* Since #18411, we started to inform about extracted gem (record_tag_helper)
  to developers who use `ActionView::Helpers::RecordTagHelper` 's methods.

* Currently, it seems no problem that we don't have to support no longer.

All of Date, DateTime and Time respond to `iso8601`.

Add ujs desc to rakefile in actionview

* Seems the desc of Rake::TestTask.new is not displayed
* Use comment instead of desc

Add custom RuboCop for `assert_not` over `refute`

Stop mutating body response

Add #dig to ActionDispatch::Request::Session

2.6 warnings: passing splat keyword arguments as a single Hash

[ci skip] Add :private option to delegation doc

Optimize the code inside AR::QueryCache middleware

Remove superfluous `ActionController::`

Add missing dots at the end of comments in environment file templates

[ci skip] Add :private option to delegate section in guide

Ruby 2.6.0 warns about this.

### Summary

The `session` object is not a real Hash but responds to many methods of Hash
such as `[]`, `[]`, `fetch`, `has_key?`.

Since Ruby 2.3, Hash also supports a `dig` method.

This commit adds a `dig` method to `ActionDispatch::Request::Session` with the
same behavior as `Hash#dig`.

This is useful if you store a hash in your session, such as:

```ruby
session[:user] = { id: 1, avatar_url: "http://example.org/nyancat.jpg" }
```

Then you can shorten your code from `session[:user][:avatar_url]` to `session.dig :user, :avatar_url`.

### Other Information

I cherry-picked a commit from https://github.com/rails/rails/pull/23864, and modify a bit.
The changes are below:

* Converts only the first key to a string adjust to the `fetch` method.
* Fixes a test case because we cannot use the indifferent access since ee5b621e.

If @app.call returns an object that is saved (for e.g., in a constant), the mutation results in a continuing cycle of wrapping the body in Rack::BodyProxy, eventually leading to SystemStackError
------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
On branch fix-return-response-mutation-rack-logger - Tue  3 Apr 2018 19:54:28 PDT by Geoff Lee <geoff.lee@lendesk.com>

73e7aab behaved as expected on codeship, failing the build with
exactly these RuboCop violations. Hopefully `rubocop -a` will
have been enough to get a passing build!

Since at least cf4afc4d we have preferred `assert_not` methods over
`refute` methods. I have seen plenty of comments in PRs about this,
and we have tried to fix it a few times (5294ad82, e45f176b, 8910f12f,
41f50beb, d4cfd543, 48a183ec, and 211adb47), but the `refute` methods
keep sneaking back in.

This custom RuboCop will take care of enforcing this preference, so we
don't have to think about it again. I suspect there are other similar
stylistic preferences that could be solved with some custom RuboCops, so
I will definitely keep my eyes open. `assert_not` over `assert !` might
be a good candidate, for example.

I wasn't totally sure if `ci/custom_cops` was the best place to put
this, but nothing else seemed quite right. At one point I had it set up
as a gem, but I think custom cops like this would have limited value
in another context.

I want to see how code climate handles the new cops before
autocorrecting the existing violations. If things go as expected, I will
push another commit with those corrections.

It is wrongly appeared as instance public methods in the doc.

http://api.rubyonrails.org/v5.1.6/classes/ActionCable/Channel/Callbacks.html
http://api.rubyonrails.org/v5.1.6/classes/ActiveRecord/Timestamp.html

Fix RenderingTest in railtie

Use `:default` option in order to set default value of `finalize_compiled_template_methods`

Test class name is not `RoutingTest` but `RenderingTest`

Note that tab_yellow.gif exists, that is the one being used.

The credits page is gone after #32429, these images are now orphan.

Add dots in order to keep consistency between other comments in these files.

Since we introduced default option for `class_attribute` and
`mattr_accessor` family of methods and changed all occurrences of setting
default values by using of `:default` option I think it would be fine to use
`:default` option in order to set default value of `finalize_compiled_template_methods`
since it expresses itself very well.

Related to #29294, #32418

Cleanup credits related code as it is removed now in PR #32429

Remove credits.html from Rails Guides

Display db:drop description in ActiveRecord

Small doc fixes

[ci skip]

Doc fix added missing quote

[ci skip]

* refs #32420