Follow up to #35154.

Specify sqlite gem version explicitly in bug report templates

Support before_reset callback in CurrentAttributes

Relax sqlite3 version dependency

Refs: rails/jbuilder#452

With this benchmark:

    require "bundler/setup"

    require "active_record"
    require "benchmark/ips"

    # This connection will do for database-independent bug reports.
    ActiveRecord::Base.establish_connection(adapter: "sqlite3", database: ":memory:")

    ActiveRecord::Schema.define do
      create_table :posts, force: true do |t|
      end
    end

    class Post < ActiveRecord::Base
    end

    new_post = Post.new

    Benchmark.ips do |b|
      b.report("present?") do
        new_post.present?
      end

      b.report("blank?") do
        new_post.blank?
      end
    end

Before:

    Warming up --------------------------------------
                present?    52.147k i/100ms
                  blank?    53.077k i/100ms
    Calculating -------------------------------------
                present?    580.184k (±21.8%) i/s -      2.555M in   5.427085s
                  blank?    601.537k (± 9.2%) i/s -      2.972M in   5.003503s

After:

    Warming up --------------------------------------
                present?   378.235k i/100ms
                  blank?   375.476k i/100ms
    Calculating -------------------------------------
                present?     17.381M (± 7.5%) i/s -     86.238M in   5.001815s
                  blank?     17.877M (± 6.4%) i/s -     88.988M in   5.004634s

This improvement is mostly because those methods were hitting
`method_missing` on a lot of levels to be able to return the value.

To avoid all this stack walking we are short-circuiting those methods.

Closes #35059.

Cookie doesn't expire anymore unless a flag is set:

Add a link to where "inquiry" is defined in ActiveSupport docs [ci skip]

[ci skip]

Cleanup the whitelisting references after #33145

Update template missing text in Guide

Changes the Getting Started guide explanation for
`ActionController::MissingExactTemplate` error, to reflect the current message.

Follow up for #29286, #35148

[ci skip]

Invalidate all query caches for current thread

Add ability to change the names of the default handlers

- There is a regression in 6.0 introduced by #32937 where cookie
  doesn't expire anymore unless the new `use_cookies_with_metadata`
  configuration is set to `true`.

  This causes issue for app migration from 5.2 to 6.0 because the
  `use_cookies_with_metadata` flag can't be set to true until all
  servers are running on 6.0.

  Here is a small reproduction script that you can run in the console

  ```ruby
  ActionDispatch::Cookies

  request = ActionDispatch::Request.empty
  request.env["action_dispatch.key_generator"] = ActiveSupport::KeyGenerator.new('1234567890')
  request.env["action_dispatch.signed_cookie_salt"] = 'signed cookie'
  request.env["action_dispatch.cookies_rotations"] = ActiveSupport::Messages::RotationConfiguration.new
  request.env["action_dispatch.use_authenticated_cookie_encryption"] = true
  signed_cookie = request.cookie_jar.signed
  signed_cookie[:foobar] = { value: '123', expires: 1.day.ago }
  p signed_cookie[:foobar]
  ```

Fix and update template_missing image in guide

This commit updates the Missing Template image in the Getting Started
Guide.

The new image displays the current error message. Additionally:
- New image is a PNG file, whereas the current image is in fact a misnamed JPEG.
- New image is smaller.

Before:

```
$ file --mime -b template_is_missing_articles_new.png
image/jpeg; charset=binary
$ du -h template_is_missing_articles_new.png
464K    template_is_missing_articles_new.png
```

After:

```
$ file --mime -b template_is_missing_articles_new.png
image/png; charset=binary
$ du -h template_is_missing_articles_new.png
28K     template_is_missing_articles_new.png
```

[ci-skip]

During the development of #33145, I have named a few concepts in the
code as `whitelisted`. We decided to stay away from the term and I
adjusted most of the code afterwards, but here are the cases I forgot to
change.

I also found a case in the API guide that we could have cleaned up as
well.

[ci skip]

Hint at advanced options for foreign_key

Use consistent hash syntax for routes

Introduce a file type template

[ci skip]

This commit passes the mutated source to the template handler as a
parameter and deprecates the old handlers.  Old handlers required that
templates contain a reference to mutated source code, but we would like
to make template objects "read only".  This change lets the template
remain "read only" while still giving template handlers access to the
source code after mutations.

Make Notifications::Fanout faster after changing subscriptions

This change ensures that all query cahces are cleared across all
connections per handler for the current thread so if you write on one
connection the read will have the query cache cleared.

When adding/removing a subscription with a string pattern, it isn't
necessary to clear the entire cache, only the cache for the key being
added.

When adding/removing subscriptions for a regex or to match all events,
the full cache is still cleared.

Previously we stored all subscribers in an array, and every time a new
message name came in asked each subscriber if they responded to the
message.

This commit changes Fanout to hash subscribers with a String pattern by
their pattern. This way we can look them up directly and only do the
O(N) scan over the non-string (Regex or any) patterns.

Every template that specifies a "virtual path" loses the template source
when the template gets compiled:

  https://github.com/rails/rails/blob/eda0f574f129fcd5ad1fc58b55cb6d1db71ea95c/actionview/lib/action_view/template.rb#L275

The "refresh" method seems to think that the source code for a template
can be recovered if there is a virtual path:

  https://github.com/rails/rails/blob/eda0f574f129fcd5ad1fc58b55cb6d1db71ea95c/actionview/lib/action_view/template.rb#L171-L188

Every call site that allocates a template object *and* provides a
"virtual path" reads the template contents from the filesystem:

  https://github.com/rails/rails/blob/eda0f574f129fcd5ad1fc58b55cb6d1db71ea95c/actionview/lib/action_view/template/resolver.rb#L229-L231

Templates that are inline or literals don't provide a "virtual path":

  https://github.com/rails/rails/blob/eda0f574f129fcd5ad1fc58b55cb6d1db71ea95c/actionview/lib/action_view/renderer/template_renderer.rb#L34

This commit introduces a `FileTemplate` type that subclasses `Template`.
The `FileTemplate` keeps a reference to the filename, and reads the
source from the filesystem.  This effectively makes the template source
immutable.

Other classes depended on the source to be mutated while being compiled,
so this commit also introduces a temporary way to pass the mutated
source to the ERB (or whatever) compiler.  See `LegacyTemplate`.

I think we should consider it an error to provide a virtual path on a
non file type template an non-file templates can't recover their source.
Here is an example:

  https://github.com/rails/rails/blob/eda0f574f129fcd5ad1fc58b55cb6d1db71ea95c/actionview/lib/action_view/testing/resolvers.rb#L53

This provides a "virtual path" so the source code (a string literal) is
thrown away after compilation.  Clearly we can't recover that string, so
I think this should be an error.

When I wrote the `connected_to` and `connects_to` API's I wrote them
with the idea in mind that it didn't really matter what the
handlers/roles were called as long as those connecting to the roles knew
which one wrote and which one read.

With the introduction of the middleware Rails begins to assume it's
`writing` and `reading` and there's no room for other roles. At GitHub
we've been using this method for a long time so we have a ton of legacy
code that uses different handler names `default` and `readonly`. We
could rename all our code but I think this is better for a few reasons:

- Legacy apps that have been using multiple databases for a long time
  can have an eaiser time switching.
- If we later find this to cause more issues than it's worth we can
  easily deprecate.
- We won't force old apps to rewrite the resolver middleware just to use
  a different handler.

Adding the writing_role/reading_role required that I move the code that
creates the first handler for writing to the railtie. If I didn't move
this the core class would assign the handler before I was able to assign
a new one in my configuration and I'd end up with 3 handlers instead of
2.

Refactor options for database selector middleware

Eagerly materialize the fixtures transaction

ActiveSupport typo fixes.

Right now we only have one option that's supported, the delay. However I
can see us supporting other options in the future.

This PR refactors the options to get passed into the resolver so whether
you're using middleware or using the config options you can pass options
to the resolver. This will also make it easy to add new options in the
future.

include the content type when uploading to S3

Railties typo fixes.