aboutsummaryrefslogtreecommitdiffstats
path: root/activerecord/CHANGELOG.md
blob: f73fc9d9a3a4e98f1b4123fb605a0aeaf0ae96f0 (plain) (blame)
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
## Rails 4.0.0 (unreleased) ##

*   Fix quoting for sqlite migrations using copy_table_contents() with binary
    columns.

    These would fail with "SQLite3::SQLException: unrecognized token" because
    the column was not being passed to quote() so the data was not quoted
    correctly.

    *Matthew M. Boedicker*

*   Promotes `change_column_null` to the migrations API. This macro sets/removes
    `NOT NULL` constraints, and accepts an optional argument to replace existing
    `NULL`s if needed. The adapters for SQLite, MySQL, PostgreSQL, and (at least)
    Oracle, already implement this method.

    *Xavier Noria*

*   Uniqueness validation allows you to pass `:conditions` to limit
    the constraint lookup.

    Example:

        validates_uniqueness_of :title, conditions: -> { where('approved = ?', true) }

    *Mattias Pfeiffer + Yves Senn*

*   `connection` is deprecated as an instance method.
    This allows end-users to have a `connection` method on their models
    without clashing with Active Record internals.

    *Ben Moss*

*   When copying migrations, preserve their magic comments and content encoding.

    *OZAWA Sakuro*

*   Fix `subclass_from_attrs` when `eager_load` is false. It cannot find
    subclass because all classes are loaded automatically when it needs.

    *Dmitry Vorotilin*

*   When `:name` option is provided to `remove_index`, use it if there is no
    index by the conventional name.

    For example, previously if an index was removed like so
    `remove_index :values, column: :value, name: 'a_different_name'`
    the generated SQL would not contain the specified index name,
    and hence the migration would fail.
    Fixes #8858.

    *Ezekiel Smithburg*

*   Created block to by-pass the prepared statement bindings.
    This will allow to compose fragments of large SQL statements to
    avoid multiple round-trips between Ruby and the DB.

    Example:

        sql = Post.connection.unprepared_statement do
          Post.first.comments.to_sql
        end

    *Cédric Fabianski*

*   Change the semantics of combining scopes to be the same as combining
    class methods which return scopes. For example:

        class User < ActiveRecord::Base
          scope :active,   -> { where state: 'active' }
          scope :inactive, -> { where state: 'inactive' }
        end

        class Post < ActiveRecord::Base
          def self.active
            where state: 'active'
          end

          def self.inactive
            where state: 'inactive'
          end
        end

        ### BEFORE ###

        User.where(state: 'active').where(state: 'inactive')
        # => SELECT * FROM users WHERE state = 'active' AND state = 'inactive'

        User.active.inactive
        # => SELECT * FROM users WHERE state = 'inactive'

        Post.active.inactive
        # => SELECT * FROM posts WHERE state = 'active' AND state = 'inactive'

        ### AFTER ###

        User.active.inactive
        # => SELECT * FROM posts WHERE state = 'active' AND state = 'inactive'

    Before this change, invoking a scope would merge it into the current
    scope and return the result. `Relation#merge` applies "last where
    wins" logic to de-duplicate the conditions, but this lead to
    confusing and inconsistent behaviour. This fixes that.

    If you really do want the "last where wins" logic, you can opt-in to
    it like so:

        User.active.merge(User.inactive)

    Fixes #7365.

    *Neeraj Singh* and *Jon Leighton*

*   Expand `#cache_key` to consult all relevant updated timestamps.

    Previously only `updated_at` column was checked, now it will
    consult other columns that received updated timestamps on save,
    such as `updated_on`.  When multiple columns are present it will
    use the most recent timestamp.
    Fixes #9033.

    *Brendon Murphy*

*   Throw `NotImplementedError` when trying to instantiate `ActiveRecord::Base` or an abstract class.

    *Aaron Weiner*

*   Warn when `rake db:structure:dump` with a MySQL database and
    `mysqldump` is not in the PATH or fails.
    Fixes #9518.

    *Yves Senn*

*   Remove `connection#structure_dump`, which is no longer used. *Yves Senn*

*   Make it possible to execute migrations without a transaction even
    if the database adapter supports DDL transactions.
    Fixes #9483.

    Example:

        class ChangeEnum < ActiveRecord::Migration
          disable_ddl_transaction!

          def up
            execute "ALTER TYPE model_size ADD VALUE 'new_value'"
          end
        end

    *Yves Senn*

*   Assigning "0.0" to a nullable numeric column does not make it dirty.
    Fixes #9034.

    Example:

        product = Product.create price: 0.0
        product.price = '0.0'
        product.changed? # => false (this used to return true)
        product.changes # => {} (this used to return { price: [0.0, 0.0] })

    *Yves Senn*

*   Added functionality to unscope relations in a relations chain. For
    instance, if you are passed in a chain of relations as follows:

        User.where(name: "John").order('id DESC')

    but you want to get rid of order, then this feature allows you to do:

        User.where(name: "John").order('id DESC').unscope(:order)
            == User.where(name: "John")

    The .unscope() function is more general than the .except() method because
    .except() only works on the relation it is acting on. However, .unscope()
    works for any relation in the entire relation chain.

    *John Wang*

*   Postgresql timestamp with time zone (timestamptz) datatype now returns a
    ActiveSupport::TimeWithZone instance instead of a string

    *Troy Kruthoff*

*   The `#append` method for collection associations behaves like`<<`.
    `#prepend` is not defined and `<<` or `#append` should be used.
    Fixes #7364.

    *Yves Senn*

*   Added support for creating a table via Rails migration generator.
    For example,

        rails g migration create_books title:string content:text

    will generate a migration that creates a table called books with
    the listed attributes, without creating a model.

    *Sammy Larbi*

*   Fix bug that raises the wrong exception when the exception handled by PostgreSQL adapter
    doesn't respond to `#result`.
    Fixes #8617.

    *kennyj*

*   Support PostgreSQL specific column types when using `change_table`.
    Fixes #9480.

    Example:

        change_table :authors do |t|
          t.hstore :books
          t.json :metadata
        end

    *Yves Senn*

*   Revert 408227d9c5ed7d, 'quote numeric'. This introduced some regressions.

    *Steve Klabnik*

*   Fix calculation of `db_runtime` property in
   `ActiveRecord::Railties::ControllerRuntime#cleanup_view_runtime`.
    Previously, after raising `ActionView::MissingTemplate`, `db_runtime` was
    not populated.
    Fixes #9215.

    *Igor Fedoronchuk*

*   Do not try to touch invalid (and thus not persisted) parent record
    for a `belongs_to :parent, touch: true` association

    *Olek Janiszewski*

*   Fix when performing an ordered join query. The bug only
    affected queries where the order was given with a symbol.
    Fixes #9275.

    Example:

        # This will expand the order :name to "authors".name.
        Author.joins(:books).where('books.published = 1').order(:name)


## Rails 4.0.0.beta1 (February 25, 2013) ##

*   Fix overriding of attributes by `default_scope` on `ActiveRecord::Base#dup`.

    *Hiroshige UMINO*

*   Update queries now use prepared statements.

    *Olli Rissanen*

*   Fixing issue #8345. Now throwing an error when one attempts to touch a
    new object that has not yet been persisted. For instance:

    Example:

        ball = Ball.new
        ball.touch :updated_at   # => raises error

    It is not until the ball object has been persisted that it can be touched.
    This follows the behavior of update_column.

    *John Wang*

*   Preloading ordered `has_many :through` associations no longer applies
    invalid ordering to the `:through` association.
    Fixes #8663.

    *Yves Senn*

*   The auto explain feature has been removed. This feature was
    activated by configuring `config.active_record.auto_explain_threshold_in_seconds`.
    The configuration option was deprecated and has no more effect.

    You can still use `ActiveRecord::Relation#explain` to see the EXPLAIN output for
    any given relation.

    *Yves Senn*

*   The `:on` option for `after_commit` and `after_rollback` now
    accepts an Array of actions.
    Fixes #988.

    Example:

        after_commit :update_cache on: [:create, :update]

    *Yves Senn*

*   Rename related indexes on `rename_table` and `rename_column`. This
    does not affect indexes with custom names.

    *Yves Senn*

*   Prevent the creation of indices with too long names, which cause
    internal operations to fail (sqlite3 adapter only). The method
    `allowed_index_name_length` defines the length limit enforced by
    rails. It's value defaults to `index_name_length` but can vary per adapter.
    Fixes #8264.

    *Yves Senn*

*   Fixing issue #776.

    Memory bloat in transactions is handled by having the transaction hold only
    the AR objects which it absolutely needs to know about. These are the AR
    objects with callbacks (they need to be updated as soon as something in the
    transaction occurs).

    All other AR objects can be updated lazily by keeping a reference to a
    TransactionState object. If an AR object gets inside a transaction, then
    the transaction will add its TransactionState to the AR object. When the
    user makes a call to some attribute on an AR object (which has no
    callbacks) associated with a transaction, the AR object will call the
    sync_with_transaction_state method and make sure it is up to date with the
    transaction. After it has synced with the transaction state, the AR object
    will return the attribute that was requested.

    Most of the logic in the changes are used to handle multiple transactions,
    in which case the AR object has to recursively follow parent pointers of
    TransactionState objects.

    *John Wang*

*   Descriptive error message when the necessary AR adapter gem was not found.
    Fixes #7313.

    *Yves Senn*

*   Active Record now raises an error when blank arguments are passed to query
    methods for which blank arguments do not make sense.

    Example:

        Post.includes()     # => raises error

    *John Wang*

*   Simplified type casting code for timezone aware attributes to use the
    `in_time_zone` method if it is available. This introduces a subtle change
    of behavior when using `Date` instances as they are directly converted to
    `ActiveSupport::TimeWithZone` instances without first being converted to
    `Time` instances. For example:

        # Rails 3.2 behavior
        >> Date.today.to_time.in_time_zone
        => Wed, 13 Feb 2013 07:00:00 UTC +00:00

        # Rails 4.0 behavior
        >> Date.today.in_time_zone
        => Wed, 13 Feb 2013 00:00:00 UTC +00:00

    On the plus side it now behaves the same whether you pass a `String` date
    or an actual `Date` instance. For example:

        # Rails 3.2 behavior
        >> Date.civil(2013, 2, 13).to_time.in_time_zone
        => Wed, 13 Feb 2013 07:00:00 UTC +00:00
        >> Time.zone.parse("2013-02-13")
        => Wed, 13 Feb 2013 00:00:00 UTC +00:00

        # Rails 4.0 behavior
        >> Date.civil(2013, 2, 13).in_time_zone
        => Wed, 13 Feb 2013 00:00:00 UTC +00:00
        >> "2013-02-13".in_time_zone
        => Wed, 13 Feb 2013 00:00:00 UTC +00:00

    If you need the old behavior you can convert the dates to times manually.
    For example:

        >> Post.new(created_at: Date.today).created_at
        => Wed, 13 Feb 2013 00:00:00 UTC +00:00

        >> Post.new(created_at: Date.today.to_time).created_at
        => Wed, 13 Feb 2013 07:00:00 UTC +00:00

    *Andrew White*

*   Preloading `has_many :through` associations with conditions won't
    cache the `:through` association. This will prevent invalid
    subsets to be cached.
    Fixes #8423.

    Example:

        class User
          has_many :posts
          has_many :recent_comments, -> { where('created_at > ?', 1.week.ago) }, :through => :posts
        end

        a_user = User.includes(:recent_comments).first

        # This is preloaded.
        a_user.recent_comments

        # This is not preloaded, fetched now.
        a_user.posts

    *Yves Senn*

*   Don't run `after_commit` callbacks when creating through an association
    if saving the record fails.

    *James Miller*

*   Allow store accessors to be overrided like other attribute methods, e.g.:

        class User < ActiveRecord::Base
          store :settings, accessors: [ :color, :homepage ], coder: JSON

          def color
            super || 'red'
          end
        end

    *Sergey Nartimov*

*   Quote numeric values being compared to non-numeric columns. Otherwise,
    in some database, the string column values will be coerced to a numeric
    allowing 0, 0.0 or false to match any string starting with a non-digit.

    Example:

        App.where(apikey: 0) # => SELECT * FROM users WHERE apikey = '0'

    *Dylan Smith*

*   Schema dumper supports dumping the enabled database extensions to `schema.rb`
    (currently only supported by postgresql).

    *Justin George*

*   The database adpters now converts the options passed thought `DATABASE_URL`
    environment variable to the proper Ruby types before using. For example, SQLite requires
    that the timeout value is an integer, and PostgreSQL requires that the
    prepared_statements option is a boolean. These now work as expected:

    Example:

        DATABASE_URL=sqlite3://localhost/test_db?timeout=500
        DATABASE_URL=postgresql://localhost/test_db?prepared_statements=false

    *Aaron Stone + Rafael Mendonça França*

*   `Relation#merge` now only overwrites where values on the LHS of the
    merge. Consider:

        left  = Person.where(age: [13, 14, 15])
        right = Person.where(age: [13, 14]).where(age: [14, 15])

    `left` results in the following SQL:

        WHERE age IN (13, 14, 15)

    `right` results in the following SQL:

        WHERE age IN (13, 14) AND age IN (14, 15)

    Previously, `left.merge(right)` would result in all but the last
    condition being removed:

        WHERE age IN (14, 15)

    Now it results in the LHS condition(s) for `age` being removed, but
    the RHS remains as it is:

        WHERE age IN (13, 14) AND age IN (14, 15)

    *Jon Leighton*

*   Fix handling of dirty time zone aware attributes

    Previously, when `time_zone_aware_attributes` were enabled, after
    changing a datetime or timestamp attribute and then changing it back
    to the original value, `changed_attributes` still tracked the
    attribute as changed. This caused `[attribute]_changed?` and
    `changed?` methods to return true incorrectly.

    Example:

        in_time_zone 'Paris' do
          order = Order.new
          original_time = Time.local(2012, 10, 10)
          order.shipped_at = original_time
          order.save
          order.changed? # => false

          # changing value
          order.shipped_at = Time.local(2013, 1, 1)
          order.changed? # => true

          # reverting to original value
          order.shipped_at = original_time
          order.changed? # => false, used to return true
        end

    *Lilibeth De La Cruz*

*   When `#count` is used in conjunction with `#uniq` we perform `count(:distinct => true)`.
    Fixes #6865.

    Example:

        relation.uniq.count # => SELECT COUNT(DISTINCT *)

    *Yves Senn + Kaspar Schiess*

*   PostgreSQL ranges type support. Includes: int4range, int8range,
    numrange, tsrange, tstzrange, daterange

    Ranges can be created with inclusive and exclusive bounds.

    Example:

        create_table :Room do |t|
          t.daterange :availability
        end

        Room.create(availability: (Date.today..Float::INFINITY))
        Room.first.availability # => Wed, 19 Sep 2012..Infinity

    One thing to note: Range class does not support exclusive lower
    bound.

    *Alexander Grebennik*

*   Added a state instance variable to each transaction. Will allow other objects
    to know whether a transaction has been committed or rolled back.

    *John Wang*

*   Collection associations `#empty?` always respects builded records.
    Fixes #8879.

    Example:

        widget = Widget.new
        widget.things.build
        widget.things.empty? # => false

    *Yves Senn*

*   Support for PostgreSQL's `ltree` data type.

    *Rob Worley*

*   Fix undefined method `to_i` when calling `new` on a scope that uses an
    Array; Fix FloatDomainError when setting integer column to NaN.
    Fixes #8718, #8734, #8757.

    *Jason Stirk + Tristan Harward*

*   Rename `update_attributes` to `update`, keep `update_attributes` as an alias for `update` method.
    This is a soft-deprecation for `update_attributes`, although it will still work without any
    deprecation message in 4.0 is recommended to start using `update` since `update_attributes` will be
    deprecated and removed in future versions of Rails.

    *Amparo Luna + Guillermo Iguaran*

*   `after_commit` and `after_rollback` now validate the `:on` option and raise an `ArgumentError`
    if it is not one of `:create`, `:destroy` or `:update`

    *Pascal Friederich*

*   Improve ways to write `change` migrations, making the old `up` & `down` methods no longer necessary.

    * The methods `drop_table` and `remove_column` are now reversible, as long as the necessary information is given.
      The method `remove_column` used to accept multiple column names; instead use `remove_columns` (which is not reversible).
      The method `change_table` is also reversible, as long as its block doesn't call `remove`, `change` or `change_default`

    * New method `reversible` makes it possible to specify code to be run when migrating up or down.
      See the [Guide on Migration](https://github.com/rails/rails/blob/master/guides/source/migrations.md#using-the-reversible-method)

    * New method `revert` will revert a whole migration or the given block.
      If migrating down, the given migration / block is run normally.
      See the [Guide on Migration](https://github.com/rails/rails/blob/master/guides/source/migrations.md#reverting-previous-migrations)

    Attempting to revert the methods `execute`, `remove_columns` and `change_column` will now
    raise an `IrreversibleMigration` instead of actually executing them without any output.

    *Marc-André Lafortune*

*   Serialized attributes can be serialized in integer columns.
    Fixes #8575.

    *Rafael Mendonça França*

*   Keep index names when using `alter_table` with sqlite3.
    Fixes #3489.

    *Yves Senn*

*   Add ability for postgresql adapter to disable user triggers in `disable_referential_integrity`.
    Fixes #5523.

    *Gary S. Weaver*

*   Added support for `validates_uniqueness_of` in PostgreSQL array columns.
    Fixes #8075.

    *Pedro Padron*

*   Allow int4range and int8range columns to be created in PostgreSQL and properly convert to/from database.

    *Alexey Vasiliev aka leopard*

*   Do not log the binding values for binary columns.

    *Matthew M. Boedicker*

*   Fix counter cache columns not updated when replacing `has_many :through`
    associations.

    *Matthew Robertson*

*   Recognize migrations placed in directories containing numbers and 'rb'.
    Fixes #8492.

    *Yves Senn*

*   Add `ActiveRecord::Base.cache_timestamp_format` class attribute to control
    the format of the timestamp value in the cache key. Defaults to `:nsec`.
    Fixes #8195.

    *Rafael Mendonça França*

*   Session variables can be set for the `mysql`, `mysql2`, and `postgresql` adapters
    in the `variables: <hash>` parameter in `config/database.yml`. The key-value pairs of this
    hash will be sent in a `SET key = value` query on new database connections. See also:
    http://dev.mysql.com/doc/refman/5.0/en/set-statement.html
    http://www.postgresql.org/docs/8.3/static/sql-set.html

    *Aaron Stone*

*   Allow `Relation#where` with no arguments to be chained with new `not` query method.

    Example:

        Developer.where.not(name: 'Aaron')

    *Akira Matsuda*

*   Unscope `update_column(s)` query to ignore default scope.

    When applying `default_scope` to a class with a where clause, using
    `update_column(s)` could generate a query that would not properly update
    the record due to the where clause from the `default_scope` being applied
    to the update query.

        class User < ActiveRecord::Base
          default_scope -> { where(active: true) }
        end

        user = User.first
        user.active = false
        user.save!

        user.update_column(:active, true) # => false

    In this situation we want to skip the default_scope clause and just
    update the record based on the primary key. With this change:

        user.update_column(:active, true) # => true

    Fixes #8436.

    *Carlos Antonio da Silva*

*   SQLite adapter no longer corrupts binary data if the data contains `%00`.

    *Chris Feist*

*   Fix performance problem with `primary_key` method in PostgreSQL adapter when having many schemas.
    Uses `pg_constraint` table instead of `pg_depend` table which has many records in general.
    Fixes #8414.

    *kennyj*

*   Do not instantiate intermediate Active Record objects when eager loading.
    These records caused `after_find` to run more than expected.
    Fixes #3313.

    *Yves Senn*

*   Add STI support to init and building associations.
    Allows you to do `BaseClass.new(type: "SubClass")` as well as
    `parent.children.build(type: "SubClass")` or `parent.build_child`
    to initialize an STI subclass. Ensures that the class name is a
    valid class and that it is in the ancestors of the super class
    that the association is expecting.

    *Jason Rush*

*   Observers was extracted from Active Record as `rails-observers` gem.

    *Rafael Mendonça França*

*   Ensure that associations take a symbol argument. *Steve Klabnik*

*   Fix dirty attribute checks for `TimeZoneConversion` with nil and blank
    datetime attributes. Setting a nil datetime to a blank string should not
    result in a change being flagged.
    Fixes #8310.

    *Alisdair McDiarmid*

*   Prevent mass assignment to the type column of polymorphic associations when using `build`
    Fixes #8265.

    *Yves Senn*

*   Deprecate calling `Relation#sum` with a block. To perform a calculation over
    the array result of the relation, use `to_a.sum(&block)`.

    *Carlos Antonio da Silva*

*   Fix postgresql adapter to handle BC timestamps correctly

        HistoryEvent.create!(name: "something", occured_at: Date.new(0) - 5.years)

    *Bogdan Gusiev*

*   When running migrations on Postgresql, the `:limit` option for `binary` and `text` columns is silently dropped.
    Previously, these migrations caused sql exceptions, because Postgresql doesn't support limits on these types.

    *Victor Costan*

*   Don't change STI type when calling `ActiveRecord::Base#becomes`.
    Add `ActiveRecord::Base#becomes!` with the previous behavior.

    See #3023 for more information.

    *Thomas Hollstegge*

*   `rename_index` can be used inside a `change_table` block.

        change_table :accounts do |t|
          t.rename_index :user_id, :account_id
        end

    *Jarek Radosz*

*   `#pluck` can be used on a relation with `select` clause. Fix #7551

    Example:

        Topic.select([:approved, :id]).order(:id).pluck(:id)

    *Yves Senn*

*   Do not create useless database transaction when building `has_one` association.

    Example:

        User.has_one :profile
        User.new.build_profile

    *Bogdan Gusiev*

*   `:counter_cache` option for `has_many` associations to support custom named counter caches.
    Fixes #7993.

    *Yves Senn*

*   Deprecate the possibility to pass a string as third argument of `add_index`.
    Pass `unique: true` instead.

        add_index(:users, :organization_id, unique: true)

    *Rafael Mendonça França*

*   Raise an `ArgumentError` when passing an invalid option to `add_index`.

    *Rafael Mendonça França*

*   Fix `find_in_batches` crashing when IDs are strings and start option is not specified.

    *Alexis Bernard*

*   `AR::Base#attributes_before_type_cast` now returns unserialized values for serialized attributes.

    *Nikita Afanasenko*

*   Use query cache/uncache when using `DATABASE_URL`.
    Fixes #6951.

    *kennyj*

*   Fix bug where `update_columns` and `update_column` would not let you update the primary key column.

    *Henrik Nyh*

*   The `create_table` method raises an `ArgumentError` when the primary key column is redefined.
    Fixes #6378.

    *Yves Senn*

*   `ActiveRecord::AttributeMethods#[]` raises `ActiveModel::MissingAttributeError`
    error if the given attribute is missing. Fixes #5433.

        class Person < ActiveRecord::Base
          belongs_to :company
        end

        # Before:
        person = Person.select('id').first
        person[:name]       # => nil
        person.name         # => ActiveModel::MissingAttributeError: missing_attribute: name
        person[:company_id] # => nil
        person.company      # => nil

        # After:
        person = Person.select('id').first
        person[:name]       # => ActiveModel::MissingAttributeError: missing_attribute: name
        person.name         # => ActiveModel::MissingAttributeError: missing_attribute: name
        person[:company_id] # => ActiveModel::MissingAttributeError: missing_attribute: company_id
        person.company      # => ActiveModel::MissingAttributeError: missing_attribute: company_id

    *Francesco Rodriguez*

*   Small binary fields use the `VARBINARY` MySQL type, instead of `TINYBLOB`.

    *Victor Costan*

*   Decode URI encoded attributes on database connection URLs.

    *Shawn Veader*

*   Add `find_or_create_by`, `find_or_create_by!` and
    `find_or_initialize_by` methods to `Relation`.

    These are similar to the `first_or_create` family of methods, but
    the behaviour when a record is created is slightly different:

        User.where(first_name: 'Penélope').first_or_create

    will execute:

        User.where(first_name: 'Penélope').create

    Causing all the `create` callbacks to execute within the context of
    the scope. This could affect queries that occur within callbacks.

        User.find_or_create_by(first_name: 'Penélope')

    will execute:

        User.create(first_name: 'Penélope')

    Which obviously does not affect the scoping of queries within
    callbacks.

    The `find_or_create_by` version also reads better, frankly.

    If you need to add extra attributes during create, you can do one of:

        User.create_with(active: true).find_or_create_by(first_name: 'Jon')
        User.find_or_create_by(first_name: 'Jon') { |u| u.active = true }

    The `first_or_create` family of methods have been nodoc'ed in favour
    of this API. They may be deprecated in the future but their
    implementation is very small and it's probably not worth putting users
    through lots of annoying deprecation warnings.

    *Jon Leighton*

*   Fix bug with presence validation of associations. Would incorrectly add duplicated errors
    when the association was blank. Bug introduced in 1fab518c6a75dac5773654646eb724a59741bc13.

    *Scott Willson*

*   Fix bug where sum(expression) returns string '0' for no matching records.
    Fixes #7439

    *Tim Macfarlane*

*   PostgreSQL adapter correctly fetches default values when using multiple schemas and domains in a db. Fixes #7914

    *Arturo Pie*

*   Learn ActiveRecord::QueryMethods#order work with hash arguments

    When symbol or hash passed we convert it to Arel::Nodes::Ordering.
    If we pass invalid direction(like name: :DeSc) ActiveRecord::QueryMethods#order will raise an exception

        User.order(:name, email: :desc)
        # SELECT "users".* FROM "users" ORDER BY "users"."name" ASC, "users"."email" DESC

    *Tima Maslyuchenko*

*   Rename `ActiveRecord::Fixtures` class to `ActiveRecord::FixtureSet`.
    Instances of this class normally hold a collection of fixtures (records)
    loaded either from a single YAML file, or from a file and a folder
    with the same name.  This change make the class name singular and makes
    the class easier to distinguish from the modules like
    `ActiveRecord::TestFixtures`, which operates on multiple fixture sets,
    or `DelegatingFixtures`, `::Fixtures`, etc.,
    and from the class `ActiveRecord::Fixture`, which corresponds to a single
    fixture.

    *Alexey Muranov*

*   The postgres adapter now supports tables with capital letters.
    Fixes #5920.

    *Yves Senn*

*   `CollectionAssociation#count` returns `0` without querying if the
    parent record is not persisted.

    Before:

        person.pets.count
        # SELECT COUNT(*) FROM "pets" WHERE "pets"."person_id" IS NULL
        # => 0

    After:

        person.pets.count
        # fires without sql query
        # => 0

    *Francesco Rodriguez*

*   Fix `reset_counters` crashing on `has_many :through` associations.
    Fixes #7822.

    *lulalala*

*   Support for partial inserts.

    When inserting new records, only the fields which have been changed
    from the defaults will actually be included in the INSERT statement.
    The other fields will be populated by the database.

    This is more efficient, and also means that it will be safe to
    remove database columns without getting subsequent errors in running
    app processes (so long as the code in those processes doesn't
    contain any references to the removed column).

    The `partial_updates` configuration option is now renamed to
    `partial_writes` to reflect the fact that it now impacts both inserts
    and updates.

    *Jon Leighton*

*   Allow before and after validations to take an array of lifecycle events

    *John Foley*

*   Support for specifying transaction isolation level

    If your database supports setting the isolation level for a transaction, you can set
    it like so:

        Post.transaction(isolation: :serializable) do
          # ...
        end

    Valid isolation levels are:

    * `:read_uncommitted`
    * `:read_committed`
    * `:repeatable_read`
    * `:serializable`

    You should consult the documentation for your database to understand the
    semantics of these different levels:

    * http://www.postgresql.org/docs/9.1/static/transaction-iso.html
    * https://dev.mysql.com/doc/refman/5.0/en/set-transaction.html

    An `ActiveRecord::TransactionIsolationError` will be raised if:

    * The adapter does not support setting the isolation level
    * You are joining an existing open transaction
    * You are creating a nested (savepoint) transaction

    The mysql, mysql2 and postgresql adapters support setting the transaction
    isolation level. However, support is disabled for mysql versions below 5,
    because they are affected by a bug (http://bugs.mysql.com/bug.php?id=39170)
    which means the isolation level gets persisted outside the transaction.

    *Jon Leighton*

*   `ActiveModel::ForbiddenAttributesProtection` is included by default
    in Active Record models. Check the docs of `ActiveModel::ForbiddenAttributesProtection`
    for more details.

    *Guillermo Iguaran*

*   Remove integration between Active Record and
    `ActiveModel::MassAssignmentSecurity`, `protected_attributes` gem
    should be added to use `attr_accessible`/`attr_protected`. Mass
    assignment options has been removed from all the AR methods that
    used it (ex. `AR::Base.new`, `AR::Base.create`, `AR::Base#update_attributes`, etc).

    *Guillermo Iguaran*

*   Fix the return of querying with an empty hash.
    Fixes #6971.

        User.where(token: {})

    Before:

        #=> SELECT * FROM users;

    After:

        #=> SELECT * FROM users WHERE 1=0;

    *Damien Mathieu*

*   Fix creation of through association models when using `collection=[]`
    on a `has_many :through` association from an unsaved model.
    Fixes #7661.

    *Ernie Miller*

*   Explain only normal CRUD sql (select / update / insert / delete).
    Fix problem that explains unexplainable sql.
    Closes #7544 #6458.

    *kennyj*

*   You can now override the generated accessor methods for stored attributes
    and reuse the original behavior with `read_store_attribute` and `write_store_attribute`,
    which are counterparts to `read_attribute` and `write_attribute`.

    *Matt Jones*

*   Accept `belongs_to` (including polymorphic) association keys in queries.

    The following queries are now equivalent:

        Post.where(author: author)
        Post.where(author_id: author)

        PriceEstimate.where(estimate_of: treasure)
        PriceEstimate.where(estimate_of_type: 'Treasure', estimate_of_id: treasure)

    *Peter Brown*

*   Use native `mysqldump` command instead of `structure_dump` method
    when dumping the database structure to a sql file. Fixes #5547.

    *kennyj*

*   PostgreSQL inet and cidr types are converted to `IPAddr` objects.

    *Dan McClain*

*   PostgreSQL array type support. Any datatype can be used to create an
    array column, with full migration and schema dumper support.

    To declare an array column, use the following syntax:

        create_table :table_with_arrays do |t|
          t.integer :int_array, array: true
          # integer[]
          t.integer :int_array, array: true, length: 2
          # smallint[]
          t.string :string_array, array: true, length: 30
          # char varying(30)[]
        end

    This respects any other migration detail (limits, defaults, etc).
    Active Record will serialize and deserialize the array columns on
    their way to and from the database.

    One thing to note: PostgreSQL does not enforce any limits on the
    number of elements, and any array can be multi-dimensional. Any
    array that is multi-dimensional must be rectangular (each sub array
    must have the same number of elements as its siblings).

    If the `pg_array_parser` gem is available, it will be used when
    parsing PostgreSQL's array representation.

    *Dan McClain*

*   Attribute predicate methods, such as `article.title?`, will now raise
    `ActiveModel::MissingAttributeError` if the attribute being queried for
    truthiness was not read from the database, instead of just returning `false`.

    *Ernie Miller*

*   `ActiveRecord::SchemaDumper` uses Ruby 1.9 style hash, which means that the
    schema.rb file will be generated using this new syntax from now on.

    *Konstantin Shabanov*

*   Map interval with precision to string datatype in PostgreSQL. Fixes #7518.

    *Yves Senn*

*   Fix eagerly loading associations without primary keys. Fixes #4976.

    *Kelley Reynolds*

*   Rails now raise an exception when you're trying to run a migration that has an invalid
    file name. Only lower case letters, numbers, and '_' are allowed in migration's file name.
    Please see #7419 for more details.

    *Jan Bernacki*

*   Fix bug when calling `store_accessor` multiple times.
    Fixes #7532.

    *Matt Jones*

*   Fix store attributes that show the changes incorrectly.
    Fixes #7532.

    *Matt Jones*

*   Fix `ActiveRecord::Relation#pluck` when columns or tables are reserved words.

    *Ian Lesperance*

*   Allow JSON columns to be created in PostgreSQL and properly encoded/decoded.
    to/from database.

    *Dickson S. Guedes*

*   Fix time column type casting for invalid time string values to correctly return `nil`.

    *Adam Meehan*

*   Allow to pass Symbol or Proc into `:limit` option of #accepts_nested_attributes_for.

    *Mikhail Dieterle*

*   ActiveRecord::SessionStore has been extracted from Active Record as `activerecord-session_store`
    gem. Please read the `README.md` file on the gem for the usage.

    *Prem Sichanugrist*

*   Fix `reset_counters` when there are multiple `belongs_to` association with the
    same foreign key and one of them have a counter cache.
    Fixes #5200.

    *Dave Desrochers*

*   `serialized_attributes` and `_attr_readonly` become class method only. Instance reader methods are deprecated.

    *kennyj*

*   Round usec when comparing timestamp attributes in the dirty tracking.
    Fixes #6975.

    *kennyj*

*   Use inversed parent for first and last child of `has_many` association.

    *Ravil Bayramgalin*

*   Fix `Column.microseconds` and `Column.fast_string_to_time` to avoid converting
    timestamp seconds to a float, since it occasionally results in inaccuracies
    with microsecond-precision times. Fixes #7352.

    *Ari Pollak*

*   Fix AR#dup to nullify the validation errors in the dup'ed object. Previously the original
    and the dup'ed object shared the same errors.

    *Christian Seiler*

*   Raise `ArgumentError` if list of attributes to change is empty in `update_all`.

    *Roman Shatsov*

*   Fix AR#create to return an unsaved record when AR::RecordInvalid is
    raised. Fixes #3217.

    *Dave Yeu*

*   Fixed table name prefix that is generated in engines for namespaced models.

    *Wojciech Wnętrzak*

*   Make sure `:environment` task is executed before `db:schema:load` or `db:structure:load`.
    Fixes #4772.

    *Seamus Abshere*

*   Allow Relation#merge to take a proc.

    This was requested by DHH to allow creating of one's own custom
    association macros.

    For example:

        module Commentable
          def has_many_comments(extra)
            has_many :comments, -> { where(:foo).merge(extra) }
          end
        end

        class Post < ActiveRecord::Base
          extend Commentable
          has_many_comments -> { where(:bar) }
        end

    *Jon Leighton*

*   Add CollectionProxy#scope.

    This can be used to get a Relation from an association.

    Previously we had a #scoped method, but we're deprecating that for
    AR::Base, so it doesn't make sense to have it here.

    This was requested by DHH, to facilitate code like this:

        Project.scope.order('created_at DESC').page(current_page).tagged_with(@tag).limit(5).scoping do
          @topics      = @project.topics.scope
          @todolists   = @project.todolists.scope
          @attachments = @project.attachments.scope
          @documents   = @project.documents.scope
        end

    *Jon Leighton*

*   Add `Relation#load`.

    This method explicitly loads the records and then returns `self`.

    Rather than deciding between "do I want an array or a relation?",
    most people are actually asking themselves "do I want to eager load
    or lazy load?" Therefore, this method provides a way to explicitly
    eager-load without having to switch from a `Relation` to an array.

    Example:

        @posts = Post.where(published: true).load

    *Jon Leighton*

*   `Relation#order`: make new order prepend old one.

        User.order("name asc").order("created_at desc")
        # SELECT * FROM users ORDER BY created_at desc, name asc

    This also affects order defined in `default_scope` or any kind of associations.

    *Bogdan Gusiev*

*   `Model.all` now returns an `ActiveRecord::Relation`, rather than an
    array of records. Use `Relation#to_a` if you really want an array.

    In some specific cases, this may cause breakage when upgrading.
    However in most cases the `ActiveRecord::Relation` will just act as a
    lazy-loaded array and there will be no problems.

    Note that calling `Model.all` with options (e.g.
    `Model.all(conditions: '...')` was already deprecated, but it will
    still return an array in order to make the transition easier.

    `Model.scoped` is deprecated in favour of `Model.all`.

    `Relation#all` still returns an array, but is deprecated (since it
    would serve no purpose if we made it return a `Relation`).

    *Jon Leighton*

*   `:finder_sql` and `:counter_sql` options on collection associations
    are deprecated. Please transition to using scopes.

    *Jon Leighton*

*   `:insert_sql` and `:delete_sql` options on `has_and_belongs_to_many`
    associations are deprecated. Please transition to using `has_many
    :through`.

    *Jon Leighton*

*   Added `#update_columns` method which updates the attributes from
    the passed-in hash without calling save, hence skipping validations and
    callbacks. `ActiveRecordError` will be raised when called on new objects
    or when at least one of the attributes is marked as read only.

        post.attributes # => {"id"=>2, "title"=>"My title", "body"=>"My content", "author"=>"Peter"}
        post.update_columns(title: 'New title', author: 'Sebastian') # => true
        post.attributes # => {"id"=>2, "title"=>"New title", "body"=>"My content", "author"=>"Sebastian"}

    *Sebastian Martinez + Rafael Mendonça França*

*   The migration generator now creates a join table with (commented) indexes every time
    the migration name contains the word `join_table`:

        rails g migration create_join_table_for_artists_and_musics artist_id:index music_id

    *Aleksey Magusev*

*   Add `add_reference` and `remove_reference` schema statements. Aliases, `add_belongs_to`
    and `remove_belongs_to` are acceptable. References are reversible.

    Examples:

        # Create a user_id column
        add_reference(:products, :user)
        # Create a supplier_id, supplier_type columns and appropriate index
        add_reference(:products, :supplier, polymorphic: true, index: true)
        # Remove polymorphic reference
        remove_reference(:products, :supplier, polymorphic: true)

    *Aleksey Magusev*

*   Add `:default` and `:null` options to `column_exists?`.

        column_exists?(:testings, :taggable_id, :integer, null: false)
        column_exists?(:testings, :taggable_type, :string, default: 'Photo')

    *Aleksey Magusev*

*   `ActiveRecord::Relation#inspect` now makes it clear that you are
    dealing with a `Relation` object rather than an array:.

        User.where(age: 30).inspect
        # => <ActiveRecord::Relation [#<User ...>, #<User ...>, ...]>

        User.where(age: 30).to_a.inspect
        # => [#<User ...>, #<User ...>]

    The number of records displayed will be limited to 10.

    *Brian Cardarella, Jon Leighton & Damien Mathieu*

*   Add `collation` and `ctype` support to PostgreSQL. These are available for PostgreSQL 8.4 or later.
    Example:

        development:
          adapter: postgresql
          host: localhost
          database: rails_development
          username: foo
          password: bar
          encoding: UTF8
          collation: ja_JP.UTF8
          ctype: ja_JP.UTF8

    *kennyj*

*   Changed `validates_presence_of` on an association so that children objects
    do not validate as being present if they are marked for destruction. This
    prevents you from saving the parent successfully and thus putting the parent
    in an invalid state.

    *Nick Monje & Brent Wheeldon*

*   `FinderMethods#exists?` now returns `false` with the `false` argument.

    *Egor Lynko*

*   Added support for specifying the precision of a timestamp in the postgresql
    adapter. So, instead of having to incorrectly specify the precision using the
    `:limit` option, you may use `:precision`, as intended. For example, in a migration:

        def change
          create_table :foobars do |t|
            t.timestamps precision: 0
          end
        end

    *Tony Schneider*

*   Allow `ActiveRecord::Relation#pluck` to accept multiple columns. Returns an
    array of arrays containing the typecasted values:

        Person.pluck(:id, :name)
        # SELECT people.id, people.name FROM people
        # [[1, 'David'], [2, 'Jeremy'], [3, 'Jose']]

    *Jeroen van Ingen & Carlos Antonio da Silva*

*   Improve the derivation of HABTM join table name to take account of nesting.
    It now takes the table names of the two models, sorts them lexically and
    then joins them, stripping any common prefix from the second table name.

    Some examples:

        Top level models (Category <=> Product)
        Old: categories_products
        New: categories_products

        Top level models with a global table_name_prefix (Category <=> Product)
        Old: site_categories_products
        New: site_categories_products

        Nested models in a module without a table_name_prefix method (Admin::Category <=> Admin::Product)
        Old: categories_products
        New: categories_products

        Nested models in a module with a table_name_prefix method (Admin::Category <=> Admin::Product)
        Old: categories_products
        New: admin_categories_products

        Nested models in a parent model (Catalog::Category <=> Catalog::Product)
        Old: categories_products
        New: catalog_categories_products

        Nested models in different parent models (Catalog::Category <=> Content::Page)
        Old: categories_pages
        New: catalog_categories_content_pages

    *Andrew White*

*   Move HABTM validity checks to `ActiveRecord::Reflection`. One side effect of
    this is to move when the exceptions are raised from the point of declaration
    to when the association is built. This is consistant with other association
    validity checks.

    *Andrew White*

*   Added `stored_attributes` hash which contains the attributes stored using
    `ActiveRecord::Store`. This allows you to retrieve the list of attributes
    you've defined.

       class User < ActiveRecord::Base
         store :settings, accessors: [:color, :homepage]
       end

       User.stored_attributes[:settings] # [:color, :homepage]

    *Joost Baaij & Carlos Antonio da Silva*

*   PostgreSQL default log level is now 'warning', to bypass the noisy notice
    messages. You can change the log level using the `min_messages` option
    available in your config/database.yml.

    *kennyj*

*   Add uuid datatype support to PostgreSQL adapter.

    *Konstantin Shabanov*

*   Added `ActiveRecord::Migration.check_pending!` that raises an error if
    migrations are pending.

    *Richard Schneeman*

*   Added `#destroy!` which acts like `#destroy` but will raise an
    `ActiveRecord::RecordNotDestroyed` exception instead of returning `false`.

    *Marc-André Lafortune*

*   Added support to `CollectionAssociation#delete` for passing `fixnum`
    or `string` values as record ids. This finds the records responding
    to the `id` and executes delete on them.

        class Person < ActiveRecord::Base
          has_many :pets
        end

        person.pets.delete("1") # => [#<Pet id: 1>]
        person.pets.delete(2, 3) # => [#<Pet id: 2>, #<Pet id: 3>]

    *Francesco Rodriguez*

*   Deprecated most of the 'dynamic finder' methods. All dynamic methods
    except for `find_by_...` and `find_by_...!` are deprecated. Here's
    how you can rewrite the code:

      * `find_all_by_...` can be rewritten using `where(...)`
      * `find_last_by_...` can be rewritten using `where(...).last`
      * `scoped_by_...` can be rewritten using `where(...)`
      * `find_or_initialize_by_...` can be rewritten using
        `where(...).first_or_initialize`
      * `find_or_create_by_...` can be rewritten using
        `find_or_create_by(...)` or where(...).first_or_create`
      * `find_or_create_by_...!` can be rewritten using
        `find_or_create_by!(...) or `where(...).first_or_create!`

    The implementation of the deprecated dynamic finders has been moved
    to the `activerecord-deprecated_finders` gem. See below for details.

    *Jon Leighton*

*   Deprecated the old-style hash based finder API. This means that
    methods which previously accepted "finder options" no longer do. For
    example this:

        Post.find(:all, conditions: { comments_count: 10 }, limit: 5)

    Should be rewritten in the new style which has existed since Rails 3:

        Post.where(comments_count: 10).limit(5)

    Note that as an interim step, it is possible to rewrite the above as:

        Post.all.merge(where: { comments_count: 10 }, limit: 5)

    This could save you a lot of work if there is a lot of old-style
    finder usage in your application.

    `Relation#merge` now accepts a hash of
    options, but they must be identical to the names of the equivalent
    finder method. These are mostly identical to the old-style finder
    option names, except in the following cases:

      * `:conditions` becomes `:where`.
      * `:include` becomes `:includes`.

    The code to implement the deprecated features has been moved out to the
    `activerecord-deprecated_finders` gem. This gem is a dependency of Active
    Record in Rails 4.0, so the interface works out of the box. It will no
    longer be a dependency from Rails 4.1 (you'll need to add it to the
    `Gemfile` in 4.1), and will be maintained until Rails 5.0.

    *Jon Leighton*

*   It's not possible anymore to destroy a model marked as read only.

    *Johannes Barre*

*   Added ability to ActiveRecord::Relation#from to accept other ActiveRecord::Relation objects.

      Record.from(subquery)
      Record.from(subquery, :a)

    *Radoslav Stankov*

*   Added custom coders support for ActiveRecord::Store. Now you can set
    your custom coder like this:

        store :settings, accessors: [ :color, :homepage ], coder: JSON

    *Andrey Voronkov*

*   `mysql` and `mysql2` connections will set `SQL_MODE=STRICT_ALL_TABLES` by
    default to avoid silent data loss. This can be disabled by specifying
    `strict: false` in your `database.yml`.

    *Michael Pearson*

*   Added default order to `first` to assure consistent results among
    different database engines. Introduced `take` as a replacement to
    the old behavior of `first`.

    *Marcelo Silveira*

*   Added an `:index` option to automatically create indexes for references
    and belongs_to statements in migrations.

    The `references` and `belongs_to` methods now support an `index`
    option that receives either a boolean value or an options hash
    that is identical to options available to the add_index method:

      create_table :messages do |t|
        t.references :person, index: true
      end

      Is the same as:

      create_table :messages do |t|
        t.references :person
      end
      add_index :messages, :person_id

    Generators have also been updated to use the new syntax.

    *Joshua Wood*

*   Added `#find_by` and `#find_by!` to mirror the functionality
    provided by dynamic finders in a way that allows dynamic input more
    easily:

        Post.find_by name: 'Spartacus', rating: 4
        Post.find_by "published_at < ?", 2.weeks.ago
        Post.find_by! name: 'Spartacus'

    *Jon Leighton*

*   Added ActiveRecord::Base#slice to return a hash of the given methods with
    their names as keys and returned values as values.

    *Guillermo Iguaran*

*   Deprecate eager-evaluated scopes.

    Don't use this:

        scope :red, where(color: 'red')
        default_scope where(color: 'red')

    Use this:

        scope :red, -> { where(color: 'red') }
        default_scope { where(color: 'red') }

    The former has numerous issues. It is a common newbie gotcha to do
    the following:

        scope :recent, where(published_at: Time.now - 2.weeks)

    Or a more subtle variant:

        scope :recent, -> { where(published_at: Time.now - 2.weeks) }
        scope :recent_red, recent.where(color: 'red')

    Eager scopes are also very complex to implement within Active
    Record, and there are still bugs. For example, the following does
    not do what you expect:

        scope :remove_conditions, except(:where)
        where(...).remove_conditions # => still has conditions

    *Jon Leighton*

*   Remove IdentityMap

    IdentityMap has never graduated to be an "enabled-by-default" feature, due
    to some inconsistencies with associations, as described in this commit:

       https://github.com/rails/rails/commit/302c912bf6bcd0fa200d964ec2dc4a44abe328a6

    Hence the removal from the codebase, until such issues are fixed.

    *Carlos Antonio da Silva*

*   Added the schema cache dump feature.

    `Schema cache dump` feature was implemetend. This feature can dump/load internal state of `SchemaCache` instance
    because we want to boot rails more quickly when we have many models.

    Usage notes:

      1) execute rake task.
      RAILS_ENV=production bundle exec rake db:schema:cache:dump
      => generate db/schema_cache.dump

      2) add config.active_record.use_schema_cache_dump = true in config/production.rb. BTW, true is default.

      3) boot rails.
      RAILS_ENV=production bundle exec rails server
      => use db/schema_cache.dump

      4) If you remove clear dumped cache, execute rake task.
      RAILS_ENV=production bundle exec rake db:schema:cache:clear
      => remove db/schema_cache.dump

    *kennyj*

*   Added support for partial indices to PostgreSQL adapter.

    The `add_index` method now supports a `where` option that receives a
    string with the partial index criteria.

        add_index(:accounts, :code, where: 'active')

        Generates

        CREATE INDEX index_accounts_on_code ON accounts(code) WHERE active

    *Marcelo Silveira*

*   Implemented ActiveRecord::Relation#none method.

    The `none` method returns a chainable relation with zero records
    (an instance of the NullRelation class).

    Any subsequent condition chained to the returned relation will continue
    generating an empty relation and will not fire any query to the database.

    *Juanjo Bazán*

*   Added the `ActiveRecord::NullRelation` class implementing the null
    object pattern for the Relation class.

    *Juanjo Bazán*

*   Added new `dependent: :restrict_with_error` option. This will add
    an error to the model, rather than raising an exception.

    The `:restrict` option is renamed to `:restrict_with_exception` to
    make this distinction explicit.

    *Manoj Kumar & Jon Leighton*

*   Added `create_join_table` migration helper to create HABTM join tables.

        create_join_table :products, :categories
        # =>
        # create_table :categories_products, id: false do |td|
        #   td.integer :product_id,  null: false
        #   td.integer :category_id, null: false
        # end

    *Rafael Mendonça França*

*   The primary key is always initialized in the @attributes hash to `nil` (unless
    another value has been specified).

    *Aaron Paterson*

*   In previous releases, the following would generate a single query with
    an `OUTER JOIN comments`, rather than two separate queries:

        Post.includes(:comments)
            .where("comments.name = 'foo'")

    This behaviour relies on matching SQL string, which is an inherently
    flawed idea unless we write an SQL parser, which we do not wish to
    do.

    Therefore, it is now deprecated.

    To avoid deprecation warnings and for future compatibility, you must
    explicitly state which tables you reference, when using SQL snippets:

        Post.includes(:comments)
            .where("comments.name = 'foo'")
            .references(:comments)

    Note that you do not need to explicitly specify references in the
    following cases, as they can be automatically inferred:

        Post.includes(:comments).where(comments: { name: 'foo' })
        Post.includes(:comments).where('comments.name' => 'foo')
        Post.includes(:comments).order('comments.name')

    You do not need to worry about this unless you are doing eager
    loading. Basically, don't worry unless you see a deprecation warning
    or (in future releases) an SQL error due to a missing JOIN.

    *Jon Leighton*

*   Support for the `schema_info` table has been dropped. Please
    switch to `schema_migrations`.

    *Aaron Patterson*

*   Connections *must* be closed at the end of a thread. If not, your
    connection pool can fill and an exception will be raised.

    *Aaron Patterson*

*   PostgreSQL hstore records can be created.

    *Aaron Patterson*

*   PostgreSQL hstore types are automatically deserialized from the database.

    *Aaron Patterson*


Please check [3-2-stable](https://github.com/rails/rails/blob/3-2-stable/activerecord/CHANGELOG.md) for previous changes.