| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|\
| |
| |
| |
| |
| |
| | |
Checks to see if the record contains the foreign_key to set the inverse automatically
Conflicts:
activerecord/CHANGELOG.md
|
| |
| |
| |
| | |
automatically
|
|\ \
| | |
| | |
| | |
| | | |
dougbarth/dont_swallow_exceptions_during_transactional_statements_in_mysql
Don't swallow exceptions in transctional statements
|
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
The MySQL connection adapater swallows all StandardError exceptions,
which includes Mysql::Error and Mysql2::Error. The comment in the
exception clause claims errors thrown here indicate that transactions
aren't supported by the server but that isn't necessarily true. It's
possible the MySQL server has gone away and swallowing a failed commit
may let the application return a successful response when the data has
not been saved. Also, replication libraries like Galera require that the
application handle exceptions thrown at BEGIN/COMMIT.
I'm unable to determine what version of MySQL threw an exception for
transactional statements. I tried as far back as 3.23.49 with InnoDB
disabled but BEGIN & COMMIT statements do not throw an error. If there's
a real case for this logic to continue, we could instead push this
behavior into a configuration setting.
The exception swallowing has been there since the beginning:
db045dbbf60b53dbe013ef25554fd013baf88134
|
|\ \ \
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Explicitly exit with status "1" for create and drop task failures
Conflicts:
activerecord/CHANGELOG.md
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
* A non-zero exit status allows subsequent shell commands to be chained
together such as: `rake db:reset test:prepare && rspec && cap deploy`
(if you're feeling brave :)
* Any exceptions raised during the `create` and `drop` tasks are caught
in order to print a "pretty" message to the user. Unfortunately doing
so prevents rake from aborting with a non-zero exit status to the shell.
* Therefore we re-raise the exceptions after the "pretty" message and
re-catch them in the task.
* From the task we explicitly exit with a non-zero status. This method
was chosen (rather than just letting rake fail from the exception) so
that the backtrace is suppressed and the output to stderr is
unchanged.
* Update activerecord CHANGELOG
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | | |
| | | |
| | | | |
association empty?/any? predicates any more (there is still a problem when select is applied to a relation, or if you try association#exists? -- but its easier to work around)
|
|\ \ \ \
| | | | |
| | | | | |
Addendum to #12891
|
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* Fix incorrectly named tests
* Restore Object#to_param behavior
* Ensure param is derived from a squished and truncated string
|
|\ \ \ \ \
| |/ / / /
|/| | | | |
Also dup `column_types` in AR::Result `initialize_copy`
|
| | | | | |
|
| | | | |
| | | | |
| | | | |
| | | | | |
attribute or method.
|
|\ \ \ \ \
| | | | | |
| | | | | | |
Move `SchemaCreation` to its own file instead of `AbstractAdapter`.
|
| |/ / / / |
|
|\ \ \ \ \
| | | | | |
| | | | | | |
Refer to Rails.env from AR only when Rails is defined
|
| | | | | | |
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This has returned true since 3cc9b5f1, and is not used internally.
|
| | | | | |
| | | | | |
| | | | | | |
Drop some comments that document the implementation rather than the interface.
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
when destroying a record on a has_many :through association.
:destroy method has own counter_cache callbacks.
|
| | | | | | |
|
|\ \ \ \ \ \
| | | | | | |
| | | | | | | |
Fixes problem with replacing has_one association record with itself
|
| |/ / / / / |
|
|/ / / / / |
|
| | | | | |
|
|\ \ \ \ \
| | | | | |
| | | | | | |
clarified missleading message
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
(No database connection) sounds not be
able to connect database.
see more #12804
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This is a follow up to: 97f0d9a0dd12e7ad634815eecfeff866f64aad92
This change is also related to: df0edafac9eb47cd971970a9e7b13a3eaddf214e
|
| | | | | | |
|
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
This will make the conditional and to code clear
|
| | | | | | |
|
|/ / / / / |
|
|/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
The log output used to be confusing in situation where type casting has
"unexpected" effects. For example when finding records with a `String`.
BEFORE:
irb(main):002:0> Event.find("im-no-integer")
D, [2013-11-09T11:10:28.998857 #1706] DEBUG -- : Event Load (4.5ms) SELECT "events".* FROM "events" WHERE "events"."id" = $1 LIMIT 1 [["id", "im-no-integer"]]
AFTER:
irb(main):002:0> Event.find("im-no-integer")
D, [2013-11-09T11:10:28.998857 #1706] DEBUG -- : Event Load (4.5ms) SELECT "events".* FROM "events" WHERE "events"."id" = $1 LIMIT 1 [["id", 0]]
|
|\ \ \ \
| | | | |
| | | | | |
add autoload :TransactionState to fix Marshal.load issues
|
| | |_|/
| |/| | |
|
|\ \ \ \
| |/ / /
|/| | | |
|
| | | |
| | | |
| | | |
| | | | |
refs #11561 #12700
|
| | | |
| | | |
| | | |
| | | |
| | | | |
The use of `store` on Hstore fields (for instance) is not needed
because serialization aimed to interact with Ruby objects and not
with provided field types
|
| | | |
| | | |
| | | |
| | | | |
See #8328
|
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This saved about 46 array allocations per request on an extremely simple
application. The delegation happened in the notification subsystem
which is a hotspot, so this should result in even more savings with
larger apps.
Squashed commit of the following:
commit 41eef0d1479526f7de25fd4391d98e61c126d9f5
Author: Aaron Patterson <aaron.patterson@gmail.com>
Date: Wed Nov 6 16:32:31 2013 -0800
speed up notifications
commit 586b4a18656f66fb2c518fb8e8fee66a016e8ae6
Author: Aaron Patterson <aaron.patterson@gmail.com>
Date: Wed Nov 6 16:31:05 2013 -0800
speed up runtime registry methods
commit b67d074cb4314df9a88438f785868cef77e583d7
Author: Aaron Patterson <aaron.patterson@gmail.com>
Date: Wed Nov 6 16:28:12 2013 -0800
change method name and make it public
|
| | | | |
|
| |_|/
|/| | |
|
| | |
| | |
| | |
| | |
| | | |
This allows to assign both `String` and `Symbol` values to the enum
without having to call `to_sym`, which is a security problem.
|