A lot of scripts assumes the existence of this folder and most would fail if it
is absent.

One example of this is `rake restart` (before the previous commit) – it tries to
`touch tmp/restart.txt`, which would fail if `tmp` does not exist, which was the
case for a freshly-cloned project as `tmp` is `.gitignored` by default.

See #20299.

[Yoong Kang Lim, Sunny Juneja]

Set `config.static_index` to serve a static directory index file not
named `index`. For example, to serve `main.html` instead of `index.html`
for directory requests, set `config.static_index` to `"main"`.

Use the mail gem's own API to locate the correct part.

Fixes #14435.

Punctuation fix

Support for versions of SQLite less than 3 was removed in #6011 as part
of the Rails 4.0 release. Therefore there is no need to have support for
it in the `rails dbconsole` command anymore.

This is a squash of the following commits, from first to last:

-

Fix minor, random things I’ve come across lately that individually
did not seem worth making a PR for, so I saved them for one commit.

One common error is using “it’s” (which is an abbreviation of “it is”)
when the possessive “its” should be used for indicating possession.

-

Changes include the name of a test, so remove the `[skip ci]` (thanks @senny).

-

Line wrap the changes at 80 chars and add one more doc fix.

-

Add a missing line wrap in the Contributing to Ruby on Rails Guide.

-

Line wrap the `TIP` section in the Contributing to Ruby on Rails Guide as well.

Rendering the guide locally with `bundle exec rake guides:generate` did
not show any change in on-screen formatting after adding the line wrap.

The HTML generated is (extra line added to illustrate where the line
wrap takes place):

    <div class="info"><p>Please squash your commits into a single commit
when appropriate. This

    simplifies future cherry picks and also keeps the git log
clean.</p></div>

-

Squash commits.

That will match the new Bundler executables convention.
Bundler Blog Post: http://bundler.io/blog/2015/03/20/moving-bins-to-exe.html
Also updated the necessary tests.

Previously, running `rails new` would not print any of the output from
`bundle install` until all the gems had finished installing. This made
it look like the generator was hanging at the `bundle install` step.

This commit switches to using `system` so that the bundle command can
output as it needs to.

This has the added benefit of including output bundler produces on
standard error, which the previous code ignored since backticks only
capture standard out. This is not a big deal right now since bundler
does not currently print errors to standard error, but that may change
in the future (see: bundler/bundler/issues/3353).

Also edited the copy to be more consistent with `rake routes`.

See #19323.

Preserving RACK_ENV behavior.

This reverts commit b19990c8, reversing
changes made to ac291b76.

[ci skip]

Fix regression on route method that was added by
bac812a7. The regression was that when
calling the `route` method, we were not appending a \n anymore.

[fixes #19316]

This task prints out initializers for an application. It is useful to
develop a rubygem which involves the initialization process.

This introduces undesirable `Rails.logger` formatters (such as the syslog
formatter) onto a `Logger.new(STDERR)` for the console. The production
logger may be going elsewhere than standard io, so we can't presume to
reuse its formatter.

With syslog, this causes missing newlines in the console, so irb prompts
start at the end of the last log message.

We can work to expose the console formatter in another way to address
the original issue.

This reverts commit 026ce5dd, reversing
changes made to 6f0a69c5.

Fixes #18876. Rake restart touches `tmp/restart.txt` to restart
application on next request. Updated tests and documentation
accordingly.

Deprecate `required` option in favor of `optional` for belongs_to.

Fixes: https://github.com/rails/rails/issues/16433.

README.rdoc was generated to support the doc:app task. Now that
this task is gone we can switch to Markdown, which is nowadays
a better default.

This patch removes the tasks doc:app, doc:rails, and doc:guides.

In our experience applications do not generate APIs using doc:app.
Methods may be certainly documented for maintainers, annotated
with YARD tags, etc. but that is intended to be read with the
source code, not in a separate website. Then, teams also have
typically selected topics written down in Markdown files, or in
a GitHub wiki... that kind of thing.

If a team absolutely needs to generate application documentation
for internal purposes, they can still easily write their own task.

Regarding doc:rails and doc:guides, we live in 2015. We are used
to go to online docs all the time. If you really want access to the
API offline RubyGems generates it for every Rails component unless
you tell it not to, and you can checkout the Rails source code to
read the guides as Markdown, or download them for a Kindle reader.

All in all, maintaining this code does not seem to be worthwhile
anymore.

As a consequence of this, guides (+3 MB uncompressed) won't be
distributed with the rails gem anymore. Of course, guides and API
are going to be still part of releases, since documentation is
maintained alongside code and tests.

Also, time permitting, this will allow us to experiment with novel
ways to generate documentation in the Rails docs server, since
right now we were constrained by being able to generate them in
the user's environment.

Thor isn't very discerning over whether some content is present when passed
to `inject_into_file`, e.g. a commented out route is detected as being present.
So to prevent people scratching their heads as to why a route hasn't appeared
it's better to fall on the side of having duplicate routes.

When there is a single scaffold in the routes.rb with no other lines
then revoking/destroying it will create a routes.rb file with a syntax
error. This is because the sentinel for the Thor `route` action didn't
include the newline but the logged route code did.

The fix is to add the newline to the sentinel and remove it from the
the logged route code.

Fixes #15913.

Rails no longer generates Test::Unit files by default.

Fixes #18473

- test/functional
- test/unit

This stems from [a comment](rails#17227 (comment)) by @dhh.
In summary:

* New Rails 5.0 apps will not accept `return false` as a way to halt callback chains, and will not display a deprecation warning.
* Existing apps ported to Rails 5.0 will still accept `return false` as a way to halt callback chains, albeit with a deprecation warning.

For this purpose, this commit introduces a Rails configuration option:

```ruby
config.active_support.halt_callback_chains_on_return_false
```

For new Rails 5.0 apps, this option will be set to `false` by a new initializer
`config/initializers/callback_terminator.rb`:

```ruby
Rails.application.config.active_support.halt_callback_chains_on_return_false = false
```

For existing apps ported to Rails 5.0, the initializers above will not exist.
Even running `rake rails:update` will not create this initializer.

Since the default value of `halt_callback_chains_on_return_false` is set to
`true`, these apps will still accept `return true` as a way to halt callback
chains, displaying a deprecation warning.

Developers will be able to switch to the new behavior (and stop the warning)
by manually adding the line above to their `config/application.rb`.

A gist with the suggested release notes to add to Rails 5.0 after this
commit is available at https://gist.github.com/claudiob/614c59409fb7d11f2931

We will support only Ruby >= 2.1.

But right now we don't accept pull requests with syntax changes to drop
support to Ruby 1.9.

If you run a generator such as:

```
rails generate model accounts supplier:references
```

The resulting migration will now add the corresponding foreign key
constraint unless the reference was specified to be polymorphic.