_____ __
/ ___/___ / /_ __ _____ __ ______
/ (_ // -_) __/ / // / _ | // / __/
\___/ \__/\__/ \_, /\___|_,_/_/
'-(+++++... /___/ ............ s++++++++.
.(++++++++++++( -+++~~~~~ 'N++++++++++= B++++++++s
'+++++++++++++++( DN=++++++< -NB+++++++++s Bz++++++++
+++++++++++++++++. 'NNN=++++++(-BNB+++++++++'BN========-
=B=+++++++++sDBBBBs<. +NNNN=+++++(<NNNBDBDBNNNNDBNNBBBBBBBz
'NN+s+++++=BBBh=+((+sB=DNNNNNBBBNNDhNNNN' .. -BNs (z
sN+N+z+++sBs- .hNNNNBNh(-' NNNN. .NB- <Ns +BBBz
D+N+N+z=Bs. .~. 'NNNNNB~ DNNN .NN+ (Ns =B+~
BN+N+NNB+ ~NBD BNNNBN h' hNNN .Bs< <Ns =Dz~
BNNNN+ND +BhB DNNNBz 'N~ sNNN 'ND' hNs -(z-
NNNNNNB( sDDN hNNNB( +N~ +NNN ..-<hBNs ~=h-
NNNNNNB- sDBN 'DNNNN. DN~ <BNN 'NNhs(NNs =NN-
NNNNNNB' zNNN(+sDNNBNNNz .NB( ~BNN 'D NNs =Bh~
NNNNNNN. zNNNBNBBNNNNNB( <NB~ 'BNB 'D <Bs =NNNN
NNNNNNN. zNNNNNNNNNNBNN. . .NDB -D hs N
NNNNNNN. zNNNNNhBNBN=+z +BN- NDB '(D <BhhhhhhhN
BNNNNNN. zNNNh=<~. h-s< D~D'.-~B<NBNBDs ''``
DNNNNNN. zNNN D-B-~(+B DNBDzs-`
sNNNNNN. zNNN B(NDhs=( ''`
(BNNNNN' zNDB ~D.
sNNNNB- +hsz h< O___
<NNNB( -=s' 'hz ___ ___ < s > \____
~BNh .+B+ / _ \/ _ \ = \ /
~Dz~. .~+zB=' \___/_//_/ / \ \__/
<hNNBNBh=(. / \ `
..
If
- You use Capistrano to deploy your application, and
- You have Rake tasks that you want to execute remotely — but you don’t want to hand-code Capistrano recipes for each Rake task —
Then
- You can use the Cape DSL within Capistrano recipes to dynamically add recipes for your application’s Rake tasks, and
- You can execute your Rake tasks on your deployed servers, friction-free, and look like a superhero. [cue fanfare]
- Mirror Rake tasks as Capistrano recipes, optionally filtered by namespace or name
- Embed Rake tasks in Capistrano namespaces
- Pass arguments to Rake tasks by setting environment variables with the same names
- Specify the Rake executable for local and remote Rake installations
- Enumerate Rake tasks for your own purposes
See what’s changed lately by reading the project history.
Install the RubyGem.
$ gem install cape
Or you may want to make Cape a dependency of your project by using Bundler.
# Gemfile
source 'http://rubygems.org'
gem 'cape', '~> 1'
Assume the following Rake tasks are defined.
desc 'Rakes the leaves'
task :leaves do
$stdout.puts "Raking the leaves"
end
desc 'Rakes and bags the leaves'
task :bag_leaves, [:paper_or_plastic] => :leaves do |task, arguments|
$stdout.puts "Putting the leaves in a #{arguments[:paper_or_plastic]} bag"
end
Rake lists these tasks in the expected fashion.
$ rake --tasks
rake bag_leaves[paper_or_plastic] # Rakes and bags the leaves
rake leaves # Rakes the leaves
$ rake --describe bag_leaves
rake bag_leaves[paper_or_plastic]
Rakes and bags the leaves
Add the following to your Capistrano recipes. Note that Cape statements must be contained in a Cape block.
# config/deploy.rb
require 'cape'
Cape do
# Create Capistrano recipes for all Rake tasks.
mirror_rake_tasks
end
Now all your Rake tasks appear alongside your Capistrano recipes.
$ cap --tasks
cap deploy # Deploys your project.
...
[other built-in Capistrano recipes]
...
cap bag_leaves # Rakes and bags the leaves.
cap leaves # Rakes the leaves.
Some tasks were not listed, either because they have no description,
or because they are only used internally by other tasks. To see all
tasks, type `cap -vT'.
Extended help may be available for these tasks.
Type `cap -e taskname' to view it.
Let’s use Capistrano to view the unabbreviated description of a Rake task recipe, including instructions for how to pass arguments to it. Note that Rake task parameters are automatically converted to environment variables.
$ cap --explain bag_leaves
------------------------------------------------------------
cap bag_leaves
------------------------------------------------------------
Bags the leaves.
Set environment variable PAPER_OR_PLASTIC if you want to pass a Rake task argument.
Here’s how to invoke a task/recipe with arguments. On the local computer, via Rake:
$ rake bag_leaves[plastic]
(in /current/working/directory)
Raking the leaves
Putting the leaves in a plastic bag
On remote computers, via Capistrano:
$ cap bag_leaves PAPER_OR_PLASTIC=plastic
* executing `bag_leaves'
* executing "cd /path/to/currently/deployed/version/of/your/app && /usr/bin/env `/usr/bin/env bundle check >/dev/null 2>&1; case $? in 0|1 ) echo bundle exec ;; esac` rake bag_leaves[plastic]"
servers: ["your.server.name"]
[your.server.name] executing command
** [out :: your.server.name] (in /path/to/currently/deployed/version/of/your/app)
** [out :: your.server.name] Raking the leaves
** [out :: your.server.name] Putting the leaves in a plastic bag
command finished in 1000ms
Cape lets you filter the Rake tasks to be mirrored. Note that Cape statements must be contained in a Cape block.
# config/deploy.rb
require 'cape'
Cape do
# Create Capistrano recipes for the Rake task 'foo' and/or for the tasks in
# the 'foo' namespace.
mirror_rake_tasks :foo
end
Cape lets you specify options used for defining Capistrano recipes. You can also specify remote environment variables to be set when executing Rake tasks. Note that Cape statements must be contained in a Cape block.
# config/deploy.rb
require 'cape'
Cape do
# Display defined Rails routes on application server remote machines only.
mirror_rake_tasks :routes, :roles => :app
# Execute database migration on application server remote machines only,
# and set the 'RAILS_ENV' environment variable to the value of the
# Capistrano variable 'rails_env'.
mirror_rake_tasks 'db:migrate', :roles => :app do |env|
env['RAILS_ENV'] = rails_env
end
end
The above is equivalent to the following manually-defined Capistrano recipes.
# config/deploy.rb
# These translations to Capistrano are just for illustration.
RAKE = '/usr/bin/env ' +
'`' +
'/usr/bin/env bundle check >/dev/null 2>&1; ' +
'case $? in ' +
# Exit code 0: bundle is defined and installed
# Exit code 1: bundle is defined but not installed
'0|1 ) ' +
'echo bundle exec ' +
';; ' +
'esac' +
'` ' +
'rake'
task :routes, :roles => :app do
run "cd #{current_path} && #{RAKE} routes"
end
namespace :db do
task :migrate, :roles => :app do
run "cd #{current_path} && #{RAKE} db:migrate RAILS_ENV=#{rails_env}"
end
end
Cape plays friendly with the Capistrano DSL for organizing Rake tasks in Capistrano namespaces. Note that Cape statements must be contained in a Cape block.
# config/deploy.rb
require 'cape'
namespace :rake_tasks do
# Use an argument with the Cape block, if you want to or need to.
Cape do |cape|
cape.mirror_rake_tasks
end
end
Cape lets you enumerate Rake tasks, optionally filtering them by task name or namespace. Note that Cape statements must be contained in a Cape block.
# config/deploy.rb
require 'cape'
Cape do
# Enumerate all Rake tasks.
each_rake_task do |t|
# Do something interesting with this hash:
# * t[:name] -- the full name of the task
# * t[:parameters] -- the names of task arguments
# * t[:description] -- documentation on the task, including parameters
end
# Enumerate the Rake task 'foo' and/or the tasks in the 'foo' namespace.
each_rake_task 'foo' do |t|
# ...
end
end
Cape lets you specify how Rake should be run on the local computer and on remote computers. But the default behavior is most likely just right for your needs:
- It detects whether Bundler is installed on the computer
- It detects whether the project uses Bundler to manage its dependencies
- It runs Rake via Bundler if the above conditions are true; otherwise, it runs Rake directly
Note that Cape statements must be contained in a Cape block.
# config/deploy.rb
require 'cape'
# Configure Cape never to run Rake via Bundler, neither locally nor remotely.
Cape.local_rake_executable = '/usr/bin/env rake'
Cape.remote_rake_executable = '/usr/bin/env rake'
Cape do
# Create Capistrano recipes for all Rake tasks.
mirror_rake_tasks
end
A Rake task that lacks a description can be mirrored or enumerated only if Rake v0.9.3 or newer is installed. Older versions of Rake did not support enumerating such tasks. You may want to make Rake v0.9.3 a dependency of your project.
# Gemfile
source 'http://rubygems.org'
gem 'rake', '>= 0.9.3'
A Rake task whose name collides with a Ruby method cannot be mirrored. For example, the name of Rails’s db:fixtures:load task collides with the Ruby Core Library’s Kernel::load method because that method is mixed into all objects. If you try to mirror db:fixtures:load, Capistrano will raise an exception. There is a questionable workaround for this.
A Rake task is always executed in the Capistrano deployment’s current_path. You may need to execute a task under release_path or another remote filesystem location, but this is not possible at present. Discuss this.
Report defects and feature requests on GitHub Issues.
Your patches are welcome, and you will receive attribution here for good stuff.
Released under the MIT License.