I incorrectly changed behavior of `dup`. Reading the original issue I
thought that `dup` should retain the original contents of the record
and it's associations but it is in fact supposed to be a copy as if a
record had been reinitialized.

This reverts commit ca8c21df.

Use `tables` instead of `data_sources - views`

Take failed screenshot before reset driver

Improve documentation for Testing Your Mailers [ci skip]

The current Basic Test Case example has the following assertion

```
  assert_equal read_fixture('invite').join, email.body.to_s

```

email.body.to_s returns an empty string if both HTML and text templates
exist for a given mailer. This commit adds a note to section 11.2.2
explaining this and also suggests using email.text_part.body.to_s
and email.html_part.body.to_s as alternatives.

Fix http -> https [ci skip]

In Rails 3.2 dupping a `CollectionProxy` would dup it's `load_target` as
well. That functionality has been broken since the release of Rails 4.0.
I hit this in an application upgrade and wondered why duplicating a
CollectionProxy and assigning it to a variable stopped working.

When calling `dup` on a `CollectionProxy` only the owner (ex.
topic) was getting duplicated and the `load_target` would remain in tact
with it's original object ID. Dupping the `load_target` is useful for performing
a logging operation after records have been destroyed in a method.

For example:

```
def transfer_operation
  saved_replies = topic.replies

  topic.replies.clear

  saved_replies.each do |reply|
    user.update_replies_count!
  end
end
```

This change adds a `initialize_dup` method that performs a `deep_dup` on
the `@associatiation` so that the `load_target` is dupped as well.

Fixes #17117

Oracle database also does not allow aliases in the having clause

Now reset the driver before take failed screenshot since #28144.
However, I think that failed screenshot should be take with the driver
actually used in the test.
So, fixed to take screenshot before reset driver.

`tables` returns only tables now.

Use released webpacker in new applications

Remove `encode_special_chars` option from `strip_tags`

Because webpacker 1.0 already released.

Follow up #28183

Set correct host except development environment

Handle non-existing $ARCONN

Only load SystemTestCase if Puma is defined

Currently `localhost` is used for the default host in all environments.
But up to Rails 5.0, `0.0.0.0` is used except for development.
So fixed to use the same value as 5.0.

Fixes #28184

SystemTestCase supports only Puma, and always load puma's file.
https://github.com/rails/rails/blob/master/actionpack/lib/action_dispatch/system_testing/server.rb#L1

For that reason, the case of use Capybara but do not use Puma, it will cause an error.
So we need to check about Puma is defined as well.

Use `max_identifier_length` for `index_name_length`

Fix `test_apply_distinct_in_count` failure in PG adapter

Fix `find_nth` with `limit_value`

Update i18n guide to cover :zero key support in pluralization [ci skip]

```
% ARCONN=postgresql be ruby -w -Itest test/cases/calculations_test.rb -n test_apply_distinct_in_count
/Users/kamipo/src/github.com/rails/rails/activesupport/lib/active_support/core_ext/enumerable.rb:20: warning: method redefined; discarding old sum
Using postgresql
Run options: -n test_apply_distinct_in_count --seed 27731

F

Finished in 0.163682s, 6.1094 runs/s, 30.5470 assertions/s.

  1) Failure:
CalculationsTest#test_apply_distinct_in_count [test/cases/calculations_test.rb:238]:
Expected /\ASELECT(?! DISTINCT) COUNT\(DISTINCT\b/ to match "SHOW max_identifier_length".

1 runs, 5 assertions, 1 failures, 0 errors, 0 skips
```

https://travis-ci.org/rails/rails/jobs/205493811#L1933
https://travis-ci.org/rails/rails/jobs/205496299#L1562
https://travis-ci.org/rails/rails/jobs/205514765#L1562

Actually `index_name_length` depend on `max_identifier_length`, not
always 63.

Deprecate `supports_migrations?` on connection adapters

Refactor connection_pool_test

`supports_migrations?` was added at 4160b518 to determine if schema
statements (`create_table`, `drop_table`, etc) are implemented in the
adapter. But all tested databases has been supported migrations since
a4fc93c3 at least.

Include selects in group query with having clause

When a grouped calculation contains a having clause that references a
selected value, we need to include that selected value in the query.

Postgres doesn't support referencing a selected value in a having
clause, but other databases do; we can skip the test on the pg adapter
but run it for the others.

This was fixed before in 9a298a16, but
the test coverage was lost in 5a05207d.
The fix regressed in 6311975f and was
removed in 97d46c17.

Remove unused params in RouteSet#add_route

If the `index` exceeds a `limit`, simply return an empty result without
querying the database.

Push `valid_type?` up to abstract adapter