Follow up of #35838.

And also this refactors `in_clause_length` handling is entirely
integrated in Arel visitor.

* Avoid duplicated `@new_record` assignment
* Extract `define_attribute_methods` into `init_internals`

Follow up to b1458218.

Related 0ee96d13.

Those helper methods makes relation values access 15% slower.

https://gist.github.com/kamipo/e64439f7a206e1c5b5c69d92d982828e

Before (02b5b8cb):

```
Warming up --------------------------------------
        #limit_value   237.074k i/100ms
    #limit_value = 1   222.052k i/100ms
Calculating -------------------------------------
        #limit_value      6.477M (± 2.9%) i/s -     32.479M in   5.019475s
    #limit_value = 1      5.297M (± 4.3%) i/s -     26.424M in   4.999933s
```

After (this change):

```
Warming up --------------------------------------
        #limit_value   261.109k i/100ms
    #limit_value = 1   239.646k i/100ms
Calculating -------------------------------------
        #limit_value      7.412M (± 1.6%) i/s -     37.077M in   5.003345s
    #limit_value = 1      6.134M (± 1.0%) i/s -     30.675M in   5.000908s
```

I've realized that `user.id` is 20% slower than `user.name` in the
benchmark (https://github.com/rails/rails/pull/35987#issuecomment-483882480).

The reason that performance difference is that `self.class.primary_key`
method call is a bit slow.

Avoiding that method call will make almost attribute access faster and
`user.id` will be completely the same performance with `user.name`.

Before (02b5b8cb):

```
Warming up --------------------------------------
             user.id   140.535k i/100ms
          user['id']    96.549k i/100ms
           user.name   158.110k i/100ms
        user['name']    94.507k i/100ms
       user.changed?    19.003k i/100ms
 user.saved_changes?    25.404k i/100ms
Calculating -------------------------------------
             user.id      2.231M (± 0.9%) i/s -     11.243M in   5.040066s
          user['id']      1.310M (± 1.3%) i/s -      6.565M in   5.012607s
           user.name      2.683M (± 1.2%) i/s -     13.439M in   5.009392s
        user['name']      1.322M (± 0.9%) i/s -      6.615M in   5.003239s
       user.changed?    201.999k (±10.9%) i/s -      1.007M in   5.091195s
 user.saved_changes?    258.214k (±17.1%) i/s -      1.245M in   5.007421s
```

After (this change):

```
Warming up --------------------------------------
             user.id   158.364k i/100ms
          user['id']   106.412k i/100ms
           user.name   158.644k i/100ms
        user['name']   107.518k i/100ms
       user.changed?    19.082k i/100ms
 user.saved_changes?    24.886k i/100ms
Calculating -------------------------------------
             user.id      2.768M (± 1.1%) i/s -     13.936M in   5.034957s
          user['id']      1.507M (± 2.1%) i/s -      7.555M in   5.017211s
           user.name      2.727M (± 1.5%) i/s -     13.643M in   5.004766s
        user['name']      1.521M (± 1.3%) i/s -      7.634M in   5.018321s
       user.changed?    200.865k (±11.1%) i/s -    992.264k in   5.044868s
 user.saved_changes?    269.652k (±10.5%) i/s -      1.344M in   5.077972s
```

It was never used from the beginning.

Method call in Ruby is a bit slow.

This makes attribute access 10% faster by avoiding method call
(`sync_with_transaction_state`).

Before (96cf7e0e):

```
Warming up --------------------------------------
             user.id   131.291k i/100ms
          user['id']    91.786k i/100ms
           user.name   151.605k i/100ms
        user['name']    92.664k i/100ms
       user.changed?    17.772k i/100ms
 user.saved_changes?    23.909k i/100ms
Calculating -------------------------------------
             user.id      1.988M (± 7.0%) i/s -      9.978M in   5.051474s
          user['id']      1.155M (± 5.8%) i/s -      5.783M in   5.022672s
           user.name      2.450M (± 4.3%) i/s -     12.280M in   5.021234s
        user['name']      1.263M (± 2.1%) i/s -      6.394M in   5.066638s
       user.changed?    175.070k (±13.3%) i/s -    853.056k in   5.011555s
 user.saved_changes?    259.114k (±11.8%) i/s -      1.267M in   5.001260s
```

After (this change):

```
Warming up --------------------------------------
             user.id   137.625k i/100ms
          user['id']    96.054k i/100ms
           user.name   156.379k i/100ms
        user['name']    94.795k i/100ms
       user.changed?    18.172k i/100ms
 user.saved_changes?    24.337k i/100ms
Calculating -------------------------------------
             user.id      2.201M (± 0.5%) i/s -     11.010M in   5.002955s
          user['id']      1.320M (± 1.0%) i/s -      6.628M in   5.021293s
           user.name      2.677M (± 1.6%) i/s -     13.449M in   5.024399s
        user['name']      1.314M (± 1.8%) i/s -      6.636M in   5.051444s
       user.changed?    190.588k (±11.1%) i/s -    944.944k in   5.065848s
 user.saved_changes?    262.782k (±12.1%) i/s -      1.290M in   5.028080s
```

This was accidentally left in, the standard `db:migrate:up` doesn't have
a description so `db:migrate:up:namespace` shouldn't have one either.

This change adds the ability to run up/down for a database in a multi-db
environment.

If you have an app with a primary and animals database the following
tasks will be generated:

```
VERSION=123 rake db:migrate:up:primary
VERSION=123 rake db:migrate:up:primary

VERSION=123 rake db:migrate:down:primary
VERSION=123 rake db:migrate:up:animals
```

I didn't generate descriptions with them since we don't generate a
description for a single database application.

In addition to this change I've made it so if your application has
multiple databases Rails will raise if you try to run `up` or `down`
without a namespace. This is because we don't know which DB you want to
run `up` or `down` against unless the app tells us, so it's safer to
just block it and recommend using namespaced versions of up/down
respectively.

The output for the raise looks like:

```
You're using a multiple database application. To use `db:migrate:down`
you must run the namespaced task with a VERSION. Available tasks are
db:migrate:down:primary and db:migrate:down:animals.
```

This reverts commit f656bb30.

Reason: Test in Action View expects the `collection_cache_key` working...

https://github.com/rails/rails/blob/ff6b713f5e729859995f204093ad3f8e08f39ea8/actionview/test/activerecord/relation_cache_test.rb#L21
https://github.com/rails/rails/blob/ff6b713f5e729859995f204093ad3f8e08f39ea8/actionview/test/fixtures/project.rb#L6

https://buildkite.com/rails/rails/builds/60609#d19181fb-fe80-4d1e-891c-1109b540fb4b/981-1009

The `collection_cache_key` is private API for a long time, but I've
maintained it in #35848 since it is mentioned in the doc
(https://github.com/rails/rails/pull/35848#discussion_r272011475).

The doc has removed at 1da9a7e4, so there is no longer a reason to
maintain that private API.

This follows #35595.

This partly reverts 8ca3c286.

`.is_a?(GeneratedAttributeMethodsBuilder)` doesn't make sense to me, the
module class's `new` creates just an instance of the module class itself.

Actionable errors let's you dispatch actions from Rails' error pages. This
can help you save time if you have a clear action for the resolution of
common development errors.

The de-facto example are pending migrations. Every time pending migrations
are found, a middleware raises an error. With actionable errors, you can
run the migrations right from the error page. Other examples include Rails
plugins that need to run a rake task to setup themselves. They can now
raise actionable errors to run the setup straight from the error pages.

Here is how to define an actionable error:

```ruby
class PendingMigrationError < MigrationError #:nodoc:
  include ActiveSupport::ActionableError

  action "Run pending migrations" do
    ActiveRecord::Tasks::DatabaseTasks.migrate
  end
end
```

To make an error actionable, include the `ActiveSupport::ActionableError`
module and invoke the `action` class macro to define the action. An action
needs a name and a procedure to execute. The name is shown as the name of a
button on the error pages. Once clicked, it will invoke the given
procedure.

`where.not` with polymorphic association is partly fixed incidentally at
213796fb (refer #33493, #26207, #17010, #16983, #14161), and I've added
test case e9ba12f7 to avoid lose that fix accidentally in the future.

In Rails 5.2, `where.not(polymorphic: object)` works as expected as
NAND, but `where.not(polymorphic_type: object.class.polymorphic_name,
polymorphic_id: object.id)` still unexpectedly works as NOR.

To will make `where.not` working desiredly as NAND in Rails 6.1, this
deprecates `where.not` working as NOR. If people want to continue NOR
conditions, we'd encourage to them to `where.not` each conditions
manually.

```ruby
all = [treasures(:diamond), treasures(:sapphire), cars(:honda), treasures(:sapphire)]
assert_equal all, PriceEstimate.all.map(&:estimate_of)
```

In Rails 6.0:

```ruby
sapphire = treasures(:sapphire)

nor = all.reject { |e|
  e.estimate_of_type == sapphire.class.polymorphic_name
}.reject { |e|
  e.estimate_of_id == sapphire.id
}
assert_equal [cars(:honda)], nor

without_sapphire = PriceEstimate.where.not(
  estimate_of_type: sapphire.class.polymorphic_name, estimate_of_id: sapphire.id
)
assert_equal nor, without_sapphire.map(&:estimate_of)
```

In Rails 6.1:

```ruby
sapphire = treasures(:sapphire)

nand = all - [sapphire]
assert_equal [treasures(:diamond), cars(:honda)], nand

without_sapphire = PriceEstimate.where.not(
  estimate_of_type: sapphire.class.polymorphic_name, estimate_of_id: sapphire.id
)
assert_equal nand, without_sapphire.map(&:estimate_of)
```

Resolves #31209.

When I added the rake tasks for multiple databases I accidentally made
it so you couldn't use Active Record without Railties.

Since getting the database configuration depends on being able to load
Railties (there is no other way to read and parse the yaml file and load
it) I've decided that using multiple databases outside of Rails is a
no-op.

I've changed the code here to return if Rails isn't defined. Then I
changed the other tasks to use `ActiveRecord::Tasks::DatabaseTasks.env`
instead of `Rails.env`. The multi-db tasks can keep using `Rails.env`
because they'll only be generated if we're using Rails and not just
Active Record.

Cache versioning enables the same cache key to be reused when the object
being cached changes by moving the volatile part of the cache key out of
the cache key and into a version that is embedded in the cache entry.

This is already occurring when the object being cached is an
`ActiveRecord::Base`, but when caching an `ActiveRecord::Relation`
we are currently still putting the volatile information (max updated at
and count) as part of the cache key.

This PR moves the volatile part of the relations `cache_key` into the
`cache_version` to support recycling cache keys for
`ActiveRecord::Relation`s.

Currently the rollback only restores primary key value, `new_record?`,
`destroyed?`, and `frozen?`. Since the `save` clears current dirty
attribute states, retrying save after rollback will causes no change
saved if partial writes is enabled (by default).

This makes `remember_transaction_record_state` remembers original values
then restores dirty attribute states after rollback.

Fixes #15018.
Fixes #30167.
Fixes #33868.
Fixes #33443.
Closes #33444.
Closes #34504.

Since 8180c396, remaining transaction state is cleared in
`force_clear_transaction_record_state` to less work
`sync_with_transaction_state`. But it caused a race condition that
`@transaction_state` would be cleared by other threads if the state is
finalized.

To work as before, snapshot `@transaction_state` to local variable not
to refer `@transaction_state` directly.

Fixes #35983.

This way, we only need to filter the backtrace up to the first non-noise
stack frame.

This also updates noise to be able to deal with being passed a lazy
enum. We don't need this anywhere, but it seemed better for this to be
consistent.

`belongs_to` association have `required: true` by default
https://github.com/rails/rails/pull/18937 onwards so we don't need it
in the generator template.
We still need the code for required in the command line generator as
it adds `null: false` in the migration.

Before this fix, `touch` only clears dirty tracking for touched
attributes, doesn't track saved (touched) changes.

This fixes that tracks saved changes and carry over remaining changes.

Fixes #33429.
Closes #34306.

We can revert migrations using `change_column_comment` or
`change_table_comment` at current master.
However, results are not what we expect: comments are remained in new
status.
This change tells previous comment to these methods in a way like
`change_column_default`.

All transaction states (:committed, :fully_committed, :rolledback,
:fully_rolledback) are mutually exclusive.
And also, `force_clear_transaction_record_state` should clear
`@transaction_state` as well.

`add_to_transaction` was added at da840d13, but it should not be called
by except internal, since `remember_transaction_record_state` should be
called only once before saving.

And also, currently `add_to_transaction` doesn't always add the record
to transaction since da8de911, that is the reason hard to use that even
in internal.

Even if `add_to_transaction` ensure to add the record to transaction,
that is an internal concern, people don't need to explicitly call
`add_to_transaction`.

Running this migration on mysql at current master fails
because `add_references_for_alter` is missing.

```
change_table :users, bulk: true do |t|
  t.references :article
end
```

This is also true for postgresql adapter,
but its `bulk_alter_table` implementation can fallback in such case.
postgresql's implementation is desirable to prevent unknown failure like this.

Most of the time, these methods are called from actual methods defined
from columns in the schema, not from method_missing, so the current
wording is misleading.

  * Remove AbstractMysqlAdapter::Version since full_version_string will always
    be set.
  * Remove nodoc on private methods because private methods are not exposed in
    the docs.

* remove useless `@type_metadata` and `@array`
* move the compatibility code (for array) into column
* etc.