Curate is a Rails engine leveraging ProjectHydra and ProjectBlacklight components to deliver a foundation for an Institutional Repositories. It is released under the Apache 2 License

• Starting a new Curate-based Rails application
  • or install by hand
• Developing and contributing for the Curate gem
  • Prerequisites
  • Clone the Curate repository
  • Jetty
  • Running the specs
    • All of them
    • Some of them
    • With Zeus
  • Contributing back
• Working on Curate while working on my Application
• Integration with CAS, LDAP, Shibboleth
• Standing up your Curate-based Rails application in Production

When you generate your new Rails application, you can use Curate's application template:

$ rails new my_curate_application -m https://raw.github.com/projecthydra/curate/master/lib/generators/curate/application_template.rb

Add the following line to your application's Gemfile:

gem 'curate'

And then execute:

$ bundle
$ rails generate curate

You may need to have the following installed ‡

• imagemagick (http://www.imagemagick.org/script/index.php)
• fits.sh (https://code.google.com/p/fits/wiki/installing)
• ffmpeg (optional, http://www.ffmpeg.org/)
• Redis (http://redis.io/)
• rubygems (http://rubygems.org/pages/download)
• ClamAV (http://www.clamav.net/)

‡ - Why "you may need"? Some of these are only optionally used in development and tests; But production will need it.

From the command line:

git clone https://github.com/projecthydra/curate.git ./path/to/local

Curate uses Jetty for development and testing. You could configure it to use an alternate Fedora and SOLR location, but that is an exercise for the reader. The hydra-jetty package should not be use for secure production installations

Install jetty, you should only need to do this once (unless you remove the ./jetty directory)

$ rake jetty:unzip

Inside the Curate directory (i.e. ./path/to/local):

1. Make sure jetty is running (rake jetty:start); It will take a bit to spin up jetty.
2. Make sure you have a dummy app ‡
3. Run rake regenerate to build a clean app in ./spec/dummy
4. Then run rake spec; The tests will take quite a while ‡‡

‡ - A Rails engine requires a Rails application to run. The dummy app is a generated application inside Curate in the ./spec/internal directory ‡‡ - Slow tests are a big problem and we are working on speeding them up, but its complicated.

In some cases you want to know the results of a single test. Here's how you do it.

1. Make sure jetty is running (rake jetty:start); It will take a bit to spin up jetty.
2. Make sure you have a dummy app ‡
3. Run rake regenerate to build a clean app in ./spec/dummy
4. Then run BUNDLE_GEMFILE=spec/internal/Gemfile bundle exec rspec path/to/spec.rb:LINE ‡

‡ - With Curate being an Engine we need to point to the Dummy Application's Gemfile. In full Rails applications you can normally run the following rspec path/to/spec.rb:LINE

Zeus pre-loads your Rails app so that your normal development tasks such as console, server, generate, and specs/tests take less than one second. ‡

1. In terminal window #1 run zeus start
2. In terminal window #2, once Zeus is started: run zeus rake spec for all tests; or zeus test path/to/spec.rb:LINE for one

‡ - Loading the environment to run your tests takes less than a second. So running an individual test will take less time.

If you are interested in helping us make Curate better, please take a look at our Contributing resources and guidelines.

Assuming you are wanting to work on your Curate-based application and make modifications to the Curate gem, follow these instructions.

Replace the following line in the Gemfile of your Curate-based application (see Starting a New Curate Based Application):

gem 'curate' ...

with

gem 'curate', path: './path/to/my/clone/of/curate'

More information about Gemfile management at Bundler.io

You can then do concurrent development on both your clone of the Curate gem and your Curate-based application.

NOTE: Any changes you make in the Curate gem will likely require you to restart your web-server.

For integration with CAS, refer: https://github.com/nbudin/devise_cas_authenticatable

For integration with LDAP, refer: https://github.com/cschiewek/devise_ldap_authenticatable

For integration with Shibboleth, refer: https://github.com/jgeorge300/devise_shibboleth_authenticatable

For integration with multiple authentication systems (and for twitter, facebook, etc.), consider using OmniAuth: https://github.com/intridea/omniauth

We are working on this and have more to come.