~drizzle-trunk/drizzle/development

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
1001
1002
1003
1004
1005
1006
1007
1008
1009
1010
1011
1012
1013
1014
1015
1016
1017
1018
1019
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048
1049
1050
1051
1052
1053
1054
1055
1056
1057
1058
1059
1060
1061
1062
1063
1064
1065
1066
1067
1068
1069
1070
1071
1072
1073
1074
1075
1076
1077
1078
1079
1080
1081
1082
1083
1084
1085
1086
1087
1088
1089
1090
1091
1092
1093
1094
1095
1096
1097
1098
1099
1100
1101
1102
1103
1104
1105
1106
1107
1108
1109
1110
1111
1112
1113
1114
1115
1116
1117
1118
1119
1120
1121
1122
1123
1124
1125
1126
1127
1128
1129
1130
1131
1132
1133
1134
1135
1136
1137
1138
1139
1140
1141
1142
1143
1144
1145
1146
1147
1148
1149
1150
1151
1152
1153
1154
1155
1156
1157
1158
1159
1160
1161
1162
1163
1164
1165
1166
1167
1168
1169
1170
1171
1172
1173
1174
1175
1176
1177
1178
1179
1180
1181
1182
1183
1184
1185
1186
1187
1188
1189
1190
1191
1192
1193
1194
1195
1196
1197
1198
1199
1200
1201
1202
1203
1204
1205
1206
1207
1208
1209
1210
1211
1212
1213
1214
1215
1216
1217
1218
1219
1220
1221
1222
1223
1224
1225
1226
1227
1228
1229
1230
1231
1232
1233
1234
1235
1236
1237
1238
1239
1240
1241
1242
1243
1244
1245
1246
1247
1248
1249
1250
1251
1252
1253
1254
1255
1256
1257
1258
1259
1260
1261
1262
1263
1264
1265
1266
1267
1268
1269
1270
1271
1272
1273
1274
1275
1276
1277
1278
1279
1280
1281
1282
1283
1284
1285
1286
1287
1288
1289
1290
1291
1292
1293
1294
1295
1296
1297
1298
1299
1300
1301
1302
1303
1304
1305
1306
1307
1308
1309
1310
1311
1312
1313
1314
1315
1316
1317
1318
1319
1320
1321
1322
1323
1324
1325
1326
1327
1328
1329
1330
1331
1332
1333
1334
1335
1336
1337
1338
1339
1340
1341
1342
1343
1344
1345
1346
1347
1348
1349
1350
1351
1352
1353
1354
1355
1356
1357
1358
1359
1360
1361
1362
1363
1364
1365
1366
1367
1368
1369
1370
1371
1372
1373
1374
1375
1376
1377
1378
1379
1380
1381
1382
1383
1384
1385
1386
1387
1388
1389
1390
1391
1392
1393
1394
1395
1396
1397
1398
1399
1400
1401
1402
1403
1404
1405
1406
1407
1408
1409
1410
1411
1412
1413
1414
1415
1416
1417
1418
1419
1420
1421
1422
1423
1424
1425
1426
1427
1428
1429
1430
1431
1432
1433
1434
1435
1436
1437
1438
1439
1440
1441
1442
1443
1444
1445
1446
1447
1448
1449
1450
1451
1452
1453
1454
1455
1456
1457
1458
1459
1460
1461
1462
1463
1464
1465
1466
1467
1468
1469
1470
1471
1472
1473
1474
1475
1476
1477
1478
1479
1480
1481
1482
1483
1484
1485
1486
1487
1488
1489
1490
1491
1492
1493
1494
1495
1496
1497
1498
1499
1500
1501
1502
1503
1504
1505
1506
1507
1508
1509
1510
1511
1512
1513
1514
1515
1516
1517
1518
1519
1520
1521
1522
1523
1524
1525
1526
1527
1528
1529
1530
1531
1532
1533
1534
1535
1536
1537
1538
1539
1540
1541
1542
1543
1544
1545
1546
1547
1548
1549
1550
1551
1552
1553
1554
1555
1556
1557
1558
1559
1560
1561
1562
1563
1564
1565
1566
1567
1568
1569
1570
1571
1572
1573
1574
1575
1576
1577
1578
1579
1580
1581
1582
1583
1584
1585
1586
1587
1588
1589
1590
1591
1592
1593
1594
1595
1596
1597
1598
1599
1600
1601
1602
1603
1604
1605
1606
1607
1608
1609
1610
1611
1612
1613
1614
1615
1616
1617
1618
1619
1620
1621
1622
1623
1624
1625
1626
1627
1628
1629
1630
1631
1632
1633
1634
1635
1636
1637
1638
1639
1640
1641
1642
1643
1644
1645
1646
1647
1648
1649
1650
1651
1652
1653
1654
1655
1656
1657
1658
1659
1660
1661
1662
1663
1664
1665
1666
1667
1668
1669
1670
1671
1672
1673
1674
1675
1676
1677
1678
1679
1680
1681
1682
1683
1684
1685
1686
1687
1688
1689
1690
1691
1692
1693
1694
1695
1696
1697
1698
1699
1700
1701
1702
1703
1704
1705
1706
1707
1708
1709
1710
1711
1712
1713
1714
1715
1716
1717
1718
1719
1720
1721
1722
1723
1724
1725
1726
1727
1728
1729
1730
1731
1732
1733
1734
1735
1736
1737
1738
1739
1740
1741
1742
1743
1744
1745
1746
1747
1748
1749
1750
1751
1752
1753
1754
1755
1756
1757
1758
1759
1760
1761
1762
1763
1764
1765
1766
1767
1768
/* Copyright (C) 2000-2006 MySQL AB

   This program is free software; you can redistribute it and/or modify
   it under the terms of the GNU General Public License as published by
   the Free Software Foundation; version 2 of the License.

   This program is distributed in the hope that it will be useful,
   but WITHOUT ANY WARRANTY; without even the implied warranty of
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
   GNU General Public License for more details.

   You should have received a copy of the GNU General Public License
   along with this program; if not, write to the Free Software
   Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA */


/* Some general useful functions */

#include "config.h"

#include <float.h>
#include <fcntl.h>

#include <string>
#include <vector>
#include <algorithm>

#include <drizzled/error.h>
#include <drizzled/gettext.h>

#include "drizzled/plugin/transactional_storage_engine.h"
#include "drizzled/plugin/authorization.h"
#include <drizzled/nested_join.h>
#include <drizzled/sql_parse.h>
#include <drizzled/item/sum.h>
#include <drizzled/table_list.h>
#include <drizzled/session.h>
#include <drizzled/sql_base.h>
#include <drizzled/sql_select.h>
#include <drizzled/field/blob.h>
#include <drizzled/field/varstring.h>
#include <drizzled/field/double.h>
#include <drizzled/unireg.h>
#include <drizzled/message/table.pb.h>
#include "drizzled/sql_table.h"
#include "drizzled/charset.h"
#include "drizzled/internal/m_string.h"
#include "plugin/myisam/myisam.h"

#include <drizzled/item/string.h>
#include <drizzled/item/int.h>
#include <drizzled/item/decimal.h>
#include <drizzled/item/float.h>
#include <drizzled/item/null.h>
#include <drizzled/temporal.h>

#include "drizzled/table/instance.h"

#include "drizzled/table_proto.h"

using namespace std;

namespace drizzled
{

extern pid_t current_pid;
extern plugin::StorageEngine *heap_engine;
extern plugin::StorageEngine *myisam_engine;

/* Functions defined in this cursor */

void open_table_error(TableShare *share, int error, int db_errno,
                      myf errortype, int errarg);

/*************************************************************************/

// @note this should all be the destructor
int Table::delete_table(bool free_share)
{
  int error= 0;

  if (db_stat)
    error= cursor->close();
  _alias.clear();
  if (field)
  {
    for (Field **ptr=field ; *ptr ; ptr++)
    {
      delete *ptr;
    }
    field= 0;
  }
  delete cursor;
  cursor= 0;				/* For easier errorchecking */

  if (free_share)
  {
    release();
  }

  return error;
}

Table::~Table()
{
  mem_root.free_root(MYF(0));
}


void Table::resetTable(Session *session,
                       TableShare *share,
                       uint32_t db_stat_arg)
{
  setShare(share);
  field= NULL;

  cursor= NULL;
  next= NULL;
  prev= NULL;

  read_set= NULL;
  write_set= NULL;

  tablenr= 0;
  db_stat= db_stat_arg;

  in_use= session;
  record[0]= (unsigned char *) NULL;
  record[1]= (unsigned char *) NULL;

  insert_values.clear();
  key_info= NULL;
  next_number_field= NULL;
  found_next_number_field= NULL;
  timestamp_field= NULL;

  pos_in_table_list= NULL;
  group= NULL;
  _alias.clear();
  null_flags= NULL;

  lock_position= 0;
  lock_data_start= 0;
  lock_count= 0;
  used_fields= 0;
  status= 0;
  derived_select_number= 0;
  current_lock= F_UNLCK;
  copy_blobs= false;

  maybe_null= false;

  null_row= false;

  force_index= false;
  distinct= false;
  const_table= false;
  no_rows= false;
  key_read= false;
  no_keyread= false;

  open_placeholder= false;
  locked_by_name= false;
  no_cache= false;

  auto_increment_field_not_null= false;
  alias_name_used= false;

  query_id= 0;
  quick_condition_rows= 0;

  timestamp_field_type= TIMESTAMP_NO_AUTO_SET;
  map= 0;

  reginfo.reset();

  covering_keys.reset();

  quick_keys.reset();
  merge_keys.reset();

  keys_in_use_for_query.reset();
  keys_in_use_for_group_by.reset();
  keys_in_use_for_order_by.reset();

  memset(quick_rows, 0, sizeof(ha_rows) * MAX_KEY);
  memset(const_key_parts, 0, sizeof(ha_rows) * MAX_KEY);

  memset(quick_key_parts, 0, sizeof(unsigned int) * MAX_KEY);
  memset(quick_n_ranges, 0, sizeof(unsigned int) * MAX_KEY);

  memory::init_sql_alloc(&mem_root, TABLE_ALLOC_BLOCK_SIZE, 0);
}



/* Deallocate temporary blob storage */

void free_blobs(register Table *table)
{
  uint32_t *ptr, *end;
  for (ptr= table->getBlobField(), end=ptr + table->sizeBlobFields();
       ptr != end ;
       ptr++)
  {
    ((Field_blob*) table->getField(*ptr))->free();
  }
}


TYPELIB *typelib(memory::Root *mem_root, List<String> &strings)
{
  TYPELIB *result= (TYPELIB*) mem_root->alloc_root(sizeof(TYPELIB));
  if (!result)
    return 0;
  result->count= strings.elements;
  result->name= "";
  uint32_t nbytes= (sizeof(char*) + sizeof(uint32_t)) * (result->count + 1);
  
  if (!(result->type_names= (const char**) mem_root->alloc_root(nbytes)))
    return 0;
    
  result->type_lengths= (uint*) (result->type_names + result->count + 1);

  List_iterator<String> it(strings);
  String *tmp;
  for (uint32_t i= 0; (tmp= it++); i++)
  {
    result->type_names[i]= tmp->ptr();
    result->type_lengths[i]= tmp->length();
  }

  result->type_names[result->count]= 0;   // End marker
  result->type_lengths[result->count]= 0;

  return result;
}

	/* Check that the integer is in the internal */

int set_zone(register int nr, int min_zone, int max_zone)
{
  if (nr<=min_zone)
    return (min_zone);
  if (nr>=max_zone)
    return (max_zone);
  return (nr);
} /* set_zone */


/*
  Store an SQL quoted string.

  SYNOPSIS
    append_unescaped()
    res		result String
    pos		string to be quoted
    length	it's length

  NOTE
    This function works correctly with utf8 or single-byte charset strings.
    May fail with some multibyte charsets though.
*/

void append_unescaped(String *res, const char *pos, uint32_t length)
{
  const char *end= pos+length;
  res->append('\'');

  for (; pos != end ; pos++)
  {
    uint32_t mblen;
    if (use_mb(default_charset_info) &&
        (mblen= my_ismbchar(default_charset_info, pos, end)))
    {
      res->append(pos, mblen);
      pos+= mblen - 1;
      if (pos >= end)
        break;
      continue;
    }

    switch (*pos) {
    case 0:				/* Must be escaped for 'mysql' */
      res->append('\\');
      res->append('0');
      break;
    case '\n':				/* Must be escaped for logs */
      res->append('\\');
      res->append('n');
      break;
    case '\r':
      res->append('\\');		/* This gives better readability */
      res->append('r');
      break;
    case '\\':
      res->append('\\');		/* Because of the sql syntax */
      res->append('\\');
      break;
    case '\'':
      res->append('\'');		/* Because of the sql syntax */
      res->append('\'');
      break;
    default:
      res->append(*pos);
      break;
    }
  }
  res->append('\'');
}


int rename_file_ext(const char * from,const char * to,const char * ext)
{
  string from_s, to_s;

  from_s.append(from);
  from_s.append(ext);
  to_s.append(to);
  to_s.append(ext);
  return (internal::my_rename(from_s.c_str(),to_s.c_str(),MYF(MY_WME)));
}

/*
  Check if database name is valid

  SYNPOSIS
    check_db_name()
    org_name		Name of database and length

  RETURN
    false error
    true ok
*/

bool check_db_name(Session *session, SchemaIdentifier &schema_identifier)
{
  if (not plugin::Authorization::isAuthorized(session->user(), schema_identifier))
  {
    return false;
  }

  return schema_identifier.isValid();
}

/*
  Allow anything as a table name, as long as it doesn't contain an
  ' ' at the end
  returns 1 on error
*/
bool check_table_name(const char *name, uint32_t length)
{
  if (!length || length > NAME_LEN || name[length - 1] == ' ')
    return 1;
  LEX_STRING ident;
  ident.str= (char*) name;
  ident.length= length;
  return check_identifier_name(&ident);
}


/*
  Eventually, a "length" argument should be added
  to this function, and the inner loop changed to
  check_identifier_name() call.
*/
bool check_column_name(const char *name)
{
  uint32_t name_length= 0;  // name length in symbols
  bool last_char_is_space= true;

  while (*name)
  {
    last_char_is_space= my_isspace(system_charset_info, *name);
    if (use_mb(system_charset_info))
    {
      int len=my_ismbchar(system_charset_info, name,
                          name+system_charset_info->mbmaxlen);
      if (len)
      {
        if (len > 3) /* Disallow non-BMP characters */
          return 1;
        name += len;
        name_length++;
        continue;
      }
    }
    /*
      NAMES_SEP_CHAR is used in FRM format to separate SET and ENUM values.
      It is defined as 0xFF, which is a not valid byte in utf8.
      This assert is to catch use of this byte if we decide to
      use non-utf8 as system_character_set.
    */
    assert(*name != NAMES_SEP_CHAR);
    name++;
    name_length++;
  }
  /* Error if empty or too long column name */
  return last_char_is_space || (uint32_t) name_length > NAME_CHAR_LEN;
}


/*****************************************************************************
  Functions to handle column usage bitmaps (read_set, write_set etc...)
*****************************************************************************/

/* Reset all columns bitmaps */

void Table::clear_column_bitmaps()
{
  /*
    Reset column read/write usage. It's identical to:
    bitmap_clear_all(&table->def_read_set);
    bitmap_clear_all(&table->def_write_set);
  */
  def_read_set.reset();
  def_write_set.reset();
  column_bitmaps_set(def_read_set, def_write_set);
}


/*
  Tell Cursor we are going to call position() and rnd_pos() later.

  NOTES:
  This is needed for handlers that uses the primary key to find the
  row. In this case we have to extend the read bitmap with the primary
  key fields.
*/

void Table::prepare_for_position()
{

  if ((cursor->getEngine()->check_flag(HTON_BIT_PRIMARY_KEY_IN_READ_INDEX)) &&
      getShare()->hasPrimaryKey())
  {
    mark_columns_used_by_index_no_reset(getShare()->getPrimaryKey());
  }
  return;
}


/*
  Mark that only fields from one key is used

  NOTE:
    This changes the bitmap to use the tmp bitmap
    After this, you can't access any other columns in the table until
    bitmaps are reset, for example with Table::clear_column_bitmaps()
    or Table::restore_column_maps_after_mark_index()
*/

void Table::mark_columns_used_by_index(uint32_t index)
{
  boost::dynamic_bitset<> *bitmap= &tmp_set;

  (void) cursor->extra(HA_EXTRA_KEYREAD);
  bitmap->reset();
  mark_columns_used_by_index_no_reset(index, *bitmap);
  column_bitmaps_set(*bitmap, *bitmap);
  return;
}


/*
  Restore to use normal column maps after key read

  NOTES
    This reverse the change done by mark_columns_used_by_index

  WARNING
    For this to work, one must have the normal table maps in place
    when calling mark_columns_used_by_index
*/

void Table::restore_column_maps_after_mark_index()
{

  key_read= 0;
  (void) cursor->extra(HA_EXTRA_NO_KEYREAD);
  default_column_bitmaps();
  return;
}


/*
  mark columns used by key, but don't reset other fields
*/

void Table::mark_columns_used_by_index_no_reset(uint32_t index)
{
    mark_columns_used_by_index_no_reset(index, *read_set);
}


void Table::mark_columns_used_by_index_no_reset(uint32_t index,
                                                boost::dynamic_bitset<>& bitmap)
{
  KeyPartInfo *key_part= key_info[index].key_part;
  KeyPartInfo *key_part_end= (key_part + key_info[index].key_parts);
  for (; key_part != key_part_end; key_part++)
  {
    if (! bitmap.empty())
      bitmap.set(key_part->fieldnr-1);
  }
}


/*
  Mark auto-increment fields as used fields in both read and write maps

  NOTES
    This is needed in insert & update as the auto-increment field is
    always set and sometimes read.
*/

void Table::mark_auto_increment_column()
{
  assert(found_next_number_field);
  /*
    We must set bit in read set as update_auto_increment() is using the
    store() to check overflow of auto_increment values
  */
  setReadSet(found_next_number_field->position());
  setWriteSet(found_next_number_field->position());
  if (getShare()->next_number_keypart)
    mark_columns_used_by_index_no_reset(getShare()->next_number_index);
}


/*
  Mark columns needed for doing an delete of a row

  DESCRIPTON
    Some table engines don't have a cursor on the retrieve rows
    so they need either to use the primary key or all columns to
    be able to delete a row.

    If the engine needs this, the function works as follows:
    - If primary key exits, mark the primary key columns to be read.
    - If not, mark all columns to be read

    If the engine has HA_REQUIRES_KEY_COLUMNS_FOR_DELETE, we will
    mark all key columns as 'to-be-read'. This allows the engine to
    loop over the given record to find all keys and doesn't have to
    retrieve the row again.
*/

void Table::mark_columns_needed_for_delete()
{
  /*
    If the Cursor has no cursor capabilites, or we have row-based
    replication active for the current statement, we have to read
    either the primary key, the hidden primary key or all columns to
    be able to do an delete

  */
  if (not getShare()->hasPrimaryKey())
  {
    /* fallback to use all columns in the table to identify row */
    use_all_columns();
    return;
  }
  else
    mark_columns_used_by_index_no_reset(getShare()->getPrimaryKey());

  /* If we the engine wants all predicates we mark all keys */
  if (cursor->getEngine()->check_flag(HTON_BIT_REQUIRES_KEY_COLUMNS_FOR_DELETE))
  {
    Field **reg_field;
    for (reg_field= field ; *reg_field ; reg_field++)
    {
      if ((*reg_field)->flags & PART_KEY_FLAG)
        setReadSet((*reg_field)->position());
    }
  }
}


/*
  Mark columns needed for doing an update of a row

  DESCRIPTON
    Some engines needs to have all columns in an update (to be able to
    build a complete row). If this is the case, we mark all not
    updated columns to be read.

    If this is no the case, we do like in the delete case and mark
    if neeed, either the primary key column or all columns to be read.
    (see mark_columns_needed_for_delete() for details)

    If the engine has HTON_BIT_REQUIRES_KEY_COLUMNS_FOR_DELETE, we will
    mark all USED key columns as 'to-be-read'. This allows the engine to
    loop over the given record to find all changed keys and doesn't have to
    retrieve the row again.
*/

void Table::mark_columns_needed_for_update()
{
  /*
    If the Cursor has no cursor capabilites, or we have row-based
    logging active for the current statement, we have to read either
    the primary key, the hidden primary key or all columns to be
    able to do an update
  */
  if (not getShare()->hasPrimaryKey())
  {
    /* fallback to use all columns in the table to identify row */
    use_all_columns();
    return;
  }
  else
    mark_columns_used_by_index_no_reset(getShare()->getPrimaryKey());

  if (cursor->getEngine()->check_flag(HTON_BIT_REQUIRES_KEY_COLUMNS_FOR_DELETE))
  {
    /* Mark all used key columns for read */
    Field **reg_field;
    for (reg_field= field ; *reg_field ; reg_field++)
    {
      /* Merge keys is all keys that had a column refered to in the query */
      if (is_overlapping(merge_keys, (*reg_field)->part_of_key))
        setReadSet((*reg_field)->position());
    }
  }

}


/*
  Mark columns the Cursor needs for doing an insert

  For now, this is used to mark fields used by the trigger
  as changed.
*/

void Table::mark_columns_needed_for_insert()
{
  if (found_next_number_field)
    mark_auto_increment_column();
}



size_t Table::max_row_length(const unsigned char *data)
{
  size_t length= getRecordLength() + 2 * sizeFields();
  uint32_t *const beg= getBlobField();
  uint32_t *const end= beg + sizeBlobFields();

  for (uint32_t *ptr= beg ; ptr != end ; ++ptr)
  {
    Field_blob* const blob= (Field_blob*) field[*ptr];
    length+= blob->get_length((const unsigned char*)
                              (data + blob->offset(getInsertRecord()))) +
      HA_KEY_BLOB_LENGTH;
  }
  return length;
}

void Table::setVariableWidth(void)
{
  assert(in_use);
  if (in_use && in_use->lex->sql_command == SQLCOM_CREATE_TABLE)
  {
    getMutableShare()->setVariableWidth();
    return;
  }

  assert(0); // Programming error, you can't set this on a plain old Table.
}

/****************************************************************************
 Functions for creating temporary tables.
****************************************************************************/
/**
  Create field for temporary table from given field.

  @param session	       Thread Cursor
  @param org_field    field from which new field will be created
  @param name         New field name
  @param table	       Temporary table
  @param item	       !=NULL if item->result_field should point to new field.
                      This is relevant for how fill_record() is going to work:
                      If item != NULL then fill_record() will update
                      the record in the original table.
                      If item == NULL then fill_record() will update
                      the temporary table
  @param convert_blob_length   If >0 create a varstring(convert_blob_length)
                               field instead of blob.

  @retval
    NULL		on error
  @retval
    new_created field
*/

Field *create_tmp_field_from_field(Session *session, Field *org_field,
                                   const char *name, Table *table,
                                   Item_field *item, uint32_t convert_blob_length)
{
  Field *new_field;

  /*
    Make sure that the blob fits into a Field_varstring which has
    2-byte lenght.
  */
  if (convert_blob_length && convert_blob_length <= Field_varstring::MAX_SIZE &&
      (org_field->flags & BLOB_FLAG))
  {
    table->setVariableWidth();
    new_field= new Field_varstring(convert_blob_length,
                                   org_field->maybe_null(),
                                   org_field->field_name,
                                   org_field->charset());
  }
  else
  {
    new_field= org_field->new_field(session->mem_root, table,
                                    table == org_field->getTable());
  }
  if (new_field)
  {
    new_field->init(table);
    new_field->orig_table= org_field->orig_table;
    if (item)
      item->result_field= new_field;
    else
      new_field->field_name= name;
    new_field->flags|= (org_field->flags & NO_DEFAULT_VALUE_FLAG);
    if (org_field->maybe_null() || (item && item->maybe_null))
      new_field->flags&= ~NOT_NULL_FLAG;	// Because of outer join
    if (org_field->type() == DRIZZLE_TYPE_VARCHAR)
      table->getMutableShare()->db_create_options|= HA_OPTION_PACK_RECORD;
    else if (org_field->type() == DRIZZLE_TYPE_DOUBLE)
      ((Field_double *) new_field)->not_fixed= true;
  }
  return new_field;
}


/**
  Create a temp table according to a field list.

  Given field pointers are changed to point at tmp_table for
  send_fields. The table object is self contained: it's
  allocated in its own memory root, as well as Field objects
  created for table columns.
  This function will replace Item_sum items in 'fields' list with
  corresponding Item_field items, pointing at the fields in the
  temporary table, unless this was prohibited by true
  value of argument save_sum_fields. The Item_field objects
  are created in Session memory root.

  @param session                  thread handle
  @param param                a description used as input to create the table
  @param fields               list of items that will be used to define
                              column types of the table (also see NOTES)
  @param group                TODO document
  @param distinct             should table rows be distinct
  @param save_sum_fields      see NOTES
  @param select_options
  @param rows_limit
  @param table_alias          possible name of the temporary table that can
                              be used for name resolving; can be "".
*/

#define STRING_TOTAL_LENGTH_TO_PACK_ROWS 128
#define AVG_STRING_LENGTH_TO_PACK_ROWS   64
#define RATIO_TO_PACK_ROWS	       2

Table *
create_tmp_table(Session *session,Tmp_Table_Param *param,List<Item> &fields,
		 Order *group, bool distinct, bool save_sum_fields,
		 uint64_t select_options, ha_rows rows_limit,
		 const char *table_alias)
{
  memory::Root *mem_root_save;
  uint	i,field_count,null_count,null_pack_length;
  uint32_t  copy_func_count= param->func_count;
  uint32_t  hidden_null_count, hidden_null_pack_length, hidden_field_count;
  uint32_t  blob_count,group_null_items, string_count;
  uint32_t fieldnr= 0;
  ulong reclength, string_total_length;
  bool  using_unique_constraint= false;
  bool  use_packed_rows= true;
  bool  not_all_columns= !(select_options & TMP_TABLE_ALL_COLUMNS);
  unsigned char	*pos, *group_buff;
  unsigned char *null_flags;
  Field **reg_field, **from_field, **default_field;
  CopyField *copy= 0;
  KeyInfo *keyinfo;
  KeyPartInfo *key_part_info;
  Item **copy_func;
  MI_COLUMNDEF *recinfo;
  uint32_t total_uneven_bit_length= 0;
  bool force_copy_fields= param->force_copy_fields;
  uint64_t max_rows= 0;

  session->status_var.created_tmp_tables++;

  if (group)
  {
    if (! param->quick_group)
    {
      group= 0;					// Can't use group key
    }
    else for (Order *tmp=group ; tmp ; tmp=tmp->next)
    {
      /*
        marker == 4 means two things:
        - store NULLs in the key, and
        - convert BIT fields to 64-bit long, needed because MEMORY tables
          can't index BIT fields.
      */
      (*tmp->item)->marker= 4;
      if ((*tmp->item)->max_length >= CONVERT_IF_BIGGER_TO_BLOB)
	using_unique_constraint= true;
    }
    if (param->group_length >= MAX_BLOB_WIDTH)
      using_unique_constraint= true;
    if (group)
      distinct= 0;				// Can't use distinct
  }

  field_count=param->field_count+param->func_count+param->sum_func_count;
  hidden_field_count=param->hidden_field_count;

  /*
    When loose index scan is employed as access method, it already
    computes all groups and the result of all aggregate functions. We
    make space for the items of the aggregate function in the list of
    functions Tmp_Table_Param::items_to_copy, so that the values of
    these items are stored in the temporary table.
  */
  if (param->precomputed_group_by)
  {
    copy_func_count+= param->sum_func_count;
  }

  table::Instance *table;
  table= session->getInstanceTable(); // This will not go into the tableshare cache, so no key is used.

  if (not table->getMemRoot()->multi_alloc_root(0,
                                                &default_field, sizeof(Field*) * (field_count),
                                                &from_field, sizeof(Field*)*field_count,
                                                &copy_func, sizeof(*copy_func)*(copy_func_count+1),
                                                &param->keyinfo, sizeof(*param->keyinfo),
                                                &key_part_info, sizeof(*key_part_info)*(param->group_parts+1),
                                                &param->start_recinfo, sizeof(*param->recinfo)*(field_count*2+4),
                                                &group_buff, (group && ! using_unique_constraint ?
                                                              param->group_length : 0),
                                                NULL))
  {
    return NULL;
  }
  /* CopyField belongs to Tmp_Table_Param, allocate it in Session mem_root */
  if (!(param->copy_field= copy= new (session->mem_root) CopyField[field_count]))
  {
    return NULL;
  }
  param->items_to_copy= copy_func;
  /* make table according to fields */

  memset(default_field, 0, sizeof(Field*) * (field_count));
  memset(from_field, 0, sizeof(Field*)*field_count);

  mem_root_save= session->mem_root;
  session->mem_root= table->getMemRoot();

  table->getMutableShare()->setFields(field_count+1);
  table->setFields(table->getMutableShare()->getFields(true));
  reg_field= table->getMutableShare()->getFields(true);
  table->setAlias(table_alias);
  table->reginfo.lock_type=TL_WRITE;	/* Will be updated */
  table->db_stat=HA_OPEN_KEYFILE+HA_OPEN_RNDFILE;
  table->map=1;
  table->copy_blobs= 1;
  assert(session);
  table->in_use= session;
  table->quick_keys.reset();
  table->covering_keys.reset();
  table->keys_in_use_for_query.reset();

  table->getMutableShare()->blob_field.resize(field_count+1);
  uint32_t *blob_field= &table->getMutableShare()->blob_field[0];
  table->getMutableShare()->blob_ptr_size= portable_sizeof_char_ptr;
  table->getMutableShare()->db_low_byte_first=1;                // True for HEAP and MyISAM
  table->getMutableShare()->table_charset= param->table_charset;
  table->getMutableShare()->keys_for_keyread.reset();
  table->getMutableShare()->keys_in_use.reset();

  /* Calculate which type of fields we will store in the temporary table */

  reclength= string_total_length= 0;
  blob_count= string_count= null_count= hidden_null_count= group_null_items= 0;
  param->using_indirect_summary_function= 0;

  List_iterator_fast<Item> li(fields);
  Item *item;
  Field **tmp_from_field=from_field;
  while ((item=li++))
  {
    Item::Type type=item->type();
    if (not_all_columns)
    {
      if (item->with_sum_func && type != Item::SUM_FUNC_ITEM)
      {
        if (item->used_tables() & OUTER_REF_TABLE_BIT)
          item->update_used_tables();
        if (type == Item::SUBSELECT_ITEM ||
            (item->used_tables() & ~OUTER_REF_TABLE_BIT))
        {
	  /*
	    Mark that the we have ignored an item that refers to a summary
	    function. We need to know this if someone is going to use
	    DISTINCT on the result.
	  */
	  param->using_indirect_summary_function=1;
	  continue;
        }
      }
      if (item->const_item() && (int) hidden_field_count <= 0)
        continue; // We don't have to store this
    }
    if (type == Item::SUM_FUNC_ITEM && !group && !save_sum_fields)
    {						/* Can't calc group yet */
      ((Item_sum*) item)->result_field= 0;
      for (i= 0 ; i < ((Item_sum*) item)->arg_count ; i++)
      {
	Item **argp= ((Item_sum*) item)->args + i;
	Item *arg= *argp;
	if (!arg->const_item())
	{
	  Field *new_field=
            create_tmp_field(session, table, arg, arg->type(), &copy_func,
                             tmp_from_field, &default_field[fieldnr],
                             group != 0,not_all_columns,
                             false,
                             param->convert_blob_length);
	  if (!new_field)
	    goto err;					// Should be OOM
	  tmp_from_field++;
	  reclength+=new_field->pack_length();
	  if (new_field->flags & BLOB_FLAG)
	  {
	    *blob_field++= fieldnr;
	    blob_count++;
	  }
	  *(reg_field++)= new_field;
          if (new_field->real_type() == DRIZZLE_TYPE_VARCHAR)
          {
            string_count++;
            string_total_length+= new_field->pack_length();
          }
          session->mem_root= mem_root_save;
          session->change_item_tree(argp, new Item_field(new_field));
          session->mem_root= table->getMemRoot();
	  if (!(new_field->flags & NOT_NULL_FLAG))
          {
	    null_count++;
            /*
              new_field->maybe_null() is still false, it will be
              changed below. But we have to setup Item_field correctly
            */
            (*argp)->maybe_null=1;
          }
          new_field->setPosition(fieldnr++);
	}
      }
    }
    else
    {
      /*
	The last parameter to create_tmp_field() is a bit tricky:

	We need to set it to 0 in union, to get fill_record() to modify the
	temporary table.
	We need to set it to 1 on multi-table-update and in select to
	write rows to the temporary table.
	We here distinguish between UNION and multi-table-updates by the fact
	that in the later case group is set to the row pointer.
      */
      Field *new_field=
        create_tmp_field(session, table, item, type, &copy_func,
                         tmp_from_field, &default_field[fieldnr],
                         group != 0,
                         !force_copy_fields &&
                           (not_all_columns || group != 0),
                         force_copy_fields,
                         param->convert_blob_length);

      if (!new_field)
      {
	if (session->is_fatal_error)
	  goto err;				// Got OOM
	continue;				// Some kindf of const item
      }
      if (type == Item::SUM_FUNC_ITEM)
	((Item_sum *) item)->result_field= new_field;
      tmp_from_field++;
      reclength+=new_field->pack_length();
      if (!(new_field->flags & NOT_NULL_FLAG))
	null_count++;
      if (new_field->flags & BLOB_FLAG)
      {
        *blob_field++= fieldnr;
	blob_count++;
      }
      if (item->marker == 4 && item->maybe_null)
      {
	group_null_items++;
	new_field->flags|= GROUP_FLAG;
      }
      new_field->setPosition(fieldnr++);
      *(reg_field++)= new_field;
    }
    if (!--hidden_field_count)
    {
      /*
        This was the last hidden field; Remember how many hidden fields could
        have null
      */
      hidden_null_count=null_count;
      /*
	We need to update hidden_field_count as we may have stored group
	functions with constant arguments
      */
      param->hidden_field_count= fieldnr;
      null_count= 0;
    }
  }
  assert(fieldnr == (uint32_t) (reg_field - table->getFields()));
  assert(field_count >= (uint32_t) (reg_field - table->getFields()));
  field_count= fieldnr;
  *reg_field= 0;
  *blob_field= 0;				// End marker
  table->getMutableShare()->fields= field_count;

  /* If result table is small; use a heap */
  /* future: storage engine selection can be made dynamic? */
  if (blob_count || using_unique_constraint || 
      (session->lex->select_lex.options & SELECT_BIG_RESULT) ||
      (session->lex->current_select->olap == ROLLUP_TYPE) ||
      (select_options & (OPTION_BIG_TABLES | SELECT_SMALL_RESULT)) == OPTION_BIG_TABLES)
  {
    table->getMutableShare()->storage_engine= myisam_engine;
    table->cursor= table->getMutableShare()->db_type()->getCursor(*table);
    if (group &&
	(param->group_parts > table->cursor->getEngine()->max_key_parts() ||
	 param->group_length > table->cursor->getEngine()->max_key_length()))
    {
      using_unique_constraint= true;
    }
  }
  else
  {
    table->getMutableShare()->storage_engine= heap_engine;
    table->cursor= table->getMutableShare()->db_type()->getCursor(*table);
  }
  if (! table->cursor)
    goto err;


  if (! using_unique_constraint)
    reclength+= group_null_items;	// null flag is stored separately

  table->getMutableShare()->blob_fields= blob_count;
  if (blob_count == 0)
  {
    /* We need to ensure that first byte is not 0 for the delete link */
    if (param->hidden_field_count)
      hidden_null_count++;
    else
      null_count++;
  }
  hidden_null_pack_length=(hidden_null_count+7)/8;
  null_pack_length= (hidden_null_pack_length +
                     (null_count + total_uneven_bit_length + 7) / 8);
  reclength+=null_pack_length;
  if (!reclength)
    reclength=1;				// Dummy select
  /* Use packed rows if there is blobs or a lot of space to gain */
  if (blob_count || ((string_total_length >= STRING_TOTAL_LENGTH_TO_PACK_ROWS) && (reclength / string_total_length <= RATIO_TO_PACK_ROWS || (string_total_length / string_count) >= AVG_STRING_LENGTH_TO_PACK_ROWS)))
    use_packed_rows= 1;

  table->getMutableShare()->setRecordLength(reclength);
  {
    uint32_t alloc_length=ALIGN_SIZE(reclength+MI_UNIQUE_HASH_LENGTH+1);
    table->getMutableShare()->rec_buff_length= alloc_length;
    if (!(table->record[0]= (unsigned char*) table->alloc_root(alloc_length*2)))
    {
      goto err;
    }
    table->record[1]= table->getInsertRecord()+alloc_length;
    table->getMutableShare()->resizeDefaultValues(alloc_length);
  }
  copy_func[0]= 0;				// End marker
  param->func_count= copy_func - param->items_to_copy;

  table->setup_tmp_table_column_bitmaps();

  recinfo=param->start_recinfo;
  null_flags=(unsigned char*) table->getInsertRecord();
  pos=table->getInsertRecord()+ null_pack_length;
  if (null_pack_length)
  {
    memset(recinfo, 0, sizeof(*recinfo));
    recinfo->type=FIELD_NORMAL;
    recinfo->length=null_pack_length;
    recinfo++;
    memset(null_flags, 255, null_pack_length);	// Set null fields

    table->null_flags= (unsigned char*) table->getInsertRecord();
    table->getMutableShare()->null_fields= null_count+ hidden_null_count;
    table->getMutableShare()->null_bytes= null_pack_length;
  }
  null_count= (blob_count == 0) ? 1 : 0;
  hidden_field_count=param->hidden_field_count;
  for (i= 0,reg_field= table->getFields(); i < field_count; i++,reg_field++,recinfo++)
  {
    Field *field= *reg_field;
    uint32_t length;
    memset(recinfo, 0, sizeof(*recinfo));

    if (!(field->flags & NOT_NULL_FLAG))
    {
      if (field->flags & GROUP_FLAG && !using_unique_constraint)
      {
	/*
	  We have to reserve one byte here for NULL bits,
	  as this is updated by 'end_update()'
	*/
	*pos++= '\0';				// Null is stored here
	recinfo->length= 1;
	recinfo->type=FIELD_NORMAL;
	recinfo++;
	memset(recinfo, 0, sizeof(*recinfo));
      }
      else
      {
	recinfo->null_bit= 1 << (null_count & 7);
	recinfo->null_pos= null_count/8;
      }
      field->move_field(pos,null_flags+null_count/8,
			1 << (null_count & 7));
      null_count++;
    }
    else
      field->move_field(pos,(unsigned char*) 0,0);
    field->reset();

    /*
      Test if there is a default field value. The test for ->ptr is to skip
      'offset' fields generated by initalize_tables
    */
    if (default_field[i] && default_field[i]->ptr)
    {
      /*
         default_field[i] is set only in the cases  when 'field' can
         inherit the default value that is defined for the field referred
         by the Item_field object from which 'field' has been created.
      */
      ptrdiff_t diff;
      Field *orig_field= default_field[i];
      /* Get the value from default_values */
      diff= (ptrdiff_t) (orig_field->getTable()->getDefaultValues() - orig_field->getTable()->getInsertRecord());
      orig_field->move_field_offset(diff);      // Points now at default_values
      if (orig_field->is_real_null())
        field->set_null();
      else
      {
        field->set_notnull();
        memcpy(field->ptr, orig_field->ptr, field->pack_length());
      }
      orig_field->move_field_offset(-diff);     // Back to getInsertRecord()
    }

    if (from_field[i])
    {						/* Not a table Item */
      copy->set(field,from_field[i],save_sum_fields);
      copy++;
    }
    length=field->pack_length();
    pos+= length;

    /* Make entry for create table */
    recinfo->length=length;
    if (field->flags & BLOB_FLAG)
      recinfo->type= (int) FIELD_BLOB;
    else
      recinfo->type=FIELD_NORMAL;
    if (!--hidden_field_count)
      null_count=(null_count+7) & ~7;		// move to next byte
  }

  param->copy_field_end=copy;
  param->recinfo=recinfo;
  table->storeRecordAsDefault();        // Make empty default record

  if (session->variables.tmp_table_size == ~ (uint64_t) 0)		// No limit
  {
    max_rows= ~(uint64_t) 0;
  }
  else
  {
    max_rows= (uint64_t) (((table->getMutableShare()->db_type() == heap_engine) ?
                           min(session->variables.tmp_table_size,
                               session->variables.max_heap_table_size) :
                           session->variables.tmp_table_size) /
                          table->getMutableShare()->getRecordLength());
  }

  set_if_bigger(max_rows, (uint64_t)1);	// For dummy start options
  /*
    Push the LIMIT clause to the temporary table creation, so that we
    materialize only up to 'rows_limit' records instead of all result records.
  */
  set_if_smaller(max_rows, rows_limit);

  table->getMutableShare()->setMaxRows(max_rows);

  param->end_write_records= rows_limit;

  keyinfo= param->keyinfo;

  if (group)
  {
    table->group=group;				/* Table is grouped by key */
    param->group_buff=group_buff;
    table->getMutableShare()->keys=1;
    table->getMutableShare()->uniques= test(using_unique_constraint);
    table->key_info=keyinfo;
    keyinfo->key_part=key_part_info;
    keyinfo->flags=HA_NOSAME;
    keyinfo->usable_key_parts=keyinfo->key_parts= param->group_parts;
    keyinfo->key_length= 0;
    keyinfo->rec_per_key= 0;
    keyinfo->algorithm= HA_KEY_ALG_UNDEF;
    keyinfo->name= (char*) "group_key";
    Order *cur_group= group;
    for (; cur_group ; cur_group= cur_group->next, key_part_info++)
    {
      Field *field=(*cur_group->item)->get_tmp_table_field();
      bool maybe_null=(*cur_group->item)->maybe_null;
      key_part_info->null_bit= 0;
      key_part_info->field=  field;
      key_part_info->offset= field->offset(table->getInsertRecord());
      key_part_info->length= (uint16_t) field->key_length();
      key_part_info->type=   (uint8_t) field->key_type();
      key_part_info->key_type= 
	((ha_base_keytype) key_part_info->type == HA_KEYTYPE_TEXT ||
	 (ha_base_keytype) key_part_info->type == HA_KEYTYPE_VARTEXT1 ||
	 (ha_base_keytype) key_part_info->type == HA_KEYTYPE_VARTEXT2) ?
	0 : 1;
      if (!using_unique_constraint)
      {
	cur_group->buff=(char*) group_buff;
	if (!(cur_group->field= field->new_key_field(session->mem_root,table,
                                                     group_buff +
                                                     test(maybe_null),
                                                     field->null_ptr,
                                                     field->null_bit)))
	  goto err;
	if (maybe_null)
	{
	  /*
	    To be able to group on NULL, we reserved place in group_buff
	    for the NULL flag just before the column. (see above).
	    The field data is after this flag.
	    The NULL flag is updated in 'end_update()' and 'end_write()'
	  */
	  keyinfo->flags|= HA_NULL_ARE_EQUAL;	// def. that NULL == NULL
	  key_part_info->null_bit=field->null_bit;
	  key_part_info->null_offset= (uint32_t) (field->null_ptr -
					      (unsigned char*) table->getInsertRecord());
          cur_group->buff++;                        // Pointer to field data
	  group_buff++;                         // Skipp null flag
	}
        /* In GROUP BY 'a' and 'a ' are equal for VARCHAR fields */
        key_part_info->key_part_flag|= HA_END_SPACE_ARE_EQUAL;
	group_buff+= cur_group->field->pack_length();
      }
      keyinfo->key_length+=  key_part_info->length;
    }
  }

  if (distinct && field_count != param->hidden_field_count)
  {
    /*
      Create an unique key or an unique constraint over all columns
      that should be in the result.  In the temporary table, there are
      'param->hidden_field_count' extra columns, whose null bits are stored
      in the first 'hidden_null_pack_length' bytes of the row.
    */
    if (blob_count)
    {
      /*
        Special mode for index creation in MyISAM used to support unique
        indexes on blobs with arbitrary length. Such indexes cannot be
        used for lookups.
      */
      table->getMutableShare()->uniques= 1;
    }
    null_pack_length-=hidden_null_pack_length;
    keyinfo->key_parts= ((field_count-param->hidden_field_count)+
			 (table->getMutableShare()->uniques ? test(null_pack_length) : 0));
    table->distinct= 1;
    table->getMutableShare()->keys= 1;
    if (!(key_part_info= (KeyPartInfo*)
         table->alloc_root(keyinfo->key_parts * sizeof(KeyPartInfo))))
      goto err;
    memset(key_part_info, 0, keyinfo->key_parts * sizeof(KeyPartInfo));
    table->key_info=keyinfo;
    keyinfo->key_part=key_part_info;
    keyinfo->flags=HA_NOSAME | HA_NULL_ARE_EQUAL;
    keyinfo->key_length=(uint16_t) reclength;
    keyinfo->name= (char*) "distinct_key";
    keyinfo->algorithm= HA_KEY_ALG_UNDEF;
    keyinfo->rec_per_key= 0;

    /*
      Create an extra field to hold NULL bits so that unique indexes on
      blobs can distinguish NULL from 0. This extra field is not needed
      when we do not use UNIQUE indexes for blobs.
    */
    if (null_pack_length && table->getMutableShare()->uniques)
    {
      key_part_info->null_bit= 0;
      key_part_info->offset=hidden_null_pack_length;
      key_part_info->length=null_pack_length;
      table->setVariableWidth();
      key_part_info->field= new Field_varstring(table->getInsertRecord(),
                                                (uint32_t) key_part_info->length,
                                                0,
                                                (unsigned char*) 0,
                                                (uint32_t) 0,
                                                NULL,
                                                &my_charset_bin);
      if (!key_part_info->field)
        goto err;
      key_part_info->field->init(table);
      key_part_info->key_type= 1; /* binary comparison */
      key_part_info->type=    HA_KEYTYPE_BINARY;
      key_part_info++;
    }
    /* Create a distinct key over the columns we are going to return */
    for (i=param->hidden_field_count, reg_field=table->getFields() + i ;
	 i < field_count;
	 i++, reg_field++, key_part_info++)
    {
      key_part_info->null_bit= 0;
      key_part_info->field=    *reg_field;
      key_part_info->offset=   (*reg_field)->offset(table->getInsertRecord());
      key_part_info->length=   (uint16_t) (*reg_field)->pack_length();
      /* @todo The below method of computing the key format length of the
        key part is a copy/paste from optimizer/range.cc, and table.cc.
        This should be factored out, e.g. as a method of Field.
        In addition it is not clear if any of the Field::*_length
        methods is supposed to compute the same length. If so, it
        might be reused.
      */
      key_part_info->store_length= key_part_info->length;

      if ((*reg_field)->real_maybe_null())
        key_part_info->store_length+= HA_KEY_NULL_LENGTH;
      if ((*reg_field)->type() == DRIZZLE_TYPE_BLOB ||
          (*reg_field)->real_type() == DRIZZLE_TYPE_VARCHAR)
        key_part_info->store_length+= HA_KEY_BLOB_LENGTH;

      key_part_info->type=     (uint8_t) (*reg_field)->key_type();
      key_part_info->key_type =
	((ha_base_keytype) key_part_info->type == HA_KEYTYPE_TEXT ||
	 (ha_base_keytype) key_part_info->type == HA_KEYTYPE_VARTEXT1 ||
	 (ha_base_keytype) key_part_info->type == HA_KEYTYPE_VARTEXT2) ?
	0 : 1;
    }
  }

  if (session->is_fatal_error)				// If end of memory
    goto err;
  table->getMutableShare()->db_record_offset= 1;
  if (table->getShare()->db_type() == myisam_engine)
  {
    if (table->create_myisam_tmp_table(param->keyinfo, param->start_recinfo,
				       &param->recinfo, select_options))
      goto err;
  }
  assert(table->in_use);
  if (table->open_tmp_table())
    goto err;

  session->mem_root= mem_root_save;

  return(table);

err:
  session->mem_root= mem_root_save;
  table= NULL;

  return NULL;
}

/****************************************************************************/

void Table::column_bitmaps_set(boost::dynamic_bitset<>& read_set_arg,
                               boost::dynamic_bitset<>& write_set_arg)
{
  read_set= &read_set_arg;
  write_set= &write_set_arg;
}


const boost::dynamic_bitset<> Table::use_all_columns(boost::dynamic_bitset<>& in_map)
{
  const boost::dynamic_bitset<> old= in_map;
  in_map= getShare()->all_set;
  return old;
}

void Table::restore_column_map(const boost::dynamic_bitset<>& old)
{
  for (boost::dynamic_bitset<>::size_type i= 0; i < old.size(); i++)
  {
    if (old.test(i))
    {
      read_set->set(i);
    }
    else
    {
      read_set->reset(i);
    }
  }
}

uint32_t Table::find_shortest_key(const key_map *usable_keys)
{
  uint32_t min_length= UINT32_MAX;
  uint32_t best= MAX_KEY;
  if (usable_keys->any())
  {
    for (uint32_t nr= 0; nr < getShare()->sizeKeys() ; nr++)
    {
      if (usable_keys->test(nr))
      {
        if (key_info[nr].key_length < min_length)
        {
          min_length= key_info[nr].key_length;
          best=nr;
        }
      }
    }
  }
  return best;
}

/*****************************************************************************
  Remove duplicates from tmp table
  This should be recoded to add a unique index to the table and remove
  duplicates
  Table is a locked single thread table
  fields is the number of fields to check (from the end)
*****************************************************************************/

bool Table::compare_record(Field **ptr)
{
  for (; *ptr ; ptr++)
  {
    if ((*ptr)->cmp_offset(getShare()->rec_buff_length))
      return true;
  }
  return false;
}

/**
   True if the table's input and output record buffers are comparable using
   compare_records(TABLE*).
 */
bool Table::records_are_comparable()
{
  return ((getEngine()->check_flag(HTON_BIT_PARTIAL_COLUMN_READ) == 0) ||
          write_set->is_subset_of(*read_set));
}

/**
   Compares the input and outbut record buffers of the table to see if a row
   has changed. The algorithm iterates over updated columns and if they are
   nullable compares NULL bits in the buffer before comparing actual
   data. Special care must be taken to compare only the relevant NULL bits and
   mask out all others as they may be undefined. The storage engine will not
   and should not touch them.

   @param table The table to evaluate.

   @return true if row has changed.
   @return false otherwise.
*/
bool Table::compare_records()
{
  if (getEngine()->check_flag(HTON_BIT_PARTIAL_COLUMN_READ) != 0)
  {
    /*
      Storage engine may not have read all columns of the record.  Fields
      (including NULL bits) not in the write_set may not have been read and
      can therefore not be compared.
    */
    for (Field **ptr= this->field ; *ptr != NULL; ptr++)
    {
      Field *f= *ptr;
      if (write_set->test(f->position()))
      {
        if (f->real_maybe_null())
        {
          unsigned char null_byte_index= f->null_ptr - record[0];

          if (((record[0][null_byte_index]) & f->null_bit) !=
              ((record[1][null_byte_index]) & f->null_bit))
            return true;
        }
        if (f->cmp_binary_offset(getShare()->rec_buff_length))
          return true;
      }
    }
    return false;
  }

  /*
    The storage engine has read all columns, so it's safe to compare all bits
    including those not in the write_set. This is cheaper than the
    field-by-field comparison done above.
  */
  if (not getShare()->blob_fields + getShare()->hasVariableWidth())
    // Fixed-size record: do bitwise comparison of the records
    return memcmp(this->getInsertRecord(), this->getUpdateRecord(), (size_t) getShare()->getRecordLength());

  /* Compare null bits */
  if (memcmp(null_flags, null_flags + getShare()->rec_buff_length, getShare()->null_bytes))
    return true; /* Diff in NULL value */

  /* Compare updated fields */
  for (Field **ptr= field ; *ptr ; ptr++)
  {
    if (isWriteSet((*ptr)->position()) &&
	(*ptr)->cmp_binary_offset(getShare()->rec_buff_length))
      return true;
  }
  return false;
}

/*
 * Store a record from previous record into next
 *
 */
void Table::storeRecord()
{
  memcpy(getUpdateRecord(), getInsertRecord(), (size_t) getShare()->getRecordLength());
}

/*
 * Store a record as an insert
 *
 */
void Table::storeRecordAsInsert()
{
  assert(insert_values.size() >= getShare()->getRecordLength());
  memcpy(&insert_values[0], getInsertRecord(), (size_t) getShare()->getRecordLength());
}

/*
 * Store a record with default values
 *
 */
void Table::storeRecordAsDefault()
{
  memcpy(getMutableShare()->getDefaultValues(), getInsertRecord(), (size_t) getShare()->getRecordLength());
}

/*
 * Restore a record from previous record into next
 *
 */
void Table::restoreRecord()
{
  memcpy(getInsertRecord(), getUpdateRecord(), (size_t) getShare()->getRecordLength());
}

/*
 * Restore a record with default values
 *
 */
void Table::restoreRecordAsDefault()
{
  memcpy(getInsertRecord(), getMutableShare()->getDefaultValues(), (size_t) getShare()->getRecordLength());
}

/*
 * Empty a record
 *
 */
void Table::emptyRecord()
{
  restoreRecordAsDefault();
  memset(null_flags, 255, getShare()->null_bytes);
}

Table::Table() : 
  field(NULL),
  cursor(NULL),
  next(NULL),
  prev(NULL),
  read_set(NULL),
  write_set(NULL),
  tablenr(0),
  db_stat(0),
  def_read_set(),
  def_write_set(),
  tmp_set(),
  in_use(NULL),
  key_info(NULL),
  next_number_field(NULL),
  found_next_number_field(NULL),
  timestamp_field(NULL),
  pos_in_table_list(NULL),
  group(NULL),
  null_flags(NULL),
  lock_position(0),
  lock_data_start(0),
  lock_count(0),
  used_fields(0),
  status(0),
  derived_select_number(0),
  current_lock(F_UNLCK),
  copy_blobs(false),
  maybe_null(false),
  null_row(false),
  force_index(false),
  distinct(false),
  const_table(false),
  no_rows(false),
  key_read(false),
  no_keyread(false),
  open_placeholder(false),
  locked_by_name(false),
  no_cache(false),
  auto_increment_field_not_null(false),
  alias_name_used(false),
  query_id(0),
  quick_condition_rows(0),
  timestamp_field_type(TIMESTAMP_NO_AUTO_SET),
  map(0)
{
  record[0]= (unsigned char *) 0;
  record[1]= (unsigned char *) 0;

  reginfo.reset();
  covering_keys.reset();
  quick_keys.reset();
  merge_keys.reset();

  keys_in_use_for_query.reset();
  keys_in_use_for_group_by.reset();
  keys_in_use_for_order_by.reset();

  memset(quick_rows, 0, sizeof(ha_rows) * MAX_KEY);
  memset(const_key_parts, 0, sizeof(ha_rows) * MAX_KEY);

  memset(quick_key_parts, 0, sizeof(unsigned int) * MAX_KEY);
  memset(quick_n_ranges, 0, sizeof(unsigned int) * MAX_KEY);
}

/*****************************************************************************
  The different ways to read a record
  Returns -1 if row was not found, 0 if row was found and 1 on errors
*****************************************************************************/

/** Help function when we get some an error from the table Cursor. */

int Table::report_error(int error)
{
  if (error == HA_ERR_END_OF_FILE || error == HA_ERR_KEY_NOT_FOUND)
  {
    status= STATUS_GARBAGE;
    return -1;					// key not found; ok
  }
  /*
    Locking reads can legally return also these errors, do not
    print them to the .err log
  */
  if (error != HA_ERR_LOCK_DEADLOCK && error != HA_ERR_LOCK_WAIT_TIMEOUT)
    errmsg_printf(ERRMSG_LVL_ERROR, _("Got error %d when reading table '%s'"),
                  error, getShare()->getPath());
  print_error(error, MYF(0));

  return 1;
}


void Table::setup_table_map(TableList *table_list, uint32_t table_number)
{
  used_fields= 0;
  const_table= 0;
  null_row= 0;
  status= STATUS_NO_RECORD;
  maybe_null= table_list->outer_join;
  TableList *embedding= table_list->getEmbedding();
  while (!maybe_null && embedding)
  {
    maybe_null= embedding->outer_join;
    embedding= embedding->getEmbedding();
  }
  tablenr= table_number;
  map= (table_map) 1 << table_number;
  force_index= table_list->force_index;
  covering_keys= getShare()->keys_for_keyread;
  merge_keys.reset();
}


bool Table::fill_item_list(List<Item> *item_list) const
{
  /*
    All Item_field's created using a direct pointer to a field
    are fixed in Item_field constructor.
  */
  for (Field **ptr= field; *ptr; ptr++)
  {
    Item_field *item= new Item_field(*ptr);
    if (!item || item_list->push_back(item))
      return true;
  }
  return false;
}


void Table::filesort_free_buffers(bool full)
{
  if (sort.record_pointers)
  {
    free((unsigned char*) sort.record_pointers);
    sort.record_pointers=0;
  }
  if (full)
  {
    if (sort.sort_keys )
    {
      if ((unsigned char*) sort.sort_keys)
        free((unsigned char*) sort.sort_keys);
      sort.sort_keys= 0;
    }
    if (sort.buffpek)
    {
      if ((unsigned char*) sort.buffpek)
        free((unsigned char*) sort.buffpek);
      sort.buffpek= 0;
      sort.buffpek_len= 0;
    }
  }

  if (sort.addon_buf)
  {
    free((char *) sort.addon_buf);
    free((char *) sort.addon_field);
    sort.addon_buf=0;
    sort.addon_field=0;
  }
}

} /* namespace drizzled */