Fix tests on Mail 2.7.1

Currently `ids_reader` doesn't respect dirty target when the target is
not loaded yet unlike `collection.size`. I believe the inconsistency is
a bug, fixes the `ids_reader` to behave consistently regardless of
whether target is loaded or not.

Generate delegation methods to named scope in the definition time

When Content-Type is "application/x-gzip", request.content_type resolves to "application/gzip", because application/x-gzip is a synonym of application/gzip by default. This causes the acceptable_content? check in ActiveStorage::DiskController to fail, because the direct upload token contains application/x-gzip, which is not equal to application/gzip.

Fix by comparing the token content type with the request content type *and its synonyms*.

Remove bundler warnings from bin/setup output

Since #31575, `set_inverse_instance` replaces the foreign key by the
current owner immediately to make it happen when a record is added to
collection association.

But `set_inverse_instance` is not only called when a record is added,
but also when a record is loaded from queries. And also, that loaded
records are not always associated records for some reason (using `or`,
`unscope`, `rewhere`, etc).

It is hard to distinguish whether or not we should invoke
`set_inverse_instance`, but at least we should avoid the undesired
side-effect which was brought from #31575.

Fixes #34108.

getting started page rails version update to 5.2.1

* Fix reading comment only encrypted files

When a encrypted file contains only comments then reading that files raises an error:

    NoMethodError: undefined method `deep_symbolize_keys' for false:FalseClass
        activesupport/lib/active_support/encrypted_configuration.rb:33:in `config'
        test/encrypted_configuration_test.rb:52:in `block in <class:EncryptedConfigurationTest>'

This happens because the previous implementation returned a `{}` fallback for blank YAML strings. But it did not handle YAML strings that are present but still do not contain any _usefull_ YAML - like the file created by `Rails::Generators::EncryptedFileGenerator` which looks like this:

    # aws:
    #   access_key_id: 123
    #   secret_access_key: 345

* Fix coding style violation

* Add backwardscompatible with Psych versions that were shipped with Ruby <2.5

* Do not rely on railties for Active Support test

* Simplify error handling

* Improve test naming

* Simplify file creation in test

Fix optionally scoped root route unscoped access

This reverts commit 130614a8, reversing
changes made to 1a3f963f.

This is a breaking change even for stores that supports the new schema.
Since they don't implement `supports_cache_versioning?` they will now
start to fail where they would just work before.

Backport of #33932

- Moving the `supports_cache_versioning?` check to a class method. 
- Shorten the method doc. 
- Expand on the error message.

If you are using the "in cache versioning" also known as "recyclable cache keys" the cache store must be aware of this scheme, otherwise you will generate cache entries that never invalidate.

This PR adds a check to the initialization process to ensure that if recyclable cache keys are being used via

```
config.active_record.cache_versioning = true
```

Then the cache store needs to show that it supports this versioning scheme. Cache stores can let Rails know that they support this scheme by adding a method `supports_in_cache_versioning?` and returning true.

Make sure `assert_enqueued_with()` & `assert_performed_with()` work reliably with hash arguments

Contributing to Ruby on Rails Guide: standardize git remote names [ci skip].

Make a deep copy of the _default_attributes in column_defaults

When you pass an empty array to find we know we shoudl return an empty
array but it is surprising that we are returning the original empty
array instead of a new one.

Fix deprecation warning of `ActiveRecord::Migrator.migrations_path=`

Fix av-ujs build

PostgreSQL: prepare for pg-1.1

Raise `ActiveRecord::InvalidForeignKey` in `before_destroy` for a blob if attachments exist

Fallback to unprepared statement only when bind params limit is exceeded

Backport documentation fixes

Eager loading/preloading should be worked regardless of large number of records

Fix query cache to load before first request

Disable content security policy for mailer previews

Fix zero-byte files upload in #33630

Fix non_numeric_string?

Make sure there are no duplicated nested records with create_with

`spec` is the same variable name as gemspec generated by bundler, and its
intention is easier to understand than a one-letter variable.
https://github.com/bundler/bundler/blob/00fd58eaa69015092ee272c4cb5aa92a5e7ee45c/lib/bundler/templates/newgem/newgem.gemspec.tt#L11

This is follow up 685cf0d5. `spec` is already in use there.

This will avoid gems that are made to be private to be pushed to public
repositories.

update I18n fallbacks configuration to be compatible with i18n 1.1.0

Fix regression setting children record in parent before_save callback.

This restores an ability that `update` with ids on a relation which is
described at https://github.com/rails/rails/issues/33470#issuecomment-411203013.

I personally think that the `update` with two arguments on a relation is
not a designed feature, since that is totally not using a relation
state, and also is not documented.

But removing any feature should not be suddenly happened in a stable
version even if that is not documented.

In the system test template, enter a value based on label.
However, since `label` method does not use `titleize` by default.
If generate a value including underscore, cannot find a label and the test
will fail.

```
$ ./bin/rails g scaffold user name:string phone_number:string
$ ./bin/rails t test/system/users_test.rb

E

Error:
UsersTest#test_creating_a_User:
Capybara::ElementNotFound: Unable to find field "Phone Number"
    test/system/users_test.rb:18:in `block in <class:UsersTest>'
```

This removes unnecessary `titleize` so that the generated file will pass
even if the attribute contains an underscore.

Prevent leaking of user's DB credentials on `rails db:create` failure

Mutation tracker should be cleared before continuing around callbacks