The code incorrectly assumes that the option was written as
`foreign_key: true`, but that is not always the case. This now mirrors
the behavior of reverting `add_foreign_key`. The code was changed to use
kwargs while I was touching it, as well.

This could really use a refactoring to go through the same code paths as
`add_refernce` in the future, so we don't duplicate default values.

Fixes #25169

Action Cable: Tune Sauce Labs tests

Do not include default column limit in schema.rb

Prior to this change, we would get collisions if Active Record objects
of different classes with the same ID were used as keys of the same
hash. It bothers me slightly that we have to allocate inside of this
method, but Ruby doesn't provide any way to hash multiple values without
allocation

- Remove dead classes / dead code
- Move class definitions to where they are used, don't define in a
  shared space

Cuts down the number of concurrent Sauce Labs VMs we need to spin up. Can add specific versions back in to target regressions if need be.

Fix Typo and Grammar [ci skip]

Follow up of #20815.

```ruby
class CreatePeople < ActiveRecord::Migration[5.0]
  def change
    create_table :people do |t|
      t.integer :int
      t.bigint :bint
      t.text :txt
      t.binary :bin
    end
  end
end
```

Result.

In postgresql and sqlite3 adapters:

```ruby
ActiveRecord::Schema.define(version: 20160531141018) do

  create_table "people", force: :cascade do |t|
    t.integer "int"
    t.bigint  "bint"
    t.text    "txt"
    t.binary  "bin"
  end

end
```

In mysql2 adapter:

```ruby
ActiveRecord::Schema.define(version: 20160531141018) do

  create_table "people", force: :cascade, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8mb4" do |t|
    t.integer "int"
    t.bigint  "bint"
    t.text    "txt",  limit: 65535
    t.binary  "bin",  limit: 65535
  end

end
```

After this patch:

```ruby
ActiveRecord::Schema.define(version: 20160531141018) do

  create_table "people", force: :cascade, options: "ENGINE=InnoDB DEFAULT CHARSET=utf8mb4" do |t|
    t.integer "int"
    t.bigint  "bint"
    t.text    "txt"
    t.binary  "bin"
  end

end
```

This behavior was broken by 36e9be85. When the value is assigned
directly, either through mass assignment or directly assigning a hash,
the hash gets passed through to this writer method directly. While this
is intended to handle certain cases, when an explicit converter has been
provided, we should continue to use that instead. The positioning of the
added guard caused the new behavior to override that case.

Fixes #25210

Move `Workshop` class definition

We should define it only where we need it, not in the global abstract
unit 😬

Reuse a result of `table.associated_table(column)` in `AssociationQueryHandler.value_for`

Currently `exists?` does some hackery where it assumes that we can join
onto anything that we passed to `eager_load` or `includes`, which
doesn't work if we are joining onto a polymorphic association.

Actually figuring out if we want to include something would require
knowledge deep within the join dependency module, which is hard to pull
up. The simplest solution is just to pass a flag down that says we're
not actually going to try to eager load any of the data. It's not the
solution I'd like, but that code really needs to be untangled before we
can do much with it.

This is another attempt at 6d5b1fdf which should address the concerns
that led to reverting it in 4ecabed2.

use OpenSSL::Cipher instead of deprecated OpenSSL::Cipher::Cipher

We now use default procs inside of the errors object, which gets
included by default when marshaling anything that includes
`ActiveModel::Validations`. This means that Active Record objects cannot
be marshalled. We strip and apply the default proc ourselves. This will
ensure the objects are YAML serializable as well, since YAML falls back
to marshal implementations now. This is less important, however, as the
errors aren't included when dumping Active Record objects.

This commit does not include a changelog entry, as 5.0 is still in RC
status at the time of writing, and 5.0.0 will not release with the bug
this fixes.

Fixes #25165

Fix migration class names in tests

Add a missing capital letter and avoid using absolute links to the
API because they may refer to out-dated documentation on the Edge
site.

Based on https://github.com/rails/rails/pull/25192#discussion_r65018222 and http://ruby-doc.org/stdlib-1.9.3/libdoc/openssl/rdoc/OpenSSL/Cipher/Cipher.html

Requesting documentation update

Change some establish_connection logic

Move new default initializers into separate directory

[ci skip]Fix wrong require path raising LoadError

ActionController::Parameters: Fix #dig doc code

This example code wasn't getting wrapped in a `<code>` tag due to incorrect indentation.

rename enum_prefix and enum_suffix options to _prefix and _suffix [ci skip]

`:enum_prefix` and `:enum_suffix` was changed to `:_prefix` and `:_suffix`
in #20999.

- These initializers are new defaults for Rails 5 and newly generated
  apps do not need to change them.
- Fixes #25181

Follow up to e45f388e.

When changing the generated integration tests to assert redirects with
an absolute path this redirection check in `assert_redirected_to` against
the absolute response location would now pass:
https://github.com/rails/rails/blob/af245aaf3a1c447752a1b5895adddc66e5f73c00/actionpack/lib/action_dispatch/testing/assertions/response.rb#L55

Thus we'd break early instead of hitting `assert_operator` and have
2 fewer assertions per `assert_redirected_to` as `assert_operator`
is composed of 2 assertions internally:
https://github.com/seattlerb/minitest/blob/4e146b1515b19a5c474e39c174b037510f6dbc6e/lib/minitest/assertions.rb#L254-L258

Deduct 2 assertions for 2 redirects taking the expected count down by 4
in total for the two failing tests.

Remove public/ files for API apps

The first comment does not explain anything that the code doesnt and the second is now irrelevant with the dedicated API mode

- Fixes #25180

fix incorrect class name [ci skip]