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 static private final String SCAN_ATTRIBUTES_METRICS_BY_REGION_ENABLE = 119 "scan.attributes.metrics.byregion.enable"; 120 121 /** 122 * -1 means no caching specified and the value of {@link HConstants#HBASE_CLIENT_SCANNER_CACHING} 123 * (default to {@link HConstants#DEFAULT_HBASE_CLIENT_SCANNER_CACHING}) will be used 124 */ 125 private int caching = -1; 126 private long maxResultSize = -1; 127 private boolean cacheBlocks = true; 128 private boolean reversed = false; 129 private TimeRange tr = TimeRange.allTime(); 130 private Map<byte[], NavigableSet<byte[]>> familyMap = 131 new TreeMap<byte[], NavigableSet<byte[]>>(Bytes.BYTES_COMPARATOR); 132 private Boolean asyncPrefetch = null; 133 134 /** 135 * Parameter name for client scanner sync/async prefetch toggle. When using async scanner, 136 * prefetching data from the server is done at the background. The parameter currently won't have 137 * any effect in the case that the user has set Scan#setSmall or Scan#setReversed 138 */ 139 public static final String HBASE_CLIENT_SCANNER_ASYNC_PREFETCH = 140 "hbase.client.scanner.async.prefetch"; 141 142 /** 143 * Default value of {@link #HBASE_CLIENT_SCANNER_ASYNC_PREFETCH}. 144 */ 145 public static final boolean DEFAULT_HBASE_CLIENT_SCANNER_ASYNC_PREFETCH = false; 146 147 /** 148 * The mvcc read point to use when open a scanner. Remember to clear it after switching regions as 149 * the mvcc is only valid within region scope. 150 */ 151 private long mvccReadPoint = -1L; 152 153 /** 154 * The number of rows we want for this scan. We will terminate the scan if the number of return 155 * rows reaches this value. 156 */ 157 private int limit = -1; 158 159 /** 160 * Control whether to use pread at server side. 161 */ 162 private ReadType readType = ReadType.DEFAULT; 163 164 private boolean needCursorResult = false; 165 166 /** 167 * Create a Scan operation across all rows. 168 */ 169 public Scan() { 170 } 171 172 /** 173 * Creates a new instance of this class while copying all values. 174 * @param scan The scan instance to copy from. 175 * @throws IOException When copying the values fails. 176 */ 177 public Scan(Scan scan) throws IOException { 178 startRow = scan.getStartRow(); 179 includeStartRow = scan.includeStartRow(); 180 stopRow = scan.getStopRow(); 181 includeStopRow = scan.includeStopRow(); 182 maxVersions = scan.getMaxVersions(); 183 batch = scan.getBatch(); 184 storeLimit = scan.getMaxResultsPerColumnFamily(); 185 storeOffset = scan.getRowOffsetPerColumnFamily(); 186 caching = scan.getCaching(); 187 maxResultSize = scan.getMaxResultSize(); 188 cacheBlocks = scan.getCacheBlocks(); 189 filter = scan.getFilter(); // clone? 190 loadColumnFamiliesOnDemand = scan.getLoadColumnFamiliesOnDemandValue(); 191 consistency = scan.getConsistency(); 192 this.setIsolationLevel(scan.getIsolationLevel()); 193 reversed = scan.isReversed(); 194 asyncPrefetch = scan.isAsyncPrefetch(); 195 allowPartialResults = scan.getAllowPartialResults(); 196 tr = scan.getTimeRange(); // TimeRange is immutable 197 Map<byte[], NavigableSet<byte[]>> fams = scan.getFamilyMap(); 198 for (Map.Entry<byte[], NavigableSet<byte[]>> entry : fams.entrySet()) { 199 byte[] fam = entry.getKey(); 200 NavigableSet<byte[]> cols = entry.getValue(); 201 if (cols != null && cols.size() > 0) { 202 for (byte[] col : cols) { 203 addColumn(fam, col); 204 } 205 } else { 206 addFamily(fam); 207 } 208 } 209 for (Map.Entry<String, byte[]> attr : scan.getAttributesMap().entrySet()) { 210 setAttribute(attr.getKey(), attr.getValue()); 211 } 212 for (Map.Entry<byte[], TimeRange> entry : scan.getColumnFamilyTimeRange().entrySet()) { 213 TimeRange tr = entry.getValue(); 214 setColumnFamilyTimeRange(entry.getKey(), tr.getMin(), tr.getMax()); 215 } 216 this.mvccReadPoint = scan.getMvccReadPoint(); 217 this.limit = scan.getLimit(); 218 this.needCursorResult = scan.isNeedCursorResult(); 219 setPriority(scan.getPriority()); 220 readType = scan.getReadType(); 221 super.setReplicaId(scan.getReplicaId()); 222 super.setQueryMetricsEnabled(scan.isQueryMetricsEnabled()); 223 } 224 225 /** 226 * Builds a scan object with the same specs as get. 227 * @param get get to model scan after 228 */ 229 public Scan(Get get) { 230 this.startRow = get.getRow(); 231 this.includeStartRow = true; 232 this.stopRow = get.getRow(); 233 this.includeStopRow = true; 234 this.filter = get.getFilter(); 235 this.cacheBlocks = get.getCacheBlocks(); 236 this.maxVersions = get.getMaxVersions(); 237 this.storeLimit = get.getMaxResultsPerColumnFamily(); 238 this.storeOffset = get.getRowOffsetPerColumnFamily(); 239 this.tr = get.getTimeRange(); 240 this.familyMap = get.getFamilyMap(); 241 this.asyncPrefetch = false; 242 this.consistency = get.getConsistency(); 243 this.setIsolationLevel(get.getIsolationLevel()); 244 this.loadColumnFamiliesOnDemand = get.getLoadColumnFamiliesOnDemandValue(); 245 for (Map.Entry<String, byte[]> attr : get.getAttributesMap().entrySet()) { 246 setAttribute(attr.getKey(), attr.getValue()); 247 } 248 for (Map.Entry<byte[], TimeRange> entry : get.getColumnFamilyTimeRange().entrySet()) { 249 TimeRange tr = entry.getValue(); 250 setColumnFamilyTimeRange(entry.getKey(), tr.getMin(), tr.getMax()); 251 } 252 this.mvccReadPoint = -1L; 253 setPriority(get.getPriority()); 254 super.setReplicaId(get.getReplicaId()); 255 super.setQueryMetricsEnabled(get.isQueryMetricsEnabled()); 256 } 257 258 public boolean isGetScan() { 259 return includeStartRow && includeStopRow 260 && ClientUtil.areScanStartRowAndStopRowEqual(this.startRow, this.stopRow); 261 } 262 263 /** 264 * Get all columns from the specified family. 265 * <p> 266 * Overrides previous calls to addColumn for this family. 267 * @param family family name 268 */ 269 public Scan addFamily(byte[] family) { 270 familyMap.remove(family); 271 familyMap.put(family, null); 272 return this; 273 } 274 275 /** 276 * Get the column from the specified family with the specified qualifier. 277 * <p> 278 * Overrides previous calls to addFamily for this family. 279 * @param family family name 280 * @param qualifier column qualifier 281 */ 282 public Scan addColumn(byte[] family, byte[] qualifier) { 283 NavigableSet<byte[]> set = familyMap.get(family); 284 if (set == null) { 285 set = new TreeSet<>(Bytes.BYTES_COMPARATOR); 286 familyMap.put(family, set); 287 } 288 if (qualifier == null) { 289 qualifier = HConstants.EMPTY_BYTE_ARRAY; 290 } 291 set.add(qualifier); 292 return this; 293 } 294 295 /** 296 * Get versions of columns only within the specified timestamp range, [minStamp, maxStamp). Note, 297 * default maximum versions to return is 1. If your time range spans more than one version and you 298 * want all versions returned, up the number of versions beyond the default. 299 * @param minStamp minimum timestamp value, inclusive 300 * @param maxStamp maximum timestamp value, exclusive 301 * @see #readAllVersions() 302 * @see #readVersions(int) 303 */ 304 public Scan setTimeRange(long minStamp, long maxStamp) throws IOException { 305 tr = TimeRange.between(minStamp, maxStamp); 306 return this; 307 } 308 309 /** 310 * Get versions of columns with the specified timestamp. Note, default maximum versions to return 311 * is 1. If your time range spans more than one version and you want all versions returned, up the 312 * number of versions beyond the defaut. 313 * @param timestamp version timestamp 314 * @see #readAllVersions() 315 * @see #readVersions(int) 316 */ 317 public Scan setTimestamp(long timestamp) { 318 try { 319 tr = TimeRange.at(timestamp); 320 } catch (Exception e) { 321 // This should never happen, unless integer overflow or something extremely wrong... 322 LOG.error("TimeRange failed, likely caused by integer overflow. ", e); 323 throw e; 324 } 325 326 return this; 327 } 328 329 @Override 330 public Scan setColumnFamilyTimeRange(byte[] cf, long minStamp, long maxStamp) { 331 super.setColumnFamilyTimeRange(cf, minStamp, maxStamp); 332 return this; 333 } 334 335 /** 336 * Set the start row of the scan. 337 * <p> 338 * If the specified row does not exist, the Scanner will start from the next closest row after the 339 * specified row. 340 * <p> 341 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 342 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 343 * unexpected or even undefined. 344 * </p> 345 * @param startRow row to start scanner at or after 346 * @throws IllegalArgumentException if startRow does not meet criteria for a row key (when length 347 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 348 */ 349 public Scan withStartRow(byte[] startRow) { 350 return withStartRow(startRow, true); 351 } 352 353 /** 354 * Set the start row of the scan. 355 * <p> 356 * If the specified row does not exist, or the {@code inclusive} is {@code false}, the Scanner 357 * will start from the next closest row after the specified row. 358 * <p> 359 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 360 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 361 * unexpected or even undefined. 362 * </p> 363 * @param startRow row to start scanner at or after 364 * @param inclusive whether we should include the start row when scan 365 * @throws IllegalArgumentException if startRow does not meet criteria for a row key (when length 366 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 367 */ 368 public Scan withStartRow(byte[] startRow, boolean inclusive) { 369 if (Bytes.len(startRow) > HConstants.MAX_ROW_LENGTH) { 370 throw new IllegalArgumentException("startRow's length must be less than or equal to " 371 + HConstants.MAX_ROW_LENGTH + " to meet the criteria" + " for a row key."); 372 } 373 this.startRow = startRow; 374 this.includeStartRow = inclusive; 375 return this; 376 } 377 378 /** 379 * Set the stop row of the scan. 380 * <p> 381 * The scan will include rows that are lexicographically less than the provided stopRow. 382 * <p> 383 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 384 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 385 * unexpected or even undefined. 386 * </p> 387 * @param stopRow row to end at (exclusive) 388 * @throws IllegalArgumentException if stopRow does not meet criteria for a row key (when length 389 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 390 */ 391 public Scan withStopRow(byte[] stopRow) { 392 return withStopRow(stopRow, false); 393 } 394 395 /** 396 * Set the stop row of the scan. 397 * <p> 398 * The scan will include rows that are lexicographically less than (or equal to if 399 * {@code inclusive} is {@code true}) the provided stopRow. 400 * <p> 401 * <b>Note:</b> <strong>Do NOT use this in combination with {@link #setRowPrefixFilter(byte[])} or 402 * {@link #setStartStopRowForPrefixScan(byte[])}.</strong> Doing so will make the scan result 403 * unexpected or even undefined. 404 * </p> 405 * @param stopRow row to end at 406 * @param inclusive whether we should include the stop row when scan 407 * @throws IllegalArgumentException if stopRow does not meet criteria for a row key (when length 408 * exceeds {@link HConstants#MAX_ROW_LENGTH}) 409 */ 410 public Scan withStopRow(byte[] stopRow, boolean inclusive) { 411 if (Bytes.len(stopRow) > HConstants.MAX_ROW_LENGTH) { 412 throw new IllegalArgumentException("stopRow's length must be less than or equal to " 413 + HConstants.MAX_ROW_LENGTH + " to meet the criteria" + " for a row key."); 414 } 415 this.stopRow = stopRow; 416 this.includeStopRow = inclusive; 417 return this; 418 } 419 420 /** 421 * <p> 422 * Set a filter (using stopRow and startRow) so the result set only contains rows where the rowKey 423 * starts with the specified prefix. 424 * </p> 425 * <p> 426 * This is a utility method that converts the desired rowPrefix into the appropriate values for 427 * the startRow and stopRow to achieve the desired result. 428 * </p> 429 * <p> 430 * This can safely be used in combination with setFilter. 431 * </p> 432 * <p> 433 * <strong>This CANNOT be used in combination with withStartRow and/or withStopRow.</strong> Such 434 * a combination will yield unexpected and even undefined results. 435 * </p> 436 * @param rowPrefix the prefix all rows must start with. (Set <i>null</i> to remove the filter.) 437 * @deprecated since 2.5.0, will be removed in 4.0.0. The name of this method is considered to be 438 * confusing as it does not use a {@link Filter} but uses setting the startRow and 439 * stopRow instead. Use {@link #setStartStopRowForPrefixScan(byte[])} instead. 440 */ 441 @Deprecated 442 public Scan setRowPrefixFilter(byte[] rowPrefix) { 443 return setStartStopRowForPrefixScan(rowPrefix); 444 } 445 446 /** 447 * <p> 448 * Set a filter (using stopRow and startRow) so the result set only contains rows where the rowKey 449 * starts with the specified prefix. 450 * </p> 451 * <p> 452 * This is a utility method that converts the desired rowPrefix into the appropriate values for 453 * the startRow and stopRow to achieve the desired result. 454 * </p> 455 * <p> 456 * This can safely be used in combination with setFilter. 457 * </p> 458 * <p> 459 * <strong>This CANNOT be used in combination with withStartRow and/or withStopRow.</strong> Such 460 * a combination will yield unexpected and even undefined results. 461 * </p> 462 * @param rowPrefix the prefix all rows must start with. (Set <i>null</i> to remove the filter.) 463 */ 464 public Scan setStartStopRowForPrefixScan(byte[] rowPrefix) { 465 if (rowPrefix == null) { 466 withStartRow(HConstants.EMPTY_START_ROW); 467 withStopRow(HConstants.EMPTY_END_ROW); 468 } else { 469 this.withStartRow(rowPrefix); 470 this.withStopRow(ClientUtil.calculateTheClosestNextRowKeyForPrefix(rowPrefix)); 471 } 472 return this; 473 } 474 475 /** 476 * Get all available versions. 477 */ 478 public Scan readAllVersions() { 479 this.maxVersions = Integer.MAX_VALUE; 480 return this; 481 } 482 483 /** 484 * Get up to the specified number of versions of each column. 485 * @param versions specified number of versions for each column 486 */ 487 public Scan readVersions(int versions) { 488 this.maxVersions = versions; 489 return this; 490 } 491 492 /** 493 * Set the maximum number of cells to return for each call to next(). Callers should be aware that 494 * this is not equivalent to calling {@link #setAllowPartialResults(boolean)}. If you don't allow 495 * partial results, the number of cells in each Result must equal to your batch setting unless it 496 * is the last Result for current row. So this method is helpful in paging queries. If you just 497 * want to prevent OOM at client, use setAllowPartialResults(true) is better. 498 * @param batch the maximum number of values 499 * @see Result#mayHaveMoreCellsInRow() 500 */ 501 public Scan setBatch(int batch) { 502 if (this.hasFilter() && this.filter.hasFilterRow()) { 503 throw new IncompatibleFilterException( 504 "Cannot set batch on a scan using a filter" + " that returns true for filter.hasFilterRow"); 505 } 506 this.batch = batch; 507 return this; 508 } 509 510 /** 511 * Set the maximum number of values to return per row per Column Family 512 * @param limit the maximum number of values returned / row / CF 513 */ 514 public Scan setMaxResultsPerColumnFamily(int limit) { 515 this.storeLimit = limit; 516 return this; 517 } 518 519 /** 520 * Set offset for the row per Column Family. 521 * @param offset is the number of kvs that will be skipped. 522 */ 523 public Scan setRowOffsetPerColumnFamily(int offset) { 524 this.storeOffset = offset; 525 return this; 526 } 527 528 /** 529 * Set the number of rows for caching that will be passed to scanners. If not set, the 530 * Configuration setting {@link HConstants#HBASE_CLIENT_SCANNER_CACHING} will apply. Higher 531 * caching values will enable faster scanners but will use more memory. 532 * @param caching the number of rows for caching 533 */ 534 public Scan setCaching(int caching) { 535 this.caching = caching; 536 return this; 537 } 538 539 /** Returns the maximum result size in bytes. See {@link #setMaxResultSize(long)} */ 540 public long getMaxResultSize() { 541 return maxResultSize; 542 } 543 544 /** 545 * Set the maximum result size. The default is -1; this means that no specific maximum result size 546 * will be set for this scan, and the global configured value will be used instead. (Defaults to 547 * unlimited). 548 * @param maxResultSize The maximum result size in bytes. 549 */ 550 public Scan setMaxResultSize(long maxResultSize) { 551 this.maxResultSize = maxResultSize; 552 return this; 553 } 554 555 @Override 556 public Scan setFilter(Filter filter) { 557 if (filter != null && filter.hasFilterRow() && this.batch > 0) { 558 throw new IncompatibleFilterException( 559 "Cannot set a filter that returns true for filter.hasFilterRow on a scan with batch set"); 560 } 561 super.setFilter(filter); 562 return this; 563 } 564 565 /** 566 * Setting the familyMap 567 * @param familyMap map of family to qualifier 568 */ 569 public Scan setFamilyMap(Map<byte[], NavigableSet<byte[]>> familyMap) { 570 this.familyMap = familyMap; 571 return this; 572 } 573 574 /** 575 * Getting the familyMap 576 */ 577 public Map<byte[], NavigableSet<byte[]>> getFamilyMap() { 578 return this.familyMap; 579 } 580 581 /** Returns the number of families in familyMap */ 582 public int numFamilies() { 583 if (hasFamilies()) { 584 return this.familyMap.size(); 585 } 586 return 0; 587 } 588 589 /** Returns true if familyMap is non empty, false otherwise */ 590 public boolean hasFamilies() { 591 return !this.familyMap.isEmpty(); 592 } 593 594 /** Returns the keys of the familyMap */ 595 public byte[][] getFamilies() { 596 if (hasFamilies()) { 597 return this.familyMap.keySet().toArray(new byte[0][0]); 598 } 599 return null; 600 } 601 602 /** Returns the startrow */ 603 public byte[] getStartRow() { 604 return this.startRow; 605 } 606 607 /** Returns if we should include start row when scan */ 608 public boolean includeStartRow() { 609 return includeStartRow; 610 } 611 612 /** Returns the stoprow */ 613 public byte[] getStopRow() { 614 return this.stopRow; 615 } 616 617 /** Returns if we should include stop row when scan */ 618 public boolean includeStopRow() { 619 return includeStopRow; 620 } 621 622 /** Returns the max number of versions to fetch */ 623 public int getMaxVersions() { 624 return this.maxVersions; 625 } 626 627 /** Returns maximum number of values to return for a single call to next() */ 628 public int getBatch() { 629 return this.batch; 630 } 631 632 /** Returns maximum number of values to return per row per CF */ 633 public int getMaxResultsPerColumnFamily() { 634 return this.storeLimit; 635 } 636 637 /** 638 * Method for retrieving the scan's offset per row per column family (#kvs to be skipped) 639 * @return row offset 640 */ 641 public int getRowOffsetPerColumnFamily() { 642 return this.storeOffset; 643 } 644 645 /** Returns caching the number of rows fetched when calling next on a scanner */ 646 public int getCaching() { 647 return this.caching; 648 } 649 650 /** Returns TimeRange */ 651 public TimeRange getTimeRange() { 652 return this.tr; 653 } 654 655 /** Returns RowFilter */ 656 @Override 657 public Filter getFilter() { 658 return filter; 659 } 660 661 /** Returns true is a filter has been specified, false if not */ 662 public boolean hasFilter() { 663 return filter != null; 664 } 665 666 /** 667 * Set whether blocks should be cached for this Scan. 668 * <p> 669 * This is true by default. When true, default settings of the table and family are used (this 670 * will never override caching blocks if the block cache is disabled for that family or entirely). 671 * @param cacheBlocks if false, default settings are overridden and blocks will not be cached 672 */ 673 public Scan setCacheBlocks(boolean cacheBlocks) { 674 this.cacheBlocks = cacheBlocks; 675 return this; 676 } 677 678 /** 679 * Get whether blocks should be cached for this Scan. 680 * @return true if default caching should be used, false if blocks should not be cached 681 */ 682 public boolean getCacheBlocks() { 683 return cacheBlocks; 684 } 685 686 /** 687 * Set whether this scan is a reversed one 688 * <p> 689 * This is false by default which means forward(normal) scan. 690 * @param reversed if true, scan will be backward order 691 */ 692 public Scan setReversed(boolean reversed) { 693 this.reversed = reversed; 694 return this; 695 } 696 697 /** 698 * Get whether this scan is a reversed one. 699 * @return true if backward scan, false if forward(default) scan 700 */ 701 public boolean isReversed() { 702 return reversed; 703 } 704 705 /** 706 * Setting whether the caller wants to see the partial results when server returns 707 * less-than-expected cells. It is helpful while scanning a huge row to prevent OOM at client. By 708 * default this value is false and the complete results will be assembled client side before being 709 * delivered to the caller. 710 * @see Result#mayHaveMoreCellsInRow() 711 * @see #setBatch(int) 712 */ 713 public Scan setAllowPartialResults(final boolean allowPartialResults) { 714 this.allowPartialResults = allowPartialResults; 715 return this; 716 } 717 718 /** 719 * Returns true when the constructor of this scan understands that the results they will see may 720 * only represent a partial portion of a row. The entire row would be retrieved by subsequent 721 * calls to {@link ResultScanner#next()} 722 */ 723 public boolean getAllowPartialResults() { 724 return allowPartialResults; 725 } 726 727 @Override 728 public Scan setLoadColumnFamiliesOnDemand(boolean value) { 729 super.setLoadColumnFamiliesOnDemand(value); 730 return this; 731 } 732 733 /** 734 * Compile the table and column family (i.e. schema) information into a String. Useful for parsing 735 * and aggregation by debugging, logging, and administration tools. 736 */ 737 @Override 738 public Map<String, Object> getFingerprint() { 739 Map<String, Object> map = new HashMap<>(); 740 List<String> families = new ArrayList<>(); 741 if (this.familyMap.isEmpty()) { 742 map.put("families", "ALL"); 743 return map; 744 } else { 745 map.put("families", families); 746 } 747 for (Map.Entry<byte[], NavigableSet<byte[]>> entry : this.familyMap.entrySet()) { 748 families.add(Bytes.toStringBinary(entry.getKey())); 749 } 750 return map; 751 } 752 753 /** 754 * Compile the details beyond the scope of getFingerprint (row, columns, timestamps, etc.) into a 755 * Map along with the fingerprinted information. Useful for debugging, logging, and administration 756 * tools. 757 * @param maxCols a limit on the number of columns output prior to truncation 758 */ 759 @Override 760 public Map<String, Object> toMap(int maxCols) { 761 // start with the fingerprint map and build on top of it 762 Map<String, Object> map = getFingerprint(); 763 // map from families to column list replaces fingerprint's list of families 764 Map<String, List<String>> familyColumns = new HashMap<>(); 765 map.put("families", familyColumns); 766 // add scalar information first 767 map.put("startRow", Bytes.toStringBinary(this.startRow)); 768 map.put("stopRow", Bytes.toStringBinary(this.stopRow)); 769 map.put("maxVersions", this.maxVersions); 770 map.put("batch", this.batch); 771 map.put("caching", this.caching); 772 map.put("maxResultSize", this.maxResultSize); 773 map.put("cacheBlocks", this.cacheBlocks); 774 map.put("loadColumnFamiliesOnDemand", this.loadColumnFamiliesOnDemand); 775 List<Long> timeRange = new ArrayList<>(2); 776 timeRange.add(this.tr.getMin()); 777 timeRange.add(this.tr.getMax()); 778 map.put("timeRange", timeRange); 779 int colCount = 0; 780 // iterate through affected families and list out up to maxCols columns 781 for (Map.Entry<byte[], NavigableSet<byte[]>> entry : this.familyMap.entrySet()) { 782 List<String> columns = new ArrayList<>(); 783 familyColumns.put(Bytes.toStringBinary(entry.getKey()), columns); 784 if (entry.getValue() == null) { 785 colCount++; 786 --maxCols; 787 columns.add("ALL"); 788 } else { 789 colCount += entry.getValue().size(); 790 if (maxCols <= 0) { 791 continue; 792 } 793 for (byte[] column : entry.getValue()) { 794 if (--maxCols <= 0) { 795 continue; 796 } 797 columns.add(Bytes.toStringBinary(column)); 798 } 799 } 800 } 801 map.put("totalColumns", colCount); 802 if (this.filter != null) { 803 map.put("filter", this.filter.toString()); 804 } 805 // add the id if set 806 if (getId() != null) { 807 map.put("id", getId()); 808 } 809 map.put("includeStartRow", includeStartRow); 810 map.put("includeStopRow", includeStopRow); 811 map.put("allowPartialResults", allowPartialResults); 812 map.put("storeLimit", storeLimit); 813 map.put("storeOffset", storeOffset); 814 map.put("reversed", reversed); 815 if (null != asyncPrefetch) { 816 map.put("asyncPrefetch", asyncPrefetch); 817 } 818 map.put("mvccReadPoint", mvccReadPoint); 819 map.put("limit", limit); 820 map.put("readType", readType); 821 map.put("needCursorResult", needCursorResult); 822 map.put("targetReplicaId", targetReplicaId); 823 map.put("consistency", consistency); 824 if (!colFamTimeRangeMap.isEmpty()) { 825 Map<String, List<Long>> colFamTimeRangeMapStr = colFamTimeRangeMap.entrySet().stream() 826 .collect(Collectors.toMap((e) -> Bytes.toStringBinary(e.getKey()), e -> { 827 TimeRange value = e.getValue(); 828 List<Long> rangeList = new ArrayList<>(); 829 rangeList.add(value.getMin()); 830 rangeList.add(value.getMax()); 831 return rangeList; 832 })); 833 834 map.put("colFamTimeRangeMap", colFamTimeRangeMapStr); 835 } 836 map.put("priority", getPriority()); 837 map.put("queryMetricsEnabled", queryMetricsEnabled); 838 return map; 839 } 840 841 /** 842 * Enable/disable "raw" mode for this scan. If "raw" is enabled the scan will return all delete 843 * marker and deleted rows that have not been collected, yet. This is mostly useful for Scan on 844 * column families that have KEEP_DELETED_ROWS enabled. It is an error to specify any column when 845 * "raw" is set. 846 * @param raw True/False to enable/disable "raw" mode. 847 */ 848 public Scan setRaw(boolean raw) { 849 setAttribute(RAW_ATTR, Bytes.toBytes(raw)); 850 return this; 851 } 852 853 /** Returns True if this Scan is in "raw" mode. */ 854 public boolean isRaw() { 855 byte[] attr = getAttribute(RAW_ATTR); 856 return attr == null ? false : Bytes.toBoolean(attr); 857 } 858 859 @Override 860 public Scan setAttribute(String name, byte[] value) { 861 super.setAttribute(name, value); 862 return this; 863 } 864 865 @Override 866 public Scan setId(String id) { 867 super.setId(id); 868 return this; 869 } 870 871 @Override 872 public Scan setAuthorizations(Authorizations authorizations) { 873 super.setAuthorizations(authorizations); 874 return this; 875 } 876 877 @Override 878 public Scan setACL(Map<String, Permission> perms) { 879 super.setACL(perms); 880 return this; 881 } 882 883 @Override 884 public Scan setACL(String user, Permission perms) { 885 super.setACL(user, perms); 886 return this; 887 } 888 889 @Override 890 public Scan setConsistency(Consistency consistency) { 891 super.setConsistency(consistency); 892 return this; 893 } 894 895 @Override 896 public Scan setReplicaId(int Id) { 897 super.setReplicaId(Id); 898 return this; 899 } 900 901 @Override 902 public Scan setIsolationLevel(IsolationLevel level) { 903 super.setIsolationLevel(level); 904 return this; 905 } 906 907 @Override 908 public Scan setPriority(int priority) { 909 super.setPriority(priority); 910 return this; 911 } 912 913 /** 914 * Enable collection of {@link ScanMetrics}. For advanced users. While disabling scan metrics, 915 * will also disable region level scan metrics. 916 * @param enabled Set to true to enable accumulating scan metrics 917 */ 918 public Scan setScanMetricsEnabled(final boolean enabled) { 919 setAttribute(Scan.SCAN_ATTRIBUTES_METRICS_ENABLE, Bytes.toBytes(Boolean.valueOf(enabled))); 920 if (!enabled) { 921 setEnableScanMetricsByRegion(false); 922 } 923 return this; 924 } 925 926 /** Returns True if collection of scan metrics is enabled. For advanced users. */ 927 public boolean isScanMetricsEnabled() { 928 byte[] attr = getAttribute(Scan.SCAN_ATTRIBUTES_METRICS_ENABLE); 929 return attr == null ? false : Bytes.toBoolean(attr); 930 } 931 932 public Boolean isAsyncPrefetch() { 933 return asyncPrefetch; 934 } 935 936 /** 937 * @deprecated Since 3.0.0, will be removed in 4.0.0. After building sync client upon async 938 * client, the implementation is always 'async prefetch', so this flag is useless now. 939 */ 940 @Deprecated 941 public Scan setAsyncPrefetch(boolean asyncPrefetch) { 942 this.asyncPrefetch = asyncPrefetch; 943 return this; 944 } 945 946 /** Returns the limit of rows for this scan */ 947 public int getLimit() { 948 return limit; 949 } 950 951 /** 952 * Set the limit of rows for this scan. We will terminate the scan if the number of returned rows 953 * reaches this value. 954 * <p> 955 * This condition will be tested at last, after all other conditions such as stopRow, filter, etc. 956 * @param limit the limit of rows for this scan 957 */ 958 public Scan setLimit(int limit) { 959 this.limit = limit; 960 return this; 961 } 962 963 /** 964 * Call this when you only want to get one row. It will set {@code limit} to {@code 1}, and also 965 * set {@code readType} to {@link ReadType#PREAD}. 966 */ 967 public Scan setOneRowLimit() { 968 return setLimit(1).setReadType(ReadType.PREAD); 969 } 970 971 @InterfaceAudience.Public 972 public enum ReadType { 973 DEFAULT, 974 STREAM, 975 PREAD 976 } 977 978 /** Returns the read type for this scan */ 979 public ReadType getReadType() { 980 return readType; 981 } 982 983 /** 984 * Set the read type for this scan. 985 * <p> 986 * Notice that we may choose to use pread even if you specific {@link ReadType#STREAM} here. For 987 * example, we will always use pread if this is a get scan. 988 */ 989 public Scan setReadType(ReadType readType) { 990 this.readType = readType; 991 return this; 992 } 993 994 /** 995 * Get the mvcc read point used to open a scanner. 996 */ 997 long getMvccReadPoint() { 998 return mvccReadPoint; 999 } 1000 1001 /** 1002 * Set the mvcc read point used to open a scanner. 1003 */ 1004 Scan setMvccReadPoint(long mvccReadPoint) { 1005 this.mvccReadPoint = mvccReadPoint; 1006 return this; 1007 } 1008 1009 /** 1010 * Set the mvcc read point to -1 which means do not use it. 1011 */ 1012 Scan resetMvccReadPoint() { 1013 return setMvccReadPoint(-1L); 1014 } 1015 1016 /** 1017 * When the server is slow or we scan a table with many deleted data or we use a sparse filter, 1018 * the server will response heartbeat to prevent timeout. However the scanner will return a Result 1019 * only when client can do it. So if there are many heartbeats, the blocking time on 1020 * ResultScanner#next() may be very long, which is not friendly to online services. Set this to 1021 * true then you can get a special Result whose #isCursor() returns true and is not contains any 1022 * real data. It only tells you where the server has scanned. You can call next to continue 1023 * scanning or open a new scanner with this row key as start row whenever you want. Users can get 1024 * a cursor when and only when there is a response from the server but we can not return a Result 1025 * to users, for example, this response is a heartbeat or there are partial cells but users do not 1026 * allow partial result. Now the cursor is in row level which means the special Result will only 1027 * contains a row key. {@link Result#isCursor()} {@link Result#getCursor()} {@link Cursor} 1028 */ 1029 public Scan setNeedCursorResult(boolean needCursorResult) { 1030 this.needCursorResult = needCursorResult; 1031 return this; 1032 } 1033 1034 public boolean isNeedCursorResult() { 1035 return needCursorResult; 1036 } 1037 1038 /** 1039 * Create a new Scan with a cursor. It only set the position information like start row key. The 1040 * others (like cfs, stop row, limit) should still be filled in by the user. 1041 * {@link Result#isCursor()} {@link Result#getCursor()} {@link Cursor} 1042 */ 1043 public static Scan createScanFromCursor(Cursor cursor) { 1044 return new Scan().withStartRow(cursor.getRow()); 1045 } 1046 1047 /** 1048 * Enables region level scan metrics. If scan metrics are disabled then first enables scan metrics 1049 * followed by region level scan metrics. 1050 * @param enable Set to true to enable region level scan metrics. 1051 */ 1052 public Scan setEnableScanMetricsByRegion(final boolean enable) { 1053 if (enable) { 1054 setScanMetricsEnabled(true); 1055 } 1056 setAttribute(Scan.SCAN_ATTRIBUTES_METRICS_BY_REGION_ENABLE, Bytes.toBytes(enable)); 1057 return this; 1058 } 1059 1060 public boolean isScanMetricsByRegionEnabled() { 1061 byte[] attr = getAttribute(Scan.SCAN_ATTRIBUTES_METRICS_BY_REGION_ENABLE); 1062 return attr != null && Bytes.toBoolean(attr); 1063 } 1064}