| 1 | # |
| 2 | # Tie/IxHash.pm |
| 3 | # |
| 4 | # Indexed hash implementation for Perl |
| 5 | # |
| 6 | # See below for documentation. |
| 7 | # |
| 8 | |
| 9 | require 5.003; |
| 10 | |
| 11 | package Tie::IxHash; |
| 12 | use integer; |
| 13 | require Tie::Hash; |
| 14 | @ISA = qw(Tie::Hash); |
| 15 | |
| 16 | $VERSION = $VERSION = '1.21'; |
| 17 | |
| 18 | # |
| 19 | # standard tie functions |
| 20 | # |
| 21 | |
| 22 | sub TIEHASH { |
| 23 | my($c) = shift; |
| 24 | my($s) = []; |
| 25 | $s->[0] = {}; # hashkey index |
| 26 | $s->[1] = []; # array of keys |
| 27 | $s->[2] = []; # array of data |
| 28 | $s->[3] = 0; # iter count |
| 29 | |
| 30 | bless $s, $c; |
| 31 | |
| 32 | $s->Push(@_) if @_; |
| 33 | |
| 34 | return $s; |
| 35 | } |
| 36 | |
| 37 | #sub DESTROY {} # costly if there's nothing to do |
| 38 | |
| 39 | sub FETCH { |
| 40 | my($s, $k) = (shift, shift); |
| 41 | return exists( $s->[0]{$k} ) ? $s->[2][ $s->[0]{$k} ] : undef; |
| 42 | } |
| 43 | |
| 44 | sub STORE { |
| 45 | my($s, $k, $v) = (shift, shift, shift); |
| 46 | |
| 47 | if (exists $s->[0]{$k}) { |
| 48 | my($i) = $s->[0]{$k}; |
| 49 | $s->[1][$i] = $k; |
| 50 | $s->[2][$i] = $v; |
| 51 | $s->[0]{$k} = $i; |
| 52 | } |
| 53 | else { |
| 54 | push(@{$s->[1]}, $k); |
| 55 | push(@{$s->[2]}, $v); |
| 56 | $s->[0]{$k} = $#{$s->[1]}; |
| 57 | } |
| 58 | } |
| 59 | |
| 60 | sub DELETE { |
| 61 | my($s, $k) = (shift, shift); |
| 62 | |
| 63 | if (exists $s->[0]{$k}) { |
| 64 | my($i) = $s->[0]{$k}; |
| 65 | for ($i+1..$#{$s->[1]}) { # reset higher elt indexes |
| 66 | $s->[0]{$s->[1][$_]}--; # timeconsuming, is there is better way? |
| 67 | } |
| 68 | delete $s->[0]{$k}; |
| 69 | splice @{$s->[1]}, $i, 1; |
| 70 | return (splice(@{$s->[2]}, $i, 1))[0]; |
| 71 | } |
| 72 | return undef; |
| 73 | } |
| 74 | |
| 75 | sub EXISTS { |
| 76 | exists $_[0]->[0]{ $_[1] }; |
| 77 | } |
| 78 | |
| 79 | sub FIRSTKEY { |
| 80 | $_[0][3] = 0; |
| 81 | &NEXTKEY; |
| 82 | } |
| 83 | |
| 84 | sub NEXTKEY { |
| 85 | return $_[0][1][$_[0][3]++] if ($_[0][3] <= $#{$_[0][1]}); |
| 86 | return undef; |
| 87 | } |
| 88 | |
| 89 | |
| 90 | |
| 91 | # |
| 92 | # |
| 93 | # class functions that provide additional capabilities |
| 94 | # |
| 95 | # |
| 96 | |
| 97 | sub new { TIEHASH(@_) } |
| 98 | |
| 99 | # |
| 100 | # add pairs to end of indexed hash |
| 101 | # note that if a supplied key exists, it will not be reordered |
| 102 | # |
| 103 | sub Push { |
| 104 | my($s) = shift; |
| 105 | while (@_) { |
| 106 | $s->STORE(shift, shift); |
| 107 | } |
| 108 | return scalar(@{$s->[1]}); |
| 109 | } |
| 110 | |
| 111 | sub Push2 { |
| 112 | my($s) = shift; |
| 113 | $s->Splice($#{$s->[1]}+1, 0, @_); |
| 114 | return scalar(@{$s->[1]}); |
| 115 | } |
| 116 | |
| 117 | # |
| 118 | # pop last k-v pair |
| 119 | # |
| 120 | sub Pop { |
| 121 | my($s) = shift; |
| 122 | my($k, $v, $i); |
| 123 | $k = pop(@{$s->[1]}); |
| 124 | $v = pop(@{$s->[2]}); |
| 125 | if (defined $k) { |
| 126 | delete $s->[0]{$k}; |
| 127 | return ($k, $v); |
| 128 | } |
| 129 | return undef; |
| 130 | } |
| 131 | |
| 132 | sub Pop2 { |
| 133 | return $_[0]->Splice(-1); |
| 134 | } |
| 135 | |
| 136 | # |
| 137 | # shift |
| 138 | # |
| 139 | sub Shift { |
| 140 | my($s) = shift; |
| 141 | my($k, $v, $i); |
| 142 | $k = shift(@{$s->[1]}); |
| 143 | $v = shift(@{$s->[2]}); |
| 144 | if (defined $k) { |
| 145 | delete $s->[0]{$k}; |
| 146 | for (keys %{$s->[0]}) { |
| 147 | $s->[0]{$_}--; |
| 148 | } |
| 149 | return ($k, $v); |
| 150 | } |
| 151 | return undef; |
| 152 | } |
| 153 | |
| 154 | sub Shift2 { |
| 155 | return $_[0]->Splice(0, 1); |
| 156 | } |
| 157 | |
| 158 | # |
| 159 | # unshift |
| 160 | # if a supplied key exists, it will not be reordered |
| 161 | # |
| 162 | sub Unshift { |
| 163 | my($s) = shift; |
| 164 | my($k, $v, @k, @v, $len, $i); |
| 165 | |
| 166 | while (@_) { |
| 167 | ($k, $v) = (shift, shift); |
| 168 | if (exists $s->[0]{$k}) { |
| 169 | $i = $s->[0]{$k}; |
| 170 | $s->[1][$i] = $k; |
| 171 | $s->[2][$i] = $v; |
| 172 | $s->[0]{$k} = $i; |
| 173 | } |
| 174 | else { |
| 175 | push(@k, $k); |
| 176 | push(@v, $v); |
| 177 | $len++; |
| 178 | } |
| 179 | } |
| 180 | if (defined $len) { |
| 181 | for (keys %{$s->[0]}) { |
| 182 | $s->[0]{$_} += $len; |
| 183 | } |
| 184 | $i = 0; |
| 185 | for (@k) { |
| 186 | $s->[0]{$_} = $i++; |
| 187 | } |
| 188 | unshift(@{$s->[1]}, @k); |
| 189 | return unshift(@{$s->[2]}, @v); |
| 190 | } |
| 191 | return scalar(@{$s->[1]}); |
| 192 | } |
| 193 | |
| 194 | sub Unshift2 { |
| 195 | my($s) = shift; |
| 196 | $s->Splice(0,0,@_); |
| 197 | return scalar(@{$s->[1]}); |
| 198 | } |
| 199 | |
| 200 | # |
| 201 | # splice |
| 202 | # |
| 203 | # any existing hash key order is preserved. the value is replaced for |
| 204 | # such keys, and the new keys are spliced in the regular fashion. |
| 205 | # |
| 206 | # supports -ve offsets but only +ve lengths |
| 207 | # |
| 208 | # always assumes a 0 start offset |
| 209 | # |
| 210 | sub Splice { |
| 211 | my($s, $start, $len) = (shift, shift, shift); |
| 212 | my($k, $v, @k, @v, @r, $i, $siz); |
| 213 | my($end); # inclusive |
| 214 | |
| 215 | # XXX inline this |
| 216 | ($start, $end, $len) = $s->_lrange($start, $len); |
| 217 | |
| 218 | if (defined $start) { |
| 219 | if ($len > 0) { |
| 220 | my(@k) = splice(@{$s->[1]}, $start, $len); |
| 221 | my(@v) = splice(@{$s->[2]}, $start, $len); |
| 222 | while (@k) { |
| 223 | $k = shift(@k); |
| 224 | delete $s->[0]{$k}; |
| 225 | push(@r, $k, shift(@v)); |
| 226 | } |
| 227 | for ($start..$#{$s->[1]}) { |
| 228 | $s->[0]{$s->[1][$_]} -= $len; |
| 229 | } |
| 230 | } |
| 231 | while (@_) { |
| 232 | ($k, $v) = (shift, shift); |
| 233 | if (exists $s->[0]{$k}) { |
| 234 | # $s->STORE($k, $v); |
| 235 | $i = $s->[0]{$k}; |
| 236 | $s->[1][$i] = $k; |
| 237 | $s->[2][$i] = $v; |
| 238 | $s->[0]{$k} = $i; |
| 239 | } |
| 240 | else { |
| 241 | push(@k, $k); |
| 242 | push(@v, $v); |
| 243 | $siz++; |
| 244 | } |
| 245 | } |
| 246 | if (defined $siz) { |
| 247 | for ($start..$#{$s->[1]}) { |
| 248 | $s->[0]{$s->[1][$_]} += $siz; |
| 249 | } |
| 250 | $i = $start; |
| 251 | for (@k) { |
| 252 | $s->[0]{$_} = $i++; |
| 253 | } |
| 254 | splice(@{$s->[1]}, $start, 0, @k); |
| 255 | splice(@{$s->[2]}, $start, 0, @v); |
| 256 | } |
| 257 | } |
| 258 | return @r; |
| 259 | } |
| 260 | |
| 261 | # |
| 262 | # delete elements specified by key |
| 263 | # other elements higher than the one deleted "slide" down |
| 264 | # |
| 265 | sub Delete { |
| 266 | my($s) = shift; |
| 267 | |
| 268 | for (@_) { |
| 269 | # |
| 270 | # XXX potential optimization: could do $s->DELETE only if $#_ < 4. |
| 271 | # otherwise, should reset all the hash indices in one loop |
| 272 | # |
| 273 | $s->DELETE($_); |
| 274 | } |
| 275 | } |
| 276 | |
| 277 | # |
| 278 | # replace hash element at specified index |
| 279 | # |
| 280 | # if the optional key is not supplied the value at index will simply be |
| 281 | # replaced without affecting the order. |
| 282 | # |
| 283 | # if an element with the supplied key already exists, it will be deleted first. |
| 284 | # |
| 285 | # returns the key of replaced value if it succeeds. |
| 286 | # |
| 287 | sub Replace { |
| 288 | my($s) = shift; |
| 289 | my($i, $v, $k) = (shift, shift, shift); |
| 290 | if (defined $i and $i <= $#{$s->[1]} and $i >= 0) { |
| 291 | if (defined $k) { |
| 292 | delete $s->[0]{ $s->[1][$i] }; |
| 293 | $s->DELETE($k) ; #if exists $s->[0]{$k}; |
| 294 | $s->[1][$i] = $k; |
| 295 | $s->[2][$i] = $v; |
| 296 | $s->[0]{$k} = $i; |
| 297 | return $k; |
| 298 | } |
| 299 | else { |
| 300 | $s->[2][$i] = $v; |
| 301 | return $s->[1][$i]; |
| 302 | } |
| 303 | } |
| 304 | return undef; |
| 305 | } |
| 306 | |
| 307 | # |
| 308 | # Given an $start and $len, returns a legal start and end (where start <= end) |
| 309 | # for the current hash. |
| 310 | # Legal range is defined as 0 to $#s+1 |
| 311 | # $len defaults to number of elts upto end of list |
| 312 | # |
| 313 | # 0 1 2 ... |
| 314 | # | X | X | X ... X | X | X | |
| 315 | # -2 -1 (no -0 alas) |
| 316 | # X's above are the elements |
| 317 | # |
| 318 | sub _lrange { |
| 319 | my($s) = shift; |
| 320 | my($offset, $len) = @_; |
| 321 | my($start, $end); # both inclusive |
| 322 | my($size) = $#{$s->[1]}+1; |
| 323 | |
| 324 | return undef unless defined $offset; |
| 325 | if($offset < 0) { |
| 326 | $start = $offset + $size; |
| 327 | $start = 0 if $start < 0; |
| 328 | } |
| 329 | else { |
| 330 | ($offset > $size) ? ($start = $size) : ($start = $offset); |
| 331 | } |
| 332 | |
| 333 | if (defined $len) { |
| 334 | $len = -$len if $len < 0; |
| 335 | $len = $size - $start if $len > $size - $start; |
| 336 | } |
| 337 | else { |
| 338 | $len = $size - $start; |
| 339 | } |
| 340 | $end = $start + $len - 1; |
| 341 | |
| 342 | return ($start, $end, $len); |
| 343 | } |
| 344 | |
| 345 | # |
| 346 | # Return keys at supplied indices |
| 347 | # Returns all keys if no args. |
| 348 | # |
| 349 | sub Keys { |
| 350 | my($s) = shift; |
| 351 | return ( @_ == 1 |
| 352 | ? $s->[1][$_[0]] |
| 353 | : ( @_ |
| 354 | ? @{$s->[1]}[@_] |
| 355 | : @{$s->[1]} ) ); |
| 356 | } |
| 357 | |
| 358 | # |
| 359 | # Returns values at supplied indices |
| 360 | # Returns all values if no args. |
| 361 | # |
| 362 | sub Values { |
| 363 | my($s) = shift; |
| 364 | return ( @_ == 1 |
| 365 | ? $s->[2][$_[0]] |
| 366 | : ( @_ |
| 367 | ? @{$s->[2]}[@_] |
| 368 | : @{$s->[2]} ) ); |
| 369 | } |
| 370 | |
| 371 | # |
| 372 | # get indices of specified hash keys |
| 373 | # |
| 374 | sub Indices { |
| 375 | my($s) = shift; |
| 376 | return ( @_ == 1 ? $s->[0]{$_[0]} : @{$s->[0]}{@_} ); |
| 377 | } |
| 378 | |
| 379 | # |
| 380 | # number of k-v pairs in the ixhash |
| 381 | # note that this does not equal the highest index |
| 382 | # owing to preextended arrays |
| 383 | # |
| 384 | sub Length { |
| 385 | return scalar @{$_[0]->[1]}; |
| 386 | } |
| 387 | |
| 388 | # |
| 389 | # Reorder the hash in the supplied key order |
| 390 | # |
| 391 | # warning: any unsupplied keys will be lost from the hash |
| 392 | # any supplied keys that dont exist in the hash will be ignored |
| 393 | # |
| 394 | sub Reorder { |
| 395 | my($s) = shift; |
| 396 | my(@k, @v, %x, $i); |
| 397 | return unless @_; |
| 398 | |
| 399 | $i = 0; |
| 400 | for (@_) { |
| 401 | if (exists $s->[0]{$_}) { |
| 402 | push(@k, $_); |
| 403 | push(@v, $s->[2][ $s->[0]{$_} ] ); |
| 404 | $x{$_} = $i++; |
| 405 | } |
| 406 | } |
| 407 | $s->[1] = \@k; |
| 408 | $s->[2] = \@v; |
| 409 | $s->[0] = \%x; |
| 410 | return $s; |
| 411 | } |
| 412 | |
| 413 | sub SortByKey { |
| 414 | my($s) = shift; |
| 415 | $s->Reorder(sort $s->Keys); |
| 416 | } |
| 417 | |
| 418 | sub SortByValue { |
| 419 | my($s) = shift; |
| 420 | $s->Reorder(sort { $s->FETCH($a) cmp $s->FETCH($b) } $s->Keys) |
| 421 | } |
| 422 | |
| 423 | 1; |
| 424 | __END__ |
| 425 | |
| 426 | =head1 NAME |
| 427 | |
| 428 | Tie::IxHash - ordered associative arrays for Perl |
| 429 | |
| 430 | |
| 431 | =head1 SYNOPSIS |
| 432 | |
| 433 | # simple usage |
| 434 | use Tie::IxHash; |
| 435 | tie HASHVARIABLE, Tie::IxHash [, LIST]; |
| 436 | |
| 437 | # OO interface with more powerful features |
| 438 | use Tie::IxHash; |
| 439 | TIEOBJECT = Tie::IxHash->new( [LIST] ); |
| 440 | TIEOBJECT->Splice( OFFSET [, LENGTH [, LIST]] ); |
| 441 | TIEOBJECT->Push( LIST ); |
| 442 | TIEOBJECT->Pop; |
| 443 | TIEOBJECT->Shift; |
| 444 | TIEOBJECT->Unshift( LIST ); |
| 445 | TIEOBJECT->Keys( [LIST] ); |
| 446 | TIEOBJECT->Values( [LIST] ); |
| 447 | TIEOBJECT->Indices( LIST ); |
| 448 | TIEOBJECT->Delete( [LIST] ); |
| 449 | TIEOBJECT->Replace( OFFSET, VALUE, [KEY] ); |
| 450 | TIEOBJECT->Reorder( LIST ); |
| 451 | TIEOBJECT->SortByKey; |
| 452 | TIEOBJECT->SortByValue; |
| 453 | TIEOBJECT->Length; |
| 454 | |
| 455 | |
| 456 | =head1 DESCRIPTION |
| 457 | |
| 458 | This Perl module implements Perl hashes that preserve the order in which the |
| 459 | hash elements were added. The order is not affected when values |
| 460 | corresponding to existing keys in the IxHash are changed. The elements can |
| 461 | also be set to any arbitrary supplied order. The familiar perl array |
| 462 | operations can also be performed on the IxHash. |
| 463 | |
| 464 | |
| 465 | =head2 Standard C<TIEHASH> Interface |
| 466 | |
| 467 | The standard C<TIEHASH> mechanism is available. This interface is |
| 468 | recommended for simple uses, since the usage is exactly the same as |
| 469 | regular Perl hashes after the C<tie> is declared. |
| 470 | |
| 471 | |
| 472 | =head2 Object Interface |
| 473 | |
| 474 | This module also provides an extended object-oriented interface that can be |
| 475 | used for more powerful operations with the IxHash. The following methods |
| 476 | are available: |
| 477 | |
| 478 | =over 8 |
| 479 | |
| 480 | =item FETCH, STORE, DELETE, EXISTS |
| 481 | |
| 482 | These standard C<TIEHASH> methods mandated by Perl can be used directly. |
| 483 | See the C<tie> entry in perlfunc(1) for details. |
| 484 | |
| 485 | =item Push, Pop, Shift, Unshift, Splice |
| 486 | |
| 487 | These additional methods resembling Perl functions are available for |
| 488 | operating on key-value pairs in the IxHash. The behavior is the same as the |
| 489 | corresponding perl functions, except when a supplied hash key already exists |
| 490 | in the hash. In that case, the existing value is updated but its order is |
| 491 | not affected. To unconditionally alter the order of a supplied key-value |
| 492 | pair, first C<DELETE> the IxHash element. |
| 493 | |
| 494 | =item Keys |
| 495 | |
| 496 | Returns an array of IxHash element keys corresponding to the list of supplied |
| 497 | indices. Returns an array of all the keys if called without arguments. |
| 498 | Note the return value is mostly only useful when used in a list context |
| 499 | (since perl will convert it to the number of elements in the array when |
| 500 | used in a scalar context, and that may not be very useful). |
| 501 | |
| 502 | If a single argument is given, returns the single key corresponding to |
| 503 | the index. This is usable in either scalar or list context. |
| 504 | |
| 505 | =item Values |
| 506 | |
| 507 | Returns an array of IxHash element values corresponding to the list of supplied |
| 508 | indices. Returns an array of all the values if called without arguments. |
| 509 | Note the return value is mostly only useful when used in a list context |
| 510 | (since perl will convert it to the number of elements in the array when |
| 511 | used in a scalar context, and that may not be very useful). |
| 512 | |
| 513 | If a single argument is given, returns the single value corresponding to |
| 514 | the index. This is usable in either scalar or list context. |
| 515 | |
| 516 | =item Indices |
| 517 | |
| 518 | Returns an array of indices corresponding to the supplied list of keys. |
| 519 | Note the return value is mostly only useful when used in a list context |
| 520 | (since perl will convert it to the number of elements in the array when |
| 521 | used in a scalar context, and that may not be very useful). |
| 522 | |
| 523 | If a single argument is given, returns the single index corresponding to |
| 524 | the key. This is usable in either scalar or list context. |
| 525 | |
| 526 | =item Delete |
| 527 | |
| 528 | Removes elements with the supplied keys from the IxHash. |
| 529 | |
| 530 | =item Replace |
| 531 | |
| 532 | Substitutes the IxHash element at the specified index with the supplied |
| 533 | value-key pair. If a key is not supplied, simply substitutes the value at |
| 534 | index with the supplied value. If an element with the supplied key already |
| 535 | exists, it will be removed from the IxHash first. |
| 536 | |
| 537 | =item Reorder |
| 538 | |
| 539 | This method can be used to manipulate the internal order of the IxHash |
| 540 | elements by supplying a list of keys in the desired order. Note however, |
| 541 | that any IxHash elements whose keys are not in the list will be removed from |
| 542 | the IxHash. |
| 543 | |
| 544 | =item Length |
| 545 | |
| 546 | Returns the number of IxHash elements. |
| 547 | |
| 548 | =item SortByKey |
| 549 | |
| 550 | Reorders the IxHash elements by textual comparison of the keys. |
| 551 | |
| 552 | =item SortByValue |
| 553 | |
| 554 | Reorders the IxHash elements by textual comparison of the values. |
| 555 | |
| 556 | =back |
| 557 | |
| 558 | |
| 559 | =head1 EXAMPLE |
| 560 | |
| 561 | use Tie::IxHash; |
| 562 | |
| 563 | # simple interface |
| 564 | $t = tie(%myhash, Tie::IxHash, 'a' => 1, 'b' => 2); |
| 565 | %myhash = (first => 1, second => 2, third => 3); |
| 566 | $myhash{fourth} = 4; |
| 567 | @keys = keys %myhash; |
| 568 | @values = values %myhash; |
| 569 | print("y") if exists $myhash{third}; |
| 570 | |
| 571 | # OO interface |
| 572 | $t = Tie::IxHash->new(first => 1, second => 2, third => 3); |
| 573 | $t->Push(fourth => 4); # same as $myhash{'fourth'} = 4; |
| 574 | ($k, $v) = $t->Pop; # $k is 'fourth', $v is 4 |
| 575 | $t->Unshift(neg => -1, zeroth => 0); |
| 576 | ($k, $v) = $t->Shift; # $k is 'neg', $v is -1 |
| 577 | @oneandtwo = $t->Splice(1, 2, foo => 100, bar => 101); |
| 578 | |
| 579 | @keys = $t->Keys; |
| 580 | @values = $t->Values; |
| 581 | @indices = $t->Indices('foo', 'zeroth'); |
| 582 | @itemkeys = $t->Keys(@indices); |
| 583 | @itemvals = $t->Values(@indices); |
| 584 | $t->Replace(2, 0.3, 'other'); |
| 585 | $t->Delete('second', 'zeroth'); |
| 586 | $len = $t->Length; # number of key-value pairs |
| 587 | |
| 588 | $t->Reorder(reverse @keys); |
| 589 | $t->SortByKey; |
| 590 | $t->SortByValue; |
| 591 | |
| 592 | |
| 593 | =head1 BUGS |
| 594 | |
| 595 | You cannot specify a negative length to C<Splice>. Negative indexes are OK, |
| 596 | though. |
| 597 | |
| 598 | Indexing always begins at 0 (despite the current C<$[> setting) for |
| 599 | all the functions. |
| 600 | |
| 601 | |
| 602 | =head1 TODO |
| 603 | |
| 604 | Addition of elements with keys that already exist to the end of the IxHash |
| 605 | must be controlled by a switch. |
| 606 | |
| 607 | Provide C<TIEARRAY> interface when it stabilizes in Perl. |
| 608 | |
| 609 | Rewrite using XSUBs for efficiency. |
| 610 | |
| 611 | |
| 612 | =head1 AUTHOR |
| 613 | |
| 614 | Gurusamy Sarathy gsar@umich.edu |
| 615 | |
| 616 | Copyright (c) 1995 Gurusamy Sarathy. All rights reserved. |
| 617 | This program is free software; you can redistribute it and/or |
| 618 | modify it under the same terms as Perl itself. |
| 619 | |
| 620 | |
| 621 | =head1 VERSION |
| 622 | |
| 623 | Version 1.21 20 Nov 1997 |
| 624 | |
| 625 | |
| 626 | =head1 SEE ALSO |
| 627 | |
| 628 | perl(1) |
| 629 | |
| 630 | =cut |