activerecord 7.0.8.1 → 7.1.0.beta1
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +4 -4
- data/CHANGELOG.md +1348 -1551
- data/MIT-LICENSE +1 -1
- data/README.rdoc +15 -16
- data/lib/active_record/aggregations.rb +16 -13
- data/lib/active_record/association_relation.rb +1 -1
- data/lib/active_record/associations/association.rb +18 -3
- data/lib/active_record/associations/association_scope.rb +16 -9
- data/lib/active_record/associations/belongs_to_association.rb +14 -6
- data/lib/active_record/associations/builder/association.rb +3 -3
- data/lib/active_record/associations/builder/belongs_to.rb +21 -8
- data/lib/active_record/associations/builder/has_and_belongs_to_many.rb +1 -5
- data/lib/active_record/associations/builder/singular_association.rb +4 -0
- data/lib/active_record/associations/collection_association.rb +17 -9
- data/lib/active_record/associations/collection_proxy.rb +16 -11
- data/lib/active_record/associations/foreign_association.rb +10 -3
- data/lib/active_record/associations/has_many_association.rb +20 -13
- data/lib/active_record/associations/has_many_through_association.rb +10 -6
- data/lib/active_record/associations/has_one_association.rb +10 -3
- data/lib/active_record/associations/join_dependency.rb +10 -8
- data/lib/active_record/associations/preloader/association.rb +27 -6
- data/lib/active_record/associations/preloader.rb +12 -9
- data/lib/active_record/associations/singular_association.rb +1 -1
- data/lib/active_record/associations/through_association.rb +22 -11
- data/lib/active_record/associations.rb +193 -97
- data/lib/active_record/attribute_assignment.rb +0 -2
- data/lib/active_record/attribute_methods/before_type_cast.rb +17 -0
- data/lib/active_record/attribute_methods/dirty.rb +40 -26
- data/lib/active_record/attribute_methods/primary_key.rb +76 -24
- data/lib/active_record/attribute_methods/query.rb +28 -16
- data/lib/active_record/attribute_methods/read.rb +18 -5
- data/lib/active_record/attribute_methods/serialization.rb +150 -31
- data/lib/active_record/attribute_methods/write.rb +3 -3
- data/lib/active_record/attribute_methods.rb +105 -21
- data/lib/active_record/attributes.rb +3 -3
- data/lib/active_record/autosave_association.rb +55 -9
- data/lib/active_record/base.rb +7 -2
- data/lib/active_record/callbacks.rb +10 -24
- data/lib/active_record/coders/column_serializer.rb +61 -0
- data/lib/active_record/coders/json.rb +1 -1
- data/lib/active_record/coders/yaml_column.rb +70 -42
- data/lib/active_record/connection_adapters/abstract/connection_handler.rb +163 -88
- data/lib/active_record/connection_adapters/abstract/connection_pool/queue.rb +2 -0
- data/lib/active_record/connection_adapters/abstract/connection_pool/reaper.rb +3 -1
- data/lib/active_record/connection_adapters/abstract/connection_pool.rb +63 -43
- data/lib/active_record/connection_adapters/abstract/database_limits.rb +5 -0
- data/lib/active_record/connection_adapters/abstract/database_statements.rb +109 -32
- data/lib/active_record/connection_adapters/abstract/query_cache.rb +60 -22
- data/lib/active_record/connection_adapters/abstract/quoting.rb +41 -6
- data/lib/active_record/connection_adapters/abstract/savepoints.rb +4 -3
- data/lib/active_record/connection_adapters/abstract/schema_creation.rb +18 -4
- data/lib/active_record/connection_adapters/abstract/schema_definitions.rb +137 -11
- data/lib/active_record/connection_adapters/abstract/schema_statements.rb +289 -122
- data/lib/active_record/connection_adapters/abstract/transaction.rb +280 -58
- data/lib/active_record/connection_adapters/abstract_adapter.rb +502 -91
- data/lib/active_record/connection_adapters/abstract_mysql_adapter.rb +200 -108
- data/lib/active_record/connection_adapters/column.rb +9 -0
- data/lib/active_record/connection_adapters/mysql/column.rb +1 -0
- data/lib/active_record/connection_adapters/mysql/database_statements.rb +22 -143
- data/lib/active_record/connection_adapters/mysql/quoting.rb +16 -12
- data/lib/active_record/connection_adapters/mysql/schema_creation.rb +9 -0
- data/lib/active_record/connection_adapters/mysql/schema_definitions.rb +6 -0
- data/lib/active_record/connection_adapters/mysql/schema_dumper.rb +1 -1
- data/lib/active_record/connection_adapters/mysql/schema_statements.rb +17 -12
- data/lib/active_record/connection_adapters/mysql2/database_statements.rb +148 -0
- data/lib/active_record/connection_adapters/mysql2_adapter.rb +98 -53
- data/lib/active_record/connection_adapters/pool_config.rb +14 -5
- data/lib/active_record/connection_adapters/pool_manager.rb +19 -9
- data/lib/active_record/connection_adapters/postgresql/column.rb +1 -2
- data/lib/active_record/connection_adapters/postgresql/database_statements.rb +76 -29
- data/lib/active_record/connection_adapters/postgresql/oid/range.rb +11 -2
- data/lib/active_record/connection_adapters/postgresql/quoting.rb +9 -6
- data/lib/active_record/connection_adapters/postgresql/referential_integrity.rb +3 -9
- data/lib/active_record/connection_adapters/postgresql/schema_creation.rb +76 -6
- data/lib/active_record/connection_adapters/postgresql/schema_definitions.rb +131 -2
- data/lib/active_record/connection_adapters/postgresql/schema_dumper.rb +42 -0
- data/lib/active_record/connection_adapters/postgresql/schema_statements.rb +351 -54
- data/lib/active_record/connection_adapters/postgresql_adapter.rb +336 -168
- data/lib/active_record/connection_adapters/schema_cache.rb +287 -59
- data/lib/active_record/connection_adapters/sqlite3/column.rb +49 -0
- data/lib/active_record/connection_adapters/sqlite3/database_statements.rb +42 -36
- data/lib/active_record/connection_adapters/sqlite3/quoting.rb +4 -3
- data/lib/active_record/connection_adapters/sqlite3/schema_definitions.rb +1 -0
- data/lib/active_record/connection_adapters/sqlite3/schema_statements.rb +26 -7
- data/lib/active_record/connection_adapters/sqlite3_adapter.rb +162 -77
- data/lib/active_record/connection_adapters/statement_pool.rb +7 -0
- data/lib/active_record/connection_adapters/trilogy/database_statements.rb +98 -0
- data/lib/active_record/connection_adapters/trilogy_adapter.rb +254 -0
- data/lib/active_record/connection_adapters.rb +3 -1
- data/lib/active_record/connection_handling.rb +71 -94
- data/lib/active_record/core.rb +128 -138
- data/lib/active_record/counter_cache.rb +46 -25
- data/lib/active_record/database_configurations/database_config.rb +9 -3
- data/lib/active_record/database_configurations/hash_config.rb +22 -12
- data/lib/active_record/database_configurations/url_config.rb +17 -11
- data/lib/active_record/database_configurations.rb +86 -33
- data/lib/active_record/delegated_type.rb +8 -3
- data/lib/active_record/deprecator.rb +7 -0
- data/lib/active_record/destroy_association_async_job.rb +2 -0
- data/lib/active_record/encryption/auto_filtered_parameters.rb +66 -0
- data/lib/active_record/encryption/cipher/aes256_gcm.rb +4 -1
- data/lib/active_record/encryption/config.rb +25 -1
- data/lib/active_record/encryption/configurable.rb +12 -19
- data/lib/active_record/encryption/context.rb +10 -3
- data/lib/active_record/encryption/contexts.rb +5 -1
- data/lib/active_record/encryption/derived_secret_key_provider.rb +8 -2
- data/lib/active_record/encryption/encryptable_record.rb +36 -18
- data/lib/active_record/encryption/encrypted_attribute_type.rb +17 -6
- data/lib/active_record/encryption/extended_deterministic_queries.rb +66 -54
- data/lib/active_record/encryption/extended_deterministic_uniqueness_validator.rb +2 -2
- data/lib/active_record/encryption/key_generator.rb +12 -1
- data/lib/active_record/encryption/message_serializer.rb +2 -0
- data/lib/active_record/encryption/properties.rb +3 -3
- data/lib/active_record/encryption/scheme.rb +19 -22
- data/lib/active_record/encryption.rb +1 -0
- data/lib/active_record/enum.rb +113 -26
- data/lib/active_record/errors.rb +89 -15
- data/lib/active_record/explain.rb +23 -3
- data/lib/active_record/fixture_set/model_metadata.rb +14 -4
- data/lib/active_record/fixture_set/render_context.rb +2 -0
- data/lib/active_record/fixture_set/table_row.rb +29 -8
- data/lib/active_record/fixtures.rb +119 -71
- data/lib/active_record/future_result.rb +30 -5
- data/lib/active_record/gem_version.rb +4 -4
- data/lib/active_record/inheritance.rb +30 -16
- data/lib/active_record/insert_all.rb +55 -8
- data/lib/active_record/integration.rb +8 -8
- data/lib/active_record/internal_metadata.rb +118 -30
- data/lib/active_record/locking/pessimistic.rb +5 -2
- data/lib/active_record/log_subscriber.rb +29 -12
- data/lib/active_record/marshalling.rb +56 -0
- data/lib/active_record/message_pack.rb +124 -0
- data/lib/active_record/middleware/database_selector/resolver.rb +4 -0
- data/lib/active_record/middleware/database_selector.rb +5 -7
- data/lib/active_record/middleware/shard_selector.rb +3 -1
- data/lib/active_record/migration/command_recorder.rb +100 -4
- data/lib/active_record/migration/compatibility.rb +131 -5
- data/lib/active_record/migration/default_strategy.rb +23 -0
- data/lib/active_record/migration/execution_strategy.rb +19 -0
- data/lib/active_record/migration.rb +213 -109
- data/lib/active_record/model_schema.rb +47 -27
- data/lib/active_record/nested_attributes.rb +28 -3
- data/lib/active_record/normalization.rb +158 -0
- data/lib/active_record/persistence.rb +183 -33
- data/lib/active_record/promise.rb +84 -0
- data/lib/active_record/query_cache.rb +3 -21
- data/lib/active_record/query_logs.rb +77 -52
- data/lib/active_record/query_logs_formatter.rb +41 -0
- data/lib/active_record/querying.rb +15 -2
- data/lib/active_record/railtie.rb +107 -45
- data/lib/active_record/railties/controller_runtime.rb +10 -5
- data/lib/active_record/railties/databases.rake +139 -145
- data/lib/active_record/railties/job_runtime.rb +23 -0
- data/lib/active_record/readonly_attributes.rb +32 -5
- data/lib/active_record/reflection.rb +169 -45
- data/lib/active_record/relation/batches/batch_enumerator.rb +5 -3
- data/lib/active_record/relation/batches.rb +190 -61
- data/lib/active_record/relation/calculations.rb +152 -63
- data/lib/active_record/relation/delegation.rb +22 -8
- data/lib/active_record/relation/finder_methods.rb +85 -15
- data/lib/active_record/relation/merger.rb +2 -0
- data/lib/active_record/relation/predicate_builder/association_query_value.rb +11 -2
- data/lib/active_record/relation/predicate_builder/relation_handler.rb +5 -1
- data/lib/active_record/relation/predicate_builder.rb +26 -14
- data/lib/active_record/relation/query_attribute.rb +2 -1
- data/lib/active_record/relation/query_methods.rb +351 -62
- data/lib/active_record/relation/spawn_methods.rb +18 -1
- data/lib/active_record/relation.rb +76 -35
- data/lib/active_record/result.rb +19 -5
- data/lib/active_record/runtime_registry.rb +10 -1
- data/lib/active_record/sanitization.rb +51 -11
- data/lib/active_record/schema.rb +2 -3
- data/lib/active_record/schema_dumper.rb +41 -7
- data/lib/active_record/schema_migration.rb +68 -33
- data/lib/active_record/scoping/default.rb +15 -5
- data/lib/active_record/scoping/named.rb +2 -2
- data/lib/active_record/scoping.rb +2 -1
- data/lib/active_record/secure_password.rb +60 -0
- data/lib/active_record/secure_token.rb +21 -3
- data/lib/active_record/signed_id.rb +7 -5
- data/lib/active_record/store.rb +8 -8
- data/lib/active_record/suppressor.rb +3 -1
- data/lib/active_record/table_metadata.rb +10 -1
- data/lib/active_record/tasks/database_tasks.rb +127 -105
- data/lib/active_record/tasks/mysql_database_tasks.rb +15 -6
- data/lib/active_record/tasks/postgresql_database_tasks.rb +16 -13
- data/lib/active_record/tasks/sqlite_database_tasks.rb +14 -7
- data/lib/active_record/test_fixtures.rb +113 -96
- data/lib/active_record/timestamp.rb +26 -14
- data/lib/active_record/token_for.rb +113 -0
- data/lib/active_record/touch_later.rb +11 -6
- data/lib/active_record/transactions.rb +36 -10
- data/lib/active_record/type/adapter_specific_registry.rb +1 -8
- data/lib/active_record/type/internal/timezone.rb +7 -2
- data/lib/active_record/type/time.rb +4 -0
- data/lib/active_record/validations/absence.rb +1 -1
- data/lib/active_record/validations/numericality.rb +5 -4
- data/lib/active_record/validations/presence.rb +5 -28
- data/lib/active_record/validations/uniqueness.rb +47 -2
- data/lib/active_record/validations.rb +8 -4
- data/lib/active_record/version.rb +1 -1
- data/lib/active_record.rb +121 -16
- data/lib/arel/errors.rb +10 -0
- data/lib/arel/factory_methods.rb +4 -0
- data/lib/arel/nodes/binary.rb +6 -1
- data/lib/arel/nodes/bound_sql_literal.rb +61 -0
- data/lib/arel/nodes/cte.rb +36 -0
- data/lib/arel/nodes/fragments.rb +35 -0
- data/lib/arel/nodes/homogeneous_in.rb +0 -8
- data/lib/arel/nodes/leading_join.rb +8 -0
- data/lib/arel/nodes/node.rb +111 -2
- data/lib/arel/nodes/sql_literal.rb +6 -0
- data/lib/arel/nodes/table_alias.rb +4 -0
- data/lib/arel/nodes.rb +4 -0
- data/lib/arel/predications.rb +2 -0
- data/lib/arel/table.rb +9 -5
- data/lib/arel/visitors/mysql.rb +8 -1
- data/lib/arel/visitors/to_sql.rb +81 -17
- data/lib/arel/visitors/visitor.rb +2 -2
- data/lib/arel.rb +16 -2
- data/lib/rails/generators/active_record/application_record/USAGE +8 -0
- data/lib/rails/generators/active_record/migration.rb +3 -1
- data/lib/rails/generators/active_record/model/USAGE +113 -0
- data/lib/rails/generators/active_record/model/model_generator.rb +15 -6
- metadata +49 -14
- data/lib/active_record/connection_adapters/legacy_pool_manager.rb +0 -35
- data/lib/active_record/null_relation.rb +0 -63
data/CHANGELOG.md
CHANGED
@@ -1,2381 +1,2178 @@
|
|
1
|
-
## Rails 7.0.
|
1
|
+
## Rails 7.1.0.beta1 (September 13, 2023) ##
|
2
2
|
|
3
|
-
*
|
3
|
+
* Encryption now supports `support_unencrypted_data` being set per-attribute.
|
4
4
|
|
5
|
+
You can now opt out of `support_unencrypted_data` on a specific encrypted attribute.
|
6
|
+
This only has an effect if `ActiveRecord::Encryption.config.support_unencrypted_data == true`.
|
5
7
|
|
6
|
-
## Rails 7.0.8 (September 09, 2023) ##
|
7
|
-
|
8
|
-
* Fix `change_column` not setting `precision: 6` on `datetime` columns when
|
9
|
-
using 7.0+ Migrations and SQLite.
|
10
|
-
|
11
|
-
*Hartley McGuire*
|
12
|
-
|
13
|
-
* Fix unscope is not working in specific case
|
14
|
-
|
15
|
-
Before:
|
16
8
|
```ruby
|
17
|
-
|
18
|
-
|
19
|
-
|
20
|
-
|
21
|
-
After:
|
22
|
-
```ruby
|
23
|
-
Post.where(id: 1...3).unscope(where: :id).to_sql # "SELECT `posts`.* FROM `posts`"
|
9
|
+
class User < ActiveRecord::Base
|
10
|
+
encrypts :name, deterministic: true, support_unencrypted_data: false
|
11
|
+
encrypts :email, deterministic: true
|
12
|
+
end
|
24
13
|
```
|
25
14
|
|
26
|
-
|
15
|
+
*Alex Ghiculescu*
|
27
16
|
|
28
|
-
|
17
|
+
* Add instrumentation for Active Record transactions
|
29
18
|
|
30
|
-
|
19
|
+
Allows subscribing to transaction events for tracking/instrumentation. The event payload contains the connection, as well as timing details.
|
31
20
|
|
32
21
|
```ruby
|
33
|
-
|
34
|
-
|
35
|
-
|
36
|
-
|
37
|
-
# STI tables
|
38
|
-
class Request < ApplicationRecord
|
39
|
-
belongs_to :requestable, polymorphic: true
|
40
|
-
|
41
|
-
validate :request_type, presence: true
|
42
|
-
end
|
43
|
-
|
44
|
-
class ProductRequest < Request
|
45
|
-
belongs_to :user
|
22
|
+
ActiveSupport::Notifications.subscribe("transaction.active_record") do |event|
|
23
|
+
puts "Transaction event occurred!"
|
24
|
+
connection = event.payload[:connection]
|
25
|
+
puts "Connection: #{connection.inspect}"
|
46
26
|
end
|
47
27
|
```
|
48
28
|
|
49
|
-
|
50
|
-
|
51
|
-
```
|
52
|
-
table_metadata.rb:22:in `has_column?': undefined method `key?' for nil:NilClass (NoMethodError)
|
53
|
-
```
|
54
|
-
|
55
|
-
*Romain Filinto*
|
56
|
-
|
57
|
-
* Fix `change_table` setting datetime precision for 6.1 Migrations
|
58
|
-
|
59
|
-
*Hartley McGuire*
|
60
|
-
|
61
|
-
* Fix change_column setting datetime precision for 6.1 Migrations
|
62
|
-
|
63
|
-
*Hartley McGuire*
|
64
|
-
|
65
|
-
## Rails 7.0.7.2 (August 22, 2023) ##
|
66
|
-
|
67
|
-
* No changes.
|
68
|
-
|
69
|
-
|
70
|
-
## Rails 7.0.7.1 (August 22, 2023) ##
|
71
|
-
|
72
|
-
* No changes.
|
73
|
-
|
74
|
-
|
75
|
-
## Rails 7.0.7 (August 09, 2023) ##
|
76
|
-
|
77
|
-
* Restores functionality to the missing method when using enums and fixes.
|
78
|
-
|
79
|
-
*paulreece*
|
80
|
-
|
81
|
-
* Fix `StatementCache::Substitute` with serialized type.
|
82
|
-
|
83
|
-
*ywenc*
|
84
|
-
|
85
|
-
* Fix `:db_runtime` on notification payload when application have multiple databases.
|
29
|
+
*Daniel Colson*, *Ian Candy*
|
86
30
|
|
87
|
-
|
31
|
+
* Support composite foreign keys via migration helpers.
|
88
32
|
|
89
|
-
|
33
|
+
```ruby
|
34
|
+
# Assuming "carts" table has "(shop_id, user_id)" as a primary key.
|
90
35
|
|
91
|
-
|
36
|
+
add_foreign_key(:orders, :carts, primary_key: [:shop_id, :user_id])
|
92
37
|
|
93
|
-
|
94
|
-
|
95
|
-
|
96
|
-
For example, `Post.in_order_of(:title, [nil, "foo"])` will now include posts
|
97
|
-
with `nil` titles, the same as `Post.all.to_a.in_order_of(:title, [nil, "foo"])`.
|
38
|
+
remove_foreign_key(:orders, :carts, primary_key: [:shop_id, :user_id])
|
39
|
+
foreign_key_exists?(:orders, :carts, primary_key: [:shop_id, :user_id])
|
40
|
+
```
|
98
41
|
|
99
42
|
*fatkodima*
|
100
43
|
|
101
|
-
*
|
102
|
-
|
103
|
-
This change intended to remove the :base attribute from the message,
|
104
|
-
but broke many assumptions which key these errors were stored.
|
105
|
-
|
106
|
-
*zzak*
|
107
|
-
|
108
|
-
* Fix `#previously_new_record?` to return true for destroyed records.
|
109
|
-
|
110
|
-
Before, if a record was created and then destroyed, `#previously_new_record?` would return true.
|
111
|
-
Now, any UPDATE or DELETE to a record is considered a change, and will result in `#previously_new_record?`
|
112
|
-
returning false.
|
113
|
-
|
114
|
-
*Adrianna Chang*
|
115
|
-
|
116
|
-
* Revert breaking changes to `has_one` relationship deleting the old record before the new one is validated.
|
117
|
-
|
118
|
-
*zzak*
|
119
|
-
|
120
|
-
* Fix support for Active Record instances being uses in queries.
|
121
|
-
|
122
|
-
As of `7.0.5`, query arguments were deep duped to avoid mutations impacting
|
123
|
-
the query cache, but this had the adverse effect to clearing the primary key when
|
124
|
-
the query argument contained an `ActiveRecord::Base` instance.
|
125
|
-
|
126
|
-
This broke the `noticed` gem.
|
127
|
-
|
128
|
-
*Jean Boussier*
|
44
|
+
* Adds support for `if_not_exists` when adding a check constraint.
|
129
45
|
|
46
|
+
```ruby
|
47
|
+
add_check_constraint :posts, "post_type IN ('blog', 'comment', 'share')", if_not_exists: true
|
48
|
+
```
|
130
49
|
|
131
|
-
|
132
|
-
|
133
|
-
* Fix autosave associations with validations added on `:base` of the associated objects.
|
134
|
-
|
135
|
-
*fatkodima*
|
136
|
-
|
137
|
-
* Fix result with anonymous PostgreSQL columns of different type from json.
|
50
|
+
*Cody Cutrer*
|
138
51
|
|
139
|
-
|
52
|
+
* Raise an `ArgumentError` when `#accepts_nested_attributes_for` is declared more than once for an association in
|
53
|
+
the same class. Previously, the last declaration would silently override the previous one. Overriding in a subclass
|
54
|
+
is still allowed.
|
140
55
|
|
141
|
-
*
|
56
|
+
*Joshua Young*
|
142
57
|
|
143
|
-
|
58
|
+
* Deprecate `rewhere` argument on `#merge`.
|
144
59
|
|
145
|
-
|
60
|
+
The `rewhere` argument on `#merge`is deprecated without replacement and
|
61
|
+
will be removed in Rails 7.2.
|
146
62
|
|
147
|
-
*
|
63
|
+
*Adam Hess*
|
148
64
|
|
149
|
-
* Fix
|
65
|
+
* Fix unscope is not working in specific case
|
150
66
|
|
151
67
|
Before:
|
152
68
|
```ruby
|
153
|
-
|
154
|
-
|
69
|
+
Post.where(id: 1...3).unscope(where: :id).to_sql # "SELECT `posts`.* FROM `posts` WHERE `posts`.`id` >= 1 AND `posts`.`id` < 3"
|
70
|
+
|
155
71
|
```
|
156
72
|
|
157
|
-
|
73
|
+
After:
|
158
74
|
```ruby
|
159
|
-
|
160
|
-
#=> SELECT (...) WHERE "treasures"."id" IN (SELECT "price_estimates"."estimate_of_id" FROM "price_estimates" WHERE "price_estimates"."estimate_of_type" = 'Treasure')
|
75
|
+
Post.where(id: 1...3).unscope(where: :id).to_sql # "SELECT `posts`.* FROM `posts`"
|
161
76
|
```
|
162
77
|
|
163
|
-
|
164
|
-
|
165
|
-
* Fix decrementing counter caches on optimistically locked record deletion
|
166
|
-
|
167
|
-
*fatkodima*
|
168
|
-
|
169
|
-
* Ensure binary-destined values have binary encoding during type cast.
|
170
|
-
|
171
|
-
*Matthew Draper*
|
172
|
-
|
173
|
-
* Preserve existing column default functions when altering table in SQLite.
|
174
|
-
|
175
|
-
*fatkodima*
|
176
|
-
|
177
|
-
* Remove table alias added when using `where.missing` or `where.associated`.
|
178
|
-
|
179
|
-
*fatkodima*
|
180
|
-
|
181
|
-
* Fix `Enumerable#in_order_of` to only flatten first level to preserve nesting.
|
182
|
-
|
183
|
-
*Miha Rekar*
|
184
|
-
|
185
|
-
|
186
|
-
## Rails 7.0.5.1 (June 26, 2023) ##
|
187
|
-
|
188
|
-
* No changes.
|
189
|
-
|
190
|
-
## Rails 7.0.5 (May 24, 2023) ##
|
78
|
+
Fixes #48094.
|
191
79
|
|
192
|
-
*
|
80
|
+
*Kazuya Hatanaka*
|
193
81
|
|
194
|
-
|
82
|
+
* Change `has_secure_token` default to `on: :initialize`
|
195
83
|
|
196
|
-
|
84
|
+
Change the new default value from `on: :create` to `on: :initialize`
|
197
85
|
|
198
|
-
|
86
|
+
Can be controlled by the `config.active_record.generate_secure_token_on`
|
87
|
+
configuration:
|
199
88
|
|
200
|
-
|
89
|
+
```ruby
|
90
|
+
config.active_record.generate_secure_token_on = :create
|
91
|
+
```
|
201
92
|
|
202
|
-
*
|
93
|
+
*Sean Doyle*
|
203
94
|
|
204
|
-
* Fix
|
95
|
+
* Fix `change_column` not setting `precision: 6` on `datetime` columns when
|
96
|
+
using 7.0+ Migrations and SQLite.
|
205
97
|
|
206
98
|
*Hartley McGuire*
|
207
99
|
|
208
|
-
*
|
209
|
-
|
210
|
-
*Jonathan Hefner*
|
211
|
-
|
212
|
-
* Store correct environment in `internal_metadata` when run rails `db:prepare`.
|
213
|
-
|
214
|
-
*fatkodima*
|
215
|
-
|
216
|
-
* Make sure `ActiveRecord::Relation#sum` works with objects that implement `#coerce` without deprecation.
|
217
|
-
|
218
|
-
*Alex Ghiculescu*
|
219
|
-
|
220
|
-
* Fix retrieving foreign keys referencing tables named like keywords in PostgreSQL and MySQL.
|
221
|
-
|
222
|
-
*fatkodima*
|
223
|
-
|
224
|
-
* Support UUIDs in Disable Joins.
|
225
|
-
|
226
|
-
*Samuel Cochran*
|
227
|
-
|
228
|
-
* Fix Active Record's explain for queries starting with comments.
|
229
|
-
|
230
|
-
*fatkodima*
|
231
|
-
|
232
|
-
* Fix incorrectly preloading through association records when middle association has been loaded.
|
233
|
-
|
234
|
-
*Joshua Young*
|
235
|
-
|
236
|
-
* Fix where.missing and where.associated for parent/child associations.
|
237
|
-
|
238
|
-
*fatkodima*
|
239
|
-
|
240
|
-
* Fix Enumerable#in_order_of to preserve duplicates.
|
241
|
-
|
242
|
-
*fatkodima*
|
243
|
-
|
244
|
-
* Fix autoincrement on primary key for mysql.
|
245
|
-
|
246
|
-
*Eileen M. Uchitelle*
|
100
|
+
* Support composite identifiers in `to_key`
|
247
101
|
|
248
|
-
|
102
|
+
`to_key` avoids wrapping `#id` value into an `Array` if `#id` already an array
|
249
103
|
|
250
|
-
*
|
251
|
-
|
252
|
-
* Fix schema cache dumping of virtual columns.
|
253
|
-
|
254
|
-
*fatkodima*
|
104
|
+
*Nikita Vasilevsky*
|
255
105
|
|
256
|
-
*
|
106
|
+
* Add validation option for `enum`
|
257
107
|
|
258
|
-
|
108
|
+
```ruby
|
109
|
+
class Contract < ApplicationRecord
|
110
|
+
enum :status, %w[in_progress completed], validate: true
|
111
|
+
end
|
112
|
+
Contract.new(status: "unknown").valid? # => false
|
113
|
+
Contract.new(status: nil).valid? # => false
|
114
|
+
Contract.new(status: "completed").valid? # => true
|
259
115
|
|
260
|
-
|
116
|
+
class Contract < ApplicationRecord
|
117
|
+
enum :status, %w[in_progress completed], validate: { allow_nil: true }
|
118
|
+
end
|
119
|
+
Contract.new(status: "unknown").valid? # => false
|
120
|
+
Contract.new(status: nil).valid? # => true
|
121
|
+
Contract.new(status: "completed").valid? # => true
|
122
|
+
```
|
261
123
|
|
262
|
-
*
|
124
|
+
*Edem Topuzov*, *Ryuta Kamizono*
|
263
125
|
|
264
|
-
*
|
126
|
+
* Allow batching methods to use already loaded relation if available
|
265
127
|
|
266
|
-
|
128
|
+
Calling batch methods on already loaded relations will use the records previously loaded instead of retrieving
|
129
|
+
them from the database again.
|
267
130
|
|
268
|
-
*
|
131
|
+
*Adam Hess*
|
269
132
|
|
270
|
-
|
271
|
-
default precision as regular datetime columns, resulting in the following
|
272
|
-
being erroneously equivalent:
|
133
|
+
* Deprecate `read_attribute(:id)` returning the primary key if the primary key is not `:id`.
|
273
134
|
|
274
|
-
|
275
|
-
|
135
|
+
Starting in Rails 7.2, `read_attribute(:id)` will return the value of the id column, regardless of the model's
|
136
|
+
primary key. To retrieve the value of the primary key, use `#id` instead. `read_attribute(:id)` for composite
|
137
|
+
primary key models will now return the value of the id column.
|
276
138
|
|
277
|
-
|
278
|
-
datetime column default precisions match.
|
279
|
-
|
280
|
-
*Sam Bostock*
|
281
|
-
|
282
|
-
* Fix a case where the query cache can return wrong values. See #46044
|
283
|
-
|
284
|
-
*Aaron Patterson*
|
139
|
+
*Adrianna Chang*
|
285
140
|
|
141
|
+
* Fix `change_table` setting datetime precision for 6.1 Migrations
|
286
142
|
|
287
|
-
|
143
|
+
*Hartley McGuire*
|
288
144
|
|
289
|
-
*
|
145
|
+
* Fix change_column setting datetime precision for 6.1 Migrations
|
290
146
|
|
147
|
+
*Hartley McGuire*
|
291
148
|
|
292
|
-
|
149
|
+
* Add `ActiveRecord::Base#id_value` alias to access the raw value of a record's id column.
|
293
150
|
|
294
|
-
|
151
|
+
This alias is only provided for models that declare an `:id` column.
|
295
152
|
|
153
|
+
*Adrianna Chang*
|
296
154
|
|
297
|
-
|
155
|
+
* Fix previous change tracking for `ActiveRecord::Store` when using a column with JSON structured database type
|
298
156
|
|
299
|
-
|
157
|
+
Before, the methods to access the changes made during the last save `#saved_change_to_key?`, `#saved_change_to_key`, and `#key_before_last_save` did not work if the store was defined as a `store_accessor` on a column with a JSON structured database type
|
300
158
|
|
301
|
-
|
302
|
-
attempting sanitization. That sanitization could be bypassed with
|
303
|
-
carefully crafted input.
|
159
|
+
*Robert DiMartino*
|
304
160
|
|
305
|
-
|
306
|
-
occurrences of "/*" or "*/" with "/ *" or "* /". It also performs a
|
307
|
-
first pass to remove one surrounding comment to avoid compatibility
|
308
|
-
issues for users relying on the existing removal.
|
161
|
+
* Fully support `NULLS [NOT] DISTINCT` for PostgreSQL 15+ indexes.
|
309
162
|
|
310
|
-
|
311
|
-
|
163
|
+
Previous work was done to allow the index to be created in a migration, but it was not
|
164
|
+
supported in schema.rb. Additionally, the matching for `NULLS [NOT] DISTINCT` was not
|
165
|
+
in the correct order, which could have resulted in inconsistent schema detection.
|
312
166
|
|
313
|
-
|
167
|
+
*Gregory Jones*
|
314
168
|
|
315
|
-
*
|
169
|
+
* Allow escaping of literal colon characters in `sanitize_sql_*` methods when named bind variables are used
|
316
170
|
|
317
|
-
|
318
|
-
PostgreSQL will treat the column type as numeric. Comparing
|
319
|
-
integer values against numeric values can result in a slow
|
320
|
-
sequential scan.
|
171
|
+
*Justin Bull*
|
321
172
|
|
322
|
-
|
323
|
-
ActiveRecord::Base.raise_int_wider_than_64bit which defaults to true.
|
173
|
+
* Fix `#previously_new_record?` to return true for destroyed records.
|
324
174
|
|
325
|
-
|
175
|
+
Before, if a record was created and then destroyed, `#previously_new_record?` would return true.
|
176
|
+
Now, any UPDATE or DELETE to a record is considered a change, and will result in `#previously_new_record?`
|
177
|
+
returning false.
|
326
178
|
|
179
|
+
*Adrianna Chang*
|
327
180
|
|
328
|
-
|
181
|
+
* Specify callback in `has_secure_token`
|
329
182
|
|
330
|
-
|
183
|
+
```ruby
|
184
|
+
class User < ApplicationRecord
|
185
|
+
has_secure_token on: :initialize
|
186
|
+
end
|
331
187
|
|
332
|
-
|
188
|
+
User.new.token # => "abc123...."
|
189
|
+
```
|
333
190
|
|
334
|
-
*
|
191
|
+
*Sean Doyle*
|
335
192
|
|
336
|
-
|
337
|
-
which is wasteful and cause problem with YAML safe_load.
|
193
|
+
* Fix incrementation of in memory counter caches when associations overlap
|
338
194
|
|
339
|
-
|
195
|
+
When two associations had a similarly named counter cache column, Active Record
|
196
|
+
could sometime increment the wrong one.
|
340
197
|
|
341
|
-
*
|
198
|
+
*Jacopo Beschi*, *Jean Boussier*
|
342
199
|
|
343
|
-
|
200
|
+
* Don't show secrets for Active Record's `Cipher::Aes256Gcm#inspect`.
|
344
201
|
|
345
|
-
|
202
|
+
Before:
|
346
203
|
|
347
204
|
```ruby
|
348
|
-
ActiveRecord::
|
205
|
+
ActiveRecord::Encryption::Cipher::Aes256Gcm.new(secret).inspect
|
206
|
+
"#<ActiveRecord::Encryption::Cipher::Aes256Gcm:0x0000000104888038 ... @secret=\"\\xAF\\bFh]LV}q\\nl\\xB2U\\xB3 ... >"
|
349
207
|
```
|
350
208
|
|
351
|
-
|
352
|
-
|
353
|
-
* Fix supporting timezone awareness for `tsrange` and `tstzrange` array columns.
|
209
|
+
After:
|
354
210
|
|
355
211
|
```ruby
|
356
|
-
|
357
|
-
|
358
|
-
# In app config
|
359
|
-
ActiveRecord::Base.time_zone_aware_types += [:tsrange]
|
360
|
-
# In the code times are properly converted to app time zone
|
361
|
-
Shop.create!(open_hours: [Time.current..8.hour.from_now])
|
212
|
+
ActiveRecord::Encryption::Cipher::Aes256Gcm(secret).inspect
|
213
|
+
"#<ActiveRecord::Encryption::Cipher::Aes256Gcm:0x0000000104888038>"
|
362
214
|
```
|
363
215
|
|
364
|
-
*
|
365
|
-
|
366
|
-
* Resolve issue where a relation cache_version could be left stale.
|
367
|
-
|
368
|
-
Previously, when `reset` was called on a relation object it did not reset the cache_versions
|
369
|
-
ivar. This led to a confusing situation where despite having the correct data the relation
|
370
|
-
still reported a stale cache_version.
|
216
|
+
*Petrik de Heus*
|
371
217
|
|
372
|
-
|
218
|
+
* Bring back the historical behavior of committing transaction on non-local return.
|
373
219
|
|
374
220
|
```ruby
|
375
|
-
|
376
|
-
|
377
|
-
|
378
|
-
|
379
|
-
|
380
|
-
developers.cache_version # Stale cache_version
|
381
|
-
developers.reset
|
382
|
-
developers.cache_version # Returns the current correct cache_version
|
221
|
+
Model.transaction do
|
222
|
+
model.save
|
223
|
+
return
|
224
|
+
other_model.save # not executed
|
225
|
+
end
|
383
226
|
```
|
384
227
|
|
385
|
-
|
228
|
+
Historically only raised errors would trigger a rollback, but in Ruby `2.3`, the `timeout` library
|
229
|
+
started using `throw` to interrupt execution which had the adverse effect of committing open transactions.
|
386
230
|
|
387
|
-
|
231
|
+
To solve this, in Active Record 6.1 the behavior was changed to instead rollback the transaction as it was safer
|
232
|
+
than to potentially commit an incomplete transaction.
|
388
233
|
|
389
|
-
|
234
|
+
Using `return`, `break` or `throw` inside a `transaction` block was essentially deprecated from Rails 6.1 onwards.
|
390
235
|
|
391
|
-
|
392
|
-
|
236
|
+
However with the release of `timeout 0.4.0`, `Timeout.timeout` now raises an error again, and Active Record is able
|
237
|
+
to return to its original, less surprising, behavior.
|
393
238
|
|
394
|
-
|
395
|
-
|
396
|
-
|
239
|
+
This historical behavior can now be opt-ed in via:
|
240
|
+
|
241
|
+
```
|
242
|
+
Rails.application.config.active_record.commit_transaction_on_non_local_return = true
|
397
243
|
```
|
398
244
|
|
399
|
-
|
400
|
-
cause the association to be loaded.
|
245
|
+
And is the default for new applications created in Rails 7.1.
|
401
246
|
|
402
247
|
*Jean Boussier*
|
403
248
|
|
404
|
-
*
|
405
|
-
|
406
|
-
*Anmol Chopra*, *Matt Lawrence*, and *Jonathan Hefner*
|
407
|
-
|
408
|
-
* `rails db:schema:{dump,load}` now checks `ENV["SCHEMA_FORMAT"]` before config
|
409
|
-
|
410
|
-
Since `rails db:structure:{dump,load}` was deprecated there wasn't a simple
|
411
|
-
way to dump a schema to both SQL and Ruby formats. You can now do this with
|
412
|
-
an environment variable. For example:
|
413
|
-
|
414
|
-
```
|
415
|
-
SCHEMA_FORMAT=sql rake db:schema:dump
|
416
|
-
```
|
249
|
+
* Deprecate `name` argument on `#remove_connection`.
|
417
250
|
|
418
|
-
|
251
|
+
The `name` argument is deprecated on `#remove_connection` without replacement. `#remove_connection` should be called directly on the class that established the connection.
|
419
252
|
|
420
|
-
*
|
421
|
-
|
422
|
-
*edsharp*
|
253
|
+
*Eileen M. Uchitelle*
|
423
254
|
|
255
|
+
* Fix has_one through singular building with inverse.
|
424
256
|
|
425
|
-
|
257
|
+
Allows building of records from an association with a has_one through a
|
258
|
+
singular association with inverse. For belongs_to through associations,
|
259
|
+
linking the foreign key to the primary key model isn't needed.
|
260
|
+
For has_one, we cannot build records due to the association not being mutable.
|
426
261
|
|
427
|
-
*
|
262
|
+
*Gannon McGibbon*
|
428
263
|
|
429
|
-
|
430
|
-
follows:
|
264
|
+
* Disable database prepared statements when query logs are enabled
|
431
265
|
|
432
|
-
|
266
|
+
Prepared Statements and Query Logs are incompatible features due to query logs making every query unique.
|
433
267
|
|
434
|
-
|
435
|
-
"unsafe" YAML loading strategy, maintaining the existing behavior but leaving
|
436
|
-
the possible escalation vulnerability in place. Setting this option to true
|
437
|
-
is *not* recommended, but can aid in upgrading.
|
268
|
+
*zzak, Jean Boussier*
|
438
269
|
|
439
|
-
|
270
|
+
* Support decrypting data encrypted non-deterministically with a SHA1 hash digest.
|
440
271
|
|
441
|
-
|
442
|
-
|
443
|
-
application. For example, if your application uses Symbol and Time in
|
444
|
-
serialized data, you can add Symbol and Time to the allowed list as follows:
|
272
|
+
This adds a new Active Record encryption option to support decrypting data encrypted
|
273
|
+
non-deterministically with a SHA1 hash digest:
|
445
274
|
|
446
275
|
```
|
447
|
-
config.active_record.
|
276
|
+
Rails.application.config.active_record.encryption.support_sha1_for_non_deterministic_encryption = true
|
448
277
|
```
|
449
278
|
|
450
|
-
|
451
|
-
|
279
|
+
The new option addresses a problem when upgrading from 7.0 to 7.1. Due to a bug in how Active Record
|
280
|
+
Encryption was getting initialized, the key provider used for non-deterministic encryption were using
|
281
|
+
SHA-1 as its digest class, instead of the one configured globally by Rails via
|
282
|
+
`Rails.application.config.active_support.key_generator_hash_digest_class`.
|
452
283
|
|
453
|
-
|
284
|
+
*Cadu Ribeiro and Jorge Manrubia*
|
454
285
|
|
455
|
-
*
|
456
|
-
methods in class objects that referenced reloadable constants. See
|
457
|
-
[#44125](https://github.com/rails/rails/issues/44125) for details.
|
286
|
+
* Added PostgreSQL migration commands for enum rename, add value, and rename value.
|
458
287
|
|
459
|
-
|
460
|
-
|
461
|
-
|
462
|
-
|
463
|
-
Defaults would be written wrong in "db/schema.rb" and not work correctly
|
464
|
-
if using `db:schema:load`. Further more the function name would be
|
465
|
-
added as string content when saving new records.
|
466
|
-
|
467
|
-
*kaspernj*
|
468
|
-
|
469
|
-
* Fix `remove_foreign_key` with `:if_exists` option when foreign key actually exists.
|
470
|
-
|
471
|
-
*fatkodima*
|
472
|
-
|
473
|
-
* Remove `--no-comments` flag in structure dumps for PostgreSQL
|
474
|
-
|
475
|
-
This broke some apps that used custom schema comments. If you don't want
|
476
|
-
comments in your structure dump, you can use:
|
288
|
+
`rename_enum` and `rename_enum_value` are reversible. Due to Postgres
|
289
|
+
limitation, `add_enum_value` is not reversible since you cannot delete enum
|
290
|
+
values. As an alternative you should drop and recreate the enum entirely.
|
477
291
|
|
478
292
|
```ruby
|
479
|
-
|
293
|
+
rename_enum :article_status, to: :article_state
|
480
294
|
```
|
481
295
|
|
482
|
-
|
296
|
+
```ruby
|
297
|
+
add_enum_value :article_state, "archived" # will be at the end of existing values
|
298
|
+
add_enum_value :article_state, "in review", before: "published"
|
299
|
+
add_enum_value :article_state, "approved", after: "in review"
|
300
|
+
```
|
483
301
|
|
484
|
-
|
302
|
+
```ruby
|
303
|
+
rename_enum_value :article_state, from: "archived", to: "deleted"
|
304
|
+
```
|
485
305
|
|
486
|
-
|
487
|
-
parameter, instead of just `name`. This prevents unintended filtering of parameters
|
488
|
-
with a matching name in other models.
|
306
|
+
*Ray Faddis*
|
489
307
|
|
490
|
-
|
308
|
+
* Allow composite primary key to be derived from schema
|
491
309
|
|
492
|
-
|
310
|
+
Booting an application with a schema that contains composite primary keys
|
311
|
+
will not issue warning and won't `nil`ify the `ActiveRecord::Base#primary_key` value anymore.
|
493
312
|
|
494
|
-
|
313
|
+
Given a `travel_routes` table definition and a `TravelRoute` model like:
|
314
|
+
```ruby
|
315
|
+
create_table :travel_routes, primary_key: [:origin, :destination], force: true do |t|
|
316
|
+
t.string :origin
|
317
|
+
t.string :destination
|
318
|
+
end
|
495
319
|
|
496
|
-
|
320
|
+
class TravelRoute < ActiveRecord::Base; end
|
321
|
+
```
|
322
|
+
The `TravelRoute.primary_key` value will be automatically derived to `["origin", "destination"]`
|
497
323
|
|
498
|
-
*
|
324
|
+
*Nikita Vasilevsky*
|
499
325
|
|
500
|
-
|
326
|
+
* Include the `connection_pool` with exceptions raised from an adapter.
|
501
327
|
|
502
|
-
|
328
|
+
The `connection_pool` provides added context such as the connection used
|
329
|
+
that led to the exception as well as which role and shard.
|
503
330
|
|
331
|
+
*Luan Vieira*
|
504
332
|
|
505
|
-
|
333
|
+
* Support multiple column ordering for `find_each`, `find_in_batches` and `in_batches`.
|
506
334
|
|
507
|
-
|
335
|
+
When find_each/find_in_batches/in_batches are performed on a table with composite primary keys, ascending or descending order can be selected for each key.
|
508
336
|
|
337
|
+
```ruby
|
338
|
+
Person.find_each(order: [:desc, :asc]) do |person|
|
339
|
+
person.party_all_night!
|
340
|
+
end
|
341
|
+
```
|
509
342
|
|
510
|
-
|
343
|
+
*Takuya Kurimoto*
|
511
344
|
|
512
|
-
*
|
345
|
+
* Fix where on association with has_one/has_many polymorphic relations.
|
513
346
|
|
347
|
+
Before:
|
348
|
+
```ruby
|
349
|
+
Treasure.where(price_estimates: PriceEstimate.all)
|
350
|
+
#=> SELECT (...) WHERE "treasures"."id" IN (SELECT "price_estimates"."estimate_of_id" FROM "price_estimates")
|
351
|
+
```
|
514
352
|
|
515
|
-
|
353
|
+
Later:
|
354
|
+
```ruby
|
355
|
+
Treasure.where(price_estimates: PriceEstimate.all)
|
356
|
+
#=> SELECT (...) WHERE "treasures"."id" IN (SELECT "price_estimates"."estimate_of_id" FROM "price_estimates" WHERE "price_estimates"."estimate_of_type" = 'Treasure')
|
357
|
+
```
|
516
358
|
|
517
|
-
*
|
359
|
+
*Lázaro Nixon*
|
518
360
|
|
361
|
+
* Assign auto populated columns on Active Record record creation.
|
519
362
|
|
520
|
-
|
363
|
+
Changes record creation logic to allow for the `auto_increment` column to be assigned
|
364
|
+
immediately after creation regardless of it's relation to the model's primary key.
|
521
365
|
|
522
|
-
|
366
|
+
The PostgreSQL adapter benefits the most from the change allowing for any number of auto-populated
|
367
|
+
columns to be assigned on the object immediately after row insertion utilizing the `RETURNING` statement.
|
523
368
|
|
524
369
|
*Nikita Vasilevsky*
|
525
370
|
|
526
|
-
*
|
371
|
+
* Use the first key in the `shards` hash from `connected_to` for the `default_shard`.
|
527
372
|
|
528
|
-
|
529
|
-
|
530
|
-
* Dump the precision for datetime columns following the new defaults.
|
531
|
-
|
532
|
-
*Rafael Mendonça França*
|
533
|
-
|
534
|
-
* Make sure encrypted attributes are not being filtered twice.
|
535
|
-
|
536
|
-
*Nikita Vasilevsky*
|
373
|
+
Some applications may not want to use `:default` as a shard name in their connection model. Unfortunately Active Record expects there to be a `:default` shard because it must assume a shard to get the right connection from the pool manager. Rather than force applications to manually set this, `connects_to` can infer the default shard name from the hash of shards and will now assume that the first shard is your default.
|
537
374
|
|
538
|
-
|
375
|
+
For example if your model looked like this:
|
539
376
|
|
540
|
-
|
541
|
-
|
542
|
-
|
543
|
-
the production schema.
|
544
|
-
|
545
|
-
To avoid this the schema dumper will generate the new format which will include the Rails
|
546
|
-
version and will look like this:
|
377
|
+
```ruby
|
378
|
+
class ShardRecord < ApplicationRecord
|
379
|
+
self.abstract_class = true
|
547
380
|
|
381
|
+
connects_to shards: {
|
382
|
+
shard_one: { writing: :shard_one },
|
383
|
+
shard_two: { writing: :shard_two }
|
384
|
+
}
|
548
385
|
```
|
549
|
-
ActiveRecord::Schema[7.0].define
|
550
|
-
```
|
551
|
-
|
552
|
-
When upgrading from Rails 6.1 to Rails 7.0, you can run the `rails app:update` task that will
|
553
|
-
set the current schema version to 6.1.
|
554
|
-
|
555
|
-
*Rafael Mendonça França*
|
556
386
|
|
557
|
-
|
387
|
+
Then the `default_shard` for this class would be set to `shard_one`.
|
558
388
|
|
559
|
-
|
389
|
+
Fixes: #45390
|
560
390
|
|
561
|
-
*
|
562
|
-
when bulk-inserting fixtures that exceed `max_allowed_packet` configuration.
|
391
|
+
*Eileen M. Uchitelle*
|
563
392
|
|
564
|
-
|
393
|
+
* Fix mutation detection for serialized attributes backed by binary columns.
|
565
394
|
|
566
|
-
*
|
395
|
+
*Jean Boussier*
|
567
396
|
|
568
|
-
|
397
|
+
* Add `ActiveRecord.disconnect_all!` method to immediately close all connections from all pools.
|
569
398
|
|
570
|
-
*
|
399
|
+
*Jean Boussier*
|
571
400
|
|
572
|
-
|
401
|
+
* Discard connections which may have been left in a transaction.
|
573
402
|
|
574
|
-
|
403
|
+
There are cases where, due to an error, `within_new_transaction` may unexpectedly leave a connection in an open transaction. In these cases the connection may be reused, and the following may occur:
|
404
|
+
- Writes appear to fail when they actually succeed.
|
405
|
+
- Writes appear to succeed when they actually fail.
|
406
|
+
- Reads return stale or uncommitted data.
|
575
407
|
|
576
|
-
|
408
|
+
Previously, the following case was detected:
|
409
|
+
- An error is encountered during the transaction, then another error is encountered while attempting to roll it back.
|
577
410
|
|
578
|
-
|
411
|
+
Now, the following additional cases are detected:
|
412
|
+
- An error is encountered just after successfully beginning a transaction.
|
413
|
+
- An error is encountered while committing a transaction, then another error is encountered while attempting to roll it back.
|
414
|
+
- An error is encountered while rolling back a transaction.
|
579
415
|
|
580
|
-
|
416
|
+
*Nick Dower*
|
581
417
|
|
582
|
-
|
418
|
+
* Active Record query cache now evicts least recently used entries
|
583
419
|
|
584
|
-
|
420
|
+
By default it only keeps the `100` most recently used queries.
|
585
421
|
|
586
|
-
|
422
|
+
The cache size can be configured via `database.yml`
|
587
423
|
|
424
|
+
```yaml
|
425
|
+
development:
|
426
|
+
adapter: mysql2
|
427
|
+
query_cache: 200
|
428
|
+
```
|
588
429
|
|
589
|
-
|
430
|
+
It can also be entirely disabled:
|
590
431
|
|
432
|
+
```yaml
|
433
|
+
development:
|
434
|
+
adapter: mysql2
|
435
|
+
query_cache: false
|
436
|
+
```
|
591
437
|
|
592
|
-
*
|
438
|
+
*Jean Boussier*
|
593
439
|
|
594
|
-
|
440
|
+
* Deprecate `check_pending!` in favor of `check_all_pending!`.
|
595
441
|
|
596
|
-
|
442
|
+
`check_pending!` will only check for pending migrations on the current database connection or the one passed in. This has been deprecated in favor of `check_all_pending!` which will find all pending migrations for the database configurations in a given environment.
|
597
443
|
|
598
|
-
*
|
444
|
+
*Eileen M. Uchitelle*
|
599
445
|
|
600
|
-
|
446
|
+
* Make `increment_counter`/`decrement_counter` accept an amount argument
|
601
447
|
|
602
448
|
```ruby
|
603
|
-
|
449
|
+
Post.increment_counter(:comments_count, 5, by: 3)
|
604
450
|
```
|
605
451
|
|
606
|
-
|
607
|
-
|
608
|
-
*Oliver Günther*
|
452
|
+
*fatkodima*
|
609
453
|
|
610
|
-
*
|
454
|
+
* Add support for `Array#intersect?` to `ActiveRecord::Relation`.
|
611
455
|
|
612
|
-
|
456
|
+
`Array#intersect?` is only available on Ruby 3.1 or later.
|
613
457
|
|
614
|
-
|
458
|
+
This allows the Rubocop `Style/ArrayIntersect` cop to work with `ActiveRecord::Relation` objects.
|
615
459
|
|
616
|
-
*
|
460
|
+
*John Harry Kelly*
|
617
461
|
|
618
|
-
*
|
462
|
+
* The deferrable foreign key can be passed to `t.references`.
|
619
463
|
|
620
|
-
*
|
464
|
+
*Hiroyuki Ishii*
|
621
465
|
|
622
|
-
*
|
466
|
+
* Deprecate `deferrable: true` option of `add_foreign_key`.
|
623
467
|
|
624
|
-
|
468
|
+
`deferrable: true` is deprecated in favor of `deferrable: :immediate`, and
|
469
|
+
will be removed in Rails 7.2.
|
625
470
|
|
626
|
-
|
471
|
+
Because `deferrable: true` and `deferrable: :deferred` are hard to understand.
|
472
|
+
Both true and :deferred are truthy values.
|
473
|
+
This behavior is the same as the deferrable option of the add_unique_key method, added in #46192.
|
627
474
|
|
628
|
-
*
|
475
|
+
*Hiroyuki Ishii*
|
629
476
|
|
630
|
-
|
477
|
+
* `AbstractAdapter#execute` and `#exec_query` now clear the query cache
|
631
478
|
|
632
|
-
|
479
|
+
If you need to perform a read only SQL query without clearing the query
|
480
|
+
cache, use `AbstractAdapter#select_all`.
|
633
481
|
|
634
|
-
*
|
482
|
+
*Jean Boussier*
|
635
483
|
|
636
|
-
*
|
484
|
+
* Make `.joins` / `.left_outer_joins` work with CTEs.
|
637
485
|
|
638
|
-
|
486
|
+
For example:
|
639
487
|
|
640
|
-
|
488
|
+
```ruby
|
489
|
+
Post
|
490
|
+
.with(commented_posts: Comment.select(:post_id).distinct)
|
491
|
+
.joins(:commented_posts)
|
492
|
+
#=> WITH (...) SELECT ... INNER JOIN commented_posts on posts.id = commented_posts.post_id
|
493
|
+
```
|
641
494
|
|
642
|
-
|
643
|
-
bigint instead of integer for the SQLite Adapter.
|
495
|
+
*Vladimir Dementyev*
|
644
496
|
|
645
|
-
|
497
|
+
* Add a load hook for `ActiveRecord::ConnectionAdapters::Mysql2Adapter`
|
498
|
+
(named `active_record_mysql2adapter`) to allow for overriding aspects of the
|
499
|
+
`ActiveRecord::ConnectionAdapters::Mysql2Adapter` class. This makes `Mysql2Adapter`
|
500
|
+
consistent with `PostgreSQLAdapter` and `SQLite3Adapter` that already have load hooks.
|
646
501
|
|
647
|
-
*
|
502
|
+
*fatkodima*
|
648
503
|
|
649
|
-
|
504
|
+
* Introduce adapter for Trilogy database client
|
650
505
|
|
651
|
-
|
506
|
+
Trilogy is a MySQL-compatible database client. Rails applications can use Trilogy
|
507
|
+
by configuring their `config/database.yml`:
|
652
508
|
|
653
|
-
```
|
654
|
-
|
509
|
+
```yaml
|
510
|
+
development:
|
511
|
+
adapter: trilogy
|
512
|
+
database: blog_development
|
513
|
+
pool: 5
|
655
514
|
```
|
656
515
|
|
657
|
-
|
658
|
-
value is still ordered.
|
659
|
-
|
660
|
-
*Jean Boussier*
|
516
|
+
Or by using the `DATABASE_URL` environment variable:
|
661
517
|
|
662
|
-
|
518
|
+
```ruby
|
519
|
+
ENV['DATABASE_URL'] # => "trilogy://localhost/blog_development?pool=5"
|
520
|
+
```
|
663
521
|
|
664
|
-
*
|
522
|
+
*Adrianna Chang*
|
665
523
|
|
666
|
-
*
|
667
|
-
|
668
|
-
Since the alias is derived from the table name, we can't assume the result
|
669
|
-
is a valid identifier.
|
524
|
+
* `after_commit` callbacks defined on models now execute in the correct order.
|
670
525
|
|
671
526
|
```ruby
|
672
|
-
class
|
673
|
-
|
527
|
+
class User < ActiveRecord::Base
|
528
|
+
after_commit { puts("this gets called first") }
|
529
|
+
after_commit { puts("this gets called second") }
|
674
530
|
end
|
675
|
-
Test.group(:id).count
|
676
|
-
# syntax error at or near "1" (ActiveRecord::StatementInvalid)
|
677
|
-
# LINE 1: SELECT COUNT(*) AS count_all, "1abc"."id" AS 1abc_id FROM "1...
|
678
531
|
```
|
679
532
|
|
680
|
-
|
533
|
+
Previously, the callbacks executed in the reverse order. To opt in to the new behaviour:
|
534
|
+
|
535
|
+
```ruby
|
536
|
+
config.active_record.run_after_transaction_callbacks_in_order_defined = true
|
537
|
+
```
|
681
538
|
|
539
|
+
This is the default for new apps.
|
682
540
|
|
683
|
-
|
541
|
+
*Alex Ghiculescu*
|
684
542
|
|
685
|
-
*
|
543
|
+
* Infer `foreign_key` when `inverse_of` is present on `has_one` and `has_many` associations.
|
686
544
|
|
687
545
|
```ruby
|
688
|
-
|
546
|
+
has_many :citations, foreign_key: "book1_id", inverse_of: :book
|
689
547
|
```
|
690
548
|
|
691
|
-
|
692
|
-
responsible to detect write queries.
|
549
|
+
can be simplified to
|
693
550
|
|
694
|
-
|
695
|
-
|
696
|
-
|
697
|
-
*Jean Boussier*
|
551
|
+
```ruby
|
552
|
+
has_many :citations, inverse_of: :book
|
553
|
+
```
|
698
554
|
|
699
|
-
|
555
|
+
and the foreign_key will be read from the corresponding `belongs_to` association.
|
700
556
|
|
701
|
-
|
557
|
+
*Daniel Whitney*
|
702
558
|
|
703
|
-
|
559
|
+
* Limit max length of auto generated index names
|
704
560
|
|
561
|
+
Auto generated index names are now limited to 62 bytes, which fits within
|
562
|
+
the default index name length limits for MySQL, Postgres and SQLite.
|
705
563
|
|
706
|
-
|
564
|
+
Any index name over the limit will fallback to the new short format.
|
707
565
|
|
708
|
-
|
566
|
+
Before (too long):
|
567
|
+
```
|
568
|
+
index_testings_on_foo_and_bar_and_first_name_and_last_name_and_administrator
|
569
|
+
```
|
709
570
|
|
571
|
+
After (short format):
|
572
|
+
```
|
573
|
+
idx_on_foo_bar_first_name_last_name_administrator_5939248142
|
574
|
+
```
|
710
575
|
|
711
|
-
|
576
|
+
The short format includes a hash to ensure the name is unique database-wide.
|
712
577
|
|
713
|
-
*
|
578
|
+
*Mike Coutermarsh*
|
714
579
|
|
580
|
+
* Introduce a more stable and optimized Marshal serializer for Active Record models.
|
715
581
|
|
716
|
-
|
582
|
+
Can be enabled with `config.active_record.marshalling_format_version = 7.1`.
|
717
583
|
|
718
|
-
*
|
584
|
+
*Jean Boussier*
|
719
585
|
|
720
|
-
|
586
|
+
* Allow specifying where clauses with column-tuple syntax.
|
721
587
|
|
722
|
-
|
588
|
+
Querying through `#where` now accepts a new tuple-syntax which accepts, as
|
589
|
+
a key, an array of columns and, as a value, an array of corresponding tuples.
|
590
|
+
The key specifies a list of columns, while the value is an array of
|
591
|
+
ordered-tuples that conform to the column list.
|
723
592
|
|
724
|
-
|
593
|
+
For instance:
|
725
594
|
|
726
|
-
|
595
|
+
```ruby
|
596
|
+
# Cpk::Book => Cpk::Book(author_id: integer, number: integer, title: string, revision: integer)
|
597
|
+
# Cpk::Book.primary_key => ["author_id", "number"]
|
727
598
|
|
728
|
-
|
599
|
+
book = Cpk::Book.create!(author_id: 1, number: 1)
|
600
|
+
Cpk::Book.where(Cpk::Book.primary_key => [[1, 2]]) # => [book]
|
729
601
|
|
730
|
-
|
602
|
+
# Topic => Topic(id: integer, title: string, author_name: string...)
|
731
603
|
|
732
|
-
|
604
|
+
Topic.where([:title, :author_name] => [["The Alchemist", "Paul Coelho"], ["Harry Potter", "J.K Rowling"]])
|
605
|
+
```
|
733
606
|
|
734
|
-
*
|
607
|
+
*Paarth Madan*
|
735
608
|
|
736
|
-
|
737
|
-
Single Table Inheritance. This affects enums defined in subclasses, previously
|
738
|
-
the value of these fields was not parsed and remained `nil`
|
609
|
+
* Allow warning codes to be ignore when reporting SQL warnings.
|
739
610
|
|
740
|
-
|
611
|
+
Active Record config that can ignore warning codes
|
741
612
|
|
742
|
-
|
613
|
+
```ruby
|
614
|
+
# Configure allowlist of warnings that should always be ignored
|
615
|
+
config.active_record.db_warnings_ignore = [
|
616
|
+
"1062", # MySQL Error 1062: Duplicate entry
|
617
|
+
]
|
618
|
+
```
|
743
619
|
|
744
|
-
|
620
|
+
This is supported for the MySQL and PostgreSQL adapters.
|
745
621
|
|
746
|
-
*
|
622
|
+
*Nick Borromeo*
|
747
623
|
|
748
|
-
|
624
|
+
* Introduce `:active_record_fixtures` lazy load hook.
|
749
625
|
|
750
|
-
|
626
|
+
Hooks defined with this name will be run whenever `TestFixtures` is included
|
627
|
+
in a class.
|
751
628
|
|
752
629
|
```ruby
|
753
|
-
|
754
|
-
|
630
|
+
ActiveSupport.on_load(:active_record_fixtures) do
|
631
|
+
self.fixture_paths << "test/fixtures"
|
755
632
|
end
|
756
633
|
|
757
|
-
|
758
|
-
|
759
|
-
|
760
|
-
*Alexandre Ruban*
|
634
|
+
klass = Class.new
|
635
|
+
klass.include(ActiveRecord::TestFixtures)
|
761
636
|
|
762
|
-
|
637
|
+
klass.fixture_paths # => ["test/fixtures"]
|
638
|
+
```
|
763
639
|
|
764
|
-
*
|
640
|
+
*Andrew Novoselac*
|
765
641
|
|
766
|
-
*
|
642
|
+
* Introduce `TestFixtures#fixture_paths`.
|
767
643
|
|
768
|
-
|
644
|
+
Multiple fixture paths can now be specified using the `#fixture_paths` accessor.
|
645
|
+
Apps will continue to have `test/fixtures` as their one fixture path by default,
|
646
|
+
but additional fixture paths can be specified.
|
769
647
|
|
770
648
|
```ruby
|
771
|
-
|
772
|
-
|
773
|
-
{ id: 2, name: "Copper", price: 4.84 },
|
774
|
-
{ id: 4, name: "Gold", price: 1380.87 },
|
775
|
-
{ id: 6, name: "Aluminium", price: 0.35 }
|
776
|
-
],
|
777
|
-
update_only: [:price] # Only prices will be updated
|
778
|
-
)
|
649
|
+
ActiveSupport::TestCase.fixture_paths << "component1/test/fixtures"
|
650
|
+
ActiveSupport::TestCase.fixture_paths << "component2/test/fixtures"
|
779
651
|
```
|
780
652
|
|
781
|
-
|
653
|
+
`TestFixtures#fixture_path` is now deprecated.
|
782
654
|
|
783
|
-
*
|
784
|
-
|
785
|
-
*Rafael Mendonça França*
|
655
|
+
*Andrew Novoselac*
|
786
656
|
|
787
|
-
*
|
657
|
+
* Adds support for deferrable exclude constraints in PostgreSQL.
|
788
658
|
|
789
|
-
|
659
|
+
By default, exclude constraints in PostgreSQL are checked after each statement.
|
660
|
+
This works for most use cases, but becomes a major limitation when replacing
|
661
|
+
records with overlapping ranges by using multiple statements.
|
790
662
|
|
791
|
-
|
663
|
+
```ruby
|
664
|
+
exclusion_constraint :users, "daterange(valid_from, valid_to) WITH &&", deferrable: :immediate
|
665
|
+
```
|
792
666
|
|
793
|
-
|
667
|
+
Passing `deferrable: :immediate` checks constraint after each statement,
|
668
|
+
but allows manually deferring the check using `SET CONSTRAINTS ALL DEFERRED`
|
669
|
+
within a transaction. This will cause the excludes to be checked after the transaction.
|
794
670
|
|
795
|
-
|
671
|
+
It's also possible to change the default behavior from an immediate check
|
672
|
+
(after the statement), to a deferred check (after the transaction):
|
796
673
|
|
797
|
-
|
674
|
+
```ruby
|
675
|
+
exclusion_constraint :users, "daterange(valid_from, valid_to) WITH &&", deferrable: :deferred
|
676
|
+
```
|
798
677
|
|
799
|
-
*
|
678
|
+
*Hiroyuki Ishii*
|
800
679
|
|
801
|
-
|
680
|
+
* Respect `foreign_type` option to `delegated_type` for `{role}_class` method.
|
802
681
|
|
803
|
-
|
682
|
+
Usage of `delegated_type` with non-conventional `{role}_type` column names can now be specified with `foreign_type` option.
|
683
|
+
This option is the same as `foreign_type` as forwarded to the underlying `belongs_to` association that `delegated_type` wraps.
|
804
684
|
|
805
|
-
|
685
|
+
*Jason Karns*
|
806
686
|
|
807
|
-
|
687
|
+
* Add support for unique constraints (PostgreSQL-only).
|
808
688
|
|
689
|
+
```ruby
|
690
|
+
add_unique_key :sections, [:position], deferrable: :deferred, name: "unique_section_position"
|
691
|
+
remove_unique_key :sections, name: "unique_section_position"
|
809
692
|
```
|
810
|
-
# Before:
|
811
|
-
Foo Load (0.2ms) SELECT "foos".* FROM "foos" WHERE "foos"."passw" = ? LIMIT ? [["passw", "hello"], ["LIMIT", 1]]
|
812
693
|
|
813
|
-
#
|
814
|
-
|
694
|
+
See PostgreSQL's [Unique Constraints](https://www.postgresql.org/docs/current/ddl-constraints.html#DDL-CONSTRAINTS-UNIQUE-CONSTRAINTS) documentation for more on unique constraints.
|
695
|
+
|
696
|
+
By default, unique constraints in PostgreSQL are checked after each statement.
|
697
|
+
This works for most use cases, but becomes a major limitation when replacing
|
698
|
+
records with unique column by using multiple statements.
|
699
|
+
|
700
|
+
An example of swapping unique columns between records.
|
701
|
+
|
702
|
+
```ruby
|
703
|
+
# position is unique column
|
704
|
+
old_item = Item.create!(position: 1)
|
705
|
+
new_item = Item.create!(position: 2)
|
706
|
+
|
707
|
+
Item.transaction do
|
708
|
+
old_item.update!(position: 2)
|
709
|
+
new_item.update!(position: 1)
|
710
|
+
end
|
815
711
|
```
|
816
712
|
|
817
|
-
|
713
|
+
Using the default behavior, the transaction would fail when executing the
|
714
|
+
first `UPDATE` statement.
|
818
715
|
|
819
|
-
|
716
|
+
By passing the `:deferrable` option to the `add_unique_key` statement in
|
717
|
+
migrations, it's possible to defer this check.
|
820
718
|
|
821
|
-
|
719
|
+
```ruby
|
720
|
+
add_unique_key :items, [:position], deferrable: :immediate
|
721
|
+
```
|
822
722
|
|
823
|
-
|
723
|
+
Passing `deferrable: :immediate` does not change the behaviour of the previous example,
|
724
|
+
but allows manually deferring the check using `SET CONSTRAINTS ALL DEFERRED` within a transaction.
|
725
|
+
This will cause the unique constraints to be checked after the transaction.
|
824
726
|
|
825
|
-
|
727
|
+
It's also possible to adjust the default behavior from an immediate
|
728
|
+
check (after the statement), to a deferred check (after the transaction):
|
826
729
|
|
827
|
-
|
730
|
+
```ruby
|
731
|
+
add_unique_key :items, [:position], deferrable: :deferred
|
732
|
+
```
|
828
733
|
|
829
|
-
|
734
|
+
If you want to change an existing unique index to deferrable, you can use :using_index
|
735
|
+
to create deferrable unique constraints.
|
830
736
|
|
831
|
-
|
737
|
+
```ruby
|
738
|
+
add_unique_key :items, deferrable: :deferred, using_index: "index_items_on_position"
|
739
|
+
```
|
832
740
|
|
833
|
-
*
|
741
|
+
*Hiroyuki Ishii*
|
834
742
|
|
835
|
-
* Remove deprecated `Tasks::DatabaseTasks.
|
743
|
+
* Remove deprecated `Tasks::DatabaseTasks.schema_file_type`.
|
836
744
|
|
837
745
|
*Rafael Mendonça França*
|
838
746
|
|
839
|
-
* Remove deprecated `
|
747
|
+
* Remove deprecated `config.active_record.partial_writes`.
|
840
748
|
|
841
749
|
*Rafael Mendonça França*
|
842
750
|
|
843
|
-
*
|
844
|
-
and will be consistently replaced by the latter condition.
|
845
|
-
|
846
|
-
```ruby
|
847
|
-
# Rails 6.1 (IN clause is replaced by merger side equality condition)
|
848
|
-
Author.where(id: [david.id, mary.id]).merge(Author.where(id: bob)) # => [bob]
|
849
|
-
# Rails 6.1 (both conflict conditions exists, deprecated)
|
850
|
-
Author.where(id: david.id..mary.id).merge(Author.where(id: bob)) # => []
|
851
|
-
# Rails 6.1 with rewhere to migrate to Rails 7.0's behavior
|
852
|
-
Author.where(id: david.id..mary.id).merge(Author.where(id: bob), rewhere: true) # => [bob]
|
853
|
-
# Rails 7.0 (same behavior with IN clause, mergee side condition is consistently replaced)
|
854
|
-
Author.where(id: [david.id, mary.id]).merge(Author.where(id: bob)) # => [bob]
|
855
|
-
Author.where(id: david.id..mary.id).merge(Author.where(id: bob)) # => [bob]
|
751
|
+
* Remove deprecated `ActiveRecord::Base` config accessors.
|
856
752
|
|
857
753
|
*Rafael Mendonça França*
|
858
754
|
|
859
|
-
* Remove
|
755
|
+
* Remove the `:include_replicas` argument from `configs_for`. Use `:include_hidden` argument instead.
|
860
756
|
|
861
|
-
*
|
757
|
+
*Eileen M. Uchitelle*
|
862
758
|
|
863
|
-
*
|
759
|
+
* Allow applications to lookup a config via a custom hash key.
|
864
760
|
|
865
|
-
|
866
|
-
* `db:structure:dump`
|
867
|
-
* `db:structure:load`
|
868
|
-
* `db:structure:load_if_sql`
|
869
|
-
* `db:structure:dump:#{name}`
|
870
|
-
* `db:structure:load:#{name}`
|
871
|
-
* `db:test:load_structure`
|
872
|
-
* `db:test:load_structure:#{name}`
|
761
|
+
If you have registered a custom config or want to find configs where the hash matches a specific key, now you can pass `config_key` to `configs_for`. For example if you have a `db_config` with the key `vitess` you can look up a database configuration hash by matching that key.
|
873
762
|
|
874
|
-
|
763
|
+
```ruby
|
764
|
+
ActiveRecord::Base.configurations.configs_for(env_name: "development", name: "primary", config_key: :vitess)
|
765
|
+
ActiveRecord::Base.configurations.configs_for(env_name: "development", config_key: :vitess)
|
766
|
+
```
|
767
|
+
|
768
|
+
*Eileen M. Uchitelle*
|
875
769
|
|
876
|
-
*
|
770
|
+
* Allow applications to register a custom database configuration handler.
|
877
771
|
|
878
|
-
|
772
|
+
Adds a mechanism for registering a custom handler for cases where you want database configurations to respond to custom methods. This is useful for non-Rails database adapters or tools like Vitess that you may want to configure differently from a standard `HashConfig` or `UrlConfig`.
|
879
773
|
|
880
|
-
|
774
|
+
Given the following database YAML we want the `animals` db to create a `CustomConfig` object instead while the `primary` database will be a `UrlConfig`:
|
881
775
|
|
882
|
-
|
776
|
+
```yaml
|
777
|
+
development:
|
778
|
+
primary:
|
779
|
+
url: postgres://localhost/primary
|
780
|
+
animals:
|
781
|
+
url: postgres://localhost/animals
|
782
|
+
custom_config:
|
783
|
+
sharded: 1
|
784
|
+
```
|
883
785
|
|
884
|
-
|
885
|
-
to be committed, so in order to avoid this mistake, the transaction block is rolled back.
|
786
|
+
To register a custom handler first make a class that has your custom methods:
|
886
787
|
|
887
|
-
|
788
|
+
```ruby
|
789
|
+
class CustomConfig < ActiveRecord::DatabaseConfigurations::UrlConfig
|
790
|
+
def sharded?
|
791
|
+
custom_config.fetch("sharded", false)
|
792
|
+
end
|
888
793
|
|
889
|
-
|
794
|
+
private
|
795
|
+
def custom_config
|
796
|
+
configuration_hash.fetch(:custom_config)
|
797
|
+
end
|
798
|
+
end
|
799
|
+
```
|
890
800
|
|
891
|
-
|
801
|
+
Then register the config in an initializer:
|
892
802
|
|
893
803
|
```ruby
|
894
|
-
|
895
|
-
|
896
|
-
|
897
|
-
|
898
|
-
}
|
804
|
+
ActiveRecord::DatabaseConfigurations.register_db_config_handler do |env_name, name, url, config|
|
805
|
+
next unless config.key?(:custom_config)
|
806
|
+
CustomConfig.new(env_name, name, url, config)
|
807
|
+
end
|
899
808
|
```
|
900
809
|
|
901
|
-
|
810
|
+
When the application is booted, configuration hashes with the `:custom_config` key will be `CustomConfig` objects and respond to `sharded?`. Applications must handle the condition in which Active Record should use their custom handler.
|
902
811
|
|
903
|
-
*Eileen M. Uchitelle
|
812
|
+
*Eileen M. Uchitelle and John Crepezzi*
|
904
813
|
|
905
|
-
*
|
814
|
+
* `ActiveRecord::Base.serialize` no longer uses YAML by default.
|
906
815
|
|
907
|
-
|
816
|
+
YAML isn't particularly performant and can lead to security issues
|
817
|
+
if not used carefully.
|
908
818
|
|
909
|
-
|
819
|
+
Unfortunately there isn't really any good serializers in Ruby's stdlib
|
820
|
+
to replace it.
|
910
821
|
|
911
|
-
|
822
|
+
The obvious choice would be JSON, which is a fine format for this use case,
|
823
|
+
however the JSON serializer in Ruby's stdlib isn't strict enough, as it fallback
|
824
|
+
to casting unknown types to strings, which could lead to corrupted data.
|
912
825
|
|
913
|
-
|
826
|
+
Some third party JSON libraries like `Oj` have a suitable strict mode.
|
914
827
|
|
915
|
-
|
828
|
+
So it's preferable that users choose a serializer based on their own constraints.
|
916
829
|
|
917
|
-
|
830
|
+
The original default can be restored by setting `config.active_record.default_column_serializer = YAML`.
|
918
831
|
|
919
|
-
*
|
832
|
+
*Jean Boussier*
|
920
833
|
|
921
|
-
*
|
834
|
+
* `ActiveRecord::Base.serialize` signature changed.
|
922
835
|
|
923
|
-
|
836
|
+
Rather than a single positional argument that accepts two possible
|
837
|
+
types of values, `serialize` now accepts two distinct keyword arguments.
|
924
838
|
|
925
|
-
|
839
|
+
Before:
|
926
840
|
|
927
841
|
```ruby
|
928
|
-
|
842
|
+
serialize :content, JSON
|
843
|
+
serialize :backtrace, Array
|
929
844
|
```
|
930
845
|
|
931
|
-
|
932
|
-
|
933
|
-
* Remove deprecated support to YAML load `ActiveRecord::Base` instance in the Rails 4.2 and 4.1 formats.
|
846
|
+
After:
|
934
847
|
|
935
|
-
|
848
|
+
```ruby
|
849
|
+
serialize :content, coder: JSON
|
850
|
+
serialize :backtrace, type: Array
|
851
|
+
```
|
936
852
|
|
937
|
-
*
|
853
|
+
*Jean Boussier*
|
938
854
|
|
939
|
-
|
855
|
+
* YAML columns use `YAML.safe_dump` is available.
|
940
856
|
|
941
|
-
|
857
|
+
As of `psych 5.1.0`, `YAML.safe_dump` can now apply the same permitted
|
858
|
+
types restrictions than `YAML.safe_load`.
|
942
859
|
|
943
|
-
|
860
|
+
It's preferable to ensure the payload only use allowed types when we first
|
861
|
+
try to serialize it, otherwise you may end up with invalid records in the
|
862
|
+
database.
|
944
863
|
|
945
|
-
*
|
864
|
+
*Jean Boussier*
|
946
865
|
|
947
|
-
|
866
|
+
* `ActiveRecord::QueryLogs` better handle broken encoding.
|
948
867
|
|
949
|
-
|
868
|
+
It's not uncommon when building queries with BLOB fields to contain
|
869
|
+
binary data. Unless the call carefully encode the string in ASCII-8BIT
|
870
|
+
it generally end up being encoded in `UTF-8`, and `QueryLogs` would
|
871
|
+
end up failing on it.
|
950
872
|
|
951
|
-
|
873
|
+
`ActiveRecord::QueryLogs` no longer depend on the query to be properly encoded.
|
952
874
|
|
953
|
-
*
|
875
|
+
*Jean Boussier*
|
954
876
|
|
955
|
-
*
|
877
|
+
* Fix a bug where `ActiveRecord::Generators::ModelGenerator` would not respect create_table_migration template overrides.
|
956
878
|
|
957
|
-
```
|
958
|
-
|
959
|
-
|
960
|
-
|
961
|
-
|
879
|
+
```
|
880
|
+
rails g model create_books title:string content:text
|
881
|
+
```
|
882
|
+
will now read from the create_table_migration.rb.tt template in the following locations in order:
|
883
|
+
```
|
884
|
+
lib/templates/active_record/model/create_table_migration.rb
|
885
|
+
lib/templates/active_record/migration/create_table_migration.rb
|
886
|
+
```
|
962
887
|
|
963
|
-
|
964
|
-
# => #<Entry:0x00>
|
965
|
-
# id: 1
|
966
|
-
# entryable_id: 1,
|
967
|
-
# entryable_type: 'Message'
|
968
|
-
# ...>
|
888
|
+
*Spencer Neste*
|
969
889
|
|
970
|
-
|
971
|
-
# => #<Message:0x01>
|
972
|
-
# id: 1
|
973
|
-
# content: 'Hello world'
|
974
|
-
# ...>
|
975
|
-
```
|
890
|
+
* `ActiveRecord::Relation#explain` now accepts options.
|
976
891
|
|
977
|
-
|
892
|
+
For databases and adapters which support them (currently PostgreSQL
|
893
|
+
and MySQL), options can be passed to `explain` to provide more
|
894
|
+
detailed query plan analysis:
|
978
895
|
|
979
896
|
```ruby
|
980
|
-
|
981
|
-
# ArgumentError: Cannot build association `entryable'. Are you trying to build a polymorphic one-to-one association?
|
897
|
+
Customer.where(id: 1).joins(:orders).explain(:analyze, :verbose)
|
982
898
|
```
|
983
899
|
|
984
|
-
*
|
985
|
-
|
986
|
-
* Use subquery for DELETE with GROUP_BY and HAVING clauses.
|
900
|
+
*Reid Lynch*
|
987
901
|
|
988
|
-
|
902
|
+
* Multiple `Arel::Nodes::SqlLiteral` nodes can now be added together to
|
903
|
+
form `Arel::Nodes::Fragments` nodes. This allows joining several pieces
|
904
|
+
of SQL.
|
989
905
|
|
990
|
-
|
906
|
+
*Matthew Draper*, *Ole Friis*
|
991
907
|
|
992
|
-
|
993
|
-
DELETE FROM "posts" WHERE "posts"."id" IN (
|
994
|
-
SELECT "posts"."id" FROM "posts" INNER JOIN "comments" ON "comments"."post_id" = "posts"."id" GROUP BY "posts"."id" HAVING (count(comments.id) >= 2))
|
995
|
-
) [["flagged", "t"]]
|
996
|
-
```
|
908
|
+
* `ActiveRecord::Base#signed_id` raises if called on a new record.
|
997
909
|
|
998
|
-
|
910
|
+
Previously it would return an ID that was not usable, since it was based on `id = nil`.
|
999
911
|
|
1000
|
-
*
|
912
|
+
*Alex Ghiculescu*
|
1001
913
|
|
1002
|
-
|
914
|
+
* Allow SQL warnings to be reported.
|
1003
915
|
|
1004
|
-
|
1005
|
-
UPDATE "posts" SET "flagged" = ? WHERE "posts"."id" IN (
|
1006
|
-
SELECT "posts"."id" FROM "posts" INNER JOIN "comments" ON "comments"."post_id" = "posts"."id"
|
1007
|
-
) [["flagged", "t"]]
|
1008
|
-
```
|
916
|
+
Active Record configs can be set to enable SQL warning reporting.
|
1009
917
|
|
1010
|
-
|
918
|
+
```ruby
|
919
|
+
# Configure action to take when SQL query produces warning
|
920
|
+
config.active_record.db_warnings_action = :raise
|
1011
921
|
|
1012
|
-
|
1013
|
-
|
1014
|
-
|
1015
|
-
|
1016
|
-
|
922
|
+
# Configure allowlist of warnings that should always be ignored
|
923
|
+
config.active_record.db_warnings_ignore = [
|
924
|
+
/Invalid utf8mb4 character string/,
|
925
|
+
"An exact warning message",
|
926
|
+
]
|
1017
927
|
```
|
1018
928
|
|
1019
|
-
|
929
|
+
This is supported for the MySQL and PostgreSQL adapters.
|
1020
930
|
|
1021
|
-
*
|
931
|
+
*Adrianna Chang*, *Paarth Madan*
|
1022
932
|
|
1023
|
-
|
933
|
+
* Add `#regroup` query method as a short-hand for `.unscope(:group).group(fields)`
|
1024
934
|
|
1025
|
-
|
1026
|
-
|
1027
|
-
|
1028
|
-
|
1029
|
-
|
1030
|
-
animals:
|
1031
|
-
database: animals_db
|
1032
|
-
schema_dump: false
|
935
|
+
Example:
|
936
|
+
|
937
|
+
```ruby
|
938
|
+
Post.group(:title).regroup(:author)
|
939
|
+
# SELECT `posts`.`*` FROM `posts` GROUP BY `posts`.`author`
|
1033
940
|
```
|
1034
941
|
|
1035
|
-
|
942
|
+
*Danielius Visockas*
|
1036
943
|
|
1037
|
-
|
944
|
+
* PostgreSQL adapter method `enable_extension` now allows parameter to be `[schema_name.]<extension_name>`
|
945
|
+
if the extension must be installed on another schema.
|
1038
946
|
|
1039
|
-
|
947
|
+
Example: `enable_extension('heroku_ext.hstore')`
|
1040
948
|
|
1041
|
-
*
|
949
|
+
*Leonardo Luarte*
|
1042
950
|
|
1043
|
-
*
|
951
|
+
* Add `:include` option to `add_index`.
|
1044
952
|
|
1045
|
-
|
953
|
+
Add support for including non-key columns in indexes for PostgreSQL
|
954
|
+
with the `INCLUDE` parameter.
|
1046
955
|
|
1047
|
-
|
956
|
+
```ruby
|
957
|
+
add_index(:users, :email, include: [:id, :created_at])
|
958
|
+
```
|
1048
959
|
|
1049
|
-
|
960
|
+
will result in:
|
1050
961
|
|
1051
|
-
|
962
|
+
```sql
|
963
|
+
CREATE INDEX index_users_on_email USING btree (email) INCLUDE (id, created_at)
|
964
|
+
```
|
1052
965
|
|
1053
|
-
*
|
966
|
+
*Steve Abrams*
|
1054
967
|
|
1055
|
-
|
968
|
+
* `ActiveRecord::Relation`’s `#any?`, `#none?`, and `#one?` methods take an optional pattern
|
969
|
+
argument, more closely matching their `Enumerable` equivalents.
|
1056
970
|
|
1057
|
-
|
971
|
+
*George Claghorn*
|
1058
972
|
|
1059
|
-
|
973
|
+
* Add `ActiveRecord::Base.normalizes` for declaring attribute normalizations.
|
1060
974
|
|
1061
|
-
|
975
|
+
An attribute normalization is applied when the attribute is assigned or
|
976
|
+
updated, and the normalized value will be persisted to the database. The
|
977
|
+
normalization is also applied to the corresponding keyword argument of query
|
978
|
+
methods, allowing records to be queried using unnormalized values.
|
1062
979
|
|
1063
|
-
|
980
|
+
For example:
|
1064
981
|
|
1065
|
-
|
982
|
+
```ruby
|
983
|
+
class User < ActiveRecord::Base
|
984
|
+
normalizes :email, with: -> email { email.strip.downcase }
|
985
|
+
normalizes :phone, with: -> phone { phone.delete("^0-9").delete_prefix("1") }
|
986
|
+
end
|
1066
987
|
|
1067
|
-
|
988
|
+
user = User.create(email: " CRUISE-CONTROL@EXAMPLE.COM\n")
|
989
|
+
user.email # => "cruise-control@example.com"
|
1068
990
|
|
1069
|
-
|
991
|
+
user = User.find_by(email: "\tCRUISE-CONTROL@EXAMPLE.COM ")
|
992
|
+
user.email # => "cruise-control@example.com"
|
993
|
+
user.email_before_type_cast # => "cruise-control@example.com"
|
1070
994
|
|
1071
|
-
|
995
|
+
User.where(email: "\tCRUISE-CONTROL@EXAMPLE.COM ").count # => 1
|
996
|
+
User.where(["email = ?", "\tCRUISE-CONTROL@EXAMPLE.COM "]).count # => 0
|
1072
997
|
|
1073
|
-
|
998
|
+
User.exists?(email: "\tCRUISE-CONTROL@EXAMPLE.COM ") # => true
|
999
|
+
User.exists?(["email = ?", "\tCRUISE-CONTROL@EXAMPLE.COM "]) # => false
|
1074
1000
|
|
1075
|
-
|
1001
|
+
User.normalize_value_for(:phone, "+1 (555) 867-5309") # => "5558675309"
|
1002
|
+
```
|
1076
1003
|
|
1077
|
-
*
|
1004
|
+
*Jonathan Hefner*
|
1078
1005
|
|
1079
|
-
|
1080
|
-
example, the `comments` association here now automatically detects
|
1081
|
-
`inverse_of: :post`, so we don't need to pass that option:
|
1006
|
+
* Hide changes to before_committed! callback behaviour behind flag.
|
1082
1007
|
|
1083
|
-
|
1084
|
-
|
1085
|
-
|
1086
|
-
|
1008
|
+
In #46525, behavior around before_committed! callbacks was changed so that callbacks
|
1009
|
+
would run on every enrolled record in a transaction, not just the first copy of a record.
|
1010
|
+
This change in behavior is now controlled by a configuration option,
|
1011
|
+
`config.active_record.before_committed_on_all_records`. It will be enabled by default on Rails 7.1.
|
1087
1012
|
|
1088
|
-
|
1089
|
-
belongs_to :post
|
1090
|
-
end
|
1091
|
-
```
|
1013
|
+
*Adrianna Chang*
|
1092
1014
|
|
1093
|
-
|
1094
|
-
association has a scope. In this example a scope on the `post` association
|
1095
|
-
would still prevent Rails from finding the inverse for the `comments`
|
1096
|
-
association.
|
1015
|
+
* The `namespaced_controller` Query Log tag now matches the `controller` format
|
1097
1016
|
|
1098
|
-
|
1017
|
+
For example, a request processed by `NameSpaced::UsersController` will now log as:
|
1099
1018
|
|
1100
|
-
```
|
1101
|
-
|
1019
|
+
```
|
1020
|
+
:controller # "users"
|
1021
|
+
:namespaced_controller # "name_spaced/users"
|
1102
1022
|
```
|
1103
1023
|
|
1104
|
-
*
|
1105
|
-
|
1106
|
-
* Accept optional transaction args to `ActiveRecord::Locking::Pessimistic#with_lock`
|
1107
|
-
|
1108
|
-
`#with_lock` now accepts transaction options like `requires_new:`,
|
1109
|
-
`isolation:`, and `joinable:`
|
1110
|
-
|
1111
|
-
*John Mileham*
|
1024
|
+
*Alex Ghiculescu*
|
1112
1025
|
|
1113
|
-
*
|
1026
|
+
* Return only unique ids from ActiveRecord::Calculations#ids
|
1114
1027
|
|
1115
|
-
|
1116
|
-
|
1117
|
-
One example of this is looking up / creating a person via one or more unique alias.
|
1028
|
+
Updated ActiveRecord::Calculations#ids to only return the unique ids of the base model
|
1029
|
+
when using eager_load, preload and includes.
|
1118
1030
|
|
1119
1031
|
```ruby
|
1120
|
-
|
1121
|
-
|
1122
|
-
|
1123
|
-
|
1124
|
-
|
1125
|
-
|
1126
|
-
.create_with(name: "David Heinemeier Hansson")
|
1127
|
-
.create_or_find_by(id: alias.user_id)
|
1128
|
-
end
|
1032
|
+
Post.find_by(id: 1).comments.count
|
1033
|
+
# => 5
|
1034
|
+
Post.includes(:comments).where(id: 1).pluck(:id)
|
1035
|
+
# => [1, 1, 1, 1, 1]
|
1036
|
+
Post.includes(:comments).where(id: 1).ids
|
1037
|
+
# => [1]
|
1129
1038
|
```
|
1130
1039
|
|
1131
|
-
|
1132
|
-
|
1133
|
-
By passing the `:deferrable` option to the `add_foreign_key` statement in migrations, it's possible to defer this
|
1134
|
-
check.
|
1040
|
+
*Joshua Young*
|
1135
1041
|
|
1136
|
-
|
1137
|
-
add_foreign_key :aliases, :person, deferrable: true
|
1138
|
-
```
|
1042
|
+
* Stop using `LOWER()` for case-insensitive queries on `citext` columns
|
1139
1043
|
|
1140
|
-
|
1141
|
-
`
|
1142
|
-
|
1044
|
+
Previously, `LOWER()` was added for e.g. uniqueness validations with
|
1045
|
+
`case_sensitive: false`.
|
1046
|
+
It wasn't mentioned in the documentation that the index without `LOWER()`
|
1047
|
+
wouldn't be used in this case.
|
1143
1048
|
|
1144
|
-
|
1145
|
-
(after the transaction):
|
1049
|
+
*Phil Pirozhkov*
|
1146
1050
|
|
1147
|
-
|
1148
|
-
|
1149
|
-
```
|
1051
|
+
* Extract `#sync_timezone_changes` method in AbstractMysqlAdapter to enable subclasses
|
1052
|
+
to sync database timezone changes without overriding `#raw_execute`.
|
1150
1053
|
|
1151
|
-
*
|
1054
|
+
*Adrianna Chang*, *Paarth Madan*
|
1152
1055
|
|
1153
|
-
*
|
1056
|
+
* Do not write additional new lines when dumping sql migration versions
|
1154
1057
|
|
1155
|
-
|
1156
|
-
```ruby
|
1157
|
-
ActiveRecord::DatabaseConfigurations::UrlConfig.new(
|
1158
|
-
:production, :production, 'postgres:///?user=user&password=secret&dbname=app', {}
|
1159
|
-
).configuration_hash
|
1160
|
-
```
|
1058
|
+
This change updates the `insert_versions_sql` function so that the database insert string containing the current database migration versions does not end with two additional new lines.
|
1161
1059
|
|
1162
|
-
|
1060
|
+
*Misha Schwartz*
|
1163
1061
|
|
1164
|
-
|
1165
|
-
{ :user=>"user", :password=>"secret", :dbname=>"app", :adapter=>"postgresql" }
|
1166
|
-
```
|
1062
|
+
* Fix `composed_of` value freezing and duplication.
|
1167
1063
|
|
1168
|
-
|
1064
|
+
Previously composite values exhibited two confusing behaviors:
|
1169
1065
|
|
1170
|
-
|
1066
|
+
- When reading a compositve value it'd _NOT_ be frozen, allowing it to get out of sync with its underlying database
|
1067
|
+
columns.
|
1068
|
+
- When writing a compositve value the argument would be frozen, potentially confusing the caller.
|
1171
1069
|
|
1172
|
-
|
1070
|
+
Currently, composite values instantiated based on database columns are frozen (addressing the first issue) and
|
1071
|
+
assigned compositve values are duplicated and the duplicate is frozen (addressing the second issue).
|
1173
1072
|
|
1174
|
-
|
1175
|
-
def up
|
1176
|
-
create_enum :mood, ["happy", "sad"]
|
1073
|
+
*Greg Navis*
|
1177
1074
|
|
1178
|
-
|
1179
|
-
t.enum :current_mood, enum_type: "mood", default: "happy", null: false
|
1180
|
-
end
|
1181
|
-
end
|
1182
|
-
```
|
1075
|
+
* Fix redundant updates to the column insensitivity cache
|
1183
1076
|
|
1184
|
-
|
1185
|
-
|
1077
|
+
Fixed redundant queries checking column capability for insensitive
|
1078
|
+
comparison.
|
1186
1079
|
|
1187
|
-
*
|
1080
|
+
*Phil Pirozhkov*
|
1188
1081
|
|
1189
|
-
*
|
1082
|
+
* Allow disabling methods generated by `ActiveRecord.enum`.
|
1190
1083
|
|
1191
|
-
|
1192
|
-
This allows loading the dump without a pgsql superuser account.
|
1084
|
+
*Alfred Dominic*
|
1193
1085
|
|
1194
|
-
|
1086
|
+
* Avoid validating `belongs_to` association if it has not changed.
|
1195
1087
|
|
1196
|
-
|
1088
|
+
Previously, when updating a record, Active Record will perform an extra query to check for the presence of
|
1089
|
+
`belongs_to` associations (if the presence is configured to be mandatory), even if that attribute hasn't changed.
|
1197
1090
|
|
1198
|
-
|
1091
|
+
Currently, only `belongs_to`-related columns are checked for presence. It is possible to have orphaned records with
|
1092
|
+
this approach. To avoid this problem, you need to use a foreign key.
|
1199
1093
|
|
1200
|
-
|
1201
|
-
support of those to the PostgreSQL adapter.
|
1094
|
+
This behavior can be controlled by configuration:
|
1202
1095
|
|
1203
1096
|
```ruby
|
1204
|
-
|
1205
|
-
t.string :name
|
1206
|
-
t.virtual :name_upcased, type: :string, as: 'upper(name)', stored: true
|
1207
|
-
end
|
1097
|
+
config.active_record.belongs_to_required_validates_foreign_key = false
|
1208
1098
|
```
|
1209
1099
|
|
1210
|
-
|
1100
|
+
and will be disabled by default with `config.load_defaults 7.1`.
|
1211
1101
|
|
1102
|
+
*fatkodima*
|
1212
1103
|
|
1213
|
-
|
1104
|
+
* `has_one` and `belongs_to` associations now define a `reset_association` method
|
1105
|
+
on the owner model (where `association` is the name of the association). This
|
1106
|
+
method unloads the cached associate record, if any, and causes the next access
|
1107
|
+
to query it from the database.
|
1214
1108
|
|
1215
|
-
*
|
1109
|
+
*George Claghorn*
|
1216
1110
|
|
1111
|
+
* Allow per attribute setting of YAML permitted classes (safe load) and unsafe load.
|
1217
1112
|
|
1218
|
-
|
1113
|
+
*Carlos Palhares*
|
1219
1114
|
|
1220
|
-
*
|
1115
|
+
* Add a build persistence method
|
1221
1116
|
|
1222
|
-
|
1117
|
+
Provides a wrapper for `new`, to provide feature parity with `create`s
|
1118
|
+
ability to create multiple records from an array of hashes, using the
|
1119
|
+
same notation as the `build` method on associations.
|
1223
1120
|
|
1224
|
-
|
1225
|
-
Creating scope :my_scope_name. Overwriting existing method "MyClass.my_scope_name" when overwriting existing scopes
|
1226
|
-
```
|
1121
|
+
*Sean Denny*
|
1227
1122
|
|
1228
|
-
|
1123
|
+
* Raise on assignment to readonly attributes
|
1229
1124
|
|
1230
|
-
|
1125
|
+
```ruby
|
1126
|
+
class Post < ActiveRecord::Base
|
1127
|
+
attr_readonly :content
|
1128
|
+
end
|
1129
|
+
Post.create!(content: "cannot be updated")
|
1130
|
+
post.content # "cannot be updated"
|
1131
|
+
post.content = "something else" # => ActiveRecord::ReadonlyAttributeError
|
1132
|
+
```
|
1231
1133
|
|
1232
|
-
|
1233
|
-
and not all databases support explicitly specifying additional precision.
|
1134
|
+
Previously, assignment would succeed but silently not write to the database.
|
1234
1135
|
|
1235
|
-
|
1236
|
-
for the SQL to produce a high precision timestamp on the current database.
|
1136
|
+
This behavior can be controlled by configuration:
|
1237
1137
|
|
1238
|
-
|
1138
|
+
```ruby
|
1139
|
+
config.active_record.raise_on_assign_to_attr_readonly = true
|
1140
|
+
```
|
1239
1141
|
|
1240
|
-
|
1142
|
+
and will be enabled by default with `config.load_defaults 7.1`.
|
1241
1143
|
|
1242
|
-
*
|
1144
|
+
*Alex Ghiculescu*, *Hartley McGuire*
|
1243
1145
|
|
1244
|
-
|
1245
|
-
when running postgresql database tasks.
|
1146
|
+
* Allow unscoping of preload and eager_load associations
|
1246
1147
|
|
1247
|
-
|
1248
|
-
|
1148
|
+
Added the ability to unscope preload and eager_load associations just like
|
1149
|
+
includes, joins, etc. See ActiveRecord::QueryMethods::VALID_UNSCOPING_VALUES
|
1150
|
+
for the full list of supported unscopable scopes.
|
1249
1151
|
|
1250
|
-
|
1251
|
-
|
1252
|
-
sslcert: client.crt
|
1253
|
-
sslkey: client.key
|
1254
|
-
sslrootcert: ca.crt
|
1152
|
+
```ruby
|
1153
|
+
query.unscope(:eager_load, :preload).group(:id).select(:id)
|
1255
1154
|
```
|
1256
1155
|
|
1257
|
-
|
1156
|
+
*David Morehouse*
|
1258
1157
|
|
1259
|
-
|
1260
|
-
PGSSLMODE=verify-full
|
1261
|
-
PGSSLCERT=client.crt
|
1262
|
-
PGSSLKEY=client.key
|
1263
|
-
PGSSLROOTCERT=ca.crt
|
1264
|
-
```
|
1158
|
+
* Add automatic filtering of encrypted attributes on inspect
|
1265
1159
|
|
1266
|
-
|
1160
|
+
This feature is enabled by default but can be disabled with
|
1267
1161
|
|
1268
|
-
|
1162
|
+
```ruby
|
1163
|
+
config.active_record.encryption.add_to_filter_parameters = false
|
1164
|
+
```
|
1165
|
+
|
1166
|
+
*Hartley McGuire*
|
1269
1167
|
|
1270
|
-
*
|
1168
|
+
* Clear locking column on #dup
|
1271
1169
|
|
1272
|
-
|
1273
|
-
and not also to the callbacks from the update itself.
|
1170
|
+
This change fixes not to duplicate locking_column like id and timestamps.
|
1274
1171
|
|
1275
|
-
|
1172
|
+
```
|
1173
|
+
car = Car.create!
|
1174
|
+
car.touch
|
1175
|
+
car.lock_version #=> 1
|
1176
|
+
car.dup.lock_version #=> 0
|
1177
|
+
```
|
1276
1178
|
|
1277
|
-
*
|
1278
|
-
using `String#constantize` instead of the model's `polymorphic_class_for`.
|
1179
|
+
*Shouichi Kamiya*, *Seonggi Yang*, *Ryohei UEDA*
|
1279
1180
|
|
1280
|
-
|
1281
|
-
1. `touch`ing the previously associated record
|
1282
|
-
2. updating the previously associated record's `counter_cache`
|
1181
|
+
* Invalidate transaction as early as possible
|
1283
1182
|
|
1284
|
-
|
1183
|
+
After rescuing a `TransactionRollbackError` exception Rails invalidates transactions earlier in the flow
|
1184
|
+
allowing the framework to skip issuing the `ROLLBACK` statement in more cases.
|
1185
|
+
Only affects adapters that have `savepoint_errors_invalidate_transactions?` configured as `true`,
|
1186
|
+
which at this point is only applicable to the `mysql2` adapter.
|
1285
1187
|
|
1286
|
-
*
|
1188
|
+
*Nikita Vasilevsky*
|
1287
1189
|
|
1288
|
-
|
1190
|
+
* Allow configuring columns list to be used in SQL queries issued by an `ActiveRecord::Base` object
|
1289
1191
|
|
1290
|
-
|
1192
|
+
It is now possible to configure columns list that will be used to build an SQL query clauses when
|
1193
|
+
updating, deleting or reloading an `ActiveRecord::Base` object
|
1291
1194
|
|
1292
1195
|
```ruby
|
1293
|
-
|
1196
|
+
class Developer < ActiveRecord::Base
|
1197
|
+
query_constraints :company_id, :id
|
1198
|
+
end
|
1199
|
+
developer = Developer.first.update(name: "Bob")
|
1200
|
+
# => UPDATE "developers" SET "name" = 'Bob' WHERE "developers"."company_id" = 1 AND "developers"."id" = 1
|
1294
1201
|
```
|
1295
1202
|
|
1296
|
-
|
1203
|
+
*Nikita Vasilevsky*
|
1297
1204
|
|
1298
|
-
|
1299
|
-
config.active_record.schema_cache_ignored_tables = [/^_/]
|
1300
|
-
```
|
1205
|
+
* Adds `validate` to foreign keys and check constraints in schema.rb
|
1301
1206
|
|
1302
|
-
|
1207
|
+
Previously, `schema.rb` would not record if `validate: false` had been used when adding a foreign key or check
|
1208
|
+
constraint, so restoring a database from the schema could result in foreign keys or check constraints being
|
1209
|
+
incorrectly validated.
|
1303
1210
|
|
1304
|
-
*
|
1211
|
+
*Tommy Graves*
|
1305
1212
|
|
1306
|
-
|
1307
|
-
|
1308
|
-
database adapters. This change unifies the behavior so each method behaves the same regardless
|
1309
|
-
of adapter.
|
1213
|
+
* Adapter `#execute` methods now accept an `allow_retry` option. When set to `true`, the SQL statement will be
|
1214
|
+
retried, up to the database's configured `connection_retries` value, upon encountering connection-related errors.
|
1310
1215
|
|
1311
|
-
|
1216
|
+
*Adrianna Chang*
|
1312
1217
|
|
1313
|
-
|
1314
|
-
`columns_hash`: (unchanged) raises a db error if the table does not exist.
|
1315
|
-
`primary_keys`: (unchanged) returns `nil` if the table does not exist.
|
1316
|
-
`indexes`: (changed for mysql2) returns `[]` if the table does not exist.
|
1218
|
+
* Only trigger `after_commit :destroy` callbacks when a database row is deleted.
|
1317
1219
|
|
1318
|
-
|
1220
|
+
This prevents `after_commit :destroy` callbacks from being triggered again
|
1221
|
+
when `destroy` is called multiple times on the same record.
|
1319
1222
|
|
1320
|
-
*
|
1223
|
+
*Ben Sheldon*
|
1321
1224
|
|
1322
|
-
|
1225
|
+
* Fix `ciphertext_for` for yet-to-be-encrypted values.
|
1323
1226
|
|
1324
|
-
|
1227
|
+
Previously, `ciphertext_for` returned the cleartext of values that had not
|
1228
|
+
yet been encrypted, such as with an unpersisted record:
|
1325
1229
|
|
1326
|
-
|
1230
|
+
```ruby
|
1231
|
+
Post.encrypts :body
|
1327
1232
|
|
1328
|
-
|
1329
|
-
|
1330
|
-
|
1233
|
+
post = Post.create!(body: "Hello")
|
1234
|
+
post.ciphertext_for(:body)
|
1235
|
+
# => "{\"p\":\"abc..."
|
1331
1236
|
|
1332
|
-
|
1333
|
-
|
1237
|
+
post.body = "World"
|
1238
|
+
post.ciphertext_for(:body)
|
1239
|
+
# => "World"
|
1240
|
+
```
|
1334
1241
|
|
1335
|
-
|
1336
|
-
|
1337
|
-
database: my_database
|
1338
|
-
adapter: mysql2
|
1339
|
-
animals:
|
1340
|
-
database: my_animals_database
|
1341
|
-
adapter: mysql2
|
1342
|
-
database_tasks: false
|
1343
|
-
```
|
1242
|
+
Now, `ciphertext_for` will always return the ciphertext of encrypted
|
1243
|
+
attributes:
|
1344
1244
|
|
1345
|
-
|
1245
|
+
```ruby
|
1246
|
+
Post.encrypts :body
|
1346
1247
|
|
1347
|
-
|
1248
|
+
post = Post.create!(body: "Hello")
|
1249
|
+
post.ciphertext_for(:body)
|
1250
|
+
# => "{\"p\":\"abc..."
|
1348
1251
|
|
1349
|
-
|
1350
|
-
|
1252
|
+
post.body = "World"
|
1253
|
+
post.ciphertext_for(:body)
|
1254
|
+
# => "{\"p\":\"xyz..."
|
1255
|
+
```
|
1351
1256
|
|
1352
|
-
|
1257
|
+
*Jonathan Hefner*
|
1353
1258
|
|
1354
|
-
*
|
1259
|
+
* Fix a bug where using groups and counts with long table names would return incorrect results.
|
1355
1260
|
|
1356
|
-
|
1261
|
+
*Shota Toguchi*, *Yusaku Ono*
|
1357
1262
|
|
1358
|
-
|
1359
|
-
# config/application.rb
|
1360
|
-
module MyApp
|
1361
|
-
class Application < Rails::Application
|
1362
|
-
config.active_record.query_log_tags_enabled = true
|
1363
|
-
end
|
1364
|
-
end
|
1365
|
-
```
|
1263
|
+
* Fix encryption of column default values.
|
1366
1264
|
|
1367
|
-
|
1265
|
+
Previously, encrypted attributes that used column default values appeared to
|
1266
|
+
be encrypted on create, but were not:
|
1368
1267
|
|
1369
|
-
|
1370
|
-
|
1371
|
-
def index
|
1372
|
-
@books = Book.all
|
1373
|
-
end
|
1374
|
-
end
|
1375
|
-
```
|
1268
|
+
```ruby
|
1269
|
+
Book.encrypts :name
|
1376
1270
|
|
1377
|
-
|
1378
|
-
|
1379
|
-
|
1380
|
-
|
1271
|
+
book = Book.create!
|
1272
|
+
book.name
|
1273
|
+
# => "<untitled>"
|
1274
|
+
book.name_before_type_cast
|
1275
|
+
# => "{\"p\":\"abc..."
|
1276
|
+
book.reload.name_before_type_cast
|
1277
|
+
# => "<untitled>"
|
1278
|
+
```
|
1381
1279
|
|
1382
|
-
|
1280
|
+
Now, attributes with column default values are encrypted:
|
1383
1281
|
|
1384
|
-
|
1385
|
-
|
1386
|
-
:application,
|
1387
|
-
:controller,
|
1388
|
-
:action,
|
1389
|
-
{
|
1390
|
-
custom_static: "foo",
|
1391
|
-
custom_dynamic: -> { Time.now }
|
1392
|
-
}
|
1393
|
-
]
|
1394
|
-
```
|
1282
|
+
```ruby
|
1283
|
+
Book.encrypts :name
|
1395
1284
|
|
1396
|
-
|
1285
|
+
book = Book.create!
|
1286
|
+
book.name
|
1287
|
+
# => "<untitled>"
|
1288
|
+
book.name_before_type_cast
|
1289
|
+
# => "{\"p\":\"abc..."
|
1290
|
+
book.reload.name_before_type_cast
|
1291
|
+
# => "{\"p\":\"abc..."
|
1292
|
+
```
|
1397
1293
|
|
1398
|
-
*
|
1294
|
+
*Jonathan Hefner*
|
1399
1295
|
|
1400
|
-
|
1296
|
+
* Deprecate delegation from `Base` to `connection_handler`.
|
1401
1297
|
|
1402
|
-
|
1298
|
+
Calling `Base.clear_all_connections!`, `Base.clear_active_connections!`, `Base.clear_reloadable_connections!` and `Base.flush_idle_connections!` is deprecated. Please call these methods on the connection handler directly. In future Rails versions, the delegation from `Base` to the `connection_handler` will be removed.
|
1403
1299
|
|
1404
|
-
|
1405
|
-
about to use for `#or` or `#and` is structurally compatible with the
|
1406
|
-
receiver.
|
1300
|
+
*Eileen M. Uchitelle*
|
1407
1301
|
|
1408
|
-
|
1302
|
+
* Allow ActiveRecord::QueryMethods#reselect to receive hash values, similar to ActiveRecord::QueryMethods#select
|
1409
1303
|
|
1410
|
-
*
|
1304
|
+
*Sampat Badhe*
|
1411
1305
|
|
1412
|
-
|
1413
|
-
returned in based on a SQL expression. By default, this will be accomplished
|
1414
|
-
using a case statement, as in:
|
1306
|
+
* Validate options when managing columns and tables in migrations.
|
1415
1307
|
|
1416
|
-
|
1417
|
-
|
1418
|
-
|
1308
|
+
If an invalid option is passed to a migration method like `create_table` and `add_column`, an error will be raised
|
1309
|
+
instead of the option being silently ignored. Validation of the options will only be applied for new migrations
|
1310
|
+
that are created.
|
1419
1311
|
|
1420
|
-
|
1312
|
+
*Guo Xiang Tan*, *George Wambold*
|
1421
1313
|
|
1422
|
-
|
1423
|
-
SELECT "posts".* FROM "posts" ORDER BY CASE "posts"."id" WHEN 3 THEN 1 WHEN 5 THEN 2 WHEN 1 THEN 3 ELSE 4 END ASC
|
1424
|
-
```
|
1314
|
+
* Update query log tags to use the [SQLCommenter](https://open-telemetry.github.io/opentelemetry-sqlcommenter/) format by default. See [#46179](https://github.com/rails/rails/issues/46179)
|
1425
1315
|
|
1426
|
-
|
1427
|
-
`FIELD` function, that connection adapter will generate the following SQL
|
1428
|
-
instead:
|
1316
|
+
To opt out of SQLCommenter-formatted query log tags, set `config.active_record.query_log_tags_format = :legacy`. By default, this is set to `:sqlcommenter`.
|
1429
1317
|
|
1430
|
-
|
1431
|
-
SELECT "posts".* FROM "posts" ORDER BY FIELD("posts"."id", 1, 5, 3) DESC
|
1432
|
-
```
|
1318
|
+
*Modulitos* and *Iheanyi*
|
1433
1319
|
|
1434
|
-
|
1320
|
+
* Allow any ERB in the database.yml when creating rake tasks.
|
1435
1321
|
|
1436
|
-
|
1322
|
+
Any ERB can be used in `database.yml` even if it accesses environment
|
1323
|
+
configurations.
|
1437
1324
|
|
1438
|
-
|
1325
|
+
Deprecates `config.active_record.suppress_multiple_database_warning`.
|
1439
1326
|
|
1440
|
-
|
1441
|
-
scope = Post.includes(:comments).order(:"comments.label")
|
1442
|
-
=> true
|
1443
|
-
```
|
1327
|
+
*Eike Send*
|
1444
1328
|
|
1445
|
-
|
1329
|
+
* Add table to error for duplicate column definitions.
|
1446
1330
|
|
1447
|
-
|
1331
|
+
If a migration defines duplicate columns for a table, the error message
|
1332
|
+
shows which table it concerns.
|
1448
1333
|
|
1449
|
-
|
1450
|
-
if a different associated object has been assigned and the foreign key will be updated in the
|
1451
|
-
next save.
|
1334
|
+
*Petrik de Heus*
|
1452
1335
|
|
1453
|
-
|
1454
|
-
association to reference a different associated object.
|
1336
|
+
* Fix erroneous nil default precision on virtual datetime columns.
|
1455
1337
|
|
1456
|
-
|
1338
|
+
Prior to this change, virtual datetime columns did not have the same
|
1339
|
+
default precision as regular datetime columns, resulting in the following
|
1340
|
+
being erroneously equivalent:
|
1457
1341
|
|
1458
|
-
|
1342
|
+
t.virtual :name, type: datetime, as: "expression"
|
1343
|
+
t.virtual :name, type: datetime, precision: nil, as: "expression"
|
1459
1344
|
|
1460
|
-
|
1461
|
-
|
1345
|
+
This change fixes the default precision lookup, so virtual and regular
|
1346
|
+
datetime column default precisions match.
|
1462
1347
|
|
1463
|
-
|
1464
|
-
# config/database.yml
|
1348
|
+
*Sam Bostock*
|
1465
1349
|
|
1466
|
-
|
1467
|
-
schema_dump: false
|
1468
|
-
```
|
1350
|
+
* Use connection from `#with_raw_connection` in `#quote_string`.
|
1469
1351
|
|
1470
|
-
|
1352
|
+
This ensures that the string quoting is wrapped in the reconnect and retry logic
|
1353
|
+
that `#with_raw_connection` offers.
|
1471
1354
|
|
1472
|
-
*
|
1355
|
+
*Adrianna Chang*
|
1473
1356
|
|
1474
|
-
|
1475
|
-
on an outer table.
|
1357
|
+
* Add `expires_in` option to `signed_id`.
|
1476
1358
|
|
1477
|
-
|
1478
|
-
Post.includes(:comments).order({ "comments.label": :ASC }).eager_loading?
|
1479
|
-
# => true
|
1480
|
-
```
|
1359
|
+
*Shouichi Kamiya*
|
1481
1360
|
|
1482
|
-
|
1361
|
+
* Allow applications to set retry deadline for query retries.
|
1483
1362
|
|
1484
|
-
|
1363
|
+
Building on the work done in #44576 and #44591, we extend the logic that automatically
|
1364
|
+
reconnects database connections to take into account a timeout limit. We won't retry
|
1365
|
+
a query if a given amount of time has elapsed since the query was first attempted. This
|
1366
|
+
value defaults to nil, meaning that all retryable queries are retried regardless of time elapsed,
|
1367
|
+
but this can be changed via the `retry_deadline` option in the database config.
|
1485
1368
|
|
1486
|
-
|
1369
|
+
*Adrianna Chang*
|
1487
1370
|
|
1488
|
-
|
1371
|
+
* Fix a case where the query cache can return wrong values. See #46044
|
1489
1372
|
|
1490
|
-
*
|
1373
|
+
*Aaron Patterson*
|
1491
1374
|
|
1492
|
-
|
1375
|
+
* Support MySQL's ssl-mode option for MySQLDatabaseTasks.
|
1493
1376
|
|
1494
|
-
|
1495
|
-
|
1496
|
-
|
1377
|
+
Verifying the identity of the database server requires setting the ssl-mode
|
1378
|
+
option to VERIFY_CA or VERIFY_IDENTITY. This option was previously ignored
|
1379
|
+
for MySQL database tasks like creating a database and dumping the structure.
|
1497
1380
|
|
1498
|
-
|
1499
|
-
would cause old processes to no longer be able to create new records.
|
1381
|
+
*Petrik de Heus*
|
1500
1382
|
|
1501
|
-
|
1502
|
-
to stop using it.
|
1383
|
+
* Move `ActiveRecord::InternalMetadata` to an independent object.
|
1503
1384
|
|
1504
|
-
|
1385
|
+
`ActiveRecord::InternalMetadata` no longer inherits from `ActiveRecord::Base` and is now an independent object that should be instantiated with a `connection`. This class is private and should not be used by applications directly. If you want to interact with the schema migrations table, please access it on the connection directly, for example: `ActiveRecord::Base.connection.schema_migration`.
|
1505
1386
|
|
1506
|
-
*
|
1387
|
+
*Eileen M. Uchitelle*
|
1507
1388
|
|
1508
|
-
|
1389
|
+
* Deprecate quoting `ActiveSupport::Duration` as an integer
|
1509
1390
|
|
1510
|
-
|
1511
|
-
|
1391
|
+
Using ActiveSupport::Duration as an interpolated bind parameter in a SQL
|
1392
|
+
string template is deprecated. To avoid this warning, you should explicitly
|
1393
|
+
convert the duration to a more specific database type. For example, if you
|
1394
|
+
want to use a duration as an integer number of seconds:
|
1395
|
+
```
|
1396
|
+
Record.where("duration = ?", 1.hour.to_i)
|
1397
|
+
```
|
1398
|
+
If you want to use a duration as an ISO 8601 string:
|
1399
|
+
```
|
1400
|
+
Record.where("duration = ?", 1.hour.iso8601)
|
1512
1401
|
```
|
1513
1402
|
|
1514
|
-
|
1515
|
-
|
1516
|
-
The feature is supported by SQLite and PostgreSQL, other adapters can also add support for it.
|
1403
|
+
*Aram Greenman*
|
1517
1404
|
|
1518
|
-
|
1405
|
+
* Allow `QueryMethods#in_order_of` to order by a string column name.
|
1519
1406
|
|
1520
|
-
|
1407
|
+
```ruby
|
1408
|
+
Post.in_order_of("id", [4,2,3,1]).to_a
|
1409
|
+
Post.joins(:author).in_order_of("authors.name", ["Bob", "Anna", "John"]).to_a
|
1410
|
+
```
|
1521
1411
|
|
1522
|
-
|
1523
|
-
the owner should still return `nil` on the `has_one` relation.
|
1412
|
+
*Igor Kasyanchuk*
|
1524
1413
|
|
1525
|
-
|
1414
|
+
* Move `ActiveRecord::SchemaMigration` to an independent object.
|
1526
1415
|
|
1527
|
-
|
1416
|
+
`ActiveRecord::SchemaMigration` no longer inherits from `ActiveRecord::Base` and is now an independent object that should be instantiated with a `connection`. This class is private and should not be used by applications directly. If you want to interact with the schema migrations table, please access it on the connection directly, for example: `ActiveRecord::Base.connection.schema_migration`.
|
1528
1417
|
|
1529
|
-
*
|
1418
|
+
*Eileen M. Uchitelle*
|
1530
1419
|
|
1531
|
-
|
1420
|
+
* Deprecate `all_connection_pools` and make `connection_pool_list` more explicit.
|
1532
1421
|
|
1533
|
-
|
1422
|
+
Following on #45924 `all_connection_pools` is now deprecated. `connection_pool_list` will either take an explicit role or applications can opt into the new behavior by passing `:all`.
|
1534
1423
|
|
1535
|
-
|
1536
|
-
that already exists or when removing a foreign key that does not exist.
|
1424
|
+
*Eileen M. Uchitelle*
|
1537
1425
|
|
1538
|
-
|
1426
|
+
* Fix connection handler methods to operate on all pools.
|
1539
1427
|
|
1540
|
-
|
1541
|
-
class AddAuthorsForeignKeyToArticles < ActiveRecord::Migration[7.0]
|
1542
|
-
def change
|
1543
|
-
add_foreign_key :articles, :authors, if_not_exists: true
|
1544
|
-
end
|
1545
|
-
end
|
1546
|
-
```
|
1428
|
+
`active_connections?`, `clear_active_connections!`, `clear_reloadable_connections!`, `clear_all_connections!`, and `flush_idle_connections!` now operate on all pools by default. Previously they would default to using the `current_role` or `:writing` role unless specified.
|
1547
1429
|
|
1548
|
-
|
1549
|
-
class RemoveAuthorsForeignKeyFromArticles < ActiveRecord::Migration[7.0]
|
1550
|
-
def change
|
1551
|
-
remove_foreign_key :articles, :authors, if_exists: true
|
1552
|
-
end
|
1553
|
-
end
|
1554
|
-
```
|
1430
|
+
*Eileen M. Uchitelle*
|
1555
1431
|
|
1556
|
-
*Roberto Miranda*
|
1557
1432
|
|
1558
|
-
*
|
1433
|
+
* Allow ActiveRecord::QueryMethods#select to receive hash values.
|
1559
1434
|
|
1560
|
-
|
1561
|
-
environment variables which persisted beyond the command being run, and may
|
1562
|
-
have caused subsequent commands and connections to fail. Tasks running
|
1563
|
-
across multiple postgresql databases like `rails db:test:prepare` may have
|
1564
|
-
been affected.
|
1435
|
+
Currently, `select` might receive only raw sql and symbols to define columns and aliases to select.
|
1565
1436
|
|
1566
|
-
|
1437
|
+
With this change we can provide `hash` as argument, for example:
|
1567
1438
|
|
1568
|
-
|
1439
|
+
```ruby
|
1440
|
+
Post.joins(:comments).select(posts: [:id, :title, :created_at], comments: [:id, :body, :author_id])
|
1441
|
+
#=> "SELECT \"posts\".\"id\", \"posts\".\"title\", \"posts\".\"created_at\", \"comments\".\"id\", \"comments\".\"body\", \"comments\".\"author_id\"
|
1442
|
+
# FROM \"posts\" INNER JOIN \"comments\" ON \"comments\".\"post_id\" = \"posts\".\"id\""
|
1569
1443
|
|
1570
|
-
|
1444
|
+
Post.joins(:comments).select(posts: { id: :post_id, title: :post_title }, comments: { id: :comment_id, body: :comment_body })
|
1445
|
+
#=> "SELECT posts.id as post_id, posts.title as post_title, comments.id as comment_id, comments.body as comment_body
|
1446
|
+
# FROM \"posts\" INNER JOIN \"comments\" ON \"comments\".\"post_id\" = \"posts\".\"id\""
|
1447
|
+
```
|
1448
|
+
*Oleksandr Holubenko*, *Josef Šimánek*, *Jean Boussier*
|
1571
1449
|
|
1572
|
-
|
1450
|
+
* Adapts virtual attributes on `ActiveRecord::Persistence#becomes`.
|
1573
1451
|
|
1574
|
-
|
1452
|
+
When source and target classes have a different set of attributes adapts
|
1453
|
+
attributes such that the extra attributes from target are added.
|
1575
1454
|
|
1576
|
-
|
1455
|
+
```ruby
|
1456
|
+
class Person < ApplicationRecord
|
1457
|
+
end
|
1577
1458
|
|
1578
|
-
|
1459
|
+
class WebUser < Person
|
1460
|
+
attribute :is_admin, :boolean
|
1461
|
+
after_initialize :set_admin
|
1579
1462
|
|
1580
|
-
|
1463
|
+
def set_admin
|
1464
|
+
write_attribute(:is_admin, email =~ /@ourcompany\.com$/)
|
1465
|
+
end
|
1466
|
+
end
|
1581
1467
|
|
1582
|
-
|
1583
|
-
|
1584
|
-
|
1585
|
-
|
1468
|
+
person = Person.find_by(email: "email@ourcompany.com")
|
1469
|
+
person.respond_to? :is_admin
|
1470
|
+
# => false
|
1471
|
+
person.becomes(WebUser).is_admin?
|
1472
|
+
# => true
|
1473
|
+
```
|
1586
1474
|
|
1587
|
-
|
1588
|
-
that is `Encoding::UTF_8` by default. It can be disabled by setting it to `nil`.
|
1475
|
+
*Jacopo Beschi*, *Sampson Crowley*
|
1589
1476
|
|
1590
|
-
|
1477
|
+
* Fix `ActiveRecord::QueryMethods#in_order_of` to include `nil`s, to match the
|
1478
|
+
behavior of `Enumerable#in_order_of`.
|
1591
1479
|
|
1592
|
-
|
1480
|
+
For example, `Post.in_order_of(:title, [nil, "foo"])` will now include posts
|
1481
|
+
with `nil` titles, the same as `Post.all.to_a.in_order_of(:title, [nil, "foo"])`.
|
1593
1482
|
|
1594
|
-
|
1595
|
-
instance `"foo" = 0`, will implicitly cast `"foo"` to `0` and will evaluate to `TRUE` which can
|
1596
|
-
lead to security vulnerabilities.
|
1483
|
+
*fatkodima*
|
1597
1484
|
|
1598
|
-
|
1599
|
-
being compared, however until now it was still vulnerable when using bind parameters:
|
1485
|
+
* Optimize `add_timestamps` to use a single SQL statement.
|
1600
1486
|
|
1601
1487
|
```ruby
|
1602
|
-
|
1488
|
+
add_timestamps :my_table
|
1603
1489
|
```
|
1604
1490
|
|
1605
|
-
|
1491
|
+
Now results in the following SQL:
|
1606
1492
|
|
1607
1493
|
```sql
|
1608
|
-
|
1494
|
+
ALTER TABLE "my_table" ADD COLUMN "created_at" datetime(6) NOT NULL, ADD COLUMN "updated_at" datetime(6) NOT NULL
|
1609
1495
|
```
|
1610
1496
|
|
1611
|
-
|
1612
|
-
|
1613
|
-
```sql
|
1614
|
-
SELECT * FROM `users` WHERE `login_token` = '0' LIMIT 1;
|
1615
|
-
```
|
1497
|
+
*Iliana Hadzhiatanasova*
|
1616
1498
|
|
1617
|
-
|
1499
|
+
* Add `drop_enum` migration command for PostgreSQL
|
1618
1500
|
|
1619
|
-
|
1501
|
+
This does the inverse of `create_enum`. Before dropping an enum, ensure you have
|
1502
|
+
dropped columns that depend on it.
|
1620
1503
|
|
1621
|
-
|
1622
|
-
would silently have no effects.
|
1504
|
+
*Alex Ghiculescu*
|
1623
1505
|
|
1624
|
-
|
1506
|
+
* Adds support for `if_exists` option when removing a check constraint.
|
1625
1507
|
|
1626
|
-
|
1508
|
+
The `remove_check_constraint` method now accepts an `if_exists` option. If set
|
1509
|
+
to true an error won't be raised if the check constraint doesn't exist.
|
1627
1510
|
|
1628
|
-
|
1511
|
+
*Margaret Parsa* and *Aditya Bhutani*
|
1629
1512
|
|
1630
|
-
|
1631
|
-
Person.update!(:all, state: "confirmed")
|
1632
|
-
```
|
1513
|
+
* `find_or_create_by` now try to find a second time if it hits a unicity constraint.
|
1633
1514
|
|
1634
|
-
|
1515
|
+
`find_or_create_by` always has been inherently racy, either creating multiple
|
1516
|
+
duplicate records or failing with `ActiveRecord::RecordNotUnique` depending on
|
1517
|
+
whether a proper unicity constraint was set.
|
1635
1518
|
|
1636
|
-
|
1519
|
+
`create_or_find_by` was introduced for this use case, however it's quite wasteful
|
1520
|
+
when the record is expected to exist most of the time, as INSERT require to send
|
1521
|
+
more data than SELECT and require more work from the database. Also on some
|
1522
|
+
databases it can actually consume a primary key increment which is undesirable.
|
1637
1523
|
|
1638
|
-
|
1524
|
+
So for case where most of the time the record is expected to exist, `find_or_create_by`
|
1525
|
+
can be made race-condition free by re-trying the `find` if the `create` failed
|
1526
|
+
with `ActiveRecord::RecordNotUnique`. This assumes that the table has the proper
|
1527
|
+
unicity constraints, if not, `find_or_create_by` will still lead to duplicated records.
|
1639
1528
|
|
1640
|
-
*
|
1529
|
+
*Jean Boussier*, *Alex Kitchens*
|
1641
1530
|
|
1642
|
-
*
|
1531
|
+
* Introduce a simpler constructor API for ActiveRecord database adapters.
|
1643
1532
|
|
1644
|
-
|
1533
|
+
Previously the adapter had to know how to build a new raw connection to
|
1534
|
+
support reconnect, but also expected to be passed an initial already-
|
1535
|
+
established connection.
|
1645
1536
|
|
1646
|
-
|
1537
|
+
When manually creating an adapter instance, it will now accept a single
|
1538
|
+
config hash, and only establish the real connection on demand.
|
1647
1539
|
|
1648
|
-
*
|
1540
|
+
*Matthew Draper*
|
1649
1541
|
|
1650
|
-
|
1542
|
+
* Avoid redundant `SELECT 1` connection-validation query during DB pool
|
1543
|
+
checkout when possible.
|
1651
1544
|
|
1652
|
-
|
1545
|
+
If the first query run during a request is known to be idempotent, it can be
|
1546
|
+
used directly to validate the connection, saving a network round-trip.
|
1653
1547
|
|
1654
|
-
|
1548
|
+
*Matthew Draper*
|
1655
1549
|
|
1656
|
-
|
1550
|
+
* Automatically reconnect broken database connections when safe, even
|
1551
|
+
mid-request.
|
1657
1552
|
|
1658
|
-
|
1553
|
+
When an error occurs while attempting to run a known-idempotent query, and
|
1554
|
+
not inside a transaction, it is safe to immediately reconnect to the
|
1555
|
+
database server and try again, so this is now the default behavior.
|
1659
1556
|
|
1660
|
-
|
1661
|
-
|
1557
|
+
This new default should always be safe -- to support that, it's consciously
|
1558
|
+
conservative about which queries are considered idempotent -- but if
|
1559
|
+
necessary it can be disabled by setting the `connection_retries` connection
|
1560
|
+
option to `0`.
|
1662
1561
|
|
1663
|
-
*
|
1562
|
+
*Matthew Draper*
|
1664
1563
|
|
1665
|
-
*
|
1564
|
+
* Avoid removing a PostgreSQL extension when there are dependent objects.
|
1666
1565
|
|
1667
|
-
|
1668
|
-
on a subclass won't change the parent's logger.
|
1566
|
+
Previously, removing an extension also implicitly removed dependent objects. Now, this will raise an error.
|
1669
1567
|
|
1670
|
-
|
1568
|
+
You can force removing the extension:
|
1671
1569
|
|
1672
|
-
|
1570
|
+
```ruby
|
1571
|
+
disable_extension :citext, force: :cascade
|
1572
|
+
```
|
1673
1573
|
|
1674
|
-
|
1574
|
+
Fixes #29091.
|
1675
1575
|
|
1676
|
-
*
|
1576
|
+
*fatkodima*
|
1677
1577
|
|
1678
|
-
|
1578
|
+
* Allow nested functions as safe SQL string
|
1679
1579
|
|
1680
|
-
*
|
1580
|
+
*Michael Siegfried*
|
1681
1581
|
|
1682
|
-
|
1683
|
-
the `column_exists?` check would look for a column with the same name and type as the migration.
|
1582
|
+
* Allow `destroy_association_async_job=` to be configured with a class string instead of a constant.
|
1684
1583
|
|
1685
|
-
|
1686
|
-
|
1687
|
-
|
1688
|
-
the type to the database type without running the migration, we opted to drop the type check from
|
1689
|
-
`add_column`. This means that migrations adding a duplicate column with a different type will no
|
1690
|
-
longer raise an error.
|
1584
|
+
Defers an autoloading dependency between `ActiveRecord::Base` and `ActiveJob::Base`
|
1585
|
+
and moves the configuration of `ActiveRecord::DestroyAssociationAsyncJob`
|
1586
|
+
from ActiveJob to ActiveRecord.
|
1691
1587
|
|
1692
|
-
|
1588
|
+
Deprecates `ActiveRecord::ActiveJobRequiredError` and now raises a `NameError`
|
1589
|
+
if the job class is unloadable or an `ActiveRecord::ConfigurationError` if
|
1590
|
+
`dependent: :destroy_async` is declared on an association but there is no job
|
1591
|
+
class configured.
|
1693
1592
|
|
1694
|
-
*
|
1593
|
+
*Ben Sheldon*
|
1695
1594
|
|
1696
|
-
|
1697
|
-
in a new Rails application.
|
1698
|
-
For the above reasons log a warning message when running SQLite in production.
|
1595
|
+
* Fix `ActiveRecord::Store` to serialize as a regular Hash
|
1699
1596
|
|
1700
|
-
|
1597
|
+
Previously it would serialize as an `ActiveSupport::HashWithIndifferentAccess`
|
1598
|
+
which is wasteful and cause problem with YAML safe_load.
|
1701
1599
|
|
1702
|
-
*
|
1600
|
+
*Jean Boussier*
|
1703
1601
|
|
1704
|
-
* Add
|
1602
|
+
* Add `timestamptz` as a time zone aware type for PostgreSQL
|
1705
1603
|
|
1706
|
-
|
1707
|
-
databases. When set, this option instructs Rails to generate 2 or
|
1708
|
-
more queries rather than generating joins for `has_one` associations.
|
1604
|
+
This is required for correctly parsing `timestamp with time zone` values in your database.
|
1709
1605
|
|
1710
|
-
|
1606
|
+
If you don't want this, you can opt out by adding this initializer:
|
1711
1607
|
|
1712
1608
|
```ruby
|
1713
|
-
|
1714
|
-
has_one :dog
|
1715
|
-
has_one :veterinarian, through: :dog, disable_joins: true
|
1716
|
-
end
|
1717
|
-
```
|
1718
|
-
|
1719
|
-
Then instead of generating join SQL, two queries are used for `@person.veterinarian`:
|
1720
|
-
|
1721
|
-
```
|
1722
|
-
SELECT "dogs"."id" FROM "dogs" WHERE "dogs"."person_id" = ? [["person_id", 1]]
|
1723
|
-
SELECT "veterinarians".* FROM "veterinarians" WHERE "veterinarians"."dog_id" = ? [["dog_id", 1]]
|
1609
|
+
ActiveRecord::Base.time_zone_aware_types -= [:timestamptz]
|
1724
1610
|
```
|
1725
1611
|
|
1726
|
-
*
|
1727
|
-
|
1728
|
-
* `Arel::Visitors::Dot` now renders a complete set of properties when visiting
|
1729
|
-
`Arel::Nodes::SelectCore`, `SelectStatement`, `InsertStatement`, `UpdateStatement`, and
|
1730
|
-
`DeleteStatement`, which fixes #42026. Previously, some properties were omitted.
|
1731
|
-
|
1732
|
-
*Mike Dalessio*
|
1612
|
+
*Alex Ghiculescu*
|
1733
1613
|
|
1734
|
-
*
|
1735
|
-
`DistinctOn`, `Else`, `Except`, `InfixOperation`, `Intersect`, `Lock`, `NotRegexp`, `Quoted`,
|
1736
|
-
`Regexp`, `UnaryOperation`, `Union`, `UnionAll`, `When`, and `With`. Previously, these node
|
1737
|
-
types caused an exception to be raised by `Arel::Visitors::Dot#accept`.
|
1614
|
+
* Add new `ActiveRecord::Base.generates_token_for` API.
|
1738
1615
|
|
1739
|
-
|
1616
|
+
Currently, `signed_id` fulfills the role of generating tokens for e.g.
|
1617
|
+
resetting a password. However, signed IDs cannot reflect record state, so
|
1618
|
+
if a token is intended to be single-use, it must be tracked in a database at
|
1619
|
+
least until it expires.
|
1740
1620
|
|
1741
|
-
|
1621
|
+
With `generates_token_for`, a token can embed data from a record. When
|
1622
|
+
using the token to fetch the record, the data from the token and the current
|
1623
|
+
data from the record will be compared. If the two do not match, the token
|
1624
|
+
will be treated as invalid, the same as if it had expired. For example:
|
1742
1625
|
|
1743
1626
|
```ruby
|
1744
|
-
|
1745
|
-
|
1627
|
+
class User < ActiveRecord::Base
|
1628
|
+
has_secure_password
|
1746
1629
|
|
1747
|
-
|
1630
|
+
generates_token_for :password_reset, expires_in: 15.minutes do
|
1631
|
+
# A password's BCrypt salt changes when the password is updated.
|
1632
|
+
# By embedding (part of) the salt in a token, the token will
|
1633
|
+
# expire when the password is updated.
|
1634
|
+
BCrypt::Password.new(password_digest).salt[-10..]
|
1635
|
+
end
|
1636
|
+
end
|
1748
1637
|
|
1749
|
-
|
1750
|
-
|
1751
|
-
```
|
1638
|
+
user = User.first
|
1639
|
+
token = user.generate_token_for(:password_reset)
|
1752
1640
|
|
1753
|
-
|
1641
|
+
User.find_by_token_for(:password_reset, token) # => user
|
1754
1642
|
|
1755
|
-
|
1643
|
+
user.update!(password: "new password")
|
1644
|
+
User.find_by_token_for(:password_reset, token) # => nil
|
1645
|
+
```
|
1756
1646
|
|
1757
|
-
|
1758
|
-
By doing this the autosave callback are made more consistent for
|
1759
|
-
all 3 cases: `has_many`, `has_one`, and `belongs_to`.
|
1647
|
+
*Jonathan Hefner*
|
1760
1648
|
|
1761
|
-
|
1649
|
+
* Optimize Active Record batching for whole table iterations.
|
1762
1650
|
|
1763
|
-
|
1651
|
+
Previously, `in_batches` got all the ids and constructed an `IN`-based query for each batch.
|
1652
|
+
When iterating over the whole tables, this approach is not optimal as it loads unneeded ids and
|
1653
|
+
`IN` queries with lots of items are slow.
|
1764
1654
|
|
1765
|
-
|
1766
|
-
|
1767
|
-
|
1655
|
+
Now, whole table iterations use range iteration (`id >= x AND id <= y`) by default which can make iteration
|
1656
|
+
several times faster. E.g., tested on a PostgreSQL table with 10 million records: querying (`253s` vs `30s`),
|
1657
|
+
updating (`288s` vs `124s`), deleting (`268s` vs `83s`).
|
1768
1658
|
|
1769
|
-
|
1659
|
+
Only whole table iterations use this style of iteration by default. You can disable this behavior by passing `use_ranges: false`.
|
1660
|
+
If you iterate over the table and the only condition is, e.g., `archived_at: nil` (and only a tiny fraction
|
1661
|
+
of the records are archived), it makes sense to opt in to this approach:
|
1770
1662
|
|
1771
1663
|
```ruby
|
1772
|
-
|
1773
|
-
|
1774
|
-
has_many :humans
|
1664
|
+
Project.where(archived_at: nil).in_batches(use_ranges: true) do |relation|
|
1665
|
+
# do something
|
1775
1666
|
end
|
1776
1667
|
```
|
1777
1668
|
|
1778
|
-
|
1779
|
-
|
1780
|
-
```
|
1781
|
-
SELECT "humans"."id" FROM "humans" WHERE "humans"."dog_id" = ? [["dog_id", 1]]
|
1782
|
-
SELECT "treats".* FROM "treats" WHERE "treats"."human_id" IN (?, ?, ?) [["human_id", 1], ["human_id", 2], ["human_id", 3]]
|
1783
|
-
```
|
1784
|
-
|
1785
|
-
*Eileen M. Uchitelle*, *Aaron Patterson*, *Lee Quarella*
|
1786
|
-
|
1787
|
-
* Add setting for enumerating column names in SELECT statements.
|
1669
|
+
See #45414 for more details.
|
1788
1670
|
|
1789
|
-
|
1790
|
-
change the result of wildcard `SELECT *` queries, which invalidates the result
|
1791
|
-
of cached prepared statements and raises a `PreparedStatementCacheExpired` error.
|
1792
|
-
|
1793
|
-
When enabled, Active Record will avoid wildcards and always include column names
|
1794
|
-
in `SELECT` queries, which will return consistent results and avoid prepared
|
1795
|
-
statement errors.
|
1671
|
+
*fatkodima*
|
1796
1672
|
|
1797
|
-
|
1673
|
+
* `.with` query method added. Construct common table expressions with ease and get `ActiveRecord::Relation` back.
|
1798
1674
|
|
1799
1675
|
```ruby
|
1800
|
-
|
1801
|
-
#
|
1676
|
+
Post.with(posts_with_comments: Post.where("comments_count > ?", 0))
|
1677
|
+
# => ActiveRecord::Relation
|
1678
|
+
# WITH posts_with_comments AS (SELECT * FROM posts WHERE (comments_count > 0)) SELECT * FROM posts
|
1802
1679
|
```
|
1803
1680
|
|
1804
|
-
|
1805
|
-
|
1806
|
-
```ruby
|
1807
|
-
# config/application.rb
|
1808
|
-
module MyApp
|
1809
|
-
class Application < Rails::Application
|
1810
|
-
config.active_record.enumerate_columns_in_select_statements = true
|
1811
|
-
end
|
1812
|
-
end
|
1681
|
+
*Vlado Cingel*
|
1813
1682
|
|
1814
|
-
|
1815
|
-
class Book < ApplicationRecord
|
1816
|
-
self.enumerate_columns_in_select_statements = true
|
1817
|
-
end
|
1818
|
-
```
|
1683
|
+
* Don't establish a new connection if an identical pool exists already.
|
1819
1684
|
|
1820
|
-
|
1821
|
-
Book.limit(5)
|
1822
|
-
# SELECT id, author_id, name, format, status, language, etc FROM books LIMIT 5
|
1823
|
-
```
|
1685
|
+
Previously, if `establish_connection` was called on a class that already had an established connection, the existing connection would be removed regardless of whether it was the same config. Now if a pool is found with the same values as the new connection, the existing connection will be returned instead of creating a new one.
|
1824
1686
|
|
1825
|
-
|
1687
|
+
This has a slight change in behavior if application code is depending on a new connection being established regardless of whether it's identical to an existing connection. If the old behavior is desirable, applications should call `ActiveRecord::Base#remove_connection` before establishing a new one. Calling `establish_connection` with a different config works the same way as it did previously.
|
1826
1688
|
|
1827
|
-
*
|
1689
|
+
*Eileen M. Uchitelle*
|
1828
1690
|
|
1829
|
-
|
1830
|
-
Book.upsert_all(
|
1831
|
-
[{ id: 1, status: 1 }, { id: 2, status: 1 }],
|
1832
|
-
on_duplicate: Arel.sql("status = GREATEST(books.status, EXCLUDED.status)")
|
1833
|
-
)
|
1834
|
-
```
|
1691
|
+
* Update `db:prepare` task to load schema when an uninitialized database exists, and dump schema after migrations.
|
1835
1692
|
|
1836
|
-
*
|
1693
|
+
*Ben Sheldon*
|
1837
1694
|
|
1838
|
-
*
|
1695
|
+
* Fix supporting timezone awareness for `tsrange` and `tstzrange` array columns.
|
1839
1696
|
|
1840
1697
|
```ruby
|
1841
|
-
|
1842
|
-
|
1843
|
-
|
1844
|
-
|
1845
|
-
|
1846
|
-
|
1847
|
-
)
|
1698
|
+
# In database migrations
|
1699
|
+
add_column :shops, :open_hours, :tsrange, array: true
|
1700
|
+
# In app config
|
1701
|
+
ActiveRecord::Base.time_zone_aware_types += [:tsrange]
|
1702
|
+
# In the code times are properly converted to app time zone
|
1703
|
+
Shop.create!(open_hours: [Time.current..8.hour.from_now])
|
1848
1704
|
```
|
1849
1705
|
|
1850
|
-
*
|
1706
|
+
*Wojciech Wnętrzak*
|
1851
1707
|
|
1852
|
-
*
|
1708
|
+
* Introduce strategy pattern for executing migrations.
|
1853
1709
|
|
1854
|
-
|
1710
|
+
By default, migrations will use a strategy object that delegates the method
|
1711
|
+
to the connection adapter. Consumers can implement custom strategy objects
|
1712
|
+
to change how their migrations run.
|
1855
1713
|
|
1856
|
-
*
|
1714
|
+
*Adrianna Chang*
|
1857
1715
|
|
1858
|
-
|
1859
|
-
are regular Active Record attributes backed by a column with
|
1860
|
-
the same name. The system will transparently encrypt these
|
1861
|
-
attributes before saving them into the database and will
|
1862
|
-
decrypt them when retrieving their values.
|
1716
|
+
* Add adapter option disallowing foreign keys
|
1863
1717
|
|
1718
|
+
This adds a new option to be added to `database.yml` which enables skipping
|
1719
|
+
foreign key constraints usage even if the underlying database supports them.
|
1864
1720
|
|
1865
|
-
|
1866
|
-
|
1867
|
-
|
1868
|
-
|
1869
|
-
|
1721
|
+
Usage:
|
1722
|
+
```yaml
|
1723
|
+
development:
|
1724
|
+
<<: *default
|
1725
|
+
database: storage/development.sqlite3
|
1726
|
+
foreign_keys: false
|
1870
1727
|
```
|
1871
1728
|
|
1872
|
-
|
1873
|
-
guide](https://edgeguides.rubyonrails.org/active_record_encryption.html).
|
1729
|
+
*Paulo Barros*
|
1874
1730
|
|
1875
|
-
|
1731
|
+
* Add configurable deprecation warning for singular associations
|
1876
1732
|
|
1877
|
-
|
1878
|
-
|
1733
|
+
This adds a deprecation warning when using the plural name of a singular associations in `where`.
|
1734
|
+
It is possible to opt into the new more performant behavior with `config.active_record.allow_deprecated_singular_associations_name = false`
|
1879
1735
|
|
1880
|
-
*
|
1736
|
+
*Adam Hess*
|
1881
1737
|
|
1882
|
-
*
|
1738
|
+
* Run transactional callbacks on the freshest instance to save a given
|
1739
|
+
record within a transaction.
|
1883
1740
|
|
1884
|
-
|
1885
|
-
|
1741
|
+
When multiple Active Record instances change the same record within a
|
1742
|
+
transaction, Rails runs `after_commit` or `after_rollback` callbacks for
|
1743
|
+
only one of them. `config.active_record.run_commit_callbacks_on_first_saved_instances_in_transaction`
|
1744
|
+
was added to specify how Rails chooses which instance receives the
|
1745
|
+
callbacks. The framework defaults were changed to use the new logic.
|
1886
1746
|
|
1887
|
-
|
1888
|
-
|
1747
|
+
When `config.active_record.run_commit_callbacks_on_first_saved_instances_in_transaction`
|
1748
|
+
is `true`, transactional callbacks are run on the first instance to save,
|
1749
|
+
even though its instance state may be stale.
|
1889
1750
|
|
1890
|
-
|
1891
|
-
|
1892
|
-
|
1751
|
+
When it is `false`, which is the new framework default starting with version
|
1752
|
+
7.1, transactional callbacks are run on the instances with the freshest
|
1753
|
+
instance state. Those instances are chosen as follows:
|
1893
1754
|
|
1894
|
-
|
1895
|
-
|
1896
|
-
|
1897
|
-
|
1755
|
+
- In general, run transactional callbacks on the last instance to save a
|
1756
|
+
given record within the transaction.
|
1757
|
+
- There are two exceptions:
|
1758
|
+
- If the record is created within the transaction, then updated by
|
1759
|
+
another instance, `after_create_commit` callbacks will be run on the
|
1760
|
+
second instance. This is instead of the `after_update_commit`
|
1761
|
+
callbacks that would naively be run based on that instance’s state.
|
1762
|
+
- If the record is destroyed within the transaction, then
|
1763
|
+
`after_destroy_commit` callbacks will be fired on the last destroyed
|
1764
|
+
instance, even if a stale instance subsequently performed an update
|
1765
|
+
(which will have affected 0 rows).
|
1898
1766
|
|
1899
|
-
*
|
1767
|
+
*Cameron Bothner and Mitch Vollebregt*
|
1900
1768
|
|
1901
|
-
*
|
1769
|
+
* Enable strict strings mode for `SQLite3Adapter`.
|
1902
1770
|
|
1903
|
-
|
1771
|
+
Configures SQLite with a strict strings mode, which disables double-quoted string literals.
|
1904
1772
|
|
1905
|
-
|
1906
|
-
|
1907
|
-
|
1773
|
+
SQLite has some quirks around double-quoted string literals.
|
1774
|
+
It first tries to consider double-quoted strings as identifier names, but if they don't exist
|
1775
|
+
it then considers them as string literals. Because of this, typos can silently go unnoticed.
|
1776
|
+
For example, it is possible to create an index for a non existing column.
|
1777
|
+
See [SQLite documentation](https://www.sqlite.org/quirks.html#double_quoted_string_literals_are_accepted) for more details.
|
1908
1778
|
|
1909
|
-
|
1910
|
-
create_table "postgresql_infinities" do |t|
|
1911
|
-
t.datetime "datetime"
|
1912
|
-
end
|
1779
|
+
If you don't want this behavior, you can disable it via:
|
1913
1780
|
|
1914
|
-
|
1915
|
-
|
1916
|
-
|
1781
|
+
```ruby
|
1782
|
+
# config/application.rb
|
1783
|
+
config.active_record.sqlite3_adapter_strict_strings_by_default = false
|
1917
1784
|
```
|
1918
1785
|
|
1919
|
-
|
1786
|
+
Fixes #27782.
|
1787
|
+
|
1788
|
+
*fatkodima*, *Jean Boussier*
|
1920
1789
|
|
1921
|
-
|
1790
|
+
* Resolve issue where a relation cache_version could be left stale.
|
1922
1791
|
|
1923
|
-
|
1792
|
+
Previously, when `reset` was called on a relation object it did not reset the cache_versions
|
1793
|
+
ivar. This led to a confusing situation where despite having the correct data the relation
|
1794
|
+
still reported a stale cache_version.
|
1924
1795
|
|
1925
|
-
|
1796
|
+
Usage:
|
1926
1797
|
|
1927
1798
|
```ruby
|
1928
|
-
|
1929
|
-
|
1930
|
-
end
|
1931
|
-
```
|
1799
|
+
developers = Developer.all
|
1800
|
+
developers.cache_version
|
1932
1801
|
|
1933
|
-
|
1802
|
+
Developer.update_all(updated_at: Time.now.utc + 1.second)
|
1934
1803
|
|
1935
|
-
|
1936
|
-
|
1937
|
-
|
1938
|
-
# => #<Book id: 1, status: "proposed", ...> (for mysql2 adapter)
|
1939
|
-
# => ActiveRecord::StatementInvalid: PG::InvalidTextRepresentation: ERROR: invalid input syntax for type integer: "prohibited" (for postgresql adapter)
|
1940
|
-
# => nil (for sqlite3 adapter)
|
1804
|
+
developers.cache_version # Stale cache_version
|
1805
|
+
developers.reset
|
1806
|
+
developers.cache_version # Returns the current correct cache_version
|
1941
1807
|
```
|
1942
1808
|
|
1943
|
-
|
1809
|
+
Fixes #45341.
|
1810
|
+
|
1811
|
+
*Austen Madden*
|
1812
|
+
|
1813
|
+
* Add support for exclusion constraints (PostgreSQL-only).
|
1944
1814
|
|
1945
1815
|
```ruby
|
1946
|
-
|
1947
|
-
|
1948
|
-
# => nil (for all adapters)
|
1816
|
+
add_exclusion_constraint :invoices, "daterange(start_date, end_date) WITH &&", using: :gist, name: "invoices_date_overlap"
|
1817
|
+
remove_exclusion_constraint :invoices, name: "invoices_date_overlap"
|
1949
1818
|
```
|
1950
1819
|
|
1951
|
-
|
1820
|
+
See PostgreSQL's [`CREATE TABLE ... EXCLUDE ...`](https://www.postgresql.org/docs/12/sql-createtable.html#SQL-CREATETABLE-EXCLUDE) documentation for more on exclusion constraints.
|
1952
1821
|
|
1953
|
-
*
|
1822
|
+
*Alex Robbin*
|
1954
1823
|
|
1955
|
-
|
1824
|
+
* `change_column_null` raises if a non-boolean argument is provided
|
1956
1825
|
|
1957
|
-
|
1958
|
-
|
1959
|
-
|
1960
|
-
name: George the Monkey
|
1961
|
-
fruits: apple
|
1826
|
+
Previously if you provided a non-boolean argument, `change_column_null` would
|
1827
|
+
treat it as truthy and make your column nullable. This could be surprising, so now
|
1828
|
+
the input must be either `true` or `false`.
|
1962
1829
|
|
1963
|
-
|
1964
|
-
|
1965
|
-
|
1830
|
+
```ruby
|
1831
|
+
change_column_null :table, :column, true # good
|
1832
|
+
change_column_null :table, :column, false # good
|
1833
|
+
change_column_null :table, :column, from: true, to: false # raises (previously this made the column nullable)
|
1966
1834
|
```
|
1967
1835
|
|
1968
|
-
If the join table (`fruit_monkeys`) contains `created_at` or `updated_at` columns,
|
1969
|
-
these will now be populated when loading the fixture. Previously, fixture loading
|
1970
|
-
would crash if these columns were required, and leave them as null otherwise.
|
1971
|
-
|
1972
1836
|
*Alex Ghiculescu*
|
1973
1837
|
|
1974
|
-
*
|
1838
|
+
* Enforce limit on table names length.
|
1975
1839
|
|
1976
|
-
|
1977
|
-
a single global thread pool for all queries. By default, Rails will set `async_query_executor`
|
1978
|
-
to `nil` which will not initialize any executor. If `load_async` is called and no executor
|
1979
|
-
has been configured, the query will be executed in the foreground.
|
1840
|
+
Fixes #45130.
|
1980
1841
|
|
1981
|
-
|
1982
|
-
`config.active_record.async_query_executor` to `:global_thread_pool` and optionally define
|
1983
|
-
`config.active_record.global_executor_concurrency`. This defaults to 4. For applications that want
|
1984
|
-
to have a thread pool for each database connection, `config.active_record.async_query_executor` can
|
1985
|
-
be set to `:multi_thread_pool`. The configuration for each thread pool is set in the database
|
1986
|
-
configuration.
|
1842
|
+
*fatkodima*
|
1987
1843
|
|
1988
|
-
|
1844
|
+
* Adjust the minimum MariaDB version for check constraints support.
|
1989
1845
|
|
1990
|
-
*
|
1846
|
+
*Eddie Lebow*
|
1991
1847
|
|
1992
|
-
|
1848
|
+
* Fix Hstore deserialize regression.
|
1993
1849
|
|
1994
|
-
|
1995
|
-
class Book < ActiveRecord::Base
|
1996
|
-
enum status: [ :proposed, :written ], _prefix: true, _scopes: false
|
1997
|
-
enum cover: [ :hard, :soft ], _suffix: true, _default: :hard
|
1998
|
-
end
|
1999
|
-
```
|
1850
|
+
*edsharp*
|
2000
1851
|
|
2001
|
-
|
1852
|
+
* Add validity for PostgreSQL indexes.
|
2002
1853
|
|
2003
1854
|
```ruby
|
2004
|
-
|
2005
|
-
|
2006
|
-
enum :cover, [ :hard, :soft ], suffix: true, default: :hard
|
2007
|
-
end
|
1855
|
+
connection.index_exists?(:users, :email, valid: true)
|
1856
|
+
connection.indexes(:users).select(&:valid?)
|
2008
1857
|
```
|
2009
1858
|
|
2010
|
-
*
|
1859
|
+
*fatkodima*
|
2011
1860
|
|
2012
|
-
*
|
1861
|
+
* Fix eager loading for models without primary keys.
|
2013
1862
|
|
2014
|
-
|
1863
|
+
*Anmol Chopra*, *Matt Lawrence*, and *Jonathan Hefner*
|
2015
1864
|
|
2016
|
-
|
2017
|
-
the query, it will be performed in the foreground.
|
1865
|
+
* Avoid validating a unique field if it has not changed and is backed by a unique index.
|
2018
1866
|
|
2019
|
-
|
2020
|
-
|
1867
|
+
Previously, when saving a record, Active Record will perform an extra query to check for the
|
1868
|
+
uniqueness of each attribute having a `uniqueness` validation, even if that attribute hasn't changed.
|
1869
|
+
If the database has the corresponding unique index, then this validation can never fail for persisted
|
1870
|
+
records, and we could safely skip it.
|
2021
1871
|
|
2022
|
-
|
2023
|
-
def index
|
2024
|
-
@categories = Category.some_complex_scope.load_async
|
2025
|
-
@posts = Post.some_complex_scope.load_async
|
2026
|
-
end
|
2027
|
-
```
|
1872
|
+
*fatkodima*
|
2028
1873
|
|
2029
|
-
|
2030
|
-
the main thread had to wait to access the result. This timing is useful to know
|
2031
|
-
whether or not it's worth to load the query asynchronously.
|
1874
|
+
* Stop setting `sql_auto_is_null`
|
2032
1875
|
|
2033
|
-
|
2034
|
-
DEBUG -- : Category Load (62.1ms) SELECT * FROM `categories` LIMIT 50
|
2035
|
-
DEBUG -- : ASYNC Post Load (64ms) (db time 126.1ms) SELECT * FROM `posts` LIMIT 100
|
2036
|
-
```
|
1876
|
+
Since version 5.5 the default has been off, we no longer have to manually turn it off.
|
2037
1877
|
|
2038
|
-
|
2039
|
-
waiting for the results, and the second set of parens with "db time" is how long
|
2040
|
-
the entire query took to execute.
|
1878
|
+
*Adam Hess*
|
2041
1879
|
|
2042
|
-
|
1880
|
+
* Fix `touch` to raise an error for readonly columns.
|
2043
1881
|
|
2044
|
-
*
|
1882
|
+
*fatkodima*
|
2045
1883
|
|
2046
|
-
|
2047
|
-
the resulting relation:
|
1884
|
+
* Add ability to ignore tables by regexp for SQL schema dumps.
|
2048
1885
|
|
2049
1886
|
```ruby
|
2050
|
-
|
2051
|
-
Post.excluding(post_one, post_two)
|
1887
|
+
ActiveRecord::SchemaDumper.ignore_tables = [/^_/]
|
2052
1888
|
```
|
2053
1889
|
|
2054
|
-
|
2055
|
-
|
2056
|
-
```ruby
|
2057
|
-
post.comments.excluding(comment)
|
2058
|
-
post.comments.excluding(comment_one, comment_two)
|
2059
|
-
```
|
1890
|
+
*fatkodima*
|
2060
1891
|
|
2061
|
-
|
2062
|
-
and `Post.where.not(id: [post_one.id, post_two.id])` (for a collection).
|
1892
|
+
* Avoid queries when performing calculations on contradictory relations.
|
2063
1893
|
|
2064
|
-
|
1894
|
+
Previously calculations would make a query even when passed a
|
1895
|
+
contradiction, such as `User.where(id: []).count`. We no longer perform a
|
1896
|
+
query in that scenario.
|
2065
1897
|
|
2066
|
-
|
2067
|
-
|
1898
|
+
This applies to the following calculations: `count`, `sum`, `average`,
|
1899
|
+
`minimum` and `maximum`
|
2068
1900
|
|
2069
|
-
|
2070
|
-
references an aliased select value would generate an error when
|
2071
|
-
#include? was called, due to an optimisation that would generate
|
2072
|
-
call #exists? on the relation instead, which effectively alters
|
2073
|
-
the select values of the query (and thus removes the aliased select
|
2074
|
-
values), but leaves the having clause intact. Because the having
|
2075
|
-
clause is then referencing an aliased column that is no longer
|
2076
|
-
present in the simplified query, an ActiveRecord::InvalidStatement
|
2077
|
-
error was raised.
|
1901
|
+
*Luan Vieira, John Hawthorn and Daniel Colson*
|
2078
1902
|
|
2079
|
-
|
1903
|
+
* Allow using aliased attributes with `insert_all`/`upsert_all`.
|
2080
1904
|
|
2081
1905
|
```ruby
|
2082
|
-
|
2083
|
-
|
2084
|
-
|
2085
|
-
|
2086
|
-
|
1906
|
+
class Book < ApplicationRecord
|
1907
|
+
alias_attribute :title, :name
|
1908
|
+
end
|
1909
|
+
|
1910
|
+
Book.insert_all [{ title: "Remote", author_id: 1 }], returning: :title
|
2087
1911
|
```
|
2088
1912
|
|
2089
|
-
|
2090
|
-
simplified #exists? query, which simply checks for the presence of
|
2091
|
-
a having clause.
|
1913
|
+
*fatkodima*
|
2092
1914
|
|
2093
|
-
|
1915
|
+
* Support encrypted attributes on columns with default db values.
|
2094
1916
|
|
2095
|
-
|
1917
|
+
This adds support for encrypted attributes defined on columns with default values.
|
1918
|
+
It will encrypt those values at creation time. Before, it would raise an
|
1919
|
+
error unless `config.active_record.encryption.support_unencrypted_data` was true.
|
2096
1920
|
|
2097
|
-
*
|
2098
|
-
without Rails knowledge (e.g., if app gets killed during long-running query or due to Rack::Timeout), app won't end
|
2099
|
-
up in perpetual crash state for being inconsistent with PostgreSQL.
|
1921
|
+
*Jorge Manrubia* and *Dima Fatko*
|
2100
1922
|
|
2101
|
-
|
1923
|
+
* Allow overriding `reading_request?` in `DatabaseSelector::Resolver`
|
2102
1924
|
|
2103
|
-
|
1925
|
+
The default implementation checks if a request is a `get?` or `head?`,
|
1926
|
+
but you can now change it to anything you like. If the method returns true,
|
1927
|
+
`Resolver#read` gets called meaning the request could be served by the
|
1928
|
+
replica database.
|
2104
1929
|
|
2105
|
-
|
2106
|
-
worked on certain types of queries. This change allows the `scoping` method to apply
|
2107
|
-
to all queries for a model in a block.
|
1930
|
+
*Alex Ghiculescu*
|
2108
1931
|
|
2109
|
-
|
2110
|
-
Post.where(blog_id: post.blog_id).scoping(all_queries: true) do
|
2111
|
-
post.update(title: "a post title") # adds `posts.blog_id = 1` to the query
|
2112
|
-
end
|
2113
|
-
```
|
1932
|
+
* Remove `ActiveRecord.legacy_connection_handling`.
|
2114
1933
|
|
2115
1934
|
*Eileen M. Uchitelle*
|
2116
1935
|
|
2117
|
-
* `
|
2118
|
-
(aliased as `ActiveRecord::Calculations.average`) will now use column-based
|
2119
|
-
type casting. This means that floating-point number columns will now be
|
2120
|
-
aggregated as `Float` and decimal columns will be aggregated as `BigDecimal`.
|
1936
|
+
* `rails db:schema:{dump,load}` now checks `ENV["SCHEMA_FORMAT"]` before config
|
2121
1937
|
|
2122
|
-
|
2123
|
-
|
1938
|
+
Since `rails db:structure:{dump,load}` was deprecated there wasn't a simple
|
1939
|
+
way to dump a schema to both SQL and Ruby formats. You can now do this with
|
1940
|
+
an environment variable. For example:
|
2124
1941
|
|
2125
|
-
```
|
2126
|
-
|
2127
|
-
|
2128
|
-
t.float "temperature"
|
2129
|
-
end
|
1942
|
+
```
|
1943
|
+
SCHEMA_FORMAT=sql rake db:schema:dump
|
1944
|
+
```
|
2130
1945
|
|
2131
|
-
|
2132
|
-
Measurement.average(:temperature).class
|
2133
|
-
# => BigDecimal
|
1946
|
+
*Alex Ghiculescu*
|
2134
1947
|
|
2135
|
-
|
2136
|
-
|
2137
|
-
|
2138
|
-
|
1948
|
+
* Fixed MariaDB default function support.
|
1949
|
+
|
1950
|
+
Defaults would be written wrong in "db/schema.rb" and not work correctly
|
1951
|
+
if using `db:schema:load`. Further more the function name would be
|
1952
|
+
added as string content when saving new records.
|
2139
1953
|
|
2140
|
-
|
2141
|
-
database adapter. This is not the case anymore. If you relied on that kind
|
2142
|
-
of magic, you now need to register your own `ActiveRecord::Type`
|
2143
|
-
(see `ActiveRecord::Attributes::ClassMethods` for documentation).
|
1954
|
+
*kaspernj*
|
2144
1955
|
|
2145
|
-
|
1956
|
+
* Add `active_record.destroy_association_async_batch_size` configuration
|
2146
1957
|
|
2147
|
-
|
1958
|
+
This allows applications to specify the maximum number of records that will
|
1959
|
+
be destroyed in a single background job by the `dependent: :destroy_async`
|
1960
|
+
association option. By default, the current behavior will remain the same:
|
1961
|
+
when a parent record is destroyed, all dependent records will be destroyed
|
1962
|
+
in a single background job. If the number of dependent records is greater
|
1963
|
+
than this configuration, the records will be destroyed in multiple
|
1964
|
+
background jobs.
|
2148
1965
|
|
2149
|
-
|
2150
|
-
a migration or schema. It takes a symbol which must correspond to one of the configured
|
2151
|
-
`NATIVE_DATABASE_TYPES`. The default is `:timestamp`, meaning `t.datetime` in a migration
|
2152
|
-
will create a "timestamp without time zone" column. To use "timestamp with time zone",
|
2153
|
-
change this to `:timestamptz` in an initializer.
|
1966
|
+
*Nick Holden*
|
2154
1967
|
|
2155
|
-
|
1968
|
+
* Fix `remove_foreign_key` with `:if_exists` option when foreign key actually exists.
|
2156
1969
|
|
2157
|
-
*
|
1970
|
+
*fatkodima*
|
1971
|
+
|
1972
|
+
* Remove `--no-comments` flag in structure dumps for PostgreSQL
|
2158
1973
|
|
2159
|
-
|
1974
|
+
This broke some apps that used custom schema comments. If you don't want
|
1975
|
+
comments in your structure dump, you can use:
|
2160
1976
|
|
2161
|
-
|
2162
|
-
|
1977
|
+
```ruby
|
1978
|
+
ActiveRecord::Tasks::DatabaseTasks.structure_dump_flags = ['--no-comments']
|
1979
|
+
```
|
2163
1980
|
|
2164
1981
|
*Alex Ghiculescu*
|
2165
1982
|
|
2166
|
-
*
|
2167
|
-
`ActiveRecord::Sanitization.disallow_raw_sql!`.
|
1983
|
+
* Reduce the memory footprint of fixtures accessors.
|
2168
1984
|
|
2169
|
-
|
1985
|
+
Until now fixtures accessors were eagerly defined using `define_method`.
|
1986
|
+
So the memory usage was directly dependent of the number of fixtures and
|
1987
|
+
test suites.
|
2170
1988
|
|
2171
|
-
|
1989
|
+
Instead fixtures accessors are now implemented with `method_missing`,
|
1990
|
+
so they incur much less memory and CPU overhead.
|
2172
1991
|
|
2173
|
-
|
2174
|
-
named `ApplicationRecord` can now set a specific class to be the `primary_abstract_class`.
|
1992
|
+
*Jean Boussier*
|
2175
1993
|
|
2176
|
-
|
2177
|
-
class PrimaryApplicationRecord
|
2178
|
-
self.primary_abstract_class
|
2179
|
-
end
|
2180
|
-
```
|
1994
|
+
* Fix `config.active_record.destroy_association_async_job` configuration
|
2181
1995
|
|
2182
|
-
|
2183
|
-
|
2184
|
-
|
2185
|
-
|
2186
|
-
|
1996
|
+
`config.active_record.destroy_association_async_job` should allow
|
1997
|
+
applications to specify the job that will be used to destroy associated
|
1998
|
+
records in the background for `has_many` associations with the
|
1999
|
+
`dependent: :destroy_async` option. Previously, that was ignored, which
|
2000
|
+
meant the default `ActiveRecord::DestroyAssociationAsyncJob` always
|
2001
|
+
destroyed records in the background.
|
2187
2002
|
|
2188
|
-
*
|
2003
|
+
*Nick Holden*
|
2189
2004
|
|
2190
|
-
*
|
2191
|
-
Now that Active Record supports multiple databases configuration,
|
2192
|
-
we need a way to pass specific flags for dump/load databases since
|
2193
|
-
the options are not the same for different adapters.
|
2194
|
-
We can use in the original way:
|
2005
|
+
* Fix `change_column_comment` to preserve column's AUTO_INCREMENT in the MySQL adapter
|
2195
2006
|
|
2196
|
-
|
2197
|
-
ActiveRecord::Tasks::DatabaseTasks.structure_dump_flags = ['--no-defaults', '--skip-add-drop-table']
|
2198
|
-
# or
|
2199
|
-
ActiveRecord::Tasks::DatabaseTasks.structure_dump_flags = '--no-defaults --skip-add-drop-table'
|
2200
|
-
```
|
2007
|
+
*fatkodima*
|
2201
2008
|
|
2202
|
-
|
2203
|
-
is the adapter
|
2009
|
+
* Fix quoting of `ActiveSupport::Duration` and `Rational` numbers in the MySQL adapter.
|
2204
2010
|
|
2205
|
-
|
2206
|
-
ActiveRecord::Tasks::DatabaseTasks.structure_dump_flags = {
|
2207
|
-
mysql2: ['--no-defaults', '--skip-add-drop-table'],
|
2208
|
-
postgres: '--no-tablespaces'
|
2209
|
-
}
|
2210
|
-
```
|
2011
|
+
*Kevin McPhillips*
|
2211
2012
|
|
2212
|
-
|
2013
|
+
* Allow column name with COLLATE (e.g., title COLLATE "C") as safe SQL string
|
2213
2014
|
|
2214
|
-
*
|
2215
|
-
adapter if the "url" protocol is "jdbc", "http", or "https". Previously only
|
2216
|
-
urls with the "jdbc" prefix were passed to the Active Record Adapter, others
|
2217
|
-
are assumed to be adapter specification urls.
|
2015
|
+
*Shugo Maeda*
|
2218
2016
|
|
2219
|
-
|
2017
|
+
* Permit underscores in the VERSION argument to database rake tasks.
|
2220
2018
|
|
2221
|
-
*
|
2019
|
+
*Eddie Lebow*
|
2222
2020
|
|
2223
|
-
*
|
2021
|
+
* Reversed the order of `INSERT` statements in `structure.sql` dumps
|
2224
2022
|
|
2225
|
-
This
|
2226
|
-
|
2023
|
+
This should decrease the likelihood of merge conflicts. New migrations
|
2024
|
+
will now be added at the top of the list.
|
2227
2025
|
|
2228
|
-
|
2229
|
-
|
2230
|
-
has_many :bookmarks
|
2231
|
-
has_many :articles, strict_loading: true
|
2232
|
-
end
|
2026
|
+
For existing apps, there will be a large diff the next time `structure.sql`
|
2027
|
+
is generated.
|
2233
2028
|
|
2234
|
-
|
2235
|
-
user.articles # => ActiveRecord::StrictLoadingViolationError
|
2236
|
-
user.bookmarks # => #<ActiveRecord::Associations::CollectionProxy>
|
2029
|
+
*Alex Ghiculescu*, *Matt Larraz*
|
2237
2030
|
|
2238
|
-
|
2239
|
-
user.bookmarks # => ActiveRecord::StrictLoadingViolationError
|
2031
|
+
* Fix PG.connect keyword arguments deprecation warning on ruby 2.7
|
2240
2032
|
|
2241
|
-
|
2242
|
-
user.bookmarks # => #<ActiveRecord::Associations::CollectionProxy>
|
2243
|
-
user.articles.strict_loading!(false) # => #<ActiveRecord::Associations::CollectionProxy>
|
2244
|
-
```
|
2033
|
+
Fixes #44307.
|
2245
2034
|
|
2246
|
-
*
|
2035
|
+
*Nikita Vasilevsky*
|
2247
2036
|
|
2248
|
-
*
|
2249
|
-
presence of exactly one record.
|
2037
|
+
* Fix dropping DB connections after serialization failures and deadlocks.
|
2250
2038
|
|
2251
|
-
|
2252
|
-
|
2253
|
-
|
2039
|
+
Prior to 6.1.4, serialization failures and deadlocks caused rollbacks to be
|
2040
|
+
issued for both real transactions and savepoints. This breaks MySQL which
|
2041
|
+
disallows rollbacks of savepoints following a deadlock.
|
2254
2042
|
|
2255
|
-
|
2256
|
-
|
2257
|
-
# => ActiveRecord::RecordNotFound (if no Product with given price)
|
2258
|
-
# => #<Product ...> (if one Product with given price)
|
2259
|
-
# => ActiveRecord::SoleRecordExceeded (if more than one Product with given price)
|
2043
|
+
6.1.4 removed these rollbacks, for both transactions and savepoints, causing
|
2044
|
+
the DB connection to be left in an unknown state and thus discarded.
|
2260
2045
|
|
2261
|
-
|
2262
|
-
# as above
|
2263
|
-
```
|
2046
|
+
These rollbacks are now restored, except for savepoints on MySQL.
|
2264
2047
|
|
2265
|
-
*
|
2048
|
+
*Thomas Morgan*
|
2266
2049
|
|
2267
|
-
*
|
2050
|
+
* Make `ActiveRecord::ConnectionPool` Fiber-safe
|
2268
2051
|
|
2269
|
-
|
2052
|
+
When `ActiveSupport::IsolatedExecutionState.isolation_level` is set to `:fiber`,
|
2053
|
+
the connection pool now supports multiple Fibers from the same Thread checking
|
2054
|
+
out connections from the pool.
|
2055
|
+
|
2056
|
+
*Alex Matchneer*
|
2057
|
+
|
2058
|
+
* Add `update_attribute!` to `ActiveRecord::Persistence`
|
2059
|
+
|
2060
|
+
Similar to `update_attribute`, but raises `ActiveRecord::RecordNotSaved` when a `before_*` callback throws `:abort`.
|
2270
2061
|
|
2271
2062
|
```ruby
|
2272
|
-
class
|
2273
|
-
|
2274
|
-
|
2063
|
+
class Topic < ActiveRecord::Base
|
2064
|
+
before_save :check_title
|
2065
|
+
|
2066
|
+
def check_title
|
2067
|
+
throw(:abort) if title == "abort"
|
2275
2068
|
end
|
2276
2069
|
end
|
2277
2070
|
|
2278
|
-
|
2279
|
-
|
2280
|
-
|
2281
|
-
|
2282
|
-
|
2071
|
+
topic = Topic.create(title: "Test Title")
|
2072
|
+
# #=> #<Topic title: "Test Title">
|
2073
|
+
topic.update_attribute!(:title, "Another Title")
|
2074
|
+
# #=> #<Topic title: "Another Title">
|
2075
|
+
topic.update_attribute!(:title, "abort")
|
2076
|
+
# raises ActiveRecord::RecordNotSaved
|
2283
2077
|
```
|
2284
2078
|
|
2285
|
-
|
2079
|
+
*Drew Tempelmeyer*
|
2286
2080
|
|
2287
|
-
|
2081
|
+
* Avoid loading every record in `ActiveRecord::Relation#pretty_print`
|
2288
2082
|
|
2289
|
-
|
2083
|
+
```ruby
|
2084
|
+
# Before
|
2085
|
+
pp Foo.all # Loads the whole table.
|
2086
|
+
|
2087
|
+
# After
|
2088
|
+
pp Foo.all # Shows 10 items and an ellipsis.
|
2089
|
+
```
|
2290
2090
|
|
2291
|
-
*
|
2091
|
+
*Ulysse Buonomo*
|
2292
2092
|
|
2293
|
-
|
2294
|
-
`singular_id` defined by it does not work when the foreign_key does
|
2295
|
-
not end with `id`. This change fixes it by taking into account
|
2296
|
-
`primary_key` and `foreign_key` in the options.
|
2093
|
+
* Change `QueryMethods#in_order_of` to drop records not listed in values.
|
2297
2094
|
|
2298
|
-
|
2095
|
+
`in_order_of` now filters down to the values provided, to match the behavior of the `Enumerable` version.
|
2299
2096
|
|
2300
|
-
*
|
2097
|
+
*Kevin Newton*
|
2301
2098
|
|
2302
|
-
|
2303
|
-
class User
|
2304
|
-
scope :active, -> { where(accepted: true, locked: false) }
|
2305
|
-
end
|
2099
|
+
* Allow named expression indexes to be revertible.
|
2306
2100
|
|
2307
|
-
|
2308
|
-
# ... WHERE `accepted` = 1 AND `locked` = 0
|
2101
|
+
Previously, the following code would raise an error in a reversible migration executed while rolling back, due to the index name not being used in the index removal.
|
2309
2102
|
|
2310
|
-
|
2311
|
-
|
2103
|
+
```ruby
|
2104
|
+
add_index(:settings, "(data->'property')", using: :gin, name: :index_settings_data_property)
|
2312
2105
|
```
|
2313
2106
|
|
2314
|
-
|
2107
|
+
Fixes #43331.
|
2108
|
+
|
2109
|
+
*Oliver Günther*
|
2315
2110
|
|
2316
|
-
*
|
2111
|
+
* Fix incorrect argument in PostgreSQL structure dump tasks.
|
2317
2112
|
|
2318
|
-
|
2319
|
-
to throw an error saying :polymorphic would not be a valid option.
|
2113
|
+
Updating the `--no-comment` argument added in Rails 7 to the correct `--no-comments` argument.
|
2320
2114
|
|
2321
|
-
*
|
2115
|
+
*Alex Dent*
|
2322
2116
|
|
2323
|
-
*
|
2117
|
+
* Fix migration compatibility to create SQLite references/belongs_to column as integer when migration version is 6.0.
|
2324
2118
|
|
2325
|
-
|
2119
|
+
Reference/belongs_to in migrations with version 6.0 were creating columns as
|
2120
|
+
bigint instead of integer for the SQLite Adapter.
|
2326
2121
|
|
2327
|
-
*
|
2122
|
+
*Marcelo Lauxen*
|
2328
2123
|
|
2329
|
-
|
2330
|
-
and the index name should be specified explicitly. Now, the index name is inferred
|
2331
|
-
automatically.
|
2124
|
+
* Fix `QueryMethods#in_order_of` to handle empty order list.
|
2332
2125
|
|
2333
2126
|
```ruby
|
2334
|
-
|
2127
|
+
Post.in_order_of(:id, []).to_a
|
2335
2128
|
```
|
2336
2129
|
|
2337
|
-
|
2338
|
-
|
2339
|
-
*fatkodima*
|
2340
|
-
|
2341
|
-
* Only warn about negative enums if a positive form that would cause conflicts exists.
|
2342
|
-
|
2343
|
-
Fixes #39065.
|
2344
|
-
|
2345
|
-
*Alex Ghiculescu*
|
2130
|
+
Also more explicitly set the column as secondary order, so that any other
|
2131
|
+
value is still ordered.
|
2346
2132
|
|
2347
|
-
*
|
2133
|
+
*Jean Boussier*
|
2348
2134
|
|
2349
|
-
|
2135
|
+
* Fix quoting of column aliases generated by calculation methods.
|
2350
2136
|
|
2351
|
-
|
2137
|
+
Since the alias is derived from the table name, we can't assume the result
|
2138
|
+
is a valid identifier.
|
2352
2139
|
|
2353
2140
|
```ruby
|
2354
|
-
class
|
2355
|
-
|
2141
|
+
class Test < ActiveRecord::Base
|
2142
|
+
self.table_name = '1abc'
|
2356
2143
|
end
|
2144
|
+
Test.group(:id).count
|
2145
|
+
# syntax error at or near "1" (ActiveRecord::StatementInvalid)
|
2146
|
+
# LINE 1: SELECT COUNT(*) AS count_all, "1abc"."id" AS 1abc_id FROM "1...
|
2357
2147
|
```
|
2358
2148
|
|
2359
|
-
*
|
2149
|
+
*Jean Boussier*
|
2360
2150
|
|
2361
|
-
* Add `
|
2151
|
+
* Add `authenticate_by` when using `has_secure_password`.
|
2362
2152
|
|
2363
|
-
|
2364
|
-
|
2365
|
-
account.users.joins(:contact).where.not(contact_id: nil)
|
2153
|
+
`authenticate_by` is intended to replace code like the following, which
|
2154
|
+
returns early when a user with a matching email is not found:
|
2366
2155
|
|
2367
|
-
|
2368
|
-
|
2156
|
+
```ruby
|
2157
|
+
User.find_by(email: "...")&.authenticate("...")
|
2369
2158
|
```
|
2370
2159
|
|
2371
|
-
|
2160
|
+
Such code is vulnerable to timing-based enumeration attacks, wherein an
|
2161
|
+
attacker can determine if a user account with a given email exists. After
|
2162
|
+
confirming that an account exists, the attacker can try passwords associated
|
2163
|
+
with that email address from other leaked databases, in case the user
|
2164
|
+
re-used a password across multiple sites (a common practice). Additionally,
|
2165
|
+
knowing an account email address allows the attacker to attempt a targeted
|
2166
|
+
phishing ("spear phishing") attack.
|
2372
2167
|
|
2373
|
-
|
2168
|
+
`authenticate_by` addresses the vulnerability by taking the same amount of
|
2169
|
+
time regardless of whether a user with a matching email is found:
|
2374
2170
|
|
2375
|
-
|
2376
|
-
|
2171
|
+
```ruby
|
2172
|
+
User.authenticate_by(email: "...", password: "...")
|
2173
|
+
```
|
2377
2174
|
|
2378
|
-
*
|
2175
|
+
*Jonathan Hefner*
|
2379
2176
|
|
2380
2177
|
|
2381
|
-
Please check [
|
2178
|
+
Please check [7-0-stable](https://github.com/rails/rails/blob/7-0-stable/activerecord/CHANGELOG.md) for previous changes.
|