| 1 | /* |
| 2 | |
| 3 | Derby - Class org.apache.derby.iapi.store.access.BackingStoreHashtable |
| 4 | |
| 5 | Copyright 1999, 2004 The Apache Software Foundation or its licensors, as applicable. |
| 6 | |
| 7 | Licensed under the Apache License, Version 2.0 (the "License"); |
| 8 | you may not use this file except in compliance with the License. |
| 9 | You may obtain a copy of the License at |
| 10 | |
| 11 | http://www.apache.org/licenses/LICENSE-2.0 |
| 12 | |
| 13 | Unless required by applicable law or agreed to in writing, software |
| 14 | distributed under the License is distributed on an "AS IS" BASIS, |
| 15 | WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. |
| 16 | See the License for the specific language governing permissions and |
| 17 | limitations under the License. |
| 18 | |
| 19 | */ |
| 20 | |
| 21 | package org.apache.derby.iapi.store.access; |
| 22 | |
| 23 | import org.apache.derby.iapi.services.sanity.SanityManager; |
| 24 | |
| 25 | import org.apache.derby.iapi.services.io.Storable; |
| 26 | |
| 27 | import org.apache.derby.iapi.error.StandardException; |
| 28 | |
| 29 | import org.apache.derby.iapi.types.CloneableObject; |
| 30 | import org.apache.derby.iapi.types.DataValueDescriptor; |
| 31 | |
| 32 | import org.apache.derby.iapi.services.cache.ClassSize; |
| 33 | |
| 34 | import java.util.Enumeration; |
| 35 | import java.util.Hashtable; |
| 36 | import java.util.Properties; |
| 37 | import java.util.Vector; |
| 38 | import java.util.NoSuchElementException; |
| 39 | |
| 40 | /** |
| 41 | A BackingStoreHashtable is a utility class which will store a set of rows into |
| 42 | an in memory hash table, or overflow the hash table to a tempory on disk |
| 43 | structure. |
| 44 | <p> |
| 45 | All rows must contain the same number of columns, and the column at position |
| 46 | N of all the rows must have the same format id. If the BackingStoreHashtable needs to be |
| 47 | overflowed to disk, then an arbitrary row will be chosen and used as a template |
| 48 | for creating the underlying overflow container. |
| 49 | |
| 50 | <p> |
| 51 | The hash table will be built logically as follows (actual implementation |
| 52 | may differ). The important points are that the hash value is the standard |
| 53 | java hash value on the row[key_column_numbers[0], if key_column_numbers.length is 1, |
| 54 | or row[key_column_numbers[0, 1, ...]] if key_column_numbers.length > 1, |
| 55 | and that duplicate detection is done by the standard java duplicate detection provided by |
| 56 | java.util.Hashtable. |
| 57 | <p> |
| 58 | import java.util.Hashtable; |
| 59 | |
| 60 | hash_table = new Hashtable(); |
| 61 | |
| 62 | Object[] row; |
| 63 | boolean needsToClone = rowSource.needsToClone(); |
| 64 | |
| 65 | while((row = rowSource.getNextRowFromRowSource()) != null) |
| 66 | { |
| 67 | if (needsToClone) |
| 68 | row = clone_row_from_row(row); |
| 69 | |
| 70 | Object key = KeyHasher.buildHashKey(row, key_column_numbers); |
| 71 | |
| 72 | if ((duplicate_value = |
| 73 | hash_table.put(key, row)) != null) |
| 74 | { |
| 75 | Vector row_vec; |
| 76 | |
| 77 | // inserted a duplicate |
| 78 | if ((duplicate_value instanceof vector)) |
| 79 | { |
| 80 | row_vec = (Vector) duplicate_value; |
| 81 | } |
| 82 | else |
| 83 | { |
| 84 | // allocate vector to hold duplicates |
| 85 | row_vec = new Vector(2); |
| 86 | |
| 87 | // insert original row into vector |
| 88 | row_vec.addElement(duplicate_value); |
| 89 | |
| 90 | // put the vector as the data rather than the row |
| 91 | hash_table.put(key, row_vec); |
| 92 | } |
| 93 | |
| 94 | // insert new row into vector |
| 95 | row_vec.addElement(row); |
| 96 | } |
| 97 | } |
| 98 | |
| 99 | **/ |
| 100 | |
| 101 | public class BackingStoreHashtable |
| 102 | { |
| 103 | |
| 104 | /************************************************************************** |
| 105 | * Fields of the class |
| 106 | ************************************************************************** |
| 107 | */ |
| 108 | private TransactionController tc; |
| 109 | private Hashtable hash_table; |
| 110 | private int[] key_column_numbers; |
| 111 | private boolean remove_duplicates; |
| 112 | private boolean skipNullKeyColumns; |
| 113 | private Properties auxillary_runtimestats; |
| 114 | private RowSource row_source; |
| 115 | /* If max_inmemory_rowcnt > 0 then use that to decide when to spill to disk. |
| 116 | * Otherwise compute max_inmemory_size based on the JVM memory size when the BackingStoreHashtable |
| 117 | * is constructed and use that to decide when to spill to disk. |
| 118 | */ |
| 119 | private long max_inmemory_rowcnt; |
| 120 | private long inmemory_rowcnt; |
| 121 | private long max_inmemory_size; |
| 122 | private boolean keepAfterCommit; |
| 123 | |
| 124 | /** |
| 125 | * The estimated number of bytes used by Vector(0) |
| 126 | */ |
| 127 | private final static int vectorSize = ClassSize.estimateBaseFromCatalog(java.util.Vector.class); |
| 128 | |
| 129 | private DiskHashtable diskHashtable; |
| 130 | |
| 131 | /************************************************************************** |
| 132 | * Constructors for This class: |
| 133 | ************************************************************************** |
| 134 | */ |
| 135 | private BackingStoreHashtable(){} |
| 136 | |
| 137 | /** |
| 138 | * Create the BackingStoreHashtable from a row source. |
| 139 | * <p> |
| 140 | * This routine drains the RowSource. The performance characteristics |
| 141 | * depends on the number of rows inserted and the parameters to the |
| 142 | * constructor. |
| 143 | * <p> |
| 144 | * If the number of rows is <= "max_inmemory_rowcnt", then the rows are |
| 145 | * inserted into a java.util.Hashtable. In this case no |
| 146 | * TransactionController is necessary, a "null" tc is valid. |
| 147 | * <p> |
| 148 | * If the number of rows is > "max_inmemory_rowcnt", then the rows will |
| 149 | * be all placed in some sort of Access temporary file on disk. This |
| 150 | * case requires a valid TransactionController. |
| 151 | * |
| 152 | * @param tc An open TransactionController to be used if the |
| 153 | * hash table needs to overflow to disk. |
| 154 | * |
| 155 | * @param row_source RowSource to read rows from. |
| 156 | * |
| 157 | * @param key_column_numbers The column numbers of the columns in the |
| 158 | * scan result row to be the key to the Hashtable. |
| 159 | * "0" is the first column in the scan result |
| 160 | * row (which may be different than the first |
| 161 | * row in the table of the scan). |
| 162 | * |
| 163 | * @param remove_duplicates Should the Hashtable automatically remove |
| 164 | * duplicates, or should it create the Vector of |
| 165 | * duplicates? |
| 166 | * |
| 167 | * @param estimated_rowcnt The estimated number of rows in the hash table. |
| 168 | * Pass in -1 if there is no estimate. |
| 169 | * |
| 170 | * @param max_inmemory_rowcnt |
| 171 | * The maximum number of rows to insert into the |
| 172 | * inmemory Hash table before overflowing to disk. |
| 173 | * Pass in -1 if there is no maximum. |
| 174 | * |
| 175 | * @param initialCapacity If not "-1" used to initialize the java |
| 176 | * Hashtable. |
| 177 | * |
| 178 | * @param loadFactor If not "-1" used to initialize the java |
| 179 | * Hashtable. |
| 180 | * |
| 181 | * @param skipNullKeyColumns Skip rows with a null key column, if true. |
| 182 | * |
| 183 | * @param keepAfterCommit If true the hash table is kept after a commit, |
| 184 | * if false the hash table is dropped on the next commit. |
| 185 | * |
| 186 | * |
| 187 | * @exception StandardException Standard exception policy. |
| 188 | **/ |
| 189 | public BackingStoreHashtable( |
| 190 | TransactionController tc, |
| 191 | RowSource row_source, |
| 192 | int[] key_column_numbers, |
| 193 | boolean remove_duplicates, |
| 194 | long estimated_rowcnt, |
| 195 | long max_inmemory_rowcnt, |
| 196 | int initialCapacity, |
| 197 | float loadFactor, |
| 198 | boolean skipNullKeyColumns, |
| 199 | boolean keepAfterCommit) |
| 200 | throws StandardException |
| 201 | { |
| 202 | this.key_column_numbers = key_column_numbers; |
| 203 | this.remove_duplicates = remove_duplicates; |
| 204 | this.row_source = row_source; |
| 205 | this.skipNullKeyColumns = skipNullKeyColumns; |
| 206 | this.max_inmemory_rowcnt = max_inmemory_rowcnt; |
| 207 | if( max_inmemory_rowcnt > 0) |
| 208 | max_inmemory_size = Long.MAX_VALUE; |
| 209 | else |
| 210 | max_inmemory_size = Runtime.getRuntime().totalMemory()/100; |
| 211 | this.tc = tc; |
| 212 | this.keepAfterCommit = keepAfterCommit; |
| 213 | |
| 214 | Object[] row; |
| 215 | |
| 216 | // use passed in capacity and loadfactor if not -1, you must specify |
| 217 | // capacity if you want to specify loadfactor. |
| 218 | if (initialCapacity != -1) |
| 219 | { |
| 220 | hash_table = |
| 221 | ((loadFactor == -1) ? |
| 222 | new Hashtable(initialCapacity) : |
| 223 | new Hashtable(initialCapacity, loadFactor)); |
| 224 | } |
| 225 | else |
| 226 | { |
| 227 | /* We want to create the hash table based on the estimated row |
| 228 | * count if a) we have an estimated row count (i.e. it's greater |
| 229 | * than zero) and b) we think we can create a hash table to |
| 230 | * hold the estimated row count without running out of memory. |
| 231 | * The check for "b" is required because, for deeply nested |
| 232 | * queries and/or queries with a high number of tables in |
| 233 | * their FROM lists, the optimizer can end up calculating |
| 234 | * some very high row count estimates--even up to the point of |
| 235 | * Double.POSITIVE_INFINITY (see DERBY-1259 for an explanation |
| 236 | * of how that can happen). In that case any attempts to |
| 237 | * create a Hashtable of size estimated_rowcnt can cause |
| 238 | * OutOfMemory errors when we try to create the Hashtable. |
| 239 | * So as a "red flag" for that kind of situation, we check to |
| 240 | * see if the estimated row count is greater than the max |
| 241 | * in-memory size for this table. Unit-wise this comparison |
| 242 | * is relatively meaningless: rows vs bytes. But if our |
| 243 | * estimated row count is greater than the max number of |
| 244 | * in-memory bytes that we're allowed to consume, then |
| 245 | * it's very likely that creating a Hashtable with a capacity |
| 246 | * of estimated_rowcnt will lead to memory problems. So in |
| 247 | * that particular case we leave hash_table null here and |
| 248 | * initialize it further below, using the estimated in-memory |
| 249 | * size of the first row to figure out what a reasonable size |
| 250 | * for the Hashtable might be. |
| 251 | */ |
| 252 | hash_table = |
| 253 | (((estimated_rowcnt <= 0) || (row_source == null)) ? |
| 254 | new Hashtable() : |
| 255 | (estimated_rowcnt < max_inmemory_size) ? |
| 256 | new Hashtable((int) estimated_rowcnt) : |
| 257 | null); |
| 258 | } |
| 259 | |
| 260 | if (row_source != null) |
| 261 | { |
| 262 | boolean needsToClone = row_source.needsToClone(); |
| 263 | |
| 264 | while ((row = getNextRowFromRowSource()) != null) |
| 265 | { |
| 266 | // If we haven't initialized the hash_table yet then that's |
| 267 | // because a Hashtable with capacity estimated_rowcnt would |
| 268 | // probably cause memory problems. So look at the first row |
| 269 | // that we found and use that to create the hash table with |
| 270 | // an initial capacity such that, if it was completely full, |
| 271 | // it would still satisfy the max_inmemory condition. Note |
| 272 | // that this isn't a hard limit--the hash table can grow if |
| 273 | // needed. |
| 274 | if (hash_table == null) |
| 275 | { |
| 276 | // Check to see how much memory we think the first row |
| 277 | // is going to take, and then use that to set the initial |
| 278 | // capacity of the Hashtable. |
| 279 | double rowUsage = getEstimatedMemUsage(row); |
| 280 | hash_table = new Hashtable((int)(max_inmemory_size / rowUsage)); |
| 281 | } |
| 282 | |
| 283 | if (needsToClone) |
| 284 | { |
| 285 | row = cloneRow(row); |
| 286 | } |
| 287 | |
| 288 | Object key = |
| 289 | KeyHasher.buildHashKey(row, key_column_numbers); |
| 290 | |
| 291 | add_row_to_hash_table(hash_table, key, row); |
| 292 | } |
| 293 | } |
| 294 | |
| 295 | // In the (unlikely) event that we received a "red flag" estimated_rowcnt |
| 296 | // that is too big (see comments above), it's possible that, if row_source |
| 297 | // was null or else didn't have any rows, hash_table could still be null |
| 298 | // at this point. So we initialize it to an empty hashtable (representing |
| 299 | // an empty result set) so that calls to other methods on this |
| 300 | // BackingStoreHashtable (ex. "size()") will have a working hash_table |
| 301 | // on which to operate. |
| 302 | if (hash_table == null) |
| 303 | hash_table = new Hashtable(); |
| 304 | } |
| 305 | |
| 306 | /************************************************************************** |
| 307 | * Private/Protected methods of This class: |
| 308 | ************************************************************************** |
| 309 | */ |
| 310 | |
| 311 | /** |
| 312 | * Call method to either get next row or next row with non-null |
| 313 | * key columns. |
| 314 | * |
| 315 | * |
| 316 | * @exception StandardException Standard exception policy. |
| 317 | */ |
| 318 | private Object[] getNextRowFromRowSource() |
| 319 | throws StandardException |
| 320 | { |
| 321 | Object[] row = row_source.getNextRowFromRowSource(); |
| 322 | |
| 323 | if (skipNullKeyColumns) |
| 324 | { |
| 325 | while (row != null) |
| 326 | { |
| 327 | // Are any key columns null? |
| 328 | int index = 0; |
| 329 | for ( ; index < key_column_numbers.length; index++) |
| 330 | { |
| 331 | if (SanityManager.DEBUG) |
| 332 | { |
| 333 | if (! (row[key_column_numbers[index]] instanceof Storable)) |
| 334 | { |
| 335 | SanityManager.THROWASSERT( |
| 336 | "row[key_column_numbers[index]] expected to be Storable, not " + |
| 337 | row[key_column_numbers[index]].getClass().getName()); |
| 338 | } |
| 339 | } |
| 340 | Storable storable = (Storable) row[key_column_numbers[index]]; |
| 341 | if (storable.isNull()) |
| 342 | { |
| 343 | break; |
| 344 | } |
| 345 | } |
| 346 | // No null key columns |
| 347 | if (index == key_column_numbers.length) |
| 348 | { |
| 349 | return row; |
| 350 | } |
| 351 | // 1 or more null key columns |
| 352 | row = row_source.getNextRowFromRowSource(); |
| 353 | } |
| 354 | } |
| 355 | return row; |
| 356 | } |
| 357 | |
| 358 | /** |
| 359 | * Return a cloned copy of the row. |
| 360 | * |
| 361 | * @return The cloned row row to use. |
| 362 | * |
| 363 | * @exception StandardException Standard exception policy. |
| 364 | **/ |
| 365 | static Object[] cloneRow(Object[] old_row) |
| 366 | throws StandardException |
| 367 | { |
| 368 | Object[] new_row = new DataValueDescriptor[old_row.length]; |
| 369 | |
| 370 | // the only difference between getClone and cloneObject is cloneObject does |
| 371 | // not objectify a stream. We use getClone here. Beetle 4896. |
| 372 | for (int i = 0; i < old_row.length; i++) |
| 373 | { |
| 374 | if( old_row[i] != null) |
| 375 | new_row[i] = ((DataValueDescriptor) old_row[i]).getClone(); |
| 376 | } |
| 377 | |
| 378 | return(new_row); |
| 379 | } |
| 380 | |
| 381 | /** |
| 382 | * Do the work to add one row to the hash table. |
| 383 | * <p> |
| 384 | * |
| 385 | * @param row Row to add to the hash table. |
| 386 | * @param hash_table The java HashTable to load into. |
| 387 | * |
| 388 | * @exception StandardException Standard exception policy. |
| 389 | **/ |
| 390 | private void add_row_to_hash_table( |
| 391 | Hashtable hash_table, |
| 392 | Object key, |
| 393 | Object[] row) |
| 394 | throws StandardException |
| 395 | { |
| 396 | if( spillToDisk( hash_table, key, row)) |
| 397 | return; |
| 398 | |
| 399 | Object duplicate_value = null; |
| 400 | |
| 401 | if ((duplicate_value = hash_table.put(key, row)) == null) |
| 402 | doSpaceAccounting( row, false); |
| 403 | else |
| 404 | { |
| 405 | if (!remove_duplicates) |
| 406 | { |
| 407 | Vector row_vec; |
| 408 | |
| 409 | // inserted a duplicate |
| 410 | if ((duplicate_value instanceof Vector)) |
| 411 | { |
| 412 | doSpaceAccounting( row, false); |
| 413 | row_vec = (Vector) duplicate_value; |
| 414 | } |
| 415 | else |
| 416 | { |
| 417 | // allocate vector to hold duplicates |
| 418 | row_vec = new Vector(2); |
| 419 | |
| 420 | // insert original row into vector |
| 421 | row_vec.addElement(duplicate_value); |
| 422 | doSpaceAccounting( row, true); |
| 423 | } |
| 424 | |
| 425 | // insert new row into vector |
| 426 | row_vec.addElement(row); |
| 427 | |
| 428 | // store vector of rows back into hash table, |
| 429 | // overwriting the duplicate key that was |
| 430 | // inserted. |
| 431 | hash_table.put(key, row_vec); |
| 432 | } |
| 433 | } |
| 434 | |
| 435 | row = null; |
| 436 | } |
| 437 | |
| 438 | private void doSpaceAccounting( Object[] row, |
| 439 | boolean firstDuplicate) |
| 440 | { |
| 441 | inmemory_rowcnt++; |
| 442 | if( max_inmemory_rowcnt <= 0) |
| 443 | { |
| 444 | max_inmemory_size -= getEstimatedMemUsage(row); |
| 445 | if( firstDuplicate) |
| 446 | max_inmemory_size -= vectorSize; |
| 447 | } |
| 448 | } // end of doSpaceAccounting |
| 449 | |
| 450 | /** |
| 451 | * Determine whether a new row should be spilled to disk and, if so, do it. |
| 452 | * |
| 453 | * @param hash_table The in-memory hash table |
| 454 | * @param key The row's key |
| 455 | * @param row |
| 456 | * |
| 457 | * @return true if the row was spilled to disk, false if not |
| 458 | * |
| 459 | * @exception StandardException Standard exception policy. |
| 460 | */ |
| 461 | private boolean spillToDisk( Hashtable hash_table, |
| 462 | Object key, |
| 463 | Object[] row) |
| 464 | throws StandardException |
| 465 | { |
| 466 | // Once we have started spilling all new rows will go to disk, even if we have freed up some |
| 467 | // memory by moving duplicates to disk. This simplifies handling of duplicates and accounting. |
| 468 | if( diskHashtable == null) |
| 469 | { |
| 470 | if( max_inmemory_rowcnt > 0) |
| 471 | { |
| 472 | if( inmemory_rowcnt < max_inmemory_rowcnt) |
| 473 | return false; // Do not spill |
| 474 | } |
| 475 | else if( max_inmemory_size > 0) |
| 476 | return false; |
| 477 | // Want to start spilling |
| 478 | if( ! (row instanceof DataValueDescriptor[])) |
| 479 | { |
| 480 | if( SanityManager.DEBUG) |
| 481 | SanityManager.THROWASSERT( "BackingStoreHashtable row is not DataValueDescriptor[]"); |
| 482 | // Do not know how to put it on disk |
| 483 | return false; |
| 484 | } |
| 485 | diskHashtable = new DiskHashtable( tc, |
| 486 | (DataValueDescriptor[]) row, |
| 487 | key_column_numbers, |
| 488 | remove_duplicates, |
| 489 | keepAfterCommit); |
| 490 | } |
| 491 | |
| 492 | Object duplicateValue = hash_table.get( key); |
| 493 | if( duplicateValue != null) |
| 494 | { |
| 495 | if( remove_duplicates) |
| 496 | return true; // a degenerate case of spilling |
| 497 | // If we are keeping duplicates then move all the duplicates from memory to disk |
| 498 | // This simplifies finding duplicates: they are either all in memory or all on disk. |
| 499 | if( duplicateValue instanceof Vector) |
| 500 | { |
| 501 | Vector duplicateVec = (Vector) duplicateValue; |
| 502 | for( int i = duplicateVec.size() - 1; i >= 0; i--) |
| 503 | { |
| 504 | Object[] dupRow = (Object[]) duplicateVec.elementAt(i); |
| 505 | diskHashtable.put( key, dupRow); |
| 506 | } |
| 507 | } |
| 508 | else |
| 509 | diskHashtable.put( key, (Object []) duplicateValue); |
| 510 | hash_table.remove( key); |
| 511 | } |
| 512 | diskHashtable.put( key, row); |
| 513 | return true; |
| 514 | } // end of spillToDisk |
| 515 | |
| 516 | /** |
| 517 | * Take a row and return an estimate as to how much memory that |
| 518 | * row will consume. |
| 519 | * |
| 520 | * @param row The row for which we want to know the memory usage. |
| 521 | * @return A guess as to how much memory the current row will |
| 522 | * use. |
| 523 | */ |
| 524 | private long getEstimatedMemUsage(Object [] row) |
| 525 | { |
| 526 | long rowMem = 0; |
| 527 | for( int i = 0; i < row.length; i++) |
| 528 | { |
| 529 | if (row[i] instanceof DataValueDescriptor) |
| 530 | rowMem += ((DataValueDescriptor) row[i]).estimateMemoryUsage(); |
| 531 | rowMem += ClassSize.refSize; |
| 532 | } |
| 533 | |
| 534 | rowMem += ClassSize.refSize; |
| 535 | return rowMem; |
| 536 | } |
| 537 | |
| 538 | /************************************************************************** |
| 539 | * Public Methods of This class: |
| 540 | ************************************************************************** |
| 541 | */ |
| 542 | |
| 543 | /** |
| 544 | * Close the BackingStoreHashtable. |
| 545 | * <p> |
| 546 | * Perform any necessary cleanup after finishing with the hashtable. Will |
| 547 | * deallocate/dereference objects as necessary. If the table has gone |
| 548 | * to disk this will drop any on disk files used to support the hash table. |
| 549 | * <p> |
| 550 | * |
| 551 | * @exception StandardException Standard exception policy. |
| 552 | **/ |
| 553 | public void close() |
| 554 | throws StandardException |
| 555 | { |
| 556 | hash_table = null; |
| 557 | if( diskHashtable != null) |
| 558 | { |
| 559 | diskHashtable.close(); |
| 560 | diskHashtable = null; |
| 561 | } |
| 562 | return; |
| 563 | } |
| 564 | |
| 565 | /** |
| 566 | * Return an Enumeration that can be used to scan entire table. |
| 567 | * <p> |
| 568 | * RESOLVE - is it worth it to support this routine when we have a |
| 569 | * disk overflow hash table? |
| 570 | * |
| 571 | * @return The Enumeration. |
| 572 | * |
| 573 | * @exception StandardException Standard exception policy. |
| 574 | **/ |
| 575 | public Enumeration elements() |
| 576 | throws StandardException |
| 577 | { |
| 578 | if( diskHashtable == null) |
| 579 | return(hash_table.elements()); |
| 580 | return new BackingStoreHashtableEnumeration(); |
| 581 | } |
| 582 | |
| 583 | /** |
| 584 | * get data associated with given key. |
| 585 | * <p> |
| 586 | * There are 2 different types of objects returned from this routine. |
| 587 | * <p> |
| 588 | * In both cases, the key value is either the object stored in |
| 589 | * row[key_column_numbers[0]], if key_column_numbers.length is 1, |
| 590 | * otherwise it is a KeyHasher containing |
| 591 | * the objects stored in row[key_column_numbers[0, 1, ...]]. |
| 592 | * For every qualifying unique row value an entry is placed into the |
| 593 | * Hashtable. |
| 594 | * <p> |
| 595 | * For row values with duplicates, the value of the data is a Vector of |
| 596 | * rows. |
| 597 | * <p> |
| 598 | * The caller will have to call "instanceof" on the data value |
| 599 | * object if duplicates are expected, to determine if the data value |
| 600 | * of the Hashtable entry is a row or is a Vector of rows. |
| 601 | * <p> |
| 602 | * The BackingStoreHashtable "owns" the objects returned from the get() |
| 603 | * routine. They remain valid until the next access to the |
| 604 | * BackingStoreHashtable. If the client needs to keep references to these |
| 605 | * objects, it should clone copies of the objects. A valid |
| 606 | * BackingStoreHashtable can place all rows into a disk based conglomerate, |
| 607 | * declare a row buffer and then reuse that row buffer for every get() |
| 608 | * call. |
| 609 | * |
| 610 | * @return The value to which the key is mapped in this hashtable; |
| 611 | * null if the key is not mapped to any value in this hashtable. |
| 612 | * |
| 613 | * @param key The key to hash on. |
| 614 | * |
| 615 | * @exception StandardException Standard exception policy. |
| 616 | **/ |
| 617 | public Object get(Object key) |
| 618 | throws StandardException |
| 619 | { |
| 620 | Object obj = hash_table.get(key); |
| 621 | if( diskHashtable == null || obj != null) |
| 622 | return obj; |
| 623 | return diskHashtable.get( key); |
| 624 | } |
| 625 | |
| 626 | /** |
| 627 | * Return runtime stats to caller by adding them to prop. |
| 628 | * <p> |
| 629 | * |
| 630 | * @param prop The set of properties to append to. |
| 631 | * |
| 632 | * @exception StandardException Standard exception policy. |
| 633 | **/ |
| 634 | public void getAllRuntimeStats(Properties prop) |
| 635 | throws StandardException |
| 636 | { |
| 637 | if (auxillary_runtimestats != null) |
| 638 | org.apache.derby.iapi.util.PropertyUtil.copyProperties(auxillary_runtimestats, prop); |
| 639 | } |
| 640 | |
| 641 | /** |
| 642 | * remove a row from the hash table. |
| 643 | * <p> |
| 644 | * a remove of a duplicate removes the entire duplicate list. |
| 645 | * |
| 646 | * @param key The key of the row to remove. |
| 647 | * |
| 648 | * @exception StandardException Standard exception policy. |
| 649 | **/ |
| 650 | public Object remove( |
| 651 | Object key) |
| 652 | throws StandardException |
| 653 | { |
| 654 | Object obj = hash_table.remove(key); |
| 655 | if( obj != null || diskHashtable == null) |
| 656 | return obj; |
| 657 | return diskHashtable.remove(key); |
| 658 | } |
| 659 | |
| 660 | /** |
| 661 | * Set the auxillary runtime stats. |
| 662 | * <p> |
| 663 | * getRuntimeStats() will return both the auxillary stats and any |
| 664 | * BackingStoreHashtable() specific stats. Note that each call to |
| 665 | * setAuxillaryRuntimeStats() overwrites the Property set that was |
| 666 | * set previously. |
| 667 | * |
| 668 | * @param prop The set of properties to append from. |
| 669 | * |
| 670 | * @exception StandardException Standard exception policy. |
| 671 | **/ |
| 672 | public void setAuxillaryRuntimeStats(Properties prop) |
| 673 | throws StandardException |
| 674 | { |
| 675 | auxillary_runtimestats = prop; |
| 676 | } |
| 677 | |
| 678 | /** |
| 679 | * Put a row into the hash table. |
| 680 | * <p> |
| 681 | * The in memory hash table will need to keep a reference to the row |
| 682 | * after the put call has returned. If "needsToClone" is true then the |
| 683 | * hash table will make a copy of the row and put that, else if |
| 684 | * "needsToClone" is false then the hash table will keep a reference to |
| 685 | * the row passed in and no copy will be made. |
| 686 | * <p> |
| 687 | * If rouine returns false, then no reference is kept to the duplicate |
| 688 | * row which was rejected (thus allowing caller to reuse the object). |
| 689 | * |
| 690 | * @param needsToClone does this routine have to make a copy of the row, |
| 691 | * in order to keep a reference to it after return? |
| 692 | * @param row The row to insert into the table. |
| 693 | * |
| 694 | * @return true if row was inserted into the hash table. Returns |
| 695 | * false if the BackingStoreHashtable is eliminating |
| 696 | * duplicates, and the row being inserted is a duplicate, |
| 697 | * or if we are skipping rows with 1 or more null key columns |
| 698 | * and we find a null key column. |
| 699 | * |
| 700 | * @exception StandardException Standard exception policy. |
| 701 | **/ |
| 702 | public boolean put( |
| 703 | boolean needsToClone, |
| 704 | Object[] row) |
| 705 | throws StandardException |
| 706 | { |
| 707 | // Are any key columns null? |
| 708 | if (skipNullKeyColumns) |
| 709 | { |
| 710 | int index = 0; |
| 711 | for ( ; index < key_column_numbers.length; index++) |
| 712 | { |
| 713 | if (SanityManager.DEBUG) |
| 714 | { |
| 715 | if (! (row[key_column_numbers[index]] instanceof Storable)) |
| 716 | { |
| 717 | SanityManager.THROWASSERT( |
| 718 | "row[key_column_numbers[index]] expected to be Storable, not " + |
| 719 | row[key_column_numbers[index]].getClass().getName()); |
| 720 | } |
| 721 | } |
| 722 | Storable storable = (Storable) row[key_column_numbers[index]]; |
| 723 | if (storable.isNull()) |
| 724 | { |
| 725 | return false; |
| 726 | } |
| 727 | } |
| 728 | } |
| 729 | |
| 730 | if (needsToClone) |
| 731 | { |
| 732 | row = cloneRow(row); |
| 733 | } |
| 734 | |
| 735 | Object key = KeyHasher.buildHashKey(row, key_column_numbers); |
| 736 | |
| 737 | if ((remove_duplicates) && (get(key) != null)) |
| 738 | { |
| 739 | return(false); |
| 740 | } |
| 741 | else |
| 742 | { |
| 743 | add_row_to_hash_table(hash_table, key, row); |
| 744 | return(true); |
| 745 | } |
| 746 | } |
| 747 | |
| 748 | /** |
| 749 | * Return number of unique rows in the hash table. |
| 750 | * <p> |
| 751 | * |
| 752 | * @return The number of unique rows in the hash table. |
| 753 | * |
| 754 | * @exception StandardException Standard exception policy. |
| 755 | **/ |
| 756 | public int size() |
| 757 | throws StandardException |
| 758 | { |
| 759 | if( diskHashtable == null) |
| 760 | return(hash_table.size()); |
| 761 | return hash_table.size() + diskHashtable.size(); |
| 762 | } |
| 763 | |
| 764 | private class BackingStoreHashtableEnumeration implements Enumeration |
| 765 | { |
| 766 | private Enumeration memoryEnumeration; |
| 767 | private Enumeration diskEnumeration; |
| 768 | |
| 769 | BackingStoreHashtableEnumeration() |
| 770 | { |
| 771 | memoryEnumeration = hash_table.elements(); |
| 772 | if( diskHashtable != null) |
| 773 | { |
| 774 | try |
| 775 | { |
| 776 | diskEnumeration = diskHashtable.elements(); |
| 777 | } |
| 778 | catch( StandardException se) |
| 779 | { |
| 780 | diskEnumeration = null; |
| 781 | } |
| 782 | } |
| 783 | } |
| 784 | |
| 785 | public boolean hasMoreElements() |
| 786 | { |
| 787 | if( memoryEnumeration != null) |
| 788 | { |
| 789 | if( memoryEnumeration.hasMoreElements()) |
| 790 | return true; |
| 791 | memoryEnumeration = null; |
| 792 | } |
| 793 | if( diskEnumeration == null) |
| 794 | return false; |
| 795 | return diskEnumeration.hasMoreElements(); |
| 796 | } |
| 797 | |
| 798 | public Object nextElement() throws NoSuchElementException |
| 799 | { |
| 800 | if( memoryEnumeration != null) |
| 801 | { |
| 802 | if( memoryEnumeration.hasMoreElements()) |
| 803 | return memoryEnumeration.nextElement(); |
| 804 | memoryEnumeration = null; |
| 805 | } |
| 806 | return diskEnumeration.nextElement(); |
| 807 | } |
| 808 | } // end of class BackingStoreHashtableEnumeration |
| 809 | } |