Previosly, `update_columns` would just take whatever keys you gave it
and tried to run the update query. Most likely this would result in an
error from the database. However, if the column actually did exist, but
was in `ignored_columns`, this would result in the method returning
successfully when it should have raised, and an attribute that should
not exist written to `@attributes`.
rails/rails#33853 and rails/rails#33929 removed `encoding: utf8mb4` from database.yml
since at that time MySQL 5.1 is supported with the master branch.
Since MySQL 5.1 has been dropped, we can restore `encoding: utf8mb4` in database.yml
When loading fixtures, Ruby 1.9's hash key ordering means that HABTM
join table rows are always loaded before the parent table rows,
violating foreign key constraints that may be in place. This very
simple change ensures that the parent table's key appears first in the
hash. Violations may still occur if fixtures are loaded in the wrong
order but those instances can be avoided unlike this one.
Given that Bundler 1.10 was released back in 2015, we can assume
that people are at least on this version or a higher one so there
is no need to ask people to upgrade.
Also, given that Rails 5.0 supports Ruby 2.2+ and given that this
version come with Minitest 5.4.3 bundled, we don't need to ensure
backward compatibility with Minitest 4.
Both `UTC_OFFSET_WITH_COLON` and `UTC_OFFSET_WITHOUT_COLON` are only
used within `ActiveSupport::TimeZone` and in my opinion they do not
provide relevant information that should appear in the docs.
If you require `nokogiri` from `app/models/user.rb`, dependencies.rb
does not mark `Nokogiri` as an autoloaded constant, as expected.
But the logic to detect these non-autoloaded constants is incomplete.
See the tests defined in the patch for some cases incorrectly handled.
We repeat the phrase "Release Notes" even though release note are listed
under a... "Release Notes" category already. With this change, instead
of repeating ourselves, we can give people a sense of the recency of
each release by simply listing the month and year of each release.
This is a follow-up to #34265 as suggested by @jeremy.
[ci skip]
At fc0e3354af,
```rb
relation = relation.where(conditions)
```
was rewritten to:
```rb
relation.where!(condition)
```
This change accidentally changed the result of `Topic.exists?({})` from true to false.
To fix this regression, first I moved the blank check logic (`opts.blank?`) from `#where` to `#where!`,
because I thought `#where!` should be identical to `#where`, except that instead of returning a new relation,
it adds the condition to the existing relation.
But on second thought after some discussion on https://github.com/rails/rails/pull/34329,
I started to think that just fixing `#construct_relation_for_exists` is more preferable
than changing `#where` and `#where!`.
The PR#31819 changed `#preloaders_on` and added some test,
then #33938 reverted changes that were added to the method in #31819.
Since changes in the method were reverted and as mentioned in the
comment https://github.com/rails/rails/pull/31819#discussion_r221847481
that titles of the tests added in #31819 don't reflect implementation I
think we can remove those test for now.
The usage of maxlength in the text_field helper adds a size attribute
to the generated text_field input with the same value as the maxlength.
This implicit addition of size attribute by the method gives a false
impression that it may be bug. By adding the implementation of the
maxlength to the api docs, we explicitly tell the reader referring the
api doc that addition of size along with maxlength is the expected behaviour.
[ci skip]
There are two main reasons why `assert_called_with` should require
`args` argument:
1) If we want to assert that some method should be called and we don't
need to check with which arguments it should be called then we should use
`assert_called`.
2) `assert_called_with` without `args` argument doesn't assert anything!
```ruby
assert_called_with(@object, :increment) do
@object.decrement
end
```
It causes false assertions in tests that could cause regressions in the project.
I found this bug by working on
[minitest-mock_expectations](https://github.com/bogdanvlviv/minitest-mock_expectations) gem.
This gem is an extension for minitest that provides almost the same method call
assertions.
I was wondering whether you would consider adding "minitest-mock_expectations"
to `rails/rails` instead of private `ActiveSupport::Testing::MethodCallAssertions` module.
If yes, I'll send a patch - a970ecc42c
ActionController::Metal provides session support by delegating `session to the request (`"@_request"`)
a3dcba42e2/actionpack/lib/action_controller/metal.rb (L149)
Though the ActionController::Cookies modules isn't included, it's really a convenience for providing a first class `cookies` object.
*all* ActionController::Metal subclasses support setting cookies via the `session` object.
Bulk lookups are sort of an edge case, I have not heard of them until a colleague of mine decided to dynamically iterate over a growing set of translations and receiving them in bulk as a hash with `I18n.t 'welcome'` as in the example above.
When passing an interpolation to these bulk lookups, they will only be performed when also passing `deep_interpolation: true`.
**Without passing `deep_interpolation` flag:**
```ruby
I18n.t 'welcome', app_name: 'book store'
# => {:title=>"Welcome!", :content=>"Welcome to the %{app_name}"}
**With passing `deep_interpolation`:**
I18n.t 'welcome', deep_interpolation: true, app_name: 'book store'
# => {:title=>"Welcome!", :content=>"Welcome to the book store"}
```
I found this digging in the I18n lookup backend, the flag is listed on [I18n's Rubydoc](https://www.rubydoc.info/github/svenfuchs/i18n/master/I18n) but not otherwise listed. Since bulk lookups are mentioned here, I suggest to add a note with this flag along with it.
The default collation for utf8mb4 is changed to utf8mb4_0900_ai_ci in
MySQL 8.0, so `test_mysql_connection_collation_is_configured` is
affected by MySQL version and/or default_collation_for_utf8mb4 system
variable.
https://dev.mysql.com/doc/refman/8.0/en/server-system-variables.html#sysvar_default_collation_for_utf8mb4
```
% ARCONN=mysql2 be ruby -w -Itest test/cases/adapters/mysql2/connection_test.rb -n test_mysql_connection_collation_is_configured
Using mysql2
Run options: -n test_mysql_connection_collation_is_configured --seed 10421
# Running:
F
Failure:
Mysql2ConnectionTest#test_mysql_connection_collation_is_configured [test/cases/adapters/mysql2/connection_test.rb:108]:
Expected: "utf8mb4_general_ci"
Actual: "utf8mb4_0900_ai_ci"
rails test test/cases/adapters/mysql2/connection_test.rb:106
Finished in 0.013842s, 72.2450 runs/s, 144.4900 assertions/s.
1 runs, 2 assertions, 1 failures, 0 errors, 0 skips
```
Since #33844, eager loading/preloading with too many and/or too large
ids won't be broken by pre-checking whether the value is constructable
or not.
But the pre-checking caused the type to be evaluated at relation build
time instead of at the query execution time, that is breaking an
expectation for some apps.
I've made the pre-cheking lazy as much as possible, that is no longer
happend at relation build time.
Sean Griffin