Fix merging left_joins to maintain its own `join_type` context

* Fallback to spawn instead of fork in jruby

This commit: b342db68 introduced a `fork` fork when running the railties
tests since this is not supported in jruby we fallback to using spawn.
Fixes: https://github.com/rails/rails/issues/35900

* Add test asserting truncate returns unfrozen string

* Ensure strings returned from truncate are not frozen

This fixes an issue where strings too short to be truncated were
returned unfrozen, where as long-enough strings were returned
frozen. Now retuned strings will not be frozen whether or not
the string returned was shortened.

* Update changelog w/ new truncate behavior description

[Jordan Thomas + Rafael Mendonça França]

Adds documentation for has_one touch option after #35869 [ci skip]

Adds missing touch option to has_one association

Need to new line to break line in the markdown.

Update changelog for a bug-fix to read better [ci skip]

Don't fail ImageAnalyzer on unsupported types

Fix: #36065

The IamgeAnalyzer passes a image to ImageMagick without checking if the
image is supported by ImageMagick. This patch checks that image is
supported and if not logs an error and returns an empty hash instead of
raising an error. This is the same error handling we do when we
encounter a LoadError when mini_magick is not installed.

If we want to get alias resolved attribute finally, we can use
`attribute_alias` directly.

For that purpose, avoiding redundant `attribute_alias?` makes alias
attribute access 40% faster.

https://gist.github.com/kamipo/e427f080a27b46f50bc508fae3612a0e

Before (2c0729d8):

```
Warming up --------------------------------------
          user['id']   102.668k i/100ms
      user['new_id']    80.660k i/100ms
        user['name']    99.368k i/100ms
    user['new_name']    81.626k i/100ms
Calculating -------------------------------------
          user['id']      1.431M (± 4.0%) i/s -      7.187M in   5.031985s
      user['new_id']      1.042M (± 4.2%) i/s -      5.243M in   5.039858s
        user['name']      1.406M (± 5.6%) i/s -      7.055M in   5.036743s
    user['new_name']      1.074M (± 3.6%) i/s -      5.387M in   5.024152s
```

After (this change):

```
Warming up --------------------------------------
          user['id']   109.775k i/100ms
      user['new_id']   103.303k i/100ms
        user['name']   105.988k i/100ms
    user['new_name']    99.618k i/100ms
Calculating -------------------------------------
          user['id']      1.520M (± 6.7%) i/s -      7.574M in   5.011496s
      user['new_id']      1.485M (± 6.2%) i/s -      7.438M in   5.036252s
        user['name']      1.538M (± 5.4%) i/s -      7.737M in   5.049765s
    user['new_name']      1.516M (± 4.6%) i/s -      7.571M in   5.007293s
```

Fix sliced IN clauses to be grouped

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.

Remove redundant test setups in log_subscriber_test

Remove action_controller.perform_caching from api app's configs

Minor API doc fix

[ci skip]

Resurrect external JS/CS generation

As suggested in https://github.com/rails/rails/issues/35602#issuecomment-485833483, because we don't provide view caching and doesn't include `ActionController::Caching` for api apps, we should also avoid generating

```ruby
config.action_controller.perform_caching = true
```

for those api apps. So it won't confuse people.

**But because `perform_caching` will be `true` if not set, the behavior of the app would still be the same without these configs.**

Because controllers' `perform_caching` config is `true` by default, it
means we actually enable the caching in all those tests implicitly (and
it works). Which also means we can avoid repeatedly declaring that and
just specify it once in the setup method (just for declaration).

Update document about perform_caching's default value [ci skip]

Clarify that both `config.action_controller.perform_caching` and
`config.action_mailer.perform_caching`'s default are `true`. This
default is set in `AbstractController::Caching`.

One can verify this by commenting out all

```
@controller.perform_caching = true
```

in `actionpack/test/controller/caching_test.rb` and run the tests. They
should still pass.

Make docs on form_with_generates_ids config option match the value

[skip ci]

Update comment in attribute_method_matchers_matching

Remove reminiscences from previous documentation

Add attribute_names to ActiveModel::Attributes

Add ActionController::Caching into api app's document [ci skip]

Use ActiveJob 5.2 retry logic for old jobs