001/* 002 * Licensed to the Apache Software Foundation (ASF) under one 003 * or more contributor license agreements. See the NOTICE file 004 * distributed with this work for additional information 005 * regarding copyright ownership. The ASF licenses this file 006 * to you under the Apache License, Version 2.0 (the 007 * "License"); you may not use this file except in compliance 008 * with the License. You may obtain a copy of the License at 009 * 010 * http://www.apache.org/licenses/LICENSE-2.0 011 * 012 * Unless required by applicable law or agreed to in writing, software 013 * distributed under the License is distributed on an "AS IS" BASIS, 014 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 015 * See the License for the specific language governing permissions and 016 * limitations under the License. 017 */ 018package org.apache.hadoop.hbase.client; 019 020import java.io.IOException; 021import java.util.ArrayList; 022import java.util.HashMap; 023import java.util.List; 024import java.util.Map; 025import java.util.NavigableSet; 026import java.util.TreeMap; 027import java.util.TreeSet; 028import java.util.stream.Collectors; 029import org.apache.hadoop.hbase.HConstants; 030import org.apache.hadoop.hbase.client.metrics.ScanMetrics; 031import org.apache.hadoop.hbase.filter.Filter; 032import org.apache.hadoop.hbase.filter.IncompatibleFilterException; 033import org.apache.hadoop.hbase.io.TimeRange; 034import org.apache.hadoop.hbase.security.access.Permission; 035import org.apache.hadoop.hbase.security.visibility.Authorizations; 036import org.apache.hadoop.hbase.util.Bytes; 037import org.apache.yetus.audience.InterfaceAudience; 038import org.slf4j.Logger; 039import org.slf4j.LoggerFactory; 040 041/** 042 * Used to perform Scan operations. 043 * <p> 044 * All operations are identical to {@link Get} with the exception of instantiation. Rather than 045 * specifying a single row, an optional startRow and stopRow may be defined. If rows are not 046 * specified, the Scanner will iterate over all rows. 047 * <p> 048 * To get all columns from all rows of a Table, create an instance with no constraints; use the 049 * {@link #Scan()} constructor. To constrain the scan to specific column families, call 050 * {@link #addFamily(byte[]) addFamily} for each family to retrieve on your Scan instance. 051 * <p> 052 * To get specific columns, call {@link #addColumn(byte[], byte[]) addColumn} for each column to 053 * retrieve. 054 * <p> 055 * To only retrieve columns within a specific range of version timestamps, call 056 * {@link #setTimeRange(long, long) setTimeRange}. 057 * <p> 058 * To only retrieve columns with a specific timestamp, call {@link #setTimestamp(long) setTimestamp} 059 * . 060 * <p> 061 * To limit the number of versions of each column to be returned, call {@link #readVersions(int)}. 062 * <p> 063 * To limit the maximum number of values returned for each call to next(), call 064 * {@link #setBatch(int) setBatch}. 065 * <p> 066 * To add a filter, call {@link #setFilter(org.apache.hadoop.hbase.filter.Filter) setFilter}. 067 * <p> 068 * For small scan, it is deprecated in 2.0.0. Now we have a {@link #setLimit(int)} method in Scan 069 * object which is used to tell RS how many rows we want. If the rows return reaches the limit, the 070 * RS will close the RegionScanner automatically. And we will also fetch data when openScanner in 071 * the new implementation, this means we can also finish a scan operation in one rpc call. And we 072 * have also introduced a {@link #setReadType(ReadType)} method. You can use this method to tell RS 073 * to use pread explicitly. 074 * <p> 075 * Expert: To explicitly disable server-side block caching for this scan, execute 076 * {@link #setCacheBlocks(boolean)}. 077 * <p> 078 * <em>Note:</em> Usage alters Scan instances. Internally, attributes are updated as the Scan runs 079 * and if enabled, metrics accumulate in the Scan instance. Be aware this is the case when you go to 080 * clone a Scan instance or if you go to reuse a created Scan instance; safer is create a Scan 081 * instance per usage. 082 */ 083@InterfaceAudience.Public 084public class Scan extends Query { 085 private static final Logger LOG = LoggerFactory.getLogger(Scan.class); 086 087 private static final String RAW_ATTR = "_raw_"; 088 089 private byte[] startRow = HConstants.EMPTY_START_ROW; 090 private boolean includeStartRow = true; 091 private byte[] stopRow = HConstants.EMPTY_END_ROW; 092 private boolean includeStopRow = false; 093 private int maxVersions = 1; 094 private int batch = -1; 095 096 /** 097 * Partial {@link Result}s are {@link Result}s must be combined to form a complete {@link Result}. 098 * The {@link Result}s had to be returned in fragments (i.e. as partials) because the size of the 099 * cells in the row exceeded max result size on the server. Typically partial results will be 100 * combined client side into complete results before being delivered to the caller. However, if 101 * this flag is set, the caller is indicating that they do not mind seeing partial results (i.e. 102 * they understand that the results returned from the Scanner may only represent part of a 103 * particular row). In such a case, any attempt to combine the partials into a complete result on 104 * the client side will be skipped, and the caller will be able to see the exact results returned 105 * from the server. 106 */ 107 private boolean allowPartialResults = false; 108 109 private int storeLimit = -1; 110 private int storeOffset = 0; 111 112 private static final String SCAN_ATTRIBUTES_METRICS_ENABLE = "scan.attributes.metrics.enable"; 113 114 // If an application wants to use multiple scans over different tables each scan must 115 // define this attribute with the appropriate table name by calling 116 // scan.setAttribute(Scan.SCAN_ATTRIBUTES_TABLE_NAME, Bytes.toBytes(tableName)) 117 static public final String SCAN_ATTRIBUTES_TABLE_NAME = "scan.attributes.table.name"; 118 119 /** 120 * -1 means no caching specified and the value of {@link HConstants#HBASE_CLIENT_SCANNER_CACHING} 121 * (default to {@link HConstants#DEFAULT_HBASE_CLIENT_SCANNER_CACHING}) will be used 122 */ 123 private int caching = -1; 124 private long maxResultSize = -1; 125 private boolean cacheBlocks = true; 126 private boolean reversed = false; 127 private TimeRange tr = TimeRange.allTime(); 128 private Map<byte[], NavigableSet<byte[]>> familyMap = 129 new TreeMap<byte[], NavigableSet<byte[]>>(Bytes.BYTES_COMPARATOR); 130 private Boolean asyncPrefetch = null; 131 132 /** 133 * Parameter name for client scanner sync/async prefetch toggle. When using async scanner, 134 * prefetching data from the server is done at the background. The parameter currently won't have 135 * any effect in the case that the user has set Scan#setSmall or Scan#setReversed 136 */ 137 public static final String HBASE_CLIENT_SCANNER_ASYNC_PREFETCH = 138 "hbase.client.scanner.async.prefetch"; 139 140 /** 141 * Default value of {@link #HBASE_CLIENT_SCANNER_ASYNC_PREFETCH}. 142 */ 143 public static final boolean DEFAULT_HBASE_CLIENT_SCANNER_ASYNC_PREFETCH = false; 144 145 /** 146 * The mvcc read point to use when open a scanner. Remember to clear it after switching regions as 147 * the mvcc is only valid within region scope. 148 */ 149 private long mvccReadPoint = -1L; 150 151 /** 152 * The number of rows we want for this scan. We will terminate the scan if the number of return 153 * rows reaches this value. 154 */ 155 private int limit = -1; 156 157 /** 158 * Control whether to use pread at server side. 159 */ 160 private ReadType readType = ReadType.DEFAULT; 161 162 private boolean needCursorResult = false; 163 164 /** 165 * Create a Scan operation across all rows. 166 */ 167 public Scan() { 168 } 169 170 /** 171 * Creates a new instance of this class while copying all values. 172 * @param scan The scan instance to copy from. 173 * @throws IOException When copying the values fails. 174 */ 175 public Scan(Scan scan) throws IOException { 176 startRow = scan.getStartRow(); 177 includeStartRow = scan.includeStartRow(); 178 stopRow = scan.getStopRow(); 179 includeStopRow = scan.includeStopRow(); 180 maxVersions = scan.getMaxVersions(); 181 batch = scan.getBatch(); 182 storeLimit = scan.getMaxResultsPerColumnFamily(); 183 storeOffset = scan.getRowOffsetPerColumnFamily(); 184 caching = scan.getCaching(); 185 maxResultSize = scan.getMaxResultSize(); 186 cacheBlocks = scan.getCacheBlocks(); 187 filter = scan.getFilter(); // clone? 188 loadColumnFamiliesOnDemand = scan.getLoadColumnFamiliesOnDemandValue(); 189 consistency = scan.getConsistency(); 190 this.setIsolationLevel(scan.getIsolationLevel()); 191 reversed = scan.isReversed(); 192 asyncPrefetch = scan.isAsyncPrefetch(); 193 allowPartialResults = scan.getAllowPartialResults(); 194 tr = scan.getTimeRange(); // TimeRange is immutable 195 Map<byte[], NavigableSet<byte[]>> fams = scan.getFamilyMap(); 196 for (Map.Entry<byte[], NavigableSet<byte[]>> entry : fams.entrySet()) { 197 byte[] fam = entry.getKey(); 198 NavigableSet<byte[]> cols = entry.getValue(); 199 if (cols != null && cols.size() > 0) { 200 for (byte[] col : cols) { 201 addColumn(fam, col); 202 } 203 } else { 204 addFamily(fam); 205 } 206 } 207 for (Map.Entry<String, byte[]> attr : scan.getAttributesMap().entrySet()) { 208 setAttribute(attr.getKey(), attr.getValue()); 209 } 210 for (Map.Entry<byte[], TimeRange> entry : scan.getColumnFamilyTimeRange().entrySet()) { 211 TimeRange tr = entry.getValue(); 212 setColumnFamilyTimeRange(entry.getKey(), tr.getMin(), tr.getMax()); 213 } 214 this.mvccReadPoint = scan.getMvccReadPoint(); 215 this.limit = scan.getLimit(); 216 this.needCursorResult = scan.isNeedCursorResult(); 217 setPriority(scan.getPriority()); 218 readType = scan.getReadType(); 219 super.setReplicaId(scan.getReplicaId()); 220 } 221 222 /** 223 * Builds a scan object with the same specs as get. 224 * @param get get to model scan after 225 */ 226 public Scan(Get get) { 227 this.startRow = get.getRow(); 228 this.includeStartRow = true; 229 this.stopRow = get.getRow(); 230 this.includeStopRow = true; 231 this.filter = get.getFilter(); 232 this.cacheBlocks = get.getCacheBlocks(); 233 this.maxVersions = get.getMaxVersions(); 234 this.storeLimit = get.getMaxResultsPerColumnFamily(); 235 this.storeOffset = get.getRowOffsetPerColumnFamily(); 236 this.tr = get.getTimeRange(); 237 this.familyMap = get.getFamilyMap(); 238 this.asyncPrefetch = false; 239 this.consistency = get.getConsistency(); 240 this.setIsolationLevel(get.getIsolationLevel()); 241 this.loadColumnFamiliesOnDemand = get.getLoadColumnFamiliesOnDemandValue(); 242 for (Map.Entry<String, byte[]> attr : get.getAttributesMap().entrySet()) { 243 setAttribute(attr.getKey(), attr.getValue()); 244 } 245 for (Map.Entry<byte[], TimeRange> entry : get.getColumnFamilyTimeRange().entrySet()) { 246 TimeRange tr = entry.getValue(); 247 setColumnFamilyTimeRange(entry.getKey(), tr.getMin(), tr.getMax()); 248 } 249 this.mvccReadPoint = -1L; 250 setPriority(get.getPriority()); 251 super.setReplicaId(get.getReplicaId()); 252 } 253 254 public boolean isGetScan() { 255 return includeStartRow && includeStopRow 256 && ClientUtil.areScanStartRowAndStopRowEqual(this.startRow, this.stopRow); 257 } 258 259 /** 260 * Get all columns from the specified family. 261 * <p> 262 * Overrides previous calls to addColumn for this family. 263 * @param family family name 264 */ 265 public Scan addFamily(byte[] family) { 266 familyMap.remove(family); 267 familyMap.put(family, null); 268 return this; 269 } 270 271 /** 272 * Get the column from the specified family with the specified qualifier. 273 * <p> 274 * Overrides previous calls to addFamily for this family. 275 * @param family family name 276 * @param qualifier column qualifier 277 */ 278 public Scan addColumn(byte[] family, byte[] qualifier) { 279 NavigableSet<byte[]> set = familyMap.get(family); 280 if (set == null) { 281 set = new TreeSet<>(Bytes.BYTES_COMPARATOR); 282 familyMap.put(family, set); 283 } 284 if (qualifier == null) { 285 qualifier = HConstants.EMPTY_BYTE_ARRAY; 286 } 287 set.add(qualifier); 288 return this; 289 } 290 291 /** 292 * Get versions of columns only within the specified timestamp range, [minStamp, maxStamp). Note, 293 * default maximum versions to return is 1. If your time range spans more than one version and you 294 * want all versions returned, up the number of versions beyond the default. 295 * @param minStamp minimum timestamp value, inclusive 296 * @param maxStamp maximum timestamp value, exclusive 297 * @see #readAllVersions() 298 * @see #readVersions(int) 299 */ 300 public Scan setTimeRange(long minStamp, long maxStamp) throws IOException { 301 tr = TimeRange.between(minStamp, maxStamp); 302 return this; 303 } 304 305 /** 306 * Get versions of columns with the specified timestamp. Note, default maximum versions to return 307 * is 1. If your time range spans more than one version and you want all versions returned, up the 308 * number of versions beyond the defaut. 309 * @param timestamp version timestamp 310 * @see #readAllVersions() 311 * @see #readVersions(int) 312 */ 313 public Scan setTimestamp(long timestamp) { 314 try { 315 tr = TimeRange.at(timestamp); 316 } catch (Exception e) { 317 // This should never happen, unless integer overflow or something extremely wrong... 318 LOG.error("TimeRange failed, likely caused by integer overflow. ", e); 319 throw e; 320 } 321 322 return this; 323 } 324 325 @Override 326 public Scan setColumnFamilyTimeRange(byte[] cf, long minStamp, long maxStamp) { 327 super.setColumnFamilyTimeRange(cf, minStamp, maxStamp); 328 return this; 329 } 330 331 /** 332 * Set the start row of the scan. 333 * <p> 334 * If the specified row does not exist, the Scanner will start from the next closest row after the 335 * specified row. 336 * <p> 337 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 338 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 339 * unexpected or even undefined. 340 * </p> 341 * @param startRow row to start scanner at or after 342 * @throws IllegalArgumentException if startRow does not meet criteria for a row key (when length 343 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 344 */ 345 public Scan withStartRow(byte[] startRow) { 346 return withStartRow(startRow, true); 347 } 348 349 /** 350 * Set the start row of the scan. 351 * <p> 352 * If the specified row does not exist, or the {@code inclusive} is {@code false}, the Scanner 353 * will start from the next closest row after the specified row. 354 * <p> 355 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 356 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 357 * unexpected or even undefined. 358 * </p> 359 * @param startRow row to start scanner at or after 360 * @param inclusive whether we should include the start row when scan 361 * @throws IllegalArgumentException if startRow does not meet criteria for a row key (when length 362 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 363 */ 364 public Scan withStartRow(byte[] startRow, boolean inclusive) { 365 if (Bytes.len(startRow) > HConstants.MAX_ROW_LENGTH) { 366 throw new IllegalArgumentException("startRow's length must be less than or equal to " 367 + HConstants.MAX_ROW_LENGTH + " to meet the criteria" + " for a row key."); 368 } 369 this.startRow = startRow; 370 this.includeStartRow = inclusive; 371 return this; 372 } 373 374 /** 375 * Set the stop row of the scan. 376 * <p> 377 * The scan will include rows that are lexicographically less than the provided stopRow. 378 * <p> 379 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 380 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 381 * unexpected or even undefined. 382 * </p> 383 * @param stopRow row to end at (exclusive) 384 * @throws IllegalArgumentException if stopRow does not meet criteria for a row key (when length 385 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 386 */ 387 public Scan withStopRow(byte[] stopRow) { 388 return withStopRow(stopRow, false); 389 } 390 391 /** 392 * Set the stop row of the scan. 393 * <p> 394 * The scan will include rows that are lexicographically less than (or equal to if 395 * {@code inclusive} is {@code true}) the provided stopRow. 396 * <p> 397 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 398 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 399 * unexpected or even undefined. 400 * </p> 401 * @param stopRow row to end at 402 * @param inclusive whether we should include the stop row when scan 403 * @throws IllegalArgumentException if stopRow does not meet criteria for a row key (when length 404 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 405 */ 406 public Scan withStopRow(byte[] stopRow, boolean inclusive) { 407 if (Bytes.len(stopRow) > HConstants.MAX_ROW_LENGTH) { 408 throw new IllegalArgumentException("stopRow's length must be less than or equal to " 409 + HConstants.MAX_ROW_LENGTH + " to meet the criteria" + " for a row key."); 410 } 411 this.stopRow = stopRow; 412 this.includeStopRow = inclusive; 413 return this; 414 } 415 416 /** 417 * <p> 418 * Set a filter (using stopRow and startRow) so the result set only contains rows where the rowKey 419 * starts with the specified prefix. 420 * </p> 421 * <p> 422 * This is a utility method that converts the desired rowPrefix into the appropriate values for 423 * the startRow and stopRow to achieve the desired result. 424 * </p> 425 * <p> 426 * This can safely be used in combination with setFilter. 427 * </p> 428 * <p> 429 * <strong>This CANNOT be used in combination with withStartRow and/or withStopRow.</strong> Such 430 * a combination will yield unexpected and even undefined results. 431 * </p> 432 * @param rowPrefix the prefix all rows must start with. (Set <i>null</i> to remove the filter.) 433 * @deprecated since 2.5.0, will be removed in 4.0.0. The name of this method is considered to be 434 * confusing as it does not use a {@link Filter} but uses setting the startRow and 435 * stopRow instead. Use {@link #setStartStopRowForPrefixScan(byte[])} instead. 436 */ 437 @Deprecated 438 public Scan setRowPrefixFilter(byte[] rowPrefix) { 439 return setStartStopRowForPrefixScan(rowPrefix); 440 } 441 442 /** 443 * <p> 444 * Set a filter (using stopRow and startRow) so the result set only contains rows where the rowKey 445 * starts with the specified prefix. 446 * </p> 447 * <p> 448 * This is a utility method that converts the desired rowPrefix into the appropriate values for 449 * the startRow and stopRow to achieve the desired result. 450 * </p> 451 * <p> 452 * This can safely be used in combination with setFilter. 453 * </p> 454 * <p> 455 * <strong>This CANNOT be used in combination with withStartRow and/or withStopRow.</strong> Such 456 * a combination will yield unexpected and even undefined results. 457 * </p> 458 * @param rowPrefix the prefix all rows must start with. (Set <i>null</i> to remove the filter.) 459 */ 460 public Scan setStartStopRowForPrefixScan(byte[] rowPrefix) { 461 if (rowPrefix == null) { 462 withStartRow(HConstants.EMPTY_START_ROW); 463 withStopRow(HConstants.EMPTY_END_ROW); 464 } else { 465 this.withStartRow(rowPrefix); 466 this.withStopRow(ClientUtil.calculateTheClosestNextRowKeyForPrefix(rowPrefix)); 467 } 468 return this; 469 } 470 471 /** 472 * Get all available versions. 473 */ 474 public Scan readAllVersions() { 475 this.maxVersions = Integer.MAX_VALUE; 476 return this; 477 } 478 479 /** 480 * Get up to the specified number of versions of each column. 481 * @param versions specified number of versions for each column 482 */ 483 public Scan readVersions(int versions) { 484 this.maxVersions = versions; 485 return this; 486 } 487 488 /** 489 * Set the maximum number of cells to return for each call to next(). Callers should be aware that 490 * this is not equivalent to calling {@link #setAllowPartialResults(boolean)}. If you don't allow 491 * partial results, the number of cells in each Result must equal to your batch setting unless it 492 * is the last Result for current row. So this method is helpful in paging queries. If you just 493 * want to prevent OOM at client, use setAllowPartialResults(true) is better. 494 * @param batch the maximum number of values 495 * @see Result#mayHaveMoreCellsInRow() 496 */ 497 public Scan setBatch(int batch) { 498 if (this.hasFilter() && this.filter.hasFilterRow()) { 499 throw new IncompatibleFilterException( 500 "Cannot set batch on a scan using a filter" + " that returns true for filter.hasFilterRow"); 501 } 502 this.batch = batch; 503 return this; 504 } 505 506 /** 507 * Set the maximum number of values to return per row per Column Family 508 * @param limit the maximum number of values returned / row / CF 509 */ 510 public Scan setMaxResultsPerColumnFamily(int limit) { 511 this.storeLimit = limit; 512 return this; 513 } 514 515 /** 516 * Set offset for the row per Column Family. 517 * @param offset is the number of kvs that will be skipped. 518 */ 519 public Scan setRowOffsetPerColumnFamily(int offset) { 520 this.storeOffset = offset; 521 return this; 522 } 523 524 /** 525 * Set the number of rows for caching that will be passed to scanners. If not set, the 526 * Configuration setting {@link HConstants#HBASE_CLIENT_SCANNER_CACHING} will apply. Higher 527 * caching values will enable faster scanners but will use more memory. 528 * @param caching the number of rows for caching 529 */ 530 public Scan setCaching(int caching) { 531 this.caching = caching; 532 return this; 533 } 534 535 /** Returns the maximum result size in bytes. See {@link #setMaxResultSize(long)} */ 536 public long getMaxResultSize() { 537 return maxResultSize; 538 } 539 540 /** 541 * Set the maximum result size. The default is -1; this means that no specific maximum result size 542 * will be set for this scan, and the global configured value will be used instead. (Defaults to 543 * unlimited). 544 * @param maxResultSize The maximum result size in bytes. 545 */ 546 public Scan setMaxResultSize(long maxResultSize) { 547 this.maxResultSize = maxResultSize; 548 return this; 549 } 550 551 @Override 552 public Scan setFilter(Filter filter) { 553 super.setFilter(filter); 554 return this; 555 } 556 557 /** 558 * Setting the familyMap 559 * @param familyMap map of family to qualifier 560 */ 561 public Scan setFamilyMap(Map<byte[], NavigableSet<byte[]>> familyMap) { 562 this.familyMap = familyMap; 563 return this; 564 } 565 566 /** 567 * Getting the familyMap 568 */ 569 public Map<byte[], NavigableSet<byte[]>> getFamilyMap() { 570 return this.familyMap; 571 } 572 573 /** Returns the number of families in familyMap */ 574 public int numFamilies() { 575 if (hasFamilies()) { 576 return this.familyMap.size(); 577 } 578 return 0; 579 } 580 581 /** Returns true if familyMap is non empty, false otherwise */ 582 public boolean hasFamilies() { 583 return !this.familyMap.isEmpty(); 584 } 585 586 /** Returns the keys of the familyMap */ 587 public byte[][] getFamilies() { 588 if (hasFamilies()) { 589 return this.familyMap.keySet().toArray(new byte[0][0]); 590 } 591 return null; 592 } 593 594 /** Returns the startrow */ 595 public byte[] getStartRow() { 596 return this.startRow; 597 } 598 599 /** Returns if we should include start row when scan */ 600 public boolean includeStartRow() { 601 return includeStartRow; 602 } 603 604 /** Returns the stoprow */ 605 public byte[] getStopRow() { 606 return this.stopRow; 607 } 608 609 /** Returns if we should include stop row when scan */ 610 public boolean includeStopRow() { 611 return includeStopRow; 612 } 613 614 /** Returns the max number of versions to fetch */ 615 public int getMaxVersions() { 616 return this.maxVersions; 617 } 618 619 /** Returns maximum number of values to return for a single call to next() */ 620 public int getBatch() { 621 return this.batch; 622 } 623 624 /** Returns maximum number of values to return per row per CF */ 625 public int getMaxResultsPerColumnFamily() { 626 return this.storeLimit; 627 } 628 629 /** 630 * Method for retrieving the scan's offset per row per column family (#kvs to be skipped) 631 * @return row offset 632 */ 633 public int getRowOffsetPerColumnFamily() { 634 return this.storeOffset; 635 } 636 637 /** Returns caching the number of rows fetched when calling next on a scanner */ 638 public int getCaching() { 639 return this.caching; 640 } 641 642 /** Returns TimeRange */ 643 public TimeRange getTimeRange() { 644 return this.tr; 645 } 646 647 /** Returns RowFilter */ 648 @Override 649 public Filter getFilter() { 650 return filter; 651 } 652 653 /** Returns true is a filter has been specified, false if not */ 654 public boolean hasFilter() { 655 return filter != null; 656 } 657 658 /** 659 * Set whether blocks should be cached for this Scan. 660 * <p> 661 * This is true by default. When true, default settings of the table and family are used (this 662 * will never override caching blocks if the block cache is disabled for that family or entirely). 663 * @param cacheBlocks if false, default settings are overridden and blocks will not be cached 664 */ 665 public Scan setCacheBlocks(boolean cacheBlocks) { 666 this.cacheBlocks = cacheBlocks; 667 return this; 668 } 669 670 /** 671 * Get whether blocks should be cached for this Scan. 672 * @return true if default caching should be used, false if blocks should not be cached 673 */ 674 public boolean getCacheBlocks() { 675 return cacheBlocks; 676 } 677 678 /** 679 * Set whether this scan is a reversed one 680 * <p> 681 * This is false by default which means forward(normal) scan. 682 * @param reversed if true, scan will be backward order 683 */ 684 public Scan setReversed(boolean reversed) { 685 this.reversed = reversed; 686 return this; 687 } 688 689 /** 690 * Get whether this scan is a reversed one. 691 * @return true if backward scan, false if forward(default) scan 692 */ 693 public boolean isReversed() { 694 return reversed; 695 } 696 697 /** 698 * Setting whether the caller wants to see the partial results when server returns 699 * less-than-expected cells. It is helpful while scanning a huge row to prevent OOM at client. By 700 * default this value is false and the complete results will be assembled client side before being 701 * delivered to the caller. 702 * @see Result#mayHaveMoreCellsInRow() 703 * @see #setBatch(int) 704 */ 705 public Scan setAllowPartialResults(final boolean allowPartialResults) { 706 this.allowPartialResults = allowPartialResults; 707 return this; 708 } 709 710 /** 711 * Returns true when the constructor of this scan understands that the results they will see may 712 * only represent a partial portion of a row. The entire row would be retrieved by subsequent 713 * calls to {@link ResultScanner#next()} 714 */ 715 public boolean getAllowPartialResults() { 716 return allowPartialResults; 717 } 718 719 @Override 720 public Scan setLoadColumnFamiliesOnDemand(boolean value) { 721 super.setLoadColumnFamiliesOnDemand(value); 722 return this; 723 } 724 725 /** 726 * Compile the table and column family (i.e. schema) information into a String. Useful for parsing 727 * and aggregation by debugging, logging, and administration tools. 728 */ 729 @Override 730 public Map<String, Object> getFingerprint() { 731 Map<String, Object> map = new HashMap<>(); 732 List<String> families = new ArrayList<>(); 733 if (this.familyMap.isEmpty()) { 734 map.put("families", "ALL"); 735 return map; 736 } else { 737 map.put("families", families); 738 } 739 for (Map.Entry<byte[], NavigableSet<byte[]>> entry : this.familyMap.entrySet()) { 740 families.add(Bytes.toStringBinary(entry.getKey())); 741 } 742 return map; 743 } 744 745 /** 746 * Compile the details beyond the scope of getFingerprint (row, columns, timestamps, etc.) into a 747 * Map along with the fingerprinted information. Useful for debugging, logging, and administration 748 * tools. 749 * @param maxCols a limit on the number of columns output prior to truncation 750 */ 751 @Override 752 public Map<String, Object> toMap(int maxCols) { 753 // start with the fingerprint map and build on top of it 754 Map<String, Object> map = getFingerprint(); 755 // map from families to column list replaces fingerprint's list of families 756 Map<String, List<String>> familyColumns = new HashMap<>(); 757 map.put("families", familyColumns); 758 // add scalar information first 759 map.put("startRow", Bytes.toStringBinary(this.startRow)); 760 map.put("stopRow", Bytes.toStringBinary(this.stopRow)); 761 map.put("maxVersions", this.maxVersions); 762 map.put("batch", this.batch); 763 map.put("caching", this.caching); 764 map.put("maxResultSize", this.maxResultSize); 765 map.put("cacheBlocks", this.cacheBlocks); 766 map.put("loadColumnFamiliesOnDemand", this.loadColumnFamiliesOnDemand); 767 List<Long> timeRange = new ArrayList<>(2); 768 timeRange.add(this.tr.getMin()); 769 timeRange.add(this.tr.getMax()); 770 map.put("timeRange", timeRange); 771 int colCount = 0; 772 // iterate through affected families and list out up to maxCols columns 773 for (Map.Entry<byte[], NavigableSet<byte[]>> entry : this.familyMap.entrySet()) { 774 List<String> columns = new ArrayList<>(); 775 familyColumns.put(Bytes.toStringBinary(entry.getKey()), columns); 776 if (entry.getValue() == null) { 777 colCount++; 778 --maxCols; 779 columns.add("ALL"); 780 } else { 781 colCount += entry.getValue().size(); 782 if (maxCols <= 0) { 783 continue; 784 } 785 for (byte[] column : entry.getValue()) { 786 if (--maxCols <= 0) { 787 continue; 788 } 789 columns.add(Bytes.toStringBinary(column)); 790 } 791 } 792 } 793 map.put("totalColumns", colCount); 794 if (this.filter != null) { 795 map.put("filter", this.filter.toString()); 796 } 797 // add the id if set 798 if (getId() != null) { 799 map.put("id", getId()); 800 } 801 map.put("includeStartRow", includeStartRow); 802 map.put("includeStopRow", includeStopRow); 803 map.put("allowPartialResults", allowPartialResults); 804 map.put("storeLimit", storeLimit); 805 map.put("storeOffset", storeOffset); 806 map.put("reversed", reversed); 807 if (null != asyncPrefetch) { 808 map.put("asyncPrefetch", asyncPrefetch); 809 } 810 map.put("mvccReadPoint", mvccReadPoint); 811 map.put("limit", limit); 812 map.put("readType", readType); 813 map.put("needCursorResult", needCursorResult); 814 map.put("targetReplicaId", targetReplicaId); 815 map.put("consistency", consistency); 816 if (!colFamTimeRangeMap.isEmpty()) { 817 Map<String, List<Long>> colFamTimeRangeMapStr = colFamTimeRangeMap.entrySet().stream() 818 .collect(Collectors.toMap((e) -> Bytes.toStringBinary(e.getKey()), e -> { 819 TimeRange value = e.getValue(); 820 List<Long> rangeList = new ArrayList<>(); 821 rangeList.add(value.getMin()); 822 rangeList.add(value.getMax()); 823 return rangeList; 824 })); 825 826 map.put("colFamTimeRangeMap", colFamTimeRangeMapStr); 827 } 828 map.put("priority", getPriority()); 829 return map; 830 } 831 832 /** 833 * Enable/disable "raw" mode for this scan. If "raw" is enabled the scan will return all delete 834 * marker and deleted rows that have not been collected, yet. This is mostly useful for Scan on 835 * column families that have KEEP_DELETED_ROWS enabled. It is an error to specify any column when 836 * "raw" is set. 837 * @param raw True/False to enable/disable "raw" mode. 838 */ 839 public Scan setRaw(boolean raw) { 840 setAttribute(RAW_ATTR, Bytes.toBytes(raw)); 841 return this; 842 } 843 844 /** Returns True if this Scan is in "raw" mode. */ 845 public boolean isRaw() { 846 byte[] attr = getAttribute(RAW_ATTR); 847 return attr == null ? false : Bytes.toBoolean(attr); 848 } 849 850 @Override 851 public Scan setAttribute(String name, byte[] value) { 852 super.setAttribute(name, value); 853 return this; 854 } 855 856 @Override 857 public Scan setId(String id) { 858 super.setId(id); 859 return this; 860 } 861 862 @Override 863 public Scan setAuthorizations(Authorizations authorizations) { 864 super.setAuthorizations(authorizations); 865 return this; 866 } 867 868 @Override 869 public Scan setACL(Map<String, Permission> perms) { 870 super.setACL(perms); 871 return this; 872 } 873 874 @Override 875 public Scan setACL(String user, Permission perms) { 876 super.setACL(user, perms); 877 return this; 878 } 879 880 @Override 881 public Scan setConsistency(Consistency consistency) { 882 super.setConsistency(consistency); 883 return this; 884 } 885 886 @Override 887 public Scan setReplicaId(int Id) { 888 super.setReplicaId(Id); 889 return this; 890 } 891 892 @Override 893 public Scan setIsolationLevel(IsolationLevel level) { 894 super.setIsolationLevel(level); 895 return this; 896 } 897 898 @Override 899 public Scan setPriority(int priority) { 900 super.setPriority(priority); 901 return this; 902 } 903 904 /** 905 * Enable collection of {@link ScanMetrics}. For advanced users. 906 * @param enabled Set to true to enable accumulating scan metrics 907 */ 908 public Scan setScanMetricsEnabled(final boolean enabled) { 909 setAttribute(Scan.SCAN_ATTRIBUTES_METRICS_ENABLE, Bytes.toBytes(Boolean.valueOf(enabled))); 910 return this; 911 } 912 913 /** Returns True if collection of scan metrics is enabled. For advanced users. */ 914 public boolean isScanMetricsEnabled() { 915 byte[] attr = getAttribute(Scan.SCAN_ATTRIBUTES_METRICS_ENABLE); 916 return attr == null ? false : Bytes.toBoolean(attr); 917 } 918 919 public Boolean isAsyncPrefetch() { 920 return asyncPrefetch; 921 } 922 923 /** 924 * @deprecated Since 3.0.0, will be removed in 4.0.0. After building sync client upon async 925 * client, the implementation is always 'async prefetch', so this flag is useless now. 926 */ 927 @Deprecated 928 public Scan setAsyncPrefetch(boolean asyncPrefetch) { 929 this.asyncPrefetch = asyncPrefetch; 930 return this; 931 } 932 933 /** Returns the limit of rows for this scan */ 934 public int getLimit() { 935 return limit; 936 } 937 938 /** 939 * Set the limit of rows for this scan. We will terminate the scan if the number of returned rows 940 * reaches this value. 941 * <p> 942 * This condition will be tested at last, after all other conditions such as stopRow, filter, etc. 943 * @param limit the limit of rows for this scan 944 */ 945 public Scan setLimit(int limit) { 946 this.limit = limit; 947 return this; 948 } 949 950 /** 951 * Call this when you only want to get one row. It will set {@code limit} to {@code 1}, and also 952 * set {@code readType} to {@link ReadType#PREAD}. 953 */ 954 public Scan setOneRowLimit() { 955 return setLimit(1).setReadType(ReadType.PREAD); 956 } 957 958 @InterfaceAudience.Public 959 public enum ReadType { 960 DEFAULT, 961 STREAM, 962 PREAD 963 } 964 965 /** Returns the read type for this scan */ 966 public ReadType getReadType() { 967 return readType; 968 } 969 970 /** 971 * Set the read type for this scan. 972 * <p> 973 * Notice that we may choose to use pread even if you specific {@link ReadType#STREAM} here. For 974 * example, we will always use pread if this is a get scan. 975 */ 976 public Scan setReadType(ReadType readType) { 977 this.readType = readType; 978 return this; 979 } 980 981 /** 982 * Get the mvcc read point used to open a scanner. 983 */ 984 long getMvccReadPoint() { 985 return mvccReadPoint; 986 } 987 988 /** 989 * Set the mvcc read point used to open a scanner. 990 */ 991 Scan setMvccReadPoint(long mvccReadPoint) { 992 this.mvccReadPoint = mvccReadPoint; 993 return this; 994 } 995 996 /** 997 * Set the mvcc read point to -1 which means do not use it. 998 */ 999 Scan resetMvccReadPoint() { 1000 return setMvccReadPoint(-1L); 1001 } 1002 1003 /** 1004 * When the server is slow or we scan a table with many deleted data or we use a sparse filter, 1005 * the server will response heartbeat to prevent timeout. However the scanner will return a Result 1006 * only when client can do it. So if there are many heartbeats, the blocking time on 1007 * ResultScanner#next() may be very long, which is not friendly to online services. Set this to 1008 * true then you can get a special Result whose #isCursor() returns true and is not contains any 1009 * real data. It only tells you where the server has scanned. You can call next to continue 1010 * scanning or open a new scanner with this row key as start row whenever you want. Users can get 1011 * a cursor when and only when there is a response from the server but we can not return a Result 1012 * to users, for example, this response is a heartbeat or there are partial cells but users do not 1013 * allow partial result. Now the cursor is in row level which means the special Result will only 1014 * contains a row key. {@link Result#isCursor()} {@link Result#getCursor()} {@link Cursor} 1015 */ 1016 public Scan setNeedCursorResult(boolean needCursorResult) { 1017 this.needCursorResult = needCursorResult; 1018 return this; 1019 } 1020 1021 public boolean isNeedCursorResult() { 1022 return needCursorResult; 1023 } 1024 1025 /** 1026 * Create a new Scan with a cursor. It only set the position information like start row key. The 1027 * others (like cfs, stop row, limit) should still be filled in by the user. 1028 * {@link Result#isCursor()} {@link Result#getCursor()} {@link Cursor} 1029 */ 1030 public static Scan createScanFromCursor(Cursor cursor) { 1031 return new Scan().withStartRow(cursor.getRow()); 1032 } 1033}