This is the codebase for the SearchWorks redesign.
You'll need common dependencies for building rails applications such as a javascript runtime (e.g. v8). If you don't want to have the mysql gem installed for local development you can run the bundle install command below with the --without production deployment flag
After cloning the repository
$ bundle install
$ rake searchworks:install
The installation script will
- Migrate the database
You will need to update the configuration in config/settings.yml for various parts of the app to work. Please check that file for more information.
Start Solr
To start Solr, you can use the solr_wrapper command. However, if starting from a fresh instance, you may first need to run rake searchworks:install or rake ci so that the CJK tokenizer gets copied into the appropriate directory. There is also a separate task (rake searchworks:copy_solr_dependencies) available if you find that you need to clean solr and the CJK tokenizer is getting removed.
$ solr_wrapper
Start the rails app
$ rails s
Given that this app is using shibboleth + devise for login, it can be tricky to get a user context set. The user itself can be set by starting rails with a REMOTE_USER env var set (e.g. REMOTE_USER=jstanford rails s).
If you need to set your affiliation attribute (e.g. stanford:staff) for things like testing article search, you can start rails with a suAffiliation env var set (e.g. REMOTE_USER=jstanford suAffiliation=stanford:staff rails s).
One caveat to this is that you may still need to go through the login path /webauth/login in order to ensure proper session setup (e.g. eds guest flag set).
There are two testing tasks: rake ci and rake jenkins
This is intended for running tests against the fixtures in the local index.
$ rake ci
This is intended for running production data integration tests against the remote index.
$ TEST_SOLR_URL=http://example-solr.stanford.edu:8983/solr rake jenkins