When nil or empty string are not allowed, they are not valid.

assert_template: validating option keys

It only handles the keys locals, partial, layout and count.

    assert_template(foo: "bar") # raises ArgumentError
    assert_template(leiaute: "test") # raises ArgumentError

Add explicit opt-out for fragment cache digesting

This add support for sending an explicit opt-out of the "Russian-doll"
cache digest feature on a case-by-case basis. This is useful when cache-
expiration needs to be performed manually and it would be otherwise
difficult to know the exact name of a digested cache key.

More information: https://github.com/rails/cache_digests/pull/16

assert_template("") will now fail no matter whether a template has been rendered or not.

Don't call will_change! for datetime nil->"".

Setting a nil datetime attribute to a blank string should not cause the
attribute to be dirty.

Fix #8310

Document :hidden_field_id option for fields_for [ci skip]

Introduced in 53ca22f2.
Thanks @rochefort. [ci skip]

This yields a small bit of performance improvement when building the
defaults from constraints, specially considering that it's rather common
for constraints to be empty.

Also, there's a bit of duplicated code in here that I have to check
before extracting.

Instead of iterating again over the options and setting one by one, we
can just merge the recover hash back to the scope one since all keys
match.

Use the same :blocks key in the recover hash to revert the scope options
later.

For instance, it prevents false positive in this case:

    file = nil
    get :index
    assert_template("#{file}")

prevent mass assignment of polymorphic type when using `build`

Conflicts:
	activerecord/CHANGELOG.md

[ci skip]

Allow setting a symbol as path in scope on routes

Closes #8265

Was surprising found that this example doesn't work:

  scope :api do
    resources :users
  end

and the right form to use it is:

  scope 'api' do
    resources :users
  end

I think this should work similary as `namespace` where both are allowed.
These two are equivalent:

  namespace :api do
    resources :users
  end

  namespace 'api' do
    resources :user
  end

Since now these objects are not cached anymore, there's no need to have
these private methods, just instantiate each of them in the necessary
place.

Follow up of the discussion from the original merge commit:
https://github.com/rails/rails/commit/f9cb645dfcb5cc89f59d2f8b58a019486c828c73#commitcomment-1414561

We want to avoid people's mistakes with methods like count and sum when
called with a block, that can easily lead to code performing poorly and
that could be way better written with a db query.

Please check the discussion there for more background.
Closes #8268

Since edd94cee, CollectionProxy
delegates all calculation methods - except count - to the scope,
which does basically what this method was doing, but since we're
delegating from the proxy, the association method was never called.

To perform a sum calculation over the array of elements, use to_a.sum(&block).

Please check the discussion in f9cb645d
for more context.

This reverts commit f9cb645d.

Conflicts:
	activerecord/CHANGELOG.md

Revert "Allow blocks for count with ActiveRecord::Relation. Document and test that sum allows blocks"

This reverts commit 9cc2bf69.

Conflicts:
	activerecord/lib/active_record/relation/calculations.rb

This improves memory and performance without having to use symbols which
present DoS problems. Thanks @headius and @tenderlove for the
suggestion.

This was originally committed in
f1765019, and then reverted in
d3494903 due to it causing problems in a
real application. This second attempt should solve that.

Benchmark
---------

require 'active_record'
require 'benchmark/ips'

ActiveRecord::Base.establish_connection(adapter: 'sqlite3', database: ':memory:')

class Post < ActiveRecord::Base
  connection.create_table :posts, force: true do |t|
    t.string :name
  end
end

post = Post.create name: 'omg'

Benchmark.ips do |r|
  r.report('Post.new')          { Post.new name: 'omg' }
  r.report('post.name')         { post.name }
  r.report('post.name=')        { post.name = 'omg' }
  r.report('Post.find(1).name') { Post.find(1).name }
end

Before
------

Calculating -------------------------------------
            Post.new      1419 i/100ms
           post.name      7538 i/100ms
          post.name=      3024 i/100ms
   Post.find(1).name       243 i/100ms
-------------------------------------------------
            Post.new    20637.6 (±12.7%) i/s -     102168 in   5.039578s
           post.name  1167897.7 (±18.2%) i/s -    5186144 in   4.983077s
          post.name=    64305.6 (±9.6%) i/s -     317520 in   4.998720s
   Post.find(1).name     2678.8 (±10.8%) i/s -      13365 in   5.051265s

After
-----

Calculating -------------------------------------
            Post.new      1431 i/100ms
           post.name      7790 i/100ms
          post.name=      3181 i/100ms
   Post.find(1).name       245 i/100ms
-------------------------------------------------
            Post.new    21308.8 (±12.2%) i/s -     105894 in   5.053879s
           post.name  1534103.8 (±2.1%) i/s -    7634200 in   4.979405s
          post.name=    67441.0 (±7.5%) i/s -     337186 in   5.037871s
   Post.find(1).name     2681.9 (±10.6%) i/s -      13475 in   5.084511s

Around 0.564359s => 0.092244s speed up in my machine.

Store FlashHashes in the session as plain hashes

Coerce strings in create_join_table.

If you accidentally pass a string and a symbol, this breaks. So
we coerce them both to strings.

Fixes #7715

Fixed a typo