| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| | | | | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | |
| | | | | | |
Suggested by @dhh.
It doesn't affect the generated SQL, so seems reasonable to continue to
allow it as an association option.
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
Fixes #8795
|
| |\ \ \ \ \ \
| | | | | | |
| | | | | | | |
`CollectionAssociation#empty?` respects newly builded records
|
| | | |_|_|_|/
| |/| | | | |
|
| | | | | | |
| | | | | |
| | | | | |
| | | | | | |
[ci skip]
|
| | | | | | | |
|
| | | | | | | |
|
| | |/ / / /
|/| | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This revision makes crystal clear that the methods go
straight to the database and update the receiver. It
also adds and example, and removes the duplication
in the singular and plural forms by referring one to
the other.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
There's no need to create a new arel table or reflect on the column
association if the value is empty, these attributes are not used.
Also no need to concat a new array, just append the query value.
|
| |\ \ \ \ \
| |_|_|/ /
|/| | | | |
spelling correction for Action Mailer interceptors
|
| |/ / / / |
|
| | | | | |
|
| |\ \ \ \
| |_|_|/
|/| | | |
Removing : warning: ambiguous first argument;
|
| |/ / / |
|
| | | |
| | |
| | |
| | | |
Also add some generic tests to ensure they're properly deprecated.
|
| |\ \ \
| | | |
| | | | |
Clear specific logs when using `rake log:clear`
|
| | | | | |
|
| |\ \ \ \
| | | | |
| | | | | |
document Intercepters in ActionMailer guide
|
| | | | | | |
|
| |\ \ \ \ \
| |/ / / /
|/| | | | |
Fix date_select :selected option so you can pass it nil
|
| |/ / / / |
|
| |\ \ \ \
| | | | |
| | | | | |
Don't rely on Hash key's ordering to configure default options
|
| |/ / / /
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
If we set encoding latin1 for a PostgreSQL database, it calls
PostgreSQLAdapter::create_database with options that have,
among other things:
{ 'encoding' => 'latin1' }
Then, we use reverse_merge(:encoding => "utf8") to setup the default
encoding. In the end, the hash looks like:
{ :encoding => 'utf8', 'encoding' => 'latin1' }
The call to options.symbolize_keys calls to_sym on each_key of this
Hash. It usually means that the encoding passed overwrites the default
utf8, but it's not guaranteed. So, we shouldn't rely on it.
The same was happening in ActiveRecord::ConnectionHandling.
|
| | | | | |
|
| |\ \ \ \
| |/ / /
|/| | | |
adding regression test in master for #8631
|
| |/ / / |
|
| |\ \ \
| | | |
| | | | |
Allow a `:dirs` option to `SourceAnnotationExtractor.enumerate`
|
| | | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
Allows custom rake tasks to be defined using:
SourceAnnotationExtractor.enumerate 'TODO|FIXME', dirs: %w(app lib), tag: true
|
| | | | | |
|
| |\ \ \ \
| | | | |
| | | | | |
Strong parameters exception handling
|
| | | | | | |
|
| |\ \ \ \ \
| |/ / / /
|/| | | | |
Remove header bloat introduced by BestStandardsSupport middleware
|
| |/ / / /
| | | |
| | | |
| | | | |
The same headers were being duplicated on every request.
|
| |\ \ \ \
| | | | |
| | | | | |
Update guides/source/active_record_validations.md
|
| | | | | |
| | | | |
| | | | | |
Duplicated `update`
|
| |\ \ \ \ \
| |_|/ / /
|/| | | | |
Cleaning up ActiveModel::Dirty tests
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
* Clarifying what the #changed method returns
* Adding tests to describe what the #changed_attributes returns
* Updating test name based on pull request comment
* Moving the test lower in the file per pull request comment
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This commit changes route defaults so that explicit defaults are no
longer required where the key is not part of the path. For example:
resources :posts, bucket_type: 'posts'
will be required whenever constructing the url from a hash such as a
functional test or using url_for directly. However using the explicit
form alters the behavior so it's not required:
resources :projects, defaults: { bucket_type: 'projects' }
This changes existing behavior slightly in that any routes which
only differ in their defaults will match the first route rather
than the closest match.
Closes #8814
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This now allows the use of arrays like this:
get '/foo/:action', to: 'foo', constraints: { subdomain: %w[www admin] }
or constraints where the request method returns an Fixnum like this:
get '/foo', to: 'foo#index', constraints: { port: 8080 }
Note that this only applies to constraints on the request - path
constraints still need to be specified as Regexps as the various
constraints are compiled into a single Regexp.
|
| | | | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
Now that Journey has been integrated into ActionDispatch we can raise
the exception ActionController::UrlGenerationError directly rather than
raising the internal Journey::Router::RoutingError and then have
ActionDispatch::Routing::RouteSet#generate re-raise the exception.
|
| | |/ / /
|/| | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | |
| | | | |
This reverts commit 8348f9ea72c9b50fc4d4462fd9ebe3bba932c783, reversing
changes made to 9dfe2d6f9fabddf9a451a995678a9648c8aaf401.
Reason: this broke railties tests as explained in the issue, and the
author is going to review and report back.
https://github.com/rails/rails/pull/8930#issuecomment-12272671
|
| |\ \ \ \
| | | | |
| | | | | |
rake railties:install:migrations respects the order of railties
|
| | | | | | |
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This reverts commit 3fa00070047b5d019d39e691598ee2890283d052.
Reason: This message is usually not accurate and annoying:
Redirected by ~/.rbenv/versions/1.9.3-p327-perf/lib/ruby/1.9.1/logger.rb:371:in `add'`
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
exception
We were previously rescuing "nil" when no exception class was found.
This does work in 1.9.3, but does not in 2.0, raising an exception
asking for a class or module to be given to the rescue clause.
Thanks @yahonda for catching this.
|
| | | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | |
| | | | | |
This reverts commit 1fc294f988e93ac59057a16b0e44b9cf93af9720, reversing
changes made to fb9ac47c286fbdfe05263d4d0346e50892090603.
Reason: these tests should run fine independent of database, since they
should only take into account the configuration options. The problem was
related to a change in the way "nil" is handled by the rescue clause, in
Ruby 2.0 it raises an exception asking for class or module, in 1.9 it
passes.
|
| |\ \ \ \ \
| | | | | |
| | | | | | |
ActiveRecord <-> Active Record [ci skip]
|
| |/ / / / / |
|
