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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558
|
/*
* Copyright (C) 2002-2024 Sebastiano Vigna
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
package PACKAGE;
import static it.unimi.dsi.fastutil.BigArrays.copy;
import static it.unimi.dsi.fastutil.BigArrays.fill;
import static it.unimi.dsi.fastutil.BigArrays.set;
import it.unimi.dsi.fastutil.BigArrays;
import it.unimi.dsi.fastutil.Hash;
import it.unimi.dsi.fastutil.Size64;
import it.unimi.dsi.fastutil.HashCommon;
import static it.unimi.dsi.fastutil.HashCommon.bigArraySize;
import static it.unimi.dsi.fastutil.HashCommon.maxFill;
#if KEYS_REFERENCE
import java.util.function.Consumer;
import java.util.stream.Collector;
#endif
import java.util.Collection;
import java.util.Iterator;
import java.util.NoSuchElementException;
/** A type-specific hash big set with with a fast, small-footprint implementation.
*
* <p>Instances of this class use a hash table to represent a big set: the number
* of elements in the set is limited only by the amount of core memory. The table
* (backed by a {@linkplain it.unimi.dsi.fastutil.BigArrays big array}) is
* filled up to a specified <em>load factor</em>, and then doubled in size to
* accommodate new entries. If the table is emptied below <em>one fourth</em>
* of the load factor, it is halved in size; however, the table is never reduced to a
* size smaller than that at creation time: this approach makes it
* possible to create sets with a large capacity in which insertions and
* deletions do not cause immediately rehashing. Moreover, halving is
* not performed when deleting entries from an iterator, as it would interfere
* with the iteration process.
*
* <p>Note that {@link #clear()} does not modify the hash table size.
* Rather, a family of {@linkplain #trim() trimming
* methods} lets you control the size of the table; this is particularly useful
* if you reuse instances of this class.
*
* <p>The methods of this class are about 30% slower than those of the corresponding non-big set.
*
* @see Hash
* @see HashCommon
*/
public class OPEN_HASH_BIG_SET KEY_GENERIC extends ABSTRACT_SET KEY_GENERIC implements java.io.Serializable, Cloneable, Hash, Size64 {
private static final long serialVersionUID = 0L;
private static final boolean ASSERTS = ASSERTS_VALUE;
/** The big array of keys. */
protected transient KEY_GENERIC_TYPE[][] key;
/** The mask for wrapping a position counter. */
protected transient long mask;
/** The mask for wrapping a segment counter. */
protected transient int segmentMask;
/** The mask for wrapping a base counter. */
protected transient int baseMask;
/** Whether this set contains the null key. */
protected transient boolean containsNull;
/** The current table size (always a power of 2). */
protected transient long n;
/** Threshold after which we rehash. It must be the table size times {@link #f}. */
protected transient long maxFill;
/** We never resize below this threshold, which is the construction-time {#n}. */
protected final transient long minN;
/** The acceptable load factor. */
protected final float f;
/** Number of entries in the set. */
protected long size;
/** Initialises the mask values. */
private void initMasks() {
mask = n - 1;
/* Note that either we have more than one segment, and in this case all segments
* are BigArrays.SEGMENT_SIZE long, or we have exactly one segment whose length
* is a power of two. */
segmentMask = key[0].length - 1;
baseMask = key.length - 1;
}
/** Creates a new hash big set.
*
* <p>The actual table size will be the least power of two greater than {@code expected}/{@code f}.
*
* @param expected the expected number of elements in the set.
* @param f the load factor.
*/
SUPPRESS_WARNINGS_KEY_UNCHECKED
public OPEN_HASH_BIG_SET(final long expected, final float f) {
if (f <= 0 || f > 1) throw new IllegalArgumentException("Load factor must be greater than 0 and smaller than or equal to 1");
if (n < 0) throw new IllegalArgumentException("The expected number of elements must be nonnegative");
this.f = f;
minN = n = bigArraySize(expected, f);
maxFill = maxFill(n, f);
key = KEY_GENERIC_BIG_ARRAY_CAST BIG_ARRAYS.newBigArray(n);
initMasks();
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor.
*
* @param expected the expected number of elements in the hash big set.
*/
public OPEN_HASH_BIG_SET(final long expected) {
this(expected, DEFAULT_LOAD_FACTOR);
}
/** Creates a new hash big set with initial expected {@link Hash#DEFAULT_INITIAL_SIZE} elements
* and {@link Hash#DEFAULT_LOAD_FACTOR} as load factor.
*/
public OPEN_HASH_BIG_SET() {
this(DEFAULT_INITIAL_SIZE, DEFAULT_LOAD_FACTOR);
}
/** Creates a new hash big set copying a given collection.
*
* @param c a {@link Collection} to be copied into the new hash big set.
* @param f the load factor.
*/
public OPEN_HASH_BIG_SET(final Collection<? extends KEY_GENERIC_CLASS> c, final float f) {
this(Size64.sizeOf(c), f);
addAll(c);
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor
* copying a given collection.
*
* @param c a {@link Collection} to be copied into the new hash big set.
*/
public OPEN_HASH_BIG_SET(final Collection<? extends KEY_GENERIC_CLASS> c) {
this(c, DEFAULT_LOAD_FACTOR);
}
/** Creates a new hash big set copying a given type-specific collection.
*
* @param c a type-specific collection to be copied into the new hash big set.
* @param f the load factor.
*/
public OPEN_HASH_BIG_SET(final COLLECTION KEY_EXTENDS_GENERIC c, final float f) {
this(Size64.sizeOf(c), f);
addAll(c);
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor
* copying a given type-specific collection.
*
* @param c a type-specific collection to be copied into the new hash big set.
*/
public OPEN_HASH_BIG_SET(final COLLECTION KEY_EXTENDS_GENERIC c) {
this(c, DEFAULT_LOAD_FACTOR);
}
/** Creates a new hash big set using elements provided by a type-specific iterator.
*
* @param i a type-specific iterator whose elements will fill the new hash big set.
* @param f the load factor.
*/
public OPEN_HASH_BIG_SET(final STD_KEY_ITERATOR KEY_EXTENDS_GENERIC i, final float f) {
this(DEFAULT_INITIAL_SIZE, f);
while(i.hasNext()) add(i.NEXT_KEY());
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor using elements provided by a type-specific iterator.
*
* @param i a type-specific iterator whose elements will fill the new hash big set.
*/
public OPEN_HASH_BIG_SET(final STD_KEY_ITERATOR KEY_EXTENDS_GENERIC i) {
this(i, DEFAULT_LOAD_FACTOR);
}
#if KEYS_PRIMITIVE
/** Creates a new hash big set using elements provided by an iterator.
*
* @param i an iterator whose elements will fill the new hash big set.
* @param f the load factor.
*/
public OPEN_HASH_BIG_SET(final Iterator<?> i, final float f) {
this(ITERATORS.AS_KEY_ITERATOR(i), f);
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor using elements provided by an iterator.
*
* @param i an iterator whose elements will fill the new hash big set.
*/
public OPEN_HASH_BIG_SET(final Iterator<?> i) {
this(ITERATORS.AS_KEY_ITERATOR(i));
}
#endif
/** Creates a new hash big set and fills it with the elements of a given array.
*
* @param a an array whose elements will be used to fill the new hash big set.
* @param offset the first element to use.
* @param length the number of elements to use.
* @param f the load factor.
*/
public OPEN_HASH_BIG_SET(final KEY_GENERIC_TYPE[] a, final int offset, final int length, final float f) {
this(length < 0 ? 0 : length, f);
ARRAYS.ensureOffsetLength(a, offset, length);
for(int i = 0; i < length; i++) add(a[offset + i]);
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor and fills it with the elements of a given array.
*
* @param a an array whose elements will be used to fill the new hash big set.
* @param offset the first element to use.
* @param length the number of elements to use.
*/
public OPEN_HASH_BIG_SET(final KEY_GENERIC_TYPE[] a, final int offset, final int length) {
this(a, offset, length, DEFAULT_LOAD_FACTOR);
}
/** Creates a new hash big set copying the elements of an array.
*
* @param a an array to be copied into the new hash big set.
* @param f the load factor.
*/
public OPEN_HASH_BIG_SET(final KEY_GENERIC_TYPE[] a, final float f) {
this(a, 0, a.length, f);
}
/** Creates a new hash big set with {@link Hash#DEFAULT_LOAD_FACTOR} as load factor
* copying the elements of an array.
*
* @param a an array to be copied into the new hash big set.
*/
public OPEN_HASH_BIG_SET(final KEY_GENERIC_TYPE[] a) {
this(a, DEFAULT_LOAD_FACTOR);
}
#ifndef Custom
#if KEYS_INT_LONG_DOUBLE
/** Collects the result of a primitive {@code Stream} into a new big hash set.
*
* <p>This method performs a terminal operation on the given {@code Stream}
*
* @apiNote Taking a primitive stream instead of returning something like a
* {@link java.util.stream.Collector Collector} is necessary because there is no
* primitive {@code Collector} equivalent in the Java API.
*/
public static KEY_GENERIC OPEN_HASH_BIG_SET KEY_GENERIC toBigSet(JDK_PRIMITIVE_STREAM stream) {
return stream.collect(
OPEN_HASH_BIG_SET::new,
OPEN_HASH_BIG_SET::add,
OPEN_HASH_BIG_SET::addAll);
}
/** Collects the result of a primitive {@code Stream} into a new big hash set.
*
* <p>This method performs a terminal operation on the given {@code Stream}
*
* @apiNote Taking a primitive stream instead returning something like a
* {@link java.util.stream.Collector Collector} is necessary because there is no
* primitive {@code Collector} equivalent in the Java API.
*/
public static KEY_GENERIC OPEN_HASH_BIG_SET KEY_GENERIC toBigSetWithExpectedSize(JDK_PRIMITIVE_STREAM stream, long expectedSize) {
return stream.collect(
() -> new OPEN_HASH_BIG_SET KEY_GENERIC(expectedSize),
OPEN_HASH_BIG_SET::add,
OPEN_HASH_BIG_SET::addAll);
}
#elif KEYS_REFERENCE
// Collector wants a function that returns the collection being added to.
private OPEN_HASH_BIG_SET KEY_GENERIC combine(OPEN_HASH_BIG_SET KEY_EXTENDS_GENERIC toAddFrom) {
addAll(toAddFrom);
return this;
}
private static final Collector<KEY_TYPE, ?, OPEN_HASH_BIG_SET<KEY_TYPE>> TO_SET_COLLECTOR =
Collector.of(
OPEN_HASH_BIG_SET::new,
OPEN_HASH_BIG_SET::add,
OPEN_HASH_BIG_SET::combine);
/** Returns a {@link Collector} that collects a {@code Stream}'s elements into a new big hash set. */
SUPPRESS_WARNINGS_KEY_UNCHECKED_RAWTYPES
public static KEY_GENERIC Collector<KEY_GENERIC_TYPE, ?, OPEN_HASH_BIG_SET KEY_GENERIC> toBigSet() {
return (Collector) TO_SET_COLLECTOR;
}
/** Returns a {@link Collector} that collects a {@code Stream}'s elements into a new big hash set. */
public static KEY_GENERIC Collector<KEY_GENERIC_TYPE, ?, OPEN_HASH_BIG_SET KEY_GENERIC> toBigSetWithExpectedSize(long expectedSize) {
return Collector.of(
() -> new OPEN_HASH_BIG_SET KEY_GENERIC(expectedSize),
OPEN_HASH_BIG_SET::add,
OPEN_HASH_BIG_SET::combine);
}
#endif
#endif
private long realSize() {
return containsNull ? size - 1 : size;
}
/** Ensures that this big set can hold a certain number of elements without rehashing.
*
* @param capacity a number of elements; there will be no rehashing unless
* the set {@linkplain #size64() size} exceeds this number.
*/
public void ensureCapacity(final long capacity) {
final long needed = bigArraySize(capacity, f);
if (needed > n) rehash(needed);
}
@Override
public boolean addAll(Collection<? extends KEY_GENERIC_CLASS> c) {
final long size = Size64.sizeOf(c);
// The resulting collection will be at least c.size() big
if (f <= .5) ensureCapacity(size); // The resulting collection will be sized for c.size() elements
else ensureCapacity(size64() + size); // The resulting collection will be sized for size() + c.size() elements
return super.addAll(c);
}
#if KEYS_PRIMITIVE
@Override
public boolean addAll(COLLECTION c) {
final long size = Size64.sizeOf(c);
if (f <= .5) ensureCapacity(size); // The resulting collection will be size for c.size() elements
else ensureCapacity(size64() + size); // The resulting collection will be sized for size() + c.size() elements
return super.addAll(c);
}
#endif
@Override
public boolean add(final KEY_GENERIC_TYPE k) {
int displ, base;
if (KEY_IS_NULL(k)) {
if (containsNull) return false;
containsNull = true;
}
else {
KEY_GENERIC_TYPE curr;
final KEY_GENERIC_TYPE[][] key = this.key;
final long h = KEY2LONGHASH(k);
// The starting point.
if (! KEY_IS_NULL(curr = key[base = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][displ = (int)(h & segmentMask)])) {
if (KEY_EQUALS_NOT_NULL(curr, k)) return false;
while(! KEY_IS_NULL(curr = key[base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0)) & baseMask][displ]))
if (KEY_EQUALS_NOT_NULL(curr, k)) return false;
}
key[base][displ] = k;
}
if (size++ >= maxFill) rehash(2 * n);
if (ASSERTS) checkTable();
return true;
}
#if KEY_CLASS_Object
/** Add a random element if not present, get the existing value if already present.
*
* This is equivalent to (but faster than) doing a:
* <pre>
* K exist = set.get(k);
* if (exist == null) {
* set.add(k);
* exist = k;
* }
* </pre>
*/
public KEY_GENERIC_TYPE addOrGet(final KEY_GENERIC_TYPE k) {
int displ, base;
if (KEY_IS_NULL(k)) {
if (containsNull) return null;
containsNull = true;
}
else {
KEY_GENERIC_TYPE curr;
final KEY_GENERIC_TYPE[][] key = this.key;
final long h = KEY2LONGHASH(k);
// The starting point.
if (! KEY_IS_NULL(curr = key[base = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][displ = (int)(h & segmentMask)])) {
if (KEY_EQUALS_NOT_NULL(curr, k)) return curr;
while(! KEY_IS_NULL(curr = key[base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0)) & baseMask][displ]))
if (KEY_EQUALS_NOT_NULL(curr, k)) return curr;
}
key[base][displ] = k;
}
if (size++ >= maxFill) rehash(2 * n);
if (ASSERTS) checkTable();
return k;
}
#endif
/** Shifts left entries with the specified hash code, starting at the specified position,
* and empties the resulting free entry.
*
* @param pos a starting position.
*/
protected final void shiftKeys(long pos) {
// Shift entries with the same hash.
long last, slot;
final KEY_GENERIC_TYPE[][] key = this.key;
for(;;) {
pos = ((last = pos) + 1) & mask;
for(;;) {
if (KEY_IS_NULL(BigArrays.get(key, pos))) {
set(key, last, KEY_NULL);
return;
}
slot = KEY2LONGHASH(BigArrays.get(key, pos)) & mask;
if (last <= pos ? last >= slot || slot > pos : last >= slot && slot > pos) break;
pos = (pos + 1) & mask;
}
set(key, last, BigArrays.get(key, pos));
}
}
private boolean removeEntry(final int base, final int displ) {
size--;
shiftKeys(base * (long)BigArrays.SEGMENT_SIZE + displ);
if (n > minN && size < maxFill / 4 && n > DEFAULT_INITIAL_SIZE) rehash(n / 2);
return true;
}
private boolean removeNullEntry() {
containsNull = false;
size--;
if (n > minN && size < maxFill / 4 && n > DEFAULT_INITIAL_SIZE) rehash(n / 2);
return true;
}
@Override
public boolean remove(final KEY_TYPE k) {
if (KEY_IS_NULL(k)) {
if (containsNull) return removeNullEntry();
return false;
}
KEY_GENERIC_TYPE curr;
final KEY_GENERIC_TYPE[][] key = this.key;
final long h = KEY2LONGHASH(k);
int displ, base;
// The starting point.
if (KEY_IS_NULL(curr = key[base = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][displ = (int)(h & segmentMask)])) return false;
if (KEY_EQUALS_NOT_NULL(curr, k)) return removeEntry(base, displ);
while(true) {
if (KEY_IS_NULL(curr = key[base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0)) & baseMask][displ])) return false;
if (KEY_EQUALS_NOT_NULL(curr, k)) return removeEntry(base, displ);
}
}
@Override
public boolean contains(final KEY_TYPE k) {
if (KEY_IS_NULL(k)) return containsNull;
KEY_GENERIC_TYPE curr;
final KEY_GENERIC_TYPE[][] key = this.key;
final long h = KEY2LONGHASH(k);
int displ, base;
// The starting point.
if (KEY_IS_NULL(curr = key[base = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][displ = (int)(h & segmentMask)])) return false;
if (KEY_EQUALS_NOT_NULL(curr, k)) return true;
while(true) {
if (KEY_IS_NULL(curr = key[base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0)) & baseMask][displ])) return false;
if (KEY_EQUALS_NOT_NULL(curr, k)) return true;
}
}
#if KEY_CLASS_Object
/** Returns the element of this set that is equal to the given key, or {@code null}.
* @return the element of this set that is equal to the given key, or {@code null}.
*/
public K get(final KEY_TYPE k) {
if (k == null) return null; // This is correct independently of the value of containsNull
KEY_GENERIC_TYPE curr;
final KEY_GENERIC_TYPE[][] key = this.key;
final long h = KEY2LONGHASH(k);
int displ, base;
// The starting point.
if (KEY_IS_NULL(curr = key[base = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][displ = (int)(h & segmentMask)])) return null;
if (KEY_EQUALS_NOT_NULL(curr, k)) return curr;
while(true) {
if (KEY_IS_NULL(curr = key[base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0)) & baseMask][displ])) return null;
if (KEY_EQUALS_NOT_NULL(curr, k)) return curr;
}
}
#endif
/* Removes all elements from this set.
*
*/
/** {@inheritDoc}
*
* <p>To increase object reuse, this method does not change the table size.
* If you want to reduce the table size, you must use {@link #trim(long)}.
*/
@Override
public void clear() {
if (size == 0) return;
size = 0;
containsNull = false;
fill(key, KEY_NULL);
}
/** An iterator over a hash big set. */
private class SetIterator implements KEY_ITERATOR KEY_GENERIC {
/** The base of the last entry returned, if positive or zero; initially, the number of components
of the key array. If negative, the last element returned was
that of index {@code - base - 1} from the {@link #wrapped} list. */
int base = key.length;
/** The displacement of the last entry returned; initially, zero. */
int displ;
/** The index of the last entry that has been returned (or {@link Long#MIN_VALUE} if {@link #base} is negative).
It is -1 if either we did not return an entry yet, or the last returned entry has been removed. */
long last = -1;
/** A downward counter measuring how many entries must still be returned. */
long c = size;
/** A boolean telling us whether we should return the null key. */
boolean mustReturnNull = OPEN_HASH_BIG_SET.this.containsNull;
/** A lazily allocated list containing elements that have wrapped around the table because of removals. */
ARRAY_LIST KEY_GENERIC wrapped;
@Override
public boolean hasNext() { return c != 0; }
@Override
public KEY_GENERIC_TYPE NEXT_KEY() {
if (! hasNext()) throw new NoSuchElementException();
c--;
if (mustReturnNull) {
mustReturnNull = false;
last = n;
return KEY_NULL;
}
final KEY_GENERIC_TYPE[][] key = OPEN_HASH_BIG_SET.this.key;
for(;;) {
if (displ == 0 && base <= 0) {
// We are just enumerating elements from the wrapped list.
last = Long.MIN_VALUE;
return wrapped.GET_KEY(- (--base) - 1);
}
if (displ-- == 0) displ = key[--base].length - 1;
final KEY_GENERIC_TYPE k = key[base][displ];
if (! KEY_IS_NULL(k)) {
last = base * (long)BigArrays.SEGMENT_SIZE + displ;
return k;
}
}
}
/** Shifts left entries with the specified hash code, starting at the specified position,
* and empties the resulting free entry.
*
* @param pos a starting position.
*/
private final void shiftKeys(long pos) {
// Shift entries with the same hash.
long last, slot;
KEY_GENERIC_TYPE curr;
final KEY_GENERIC_TYPE[][] key = OPEN_HASH_BIG_SET.this.key;
for(;;) {
pos = ((last = pos) + 1) & mask;
for(;;) {
if(KEY_IS_NULL(curr = BigArrays.get(key, pos))) {
set(key, last, KEY_NULL);
return;
}
slot = KEY2LONGHASH(curr) & mask;
if (last <= pos ? last >= slot || slot > pos : last >= slot && slot > pos) break;
pos = (pos + 1) & mask;
}
if (pos < last) { // Wrapped entry.
if (wrapped == null) wrapped = new ARRAY_LIST KEY_GENERIC_DIAMOND();
wrapped.add(BigArrays.get(key, pos));
}
set(key, last, curr);
}
}
@Override
public void remove() {
if (last == -1) throw new IllegalStateException();
if (last == n) OPEN_HASH_BIG_SET.this.containsNull = false;
else if (base >= 0) shiftKeys(last);
else {
// We're removing wrapped entries.
#if KEYS_REFERENCE
OPEN_HASH_BIG_SET.this.remove(wrapped.set(- base - 1, null));
#else
OPEN_HASH_BIG_SET.this.remove(wrapped.GET_KEY(- base - 1));
#endif
last = -1; // Note that we must not decrement size
return;
}
size--;
last = -1; // You can no longer remove this entry.
if (ASSERTS) checkTable();
}
}
@Override
public KEY_ITERATOR KEY_GENERIC iterator() {
return new SetIterator();
}
private class SetSpliterator implements KEY_SPLITERATOR KEY_GENERIC {
/* For the sake of keeping things at least somewhat simple
* (aka. my sanity), the spliterator will NOT handle the indexing
* of the subarrays directly, like iterator does. Instead, it will
* delegate to BigArrays and have only a single, unified index it
* will fence on. This is probably less effecient, but it avoids having
* to track what it means to split on two sets of indexes.
* This may change in the future if the performance hit high.
*/
private static final int POST_SPLIT_CHARACTERISTICS = SPLITERATORS.SET_SPLITERATOR_CHARACTERISTICS & ~java.util.Spliterator.SIZED;
/** The index (which bucket) of the next item to give to the action. */
long pos = 0;
/** The maximum bucket (exclusive) to iterate to */
long max = n;
/** An upwards counter counting how many we have given */
long c = 0;
/** A boolean telling us whether we should return the null key. */
boolean mustReturnNull = OPEN_HASH_BIG_SET.this.containsNull;
boolean hasSplit = false;
SetSpliterator() {}
SetSpliterator(long pos, long max, boolean mustReturnNull, boolean hasSplit) {
this.pos = pos;
this.max = max;
this.mustReturnNull = mustReturnNull;
this.hasSplit = hasSplit;
}
@Override
public boolean tryAdvance(final METHOD_ARG_KEY_CONSUMER action) {
if (mustReturnNull) {
mustReturnNull = false;
++c;
action.accept(KEY_NULL);
return true;
}
final KEY_GENERIC_TYPE key[][] = OPEN_HASH_BIG_SET.this.key;
while (pos < max) {
KEY_GENERIC_TYPE gotten = BigArrays.get(key, pos);
if (! KEY_IS_NULL(gotten)) {
++c;
++pos;
action.accept(gotten);
return true;
} else {
++pos;
}
}
return false;
}
@Override
public void forEachRemaining(final METHOD_ARG_KEY_CONSUMER action) {
if (mustReturnNull) {
mustReturnNull = false;
action.accept(KEY_NULL);
++c;
}
final KEY_GENERIC_TYPE key[][] = OPEN_HASH_BIG_SET.this.key;
while (pos < max) {
KEY_GENERIC_TYPE gotten = BigArrays.get(key, pos);
if (! KEY_IS_NULL(gotten)) {
action.accept(gotten);
++c;
}
++pos;
}
}
@Override
public int characteristics() {
return hasSplit ? POST_SPLIT_CHARACTERISTICS : SPLITERATORS.SET_SPLITERATOR_CHARACTERISTICS;
}
@Override
public long estimateSize() {
if (!hasSplit) {
// Root spliterator; we know how many are remaining.
return size - c;
} else {
// After we split, we can no longer know exactly how many we have (or at least not efficiently).
// (size / n) * (max - pos) aka currentTableDensity * numberOfBucketsLeft seems like a good estimate.
return Math.min(size - c, (long)(((double)realSize() / n) * (max - pos)) + (mustReturnNull ? 1 : 0));
}
}
@Override
public SetSpliterator trySplit() {
if (pos >= max - 1) return null;
long retLen = (max - pos) >> 1;
if (retLen <= 1) return null;
long myNewPos = pos + retLen;
// Align to an outer array boundary if possible
// We add/subtract one to the bounds to ensure the new pos will always shrink the range
myNewPos = BigArrays.nearestSegmentStart(myNewPos, pos + 1, max - 1);
long retPos = pos;
long retMax = myNewPos;
// Since null is returned first, and the convention is that the returned split is the prefix of elements,
// the split will take care of returning null (if needed), and we won't return it anymore.
SetSpliterator split = new SetSpliterator(retPos, retMax, mustReturnNull, true);
this.pos = myNewPos;
this.mustReturnNull = false;
this.hasSplit = true;
return split;
}
@Override
public long skip(long n) {
if (n < 0) throw new IllegalArgumentException("Argument must be nonnegative: " + n);
if (n == 0) return 0;
long skipped = 0;
if (mustReturnNull) {
mustReturnNull = false;
++skipped;
--n;
}
final KEY_GENERIC_TYPE key[][] = OPEN_HASH_BIG_SET.this.key;
while (pos < max && n > 0) {
if (! KEY_IS_NULL(BigArrays.get(key, pos++))) {
++skipped;
--n;
}
}
return skipped;
}
}
@Override
public KEY_SPLITERATOR KEY_GENERIC spliterator() {
return new SetSpliterator();
}
@Override
public void forEach(final METHOD_ARG_KEY_CONSUMER action) {
if (containsNull) {
action.accept(KEY_NULL);
}
long pos = 0;
final long max = n;
final KEY_GENERIC_TYPE key[][] = this.key;
while (pos < max) {
KEY_GENERIC_TYPE gotten = BigArrays.get(key, pos++);
if (! KEY_IS_NULL(gotten)) {
action.accept(gotten);
}
}
}
/** Rehashes this set, making the table as small as possible.
*
* <p>This method rehashes the table to the smallest size satisfying the
* load factor. It can be used when the set will not be changed anymore, so
* to optimize access speed and size.
*
* <p>If the table size is already the minimum possible, this method
* does nothing.
*
* @return true if there was enough memory to trim the set.
* @see #trim(long)
*/
public boolean trim() {
return trim(size);
}
/** Rehashes this set if the table is too large.
*
* <p>Let <var>N</var> be the smallest table size that can hold
* <code>max(n,{@link #size64()})</code> entries, still satisfying the load factor. If the current
* table size is smaller than or equal to <var>N</var>, this method does
* nothing. Otherwise, it rehashes this set in a table of size
* <var>N</var>.
*
* <p>This method is useful when reusing sets. {@linkplain #clear() Clearing a
* set} leaves the table size untouched. If you are reusing a set
* many times, you can call this method with a typical
* size to avoid keeping around a very large table just
* because of a few large transient sets.
*
* @param n the threshold for the trimming.
* @return true if there was enough memory to trim the set.
* @see #trim()
*/
public boolean trim(final long n) {
final long l = bigArraySize(n, f);
if (l >= this.n || size > maxFill(l, f)) return true;
try {
rehash(l);
}
catch(OutOfMemoryError cantDoIt) { return false; }
return true;
}
/** Resizes the set.
*
* <p>This method implements the basic rehashing strategy, and may be
* overriden by subclasses implementing different rehashing strategies (e.g.,
* disk-based rehashing). However, you should not override this method
* unless you understand the internal workings of this class.
*
* @param newN the new size
*/
SUPPRESS_WARNINGS_KEY_UNCHECKED
protected void rehash(final long newN) {
final KEY_GENERIC_TYPE key[][] = this.key;
final KEY_GENERIC_TYPE newKey[][] = KEY_GENERIC_BIG_ARRAY_CAST BIG_ARRAYS.newBigArray(newN);
final long mask = newN - 1; // Note that this is used by the hashing macro
final int newSegmentMask = newKey[0].length - 1;
final int newBaseMask = newKey.length - 1;
int base = 0, displ = 0, b, d;
long h;
KEY_GENERIC_TYPE k;
for(long i = realSize(); i-- != 0;) {
while(KEY_IS_NULL(key[base][displ])) base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0));
k = key[base][displ];
h = KEY2LONGHASH(k);
// The starting point.
if (! KEY_IS_NULL(newKey[b = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][d = (int)(h & newSegmentMask)]))
while(! KEY_IS_NULL(newKey[b = (b + ((d = (d + 1) & newSegmentMask) == 0 ? 1 : 0)) & newBaseMask][d]));
newKey[b][d] = k;
base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0));
}
this.n = newN;
this.key = newKey;
initMasks();
maxFill = maxFill(n, f);
}
@Deprecated
@Override
public int size() {
return (int)Math.min(Integer.MAX_VALUE, size);
}
@Override
public long size64() {
return size;
}
@Override
public boolean isEmpty() {
return size == 0;
}
/** Returns a deep copy of this big set.
*
* <p>This method performs a deep copy of this big hash set; the data stored in the
* set, however, is not cloned. Note that this makes a difference only for object keys.
*
* @return a deep copy of this big set.
*/
@Override
SUPPRESS_WARNINGS_KEY_UNCHECKED
public OPEN_HASH_BIG_SET KEY_GENERIC clone() {
OPEN_HASH_BIG_SET KEY_GENERIC c;
try {
c = (OPEN_HASH_BIG_SET KEY_GENERIC)super.clone();
}
catch(CloneNotSupportedException cantHappen) {
throw new InternalError();
}
c.key = copy(key);
c.containsNull = containsNull;
return c;
}
/** Returns a hash code for this set.
*
* This method overrides the generic method provided by the superclass.
* Since {@code equals()} is not overriden, it is important
* that the value returned by this method is the same value as
* the one returned by the overriden method.
*
* @return a hash code for this set.
*/
@Override
public int hashCode() {
final KEY_GENERIC_TYPE key[][] = this.key;
int h = 0, base = 0, displ = 0;
for(long j = realSize(); j-- != 0;) {
while(KEY_IS_NULL(key[base][displ])) base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0));
#if KEYS_REFERENCE
if (this != key[base][displ])
#endif
h += KEY2JAVAHASH_NOT_NULL(key[base][displ]);
base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0));
}
return h;
}
private void writeObject(java.io.ObjectOutputStream s) throws java.io.IOException {
final KEY_ITERATOR KEY_GENERIC i = iterator();
s.defaultWriteObject();
for(long j = size; j-- != 0;) s.WRITE_KEY(i.NEXT_KEY());
}
SUPPRESS_WARNINGS_KEY_UNCHECKED
private void readObject(java.io.ObjectInputStream s) throws java.io.IOException, ClassNotFoundException {
s.defaultReadObject();
n = bigArraySize(size, f);
maxFill = maxFill(n, f);
final KEY_GENERIC_TYPE[][] key = this.key = KEY_GENERIC_BIG_ARRAY_CAST BIG_ARRAYS.newBigArray(n);
initMasks();
long h;
KEY_GENERIC_TYPE k;
int base, displ;
for(long i = size; i-- != 0;) {
k = KEY_GENERIC_CAST s.READ_KEY();
if (KEY_IS_NULL(k)) containsNull = true;
else {
h = KEY2LONGHASH(k);
if (! KEY_IS_NULL(key[base = (int)((h & mask) >>> BigArrays.SEGMENT_SHIFT)][displ = (int)(h & segmentMask)]))
while(! KEY_IS_NULL(key[base = (base + ((displ = (displ + 1) & segmentMask) == 0 ? 1 : 0)) & baseMask][displ]));
key[base][displ] = k;
}
}
if (ASSERTS) checkTable();
}
#ifdef ASSERTS_CODE
private void checkTable() {
assert (n & -n) == n : "Table length is not a power of two: " + n;
assert n == BigArrays.length(key);
long n = this.n;
while(n-- != 0)
if (! KEY_IS_NULL(BigArrays.get(key, n)) && ! contains(BigArrays.get(key, n)))
throw new AssertionError("Hash table has key " + BigArrays.get(key, n) + " marked as occupied, but the key does not belong to the table");
#if KEYS_PRIMITIVE
java.util.HashSet<KEY_GENERIC_CLASS> s = new java.util.HashSet<KEY_GENERIC_CLASS> ();
#else
java.util.HashSet<Object> s = new java.util.HashSet<Object>();
#endif
for(long i = size(); i-- != 0;)
if (! KEY_IS_NULL(BigArrays.get(key, i)) && ! s.add(BigArrays.get(key, i))) throw new AssertionError("Key " + BigArrays.get(key, i) + " appears twice");
}
#else
private void checkTable() {}
#endif
#ifdef TEST
private static long seed = System.currentTimeMillis();
private static java.util.Random r = new java.util.Random(seed);
private static KEY_TYPE genKey() {
#if KEY_CLASS_Byte || KEY_CLASS_Short || KEY_CLASS_Character
return (KEY_TYPE)(r.nextInt());
#elif KEYS_PRIMITIVE
return r.NEXT_KEY();
#elif KEY_CLASS_Object
return Integer.toBinaryString(r.nextInt());
#else
return new java.io.Serializable() {};
#endif
}
private static final class ArrayComparator implements java.util.Comparator {
public int compare(Object a, Object b) {
byte[] aa = (byte[])a;
byte[] bb = (byte[])b;
int length = Math.min(aa.length, bb.length);
for(int i = 0; i < length; i++) {
if (aa[i] < bb[i]) return -1;
if (aa[i] > bb[i]) return 1;
}
return aa.length == bb.length ? 0 : (aa.length < bb.length ? -1 : 1);
}
}
private static final class MockSet extends java.util.TreeSet {
private java.util.List list = new java.util.ArrayList();
public MockSet(java.util.Comparator c) { super(c); }
public boolean add(Object k) {
if (! contains(k)) list.add(k);
return super.add(k);
}
public boolean addAll(Collection c) {
java.util.Iterator i = c.iterator();
boolean result = false;
while(i.hasNext()) result |= add(i.next());
return result;
}
public boolean removeAll(Collection c) {
java.util.Iterator i = c.iterator();
boolean result = false;
while(i.hasNext()) result |= remove(i.next());
return result;
}
public boolean remove(Object k) {
if (contains(k)) {
int i = list.size();
while(i-- != 0) if (comparator().compare(list.get(i), k) == 0) {
list.remove(i);
break;
}
}
return super.remove(k);
}
private void justRemove(Object k) { super.remove(k); }
public java.util.Iterator iterator() {
return new java.util.Iterator() {
final java.util.Iterator iterator = list.iterator();
Object curr;
public Object next() { return curr = iterator.next(); }
public boolean hasNext() { return iterator.hasNext(); }
public void remove() {
justRemove(curr);
iterator.remove();
}
};
}
}
private static java.text.NumberFormat format = new java.text.DecimalFormat("#,###.00");
private static java.text.FieldPosition fp = new java.text.FieldPosition(0);
private static String format(double d) {
StringBuffer s = new StringBuffer();
return format.format(d, s, fp).toString();
}
// TODO Use a ASSERTS like preprocessor variable?
private static final boolean PARALLEL_STREAMS = Boolean.getBoolean("useParallelStreams");
private static void speedTest(int n, float f, boolean comp) {
int i, j;
OPEN_HASH_BIG_SET m;
java.util.HashSet t;
KEY_TYPE k[] = new KEY_TYPE[n];
KEY_TYPE nk[] = new KEY_TYPE[n];
long ms;
for(i = 0; i < n; i++) {
k[i] = genKey();
nk[i] = genKey();
}
double totAdd = 0, totYes = 0, totNo = 0, totIter = 0, totRemYes = 0, totRemNo = 0, d;
if (comp) { for(j = 0; j < 20; j++) {
t = new java.util.HashSet(16);
/* We add pairs to t. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) t.add(KEY2OBJ(k[i]));
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totAdd += d;
System.out.print("Add: " + format(d) +" K/s ");
/* We check for pairs in t. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) t.contains(KEY2OBJ(k[i]));
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totYes += d;
System.out.print("Yes: " + format(d) +" K/s ");
/* We check for pairs not in t. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) t.contains(KEY2OBJ(nk[i]));
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totNo += d;
System.out.print("No: " + format(d) +" K/s ");
/* We iterate on t. */
ms = System.currentTimeMillis();
for(java.util.Iterator it = t.iterator(); it.hasNext(); it.next());
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totIter += d;
System.out.print("Iter: " + format(d) +" K/s ");
/* We delete pairs not in t. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) t.remove(KEY2OBJ(nk[i]));
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totRemNo += d;
System.out.print("RemNo: " + format(d) +" K/s ");
/* We delete pairs in t. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) t.remove(KEY2OBJ(k[i]));
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totRemYes += d;
System.out.print("RemYes: " + format(d) +" K/s ");
System.out.println();
}
System.out.println();
System.out.println("java.util Add: " + format(totAdd/(j-3)) + " K/s Yes: " + format(totYes/(j-3)) + " K/s No: " + format(totNo/(j-3)) + " K/s Iter: " + format(totIter/(j-3)) + " K/s RemNo: " + format(totRemNo/(j-3)) + " K/s RemYes: " + format(totRemYes/(j-3)) + "K/s");
System.out.println();
totAdd = totYes = totNo = totIter = totRemYes = totRemNo = 0;
}
for(j = 0; j < 20; j++) {
m = new OPEN_HASH_BIG_SET(16, f);
/* We add pairs to m. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) m.add(k[i]);
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totAdd += d;
System.out.print("Add: " + format(d) +" K/s ");
/* We check for pairs in m. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) m.contains(k[i]);
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totYes += d;
System.out.print("Yes: " + format(d) +" K/s ");
/* We check for pairs not in m. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) m.contains(nk[i]);
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totNo += d;
System.out.print("No: " + format(d) +" K/s ");
/* We iterate on m. */
ms = System.currentTimeMillis();
for(KEY_ITERATOR it = (KEY_ITERATOR)m.iterator(); it.hasNext(); it.NEXT_KEY());
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totIter += d;
System.out.print("Iter: " + format(d) +" K/s ");
/* We delete pairs not in m. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) m.remove(nk[i]);
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totRemNo += d;
System.out.print("RemNo: " + format(d) +" K/s ");
/* We delete pairs in m. */
ms = System.currentTimeMillis();
for(i = 0; i < n; i++) m.remove(k[i]);
d = 1.0 * n / (System.currentTimeMillis() - ms);
if (j > 2) totRemYes += d;
System.out.print("RemYes: " + format(d) +" K/s ");
System.out.println();
}
System.out.println();
System.out.println("fastutil Add: " + format(totAdd/(j-3)) + " K/s Yes: " + format(totYes/(j-3)) + " K/s No: " + format(totNo/(j-3)) + " K/s Iter: " + format(totIter/(j-3)) + " K/s RemNo: " + format(totRemNo/(j-3)) + " K/s RemYes: " + format(totRemYes/(j-3)) + " K/s");
System.out.println();
}
private static void fatal(String msg) {
throw new AssertionError(msg);
}
private static void ensure(boolean cond, String msg) {
if (cond) return;
fatal(msg);
}
private static void printProbes(OPEN_HASH_BIG_SET m) {
long totProbes = 0;
double totSquareProbes = 0;
int maxProbes = 0;
final double f = (double)m.size / m.n;
for(int i = 0, c = 0; i < m.n; i++) {
if (! KEY_IS_NULL(BigArrays.get(m.key, i))) c++;
else {
if (c != 0) {
final long p = (c + 1) * (c + 2) / 2;
totProbes += p;
totSquareProbes += (double)p * p;
}
maxProbes = Math.max(c, maxProbes);
c = 0;
totProbes++;
totSquareProbes++;
}
}
final double expected = (double)totProbes / m.n;
System.err.println("Expected probes: " + (
3 * Math.sqrt(3) * (f / ((1 - f) * (1 - f))) + 4 / (9 * f) - 1
) + "; actual: " + expected + "; stddev: " + Math.sqrt(totSquareProbes / m.n - expected * expected) + "; max probes: " + maxProbes);
}
private static void runTest(int n, float f) throws Exception {
int c;
OPEN_HASH_BIG_SET m = new OPEN_HASH_BIG_SET(Hash.DEFAULT_INITIAL_SIZE, f);
java.util.Set t = new java.util.HashSet();
/* First of all, we fill t with random data. */
for(int i=0; i<f * n; i++) t.add(KEY2OBJ(genKey()));
/* Now we add to m the same data */
m.addAll(t);
ensure(m.equals(t), "Error (" + seed + "): !m.equals(t) after insertion");;
ensure(t.equals(m), "Error (" + seed + "): !t.equals(m) after insertion");;
printProbes(m);
/* Now we check that m actually holds that data. */
for(java.util.Iterator i=t.iterator(); i.hasNext();) {
Object e = i.next();
ensure(m.contains(e), "Error (" + seed + "): m and t differ on a key ("+e+") after insertion (iterating on t)");
}
/* Now we check that m actually holds that data, but iterating on m. */
c = 0;
for(java.util.Iterator i=m.iterator(); i.hasNext();) {
Object e = i.next();
c++;
ensure(t.contains(e), "Error (" + seed + "): m and t differ on a key ("+e+") after insertion (iterating on m)");
}
ensure(c == t.size(), "Error (" + seed + "): m has only " + c + " keys instead of " + t.size() + " after insertion (iterating on m)");
/* Now we check that inquiries about random data give the same answer in m and t. For
m we use the polymorphic method. */
for(int i=0; i<n; i++) {
KEY_TYPE T = genKey();
ensure(m.contains(T) == t.contains(KEY2OBJ(T)), "Error (" + seed + "): divergence in keys between t and m (polymorphic method)");
}
/* Again, we check that inquiries about random data give the same answer in m and t, but
for m we use the standard method. */
for(int i=0; i<n; i++) {
KEY_TYPE T = genKey();
ensure(m.contains(KEY2OBJ(T)) == t.contains(KEY2OBJ(T)), "Error (" + seed + "): divergence between t and m (standard method)");
}
/* Now we put and remove random data in m and t, checking that the result is the same. */
for(int i=0; i<20*n; i++) {
KEY_TYPE T = genKey();
ensure(m.add(KEY2OBJ(T)) == t.add(KEY2OBJ(T)), "Error (" + seed + "): divergence in add() between t and m");
T = genKey();
ensure(m.remove(KEY2OBJ(T)) == t.remove(KEY2OBJ(T)), "Error (" + seed + "): divergence in remove() between t and m");
}
ensure(m.equals(t), "Error (" + seed + "): !m.equals(t) after removal");;
ensure(t.equals(m), "Error (" + seed + "): !t.equals(m) after removal");;
/* Now we check that m actually holds that data. */
for(java.util.Iterator i=t.iterator(); i.hasNext();) {
Object e = i.next();
ensure(m.contains(e), "Error (" + seed + "): m and t differ on a key ("+e+") after removal (iterating on t)");
}
/* Now we check that m actually holds that data, but iterating on m. */
for(java.util.Iterator i=m.iterator(); i.hasNext();) {
Object e = i.next();
ensure(t.contains(e), "Error (" + seed + "): m and t differ on a key ("+e+") after removal (iterating on m)");
}
printProbes(m);
/* Now we make m into an array, make it again a set and check it is OK. */
KEY_TYPE a[] = m.TO_KEY_ARRAY();
ensure(new OPEN_HASH_BIG_SET(a).equals(m), "Error (" + seed + "): toArray() output (or array-based constructor) is not OK");
/* Now we check cloning. */
ensure(m.equals(((OPEN_HASH_BIG_SET)m).clone()), "Error (" + seed + "): m does not equal m.clone()");
ensure(((OPEN_HASH_BIG_SET)m).clone().equals(m), "Error (" + seed + "): m.clone() does not equal m");
int h = m.hashCode();
/* Now we save and read m. */
{
java.io.File ff = new java.io.File("it.unimi.dsi.fastutil.test." + m.getClass().getSimpleName() + "." + n);
java.io.OutputStream os = new java.io.FileOutputStream(ff);
java.io.ObjectOutputStream oos = new java.io.ObjectOutputStream(os);
oos.writeObject(m);
oos.close();
java.io.InputStream is = new java.io.FileInputStream(ff);
java.io.ObjectInputStream ois = new java.io.ObjectInputStream(is);
m = (OPEN_HASH_BIG_SET)ois.readObject();
ois.close();
ff.delete();
}
#if !KEYS_USE_REFERENCE_EQUALITY
ensure(m.hashCode() == h, "Error (" + seed + "): hashCode() changed after save/read");;
printProbes(m);
/* Now we check that m actually holds that data, but iterating on m. */
for(java.util.Iterator i=m.iterator(); i.hasNext();) {
Object e = i.next();
ensure(t.contains(e), "Error (" + seed + "): m and t differ on a key ("+e+") after save/read");
}
#else
m.clear();
m.addAll(t);
#endif
/* Now we put and remove random data in m and t, checking that the result is the same. */
for(int i=0; i<20*n; i++) {
KEY_TYPE T = genKey();
ensure(m.add(KEY2OBJ(T)) == t.add(KEY2OBJ(T)), "Error (" + seed + "): divergence in add() between t and m after save/read");
T = genKey();
ensure(m.remove(KEY2OBJ(T)) == t.remove(KEY2OBJ(T)), "Error (" + seed + "): divergence in remove() between t and m after save/read");
}
ensure(m.equals(t), "Error (" + seed + "): !m.equals(t) after post-save/read removal");;
ensure(t.equals(m), "Error (" + seed + "): !t.equals(m) after post-save/read removal");;
/* Now test that the sets hold the same data using streams */
{
#if KEYS_REFERENCE
java.util.stream.Stream<KEY_TYPE> i = m.stream();
java.util.stream.Stream<KEY_TYPE> j = t.stream();
#elif KEY_CLASS_Boolean
java.util.stream.Stream<KEY_CLASS> i = m.stream();
java.util.stream.Stream<KEY_CLASS> j = t.stream();
#else
JDK_PRIMITIVE_STREAM i = m.KEY_WIDENED_STREAM_METHOD();
java.util.stream.Stream<KEY_CLASS> j = t.stream();
#endif
if (PARALLEL_STREAMS) {
i = i.parallel();
j = j.parallel();
}
i = i.sorted();
j = j.sorted();
#if KEYS_REFERENCE || KEY_CLASS_Boolean
Object[] iArray = i.toArray();
Object[] jArray = j.toArray();
#elif KEY_CLASS_Character
int[] iArray = i.toArray();
int[] jArray = j.mapToInt(c -> (int)c.charValue()).toArray();
#else
KEY_TYPE_WIDENED[] iArray = i.toArray();
KEY_TYPE_WIDENED[] jArray = j.MAP_TO_KEY_WIDENED(Number::KEY_WIDENED_VALUE).toArray();
#endif
ensure(java.util.Arrays.equals(iArray, jArray), "! sorted arrays equal");
}
/* Now we take out of m everything, and check that it is empty. */
for(java.util.Iterator i=m.iterator(); i.hasNext();) { i.next(); i.remove();}
ensure(m.isEmpty(), "Error (" + seed + "): m is not empty (as it should be)");
#if KEY_CLASS_Integer || KEY_CLASS_Long
m = new OPEN_HASH_BIG_SET(n, f);
t.clear();
int x;
/* Now we torture-test the hash table. This part is implemented only for integers and longs. */
int p = m.key.length - 1;
for(int i=0; i<p; i++) {
for (int j=0; j<20; j++) {
m.add(i+(r.nextInt() % 10)*p);
m.remove(i+(r.nextInt() % 10)*p);
}
for (int j=-10; j<10; j++) m.remove(i+j*p);
}
t.addAll(m);
/* Now all table entries are REMOVED. */
int k = 0;
for(int i=0; i<(p*f)/10; i++) {
for (int j=0; j<10; j++) {
k++;
x = i+(r.nextInt() % 10)*p;
ensure(m.add(x) == t.add(KEY2OBJ(x)), "Error (" + seed + "): m and t differ on a key during torture-test insertion.");
}
}
ensure(m.equals(t), "Error (" + seed + "): !m.equals(t) after torture-test insertion");;
ensure(t.equals(m), "Error (" + seed + "): !t.equals(m) after torture-test insertion");;
for(int i=0; i<(p*f)/10; i++) {
for (int j=0; j<10; j++) {
x = i+(r.nextInt() % 10)*p;
ensure(m.remove(x) == t.remove(KEY2OBJ(x)), "Error (" + seed + "): m and t differ on a key during torture-test removal.");
}
}
ensure(m.equals(t), "Error (" + seed + "): !m.equals(t) after torture-test removal");;
ensure(t.equals(m), "Error (" + seed + "): !t.equals(m) after torture-test removal");;
ensure(m.equals(m.clone()), "Error (" + seed + "): !m.equals(m.clone()) after torture-test removal");;
ensure(((OPEN_HASH_BIG_SET)m.clone()).equals(m), "Error (" + seed + "): !m.clone().equals(m) after torture-test removal");;
m.trim();
ensure(m.equals(t), "Error (" + seed + "): !m.equals(t) after trim()");;
ensure(t.equals(m), "Error (" + seed + "): !t.equals(m) after trim()");;
#endif
System.out.println("Test OK");
return;
}
public static void main(String args[]) throws Exception {
float f = Hash.DEFAULT_LOAD_FACTOR;
int n = Integer.parseInt(args[1]);
if (args.length>2) f = Float.parseFloat(args[2]);
if (args.length > 3) r = new java.util.Random(seed = Long.parseLong(args[3]));
try {
if ("speedTest".equals(args[0]) || "speedComp".equals(args[0])) speedTest(n, f, "speedComp".equals(args[0]));
else if ("test".equals(args[0])) runTest(n, f);
} catch(Throwable e) {
e.printStackTrace(System.err);
System.err.println("seed: " + seed);
throw e;
}
}
#endif
}
|