $x; 833 eval { ($z,$x,$y) = unpack 'a3/A C/a* C/Z', "003ok \003yes\004z\000abc" }; 834 is($@, ''); 835 is($z, 'ok'); 836 is($x, 'yes'); 837 is($y, 'z'); 838 undef $z; 839 eval { $z = unpack 'a3/A C/a* C/Z', "003ok \003yes\004z\000abc" }; 840 is($@, ''); 841 is($z, 'ok'); 842 843 844 undef $x; 845 eval { ($x) = pack '/a*','hello' }; 846 like($@, qr!Invalid type '/'!); 847 undef $x; 848 eval { $x = pack '/a*','hello' }; 849 like($@, qr!Invalid type '/'!); 850 851 $z = pack 'n/a* N/Z* w/A*','string','hi there ','etc'; 852 my $expect = "\000\006string\0\0\0\012hi there \000\003etc"; 853 is($z, $expect); 854 855 undef $x; 856 $expect = 'hello world'; 857 eval { ($x) = unpack ("w/a", chr (11) . "hello world!")}; 858 is($x, $expect); 859 is($@, ''); 860 861 undef $x; 862 # Doing this in scalar context used to fail. 863 eval { $x = unpack ("w/a", chr (11) . "hello world!")}; 864 is($@, ''); 865 is($x, $expect); 866 867 foreach ( 868 ['a/a*/a*', '212ab345678901234567','ab3456789012'], 869 ['a/a*/a*', '3012ab345678901234567', 'ab3456789012'], 870 ['a/a*/b*', '212ab', $Is_EBCDIC ? '100000010100' : '100001100100'], 871 ) 872 { 873 my ($pat, $in, $expect) = @$_; 874 undef $x; 875 eval { ($x) = unpack $pat, $in }; 876 is($@, ''); 877 is($x, $expect) || 878 printf "# list unpack ('$pat', '$in') gave %s, expected '$expect'\n", 879 encode_list ($x); 880 881 undef $x; 882 eval { $x = unpack $pat, $in }; 883 is($@, ''); 884 is($x, $expect) || 885 printf "# scalar unpack ('$pat', '$in') gave %s, expected '$expect'\n", 886 encode_list ($x); 887 } 888 889 # / with # 890 891 my $pattern = <<'EOU'; 892 a3/A # Count in ASCII 893 C/a* # Count in a C char 894 C/Z # Count in a C char but skip after \0 895EOU 896 897 $x = $y = $z =undef; 898 eval { ($z,$x,$y) = unpack $pattern, "003ok \003yes\004z\000abc" }; 899 is($@, ''); 900 is($z, 'ok'); 901 is($x, 'yes'); 902 is($y, 'z'); 903 undef $x; 904 eval { $z = unpack $pattern, "003ok \003yes\004z\000abc" }; 905 is($@, ''); 906 is($z, 'ok'); 907 908 $pattern = <<'EOP'; 909 n/a* # Count as network short 910 w/A* # Count a BER integer 911EOP 912 $expect = "\000\006string\003etc"; 913 $z = pack $pattern,'string','etc'; 914 is($z, $expect); 915} 916 917 918SKIP: { 919 skip("(EBCDIC and) version strings are bad idea", 2) if $Is_EBCDIC; 920 921 is("1.20.300.4000", sprintf "%vd", pack("U*",1,20,300,4000)); 922 is("1.20.300.4000", sprintf "%vd", pack(" U*",1,20,300,4000)); 923} 924isnt(v1.20.300.4000, sprintf "%vd", pack("C0U*",1,20,300,4000)); 925 926my $rslt = $Is_EBCDIC ? "156 67" : "199 162"; 927is(join(" ", unpack("U0 C*", chr(0x1e2))), $rslt); 928 929# does pack U create Unicode? 930is(ord(pack('U', 300)), 300); 931 932# does unpack U deref Unicode? 933is((unpack('U', chr(300)))[0], 300); 934 935# is unpack U the reverse of pack U for Unicode string? 936is("@{[unpack('U*', pack('U*', 100, 200, 300))]}", "100 200 300"); 937 938# is unpack U the reverse of pack U for byte string? 939is("@{[unpack('U*', pack('U*', 100, 200))]}", "100 200"); 940 941 942SKIP: { 943 skip "Not for EBCDIC", 4 if $Is_EBCDIC; 944 945 # does pack U0C create Unicode? 946 is("@{[pack('U0C*', 100, 195, 136)]}", v100.v200); 947 948 # does pack C0U create characters? 949 is("@{[pack('C0U*', 100, 200)]}", pack("C*", 100, 195, 136)); 950 951 # does unpack U0U on byte data warn? 952 { 953 use warnings qw(NONFATAL all);; 954 955 my $bad = pack("U0C", 255); 956 local $SIG{__WARN__} = sub { $@ = "@_" }; 957 my @null = unpack('U0U', $bad); 958 like($@, qr/^Malformed UTF-8 character /); 959 } 960} 961 962{ 963 my $p = pack 'i*', -2147483648, ~0, 0, 1, 2147483647; 964 my (@a); 965 # bug - % had to be at the start of the pattern, no leading whitespace or 966 # comments. %i! didn't work at all. 967 foreach my $pat ('%32i*', ' %32i*', "# Muhahahaha\n%32i*", '%32i* ', 968 '%32i!*', ' %32i!*', "\n#\n#\n\r \t\f%32i!*", '%32i!*#') { 969 @a = unpack $pat, $p; 970 is($a[0], 0xFFFFFFFF) || print "# $pat\n"; 971 @a = scalar unpack $pat, $p; 972 is($a[0], 0xFFFFFFFF) || print "# $pat\n"; 973 } 974 975 976 $p = pack 'I*', 42, 12; 977 # Multiline patterns in scalar context failed. 978 foreach my $pat ('I', <<EOPOEMSNIPPET, 'I#I', 'I # I', 'I # !!!') { 979# On the Ning Nang Nong 980# Where the Cows go Bong! 981# And the Monkeys all say Boo! 982I 983EOPOEMSNIPPET 984 @a = unpack $pat, $p; 985 is(scalar @a, 1); 986 is($a[0], 42); 987 @a = scalar unpack $pat, $p; 988 is(scalar @a, 1); 989 is($a[0], 42); 990 } 991 992 # shorts (of all flavours) didn't calculate checksums > 32 bits with floating 993 # point, so a pathologically long pattern would wrap at 32 bits. 994 my $pat = "\xff\xff"x65538; # Start with it long, to save any copying. 995 foreach (4,3,2,1,0) { 996 my $len = 65534 + $_; 997 is(unpack ("%33n$len", $pat), 65535 * $len); 998 } 999} 1000 1001 1002# pack x X @ 1003foreach ( 1004 ['x', "N", "\0"], 1005 ['x4', "N", "\0"x4], 1006 ['xX', "N", ""], 1007 ['xXa*', "Nick", "Nick"], 1008 ['a5Xa5', "cameL", "llama", "camellama"], 1009 ['@4', 'N', "\0"x4], 1010 ['a*@8a*', 'Camel', 'Dromedary', "Camel\0\0\0Dromedary"], 1011 ['a*@4a', 'Perl rules', '!', 'Perl!'], 1012) 1013{ 1014 my ($template, @in) = @$_; 1015 my $out = pop @in; 1016 my $got = eval {pack $template, @in}; 1017 is($@, ''); 1018 is($out, $got) || 1019 printf "# pack ('$template', %s) gave %s expected %s\n", 1020 encode_list (@in), encode_list ($got), encode_list ($out); 1021} 1022 1023# unpack x X @ 1024foreach ( 1025 ['x', "N"], 1026 ['xX', "N"], 1027 ['xXa*', "Nick", "Nick"], 1028 ['a5Xa5', "camellama", "camel", "llama"], 1029 ['@3', "ice"], 1030 ['@2a2', "water", "te"], 1031 ['a*@1a3', "steam", "steam", "tea"], 1032) 1033{ 1034 my ($template, $in, @out) = @$_; 1035 my @got = eval {unpack $template, $in}; 1036 is($@, ''); 1037 ok (list_eq (\@got, \@out)) || 1038 printf "# list unpack ('$template', %s) gave %s expected %s\n", 1039 _qq($in), encode_list (@got), encode_list (@out); 1040 1041 my $got = eval {unpack $template, $in}; 1042 is($@, ''); 1043 @out ? is( $got, $out[0] ) # 1 or more items; should get first 1044 : ok( !defined $got ) # 0 items; should get undef 1045 or printf "# scalar unpack ('$template', %s) gave %s expected %s\n", 1046 _qq($in), encode_list ($got), encode_list ($out[0]); 1047} 1048 1049{ 1050 my $t = 'Z*Z*'; 1051 my ($u, $v) = qw(foo xyzzy); 1052 my $p = pack($t, $u, $v); 1053 my @u = unpack($t, $p); 1054 is(scalar @u, 2); 1055 is($u[0], $u); 1056 is($u[1], $v); 1057} 1058 1059{ 1060 is((unpack("w/a*", "\x02abc"))[0], "ab"); 1061 1062 # "w/a*" should be seen as one unit 1063 1064 is(scalar unpack("w/a*", "\x02abc"), "ab"); 1065} 1066 1067SKIP: { 1068 print "# group modifiers\n"; 1069 1070 skip $no_endianness, 3 * 2 + 3 * 2 + 1 if $no_endianness; 1071 1072 for my $t (qw{ (s<)< (sl>s)> (s(l(sl)<l)s)< }) { 1073 print "# testing pattern '$t'\n"; 1074 eval { ($_) = unpack($t, 'x'x18); }; 1075 is($@, ''); 1076 eval { $_ = pack($t, (0)x6); }; 1077 is($@, ''); 1078 } 1079 1080 for my $t (qw{ (s<)> (sl>s)< (s(l(sl)<l)s)> }) { 1081 print "# testing pattern '$t'\n"; 1082 eval { ($_) = unpack($t, 'x'x18); }; 1083 like($@, qr/Can't use '[<>]' in a group with different byte-order in unpack/); 1084 eval { $_ = pack($t, (0)x6); }; 1085 like($@, qr/Can't use '[<>]' in a group with different byte-order in pack/); 1086 } 1087 1088 is(pack('L<L>', (0x12345678)x2), 1089 pack('(((L1)1)<)(((L)1)1)>1', (0x12345678)x2)); 1090} 1091 1092{ 1093 sub compress_template { 1094 my $t = shift; 1095 for my $mod (qw( < > )) { 1096 $t =~ s/((?:(?:[SILQJFDP]!?$mod|[^SILQJFDP\W]!?)(?:\d+|\*|\[(?:[^]]+)\])?\/?){2,})/ 1097 my $x = $1; $x =~ s!$mod!!g ? "($x)$mod" : $x /ieg; 1098 } 1099 return $t; 1100 } 1101 1102 my %templates = ( 1103 's<' => [-42], 1104 's<c2x![S]S<' => [-42, -11, 12, 4711], 1105 '(i<j<[s]l<)3' => [-11, -22, -33, 1000000, 1100, 2201, 3302, 1106 -1000000, 32767, -32768, 1, -123456789 ], 1107 '(I!<4(J<2L<)3)5' => [1 .. 65], 1108 'q<Q<' => [-50000000005, 60000000006], 1109 'f<F<d<' => [3.14159, 111.11, 2222.22], 1110 'D<cCD<' => [1e42, -128, 255, 1e-42], 1111 'n/a*' => ['/usr/bin/perl'], 1112 'C/a*S</A*L</Z*I</a*' => [qw(Just another Perl hacker)], 1113 ); 1114 1115 for my $tle (sort keys %templates) { 1116 my @d = @{$templates{$tle}}; 1117 my $tbe = $tle; 1118 $tbe =~ y/</>/; 1119 for my $t ($tbe, $tle) { 1120 my $c = compress_template($t); 1121 print "# '$t' -> '$c'\n"; 1122 SKIP: { 1123 my $p1 = eval { pack $t, @d }; 1124 skip "cannot pack '$t' on this perl", 5 if is_valid_error($@); 1125 my $p2 = eval { pack $c, @d }; 1126 is($@, ''); 1127 is($p1, $p2); 1128 s!(/[aAZ])\*!$1!g for $t, $c; 1129 my @u1 = eval { unpack $t, $p1 }; 1130 is($@, ''); 1131 my @u2 = eval { unpack $c, $p2 }; 1132 is($@, ''); 1133 is(join('!', @u1), join('!', @u2)); 1134 } 1135 } 1136 } 1137} 1138 1139{ 1140 # from Wolfgang Laun: fix in change #13163 1141 1142 my $s = 'ABC' x 10; 1143 my $t = '*'; 1144 my $x = ord($t); 1145 my $buf = pack( 'Z*/A* C', $s, $x ); 1146 my $y; 1147 1148 my $h = $buf; 1149 $h =~ s/[^[:print:]]/./g; 1150 ( $s, $y ) = unpack( "Z*/A* C", $buf ); 1151 is($h, "30.ABCABCABCABCABCABCABCABCABCABC$t"); 1152 is(length $buf, 34); 1153 is($s, "ABCABCABCABCABCABCABCABCABCABC"); 1154 is($y, $x); 1155} 1156 1157{ 1158 # from Wolfgang Laun: fix in change #13288 1159 1160 eval { my $t=unpack("P*", "abc") }; 1161 like($@, qr/'P' must have an explicit size/); 1162} 1163 1164{ # Grouping constructs 1165 my (@a, @b); 1166 @a = unpack '(SL)', pack 'SLSLSL', 67..90; 1167 is("@a", "67 68"); 1168 @a = unpack '(SL)3', pack 'SLSLSL', 67..90; 1169 @b = (67..72); 1170 is("@a", "@b"); 1171 @a = unpack '(SL)3', pack 'SLSLSLSL', 67..90; 1172 is("@a", "@b"); 1173 @a = unpack '(SL)[3]', pack 'SLSLSLSL', 67..90; 1174 is("@a", "@b"); 1175 @a = unpack '(SL)[2] SL', pack 'SLSLSLSL', 67..90; 1176 is("@a", "@b"); 1177 @a = unpack 'A/(SL)', pack 'ASLSLSLSL', 3, 67..90; 1178 is("@a", "@b"); 1179 @a = unpack 'A/(SL)SL', pack 'ASLSLSLSL', 2, 67..90; 1180 is("@a", "@b"); 1181 @a = unpack '(SL)*', pack 'SLSLSLSL', 67..90; 1182 @b = (67..74); 1183 is("@a", "@b"); 1184 @a = unpack '(SL)*SL', pack 'SLSLSLSL', 67..90; 1185 is("@a", "@b"); 1186 eval { @a = unpack '(*SL)', '' }; 1187 like($@, qr/\(\)-group starts with a count/); 1188 eval { @a = unpack '(3SL)', '' }; 1189 like($@, qr/\(\)-group starts with a count/); 1190 eval { @a = unpack '([3]SL)', '' }; 1191 like($@, qr/\(\)-group starts with a count/); 1192 eval { @a = pack '(*SL)' }; 1193 like($@, qr/\(\)-group starts with a count/); 1194 @a = unpack '(SL)3 SL', pack '(SL)4', 67..74; 1195 is("@a", "@b"); 1196 @a = unpack '(SL)3 SL', pack '(SL)[4]', 67..74; 1197 is("@a", "@b"); 1198 @a = unpack '(SL)3 SL', pack '(SL)*', 67..74; 1199 is("@a", "@b"); 1200} 1201 1202{ # more on grouping (W.Laun) 1203 # @ absolute within ()-group 1204 my $badc = pack( '(a)*', unpack( '(@1a @0a @2)*', 'abcd' ) ); 1205 is( $badc, 'badc' ); 1206 my @b = ( 1, 2, 3 ); 1207 my $buf = pack( '(@1c)((@2C)@3c)', @b ); 1208 is( $buf, "\0\1\0\0\2\3" ); 1209 my @a = unpack( '(@1c)((@2c)@3c)', $buf ); 1210 is( "@a", "@b" ); 1211 1212 # various unpack count/code scenarios 1213 my @Env = ( a => 'AAA', b => 'BBB' ); 1214 my $env = pack( 'S(S/A*S/A*)*', @Env/2, @Env ); 1215 1216 # unpack full length - ok 1217 my @pup = unpack( 'S/(S/A* S/A*)', $env ); 1218 is( "@pup", "@Env" ); 1219 1220 # warn when count/code goes beyond end of string 1221 # \0002 \0001 a \0003 AAA \0001 b \0003 BBB 1222 # 2 4 5 7 10 1213 1223 eval { @pup = unpack( 'S/(S/A* S/A*)', substr( $env, 0, 13 ) ) }; 1224 like( $@, qr{length/code after end of string} ); 1225 1226 # postfix repeat count 1227 $env = pack( '(S/A* S/A*)' . @Env/2, @Env ); 1228 1229 # warn when count/code goes beyond end of string 1230 # \0001 a \0003 AAA \0001 b \0003 BBB 1231 # 2 3c 5 8 10 11 13 16 1232 eval { @pup = unpack( '(S/A* S/A*)' . @Env/2, substr( $env, 0, 11 ) ) }; 1233 like( $@, qr{length/code after end of string} ); 1234 1235 # catch stack overflow/segfault 1236 eval { $_ = pack( ('(' x 105) . 'A' . (')' x 105) ); }; 1237 like( $@, qr{Too deeply nested \(\)-groups} ); 1238} 1239 1240{ # syntax checks (W.Laun) 1241 use warnings qw(NONFATAL all);; 1242 my @warning; 1243 local $SIG{__WARN__} = sub { 1244 push( @warning, $_[0] ); 1245 }; 1246 eval { my $s = pack( 'Ax![4c]A', 1..5 ); }; 1247 like( $@, qr{Malformed integer in \[\]} ); 1248 1249 eval { my $buf = pack( '(c/*a*)', 'AAA', 'BB' ); }; 1250 like( $@, qr{'/' does not take a repeat count} ); 1251 1252 eval { my @inf = unpack( 'c/1a', "\x03AAA\x02BB" ); }; 1253 like( $@, qr{'/' does not take a repeat count} ); 1254 1255 eval { my @inf = unpack( 'c/*a', "\x03AAA\x02BB" ); }; 1256 like( $@, qr{'/' does not take a repeat count} ); 1257 1258 # white space where possible 1259 my @Env = ( a => 'AAA', b => 'BBB' ); 1260 my $env = pack( ' S ( S / A* S / A* )* ', @Env/2, @Env ); 1261 my @pup = unpack( ' S / ( S / A* S / A* ) ', $env ); 1262 is( "@pup", "@Env" ); 1263 1264 # white space in 4 wrong places 1265 for my $temp ( 'A ![4]', 'A [4]', 'A *', 'A 4' ){ 1266 eval { my $s = pack( $temp, 'B' ); }; 1267 like( $@, qr{Invalid type } ); 1268 } 1269 1270 # warning for commas 1271 @warning = (); 1272 my $x = pack( 'I,A', 4, 'X' ); 1273 like( $warning[0], qr{Invalid type ','} ); 1274 1275 # comma warning only once 1276 @warning = (); 1277 $x = pack( 'C(C,C)C,C', 65..71 ); 1278 like( scalar @warning, 1 ); 1279 1280 # forbidden code in [] 1281 eval { my $x = pack( 'A[@4]', 'XXXX' ); }; 1282 like( $@, qr{Within \[\]-length '\@' not allowed} ); 1283 1284 # @ repeat default 1 1285 my $s = pack( 'AA@A', 'A', 'B', 'C' ); 1286 my @c = unpack( 'AA@A', $s ); 1287 is( $s, 'AC' ); 1288 is( "@c", "A C C" ); 1289 1290 # no unpack code after / 1291 eval { my @a = unpack( "C/", "\3" ); }; 1292 like( $@, qr{Code missing after '/'} ); 1293 1294 SKIP: { 1295 skip $no_endianness, 6 if $no_endianness; 1296 1297 # modifier warnings 1298 @warning = (); 1299 $x = pack "I>>s!!", 47, 11; 1300 ($x) = unpack "I<<l!>!>", 'x'x20; 1301 is(scalar @warning, 5); 1302 like($warning[0], qr/Duplicate modifier '>' after 'I' in pack/); 1303 like($warning[1], qr/Duplicate modifier '!' after 's' in pack/); 1304 like($warning[2], qr/Duplicate modifier '<' after 'I' in unpack/); 1305 like($warning[3], qr/Duplicate modifier '!' after 'l' in unpack/); 1306 like($warning[4], qr/Duplicate modifier '>' after 'l' in unpack/); 1307 } 1308} 1309 1310{ # Repeat count [SUBEXPR] 1311 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 1312 s! S! i! I! l! L! j J); 1313 my $G; 1314 if (eval { pack 'q', 1 } ) { 1315 push @codes, qw(q Q); 1316 } else { 1317 push @codes, qw(s S); # Keep the count the same 1318 } 1319 if (eval { pack 'D', 1 } ) { 1320 push @codes, 'D'; 1321 } else { 1322 push @codes, 'd'; # Keep the count the same 1323 } 1324 1325 push @codes, map { /^[silqjfdp]/i ? ("$_<", "$_>") : () } @codes; 1326 1327 my %val; 1328 @val{@codes} = map { / [Xx] (?{ undef }) 1329 | [AZa] (?{ 'something' }) 1330 | C (?{ 214 }) 1331 | W (?{ 8188 }) 1332 | c (?{ 114 }) 1333 | [Bb] (?{ '101' }) 1334 | [Hh] (?{ 'b8' }) 1335 | [svnSiIlVNLqQjJ] (?{ 10111 }) 1336 | [FfDd] (?{ 1.36514538e67 }) 1337 | [pP] (?{ "try this buffer" }) 1338 /x; $^R } @codes; 1339 my @end = (0x12345678, 0x23456781, 0x35465768, 0x15263748); 1340 my $end = "N4"; 1341 1342 for my $type (@codes) { 1343 my @list = $val{$type}; 1344 @list = () unless defined $list[0]; 1345 for my $count ('', '3', '[11]') { 1346 my $c = 1; 1347 $c = $1 if $count =~ /(\d+)/; 1348 my @list1 = @list; 1349 @list1 = (@list1) x $c unless $type =~ /[XxAaZBbHhP]/; 1350 for my $groupend ('', ')2', ')[8]') { 1351 my $groupbegin = ($groupend ? '(' : ''); 1352 $c = 1; 1353 $c = $1 if $groupend =~ /(\d+)/; 1354 my @list2 = (@list1) x $c; 1355 1356 SKIP: { 1357 my $junk1 = "$groupbegin $type$count $groupend"; 1358 # print "# junk1=$junk1\n"; 1359 my $p = eval { pack $junk1, @list2 }; 1360 skip "cannot pack '$type' on this perl", 12 1361 if is_valid_error($@); 1362 die "pack $junk1 failed: $@" if $@; 1363 1364 my $half = int( (length $p)/2 ); 1365 for my $move ('', "X$half", "X!$half", 'x1', 'x!8', "x$half") { 1366 my $junk = "$junk1 $move"; 1367 # print "# junk='$junk', list=(@list2)\n"; 1368 $p = pack "$junk $end", @list2, @end; 1369 my @l = unpack "x[$junk] $end", $p; 1370 is(scalar @l, scalar @end); 1371 is("@l", "@end", "skipping x[$junk]"); 1372 } 1373 } 1374 } 1375 } 1376 } 1377} 1378 1379# / is recognized after spaces in scalar context 1380# XXXX no spaces are allowed in pack... In pack only before the slash... 1381is(scalar unpack('A /A Z20', pack 'A/A* Z20', 'bcde', 'xxxxx'), 'bcde'); 1382is(scalar unpack('A /A /A Z20', '3004bcde'), 'bcde'); 1383 1384{ # X! and x! 1385 my $t = 'C[3] x!8 C[2]'; 1386 my @a = (0x73..0x77); 1387 my $p = pack($t, @a); 1388 is($p, "\x73\x74\x75\0\0\0\0\0\x76\x77"); 1389 my @b = unpack $t, $p; 1390 is(scalar @b, scalar @a); 1391 is("@b", "@a", 'x!8'); 1392 $t = 'x[5] C[6] X!8 C[2]'; 1393 @a = (0x73..0x7a); 1394 $p = pack($t, @a); 1395 is($p, "\0\0\0\0\0\x73\x74\x75\x79\x7a"); 1396 @b = unpack $t, $p; 1397 @a = (0x73..0x75, 0x79, 0x7a, 0x79, 0x7a); 1398 is(scalar @b, scalar @a); 1399 is("@b", "@a"); 1400} 1401 1402{ # struct {char c1; double d; char cc[2];} 1403 my $t = 'C x![d] d C[2]'; 1404 my @a = (173, 1.283476517e-45, 42, 215); 1405 my $p = pack $t, @a; 1406 ok( length $p); 1407 my @b = unpack "$t X[$t] $t", $p; # Extract, step back, extract again 1408 is(scalar @b, 2 * scalar @a); 1409 $b = "@b"; 1410 $b =~ s/(?:17000+|16999+)\d+(e-45) /17$1 /gi; # stringification is gamble 1411 is($b, "@a @a"); 1412 1413 use warnings qw(NONFATAL all);; 1414 my $warning; 1415 local $SIG{__WARN__} = sub { 1416 $warning = $_[0]; 1417 }; 1418 @b = unpack "x[C] x[$t] X[$t] X[C] $t", "$p\0"; 1419 1420 is($warning, undef); 1421 is(scalar @b, scalar @a); 1422 $b = "@b"; 1423 $b =~ s/(?:17000+|16999+)\d+(e-45) /17$1 /gi; # stringification is gamble 1424 is($b, "@a"); 1425} 1426 1427is(length(pack("j", 0)), $Config{ivsize}); 1428is(length(pack("J", 0)), $Config{uvsize}); 1429is(length(pack("F", 0)), $Config{nvsize}); 1430 1431numbers ('j', -2147483648, -1, 0, 1, 2147483647); 1432numbers ('J', 0, 1, 2147483647, 2147483648, 4294967295); 1433numbers ('F', -(2**34), -1, 0, 1, 2**34); 1434SKIP: { 1435 my $t = eval { unpack("D*", pack("D", 12.34)) }; 1436 1437 skip "Long doubles not in use", 166 if $@ =~ /Invalid type/; 1438 1439 is(length(pack("D", 0)), $Config{longdblsize}); 1440 numbers ('D', -(2**34), -1, 0, 1, 2**34); 1441} 1442 1443# Maybe this knowledge needs to be "global" for all of pack.t 1444# Or a "can checksum" which would effectively be all the number types" 1445my %cant_checksum = map {$_=> 1} qw(A Z u w); 1446# not a b B h H 1447foreach 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)) { 1448 SKIP: { 1449 my $packed = eval {pack "${template}4", 1, 4, 9, 16}; 1450 if ($@) { 1451 die unless $@ =~ /Invalid type '$template'/; 1452 skip ("$template not supported on this perl", 1453 $cant_checksum{$template} ? 4 : 8); 1454 } 1455 my @unpack4 = unpack "${template}4", $packed; 1456 my @unpack = unpack "${template}*", $packed; 1457 my @unpack1 = unpack "${template}", $packed; 1458 my @unpack1s = scalar unpack "${template}", $packed; 1459 my @unpack4s = scalar unpack "${template}4", $packed; 1460 my @unpacks = scalar unpack "${template}*", $packed; 1461 1462 my @tests = ( ["${template}4 vs ${template}*", \@unpack4, \@unpack], 1463 ["scalar ${template} ${template}", \@unpack1s, \@unpack1], 1464 ["scalar ${template}4 vs ${template}", \@unpack4s, \@unpack1], 1465 ["scalar ${template}* vs ${template}", \@unpacks, \@unpack1], 1466 ); 1467 1468 unless ($cant_checksum{$template}) { 1469 my @unpack4_c = unpack "\%${template}4", $packed; 1470 my @unpack_c = unpack "\%${template}*", $packed; 1471 my @unpack1_c = unpack "\%${template}", $packed; 1472 my @unpack1s_c = scalar unpack "\%${template}", $packed; 1473 my @unpack4s_c = scalar unpack "\%${template}4", $packed; 1474 my @unpacks_c = scalar unpack "\%${template}*", $packed; 1475 1476 push @tests, 1477 ( ["% ${template}4 vs ${template}*", \@unpack4_c, \@unpack_c], 1478 ["% scalar ${template} ${template}", \@unpack1s_c, \@unpack1_c], 1479 ["% scalar ${template}4 vs ${template}*", \@unpack4s_c, \@unpack_c], 1480 ["% scalar ${template}* vs ${template}*", \@unpacks_c, \@unpack_c], 1481 ); 1482 } 1483 foreach my $test (@tests) { 1484 ok (list_eq ($test->[1], $test->[2]), $test->[0]) || 1485 printf "# unpack gave %s expected %s\n", 1486 encode_list (@{$test->[1]}), encode_list (@{$test->[2]}); 1487 } 1488 } 1489} 1490 1491ok(pack('u2', 'AA'), "[perl #8026]"); # used to hang and eat RAM in perl 5.7.2 1492 1493$_ = pack('c', 65); # 'A' would not be EBCDIC-friendly 1494is(unpack('c'), 65, "one-arg unpack (change #18751)"); # defaulting to $_ 1495 1496{ 1497 my $a = "X\x0901234567\n" x 100; # \t would not be EBCDIC TAB 1498 my @a = unpack("(a1 c/a)*", $a); 1499 is(scalar @a, 200, "[perl #15288]"); 1500 is($a[-1], "01234567\n", "[perl #15288]"); 1501 is($a[-2], "X", "[perl #15288]"); 1502} 1503 1504{ 1505 use warnings qw(NONFATAL all);; 1506 my $warning; 1507 local $SIG{__WARN__} = sub { 1508 $warning = $_[0]; 1509 }; 1510 my $out = pack("u99", "foo" x 99); 1511 like($warning, qr/Field too wide in 'u' format in pack at /, 1512 "Warn about too wide uuencode"); 1513 is($out, ("_" . "9F]O" x 21 . "\n") x 4 . "M" . "9F]O" x 15 . "\n", 1514 "Use max width in case of too wide uuencode"); 1515} 1516 1517# checksums 1518{ 1519 # verify that unpack advances correctly wrt a checksum 1520 my (@x) = unpack("b10a", "abcd"); 1521 my (@y) = unpack("%b10a", "abcd"); 1522 is($x[1], $y[1], "checksum advance ok"); 1523 1524 # verify that the checksum is not overflowed with C0 1525 if (ord('A') == 193) { 1526 is(unpack("C0%128U", "/bcd"), unpack("U0%128U", "abcd"), "checksum not overflowed"); 1527 } else { 1528 is(unpack("C0%128U", "abcd"), unpack("U0%128U", "abcd"), "checksum not overflowed"); 1529 } 1530} 1531 1532{ 1533 # U0 and C0 must be scoped 1534 my (@x) = unpack("a(U0)U", "b\341\277\274"); 1535 is($x[0], 'b', 'before scope'); 1536 is($x[1], 8188, 'after scope'); 1537 1538 is(pack("a(U0)U", "b", 8188), "b\341\277\274"); 1539} 1540 1541{ 1542 # counted length prefixes shouldn't change C0/U0 mode 1543 # (note the length is actually 0 in this test) 1544 if (ord('A') == 193) { 1545 is(join(',', unpack("aU0C/UU", "b\0\341\277\274")), 'b,0'); 1546 is(join(',', unpack("aU0C/CU", "b\0\341\277\274")), 'b,0'); 1547 } else { 1548 is(join(',', unpack("aC/UU", "b\0\341\277\274")), 'b,8188'); 1549 is(join(',', unpack("aC/CU", "b\0\341\277\274")), 'b,8188'); 1550 is(join(',', unpack("aU0C/UU", "b\0\341\277\274")), 'b,225'); 1551 is(join(',', unpack("aU0C/CU", "b\0\341\277\274")), 'b,225'); 1552 } 1553} 1554 1555{ 1556 # "Z0" (bug #34062) 1557 my (@x) = unpack("C*", pack("CZ0", 1, "b")); 1558 is(join(',', @x), '1', 'pack Z0 doesn\'t destroy the character before'); 1559} 1560 1561{ 1562 # Encoding neutrality 1563 # String we will pull apart and rebuild in several ways: 1564 my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1565 my $up = $down; 1566 utf8::upgrade($up); 1567 1568 my %expect = 1569 # [expected result, 1570 # how many chars it should progress, 1571 # (optional) expected result of pack] 1572 (a5 => ["\xf8\xf9\xfa\xfb\xfc", 5], 1573 A5 => ["\xf8\xf9\xfa\xfb\xfc", 5], 1574 Z5 => ["\xf8\xf9\xfa\xfb\xfc", 5, "\xf8\xf9\xfa\xfb\x00\xfd"], 1575 b21 => ["000111111001111101011", 3, "\xf8\xf9\x1a\xfb"], 1576 B21 => ["111110001111100111111", 3, "\xf8\xf9\xf8\xfb"], 1577 H5 => ["f8f9f", 3, "\xf8\xf9\xf0\xfb"], 1578 h5 => ["8f9fa", 3, "\xf8\xf9\x0a\xfb"], 1579 "s<" => [-1544, 2], 1580 "s>" => [-1799, 2], 1581 "S<" => [0xf9f8, 2], 1582 "S>" => [0xf8f9, 2], 1583 "l<" => [-67438088, 4], 1584 "l>" => [-117835013, 4], 1585 "L>" => [0xf8f9fafb, 4], 1586 "L<" => [0xfbfaf9f8, 4], 1587 n => [0xf8f9, 2], 1588 N => [0xf8f9fafb, 4], 1589 v => [63992, 2], 1590 V => [0xfbfaf9f8, 4], 1591 c => [-8, 1], 1592 U0U => [0xf8, 1], 1593 w => ["8715569050387726213", 9], 1594 q => ["-283686952306184", 8], 1595 Q => ["18446460386757245432", 8], 1596 ); 1597 1598 for my $string ($down, $up) { 1599 for my $format (sort {lc($a) cmp lc($b) || $a cmp $b } keys %expect) { 1600 SKIP: { 1601 my $expect = $expect{$format}; 1602 # unpack upgraded and downgraded string 1603 my @result = eval { unpack("$format C0 W", $string) }; 1604 skip "cannot pack/unpack '$format C0 W' on this perl", 5 if 1605 $@ && is_valid_error($@); 1606 is(@result, 2, "Two results from unpack $format C0 W"); 1607 1608 # pack to downgraded 1609 my $new = pack("$format C0 W", @result); 1610 is(length($new), $expect->[1]+1, 1611 "pack $format C0 W should give $expect->[1]+1 chars"); 1612 is($new, $expect->[2] || substr($string, 0, length $new), 1613 "pack $format C0 W returns expected value"); 1614 1615 # pack to upgraded 1616 $new = pack("a0 $format C0 W", chr(256), @result); 1617 is(length($new), $expect->[1]+1, 1618 "pack a0 $format C0 W should give $expect->[1]+1 chars"); 1619 is($new, $expect->[2] || substr($string, 0, length $new), 1620 "pack a0 $format C0 W returns expected value"); 1621 } 1622 } 1623 } 1624} 1625 1626{ 1627 # Encoding neutrality, numbers 1628 my $val = -2.68; 1629 for my $format (qw(s S i I l L j J f d F D q Q 1630 s! S! i! I! l! L! n! N! v! V!)) { 1631 SKIP: { 1632 my $down = eval { pack($format, $val) }; 1633 skip "cannot pack/unpack $format on this perl", 9 if 1634 $@ && is_valid_error($@); 1635 ok(!utf8::is_utf8($down), "Simple $format pack doesn't get upgraded"); 1636 my $up = pack("a0 $format", chr(256), $val); 1637 ok(utf8::is_utf8($up), "a0 $format with high char leads to upgrade"); 1638 is($down, $up, "$format generated strings are equal though"); 1639 my @down_expanded = unpack("$format W", $down . chr(0xce)); 1640 is(@down_expanded, 2, "Expand to two values"); 1641 is($down_expanded[1], 0xce, 1642 "unpack $format left us at the expected position"); 1643 my @up_expanded = unpack("$format W", $up . chr(0xce)); 1644 is(@up_expanded, 2, "Expand to two values"); 1645 is($up_expanded[1], 0xce, 1646 "unpack $format left us at the expected position"); 1647 is($down_expanded[0], $up_expanded[0], "$format unpack was neutral"); 1648 is(pack($format, $down_expanded[0]), $down, "Pack $format undoes unpack $format"); 1649 } 1650 } 1651} 1652 1653{ 1654 # C *is* neutral 1655 my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1656 my $up = $down; 1657 utf8::upgrade($up); 1658 my @down = unpack("C*", $down); 1659 my @expect_down = (0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, 0x05, 0x06); 1660 is("@down", "@expect_down", "byte expand"); 1661 is(pack("C*", @down), $down, "byte join"); 1662 1663 my @up = unpack("C*", $up); 1664 my @expect_up = (0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff, 0x05, 0x06); 1665 is("@up", "@expect_up", "UTF-8 expand"); 1666 is(pack("U0C0C*", @up), $up, "UTF-8 join"); 1667} 1668 1669{ 1670 # Harder cases for the neutrality test 1671 1672 # u format 1673 my $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1674 my $up = $down; 1675 utf8::upgrade($up); 1676 is(pack("u", $down), pack("u", $up), "u pack is neutral"); 1677 is(unpack("u", pack("u", $down)), $down, "u unpack to downgraded works"); 1678 is(unpack("U0C0u", pack("u", $down)), $up, "u unpack to upgraded works"); 1679 1680 # p/P format 1681 # This actually only tests something if the address contains a byte >= 0x80 1682 my $str = "abc\xa5\x00\xfede"; 1683 $down = pack("p", $str); 1684 is(pack("P", $str), $down); 1685 is(pack("U0C0p", $str), $down); 1686 is(pack("U0C0P", $str), $down); 1687 is(unpack("p", $down), "abc\xa5", "unpack p downgraded"); 1688 $up = $down; 1689 utf8::upgrade($up); 1690 is(unpack("p", $up), "abc\xa5", "unpack p upgraded"); 1691 1692 is(unpack("P7", $down), "abc\xa5\x00\xfed", "unpack P downgraded"); 1693 is(unpack("P7", $up), "abc\xa5\x00\xfed", "unpack P upgraded"); 1694 1695 # x, X and @ 1696 $down = "\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff\x05\x06"; 1697 $up = $down; 1698 utf8::upgrade($up); 1699 1700 is(unpack('@4W', $down), 0xfc, "\@positioning on downgraded string"); 1701 is(unpack('@4W', $up), 0xfc, "\@positioning on upgraded string"); 1702 1703 is(unpack('@4x2W', $down), 0xfe, "x moving on downgraded string"); 1704 is(unpack('@4x2W', $up), 0xfe, "x moving on upgraded string"); 1705 is(unpack('@4x!4W', $down), 0xfc, "x! moving on downgraded string"); 1706 is(unpack('@4x!4W', $up), 0xfc, "x! moving on upgraded string"); 1707 is(unpack('@5x!4W', $down), 0x05, "x! moving on downgraded string"); 1708 is(unpack('@5x!4W', $up), 0x05, "x! moving on upgraded string"); 1709 1710 is(unpack('@4X2W', $down), 0xfa, "X moving on downgraded string"); 1711 is(unpack('@4X2W', $up), 0xfa, "X moving on upgraded string"); 1712 is(unpack('@4X!4W', $down), 0xfc, "X! moving on downgraded string"); 1713 is(unpack('@4X!4W', $up), 0xfc, "X! moving on upgraded string"); 1714 is(unpack('@5X!4W', $down), 0xfc, "X! moving on downgraded string"); 1715 is(unpack('@5X!4W', $up), 0xfc, "X! moving on upgraded string"); 1716 is(unpack('@5X!8W', $down), 0xf8, "X! moving on downgraded string"); 1717 is(unpack('@5X!8W', $up), 0xf8, "X! moving on upgraded string"); 1718 1719 is(pack("W2x", 0xfa, 0xe3), "\xfa\xe3\x00", "x on downgraded string"); 1720 is(pack("W2x!4", 0xfa, 0xe3), "\xfa\xe3\x00\x00", 1721 "x! on downgraded string"); 1722 is(pack("W2x!2", 0xfa, 0xe3), "\xfa\xe3", "x! on downgraded string"); 1723 is(pack("U0C0W2x", 0xfa, 0xe3), "\xfa\xe3\x00", "x on upgraded string"); 1724 is(pack("U0C0W2x!4", 0xfa, 0xe3), "\xfa\xe3\x00\x00", 1725 "x! on upgraded string"); 1726 is(pack("U0C0W2x!2", 0xfa, 0xe3), "\xfa\xe3", "x! on upgraded string"); 1727 is(pack("W2X", 0xfa, 0xe3), "\xfa", "X on downgraded string"); 1728 is(pack("U0C0W2X", 0xfa, 0xe3), "\xfa", "X on upgraded string"); 1729 is(pack("W2X!2", 0xfa, 0xe3), "\xfa\xe3", "X! on downgraded string"); 1730 is(pack("U0C0W2X!2", 0xfa, 0xe3), "\xfa\xe3", "X! on upgraded string"); 1731 is(pack("W3X!2", 0xfa, 0xe3, 0xa6), "\xfa\xe3", "X! on downgraded string"); 1732 is(pack("U0C0W3X!2", 0xfa, 0xe3, 0xa6), "\xfa\xe3", 1733 "X! on upgraded string"); 1734 1735 # backward eating through a ( moves the group starting point backwards 1736 is(pack("a*(Xa)", "abc", "q"), "abq", 1737 "eating before strbeg moves it back"); 1738 is(pack("a*(Xa)", "ab" . chr(512), "q"), "abq", 1739 "eating before strbeg moves it back"); 1740 1741 # Check marked_upgrade 1742 is(pack('W(W(Wa@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, "a", 0xa4, 0xa5, 0xa6), 1743 "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6"); 1744 $up = "a"; 1745 utf8::upgrade($up); 1746 is(pack('W(W(Wa@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, $up, 0xa4, 0xa5, 0xa6), 1747 "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6", "marked upgrade caused by a"); 1748 is(pack('W(W(WW@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, 256, 0xa4, 0xa5, 0xa6), 1749 "\xa1\xa2\xa3\x{100}\x00\xa4\x00\xa5\x00\xa6", 1750 "marked upgrade caused by W"); 1751 is(pack('W(W(WU0aC0@3W)@6W)@9W', 0xa1, 0xa2, 0xa3, "a", 0xa4, 0xa5, 0xa6), 1752 "\xa1\xa2\xa3a\x00\xa4\x00\xa5\x00\xa6", "marked upgrade caused by U0"); 1753 1754 # a, A and Z 1755 $down = "\xa4\xa6\xa7"; 1756 $up = $down; 1757 utf8::upgrade($up); 1758 utf8::upgrade(my $high = "\xfeb"); 1759 1760 for my $format ("a0", "A0", "Z0", "U0a0C0", "U0A0C0", "U0Z0C0") { 1761 is(pack("a* $format a*", "ab", $down, "cd"), "abcd", 1762 "$format format on plain string"); 1763 is(pack("a* $format a*", "ab", $up, "cd"), "abcd", 1764 "$format format on upgraded string"); 1765 is(pack("a* $format a*", $high, $down, "cd"), "\xfebcd", 1766 "$format format on plain string"); 1767 is(pack("a* $format a*", $high, $up, "cd"), "\xfebcd", 1768 "$format format on upgraded string"); 1769 my @down = unpack("a1 $format a*", "\xfeb"); 1770 is("@down", "\xfe b", "unpack $format"); 1771 my @up = unpack("a1 $format a*", $high); 1772 is("@up", "\xfe b", "unpack $format"); 1773 } 1774 is(pack("a1", $high), "\xfe"); 1775 is(pack("A1", $high), "\xfe"); 1776 is(pack("Z1", $high), "\x00"); 1777 is(pack("a2", $high), "\xfeb"); 1778 is(pack("A2", $high), "\xfeb"); 1779 is(pack("Z2", $high), "\xfe\x00"); 1780 is(pack("a5", $high), "\xfeb\x00\x00\x00"); 1781 is(pack("A5", $high), "\xfeb "); 1782 is(pack("Z5", $high), "\xfeb\x00\x00\x00"); 1783 is(pack("a*", $high), "\xfeb"); 1784 is(pack("A*", $high), "\xfeb"); 1785 is(pack("Z*", $high), "\xfeb\x00"); 1786 1787 utf8::upgrade($high = "\xc3\xbeb"); 1788 is(pack("U0a2", $high), "\xfe"); 1789 is(pack("U0A2", $high), "\xfe"); 1790 is(pack("U0Z1", $high), "\x00"); 1791 is(pack("U0a3", $high), "\xfeb"); 1792 is(pack("U0A3", $high), "\xfeb"); 1793 is(pack("U0Z3", $high), "\xfe\x00"); 1794 is(pack("U0a6", $high), "\xfeb\x00\x00\x00"); 1795 is(pack("U0A6", $high), "\xfeb "); 1796 is(pack("U0Z6", $high), "\xfeb\x00\x00\x00"); 1797 is(pack("U0a*", $high), "\xfeb"); 1798 is(pack("U0A*", $high), "\xfeb"); 1799 is(pack("U0Z*", $high), "\xfeb\x00"); 1800} 1801{ 1802 # pack / 1803 my @array = 1..14; 1804 my @out = unpack("N/S", pack("N/S", @array) . "abcd"); 1805 is("@out", "@array", "pack N/S works"); 1806 @out = unpack("N/S*", pack("N/S*", @array) . "abcd"); 1807 is("@out", "@array", "pack N/S* works"); 1808 @out = unpack("N/S*", pack("N/S14", @array) . "abcd"); 1809 is("@out", "@array", "pack N/S14 works"); 1810 @out = unpack("N/S*", pack("N/S15", @array) . "abcd"); 1811 is("@out", "@array", "pack N/S15 works"); 1812 @out = unpack("N/S*", pack("N/S13", @array) . "abcd"); 1813 is("@out", "@array[0..12]", "pack N/S13 works"); 1814 @out = unpack("N/S*", pack("N/S0", @array) . "abcd"); 1815 is("@out", "", "pack N/S0 works"); 1816 is(pack("Z*/a0", "abc"), "0\0", "pack Z*/a0 makes a short string"); 1817 is(pack("Z*/Z0", "abc"), "0\0", "pack Z*/Z0 makes a short string"); 1818 is(pack("Z*/a3", "abc"), "3\0abc", "pack Z*/a3 makes a full string"); 1819 is(pack("Z*/Z3", "abc"), "3\0ab\0", "pack Z*/Z3 makes a short string"); 1820 is(pack("Z*/a5", "abc"), "5\0abc\0\0", "pack Z*/a5 makes a long string"); 1821 is(pack("Z*/Z5", "abc"), "5\0abc\0\0", "pack Z*/Z5 makes a long string"); 1822 is(pack("Z*/Z"), "1\0\0", "pack Z*/Z makes an extended string"); 1823 is(pack("Z*/Z", ""), "1\0\0", "pack Z*/Z makes an extended string"); 1824 is(pack("Z*/a", ""), "0\0", "pack Z*/a makes an extended string"); 1825} 1826{ 1827 # unpack("A*", $unicode) strips general unicode spaces 1828 is(unpack("A*", "ab \n\xa0 \0"), "ab \n\xa0", 1829 'normal A* strip leaves \xa0'); 1830 is(unpack("U0C0A*", "ab \n\xa0 \0"), "ab \n\xa0", 1831 'normal A* strip leaves \xa0 even if it got upgraded for technical reasons'); 1832 is(unpack("A*", pack("a*(U0U)a*", "ab \n", 0xa0, " \0")), "ab", 1833 'upgraded strings A* removes \xa0'); 1834 is(unpack("A*", pack("a*(U0UU)a*", "ab \n", 0xa0, 0x1680, " \0")), "ab", 1835 'upgraded strings A* removes all unicode whitespace'); 1836 is(unpack("A5", pack("a*(U0U)a*", "ab \n", 0x1680, "def", "ab")), "ab", 1837 'upgraded strings A5 removes all unicode whitespace'); 1838 is(unpack("A*", pack("U", 0x1680)), "", 1839 'upgraded strings A* with nothing left'); 1840} 1841{ 1842 # Testing unpack . and .! 1843 is(unpack(".", "ABCD"), 0, "offset at start of string is 0"); 1844 is(unpack(".", ""), 0, "offset at start of empty string is 0"); 1845 is(unpack("x3.", "ABCDEF"), 3, "simple offset works"); 1846 is(unpack("x3.", "ABC"), 3, "simple offset at end of string works"); 1847 is(unpack("x3.0", "ABC"), 0, "self offset is 0"); 1848 is(unpack("x3(x2.)", "ABCDEF"), 2, "offset is relative to inner group"); 1849 is(unpack("x3(X2.)", "ABCDEF"), -2, 1850 "negative offset relative to inner group"); 1851 is(unpack("x3(X2.2)", "ABCDEF"), 1, "offset is relative to inner group"); 1852 is(unpack("x3(x2.0)", "ABCDEF"), 0, "self offset in group is still 0"); 1853 is(unpack("x3(x2.2)", "ABCDEF"), 5, "offset counts groups"); 1854 is(unpack("x3(x2.*)", "ABCDEF"), 5, "star offset is relative to start"); 1855 1856 my $high = chr(8188) x 6; 1857 is(unpack("x3(x2.)", $high), 2, "utf8 offset is relative to inner group"); 1858 is(unpack("x3(X2.)", $high), -2, 1859 "utf8 negative offset relative to inner group"); 1860 is(unpack("x3(X2.2)", $high), 1, "utf8 offset counts groups"); 1861 is(unpack("x3(x2.0)", $high), 0, "utf8 self offset in group is still 0"); 1862 is(unpack("x3(x2.2)", $high), 5, "utf8 offset counts groups"); 1863 is(unpack("x3(x2.*)", $high), 5, "utf8 star offset is relative to start"); 1864 1865 is(unpack("U0x3(x2.)", $high), 2, 1866 "U0 mode utf8 offset is relative to inner group"); 1867 is(unpack("U0x3(X2.)", $high), -2, 1868 "U0 mode utf8 negative offset relative to inner group"); 1869 is(unpack("U0x3(X2.2)", $high), 1, 1870 "U0 mode utf8 offset counts groups"); 1871 is(unpack("U0x3(x2.0)", $high), 0, 1872 "U0 mode utf8 self offset in group is still 0"); 1873 is(unpack("U0x3(x2.2)", $high), 5, 1874 "U0 mode utf8 offset counts groups"); 1875 is(unpack("U0x3(x2.*)", $high), 5, 1876 "U0 mode utf8 star offset is relative to start"); 1877 1878 is(unpack("x3(x2.!)", $high), 2*3, 1879 "utf8 offset is relative to inner group"); 1880 is(unpack("x3(X2.!)", $high), -2*3, 1881 "utf8 negative offset relative to inner group"); 1882 is(unpack("x3(X2.!2)", $high), 1*3, 1883 "utf8 offset counts groups"); 1884 is(unpack("x3(x2.!0)", $high), 0, 1885 "utf8 self offset in group is still 0"); 1886 is(unpack("x3(x2.!2)", $high), 5*3, 1887 "utf8 offset counts groups"); 1888 is(unpack("x3(x2.!*)", $high), 5*3, 1889 "utf8 star offset is relative to start"); 1890 1891 is(unpack("U0x3(x2.!)", $high), 2, 1892 "U0 mode utf8 offset is relative to inner group"); 1893 is(unpack("U0x3(X2.!)", $high), -2, 1894 "U0 mode utf8 negative offset relative to inner group"); 1895 is(unpack("U0x3(X2.!2)", $high), 1, 1896 "U0 mode utf8 offset counts groups"); 1897 is(unpack("U0x3(x2.!0)", $high), 0, 1898 "U0 mode utf8 self offset in group is still 0"); 1899 is(unpack("U0x3(x2.!2)", $high), 5, 1900 "U0 mode utf8 offset counts groups"); 1901 is(unpack("U0x3(x2.!*)", $high), 5, 1902 "U0 mode utf8 star offset is relative to start"); 1903} 1904{ 1905 # Testing pack . and .! 1906 is(pack("(a)5 .", 1..5, 3), "123", ". relative to string start, shorten"); 1907 eval { () = pack("(a)5 .", 1..5, -3) }; 1908 like($@, qr{'\.' outside of string in pack}, "Proper error message"); 1909 is(pack("(a)5 .", 1..5, 8), "12345\x00\x00\x00", 1910 ". relative to string start, extend"); 1911 is(pack("(a)5 .", 1..5, 5), "12345", ". relative to string start, keep"); 1912 1913 is(pack("(a)5 .0", 1..5, -3), "12", 1914 ". relative to string current, shorten"); 1915 is(pack("(a)5 .0", 1..5, 2), "12345\x00\x00", 1916 ". relative to string current, extend"); 1917 is(pack("(a)5 .0", 1..5, 0), "12345", 1918 ". relative to string current, keep"); 1919 1920 is(pack("(a)5 (.)", 1..5, -3), "12", 1921 ". relative to group, shorten"); 1922 is(pack("(a)5 (.)", 1..5, 2), "12345\x00\x00", 1923 ". relative to group, extend"); 1924 is(pack("(a)5 (.)", 1..5, 0), "12345", 1925 ". relative to group, keep"); 1926 1927 is(pack("(a)3 ((a)2 .)", 1..5, -2), "1", 1928 ". relative to group, shorten"); 1929 is(pack("(a)3 ((a)2 .)", 1..5, 2), "12345", 1930 ". relative to group, keep"); 1931 is(pack("(a)3 ((a)2 .)", 1..5, 4), "12345\x00\x00", 1932 ". relative to group, extend"); 1933 1934 is(pack("(a)3 ((a)2 .2)", 1..5, 2), "12", 1935 ". relative to counted group, shorten"); 1936 is(pack("(a)3 ((a)2 .2)", 1..5, 7), "12345\x00\x00", 1937 ". relative to counted group, extend"); 1938 is(pack("(a)3 ((a)2 .2)", 1..5, 5), "12345", 1939 ". relative to counted group, keep"); 1940 1941 is(pack("(a)3 ((a)2 .*)", 1..5, 2), "12", 1942 ". relative to start, shorten"); 1943 is(pack("(a)3 ((a)2 .*)", 1..5, 7), "12345\x00\x00", 1944 ". relative to start, extend"); 1945 is(pack("(a)3 ((a)2 .*)", 1..5, 5), "12345", 1946 ". relative to start, keep"); 1947 1948 is(pack('(a)5 (. @2 a)', 1..5, -3, "a"), "12\x00\x00a", 1949 ". based shrink properly updates group starts"); 1950 1951 is(pack("(W)3 ((W)2 .)", 0x301..0x305, -2), "\x{301}", 1952 "utf8 . relative to group, shorten"); 1953 is(pack("(W)3 ((W)2 .)", 0x301..0x305, 2), 1954 "\x{301}\x{302}\x{303}\x{304}\x{305}", 1955 "utf8 . relative to group, keep"); 1956 is(pack("(W)3 ((W)2 .)", 0x301..0x305, 4), 1957 "\x{301}\x{302}\x{303}\x{304}\x{305}\x00\x00", 1958 "utf8 . relative to group, extend"); 1959 1960 is(pack("(W)3 ((W)2 .!)", 0x301..0x305, -2), "\x{301}\x{302}", 1961 "utf8 . relative to group, shorten"); 1962 is(pack("(W)3 ((W)2 .!)", 0x301..0x305, 4), 1963 "\x{301}\x{302}\x{303}\x{304}\x{305}", 1964 "utf8 . relative to group, keep"); 1965 is(pack("(W)3 ((W)2 .!)", 0x301..0x305, 6), 1966 "\x{301}\x{302}\x{303}\x{304}\x{305}\x00\x00", 1967 "utf8 . relative to group, extend"); 1968 1969 is(pack('(W)5 (. @2 a)', 0x301..0x305, -3, "a"), 1970 "\x{301}\x{302}\x00\x00a", 1971 "utf8 . based shrink properly updates group starts"); 1972} 1973{ 1974 # Testing @! 1975 is(pack('a* @3', "abcde"), "abc", 'Test basic @'); 1976 is(pack('a* @!3', "abcde"), "abc", 'Test basic @!'); 1977 is(pack('a* @2', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{301}\x{302}", 1978 'Test basic utf8 @'); 1979 is(pack('a* @!2', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{301}", 1980 'Test basic utf8 @!'); 1981 1982 is(unpack('@4 a*', "abcde"), "e", 'Test basic @'); 1983 is(unpack('@!4 a*', "abcde"), "e", 'Test basic @!'); 1984 is(unpack('@4 a*', "\x{301}\x{302}\x{303}\x{304}\x{305}"), "\x{305}", 1985 'Test basic utf8 @'); 1986 is(unpack('@!4 a*', "\x{301}\x{302}\x{303}\x{304}\x{305}"), 1987 "\x{303}\x{304}\x{305}", 'Test basic utf8 @!'); 1988} 1989{ 1990 #50256 1991 my ($v) = split //, unpack ('(B)*', 'ab'); 1992 is($v, 0); # Doesn't SEGV :-) 1993} 1994{ 1995 #73814 1996 my $x = runperl( prog => 'print split( /,/, unpack(q(%2H*), q(hello world))), qq(\n)' ); 1997 is($x, "0\n", "split /a/, unpack('%2H*'...) didn't crash"); 1998 1999 my $y = runperl( prog => 'print split( /,/, unpack(q(%32u*), q(#,3,Q)), qq(\n)), qq(\n)' ); 2000 is($y, "0\n", "split /a/, unpack('%32u*'...) didn't crash"); 2001} 2002 2003#90160 2004is(eval { () = unpack "C0 U*", ""; "ok" }, "ok", 2005 'medial U* on empty string'); 2006