View Javadoc

1   /*
2    *
3    * Licensed to the Apache Software Foundation (ASF) under one
4    * or more contributor license agreements.  See the NOTICE file
5    * distributed with this work for additional information
6    * regarding copyright ownership.  The ASF licenses this file
7    * to you under the Apache License, Version 2.0 (the
8    * "License"); you may not use this file except in compliance
9    * with the License.  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  package org.apache.hadoop.hbase.client;
21  
22  import java.io.IOException;
23  import java.util.ArrayList;
24  import java.util.Arrays;
25  import java.util.HashMap;
26  import java.util.List;
27  import java.util.Map;
28  import java.util.NavigableSet;
29  import java.util.TreeMap;
30  import java.util.TreeSet;
31  
32  import org.apache.commons.logging.Log;
33  import org.apache.commons.logging.LogFactory;
34  import org.apache.hadoop.hbase.HConstants;
35  import org.apache.hadoop.hbase.classification.InterfaceAudience;
36  import org.apache.hadoop.hbase.classification.InterfaceStability;
37  import org.apache.hadoop.hbase.filter.Filter;
38  import org.apache.hadoop.hbase.filter.IncompatibleFilterException;
39  import org.apache.hadoop.hbase.io.TimeRange;
40  import org.apache.hadoop.hbase.security.access.Permission;
41  import org.apache.hadoop.hbase.security.visibility.Authorizations;
42  import org.apache.hadoop.hbase.util.Bytes;
43  
44  /**
45   * Used to perform Scan operations.
46   * <p>
47   * All operations are identical to {@link Get} with the exception of
48   * instantiation.  Rather than specifying a single row, an optional startRow
49   * and stopRow may be defined.  If rows are not specified, the Scanner will
50   * iterate over all rows.
51   * <p>
52   * To scan everything for each row, instantiate a Scan object.
53   * <p>
54   * To modify scanner caching for just this scan, use {@link #setCaching(int) setCaching}.
55   * If caching is NOT set, we will use the caching value of the hosting {@link HTable}.  See
56   * {@link HTable#setScannerCaching(int)}. In addition to row caching, it is possible to specify a
57   * maximum result size, using {@link #setMaxResultSize(long)}. When both are used,
58   * single server requests are limited by either number of rows or maximum result size, whichever
59   * limit comes first.
60   * <p>
61   * To further define the scope of what to get when scanning, perform additional
62   * methods as outlined below.
63   * <p>
64   * To get all columns from specific families, execute {@link #addFamily(byte[]) addFamily}
65   * for each family to retrieve.
66   * <p>
67   * To get specific columns, execute {@link #addColumn(byte[], byte[]) addColumn}
68   * for each column to retrieve.
69   * <p>
70   * To only retrieve columns within a specific range of version timestamps,
71   * execute {@link #setTimeRange(long, long) setTimeRange}.
72   * <p>
73   * To only retrieve columns with a specific timestamp, execute
74   * {@link #setTimeStamp(long) setTimestamp}.
75   * <p>
76   * To limit the number of versions of each column to be returned, execute
77   * {@link #setMaxVersions(int) setMaxVersions}.
78   * <p>
79   * To limit the maximum number of values returned for each call to next(),
80   * execute {@link #setBatch(int) setBatch}.
81   * <p>
82   * To add a filter, execute {@link #setFilter(org.apache.hadoop.hbase.filter.Filter) setFilter}.
83   * <p>
84   * Expert: To explicitly disable server-side block caching for this scan,
85   * execute {@link #setCacheBlocks(boolean)}.
86   */
87  @InterfaceAudience.Public
88  @InterfaceStability.Stable
89  public class Scan extends Query {
90    private static final Log LOG = LogFactory.getLog(Scan.class);
91  
92    private static final String RAW_ATTR = "_raw_";
93  
94    /**
95     * EXPERT ONLY.
96     * An integer (not long) indicating to the scanner logic how many times we attempt to retrieve the
97     * next KV before we schedule a reseek.
98     * The right value depends on the size of the average KV. A reseek is more efficient when
99     * it can skip 5-10 KVs or 512B-1KB, or when the next KV is likely found in another HFile block.
100    * Setting this only has any effect when columns were added with
101    * {@link #addColumn(byte[], byte[])}
102    * <pre>{@code
103    * Scan s = new Scan(...);
104    * s.addColumn(...);
105    * s.setAttribute(Scan.HINT_LOOKAHEAD, Bytes.toBytes(2));
106    * }</pre>
107    * Default is 0 (always reseek).
108    */
109   public static final String HINT_LOOKAHEAD = "_look_ahead_";
110 
111   private byte [] startRow = HConstants.EMPTY_START_ROW;
112   private byte [] stopRow  = HConstants.EMPTY_END_ROW;
113   private int maxVersions = 1;
114   private int batch = -1;
115 
116   private int storeLimit = -1;
117   private int storeOffset = 0;
118   private boolean getScan;
119 
120   // If application wants to collect scan metrics, it needs to
121   // call scan.setAttribute(SCAN_ATTRIBUTES_ENABLE, Bytes.toBytes(Boolean.TRUE))
122   static public final String SCAN_ATTRIBUTES_METRICS_ENABLE = "scan.attributes.metrics.enable";
123   static public final String SCAN_ATTRIBUTES_METRICS_DATA = "scan.attributes.metrics.data";
124 
125   // If an application wants to use multiple scans over different tables each scan must
126   // define this attribute with the appropriate table name by calling
127   // scan.setAttribute(Scan.SCAN_ATTRIBUTES_TABLE_NAME, Bytes.toBytes(tableName))
128   static public final String SCAN_ATTRIBUTES_TABLE_NAME = "scan.attributes.table.name";
129 
130   /*
131    * -1 means no caching
132    */
133   private int caching = -1;
134   private long maxResultSize = -1;
135   private boolean cacheBlocks = true;
136   private boolean reversed = false;
137   private TimeRange tr = new TimeRange();
138   private Map<byte [], NavigableSet<byte []>> familyMap =
139     new TreeMap<byte [], NavigableSet<byte []>>(Bytes.BYTES_COMPARATOR);
140   private Boolean loadColumnFamiliesOnDemand = null;
141 
142   /**
143    * Set it true for small scan to get better performance
144    *
145    * Small scan should use pread and big scan can use seek + read
146    *
147    * seek + read is fast but can cause two problem (1) resource contention (2)
148    * cause too much network io
149    *
150    * [89-fb] Using pread for non-compaction read request
151    * https://issues.apache.org/jira/browse/HBASE-7266
152    *
153    * On the other hand, if setting it true, we would do
154    * openScanner,next,closeScanner in one RPC call. It means the better
155    * performance for small scan. [HBASE-9488].
156    *
157    * Generally, if the scan range is within one data block(64KB), it could be
158    * considered as a small scan.
159    */
160   private boolean small = false;
161 
162   /**
163    * Create a Scan operation across all rows.
164    */
165   public Scan() {}
166 
167   public Scan(byte [] startRow, Filter filter) {
168     this(startRow);
169     this.filter = filter;
170   }
171 
172   /**
173    * Create a Scan operation starting at the specified row.
174    * <p>
175    * If the specified row does not exist, the Scanner will start from the
176    * next closest row after the specified row.
177    * @param startRow row to start scanner at or after
178    */
179   public Scan(byte [] startRow) {
180     this.startRow = startRow;
181   }
182 
183   /**
184    * Create a Scan operation for the range of rows specified.
185    * @param startRow row to start scanner at or after (inclusive)
186    * @param stopRow row to stop scanner before (exclusive)
187    */
188   public Scan(byte [] startRow, byte [] stopRow) {
189     this.startRow = startRow;
190     this.stopRow = stopRow;
191     //if the startRow and stopRow both are empty, it is not a Get
192     this.getScan = isStartRowAndEqualsStopRow();
193   }
194 
195   /**
196    * Creates a new instance of this class while copying all values.
197    *
198    * @param scan  The scan instance to copy from.
199    * @throws IOException When copying the values fails.
200    */
201   public Scan(Scan scan) throws IOException {
202     startRow = scan.getStartRow();
203     stopRow  = scan.getStopRow();
204     maxVersions = scan.getMaxVersions();
205     batch = scan.getBatch();
206     storeLimit = scan.getMaxResultsPerColumnFamily();
207     storeOffset = scan.getRowOffsetPerColumnFamily();
208     caching = scan.getCaching();
209     maxResultSize = scan.getMaxResultSize();
210     cacheBlocks = scan.getCacheBlocks();
211     getScan = scan.isGetScan();
212     filter = scan.getFilter(); // clone?
213     loadColumnFamiliesOnDemand = scan.getLoadColumnFamiliesOnDemandValue();
214     consistency = scan.getConsistency();
215     reversed = scan.isReversed();
216     small = scan.isSmall();
217     TimeRange ctr = scan.getTimeRange();
218     tr = new TimeRange(ctr.getMin(), ctr.getMax());
219     Map<byte[], NavigableSet<byte[]>> fams = scan.getFamilyMap();
220     for (Map.Entry<byte[],NavigableSet<byte[]>> entry : fams.entrySet()) {
221       byte [] fam = entry.getKey();
222       NavigableSet<byte[]> cols = entry.getValue();
223       if (cols != null && cols.size() > 0) {
224         for (byte[] col : cols) {
225           addColumn(fam, col);
226         }
227       } else {
228         addFamily(fam);
229       }
230     }
231     for (Map.Entry<String, byte[]> attr : scan.getAttributesMap().entrySet()) {
232       setAttribute(attr.getKey(), attr.getValue());
233     }
234   }
235 
236   /**
237    * Builds a scan object with the same specs as get.
238    * @param get get to model scan after
239    */
240   public Scan(Get get) {
241     this.startRow = get.getRow();
242     this.stopRow = get.getRow();
243     this.filter = get.getFilter();
244     this.cacheBlocks = get.getCacheBlocks();
245     this.maxVersions = get.getMaxVersions();
246     this.storeLimit = get.getMaxResultsPerColumnFamily();
247     this.storeOffset = get.getRowOffsetPerColumnFamily();
248     this.tr = get.getTimeRange();
249     this.familyMap = get.getFamilyMap();
250     this.getScan = true;
251     this.consistency = get.getConsistency();
252     for (Map.Entry<String, byte[]> attr : get.getAttributesMap().entrySet()) {
253       setAttribute(attr.getKey(), attr.getValue());
254     }
255   }
256 
257   public boolean isGetScan() {
258     return this.getScan || isStartRowAndEqualsStopRow();
259   }
260 
261   private boolean isStartRowAndEqualsStopRow() {
262     return this.startRow != null && this.startRow.length > 0 &&
263         Bytes.equals(this.startRow, this.stopRow);
264   }
265   /**
266    * Get all columns from the specified family.
267    * <p>
268    * Overrides previous calls to addColumn for this family.
269    * @param family family name
270    * @return this
271    */
272   public Scan addFamily(byte [] family) {
273     familyMap.remove(family);
274     familyMap.put(family, null);
275     return this;
276   }
277 
278   /**
279    * Get the column from the specified family with the specified qualifier.
280    * <p>
281    * Overrides previous calls to addFamily for this family.
282    * @param family family name
283    * @param qualifier column qualifier
284    * @return this
285    */
286   public Scan addColumn(byte [] family, byte [] qualifier) {
287     NavigableSet<byte []> set = familyMap.get(family);
288     if(set == null) {
289       set = new TreeSet<byte []>(Bytes.BYTES_COMPARATOR);
290     }
291     if (qualifier == null) {
292       qualifier = HConstants.EMPTY_BYTE_ARRAY;
293     }
294     set.add(qualifier);
295     familyMap.put(family, set);
296     return this;
297   }
298 
299   /**
300    * Get versions of columns only within the specified timestamp range,
301    * [minStamp, maxStamp).  Note, default maximum versions to return is 1.  If
302    * your time range spans more than one version and you want all versions
303    * returned, up the number of versions beyond the default.
304    * @param minStamp minimum timestamp value, inclusive
305    * @param maxStamp maximum timestamp value, exclusive
306    * @throws IOException if invalid time range
307    * @see #setMaxVersions()
308    * @see #setMaxVersions(int)
309    * @return this
310    */
311   public Scan setTimeRange(long minStamp, long maxStamp)
312   throws IOException {
313     tr = new TimeRange(minStamp, maxStamp);
314     return this;
315   }
316 
317   /**
318    * Get versions of columns with the specified timestamp. Note, default maximum
319    * versions to return is 1.  If your time range spans more than one version
320    * and you want all versions returned, up the number of versions beyond the
321    * defaut.
322    * @param timestamp version timestamp
323    * @see #setMaxVersions()
324    * @see #setMaxVersions(int)
325    * @return this
326    */
327   public Scan setTimeStamp(long timestamp)
328   throws IOException {
329     try {
330       tr = new TimeRange(timestamp, timestamp+1);
331     } catch(IOException e) {
332       // This should never happen, unless integer overflow or something extremely wrong...
333       LOG.error("TimeRange failed, likely caused by integer overflow. ", e);
334       throw e;
335     }
336     return this;
337   }
338 
339   /**
340    * Set the start row of the scan.
341    * @param startRow row to start scan on (inclusive)
342    * Note: In order to make startRow exclusive add a trailing 0 byte
343    * @return this
344    */
345   public Scan setStartRow(byte [] startRow) {
346     this.startRow = startRow;
347     return this;
348   }
349 
350   /**
351    * Set the stop row.
352    * @param stopRow row to end at (exclusive)
353    * <p><b>Note:</b> In order to make stopRow inclusive add a trailing 0 byte</p>
354    * <p><b>Note:</b> When doing a filter for a rowKey <u>Prefix</u>
355    * use {@link #setRowPrefixFilter(byte[])}.
356    * The 'trailing 0' will not yield the desired result.</p>
357    * @return this
358    */
359   public Scan setStopRow(byte [] stopRow) {
360     this.stopRow = stopRow;
361     return this;
362   }
363 
364   /**
365    * <p>Set a filter (using stopRow and startRow) so the result set only contains rows where the
366    * rowKey starts with the specified prefix.</p>
367    * <p>This is a utility method that converts the desired rowPrefix into the appropriate values
368    * for the startRow and stopRow to achieve the desired result.</p>
369    * <p>This can safely be used in combination with setFilter.</p>
370    * <p><b>NOTE: Doing a {@link #setStartRow(byte[])} and/or {@link #setStopRow(byte[])}
371    * after this method will yield undefined results.</b></p>
372    * @param rowPrefix the prefix all rows must start with. (Set <i>null</i> to remove the filter.)
373    * @return this
374    */
375   public Scan setRowPrefixFilter(byte[] rowPrefix) {
376     if (rowPrefix == null) {
377       setStartRow(HConstants.EMPTY_START_ROW);
378       setStopRow(HConstants.EMPTY_END_ROW);
379     } else {
380       this.setStartRow(rowPrefix);
381       this.setStopRow(calculateTheClosestNextRowKeyForPrefix(rowPrefix));
382     }
383     return this;
384   }
385 
386   /**
387    * <p>When scanning for a prefix the scan should stop immediately after the the last row that
388    * has the specified prefix. This method calculates the closest next rowKey immediately following
389    * the given rowKeyPrefix.</p>
390    * <p><b>IMPORTANT: This converts a rowKey<u>Prefix</u> into a rowKey</b>.</p>
391    * <p>If the prefix is an 'ASCII' string put into a byte[] then this is easy because you can
392    * simply increment the last byte of the array.
393    * But if your application uses real binary rowids you may run into the scenario that your
394    * prefix is something like:</p>
395    * &nbsp;&nbsp;&nbsp;<b>{ 0x12, 0x23, 0xFF, 0xFF }</b><br/>
396    * Then this stopRow needs to be fed into the actual scan<br/>
397    * &nbsp;&nbsp;&nbsp;<b>{ 0x12, 0x24 }</b> (Notice that it is shorter now)<br/>
398    * This method calculates the correct stop row value for this usecase.
399    *
400    * @param rowKeyPrefix the rowKey<u>Prefix</u>.
401    * @return the closest next rowKey immediately following the given rowKeyPrefix.
402    */
403   private byte[] calculateTheClosestNextRowKeyForPrefix(byte[] rowKeyPrefix) {
404     // Essentially we are treating it like an 'unsigned very very long' and doing +1 manually.
405     // Search for the place where the trailing 0xFFs start
406     int offset = rowKeyPrefix.length;
407     while (offset > 0) {
408       if (rowKeyPrefix[offset - 1] != (byte) 0xFF) {
409         break;
410       }
411       offset--;
412     }
413 
414     if (offset == 0) {
415       // We got an 0xFFFF... (only FFs) stopRow value which is
416       // the last possible prefix before the end of the table.
417       // So set it to stop at the 'end of the table'
418       return HConstants.EMPTY_END_ROW;
419     }
420 
421     // Copy the right length of the original
422     byte[] newStopRow = Arrays.copyOfRange(rowKeyPrefix, 0, offset);
423     // And increment the last one
424     newStopRow[newStopRow.length - 1]++;
425     return newStopRow;
426   }
427 
428   /**
429    * Get all available versions.
430    * @return this
431    */
432   public Scan setMaxVersions() {
433     this.maxVersions = Integer.MAX_VALUE;
434     return this;
435   }
436 
437   /**
438    * Get up to the specified number of versions of each column.
439    * @param maxVersions maximum versions for each column
440    * @return this
441    */
442   public Scan setMaxVersions(int maxVersions) {
443     this.maxVersions = maxVersions;
444     return this;
445   }
446 
447   /**
448    * Set the maximum number of values to return for each call to next()
449    * @param batch the maximum number of values
450    */
451   public Scan setBatch(int batch) {
452     if (this.hasFilter() && this.filter.hasFilterRow()) {
453       throw new IncompatibleFilterException(
454         "Cannot set batch on a scan using a filter" +
455         " that returns true for filter.hasFilterRow");
456     }
457     this.batch = batch;
458     return this;
459   }
460 
461   /**
462    * Set the maximum number of values to return per row per Column Family
463    * @param limit the maximum number of values returned / row / CF
464    */
465   public Scan setMaxResultsPerColumnFamily(int limit) {
466     this.storeLimit = limit;
467     return this;
468   }
469 
470   /**
471    * Set offset for the row per Column Family.
472    * @param offset is the number of kvs that will be skipped.
473    */
474   public Scan setRowOffsetPerColumnFamily(int offset) {
475     this.storeOffset = offset;
476     return this;
477   }
478 
479   /**
480    * Set the number of rows for caching that will be passed to scanners.
481    * If not set, the default setting from {@link HTable#getScannerCaching()} will apply.
482    * Higher caching values will enable faster scanners but will use more memory.
483    * @param caching the number of rows for caching
484    */
485   public Scan setCaching(int caching) {
486     this.caching = caching;
487     return this;
488   }
489 
490   /**
491    * @return the maximum result size in bytes. See {@link #setMaxResultSize(long)}
492    */
493   public long getMaxResultSize() {
494     return maxResultSize;
495   }
496 
497   /**
498    * Set the maximum result size. The default is -1; this means that no specific
499    * maximum result size will be set for this scan, and the global configured
500    * value will be used instead. (Defaults to unlimited).
501    *
502    * @param maxResultSize The maximum result size in bytes.
503    */
504   public Scan setMaxResultSize(long maxResultSize) {
505     this.maxResultSize = maxResultSize;
506     return this;
507   }
508 
509   @Override
510   public Scan setFilter(Filter filter) {
511     super.setFilter(filter);
512     return this;
513   }
514 
515   /**
516    * Setting the familyMap
517    * @param familyMap map of family to qualifier
518    * @return this
519    */
520   public Scan setFamilyMap(Map<byte [], NavigableSet<byte []>> familyMap) {
521     this.familyMap = familyMap;
522     return this;
523   }
524 
525   /**
526    * Getting the familyMap
527    * @return familyMap
528    */
529   public Map<byte [], NavigableSet<byte []>> getFamilyMap() {
530     return this.familyMap;
531   }
532 
533   /**
534    * @return the number of families in familyMap
535    */
536   public int numFamilies() {
537     if(hasFamilies()) {
538       return this.familyMap.size();
539     }
540     return 0;
541   }
542 
543   /**
544    * @return true if familyMap is non empty, false otherwise
545    */
546   public boolean hasFamilies() {
547     return !this.familyMap.isEmpty();
548   }
549 
550   /**
551    * @return the keys of the familyMap
552    */
553   public byte[][] getFamilies() {
554     if(hasFamilies()) {
555       return this.familyMap.keySet().toArray(new byte[0][0]);
556     }
557     return null;
558   }
559 
560   /**
561    * @return the startrow
562    */
563   public byte [] getStartRow() {
564     return this.startRow;
565   }
566 
567   /**
568    * @return the stoprow
569    */
570   public byte [] getStopRow() {
571     return this.stopRow;
572   }
573 
574   /**
575    * @return the max number of versions to fetch
576    */
577   public int getMaxVersions() {
578     return this.maxVersions;
579   }
580 
581   /**
582    * @return maximum number of values to return for a single call to next()
583    */
584   public int getBatch() {
585     return this.batch;
586   }
587 
588   /**
589    * @return maximum number of values to return per row per CF
590    */
591   public int getMaxResultsPerColumnFamily() {
592     return this.storeLimit;
593   }
594 
595   /**
596    * Method for retrieving the scan's offset per row per column
597    * family (#kvs to be skipped)
598    * @return row offset
599    */
600   public int getRowOffsetPerColumnFamily() {
601     return this.storeOffset;
602   }
603 
604   /**
605    * @return caching the number of rows fetched when calling next on a scanner
606    */
607   public int getCaching() {
608     return this.caching;
609   }
610 
611   /**
612    * @return TimeRange
613    */
614   public TimeRange getTimeRange() {
615     return this.tr;
616   }
617 
618   /**
619    * @return RowFilter
620    */
621   @Override
622   public Filter getFilter() {
623     return filter;
624   }
625 
626   /**
627    * @return true is a filter has been specified, false if not
628    */
629   public boolean hasFilter() {
630     return filter != null;
631   }
632 
633   /**
634    * Set whether blocks should be cached for this Scan.
635    * <p>
636    * This is true by default.  When true, default settings of the table and
637    * family are used (this will never override caching blocks if the block
638    * cache is disabled for that family or entirely).
639    *
640    * @param cacheBlocks if false, default settings are overridden and blocks
641    * will not be cached
642    */
643   public Scan setCacheBlocks(boolean cacheBlocks) {
644     this.cacheBlocks = cacheBlocks;
645     return this;
646   }
647 
648   /**
649    * Get whether blocks should be cached for this Scan.
650    * @return true if default caching should be used, false if blocks should not
651    * be cached
652    */
653   public boolean getCacheBlocks() {
654     return cacheBlocks;
655   }
656 
657   /**
658    * Set whether this scan is a reversed one
659    * <p>
660    * This is false by default which means forward(normal) scan.
661    *
662    * @param reversed if true, scan will be backward order
663    * @return this
664    */
665   public Scan setReversed(boolean reversed) {
666     this.reversed = reversed;
667     return this;
668   }
669 
670   /**
671    * Get whether this scan is a reversed one.
672    * @return true if backward scan, false if forward(default) scan
673    */
674   public boolean isReversed() {
675     return reversed;
676   }
677 
678   /**
679    * Set the value indicating whether loading CFs on demand should be allowed (cluster
680    * default is false). On-demand CF loading doesn't load column families until necessary, e.g.
681    * if you filter on one column, the other column family data will be loaded only for the rows
682    * that are included in result, not all rows like in normal case.
683    * With column-specific filters, like SingleColumnValueFilter w/filterIfMissing == true,
684    * this can deliver huge perf gains when there's a cf with lots of data; however, it can
685    * also lead to some inconsistent results, as follows:
686    * - if someone does a concurrent update to both column families in question you may get a row
687    *   that never existed, e.g. for { rowKey = 5, { cat_videos => 1 }, { video => "my cat" } }
688    *   someone puts rowKey 5 with { cat_videos => 0 }, { video => "my dog" }, concurrent scan
689    *   filtering on "cat_videos == 1" can get { rowKey = 5, { cat_videos => 1 },
690    *   { video => "my dog" } }.
691    * - if there's a concurrent split and you have more than 2 column families, some rows may be
692    *   missing some column families.
693    */
694   public Scan setLoadColumnFamiliesOnDemand(boolean value) {
695     this.loadColumnFamiliesOnDemand = value;
696     return this;
697   }
698 
699   /**
700    * Get the raw loadColumnFamiliesOnDemand setting; if it's not set, can be null.
701    */
702   public Boolean getLoadColumnFamiliesOnDemandValue() {
703     return this.loadColumnFamiliesOnDemand;
704   }
705 
706   /**
707    * Get the logical value indicating whether on-demand CF loading should be allowed.
708    */
709   public boolean doLoadColumnFamiliesOnDemand() {
710     return (this.loadColumnFamiliesOnDemand != null)
711       && this.loadColumnFamiliesOnDemand.booleanValue();
712   }
713 
714   /**
715    * Compile the table and column family (i.e. schema) information
716    * into a String. Useful for parsing and aggregation by debugging,
717    * logging, and administration tools.
718    * @return Map
719    */
720   @Override
721   public Map<String, Object> getFingerprint() {
722     Map<String, Object> map = new HashMap<String, Object>();
723     List<String> families = new ArrayList<String>();
724     if(this.familyMap.size() == 0) {
725       map.put("families", "ALL");
726       return map;
727     } else {
728       map.put("families", families);
729     }
730     for (Map.Entry<byte [], NavigableSet<byte[]>> entry :
731         this.familyMap.entrySet()) {
732       families.add(Bytes.toStringBinary(entry.getKey()));
733     }
734     return map;
735   }
736 
737   /**
738    * Compile the details beyond the scope of getFingerprint (row, columns,
739    * timestamps, etc.) into a Map along with the fingerprinted information.
740    * Useful for debugging, logging, and administration tools.
741    * @param maxCols a limit on the number of columns output prior to truncation
742    * @return Map
743    */
744   @Override
745   public Map<String, Object> toMap(int maxCols) {
746     // start with the fingerpring map and build on top of it
747     Map<String, Object> map = getFingerprint();
748     // map from families to column list replaces fingerprint's list of families
749     Map<String, List<String>> familyColumns =
750       new HashMap<String, List<String>>();
751     map.put("families", familyColumns);
752     // add scalar information first
753     map.put("startRow", Bytes.toStringBinary(this.startRow));
754     map.put("stopRow", Bytes.toStringBinary(this.stopRow));
755     map.put("maxVersions", this.maxVersions);
756     map.put("batch", this.batch);
757     map.put("caching", this.caching);
758     map.put("maxResultSize", this.maxResultSize);
759     map.put("cacheBlocks", this.cacheBlocks);
760     map.put("loadColumnFamiliesOnDemand", this.loadColumnFamiliesOnDemand);
761     List<Long> timeRange = new ArrayList<Long>();
762     timeRange.add(this.tr.getMin());
763     timeRange.add(this.tr.getMax());
764     map.put("timeRange", timeRange);
765     int colCount = 0;
766     // iterate through affected families and list out up to maxCols columns
767     for (Map.Entry<byte [], NavigableSet<byte[]>> entry :
768       this.familyMap.entrySet()) {
769       List<String> columns = new ArrayList<String>();
770       familyColumns.put(Bytes.toStringBinary(entry.getKey()), columns);
771       if(entry.getValue() == null) {
772         colCount++;
773         --maxCols;
774         columns.add("ALL");
775       } else {
776         colCount += entry.getValue().size();
777         if (maxCols <= 0) {
778           continue;
779         }
780         for (byte [] column : entry.getValue()) {
781           if (--maxCols <= 0) {
782             continue;
783           }
784           columns.add(Bytes.toStringBinary(column));
785         }
786       }
787     }
788     map.put("totalColumns", colCount);
789     if (this.filter != null) {
790       map.put("filter", this.filter.toString());
791     }
792     // add the id if set
793     if (getId() != null) {
794       map.put("id", getId());
795     }
796     return map;
797   }
798 
799   /**
800    * Enable/disable "raw" mode for this scan.
801    * If "raw" is enabled the scan will return all
802    * delete marker and deleted rows that have not
803    * been collected, yet.
804    * This is mostly useful for Scan on column families
805    * that have KEEP_DELETED_ROWS enabled.
806    * It is an error to specify any column when "raw" is set.
807    * @param raw True/False to enable/disable "raw" mode.
808    */
809   public Scan setRaw(boolean raw) {
810     setAttribute(RAW_ATTR, Bytes.toBytes(raw));
811     return this;
812   }
813 
814   /**
815    * @return True if this Scan is in "raw" mode.
816    */
817   public boolean isRaw() {
818     byte[] attr = getAttribute(RAW_ATTR);
819     return attr == null ? false : Bytes.toBoolean(attr);
820   }
821 
822 
823 
824   /**
825    * Set whether this scan is a small scan
826    * <p>
827    * Small scan should use pread and big scan can use seek + read
828    *
829    * seek + read is fast but can cause two problem (1) resource contention (2)
830    * cause too much network io
831    *
832    * [89-fb] Using pread for non-compaction read request
833    * https://issues.apache.org/jira/browse/HBASE-7266
834    *
835    * On the other hand, if setting it true, we would do
836    * openScanner,next,closeScanner in one RPC call. It means the better
837    * performance for small scan. [HBASE-9488].
838    *
839    * Generally, if the scan range is within one data block(64KB), it could be
840    * considered as a small scan.
841    *
842    * @param small
843    */
844   public Scan setSmall(boolean small) {
845     this.small = small;
846     return this;
847   }
848 
849   /**
850    * Get whether this scan is a small scan
851    * @return true if small scan
852    */
853   public boolean isSmall() {
854     return small;
855   }
856 
857   @Override
858   public Scan setAttribute(String name, byte[] value) {
859     return (Scan) super.setAttribute(name, value);
860   }
861 
862   @Override
863   public Scan setId(String id) {
864     return (Scan) super.setId(id);
865   }
866 
867   @Override
868   public Scan setAuthorizations(Authorizations authorizations) {
869     return (Scan) super.setAuthorizations(authorizations);
870   }
871 
872   @Override
873   public Scan setACL(Map<String, Permission> perms) {
874     return (Scan) super.setACL(perms);
875   }
876 
877   @Override
878   public Scan setACL(String user, Permission perms) {
879     return (Scan) super.setACL(user, perms);
880   }
881 
882   @Override
883   public Scan setConsistency(Consistency consistency) {
884     return (Scan) super.setConsistency(consistency);
885   }
886 
887   @Override
888   public Scan setReplicaId(int Id) {
889     return (Scan) super.setReplicaId(Id);
890   }
891 
892   @Override
893   public Scan setIsolationLevel(IsolationLevel level) {
894     return (Scan) super.setIsolationLevel(level);
895   }
896 
897 }