Merge pull request #37508 from rmacklin/enforce-that-activestorage-compiled-bundle-is-in-sync-with-source-code

Enforce that activestorage compiled JS is in sync with source code

(Note: this is effectively the same change that we did for actioncable
in fefc3041 applied to activestorage.)

We have run into issues in the past where the activestorage compiled
javascript bundle got out of sync with the source code. For example, in
880f9779 the compiled bundle was
modified in a way that was inconsistent with the results of running
`yarn build` in the activestorage directory. This meant that running
`yarn build` there would produce a dirty git status indicating changes
to the compiled bundle, despite not having made any changes to the
activestorage javascript source code. That particular inconsistency was
fixed in 0e77706d. However, the same
problem could reoccur.

To address this, I've added a new test to enforce that activestorage's
compiled javascript bundle is in sync with the source code. When the
compiled bundle is in sync with the source code, the test will pass.
However, if the two are not in sync, the test will fail. For example, if
you were to apply the following patch (which only updates the source
code):

```
diff --git a/activestorage/app/javascript/activestorage/index.js b/activestorage/app/javascript/activestorage/index.js
index a340008fb9..a05b7223a5 100644
--- a/activestorage/app/javascript/activestorage/index.js
+++ b/activestorage/app/javascript/activestorage/index.js
@@ -8,4 +8,4 @@ function autostart() {
   }
 }

-setTimeout(autostart, 1)
+setTimeout(autostart, 2)
```
the test would then fail. Thus, the activestorage test suite will now
prevent "the compiled bundle is out of sync" issues going forward.

We have run into issues in the past where the actioncable compiled
javascript bundle got out of sync with the source code. For example, in
30a0c7e0 only the compiled bundle was
modified. This meant that anyone who ran `yarn build` in the actioncable
directory would then see a dirty git status indicating changes to the
compiled bundle, despite not having made any changes to the actioncable
javascript source code. We fixed that particular inconsistency in
a4c27588. However, the same problem
could reoccur.

To address this, I've added a new test to enforce that actioncable's
compiled javascript bundle is in sync with the source code. When the
compiled bundle is in sync with the source code, the test will pass:

    $ bundle exec ruby -Itest test/javascript_package_test.rb
    Run options: --seed 19308

    # Running:

    yarn run v1.12.3
    $ yarn lint && bundle exec rake assets:codegen
    $ eslint app/javascript
    $ rollup --config rollup.config.js

    app/javascript/action_cable/index.js → app/assets/javascripts/action_cable.js...
    created app/assets/javascripts/action_cable.js in 762ms
    ✨  Done in 6.35s.
    .

    Finished in 7.130345s, 0.1402 runs/s, 0.1402 assertions/s.
    1 runs, 1 assertions, 0 failures, 0 errors, 0 skips

However, if the two are not in sync, the test will fail. For example, if
you were to apply the following patch (which only updates the source
code):

```
diff --git a/actioncable/app/javascript/action_cable/adapters.js b/actioncable/app/javascript/action_cable/adapters.js
index 4de8131438..d38d9a6a0b 100644
--- a/actioncable/app/javascript/action_cable/adapters.js
+++ b/actioncable/app/javascript/action_cable/adapters.js
@@ -1,4 +1,5 @@
 export default {
+  foo: self.foo,
   logger: self.console,
   WebSocket: self.WebSocket
 }
```

the test would then fail like this:

    $ bundle exec ruby -Itest test/javascript_package_test.rb
    Run options: --seed 26377

    # Running:

    yarn run v1.12.3
    $ yarn lint && bundle exec rake assets:codegen
    $ eslint app/javascript
    $ rollup --config rollup.config.js

    app/javascript/action_cable/index.js → app/assets/javascripts/action_cable.js...
    created app/assets/javascripts/action_cable.js in 776ms
    ✨  Done in 5.55s.
    F

    Failure:
    JavascriptPackageTest#test_compiled_code_is_in_sync_with_source_code [test/javascript_package_test.rb:16]:
    --- expected
    +++ actual
    @@ -3,6 +3,7 @@
     })(this, function(exports) {
       \"use strict\";
       var adapters = {
    +    foo: self.foo,
         logger: self.console,
         WebSocket: self.WebSocket
       };

    rails test test/javascript_package_test.rb:9

    Finished in 5.837403s, 0.1713 runs/s, 0.1713 assertions/s.
    1 runs, 1 assertions, 1 failures, 0 errors, 0 skips

Thus, the actioncable test suite will now prevent "the compiled bundle
is out of sync" issues going forward.

Co-authored-by: NChristian Gregg <c_arlt@hotmail.com>

When `ActiveSupport.parse_json_times` is `true`,
`ActiveSupport::Messages::Medata.verify` will fail with a `TypeError` as
it does not pass a `String` to `Time.iso8601` as is expected.

This leads to hard-failure on retrieval of signed cookies with expiries.

Add methods to the public API of ActiveRecord::Inheritance

Those methods make possible to extend STI and Polymorphic associations.

They are useful for cases where you renamed a class and the name of the
class doesn't match the data in the database.

You can now implement those methods in your model to load records with
name of classes that don't exist anymore.

A simple implementation would look like:

    class Animal < ActiveRecord::Base
      @@old_names = {
        "Lion" => "BigCat"
      }
      def self.sti_name
        name = super
        @@old_names[name] || name
      end

      def self.sti_class_for(type_name)
        @@old_names.inverse[type_name]&.constantize || super
      end
    end

Fix random CI fail due to non-deterministic order

See https://buildkite.com/rails/rails/builds/64385#78d02d3f-51b4-45fc-af67-a5b9b543d2cc

Although `in_batches` does apply an order to prevent batch overlap, the
relation it yields to its block does not.  Therefore, test results must
be sorted for comparison.

The intent of that tests is completely covered by with order version.

Enumerator should be supported by ActiveSupport::SafeBuffer

Follow-up to #33615 and #37447.

Back references cannot be set because C level Proc doesn't support Binding.
This commit fixes #37422.

Merge pull request #37489 from giraffate/fix_random_ci_failure_due_to_non-deterministic_sorting_order

Fix random CI failure due to non-deterministic sorting order

Reference:
https://buildkite.com/rails/rails/builds/64354#206d3411-0f15-4aa4-aa25-83117ef3bcc0/1006-1017

When editing credentials.yml.enc after creating a symbolic link,
the symbolic link is removed and overwritten with a normal file.
This commit changes the behavior to keep the symbolic link.

In case of using a dead symbolic link that link is revived after writing.

Fixes #36411

A class is only a direct descendant if its superclass is self.

Fixes https://github.com/rails/rails/pull/37487

We don't want users or apps to use Role, so nodoc it.

Manage connection pools in a new Role class rather than in DatabaseConfig

Use ObjectSpace.each_object enumerator in descendants

Remove patch for old JRuby versions

JRuby 9.0.5.0 (released January 26 2016 and Ruby 2.2 compatible)
supports each_object(singleton_class) (#22385). Given that Rails 6
requires Ruby 2.5.0 or newer, we should be able to drop support for
versions of JRuby older than 9.0.5.0.

Support ActiveModel::Error translation lookup on indexed attributes.

Opt into has many inverse

Documentation: Add two RDoc links to methods

Update multi_json to the latest version

Avoid `warning: tried to create a Proc object without giving a block`
in Ruby head.

Fixed in intridea/multi_json@26a94ab8c78a394cc237e2ea292c1de4f6ed30d7 .

[ci skip] Remove reference to legacy Rails 2.x command

Fixed performance regression introduced MySQL 8.0