search

rails / spring

Rails application preloader
MIT License
2.81k stars 341 forks source link

readme

Spring

Build Status Gem Version

Spring is a Rails application preloader. It speeds up development by keeping your application running in the background, so you don't need to boot it every time you run a test, rake task or migration.

Features

Compatibility

Spring makes extensive use of Process.fork, so won't be able to provide a speed up on platforms which don't support forking (Windows, JRuby).

Walkthrough

Setup

Add Spring to your Gemfile:

gem "spring", group: :development

(Note: using gem "spring", git: "..." won't work and is not a supported way of using Spring.)

It's recommended to 'springify' the executables in your bin/ directory:

$ bundle install
$ bundle exec spring binstub --all

This generates a bin/spring executable, and inserts a small snippet of code into relevant existing executables. The snippet looks like this:

begin
  load File.expand_path('../spring', __FILE__)
rescue LoadError => e
  raise unless e.message.include?('spring')
end

On platforms where Spring is installed and supported, this snippet hooks Spring into the execution of commands. In other cases, the snippet will just be silently ignored, and the lines after it will be executed as normal.

If you don't want to prefix every command you type with bin/, you can use direnv to automatically add ./bin to your PATH when you cd into your application. Simply create an .envrc file with the command PATH_add bin in your Rails directory.

Enable reloading

Spring reloads application code, and therefore needs the application to have reloading enabled.

Ensure that config.enable_reloading is true in the environments that Spring manages. That setting is typically configured in config/environments/*.rb. In particular, make sure it is true for the test environment.

Note: in versions of Rails before 7.1, the setting is called cache_classes, and it needs to be false for Spring to work.

Usage

For this walkthrough I've generated a new Rails application, and run rails generate scaffold post name:string.

Let's run a test:

$ time bin/rake test test/controllers/posts_controller_test.rb
Running via Spring preloader in process 2734
Run options:

# Running tests:

.......

Finished tests in 0.127245s, 55.0121 tests/s, 78.5887 assertions/s.

7 tests, 10 assertions, 0 failures, 0 errors, 0 skips

real    0m2.165s
user    0m0.281s
sys     0m0.066s

That wasn't particularly fast because it was the first run, so Spring had to boot the application. It's now running:

$ bin/spring status
Spring is running:

26150 spring server | spring-demo-app | started 3 secs ago
26155 spring app    | spring-demo-app | started 3 secs ago | test mode

The next run is faster:

$ time bin/rake test test/controllers/posts_controller_test.rb
Running via Spring preloader in process 8352
Run options:

# Running tests:

.......

Finished tests in 0.176896s, 39.5714 tests/s, 56.5305 assertions/s.

7 tests, 10 assertions, 0 failures, 0 errors, 0 skips

real    0m0.610s
user    0m0.276s
sys     0m0.059s

If we edit any of the application files, or test files, the changes will be picked up on the next run without the background process having to restart. This works in exactly the same way as the code reloading which allows you to refresh your browser and instantly see changes during development.

But if we edit any of the files which were used to start the application (configs, initializers, your gemfile), the application needs to be fully restarted. This happens automatically.

Let's "edit" config/application.rb:

$ touch config/application.rb
$ bin/spring status
Spring is running:

26150 spring server | spring-demo-app | started 36 secs ago
26556 spring app    | spring-demo-app | started 1 sec ago | test mode

The application detected that config/application.rb changed and automatically restarted itself.

If we run a command that uses a different environment, then that environment gets booted up:

$ bin/rake routes
Running via Spring preloader in process 2363
    posts GET    /posts(.:format)          posts#index
          POST   /posts(.:format)          posts#create
 new_post GET    /posts/new(.:format)      posts#new
edit_post GET    /posts/:id/edit(.:format) posts#edit
     post GET    /posts/:id(.:format)      posts#show
          PUT    /posts/:id(.:format)      posts#update
          DELETE /posts/:id(.:format)      posts#destroy

$ bin/spring status
Spring is running:

26150 spring server | spring-demo-app | started 1 min ago
26556 spring app    | spring-demo-app | started 42 secs ago | test mode
26707 spring app    | spring-demo-app | started 2 secs ago | development mode

There's no need to "shut down" Spring. This will happen automatically when you close your terminal. However, if you do want to do a manual shut down, use the stop command:

$ bin/spring stop
Spring stopped.

From within your code, you can check whether Spring is active with if defined?(Spring).

Removal

To remove Spring:

Deployment

You must not install Spring on your production environment. To prevent it from being installed, run the bundle config set without 'development test' before bundle install command which is used to install gems on your production machines:

$ bundle config set without 'development test'
$ bundle install

Commands

rake

Runs a rake task. Rake tasks run in the development environment by default. You can change this on the fly by using the RAILS_ENV environment variable. The environment is also configurable with the Spring::Commands::Rake.environment_matchers hash. This has sensible defaults, but if you need to match a specific task to a specific environment, you'd do it like this:

Spring::Commands::Rake.environment_matchers["perf_test"] = "test"
Spring::Commands::Rake.environment_matchers[/^perf/]     = "test"

# To change the environment when you run `rake` with no arguments
Spring::Commands::Rake.environment_matchers[:default] = "development"

rails console, rails generate, rails runner

These execute the rails command you already know and love. If you run a different sub command (e.g. rails server) then Spring will automatically pass it through to the underlying rails executable (without the speed-up).

Additional commands

You can add these to your Gemfile for additional commands:

Use without adding to bundle

If you don't want Spring-related code checked into your source repository, it's possible to use Spring without adding to your Gemfile. However, using Spring binstubs without adding Spring to the Gemfile is not supported.

To use Spring like this, do a gem install spring and then prefix commands with spring. For example, rather than running bin/rake -T, you'd run spring rake -T.

Temporarily disabling Spring

If you're using Spring binstubs, but temporarily don't want commands to run through Spring, set the DISABLE_SPRING environment variable.

Class reloading

Spring uses Rails' class reloading mechanism to keep your code up to date between test runs. This is the same mechanism which allows you to see changes during development when you refresh the page. However, you may never have used this mechanism with your test environment before, and this can cause problems.

It's important to realise that code reloading means that the constants in your application are different objects after files have changed:

$ bin/rails runner 'puts User.object_id'
70127987886040
$ touch app/models/user.rb
$ bin/rails runner 'puts User.object_id'
70127976764620

Suppose you have an initializer config/initializers/save_user_class.rb like so:

USER_CLASS = User

This saves off the first version of the User class, which will not be the same object as User after the code has been reloaded:

$ bin/rails runner 'puts User == USER_CLASS'
true
$ touch app/models/user.rb
$ bin/rails runner 'puts User == USER_CLASS'
false

So to avoid this problem, don't save off references to application constants in your initialization code.

Using Spring with a containerized development environment

As of Spring 1.7, there is some support for doing this. See this example repository for information about how to do it with Docker.

Configuration

Spring will read ~/.spring.rb and config/spring.rb for custom settings. Note that ~/.spring.rb is loaded before bundler, but config/spring.rb is loaded after bundler. So if you have any spring-commands-* gems installed that you want to be available in all projects without having to be added to the project's Gemfile, require them in your ~/.spring.rb.

config/spring_client.rb is also loaded before bundler and before a server process is started, it can be used to add new top-level commands.

Application root

Spring must know how to find your Rails application. If you have a normal app everything works out of the box. If you are working on a project with a special setup (an engine for example), you must tell Spring where your app is located:

Spring.application_root = './test/dummy'

Running code before forking

There is no Spring.before_fork callback. To run something before the fork, you can place it in ~/.spring.rb or config/spring.rb or in any of the files which get run when your application initializes, such as config/application.rb, config/environments/*.rb or config/initializers/*.rb.

Running code after forking

You might want to run code after Spring forked off the process but before the actual command is run. You might want to use an after_fork callback if you have to connect to an external service, do some general cleanup or set up dynamic configuration.

Spring.after_fork do
  # run arbitrary code
end

If you want to register multiple callbacks you can simply call Spring.after_fork multiple times with different blocks.

Watching files and directories

Spring will automatically detect file changes to any file loaded when the server boots. Changes will cause the affected environments to be restarted.

If there are additional files or directories which should trigger an application restart, you can specify them with Spring.watch:

Spring.watch "config/some_config_file.yml"

By default, Spring polls the filesystem for changes once every 0.2 seconds. This method requires zero configuration, but if you find that it's using too much CPU, then you can use event-based file system listening by installing the spring-watcher-listen gem.

Quiet output

To disable the "Running via Spring preloader" message which is shown each time a command runs:

Spring.quiet = true

You can also set the initial state of the quiet configuration option to true by setting the SPRING_QUIET environment variable before executing Spring. This is useful if you want to set quiet mode when invoking the Spring executable in a subprocess, and cannot or prefer not to set it programmatically via the Spring.quiet option in ~/.spring.rb or the app's config/spring.rb.

Environment variables

The following environment variables are used by Spring:

Troubleshooting

If you want to get more information about what Spring is doing, you can run Spring explicitly in a separate terminal:

$ spring server

Logging output will be printed to stdout. You can also send log output to a file with the SPRING_LOG environment variable.