| 1 | # |
| 2 | # Data/Dumper.pm |
| 3 | # |
| 4 | # convert perl data structures into perl syntax suitable for both printing |
| 5 | # and eval |
| 6 | # |
| 7 | # Documentation at the __END__ |
| 8 | # |
| 9 | |
| 10 | package Data::Dumper; |
| 11 | |
| 12 | $VERSION = $VERSION = '2.101'; |
| 13 | |
| 14 | #$| = 1; |
| 15 | |
| 16 | require 5.004; |
| 17 | require Exporter; |
| 18 | require DynaLoader; |
| 19 | require overload; |
| 20 | |
| 21 | use Carp; |
| 22 | |
| 23 | @ISA = qw(Exporter DynaLoader); |
| 24 | @EXPORT = qw(Dumper); |
| 25 | @EXPORT_OK = qw(DumperX); |
| 26 | |
| 27 | bootstrap Data::Dumper; |
| 28 | |
| 29 | # module vars and their defaults |
| 30 | $Indent = 2 unless defined $Indent; |
| 31 | $Purity = 0 unless defined $Purity; |
| 32 | $Pad = "" unless defined $Pad; |
| 33 | $Varname = "VAR" unless defined $Varname; |
| 34 | $Useqq = 0 unless defined $Useqq; |
| 35 | $Terse = 0 unless defined $Terse; |
| 36 | $Freezer = "" unless defined $Freezer; |
| 37 | $Toaster = "" unless defined $Toaster; |
| 38 | $Deepcopy = 0 unless defined $Deepcopy; |
| 39 | $Quotekeys = 1 unless defined $Quotekeys; |
| 40 | $Bless = "bless" unless defined $Bless; |
| 41 | #$Expdepth = 0 unless defined $Expdepth; |
| 42 | #$Maxdepth = 0 unless defined $Maxdepth; |
| 43 | |
| 44 | # |
| 45 | # expects an arrayref of values to be dumped. |
| 46 | # can optionally pass an arrayref of names for the values. |
| 47 | # names must have leading $ sign stripped. begin the name with * |
| 48 | # to cause output of arrays and hashes rather than refs. |
| 49 | # |
| 50 | sub new { |
| 51 | my($c, $v, $n) = @_; |
| 52 | |
| 53 | croak "Usage: PACKAGE->new(ARRAYREF, [ARRAYREF])" |
| 54 | unless (defined($v) && (ref($v) eq 'ARRAY')); |
| 55 | $n = [] unless (defined($n) && (ref($v) eq 'ARRAY')); |
| 56 | |
| 57 | my($s) = { |
| 58 | level => 0, # current recursive depth |
| 59 | indent => $Indent, # various styles of indenting |
| 60 | pad => $Pad, # all lines prefixed by this string |
| 61 | xpad => "", # padding-per-level |
| 62 | apad => "", # added padding for hash keys n such |
| 63 | sep => "", # list separator |
| 64 | seen => {}, # local (nested) refs (id => [name, val]) |
| 65 | todump => $v, # values to dump [] |
| 66 | names => $n, # optional names for values [] |
| 67 | varname => $Varname, # prefix to use for tagging nameless ones |
| 68 | purity => $Purity, # degree to which output is evalable |
| 69 | useqq => $Useqq, # use "" for strings (backslashitis ensues) |
| 70 | terse => $Terse, # avoid name output (where feasible) |
| 71 | freezer => $Freezer, # name of Freezer method for objects |
| 72 | toaster => $Toaster, # name of method to revive objects |
| 73 | deepcopy => $Deepcopy, # dont cross-ref, except to stop recursion |
| 74 | quotekeys => $Quotekeys, # quote hash keys |
| 75 | 'bless' => $Bless, # keyword to use for "bless" |
| 76 | # expdepth => $Expdepth, # cutoff depth for explicit dumping |
| 77 | # maxdepth => $Maxdepth, # depth beyond which we give up |
| 78 | }; |
| 79 | |
| 80 | if ($Indent > 0) { |
| 81 | $s->{xpad} = " "; |
| 82 | $s->{sep} = "\n"; |
| 83 | } |
| 84 | return bless($s, $c); |
| 85 | } |
| 86 | |
| 87 | # |
| 88 | # add-to or query the table of already seen references |
| 89 | # |
| 90 | sub Seen { |
| 91 | my($s, $g) = @_; |
| 92 | if (defined($g) && (ref($g) eq 'HASH')) { |
| 93 | my($k, $v, $id); |
| 94 | while (($k, $v) = each %$g) { |
| 95 | if (defined $v and ref $v) { |
| 96 | ($id) = (overload::StrVal($v) =~ /\((.*)\)$/); |
| 97 | if ($k =~ /^[*](.*)$/) { |
| 98 | $k = (ref $v eq 'ARRAY') ? ( "\\\@" . $1 ) : |
| 99 | (ref $v eq 'HASH') ? ( "\\\%" . $1 ) : |
| 100 | (ref $v eq 'CODE') ? ( "\\\&" . $1 ) : |
| 101 | ( "\$" . $1 ) ; |
| 102 | } |
| 103 | elsif ($k !~ /^\$/) { |
| 104 | $k = "\$" . $k; |
| 105 | } |
| 106 | $s->{seen}{$id} = [$k, $v]; |
| 107 | } |
| 108 | else { |
| 109 | carp "Only refs supported, ignoring non-ref item \$$k"; |
| 110 | } |
| 111 | } |
| 112 | return $s; |
| 113 | } |
| 114 | else { |
| 115 | return map { @$_ } values %{$s->{seen}}; |
| 116 | } |
| 117 | } |
| 118 | |
| 119 | # |
| 120 | # set or query the values to be dumped |
| 121 | # |
| 122 | sub Values { |
| 123 | my($s, $v) = @_; |
| 124 | if (defined($v) && (ref($v) eq 'ARRAY')) { |
| 125 | $s->{todump} = [@$v]; # make a copy |
| 126 | return $s; |
| 127 | } |
| 128 | else { |
| 129 | return @{$s->{todump}}; |
| 130 | } |
| 131 | } |
| 132 | |
| 133 | # |
| 134 | # set or query the names of the values to be dumped |
| 135 | # |
| 136 | sub Names { |
| 137 | my($s, $n) = @_; |
| 138 | if (defined($n) && (ref($n) eq 'ARRAY')) { |
| 139 | $s->{names} = [@$n]; # make a copy |
| 140 | return $s; |
| 141 | } |
| 142 | else { |
| 143 | return @{$s->{names}}; |
| 144 | } |
| 145 | } |
| 146 | |
| 147 | sub DESTROY {} |
| 148 | |
| 149 | # |
| 150 | # dump the refs in the current dumper object. |
| 151 | # expects same args as new() if called via package name. |
| 152 | # |
| 153 | sub Dump { |
| 154 | my($s) = shift; |
| 155 | my(@out, $val, $name); |
| 156 | my($i) = 0; |
| 157 | local(@post); |
| 158 | |
| 159 | $s = $s->new(@_) unless ref $s; |
| 160 | |
| 161 | for $val (@{$s->{todump}}) { |
| 162 | my $out = ""; |
| 163 | @post = (); |
| 164 | $name = $s->{names}[$i++]; |
| 165 | if (defined $name) { |
| 166 | if ($name =~ /^[*](.*)$/) { |
| 167 | if (defined $val) { |
| 168 | $name = (ref $val eq 'ARRAY') ? ( "\@" . $1 ) : |
| 169 | (ref $val eq 'HASH') ? ( "\%" . $1 ) : |
| 170 | (ref $val eq 'CODE') ? ( "\*" . $1 ) : |
| 171 | ( "\$" . $1 ) ; |
| 172 | } |
| 173 | else { |
| 174 | $name = "\$" . $1; |
| 175 | } |
| 176 | } |
| 177 | elsif ($name !~ /^\$/) { |
| 178 | $name = "\$" . $name; |
| 179 | } |
| 180 | } |
| 181 | else { |
| 182 | $name = "\$" . $s->{varname} . $i; |
| 183 | } |
| 184 | |
| 185 | my $valstr; |
| 186 | { |
| 187 | local($s->{apad}) = $s->{apad}; |
| 188 | $s->{apad} .= ' ' x (length($name) + 3) if $s->{indent} >= 2; |
| 189 | $valstr = $s->_dump($val, $name); |
| 190 | } |
| 191 | |
| 192 | $valstr = "$name = " . $valstr . ';' if @post or !$s->{terse}; |
| 193 | $out .= $s->{pad} . $valstr . $s->{sep}; |
| 194 | $out .= $s->{pad} . join(';' . $s->{sep} . $s->{pad}, @post) |
| 195 | . ';' . $s->{sep} if @post; |
| 196 | |
| 197 | push @out, $out; |
| 198 | } |
| 199 | return wantarray ? @out : join('', @out); |
| 200 | } |
| 201 | |
| 202 | # |
| 203 | # twist, toil and turn; |
| 204 | # and recurse, of course. |
| 205 | # |
| 206 | sub _dump { |
| 207 | my($s, $val, $name) = @_; |
| 208 | my($sname); |
| 209 | my($out, $realpack, $realtype, $type, $ipad, $id, $blesspad); |
| 210 | |
| 211 | $type = ref $val; |
| 212 | $out = ""; |
| 213 | |
| 214 | if ($type) { |
| 215 | |
| 216 | # prep it, if it looks like an object |
| 217 | if ($type =~ /[a-z_:]/) { |
| 218 | my $freezer = $s->{freezer}; |
| 219 | $val->$freezer() if $freezer && UNIVERSAL::can($val, $freezer); |
| 220 | } |
| 221 | |
| 222 | ($realpack, $realtype, $id) = |
| 223 | (overload::StrVal($val) =~ /^(?:(.*)\=)?([^=]*)\(([^\(]*)\)$/); |
| 224 | |
| 225 | # if it has a name, we need to either look it up, or keep a tab |
| 226 | # on it so we know when we hit it later |
| 227 | if (defined($name) and length($name)) { |
| 228 | # keep a tab on it so that we dont fall into recursive pit |
| 229 | if (exists $s->{seen}{$id}) { |
| 230 | # if ($s->{expdepth} < $s->{level}) { |
| 231 | if ($s->{purity} and $s->{level} > 0) { |
| 232 | $out = ($realtype eq 'HASH') ? '{}' : |
| 233 | ($realtype eq 'ARRAY') ? '[]' : |
| 234 | "''" ; |
| 235 | push @post, $name . " = " . $s->{seen}{$id}[0]; |
| 236 | } |
| 237 | else { |
| 238 | $out = $s->{seen}{$id}[0]; |
| 239 | if ($name =~ /^([\@\%])/) { |
| 240 | my $start = $1; |
| 241 | if ($out =~ /^\\$start/) { |
| 242 | $out = substr($out, 1); |
| 243 | } |
| 244 | else { |
| 245 | $out = $start . '{' . $out . '}'; |
| 246 | } |
| 247 | } |
| 248 | } |
| 249 | return $out; |
| 250 | # } |
| 251 | } |
| 252 | else { |
| 253 | # store our name |
| 254 | $s->{seen}{$id} = [ (($name =~ /^[@%]/) ? ('\\' . $name ) : |
| 255 | ($realtype eq 'CODE' and |
| 256 | $name =~ /^[*](.*)$/) ? ('\\&' . $1 ) : |
| 257 | $name ), |
| 258 | $val ]; |
| 259 | } |
| 260 | } |
| 261 | |
| 262 | $s->{level}++; |
| 263 | $ipad = $s->{xpad} x $s->{level}; |
| 264 | |
| 265 | if ($realpack) { # we have a blessed ref |
| 266 | $out = $s->{'bless'} . '( '; |
| 267 | $blesspad = $s->{apad}; |
| 268 | $s->{apad} .= ' ' if ($s->{indent} >= 2); |
| 269 | } |
| 270 | |
| 271 | if ($realtype eq 'SCALAR') { |
| 272 | if ($realpack) { |
| 273 | $out .= 'do{\\(my $o = ' . $s->_dump($$val, "\${$name}") . ')}'; |
| 274 | } |
| 275 | else { |
| 276 | $out .= '\\' . $s->_dump($$val, "\${$name}"); |
| 277 | } |
| 278 | } |
| 279 | elsif ($realtype eq 'GLOB') { |
| 280 | $out .= '\\' . $s->_dump($$val, "*{$name}"); |
| 281 | } |
| 282 | elsif ($realtype eq 'ARRAY') { |
| 283 | my($v, $pad, $mname); |
| 284 | my($i) = 0; |
| 285 | $out .= ($name =~ /^\@/) ? '(' : '['; |
| 286 | $pad = $s->{sep} . $s->{pad} . $s->{apad}; |
| 287 | ($name =~ /^\@(.*)$/) ? ($mname = "\$" . $1) : |
| 288 | # omit -> if $foo->[0]->{bar}, but not ${$foo->[0]}->{bar} |
| 289 | ($name =~ /^\\?[\%\@\*\$][^{].*[]}]$/) ? ($mname = $name) : |
| 290 | ($mname = $name . '->'); |
| 291 | $mname .= '->' if $mname =~ /^\*.+\{[A-Z]+\}$/; |
| 292 | for $v (@$val) { |
| 293 | $sname = $mname . '[' . $i . ']'; |
| 294 | $out .= $pad . $ipad . '#' . $i if $s->{indent} >= 3; |
| 295 | $out .= $pad . $ipad . $s->_dump($v, $sname); |
| 296 | $out .= "," if $i++ < $#$val; |
| 297 | } |
| 298 | $out .= $pad . ($s->{xpad} x ($s->{level} - 1)) if $i; |
| 299 | $out .= ($name =~ /^\@/) ? ')' : ']'; |
| 300 | } |
| 301 | elsif ($realtype eq 'HASH') { |
| 302 | my($k, $v, $pad, $lpad, $mname); |
| 303 | $out .= ($name =~ /^\%/) ? '(' : '{'; |
| 304 | $pad = $s->{sep} . $s->{pad} . $s->{apad}; |
| 305 | $lpad = $s->{apad}; |
| 306 | ($name =~ /^\%(.*)$/) ? ($mname = "\$" . $1) : |
| 307 | # omit -> if $foo->[0]->{bar}, but not ${$foo->[0]}->{bar} |
| 308 | ($name =~ /^\\?[\%\@\*\$][^{].*[]}]$/) ? ($mname = $name) : |
| 309 | ($mname = $name . '->'); |
| 310 | $mname .= '->' if $mname =~ /^\*.+\{[A-Z]+\}$/; |
| 311 | while (($k, $v) = each %$val) { |
| 312 | my $nk = $s->_dump($k, ""); |
| 313 | $nk = $1 if !$s->{quotekeys} and $nk =~ /^[\"\']([A-Za-z_]\w*)[\"\']$/; |
| 314 | $sname = $mname . '{' . $nk . '}'; |
| 315 | $out .= $pad . $ipad . $nk . " => "; |
| 316 | |
| 317 | # temporarily alter apad |
| 318 | $s->{apad} .= (" " x (length($nk) + 4)) if $s->{indent} >= 2; |
| 319 | $out .= $s->_dump($val->{$k}, $sname) . ","; |
| 320 | $s->{apad} = $lpad if $s->{indent} >= 2; |
| 321 | } |
| 322 | if (substr($out, -1) eq ',') { |
| 323 | chop $out; |
| 324 | $out .= $pad . ($s->{xpad} x ($s->{level} - 1)); |
| 325 | } |
| 326 | $out .= ($name =~ /^\%/) ? ')' : '}'; |
| 327 | } |
| 328 | elsif ($realtype eq 'CODE') { |
| 329 | $out .= 'sub { "DUMMY" }'; |
| 330 | carp "Encountered CODE ref, using dummy placeholder" if $s->{purity}; |
| 331 | } |
| 332 | else { |
| 333 | croak "Can\'t handle $realtype type."; |
| 334 | } |
| 335 | |
| 336 | if ($realpack) { # we have a blessed ref |
| 337 | $out .= ', \'' . $realpack . '\'' . ' )'; |
| 338 | $out .= '->' . $s->{toaster} . '()' if $s->{toaster} ne ''; |
| 339 | $s->{apad} = $blesspad; |
| 340 | } |
| 341 | $s->{level}--; |
| 342 | |
| 343 | } |
| 344 | else { # simple scalar |
| 345 | |
| 346 | my $ref = \$_[1]; |
| 347 | # first, catalog the scalar |
| 348 | if ($name ne '') { |
| 349 | ($id) = ("$ref" =~ /\(([^\(]*)\)$/); |
| 350 | if (exists $s->{seen}{$id}) { |
| 351 | if ($s->{seen}{$id}[2]) { |
| 352 | $out = $s->{seen}{$id}[0]; |
| 353 | #warn "[<$out]\n"; |
| 354 | return "\${$out}"; |
| 355 | } |
| 356 | } |
| 357 | else { |
| 358 | #warn "[>\\$name]\n"; |
| 359 | $s->{seen}{$id} = ["\\$name", $ref]; |
| 360 | } |
| 361 | } |
| 362 | if (ref($ref) eq 'GLOB' or "$ref" =~ /=GLOB\([^()]+\)$/) { # glob |
| 363 | my $name = substr($val, 1); |
| 364 | if ($name =~ /^[A-Za-z_][\w:]*$/) { |
| 365 | $name =~ s/^main::/::/; |
| 366 | $sname = $name; |
| 367 | } |
| 368 | else { |
| 369 | $sname = $s->_dump($name, ""); |
| 370 | $sname = '{' . $sname . '}'; |
| 371 | } |
| 372 | if ($s->{purity}) { |
| 373 | my $k; |
| 374 | local ($s->{level}) = 0; |
| 375 | for $k (qw(SCALAR ARRAY HASH)) { |
| 376 | my $gval = *$val{$k}; |
| 377 | next unless defined $gval; |
| 378 | next if $k eq "SCALAR" && ! defined $$gval; # always there |
| 379 | |
| 380 | # _dump can push into @post, so we hold our place using $postlen |
| 381 | my $postlen = scalar @post; |
| 382 | $post[$postlen] = "\*$sname = "; |
| 383 | local ($s->{apad}) = " " x length($post[$postlen]) if $s->{indent} >= 2; |
| 384 | $post[$postlen] .= $s->_dump($gval, "\*$sname\{$k\}"); |
| 385 | } |
| 386 | } |
| 387 | $out .= '*' . $sname; |
| 388 | } |
| 389 | elsif (!defined($val)) { |
| 390 | $out .= "undef"; |
| 391 | } |
| 392 | elsif ($val =~ /^-?[1-9]\d{0,8}$/) { # safe decimal number |
| 393 | $out .= $val; |
| 394 | } |
| 395 | else { # string |
| 396 | if ($s->{useqq}) { |
| 397 | $out .= qquote($val, $s->{useqq}); |
| 398 | } |
| 399 | else { |
| 400 | $val =~ s/([\\\'])/\\$1/g; |
| 401 | $out .= '\'' . $val . '\''; |
| 402 | } |
| 403 | } |
| 404 | } |
| 405 | if ($id) { |
| 406 | # if we made it this far, $id was added to seen list at current |
| 407 | # level, so remove it to get deep copies |
| 408 | if ($s->{deepcopy}) { |
| 409 | delete($s->{seen}{$id}); |
| 410 | } |
| 411 | elsif ($name) { |
| 412 | $s->{seen}{$id}[2] = 1; |
| 413 | } |
| 414 | } |
| 415 | return $out; |
| 416 | } |
| 417 | |
| 418 | # |
| 419 | # non-OO style of earlier version |
| 420 | # |
| 421 | sub Dumper { |
| 422 | return Data::Dumper->Dump([@_]); |
| 423 | } |
| 424 | |
| 425 | # |
| 426 | # same, only calls the XS version |
| 427 | # |
| 428 | sub DumperX { |
| 429 | return Data::Dumper->Dumpxs([@_], []); |
| 430 | } |
| 431 | |
| 432 | sub Dumpf { return Data::Dumper->Dump(@_) } |
| 433 | |
| 434 | sub Dumpp { print Data::Dumper->Dump(@_) } |
| 435 | |
| 436 | # |
| 437 | # reset the "seen" cache |
| 438 | # |
| 439 | sub Reset { |
| 440 | my($s) = shift; |
| 441 | $s->{seen} = {}; |
| 442 | return $s; |
| 443 | } |
| 444 | |
| 445 | sub Indent { |
| 446 | my($s, $v) = @_; |
| 447 | if (defined($v)) { |
| 448 | if ($v == 0) { |
| 449 | $s->{xpad} = ""; |
| 450 | $s->{sep} = ""; |
| 451 | } |
| 452 | else { |
| 453 | $s->{xpad} = " "; |
| 454 | $s->{sep} = "\n"; |
| 455 | } |
| 456 | $s->{indent} = $v; |
| 457 | return $s; |
| 458 | } |
| 459 | else { |
| 460 | return $s->{indent}; |
| 461 | } |
| 462 | } |
| 463 | |
| 464 | sub Pad { |
| 465 | my($s, $v) = @_; |
| 466 | defined($v) ? (($s->{pad} = $v), return $s) : $s->{pad}; |
| 467 | } |
| 468 | |
| 469 | sub Varname { |
| 470 | my($s, $v) = @_; |
| 471 | defined($v) ? (($s->{varname} = $v), return $s) : $s->{varname}; |
| 472 | } |
| 473 | |
| 474 | sub Purity { |
| 475 | my($s, $v) = @_; |
| 476 | defined($v) ? (($s->{purity} = $v), return $s) : $s->{purity}; |
| 477 | } |
| 478 | |
| 479 | sub Useqq { |
| 480 | my($s, $v) = @_; |
| 481 | defined($v) ? (($s->{useqq} = $v), return $s) : $s->{useqq}; |
| 482 | } |
| 483 | |
| 484 | sub Terse { |
| 485 | my($s, $v) = @_; |
| 486 | defined($v) ? (($s->{terse} = $v), return $s) : $s->{terse}; |
| 487 | } |
| 488 | |
| 489 | sub Freezer { |
| 490 | my($s, $v) = @_; |
| 491 | defined($v) ? (($s->{freezer} = $v), return $s) : $s->{freezer}; |
| 492 | } |
| 493 | |
| 494 | sub Toaster { |
| 495 | my($s, $v) = @_; |
| 496 | defined($v) ? (($s->{toaster} = $v), return $s) : $s->{toaster}; |
| 497 | } |
| 498 | |
| 499 | sub Deepcopy { |
| 500 | my($s, $v) = @_; |
| 501 | defined($v) ? (($s->{deepcopy} = $v), return $s) : $s->{deepcopy}; |
| 502 | } |
| 503 | |
| 504 | sub Quotekeys { |
| 505 | my($s, $v) = @_; |
| 506 | defined($v) ? (($s->{quotekeys} = $v), return $s) : $s->{quotekeys}; |
| 507 | } |
| 508 | |
| 509 | sub Bless { |
| 510 | my($s, $v) = @_; |
| 511 | defined($v) ? (($s->{'bless'} = $v), return $s) : $s->{'bless'}; |
| 512 | } |
| 513 | |
| 514 | # used by qquote below |
| 515 | my %esc = ( |
| 516 | "\a" => "\\a", |
| 517 | "\b" => "\\b", |
| 518 | "\t" => "\\t", |
| 519 | "\n" => "\\n", |
| 520 | "\f" => "\\f", |
| 521 | "\r" => "\\r", |
| 522 | "\e" => "\\e", |
| 523 | ); |
| 524 | |
| 525 | # put a string value in double quotes |
| 526 | sub qquote { |
| 527 | local($_) = shift; |
| 528 | s/([\\\"\@\$])/\\$1/g; |
| 529 | return qq("$_") unless /[^\040-\176]/; # fast exit |
| 530 | |
| 531 | my $high = shift || ""; |
| 532 | s/([\a\b\t\n\f\r\e])/$esc{$1}/g; |
| 533 | |
| 534 | # no need for 3 digits in escape for these |
| 535 | s/([\0-\037])(?!\d)/'\\'.sprintf('%o',ord($1))/eg; |
| 536 | |
| 537 | s/([\0-\037\177])/'\\'.sprintf('%03o',ord($1))/eg; |
| 538 | if ($high eq "iso8859") { |
| 539 | s/([\200-\240])/'\\'.sprintf('%o',ord($1))/eg; |
| 540 | } elsif ($high eq "utf8") { |
| 541 | # use utf8; |
| 542 | # $str =~ s/([^\040-\176])/sprintf "\\x{%04x}", ord($1)/ge; |
| 543 | } elsif ($high eq "8bit") { |
| 544 | # leave it as it is |
| 545 | } else { |
| 546 | s/([\0-\037\177-\377])/'\\'.sprintf('%03o',ord($1))/eg; |
| 547 | } |
| 548 | return qq("$_"); |
| 549 | } |
| 550 | |
| 551 | 1; |
| 552 | __END__ |
| 553 | |
| 554 | =head1 NAME |
| 555 | |
| 556 | Data::Dumper - stringified perl data structures, suitable for both printing and C<eval> |
| 557 | |
| 558 | |
| 559 | =head1 SYNOPSIS |
| 560 | |
| 561 | use Data::Dumper; |
| 562 | |
| 563 | # simple procedural interface |
| 564 | print Dumper($foo, $bar); |
| 565 | |
| 566 | # extended usage with names |
| 567 | print Data::Dumper->Dump([$foo, $bar], [qw(foo *ary)]); |
| 568 | |
| 569 | # configuration variables |
| 570 | { |
| 571 | local $Data::Dump::Purity = 1; |
| 572 | eval Data::Dumper->Dump([$foo, $bar], [qw(foo *ary)]); |
| 573 | } |
| 574 | |
| 575 | # OO usage |
| 576 | $d = Data::Dumper->new([$foo, $bar], [qw(foo *ary)]); |
| 577 | ... |
| 578 | print $d->Dump; |
| 579 | ... |
| 580 | $d->Purity(1)->Terse(1)->Deepcopy(1); |
| 581 | eval $d->Dump; |
| 582 | |
| 583 | |
| 584 | =head1 DESCRIPTION |
| 585 | |
| 586 | Given a list of scalars or reference variables, writes out their contents in |
| 587 | perl syntax. The references can also be objects. The contents of each |
| 588 | variable is output in a single Perl statement. Handles self-referential |
| 589 | structures correctly. |
| 590 | |
| 591 | The return value can be C<eval>ed to get back an identical copy of the |
| 592 | original reference structure. |
| 593 | |
| 594 | Any references that are the same as one of those passed in will be named |
| 595 | C<$VAR>I<n> (where I<n> is a numeric suffix), and other duplicate references |
| 596 | to substructures within C<$VAR>I<n> will be appropriately labeled using arrow |
| 597 | notation. You can specify names for individual values to be dumped if you |
| 598 | use the C<Dump()> method, or you can change the default C<$VAR> prefix to |
| 599 | something else. See C<$Data::Dumper::Varname> and C<$Data::Dumper::Terse> |
| 600 | below. |
| 601 | |
| 602 | The default output of self-referential structures can be C<eval>ed, but the |
| 603 | nested references to C<$VAR>I<n> will be undefined, since a recursive |
| 604 | structure cannot be constructed using one Perl statement. You should set the |
| 605 | C<Purity> flag to 1 to get additional statements that will correctly fill in |
| 606 | these references. |
| 607 | |
| 608 | In the extended usage form, the references to be dumped can be given |
| 609 | user-specified names. If a name begins with a C<*>, the output will |
| 610 | describe the dereferenced type of the supplied reference for hashes and |
| 611 | arrays, and coderefs. Output of names will be avoided where possible if |
| 612 | the C<Terse> flag is set. |
| 613 | |
| 614 | In many cases, methods that are used to set the internal state of the |
| 615 | object will return the object itself, so method calls can be conveniently |
| 616 | chained together. |
| 617 | |
| 618 | Several styles of output are possible, all controlled by setting |
| 619 | the C<Indent> flag. See L<Configuration Variables or Methods> below |
| 620 | for details. |
| 621 | |
| 622 | |
| 623 | =head2 Methods |
| 624 | |
| 625 | =over 4 |
| 626 | |
| 627 | =item I<PACKAGE>->new(I<ARRAYREF [>, I<ARRAYREF]>) |
| 628 | |
| 629 | Returns a newly created C<Data::Dumper> object. The first argument is an |
| 630 | anonymous array of values to be dumped. The optional second argument is an |
| 631 | anonymous array of names for the values. The names need not have a leading |
| 632 | C<$> sign, and must be comprised of alphanumeric characters. You can begin |
| 633 | a name with a C<*> to specify that the dereferenced type must be dumped |
| 634 | instead of the reference itself, for ARRAY and HASH references. |
| 635 | |
| 636 | The prefix specified by C<$Data::Dumper::Varname> will be used with a |
| 637 | numeric suffix if the name for a value is undefined. |
| 638 | |
| 639 | Data::Dumper will catalog all references encountered while dumping the |
| 640 | values. Cross-references (in the form of names of substructures in perl |
| 641 | syntax) will be inserted at all possible points, preserving any structural |
| 642 | interdependencies in the original set of values. Structure traversal is |
| 643 | depth-first, and proceeds in order from the first supplied value to |
| 644 | the last. |
| 645 | |
| 646 | =item I<$OBJ>->Dump I<or> I<PACKAGE>->Dump(I<ARRAYREF [>, I<ARRAYREF]>) |
| 647 | |
| 648 | Returns the stringified form of the values stored in the object (preserving |
| 649 | the order in which they were supplied to C<new>), subject to the |
| 650 | configuration options below. In an array context, it returns a list |
| 651 | of strings corresponding to the supplied values. |
| 652 | |
| 653 | The second form, for convenience, simply calls the C<new> method on its |
| 654 | arguments before dumping the object immediately. |
| 655 | |
| 656 | =item I<$OBJ>->Dumpxs I<or> I<PACKAGE>->Dumpxs(I<ARRAYREF [>, I<ARRAYREF]>) |
| 657 | |
| 658 | This method is available if you were able to compile and install the XSUB |
| 659 | extension to C<Data::Dumper>. It is exactly identical to the C<Dump> method |
| 660 | above, only about 4 to 5 times faster, since it is written entirely in C. |
| 661 | |
| 662 | =item I<$OBJ>->Seen(I<[HASHREF]>) |
| 663 | |
| 664 | Queries or adds to the internal table of already encountered references. |
| 665 | You must use C<Reset> to explicitly clear the table if needed. Such |
| 666 | references are not dumped; instead, their names are inserted wherever they |
| 667 | are encountered subsequently. This is useful especially for properly |
| 668 | dumping subroutine references. |
| 669 | |
| 670 | Expects a anonymous hash of name => value pairs. Same rules apply for names |
| 671 | as in C<new>. If no argument is supplied, will return the "seen" list of |
| 672 | name => value pairs, in an array context. Otherwise, returns the object |
| 673 | itself. |
| 674 | |
| 675 | =item I<$OBJ>->Values(I<[ARRAYREF]>) |
| 676 | |
| 677 | Queries or replaces the internal array of values that will be dumped. |
| 678 | When called without arguments, returns the values. Otherwise, returns the |
| 679 | object itself. |
| 680 | |
| 681 | =item I<$OBJ>->Names(I<[ARRAYREF]>) |
| 682 | |
| 683 | Queries or replaces the internal array of user supplied names for the values |
| 684 | that will be dumped. When called without arguments, returns the names. |
| 685 | Otherwise, returns the object itself. |
| 686 | |
| 687 | =item I<$OBJ>->Reset |
| 688 | |
| 689 | Clears the internal table of "seen" references and returns the object |
| 690 | itself. |
| 691 | |
| 692 | =back |
| 693 | |
| 694 | =head2 Functions |
| 695 | |
| 696 | =over 4 |
| 697 | |
| 698 | =item Dumper(I<LIST>) |
| 699 | |
| 700 | Returns the stringified form of the values in the list, subject to the |
| 701 | configuration options below. The values will be named C<$VAR>I<n> in the |
| 702 | output, where I<n> is a numeric suffix. Will return a list of strings |
| 703 | in an array context. |
| 704 | |
| 705 | =item DumperX(I<LIST>) |
| 706 | |
| 707 | Identical to the C<Dumper()> function above, but this calls the XSUB |
| 708 | implementation. Only available if you were able to compile and install |
| 709 | the XSUB extensions in C<Data::Dumper>. |
| 710 | |
| 711 | =back |
| 712 | |
| 713 | =head2 Configuration Variables or Methods |
| 714 | |
| 715 | Several configuration variables can be used to control the kind of output |
| 716 | generated when using the procedural interface. These variables are usually |
| 717 | C<local>ized in a block so that other parts of the code are not affected by |
| 718 | the change. |
| 719 | |
| 720 | These variables determine the default state of the object created by calling |
| 721 | the C<new> method, but cannot be used to alter the state of the object |
| 722 | thereafter. The equivalent method names should be used instead to query |
| 723 | or set the internal state of the object. |
| 724 | |
| 725 | The method forms return the object itself when called with arguments, |
| 726 | so that they can be chained together nicely. |
| 727 | |
| 728 | =over 4 |
| 729 | |
| 730 | =item $Data::Dumper::Indent I<or> I<$OBJ>->Indent(I<[NEWVAL]>) |
| 731 | |
| 732 | Controls the style of indentation. It can be set to 0, 1, 2 or 3. Style 0 |
| 733 | spews output without any newlines, indentation, or spaces between list |
| 734 | items. It is the most compact format possible that can still be called |
| 735 | valid perl. Style 1 outputs a readable form with newlines but no fancy |
| 736 | indentation (each level in the structure is simply indented by a fixed |
| 737 | amount of whitespace). Style 2 (the default) outputs a very readable form |
| 738 | which takes into account the length of hash keys (so the hash value lines |
| 739 | up). Style 3 is like style 2, but also annotates the elements of arrays |
| 740 | with their index (but the comment is on its own line, so array output |
| 741 | consumes twice the number of lines). Style 2 is the default. |
| 742 | |
| 743 | =item $Data::Dumper::Purity I<or> I<$OBJ>->Purity(I<[NEWVAL]>) |
| 744 | |
| 745 | Controls the degree to which the output can be C<eval>ed to recreate the |
| 746 | supplied reference structures. Setting it to 1 will output additional perl |
| 747 | statements that will correctly recreate nested references. The default is |
| 748 | 0. |
| 749 | |
| 750 | =item $Data::Dumper::Pad I<or> I<$OBJ>->Pad(I<[NEWVAL]>) |
| 751 | |
| 752 | Specifies the string that will be prefixed to every line of the output. |
| 753 | Empty string by default. |
| 754 | |
| 755 | =item $Data::Dumper::Varname I<or> I<$OBJ>->Varname(I<[NEWVAL]>) |
| 756 | |
| 757 | Contains the prefix to use for tagging variable names in the output. The |
| 758 | default is "VAR". |
| 759 | |
| 760 | =item $Data::Dumper::Useqq I<or> I<$OBJ>->Useqq(I<[NEWVAL]>) |
| 761 | |
| 762 | When set, enables the use of double quotes for representing string values. |
| 763 | Whitespace other than space will be represented as C<[\n\t\r]>, "unsafe" |
| 764 | characters will be backslashed, and unprintable characters will be output as |
| 765 | quoted octal integers. Since setting this variable imposes a performance |
| 766 | penalty, the default is 0. The C<Dumpxs()> method does not honor this |
| 767 | flag yet. |
| 768 | |
| 769 | =item $Data::Dumper::Terse I<or> I<$OBJ>->Terse(I<[NEWVAL]>) |
| 770 | |
| 771 | When set, Data::Dumper will emit single, non-self-referential values as |
| 772 | atoms/terms rather than statements. This means that the C<$VAR>I<n> names |
| 773 | will be avoided where possible, but be advised that such output may not |
| 774 | always be parseable by C<eval>. |
| 775 | |
| 776 | =item $Data::Dumper::Freezer I<or> $I<OBJ>->Freezer(I<[NEWVAL]>) |
| 777 | |
| 778 | Can be set to a method name, or to an empty string to disable the feature. |
| 779 | Data::Dumper will invoke that method via the object before attempting to |
| 780 | stringify it. This method can alter the contents of the object (if, for |
| 781 | instance, it contains data allocated from C), and even rebless it in a |
| 782 | different package. The client is responsible for making sure the specified |
| 783 | method can be called via the object, and that the object ends up containing |
| 784 | only perl data types after the method has been called. Defaults to an empty |
| 785 | string. |
| 786 | |
| 787 | =item $Data::Dumper::Toaster I<or> $I<OBJ>->Toaster(I<[NEWVAL]>) |
| 788 | |
| 789 | Can be set to a method name, or to an empty string to disable the feature. |
| 790 | Data::Dumper will emit a method call for any objects that are to be dumped |
| 791 | using the syntax C<bless(DATA, CLASS)->METHOD()>. Note that this means that |
| 792 | the method specified will have to perform any modifications required on the |
| 793 | object (like creating new state within it, and/or reblessing it in a |
| 794 | different package) and then return it. The client is responsible for making |
| 795 | sure the method can be called via the object, and that it returns a valid |
| 796 | object. Defaults to an empty string. |
| 797 | |
| 798 | =item $Data::Dumper::Deepcopy I<or> $I<OBJ>->Deepcopy(I<[NEWVAL]>) |
| 799 | |
| 800 | Can be set to a boolean value to enable deep copies of structures. |
| 801 | Cross-referencing will then only be done when absolutely essential |
| 802 | (i.e., to break reference cycles). Default is 0. |
| 803 | |
| 804 | =item $Data::Dumper::Quotekeys I<or> $I<OBJ>->Quotekeys(I<[NEWVAL]>) |
| 805 | |
| 806 | Can be set to a boolean value to control whether hash keys are quoted. |
| 807 | A false value will avoid quoting hash keys when it looks like a simple |
| 808 | string. Default is 1, which will always enclose hash keys in quotes. |
| 809 | |
| 810 | =item $Data::Dumper::Bless I<or> $I<OBJ>->Bless(I<[NEWVAL]>) |
| 811 | |
| 812 | Can be set to a string that specifies an alternative to the C<bless> |
| 813 | builtin operator used to create objects. A function with the specified |
| 814 | name should exist, and should accept the same arguments as the builtin. |
| 815 | Default is C<bless>. |
| 816 | |
| 817 | =back |
| 818 | |
| 819 | =head2 Exports |
| 820 | |
| 821 | =over 4 |
| 822 | |
| 823 | =item Dumper |
| 824 | |
| 825 | =back |
| 826 | |
| 827 | =head1 EXAMPLES |
| 828 | |
| 829 | Run these code snippets to get a quick feel for the behavior of this |
| 830 | module. When you are through with these examples, you may want to |
| 831 | add or change the various configuration variables described above, |
| 832 | to see their behavior. (See the testsuite in the Data::Dumper |
| 833 | distribution for more examples.) |
| 834 | |
| 835 | |
| 836 | use Data::Dumper; |
| 837 | |
| 838 | package Foo; |
| 839 | sub new {bless {'a' => 1, 'b' => sub { return "foo" }}, $_[0]}; |
| 840 | |
| 841 | package Fuz; # a weird REF-REF-SCALAR object |
| 842 | sub new {bless \($_ = \ 'fu\'z'), $_[0]}; |
| 843 | |
| 844 | package main; |
| 845 | $foo = Foo->new; |
| 846 | $fuz = Fuz->new; |
| 847 | $boo = [ 1, [], "abcd", \*foo, |
| 848 | {1 => 'a', 023 => 'b', 0x45 => 'c'}, |
| 849 | \\"p\q\'r", $foo, $fuz]; |
| 850 | |
| 851 | ######## |
| 852 | # simple usage |
| 853 | ######## |
| 854 | |
| 855 | $bar = eval(Dumper($boo)); |
| 856 | print($@) if $@; |
| 857 | print Dumper($boo), Dumper($bar); # pretty print (no array indices) |
| 858 | |
| 859 | $Data::Dumper::Terse = 1; # don't output names where feasible |
| 860 | $Data::Dumper::Indent = 0; # turn off all pretty print |
| 861 | print Dumper($boo), "\n"; |
| 862 | |
| 863 | $Data::Dumper::Indent = 1; # mild pretty print |
| 864 | print Dumper($boo); |
| 865 | |
| 866 | $Data::Dumper::Indent = 3; # pretty print with array indices |
| 867 | print Dumper($boo); |
| 868 | |
| 869 | $Data::Dumper::Useqq = 1; # print strings in double quotes |
| 870 | print Dumper($boo); |
| 871 | |
| 872 | |
| 873 | ######## |
| 874 | # recursive structures |
| 875 | ######## |
| 876 | |
| 877 | @c = ('c'); |
| 878 | $c = \@c; |
| 879 | $b = {}; |
| 880 | $a = [1, $b, $c]; |
| 881 | $b->{a} = $a; |
| 882 | $b->{b} = $a->[1]; |
| 883 | $b->{c} = $a->[2]; |
| 884 | print Data::Dumper->Dump([$a,$b,$c], [qw(a b c)]); |
| 885 | |
| 886 | |
| 887 | $Data::Dumper::Purity = 1; # fill in the holes for eval |
| 888 | print Data::Dumper->Dump([$a, $b], [qw(*a b)]); # print as @a |
| 889 | print Data::Dumper->Dump([$b, $a], [qw(*b a)]); # print as %b |
| 890 | |
| 891 | |
| 892 | $Data::Dumper::Deepcopy = 1; # avoid cross-refs |
| 893 | print Data::Dumper->Dump([$b, $a], [qw(*b a)]); |
| 894 | |
| 895 | |
| 896 | $Data::Dumper::Purity = 0; # avoid cross-refs |
| 897 | print Data::Dumper->Dump([$b, $a], [qw(*b a)]); |
| 898 | |
| 899 | |
| 900 | ######## |
| 901 | # object-oriented usage |
| 902 | ######## |
| 903 | |
| 904 | $d = Data::Dumper->new([$a,$b], [qw(a b)]); |
| 905 | $d->Seen({'*c' => $c}); # stash a ref without printing it |
| 906 | $d->Indent(3); |
| 907 | print $d->Dump; |
| 908 | $d->Reset->Purity(0); # empty the seen cache |
| 909 | print join "----\n", $d->Dump; |
| 910 | |
| 911 | |
| 912 | ######## |
| 913 | # persistence |
| 914 | ######## |
| 915 | |
| 916 | package Foo; |
| 917 | sub new { bless { state => 'awake' }, shift } |
| 918 | sub Freeze { |
| 919 | my $s = shift; |
| 920 | print STDERR "preparing to sleep\n"; |
| 921 | $s->{state} = 'asleep'; |
| 922 | return bless $s, 'Foo::ZZZ'; |
| 923 | } |
| 924 | |
| 925 | package Foo::ZZZ; |
| 926 | sub Thaw { |
| 927 | my $s = shift; |
| 928 | print STDERR "waking up\n"; |
| 929 | $s->{state} = 'awake'; |
| 930 | return bless $s, 'Foo'; |
| 931 | } |
| 932 | |
| 933 | package Foo; |
| 934 | use Data::Dumper; |
| 935 | $a = Foo->new; |
| 936 | $b = Data::Dumper->new([$a], ['c']); |
| 937 | $b->Freezer('Freeze'); |
| 938 | $b->Toaster('Thaw'); |
| 939 | $c = $b->Dump; |
| 940 | print $c; |
| 941 | $d = eval $c; |
| 942 | print Data::Dumper->Dump([$d], ['d']); |
| 943 | |
| 944 | |
| 945 | ######## |
| 946 | # symbol substitution (useful for recreating CODE refs) |
| 947 | ######## |
| 948 | |
| 949 | sub foo { print "foo speaking\n" } |
| 950 | *other = \&foo; |
| 951 | $bar = [ \&other ]; |
| 952 | $d = Data::Dumper->new([\&other,$bar],['*other','bar']); |
| 953 | $d->Seen({ '*foo' => \&foo }); |
| 954 | print $d->Dump; |
| 955 | |
| 956 | |
| 957 | =head1 BUGS |
| 958 | |
| 959 | Due to limitations of Perl subroutine call semantics, you cannot pass an |
| 960 | array or hash. Prepend it with a C<\> to pass its reference instead. This |
| 961 | will be remedied in time, with the arrival of prototypes in later versions |
| 962 | of Perl. For now, you need to use the extended usage form, and prepend the |
| 963 | name with a C<*> to output it as a hash or array. |
| 964 | |
| 965 | C<Data::Dumper> cheats with CODE references. If a code reference is |
| 966 | encountered in the structure being processed, an anonymous subroutine that |
| 967 | contains the string '"DUMMY"' will be inserted in its place, and a warning |
| 968 | will be printed if C<Purity> is set. You can C<eval> the result, but bear |
| 969 | in mind that the anonymous sub that gets created is just a placeholder. |
| 970 | Someday, perl will have a switch to cache-on-demand the string |
| 971 | representation of a compiled piece of code, I hope. If you have prior |
| 972 | knowledge of all the code refs that your data structures are likely |
| 973 | to have, you can use the C<Seen> method to pre-seed the internal reference |
| 974 | table and make the dumped output point to them, instead. See L<EXAMPLES> |
| 975 | above. |
| 976 | |
| 977 | The C<Useqq> flag is not honored by C<Dumpxs()> (it always outputs |
| 978 | strings in single quotes). |
| 979 | |
| 980 | SCALAR objects have the weirdest looking C<bless> workaround. |
| 981 | |
| 982 | |
| 983 | =head1 AUTHOR |
| 984 | |
| 985 | Gurusamy Sarathy gsar@umich.edu |
| 986 | |
| 987 | Copyright (c) 1996-98 Gurusamy Sarathy. All rights reserved. |
| 988 | This program is free software; you can redistribute it and/or |
| 989 | modify it under the same terms as Perl itself. |
| 990 | |
| 991 | |
| 992 | =head1 VERSION |
| 993 | |
| 994 | Version 2.10 (31 Oct 1998) |
| 995 | |
| 996 | =head1 SEE ALSO |
| 997 | |
| 998 | perl(1) |
| 999 | |
| 1000 | =cut |