`eager_load` performs a single query using a `LEFT OUTER JOIN` to load
the associations. Loading the associations in a join can result in many
rows that contain redundant data and it performs poorly at scale.
With `includes` a separate query is performed for each association,
unless a join is required by conditions.
Co-authored-by: Rafael Mendonça França <rafael@franca.dev>
* Remove pidfile in production
* Update changelog
* Update activestorage/test/dummy/config/puma.rb
Co-authored-by: Rafael Mendonça França <rafael@franca.dev>
* Update template and other dummy files
---------
Co-authored-by: Rafael Mendonça França <rafael@franca.dev>
Since `ActionText::Content` wraps an `ActionText::Fragment`, and
`ActionText::Fragment` wraps a `Nokogiri::XML::DocumentFragment`, then
`ActionText::Content` should be able to rely on the newer Ruby pattern
matching introduced by [nokogiri@1.16.0][] (mainly the
[DocumentFragment#deconstruct][] method):
```ruby
content = ActionText::Content.new <<~HTML
<h1>Hello, world</h1>
<div>The body</div>
HTML
content => [h1, div]
assert_pattern { h1 => { content: "Hello, world" } }
assert_pattern { div => { content: "The body" } }
```
The implementation change relies on delegating from `Content` to
`Fragment`, and from `Fragment` to `DocumentFragment#elements` (to
deliberately exclude text nodes).
[nokogiri@1.16.0]: https://nokogiri.org/CHANGELOG.html?h=pattern
[DocumentFragment#deconstruct]: https://nokogiri.org/rdoc/Nokogiri/XML/DocumentFragment.html?h=deconstruct#method-i-deconstruct
To assert the expected number of queries are made, Rails internally uses
`assert_queries` and `assert_no_queries`. These assertions can be
useful in applications as well.
By extracting these assertions to a module, the assertions can be
included where required.
These assertions are added to `ActiveSupport::TestCase` when
ActiveRecord is defined.
ActiveStorage, ActionView and ActionText are using this module now as
well, instead of duplicating the implementation.
The internal ActiveRecord::TestCase, used for testing ActiveRecord,
implements these assertions as well. However, these are slighlty more
advanced/complex and use the SQLCounter class. To keep things simple,
for now this implementation isn't used.
Follow-up to [#50252][]
Similar to the reliance on a `FormBuilder` in the helper methods
documentation examples, the template test coverage for `#rich_text_area`
relied on invocations through a `FormBuilder` instance.
This commit adds explicit coverage for calling the `#rich_text_area`
helper method directly with both an `object_name` and `method_name`
positional arguments.
[#50252]: https://github.com/rails/rails/pull/50252
actiontext.js is compiled as ESM bundle instead of UMD bundle.
This leads to issues when trying to use ActionText with sprockets because the ESM bundle declares variables like they are scoped to the file but sprockets will see them as scoped globally.
This is a problem, in particular, if you want to mix actiontext with
turbo-rails.
The problem got introduced in https://github.com/rails/rails/pull/46447.
I traced valid compilation back to
https://github.com/rails/rails/pull/42895.
This commit mimic changes made in
https://github.com/rails/rails/pull/42895 to ActiveStorage:
Retains app/assets/javascripts/actiontext.js as a UMD package for backwards compatibility with
bundling in the asset pipeline, but also adds app/assets/javascripts/actiontext.esm.js for use
with ESM via importmap in the browser.
Remove the option `config.public_file_server.enabled` from the generators for all environments, as the value is the same in all environments.
Co-authored-by: Jonathan Hefner <jonathan@hefner.pro>
Follow-up to #47137.
Since `config.public_file_server.enabled` is true by default, this
commit changes the `config/environments/production.rb` template to
present the setting as an opt-out.
which were being set to the HTML4 defaults before the sanitizer
configuration could be applied.
Also, backfill some light tests for sanitization.
Related to #48644
The change from `#clone` to `#dup` is necessary to work around an
issue in Nokogiri where `#clone` is not defined properly for HTML5
fragment and the fragment does not have a parent Document. `#dup`
behaves the way we expect, so this should be fine.
This clarifies that the boolean interpretation (1) is due to YAML rather
than I18n, (2) is case insensitive, and (3) affects both keys and
values.
Co-authored-by: Jonathan Hefner <jonathan@hefner.pro>
Background
----------
During integration tests, it is desirable for the application to respond
as closely as possible to the way it would in production. This improves
confidence that the application behavior acts as it should.
In Rails tests, one major mismatch between the test and production
environments is that exceptions raised during an HTTP request (e.g.
`ActiveRecord::RecordNotFound`) are re-raised within the test rather
than rescued and then converted to a 404 response.
Setting `config.action_dispatch.show_exceptions` to `true` will make the
test environment act like production, however, when an unexpected
internal server error occurs, the test will be left with a opaque 500
response rather than presenting a useful stack trace. This makes
debugging more difficult.
This leaves the developer with choosing between higher quality
integration tests or an improved debugging experience on a failure.
I propose that we can achieve both.
Solution
--------
Change the configuration option `config.action_dispatch.show_exceptions`
from a boolean to one of 3 values: `:all`, `:rescuable`, `:none`. The
values `:all` and `:none` behaves the same as the previous `true` and
`false` respectively. What was previously `true` (now `:all`) continues
to be the default for non-test environments.
The new `:rescuable` value is the new default for the test environment.
It will show exceptions in the response only for rescuable exceptions as
defined by `ActionDispatch::ExceptionWrapper.rescue_responses`. In the
event of an unexpected internal server error, the exception that caused
the error will still be raised within the test so as to provide a useful
stack trace and a good debugging experience.
`assert_queries` in Action Text's test helpers can detect queries from
a previous test if background jobs are allowed to run during its
execution.
Changing the queue_adapter to the test adapter for all tests using the
helper ensure no jobs can run during its execution.
Before this commit, using ActionText and calling `#as_json` on a
non-persisted `ActiveStorage::Blob` raised an error.
This is because `ActionText::Attachable` is included in
`ActiveStorage::Blob` and overrides the `#as_json` method to
expose a global signed id. However, a global signed id can only
be generated on a persisted instance.
This commit fixes the issue by making sure the blob is persisted
before exposing the global signed id.
* Logging to a file doesn't make sense in production
You're going to run out of space, and it doesn't play well with containers. Either you log to STDOUT, and let your container setup aggregate the logs, or you'll be switching to syslogger or whatever. You won't be logging to a file in production any more.
* Remove from Dockerfile too
* Did not mean to change this default
But we should make it easy to see how to change it.
* Restore what we had
of webpacker.
Include the js and css in the application.html.erb which now shows the
rich text editor on the page. This allows the integration tests to run.
Fixes https://github.com/rails/rails/issues/46804
This represents a +2x performance optimization when the replacement
logic is based on some condition, and it returns the same unchanged
node when it wants to skip it:
```ruby
html = <<~HTML
<div>
#{'<p>ignore me</p>' * 1000}
#{'<p>replace me</p>' * 1000}
</div>
HTML
content = content_from_html(html)
replacement_example = -> do
content.fragment.replace("p") do |node|
if node.text =~ /replace me/
"<p>replace me</p>"
else
node
end
end
end
current_implementation = -> do
class ActionText::Fragment
def replace(selector)
update do |source|
source.css(selector).each do |node|
replacement_node = yield(node)
node.replace(replacement_node.to_s)
end
end
end
end
replacement_example.call
end
new_implementation = -> do
class ActionText::Fragment
def replace(selector)
update do |source|
source.css(selector).each do |node|
replacement_node = yield(node)
node.replace(replacement_node.to_s) if node != replacement_node
end
end
end
end
replacement_example.call
end
Benchmark.ips do |x|
x.report "Current implementation", ¤t_implementation
x.report "New implementation", &new_implementation
x.compare!
end
```
Results:
```
Warming up --------------------------------------
Current implementation
2.000 i/100ms
New implementation 5.000 i/100ms
Calculating -------------------------------------
Current implementation
32.484 (±30.8%) i/s - 134.000 in 5.036419s
New implementation 74.878 (±38.7%) i/s - 250.000 in 5.052168s
Comparison:
New implementation: 74.9 i/s
Current implementation: 32.5 i/s - 2.31x (± 0.00) slower
```
When a host is not specified for an `ActionController::Renderer`'s env,
the host and related options will now be derived from the routes'
`default_url_options` and `ActionDispatch::Http::URL.secure_protocol`.
For example, with:
```ruby
Rails.application.default_url_options = { host: "rubyonrails.org" }
Rails.application.config.force_ssl = true
```
Before:
```ruby
ApplicationController.renderer.render inline: "<%= blog_url %>"
# => "http://example.org/blog"
```
After:
```ruby
ApplicationController.renderer.render inline: "<%= blog_url %>"
# => "https://rubyonrails.org/blog"
```
As a consequence, Action Text attachment URLs rendered in a background
job (a la Turbo Streams) will now use `Rails.application.default_url_options`.
Fixes#41795.
Fixeshotwired/turbo-rails#54.
Fixeshotwired/turbo-rails#155.
* Use storage/ instead of db/ for sqlite3 db files
db/ should be for configuration only, not data. This will make it easier to mount a single volume into a container for testing, development, and even sqlite3 in production.
When System Tests call `fill_in_rich_text_area`, they interact with
`<trix-editor>` elements by changing the contents programmatically.
This is unlike how end-users will interact with the element. Overhauling
the test helper to more accurately reflect Real World usage would
require a sizable effort.
With that being said, leaving the `<trix-editor>` with focus after
populating its contents is a minor change that makes it a more genuine
recreation.
Expand the `has_rich_text` signature to accept a `strict_loading:`
value. Forward that value along to the `has_one` declaration made under
the hood. When omitted, `strict_loading:` will be set to the value of
the `strict_loading_by_default` class attribute (false by default).
Commit 37d1429ab1d introduced the DummyERB to avoid loading the environment when
running `rake -T`.
The DummyCompiler simply replaced all output from `<%=` with a fixed string and
removed everything else. This worked okay when it was used for YAML values.
When using `<%=` within a YAML key, it caused an error in the YAML parser,
making it impossible to use ERB as you would expect. For example a
`database.yml` file containing the following should be possible:
development:
<% 5.times do |i| %>
shard_<%= i %>:
database: db/development_shard_<%= i %>.sqlite3
adapter: sqlite3
<% end %>
Instead of using a broken ERB compiler we can temporarily use a
`Rails.application.config` that does not raise an error when configurations are
accessed which have not been set as described in #35468.
This change removes the `DummyCompiler` and uses the standard `ERB::Compiler`.
It introduces the `DummyConfig` which delegates all known configurations to the
real `Rails::Application::Configuration` instance and returns a dummy string for
everything else. This restores the full ERB capabilities without compromising on
speed when generating the rake tasks for multiple databases.
Deprecates `config.active_record.suppress_multiple_database_warning`.
