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     TimeRange ctr = scan.getTimeRange();
217     tr = new TimeRange(ctr.getMin(), ctr.getMax());
218     Map<byte[], NavigableSet<byte[]>> fams = scan.getFamilyMap();
219     for (Map.Entry<byte[],NavigableSet<byte[]>> entry : fams.entrySet()) {
220       byte [] fam = entry.getKey();
221       NavigableSet<byte[]> cols = entry.getValue();
222       if (cols != null && cols.size() > 0) {
223         for (byte[] col : cols) {
224           addColumn(fam, col);
225         }
226       } else {
227         addFamily(fam);
228       }
229     }
230     for (Map.Entry<String, byte[]> attr : scan.getAttributesMap().entrySet()) {
231       setAttribute(attr.getKey(), attr.getValue());
232     }
233   }
234 
235   /**
236    * Builds a scan object with the same specs as get.
237    * @param get get to model scan after
238    */
239   public Scan(Get get) {
240     this.startRow = get.getRow();
241     this.stopRow = get.getRow();
242     this.filter = get.getFilter();
243     this.cacheBlocks = get.getCacheBlocks();
244     this.maxVersions = get.getMaxVersions();
245     this.storeLimit = get.getMaxResultsPerColumnFamily();
246     this.storeOffset = get.getRowOffsetPerColumnFamily();
247     this.tr = get.getTimeRange();
248     this.familyMap = get.getFamilyMap();
249     this.getScan = true;
250     this.consistency = get.getConsistency();
251     for (Map.Entry<String, byte[]> attr : get.getAttributesMap().entrySet()) {
252       setAttribute(attr.getKey(), attr.getValue());
253     }
254   }
255 
256   public boolean isGetScan() {
257     return this.getScan || isStartRowAndEqualsStopRow();
258   }
259 
260   private boolean isStartRowAndEqualsStopRow() {
261     return this.startRow != null && this.startRow.length > 0 &&
262         Bytes.equals(this.startRow, this.stopRow);
263   }
264   /**
265    * Get all columns from the specified family.
266    * <p>
267    * Overrides previous calls to addColumn for this family.
268    * @param family family name
269    * @return this
270    */
271   public Scan addFamily(byte [] family) {
272     familyMap.remove(family);
273     familyMap.put(family, null);
274     return this;
275   }
276 
277   /**
278    * Get the column from the specified family with the specified qualifier.
279    * <p>
280    * Overrides previous calls to addFamily for this family.
281    * @param family family name
282    * @param qualifier column qualifier
283    * @return this
284    */
285   public Scan addColumn(byte [] family, byte [] qualifier) {
286     NavigableSet<byte []> set = familyMap.get(family);
287     if(set == null) {
288       set = new TreeSet<byte []>(Bytes.BYTES_COMPARATOR);
289     }
290     if (qualifier == null) {
291       qualifier = HConstants.EMPTY_BYTE_ARRAY;
292     }
293     set.add(qualifier);
294     familyMap.put(family, set);
295     return this;
296   }
297 
298   /**
299    * Get versions of columns only within the specified timestamp range,
300    * [minStamp, maxStamp).  Note, default maximum versions to return is 1.  If
301    * your time range spans more than one version and you want all versions
302    * returned, up the number of versions beyond the default.
303    * @param minStamp minimum timestamp value, inclusive
304    * @param maxStamp maximum timestamp value, exclusive
305    * @throws IOException if invalid time range
306    * @see #setMaxVersions()
307    * @see #setMaxVersions(int)
308    * @return this
309    */
310   public Scan setTimeRange(long minStamp, long maxStamp)
311   throws IOException {
312     tr = new TimeRange(minStamp, maxStamp);
313     return this;
314   }
315 
316   /**
317    * Get versions of columns with the specified timestamp. Note, default maximum
318    * versions to return is 1.  If your time range spans more than one version
319    * and you want all versions returned, up the number of versions beyond the
320    * defaut.
321    * @param timestamp version timestamp
322    * @see #setMaxVersions()
323    * @see #setMaxVersions(int)
324    * @return this
325    */
326   public Scan setTimeStamp(long timestamp)
327   throws IOException {
328     try {
329       tr = new TimeRange(timestamp, timestamp+1);
330     } catch(IOException e) {
331       // This should never happen, unless integer overflow or something extremely wrong...
332       LOG.error("TimeRange failed, likely caused by integer overflow. ", e);
333       throw e;
334     }
335     return this;
336   }
337 
338   /**
339    * Set the start row of the scan.
340    * @param startRow row to start scan on (inclusive)
341    * Note: In order to make startRow exclusive add a trailing 0 byte
342    * @return this
343    */
344   public Scan setStartRow(byte [] startRow) {
345     this.startRow = startRow;
346     return this;
347   }
348 
349   /**
350    * Set the stop row.
351    * @param stopRow row to end at (exclusive)
352    * <p><b>Note:</b> In order to make stopRow inclusive add a trailing 0 byte</p>
353    * <p><b>Note:</b> When doing a filter for a rowKey <u>Prefix</u>
354    * use {@link #setRowPrefixFilter(byte[])}.
355    * The 'trailing 0' will not yield the desired result.</p>
356    * @return this
357    */
358   public Scan setStopRow(byte [] stopRow) {
359     this.stopRow = stopRow;
360     return this;
361   }
362 
363   /**
364    * <p>Set a filter (using stopRow and startRow) so the result set only contains rows where the
365    * rowKey starts with the specified prefix.</p>
366    * <p>This is a utility method that converts the desired rowPrefix into the appropriate values
367    * for the startRow and stopRow to achieve the desired result.</p>
368    * <p>This can safely be used in combination with setFilter.</p>
369    * <p><b>NOTE: Doing a {@link #setStartRow(byte[])} and/or {@link #setStopRow(byte[])}
370    * after this method will yield undefined results.</b></p>
371    * @param rowPrefix the prefix all rows must start with. (Set <i>null</i> to remove the filter.)
372    * @return this
373    */
374   public Scan setRowPrefixFilter(byte[] rowPrefix) {
375     if (rowPrefix == null) {
376       setStartRow(HConstants.EMPTY_START_ROW);
377       setStopRow(HConstants.EMPTY_END_ROW);
378     } else {
379       this.setStartRow(rowPrefix);
380       this.setStopRow(calculateTheClosestNextRowKeyForPrefix(rowPrefix));
381     }
382     return this;
383   }
384 
385   /**
386    * <p>When scanning for a prefix the scan should stop immediately after the the last row that
387    * has the specified prefix. This method calculates the closest next rowKey immediately following
388    * the given rowKeyPrefix.</p>
389    * <p><b>IMPORTANT: This converts a rowKey<u>Prefix</u> into a rowKey</b>.</p>
390    * <p>If the prefix is an 'ASCII' string put into a byte[] then this is easy because you can
391    * simply increment the last byte of the array.
392    * But if your application uses real binary rowids you may run into the scenario that your
393    * prefix is something like:</p>
394    * &nbsp;&nbsp;&nbsp;<b>{ 0x12, 0x23, 0xFF, 0xFF }</b><br/>
395    * Then this stopRow needs to be fed into the actual scan<br/>
396    * &nbsp;&nbsp;&nbsp;<b>{ 0x12, 0x24 }</b> (Notice that it is shorter now)<br/>
397    * This method calculates the correct stop row value for this usecase.
398    *
399    * @param rowKeyPrefix the rowKey<u>Prefix</u>.
400    * @return the closest next rowKey immediately following the given rowKeyPrefix.
401    */
402   private byte[] calculateTheClosestNextRowKeyForPrefix(byte[] rowKeyPrefix) {
403     // Essentially we are treating it like an 'unsigned very very long' and doing +1 manually.
404     // Search for the place where the trailing 0xFFs start
405     int offset = rowKeyPrefix.length;
406     while (offset > 0) {
407       if (rowKeyPrefix[offset - 1] != (byte) 0xFF) {
408         break;
409       }
410       offset--;
411     }
412 
413     if (offset == 0) {
414       // We got an 0xFFFF... (only FFs) stopRow value which is
415       // the last possible prefix before the end of the table.
416       // So set it to stop at the 'end of the table'
417       return HConstants.EMPTY_END_ROW;
418     }
419 
420     // Copy the right length of the original
421     byte[] newStopRow = Arrays.copyOfRange(rowKeyPrefix, 0, offset);
422     // And increment the last one
423     newStopRow[newStopRow.length - 1]++;
424     return newStopRow;
425   }
426 
427   /**
428    * Get all available versions.
429    * @return this
430    */
431   public Scan setMaxVersions() {
432     this.maxVersions = Integer.MAX_VALUE;
433     return this;
434   }
435 
436   /**
437    * Get up to the specified number of versions of each column.
438    * @param maxVersions maximum versions for each column
439    * @return this
440    */
441   public Scan setMaxVersions(int maxVersions) {
442     this.maxVersions = maxVersions;
443     return this;
444   }
445 
446   /**
447    * Set the maximum number of values to return for each call to next()
448    * @param batch the maximum number of values
449    */
450   public Scan setBatch(int batch) {
451     if (this.hasFilter() && this.filter.hasFilterRow()) {
452       throw new IncompatibleFilterException(
453         "Cannot set batch on a scan using a filter" +
454         " that returns true for filter.hasFilterRow");
455     }
456     this.batch = batch;
457     return this;
458   }
459 
460   /**
461    * Set the maximum number of values to return per row per Column Family
462    * @param limit the maximum number of values returned / row / CF
463    */
464   public Scan setMaxResultsPerColumnFamily(int limit) {
465     this.storeLimit = limit;
466     return this;
467   }
468 
469   /**
470    * Set offset for the row per Column Family.
471    * @param offset is the number of kvs that will be skipped.
472    */
473   public Scan setRowOffsetPerColumnFamily(int offset) {
474     this.storeOffset = offset;
475     return this;
476   }
477 
478   /**
479    * Set the number of rows for caching that will be passed to scanners.
480    * If not set, the default setting from {@link HTable#getScannerCaching()} will apply.
481    * Higher caching values will enable faster scanners but will use more memory.
482    * @param caching the number of rows for caching
483    */
484   public Scan setCaching(int caching) {
485     this.caching = caching;
486     return this;
487   }
488 
489   /**
490    * @return the maximum result size in bytes. See {@link #setMaxResultSize(long)}
491    */
492   public long getMaxResultSize() {
493     return maxResultSize;
494   }
495 
496   /**
497    * Set the maximum result size. The default is -1; this means that no specific
498    * maximum result size will be set for this scan, and the global configured
499    * value will be used instead. (Defaults to unlimited).
500    *
501    * @param maxResultSize The maximum result size in bytes.
502    */
503   public Scan setMaxResultSize(long maxResultSize) {
504     this.maxResultSize = maxResultSize;
505     return this;
506   }
507 
508   @Override
509   public Scan setFilter(Filter filter) {
510     super.setFilter(filter);
511     return this;
512   }
513 
514   /**
515    * Setting the familyMap
516    * @param familyMap map of family to qualifier
517    * @return this
518    */
519   public Scan setFamilyMap(Map<byte [], NavigableSet<byte []>> familyMap) {
520     this.familyMap = familyMap;
521     return this;
522   }
523 
524   /**
525    * Getting the familyMap
526    * @return familyMap
527    */
528   public Map<byte [], NavigableSet<byte []>> getFamilyMap() {
529     return this.familyMap;
530   }
531 
532   /**
533    * @return the number of families in familyMap
534    */
535   public int numFamilies() {
536     if(hasFamilies()) {
537       return this.familyMap.size();
538     }
539     return 0;
540   }
541 
542   /**
543    * @return true if familyMap is non empty, false otherwise
544    */
545   public boolean hasFamilies() {
546     return !this.familyMap.isEmpty();
547   }
548 
549   /**
550    * @return the keys of the familyMap
551    */
552   public byte[][] getFamilies() {
553     if(hasFamilies()) {
554       return this.familyMap.keySet().toArray(new byte[0][0]);
555     }
556     return null;
557   }
558 
559   /**
560    * @return the startrow
561    */
562   public byte [] getStartRow() {
563     return this.startRow;
564   }
565 
566   /**
567    * @return the stoprow
568    */
569   public byte [] getStopRow() {
570     return this.stopRow;
571   }
572 
573   /**
574    * @return the max number of versions to fetch
575    */
576   public int getMaxVersions() {
577     return this.maxVersions;
578   }
579 
580   /**
581    * @return maximum number of values to return for a single call to next()
582    */
583   public int getBatch() {
584     return this.batch;
585   }
586 
587   /**
588    * @return maximum number of values to return per row per CF
589    */
590   public int getMaxResultsPerColumnFamily() {
591     return this.storeLimit;
592   }
593 
594   /**
595    * Method for retrieving the scan's offset per row per column
596    * family (#kvs to be skipped)
597    * @return row offset
598    */
599   public int getRowOffsetPerColumnFamily() {
600     return this.storeOffset;
601   }
602 
603   /**
604    * @return caching the number of rows fetched when calling next on a scanner
605    */
606   public int getCaching() {
607     return this.caching;
608   }
609 
610   /**
611    * @return TimeRange
612    */
613   public TimeRange getTimeRange() {
614     return this.tr;
615   }
616 
617   /**
618    * @return RowFilter
619    */
620   @Override
621   public Filter getFilter() {
622     return filter;
623   }
624 
625   /**
626    * @return true is a filter has been specified, false if not
627    */
628   public boolean hasFilter() {
629     return filter != null;
630   }
631 
632   /**
633    * Set whether blocks should be cached for this Scan.
634    * <p>
635    * This is true by default.  When true, default settings of the table and
636    * family are used (this will never override caching blocks if the block
637    * cache is disabled for that family or entirely).
638    *
639    * @param cacheBlocks if false, default settings are overridden and blocks
640    * will not be cached
641    */
642   public Scan setCacheBlocks(boolean cacheBlocks) {
643     this.cacheBlocks = cacheBlocks;
644     return this;
645   }
646 
647   /**
648    * Get whether blocks should be cached for this Scan.
649    * @return true if default caching should be used, false if blocks should not
650    * be cached
651    */
652   public boolean getCacheBlocks() {
653     return cacheBlocks;
654   }
655 
656   /**
657    * Set whether this scan is a reversed one
658    * <p>
659    * This is false by default which means forward(normal) scan.
660    *
661    * @param reversed if true, scan will be backward order
662    * @return this
663    */
664   public Scan setReversed(boolean reversed) {
665     this.reversed = reversed;
666     return this;
667   }
668 
669   /**
670    * Get whether this scan is a reversed one.
671    * @return true if backward scan, false if forward(default) scan
672    */
673   public boolean isReversed() {
674     return reversed;
675   }
676 
677   /**
678    * Set the value indicating whether loading CFs on demand should be allowed (cluster
679    * default is false). On-demand CF loading doesn't load column families until necessary, e.g.
680    * if you filter on one column, the other column family data will be loaded only for the rows
681    * that are included in result, not all rows like in normal case.
682    * With column-specific filters, like SingleColumnValueFilter w/filterIfMissing == true,
683    * this can deliver huge perf gains when there's a cf with lots of data; however, it can
684    * also lead to some inconsistent results, as follows:
685    * - if someone does a concurrent update to both column families in question you may get a row
686    *   that never existed, e.g. for { rowKey = 5, { cat_videos => 1 }, { video => "my cat" } }
687    *   someone puts rowKey 5 with { cat_videos => 0 }, { video => "my dog" }, concurrent scan
688    *   filtering on "cat_videos == 1" can get { rowKey = 5, { cat_videos => 1 },
689    *   { video => "my dog" } }.
690    * - if there's a concurrent split and you have more than 2 column families, some rows may be
691    *   missing some column families.
692    */
693   public Scan setLoadColumnFamiliesOnDemand(boolean value) {
694     this.loadColumnFamiliesOnDemand = value;
695     return this;
696   }
697 
698   /**
699    * Get the raw loadColumnFamiliesOnDemand setting; if it's not set, can be null.
700    */
701   public Boolean getLoadColumnFamiliesOnDemandValue() {
702     return this.loadColumnFamiliesOnDemand;
703   }
704 
705   /**
706    * Get the logical value indicating whether on-demand CF loading should be allowed.
707    */
708   public boolean doLoadColumnFamiliesOnDemand() {
709     return (this.loadColumnFamiliesOnDemand != null)
710       && this.loadColumnFamiliesOnDemand.booleanValue();
711   }
712 
713   /**
714    * Compile the table and column family (i.e. schema) information
715    * into a String. Useful for parsing and aggregation by debugging,
716    * logging, and administration tools.
717    * @return Map
718    */
719   @Override
720   public Map<String, Object> getFingerprint() {
721     Map<String, Object> map = new HashMap<String, Object>();
722     List<String> families = new ArrayList<String>();
723     if(this.familyMap.size() == 0) {
724       map.put("families", "ALL");
725       return map;
726     } else {
727       map.put("families", families);
728     }
729     for (Map.Entry<byte [], NavigableSet<byte[]>> entry :
730         this.familyMap.entrySet()) {
731       families.add(Bytes.toStringBinary(entry.getKey()));
732     }
733     return map;
734   }
735 
736   /**
737    * Compile the details beyond the scope of getFingerprint (row, columns,
738    * timestamps, etc.) into a Map along with the fingerprinted information.
739    * Useful for debugging, logging, and administration tools.
740    * @param maxCols a limit on the number of columns output prior to truncation
741    * @return Map
742    */
743   @Override
744   public Map<String, Object> toMap(int maxCols) {
745     // start with the fingerpring map and build on top of it
746     Map<String, Object> map = getFingerprint();
747     // map from families to column list replaces fingerprint's list of families
748     Map<String, List<String>> familyColumns =
749       new HashMap<String, List<String>>();
750     map.put("families", familyColumns);
751     // add scalar information first
752     map.put("startRow", Bytes.toStringBinary(this.startRow));
753     map.put("stopRow", Bytes.toStringBinary(this.stopRow));
754     map.put("maxVersions", this.maxVersions);
755     map.put("batch", this.batch);
756     map.put("caching", this.caching);
757     map.put("maxResultSize", this.maxResultSize);
758     map.put("cacheBlocks", this.cacheBlocks);
759     map.put("loadColumnFamiliesOnDemand", this.loadColumnFamiliesOnDemand);
760     List<Long> timeRange = new ArrayList<Long>();
761     timeRange.add(this.tr.getMin());
762     timeRange.add(this.tr.getMax());
763     map.put("timeRange", timeRange);
764     int colCount = 0;
765     // iterate through affected families and list out up to maxCols columns
766     for (Map.Entry<byte [], NavigableSet<byte[]>> entry :
767       this.familyMap.entrySet()) {
768       List<String> columns = new ArrayList<String>();
769       familyColumns.put(Bytes.toStringBinary(entry.getKey()), columns);
770       if(entry.getValue() == null) {
771         colCount++;
772         --maxCols;
773         columns.add("ALL");
774       } else {
775         colCount += entry.getValue().size();
776         if (maxCols <= 0) {
777           continue;
778         }
779         for (byte [] column : entry.getValue()) {
780           if (--maxCols <= 0) {
781             continue;
782           }
783           columns.add(Bytes.toStringBinary(column));
784         }
785       }
786     }
787     map.put("totalColumns", colCount);
788     if (this.filter != null) {
789       map.put("filter", this.filter.toString());
790     }
791     // add the id if set
792     if (getId() != null) {
793       map.put("id", getId());
794     }
795     return map;
796   }
797 
798   /**
799    * Enable/disable "raw" mode for this scan.
800    * If "raw" is enabled the scan will return all
801    * delete marker and deleted rows that have not
802    * been collected, yet.
803    * This is mostly useful for Scan on column families
804    * that have KEEP_DELETED_ROWS enabled.
805    * It is an error to specify any column when "raw" is set.
806    * @param raw True/False to enable/disable "raw" mode.
807    */
808   public Scan setRaw(boolean raw) {
809     setAttribute(RAW_ATTR, Bytes.toBytes(raw));
810     return this;
811   }
812 
813   /**
814    * @return True if this Scan is in "raw" mode.
815    */
816   public boolean isRaw() {
817     byte[] attr = getAttribute(RAW_ATTR);
818     return attr == null ? false : Bytes.toBoolean(attr);
819   }
820 
821 
822 
823   /**
824    * Set whether this scan is a small scan
825    * <p>
826    * Small scan should use pread and big scan can use seek + read
827    *
828    * seek + read is fast but can cause two problem (1) resource contention (2)
829    * cause too much network io
830    *
831    * [89-fb] Using pread for non-compaction read request
832    * https://issues.apache.org/jira/browse/HBASE-7266
833    *
834    * On the other hand, if setting it true, we would do
835    * openScanner,next,closeScanner in one RPC call. It means the better
836    * performance for small scan. [HBASE-9488].
837    *
838    * Generally, if the scan range is within one data block(64KB), it could be
839    * considered as a small scan.
840    *
841    * @param small
842    */
843   public Scan setSmall(boolean small) {
844     this.small = small;
845     return this;
846   }
847 
848   /**
849    * Get whether this scan is a small scan
850    * @return true if small scan
851    */
852   public boolean isSmall() {
853     return small;
854   }
855 
856   @Override
857   public Scan setAttribute(String name, byte[] value) {
858     return (Scan) super.setAttribute(name, value);
859   }
860 
861   @Override
862   public Scan setId(String id) {
863     return (Scan) super.setId(id);
864   }
865 
866   @Override
867   public Scan setAuthorizations(Authorizations authorizations) {
868     return (Scan) super.setAuthorizations(authorizations);
869   }
870 
871   @Override
872   public Scan setACL(Map<String, Permission> perms) {
873     return (Scan) super.setACL(perms);
874   }
875 
876   @Override
877   public Scan setACL(String user, Permission perms) {
878     return (Scan) super.setACL(user, perms);
879   }
880 
881   @Override
882   public Scan setConsistency(Consistency consistency) {
883     return (Scan) super.setConsistency(consistency);
884   }
885 
886   @Override
887   public Scan setReplicaId(int Id) {
888     return (Scan) super.setReplicaId(Id);
889   }
890 
891   @Override
892   public Scan setIsolationLevel(IsolationLevel level) {
893     return (Scan) super.setIsolationLevel(level);
894   }
895 
896 }