/usr/include/sdsl/wt_gmr.hpp is in libsdsl-dev 2.0.3-4.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 | /* sdsl - succinct data structures library
Copyright (C) 2014 Simon Gog
This program is free software: you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation, either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see http://www.gnu.org/licenses/ .
*/
/*! \file wt_gmr.hpp
\brief wt_gmr.hpp contains a specialized class to support select, rank
and access on inputs over a large alphabet.
\author Alexander Diehm, Timo Beller, Simon Gog
*/
#ifndef INCLUDED_SDSL_WT_GMR
#define INCLUDED_SDSL_WT_GMR
#include <sdsl/bit_vectors.hpp>
#include <sdsl/int_vector.hpp>
#include <sdsl/vectors.hpp>
//! Namespace for the succinct data structure library.
namespace sdsl
{
//! Class inv_multi_perm_support adds access to the inverse of permutations.
/*!
* \tparam t_s Sampling parameter of the inverse permutation.
* \tparam t_rac Type of the random access container used for storing the permutation.
* \tparam t_bv Type of the bitvector used to indicate back-pointers.
* \tparam t_rank Type of rank_support to rank the indicator bitvector.
*
* This support class adds access to the inverse of permutations in at
* most \(t_s\) steps.
*
* \par References
* [1] J. Munro, R. Raman, V. Raman, S. Rao: ,,Succinct representation
* of permutations'', Proceedings of ICALP 2003
*/
template<uint64_t t_s=32,
class t_rac=int_vector<>,
class t_bv=bit_vector,
class t_rank=typename t_bv::rank_1_type>
class inv_multi_perm_support
{
public:
typedef t_rac iv_type;
typedef typename iv_type::size_type size_type;
typedef typename iv_type::value_type value_type;
typedef typename iv_type::difference_type difference_type;
typedef t_bv bit_vector_type;
typedef t_rank rank_type;
typedef random_access_const_iterator<inv_multi_perm_support> const_iterator;
private:
const iv_type* m_perm = nullptr;// pointer to supported permutation
uint64_t m_chunksize; // size of one permutation
int_vector<> m_back_pointer; // back pointers
bit_vector_type m_marked; // back pointer marking
rank_type m_marked_rank; // rank support for back pointer marking
public:
//! Default constructor
inv_multi_perm_support() {};
//! Constructor
inv_multi_perm_support(const iv_type* perm, int_vector<>& iv, uint64_t chunksize) : m_perm(perm), m_chunksize(chunksize) {
bit_vector marked(iv.size(), 0);
bit_vector done(m_chunksize, 0);
size_type max_back_pointer = 0;
for (size_type i=0, off=0; i < iv.size(); ++i) {
if (i == off+chunksize) {
off = i;
util::set_to_value(done, 0);
}
if (!done[i-off]) {
done[i-off] = 1;
size_type back_pointer=i, j = i, j_new=0;
uint64_t steps = 0, all_steps = 0;
while ((j_new=(iv[j]+off)) != i) {
j = j_new;
done[j-off] = 1;
++steps; ++all_steps;
if (t_s == steps) {
max_back_pointer = std::max(max_back_pointer, back_pointer-off);
marked[j] = 1;
steps = 0;
back_pointer = j;
}
}
if (all_steps > t_s) {
marked[i] = 1;
max_back_pointer = std::max(max_back_pointer, back_pointer-off);
}
}
}
m_marked = t_bv(std::move(marked));
util::init_support(m_marked_rank, &m_marked);
util::set_to_value(done, 0);
size_type n_bp = m_marked_rank(iv.size());
m_back_pointer = int_vector<>(n_bp, 0, bits::hi(max_back_pointer)+1);
for (size_type i=0, off=0; i < iv.size(); ++i) {
if (i == off+chunksize) {
off = i;
util::set_to_value(done, 0);
}
if (!done[i-off]) {
done[i-off] = 1;
size_type back_pointer = i, j = i, j_new=0;
uint64_t steps = 0, all_steps = 0;
while ((j_new=(iv[j]+off)) != i) {
j = j_new;
done[j-off] = 1;
++steps; ++all_steps;
if (t_s == steps) {
m_back_pointer[m_marked_rank(j)] = back_pointer-off;
steps = 0;
back_pointer = j;
}
}
if (all_steps > t_s) {
m_back_pointer[m_marked_rank(i)] = back_pointer-off;
}
}
}
}
//! Copy constructor
inv_multi_perm_support(const inv_multi_perm_support& p) : m_perm(p.m_perm),
m_chunksize(p.m_chunksize), m_back_pointer(p.m_back_pointer), m_marked(p.m_marked),
m_marked_rank(p.m_marked_rank) {
m_marked_rank.set_vector(&m_marked);
}
//! Move constructor
inv_multi_perm_support(inv_multi_perm_support&& p) {
*this = std::move(p);
}
//! Assignment operation
inv_multi_perm_support& operator=(const inv_multi_perm_support& p) {
if (this != &p) {
m_perm = p.m_perm;
m_chunksize = p.m_chunksize;
m_back_pointer = p.m_back_pointer;
m_marked = p.m_marked;
m_marked_rank = p.m_marked_rank;
m_marked_rank.set_vector(&m_marked);
}
return *this;
}
//! Assignment move operation
inv_multi_perm_support& operator=(inv_multi_perm_support&& p) {
if (this != &p) {
m_perm = std::move(p.m_perm);
m_chunksize = std::move(p.m_chunksize);
m_back_pointer = std::move(p.m_back_pointer);
m_marked = std::move(p.m_marked);
m_marked_rank = std::move(p.m_marked_rank);
m_marked_rank.set_vector(&m_marked);
}
return *this;
}
//! Swap operation
void swap(inv_multi_perm_support& p) {
if (this != &p) {
std::swap(m_chunksize, p.m_chunksize);
m_back_pointer.swap(p.m_back_pointer);
m_marked.swap(p.m_marked);
util::swap_support(m_marked_rank, p.m_marked_rank, &m_marked, &(p.m_marked));
}
}
//! Returns the size of the original vector.
size_type size() const {
return nullptr == m_perm ? 0 : m_perm->size();
}
//! Returns whether the original vector contains no data.
bool empty()const {
return size() == 0;
}
//! Access operator
/*
* \par Time complexity
* \f$ \Order{t_s} \f$
*/
value_type operator[](size_type i) const {
size_type off = (i/m_chunksize)*m_chunksize;
size_type j = i, j_new=0;
while ((j_new=((*m_perm)[j])+off) != i) {
if (m_marked[j]) {
j = m_back_pointer[m_marked_rank(j)]+off;
while ((j_new=((*m_perm)[j])+off) != i) j = j_new;
} else {
j = j_new;
}
}
return j;
}
//! Returns a const_iterator to the first element.
const_iterator begin()const {
return const_iterator(this, 0);
}
//! Returns a const_iterator to the element after the last element.
const_iterator end()const {
return const_iterator(this, size());
}
void set_vector(const iv_type* v) { m_perm = v; }
//! Serialize into stream
size_type serialize(std::ostream& out, structure_tree_node* v=nullptr, std::string name="")const {
structure_tree_node* child = structure_tree::add_child(v, name, util::class_name(*this));
size_type written_bytes = 0;
written_bytes += write_member(m_chunksize, out, child, "chunksize");
written_bytes += m_back_pointer.serialize(out, child, "back_pointer");
written_bytes += m_marked.serialize(out, child, "marked");
written_bytes += m_marked_rank.serialize(out, child, "marked_rank");
structure_tree::add_size(child, written_bytes);
return written_bytes;
}
//! Load sampling from disk
void load(std::istream& in, const iv_type* v=nullptr) {
set_vector(v);
read_member(m_chunksize, in);
m_back_pointer.load(in);
m_marked.load(in);
m_marked_rank.load(in, &m_marked);
}
};
template<class t_rac>
void
_transform_to_compressed(int_vector<>& iv, typename std::enable_if<!(std::is_same<t_rac, int_vector<>>::value),
t_rac>::type& rac, const std::string filename)
{
std::string tmp_file_name = tmp_file(filename, "_compress_int_vector");
store_to_file(iv, tmp_file_name);
util::clear(iv);
int_vector_buffer<> buf(tmp_file_name, std::ios::in, 1024*1024, iv.width());
rac = t_rac(buf);
buf.close(true); // delete tmp_file
}
template<class t_rac>
void
_transform_to_compressed(int_vector<>& iv, typename std::enable_if<std::is_same<t_rac, int_vector<>>::value,
t_rac>::type& rac, const std::string)
{
rac = std::move(iv);
}
//! A wavelet tree class for integer sequences.
/*!
* \tparam t_rac Type of the random access container used for E.
* \tparam t_bitvector Type of the bitvector used for storing B.
* \tparam t_select Type of the support structure for select on pattern `1`.
* \tparam t_select_zero Type of the support structure for select on pattern `0`.
*
* This is an implementation of the first proposal in the SODA paper of Golynski et. al.
* which support fast rank and select, but not fast access.
*
* \par References
* [1] A. Golynski, J. Munro and S. Rao:
* ,,Rank/select operations on large alphabets: a tool for text indexing''
* Proceedings of SODA 2006.
*
* @ingroup wt
*/
template<class t_rac = int_vector<>,
class t_bitvector = bit_vector,
class t_select = typename t_bitvector::select_1_type,
class t_select_zero = typename t_bitvector::select_0_type>
class wt_gmr_rs
{
public:
typedef int_vector<>::size_type size_type;
typedef int_vector<>::value_type value_type;
typedef wt_tag index_category;
typedef int_alphabet_tag alphabet_category;
enum {lex_ordered=0};
private:
t_bitvector m_bv_blocks;
t_rac m_e;
t_select m_bv_blocks_select1;
t_select_zero m_bv_blocks_select0;
uint64_t m_size; // input length
uint64_t m_block_size = 0; // size of the blocks
uint64_t m_blocks; // blocks per character
uint64_t m_sigma = 0;
public:
const size_type& sigma = m_sigma;
//! Default constructor
wt_gmr_rs() {}
//! Semi-external constructor
/*! \param buf File buffer of the int_vector for which the wt_gmr should be build.
* \param size Size of the prefix of v, which should be indexed.
*/
template<uint8_t int_width>
wt_gmr_rs(int_vector_buffer<int_width>& input, size_type size) : m_size(size) {
// Determine max. symbol
for (uint64_t i=0; i<m_size; ++i) {
if (m_block_size < input[i]) m_block_size = input[i];
}
++m_block_size;
// Create and fill m_bv_blocks
m_blocks = (m_size+m_block_size-1)/m_block_size;
bit_vector b(m_size+m_block_size*m_blocks+1, 0);
int_vector<> symbols(m_block_size, 0, bits::hi(m_size)+1);
{
int_vector<> tmp(m_block_size*m_blocks, 0, bits::hi(m_block_size)+1);
for (uint64_t i=0, offset=0, j=0; i<m_size; ++i, ++j) {
if (j==m_block_size) {
++offset;
j = 0;
}
++tmp[input[i]*m_blocks+offset];
}
for (uint64_t i=0; i<symbols.size(); ++i) {
for (uint64_t j=m_blocks*i; j<(i+1)*m_blocks; ++j) {
symbols[i] += tmp[j];
}
}
for (uint64_t i=0,l=1; i<tmp.size(); ++i,++l) {
for (uint64_t j=0; j<tmp[i]; ++j)
b[l++]=1;
}
// calc m_sigma
bool write = true;
uint64_t blocks = 0;
for (uint64_t i=1; i<b.size(); ++i) {
if (blocks==m_blocks) {
blocks = 0;
write = true;
}
if (b[i]) {
if (write) {
++m_sigma;
write = false;
}
} else ++blocks;
}
m_bv_blocks = t_bitvector(std::move(b));
}
// Create and fill e
int_vector<> positions(m_size, 0, bits::hi(m_block_size)+1);
for (uint64_t i=0, tmp=0, sum=0; i<m_block_size; ++i) {
tmp = symbols[i];
symbols[i] = sum;
sum += tmp;
}
for (uint64_t i=0; i<m_size;) {
for (uint64_t j=0; j<m_block_size and i<m_size; ++i, ++j) {
positions[symbols[input[i]]++] = j;
}
}
_transform_to_compressed<t_rac>(positions, m_e, input.filename());
util::init_support(m_bv_blocks_select0, &m_bv_blocks);
util::init_support(m_bv_blocks_select1, &m_bv_blocks);
}
//! Copy constructor
wt_gmr_rs(const wt_gmr_rs& wt) {
m_bv_blocks = wt.m_bv_blocks;
m_e = wt.m_e;
m_bv_blocks_select1 = wt.m_bv_blocks_select1;
m_bv_blocks_select1.set_vector(&m_bv_blocks);
m_bv_blocks_select0 = wt.m_bv_blocks_select0;
m_bv_blocks_select0.set_vector(&m_bv_blocks);
m_size = wt.m_size;
m_block_size = wt.m_block_size;
m_blocks = wt.m_blocks;
m_sigma = wt.m_sigma;
}
//! Assignment operator
wt_gmr_rs& operator=(const wt_gmr_rs& wt) {
wt_gmr_rs tmp(wt);
tmp.swap(*this);
return *this;
}
//! Swap operator
void swap(wt_gmr_rs& fs) {
if (this != &fs) {
m_bv_blocks.swap(fs.m_bv_blocks);
m_e.swap(fs.m_e);
util::swap_support(m_bv_blocks_select0, fs.m_bv_blocks_select0, &m_bv_blocks, &(fs.m_bv_blocks));
util::swap_support(m_bv_blocks_select1, fs.m_bv_blocks_select1, &m_bv_blocks, &(fs.m_bv_blocks));
std::swap(m_size, fs.m_size);
std::swap(m_block_size, fs.m_block_size);
std::swap(m_blocks, fs.m_blocks);
std::swap(m_sigma, fs.m_sigma);
}
}
//! Returns the size of the original vector.
size_type size()const {
return m_size;
}
//! Returns whether the wavelet tree contains no data.
bool empty()const {
return m_size == 0;
}
//! Recovers the i-th symbol of the original vector.
/*! \param i The index of the symbol in the original vector.
* \returns The i-th symbol of the original vector.
* \par Time complexity
* \f$ \Order{|\Sigma|} \f$
* \par Precondition
* \f$ i < size() \f$
*/
value_type operator[](size_type i)const {
assert(i<m_size);
size_type block=i/m_block_size+1, val=i%m_block_size, search_begin, search_end, j;
while (true) {
j = m_bv_blocks_select0(block)+1;
search_begin = j-block;
if (m_bv_blocks[j]) {
search_end = m_bv_blocks_select0(block+1)-(block);
if (search_end-search_begin<50) { // After a short test, this seems to be a good threshold
while (search_begin < search_end and m_e[search_begin] <= val) {
if (m_e[search_begin]==val) {
return (block-1)/m_blocks;
}
++search_begin;
}
} else {
if (binary_search(m_e.begin()+search_begin, m_e.begin()+search_end, val)) {
return (block-1)/m_blocks;
}
}
}
block += m_blocks;
}
}
//! Calculates how many symbols c are in the prefix [0..i-1] of the supported vector.
/*!
* \param i The exclusive index of the prefix range [0..i-1], so \f$i\in[0..size()]\f$.
* \param c The symbol to count the occurrences in the prefix.
* \returns The number of occurrences of symbol c in the prefix [0..i-1] of the supported vector.
* \par Time complexity
* \f$ \Order{\log |\Sigma|} \f$
* \par Precondition
* \f$ i \leq size() \f$
*/
size_type rank(size_type i, value_type c)const {
if (0==i or c>m_block_size-1) {
return 0;
}
size_type offset=0;
size_type ones_before_cblock = m_bv_blocks_select0(c*m_blocks+1)-c*m_blocks;
auto begin = m_e.begin()+m_bv_blocks_select0(c*m_blocks+(i-1)/m_block_size+1)-(c*m_blocks+(i-1)/m_block_size+1)+1;
auto end = m_e.begin()+m_bv_blocks_select0(c*m_blocks+(i-1)/m_block_size+2)-(c*m_blocks+(i-1)/m_block_size+1);
size_type val = (i-1)%m_block_size;
if (end-begin<50) { // After a short test, this seems to be a good threshold
offset = std::find_if(begin, end, [&val](const decltype(*begin) x) { return x > val; }) - begin;
} else {
offset = lower_bound(begin, end, val+1)-begin;
}
return (begin-m_e.begin())+offset-ones_before_cblock;
}
//! Calculates how many symbols c are in the prefix [0..i-1] of the supported vector.
/*!
* \param i The exclusive index of the prefix range [0..i-1], so \f$i\in[0..size()]\f$.
* \param c The symbol to count the occurrences in the prefix.
* \returns The number of occurrences of symbol c in the prefix [0..i-1] of the supported vector.
* \par Time complexity
* \f$ \Order{|\Sigma|} \f$
* \par Precondition
* \f$ i \leq size() \f$
*/
std::pair<size_type, value_type> inverse_select(size_type i)const {
assert(i<m_size);
size_type block = i/m_block_size+1, val = i%m_block_size, offset = 0, search_begin, search_end, j;
while (true) {
j = m_bv_blocks_select0(block)+1;
search_begin = j-block;
if (m_bv_blocks[j]) {
search_end = m_bv_blocks_select0(block+1)-(block);
offset = 0;
if (search_end-search_begin<50) { // After a short test, this seems to be a good threshold
while (search_begin < search_end and m_e[search_begin] <= val) {
if (m_e[search_begin]==val) {
value_type c = (block-1)/m_blocks;
size_type ones_before_cblock = m_bv_blocks_select0(c*m_blocks+1)-(c*m_blocks);
size_type r = search_begin-ones_before_cblock;
return std::make_pair(r,c);
}
++search_begin;
}
} else {
offset = lower_bound(m_e.begin()+search_begin, m_e.begin()+search_end, val)-m_e.begin();
if (offset<search_end) {
if (m_e[offset]==val) {
value_type c = (block-1)/m_blocks;
size_type ones_before_cblock = m_bv_blocks_select0(c*m_blocks+1)-(c*m_blocks);
size_type r = offset-ones_before_cblock;
return std::make_pair(r,c);
}
}
}
}
block+=m_blocks;
}
}
//! Calculates the i-th occurrence of the symbol c in the supported vector.
/*!
* \param i The i-th occurrence.
* \param c The symbol c.
* \par Time complexity
* \f$ \Order{1} \f$
* \par Precondition
* \f$ 1 \leq i \leq rank(size(), c) \f$
*/
size_type select(size_type i, value_type c)const {
size_type k = m_bv_blocks_select0(c*m_blocks+1)-(c*m_blocks)+i;
return (m_bv_blocks_select1(k)-k)*m_block_size+m_e[k-1]-c*m_blocks*m_block_size;
}
//! Serializes the data structure into the given ostream
size_type serialize(std::ostream& out, structure_tree_node* v=nullptr, std::string name="")const {
structure_tree_node* child = structure_tree::add_child(v, name, util::class_name(*this));
size_type written_bytes = 0;
written_bytes += write_member(m_size, out, child, "size");
written_bytes += write_member(m_block_size, out, child, "block_size");
written_bytes += write_member(m_blocks, out, child, "blocks");
written_bytes += write_member(m_sigma, out, child, "sigma");
written_bytes += m_e.serialize(out, child, "E");
written_bytes += m_bv_blocks.serialize(out, child, "bv_blocks");
written_bytes += m_bv_blocks_select0.serialize(out, child, "bv_blocks_select0");
written_bytes += m_bv_blocks_select1.serialize(out, child, "bv_blocks_select1");
structure_tree::add_size(child, written_bytes);
return written_bytes;
}
//! Loads the data structure from the given istream.
void load(std::istream& in) {
read_member(m_size, in);
read_member(m_block_size, in);
read_member(m_blocks, in);
read_member(m_sigma, in);
m_e.load(in);
m_bv_blocks.load(in);
m_bv_blocks_select0.load(in, &m_bv_blocks);
m_bv_blocks_select1.load(in, &m_bv_blocks);
}
};
//! A wavelet tree class for integer sequences.
/*!
* \tparam t_rac Type of the random access container used for storing the permutation.
* \tparam t_inv_support Type of the support structure for inverse permutation
* \tparam t_bitvector Type of the bitvector used for storing B and X.
* \tparam t_select Type of the support structure for select on pattern `1`.
* \tparam t_select_zero Type of the support structure for select on pattern `0`.
*
* This is an implementation of the second proposal in the SODA paper of Golynski et. al.
* which supports fast access, inverse select, rank, and select.
*
* \par References
* [1] A. Golynski, J. Munro and S. Rao:
* ,,Rank/select operations on large alphabets: a tool for text indexing''
* Proceedings of SODA 2006.
*
* @ingroup wt
*/
template<class t_rac = int_vector<>,
class t_inverse_support = inv_multi_perm_support<32, t_rac>,
class t_bitvector = bit_vector,
class t_select = typename t_bitvector::select_1_type,
class t_select_zero = typename t_bitvector::select_0_type
>
class wt_gmr
{
public:
typedef typename t_rac::size_type size_type;
typedef typename t_rac::value_type value_type;
typedef wt_tag index_category;
typedef int_alphabet_tag alphabet_category;
enum {lex_ordered=0};
private:
t_bitvector m_bv_blocks; // 0 indicates end of block. Corresponds to B in the paper.
t_bitvector m_bv_chunks; // 0 indicates end of symbol in chunk. Corresponds to X in the paper.
t_rac m_perm; // Contains permutation of each chunk. Corresponds to \f$ \pi \f$ in the paper.
t_inverse_support m_ips; // Support for inverse permutation
t_select m_bv_blocks_select1, m_bv_chunks_select1;
t_select_zero m_bv_blocks_select0, m_bv_chunks_select0;
uint64_t m_size; // input length
uint64_t m_max_symbol = 0; // maximum character + 1
uint64_t m_chunks; // number of chunks
uint64_t m_chunksize;
uint64_t m_sigma = 0;
public:
const size_type& sigma = m_sigma;
//! Default constructor
wt_gmr() {}
//! Semi-external constructor
/*! \param buf File buffer of the int_vector for which the wt_gmr should be build.
* \param size Size of the prefix of v, which should be indexed.
*/
template<uint8_t int_width>
wt_gmr(int_vector_buffer<int_width>& input, size_type size) : m_size(size) {
// Determine max. symbol
for (uint64_t i=0; i<m_size; ++i) {
if (m_max_symbol < input[i]) m_max_symbol = input[i];
}
++m_max_symbol;
m_chunksize = (1 << (bits::hi(m_max_symbol-1)+1)); // In some cases this is better than m_max_smbol
m_chunks = (m_size+m_chunksize-1)/m_chunksize;
// calc m_bv_blocks
{
bit_vector b(m_size+m_max_symbol*m_chunks+1, 0);
int_vector<> tmp(m_max_symbol*m_chunks, 0, bits::hi(m_max_symbol-1)+2);
for (uint64_t i=0, offset=0, j=0; i<m_size; ++i, ++j) {
if (j==m_chunksize) {
++offset;
j = 0;
}
++tmp[input[i]*m_chunks+offset];
}
for (uint64_t i=0, l=1; i<tmp.size(); ++i, ++l)
for (uint64_t j=0; j<tmp[i]; ++j)
b[l++]=1;
// calc m_sigma
bool write = true;
uint64_t blocks = 0;
for (uint64_t i=1; i<b.size(); ++i) {
if (blocks==m_chunks) {
blocks = 0;
write = true;
}
if (b[i]) {
if (write) {
++m_sigma;
write = false;
}
} else ++blocks;
}
m_bv_blocks = t_bitvector(std::move(b));
}
// Calc perm and bv_chunks
{
uint64_t x_pos = 0;
bit_vector x(m_size+m_chunks*m_max_symbol+1, 0);
// fill perm and m_bv_chunks for every chunk
int_vector<> perm(m_size, 0, bits::hi(m_max_symbol-1)+1);
for (uint64_t i=0; i<m_chunks; ++i) {
int_vector<> symbols(m_max_symbol, 0, bits::hi(m_max_symbol-1)+2);
// calc symbols
for (uint64_t j=i*m_chunksize; j<(i+1)*m_chunksize and j<m_size; ++j) {
++symbols[input[j]];
}
// calc m_bv_chunks
for (uint64_t j=0; j<m_max_symbol; ++j, ++x_pos)
for (uint64_t k=0; k<symbols[j]; ++k)
x[++x_pos]=1;
// calc symbols prefix sum
for (uint64_t j=0, tmp=0, sum=0; j<m_max_symbol; ++j) {
tmp = symbols[j];
symbols[j] = sum;
sum += tmp;
}
// calc perm
for (uint64_t j=i* m_chunksize, k=0; j<(i+1)*m_chunksize and j<m_size; ++j, ++k) {
perm[i*m_chunksize+(symbols[input[j]]++)] = k;
}
}
m_bv_chunks = t_bitvector(std::move(x));
m_ips = t_inverse_support(&m_perm, perm, m_chunksize);
_transform_to_compressed<t_rac>(perm, m_perm, input.filename());
m_ips.set_vector(&m_perm);
}
util::init_support(m_bv_chunks_select1, &m_bv_chunks);
util::init_support(m_bv_chunks_select0, &m_bv_chunks);
util::init_support(m_bv_blocks_select1, &m_bv_blocks);
util::init_support(m_bv_blocks_select0, &m_bv_blocks);
}
//! Copy constructor
wt_gmr(const wt_gmr& wt) {
m_bv_blocks = wt.m_bv_blocks;
m_bv_chunks = wt.m_bv_chunks;
m_perm = wt.m_perm;
m_ips = wt.m_ips;
m_bv_blocks_select1 = wt.m_bv_blocks_select1;
m_bv_blocks_select1.set_vector(&m_bv_blocks);
m_bv_chunks_select1 = wt.m_bv_chunks_select1;
m_bv_chunks_select1.set_vector(&m_bv_chunks);
m_bv_blocks_select0 = wt.m_bv_blocks_select0;
m_bv_blocks_select0.set_vector(&m_bv_blocks);
m_bv_chunks_select0 = wt.m_bv_chunks_select0;
m_bv_chunks_select0.set_vector(&m_bv_chunks);
m_size = wt.m_size;
m_max_symbol = wt.m_max_symbol;
m_chunks = wt.m_chunks;
m_chunksize = wt.m_chunksize;
m_sigma = wt.m_sigma;
}
//! Assignment operator
wt_gmr& operator=(const wt_gmr& wt) {
wt_gmr tmp(wt);
tmp.swap(*this);
return *this;
}
//! Swap operator
void swap(wt_gmr& fs) {
if (this != &fs) {
m_bv_blocks.swap(fs.m_bv_blocks);
m_bv_chunks.swap(fs.m_bv_chunks);
m_perm.swap(fs.m_perm);
util::swap_support(m_ips, fs.m_ips, &m_perm, &(fs.m_perm));
util::swap_support(m_bv_blocks_select0, fs.m_bv_blocks_select0, &m_bv_blocks, &(fs.m_bv_blocks));
util::swap_support(m_bv_blocks_select1, fs.m_bv_blocks_select1, &m_bv_blocks, &(fs.m_bv_blocks));
util::swap_support(m_bv_chunks_select1, fs.m_bv_chunks_select1, &m_bv_chunks, &(fs.m_bv_chunks));
util::swap_support(m_bv_chunks_select0, fs.m_bv_chunks_select0, &m_bv_chunks, &(fs.m_bv_chunks));
std::swap(m_size, fs.m_size);
std::swap(m_max_symbol, fs.m_max_symbol);
std::swap(m_chunks, fs.m_chunks);
std::swap(m_chunksize, fs.m_chunksize);
std::swap(m_sigma, fs.m_sigma);
}
}
//! Returns the size of the original vector.
size_type size()const {
return m_size;
}
//! Returns whether the wavelet tree contains no data.
bool empty()const {
return m_size == 0;
}
//! Recovers the i-th symbol of the original vector.
/*! \param i The index of the symbol in the original vector.
* \returns The i-th symbol of the original vector.
* \par Time complexity
* \f$ \Order{1} + 1 Access to the inverse permutation \f$
* \par Precondition
* \f$ i < size() \f$
*/
value_type operator[](size_type i)const {
assert(i < size());
uint64_t chunk = i/m_chunksize;
uint64_t x = m_ips[i];
return m_bv_chunks_select1(x+1)-x-(chunk*m_max_symbol)-1;
}
//! Calculates how many symbols c are in the prefix [0..i-1] of the supported vector.
/*!
* \param i The exclusive index of the prefix range [0..i-1], so \f$i\in[0..size()]\f$.
* \param c The symbol to count the occurrences in the prefix.
* \returns The number of occurrences of symbol c in the prefix [0..i-1] of the supported vector.
* \par Time complexity
* \f$ \Order{\log |\Sigma|} \f$
* \par Precondition
* \f$ i \leq size() \f$
*/
size_type rank(size_type i, value_type c)const {
assert(i <= size());
if (0==i or c>m_max_symbol-1) {
return 0;
}
uint64_t chunk = (i-1)/m_chunksize;
uint64_t ones_before_c = m_bv_blocks_select0(c*m_chunks+1)-(c*m_chunks+1)+1;
uint64_t c_ones_before_chunk = m_bv_blocks_select0(c*m_chunks+chunk+1)-(c*m_chunks+chunk+1)+1-ones_before_c;
uint64_t c_ones_in_chunk = 0;
auto begin = m_perm.begin()+m_bv_chunks_select0(chunk*m_max_symbol+1+c)-(chunk*m_max_symbol+1+c)+1;
auto end = m_perm.begin()+m_bv_chunks_select0(chunk*m_max_symbol+2+c)-(chunk*m_max_symbol+2+c)+1;
size_type val = (i-1)%m_chunksize;
if (end-begin<50) { // After a short test, this seems to be a good threshold
c_ones_in_chunk = std::find_if(begin, end, [&val](const decltype(*begin) x) { return x > val; }) - begin;
} else {
c_ones_in_chunk = lower_bound(begin, end, val+1) - begin;
}
return c_ones_before_chunk+c_ones_in_chunk;
}
//! Calculates how many occurrences of symbol input[i] are in the prefix [0..i-1] of the original input.
/*!
* \param i The index of the symbol.
* \return Pair (rank(input[i],i), input[i])
* \par Time complexity
* \f$ \Order{1} + One access to the inverse permutation \f$
* \par Precondition
* \f$ i < size() \f$
*/
std::pair<size_type, value_type> inverse_select(size_type i)const {
assert(i < size());
uint64_t chunk = i/m_chunksize;
uint64_t x = m_ips[i];
uint64_t tmp = m_bv_chunks_select1(x+1);
uint64_t c = tmp-x-(chunk*m_max_symbol)-1;
uint64_t ones_before_c = m_bv_blocks_select0(c*m_chunks+1)-(c*m_chunks+1)+1;
uint64_t c_before_chunk = m_bv_blocks_select0(c*m_chunks+chunk+1)-(c*m_chunks+chunk+1)+1-ones_before_c;
uint64_t c_in_chunk = tmp-m_bv_chunks_select0(c+1+chunk*m_max_symbol)-1;
return std::make_pair(c_before_chunk+c_in_chunk, c);
}
//! Calculates the i-th occurrence of the symbol c in the supported vector.
/*!
* \param i The i-th occurrence.
* \param c The symbol c.
* \par Time complexity
* \f$ \Order{1} \f$
* \par Precondition
* \f$ 1 \leq i \leq rank(size(), c) \f$
*/
size_type select(size_type i, value_type c)const {
assert(1 <= i and i <= rank(size(), c));
uint64_t ones_before_c = m_bv_blocks_select0(c*m_chunks+1)-(c*m_chunks);
uint64_t chunk = m_bv_blocks_select1(ones_before_c+i)-ones_before_c-(c*m_chunks+1)-i+1;
uint64_t c_ones_before_chunk = m_bv_blocks_select0(c*m_chunks+chunk+1)-(c*m_chunks+chunk)-ones_before_c;
uint64_t pi_pos = m_bv_chunks_select0(chunk*m_max_symbol+c+1)+(i-c_ones_before_chunk)-chunk*m_max_symbol-c-1;
return m_perm[pi_pos]+chunk*m_chunksize;
}
//! Serializes the data structure into the given ostream
size_type serialize(std::ostream& out, structure_tree_node* v=nullptr, std::string name="")const {
structure_tree_node* child = structure_tree::add_child(v, name, util::class_name(*this));
size_type written_bytes = 0;
written_bytes += write_member(m_size, out, child, "size");
written_bytes += write_member(m_max_symbol, out, child, "max_symbol");
written_bytes += write_member(m_chunks, out, child, "chunks");
written_bytes += write_member(m_chunksize, out, child, "chunksize");
written_bytes += write_member(m_sigma, out, child, "sigma");
written_bytes += m_bv_blocks.serialize(out, child, "bv_blocks");
written_bytes += m_bv_blocks_select0.serialize(out, child, "bv_blocks_select0");
written_bytes += m_bv_blocks_select1.serialize(out, child, "bv_blocks_select1");
written_bytes += m_bv_chunks.serialize(out, child, "bv_chunks");
written_bytes += m_bv_chunks_select0.serialize(out, child, "bv_chunks_select0");
written_bytes += m_bv_chunks_select1.serialize(out, child, "bv_chunks_select1");
written_bytes += m_perm.serialize(out, child, "permutation");
written_bytes += m_ips.serialize(out, child, "inverse_permutation_support");
structure_tree::add_size(child, written_bytes);
return written_bytes;
}
//! Loads the data structure from the given istream.
void load(std::istream& in) {
read_member(m_size, in);
read_member(m_max_symbol, in);
read_member(m_chunks, in);
read_member(m_chunksize, in);
read_member(m_sigma, in);
m_bv_blocks.load(in);
m_bv_blocks_select0.load(in, &m_bv_blocks);
m_bv_blocks_select1.load(in, &m_bv_blocks);
m_bv_chunks.load(in);
m_bv_chunks_select0.load(in, &m_bv_chunks);
m_bv_chunks_select1.load(in, &m_bv_chunks);
m_perm.load(in);
m_ips.load(in, &m_perm);
}
};
}
#endif
|