Blacklight powered by LucidWorks Search
Ruby JavaScript CSS Shell
Latest commit dc98f98 Jul 12, 2014 @jamieorc jamieorc FOCUS-5621: assets now server correctly
* jruby-rack 1.1.14 was the problem
* pinned us to jruby-rack

Blacklight powered by LucidWorks Search

LucidWorks Search

LucidWorks Search, powered by Apache Solr, provides a commercially supported and easily manageable search engine environment. Data sources include web and file crawling, database indexing, and much more. It is easy to get the supported data sources configured and feeding content into the system.

Blacklight: What is it and why use it?

In its own words, from Project Blacklight...

Blacklight is an open source Solr user interface discovery platform. You can use Blacklight to enable searching and browsing of your collections. Blacklight uses the Apache Solr search engine to search full text and/or metadata. Blacklight has a highly configurable Ruby on Rails front-end. Blacklight was originally developed at the University of Virginia Library and is made public under an Apache 2.0 license.

LucidWorks Search provides a search UI meant for administrative purposes, but is not designed or recommended for production use by end-users. Search applications generally demand a middle tier, an application logic layer, between Solr and the client presentation layer. Blacklight provides Solr search and presentation services, an attractive and sophisticated customizable user interface, and a vibrant, supportive open source community. Blacklight has been deployed in production environments for years in many sites around the world.

This project delivers the necessary adjustments and customizations needed on a default Blacklight application enhanced to leverage the value added capabilities of the LucidWorks Search platform.


  1. Install LucidWorks Search
  2. Install Flare
  3. Run Flare

Use this to add the LucidWorks website as a web data source to your starting collection:

curl -H 'Content-type: application/json' -d '{"crawler":"lucid.aperture", "type":"web","url":"","crawl_depth":2, "name":"LucidWebsite"}' http://localhost:8888/api/collections/collection1/datasources

And use this to start the crawl process for your starting collection:

curl -X PUT 'http://localhost:8888/api/collections/collection1/datasources/all/job'

Running pre-built app

If you've got a pre-built lw_flare.war, you can run it using:

java -jar lw_flare.war

The current app configuration is set to use http://localhost:8888/solr/collection1 as the LucidWorks Search Solr server endpoint.

When running the executable web application, it runs on port 8080, and is accessible at http://localhost:8080

NOTE: The first request to the application can be quite slow currently (we'll explore ways to speed that up), so be patient. Successive requests to the same application will be spiffy.

For developers

* git clone github/LucidWorks/flare

* flare git:(master) ✗ java -version
java version "1.7.0_13"
Java(TM) SE Runtime Environment (build 1.7.0_13-b20)
Java HotSpot(TM) 64-Bit Server VM (build 23.7-b01, mixed mode)
* flare git:(master) ✗ export PATH=$HOME/bin/jruby-1.7.4/bin:$PATH
* flare git:(master) ✗ jruby -v
jruby 1.7.4 (1.9.3p392) 2013-05-16 2390d3b on Java HotSpot(TM) 64-Bit Server VM 1.7.0_13-b20 [darwin-x86_64]
* flare git:(master) ✗ jruby -S rails server  
=> Booting WEBrick
=> Rails 3.2.13 application starting in development on
=> Call with -d to detach
=> Ctrl-C to shutdown server
[2013-05-31 14:26:44] INFO  WEBrick 1.3.1
[2013-05-31 14:26:46] INFO  ruby 1.9.3 (2013-05-16) [java]
[2013-05-31 14:26:46] INFO  WEBrick::HTTPServer#start: pid=744 port=3000  

Environment variables

This app will use the following environment variables to access LWS core API, user API, and Solr itself, falling back to the defaults specified:

  • LWS_USER_API_BASE: defaults to (uses /api/users/#{username} from the base)

  • LWS_CORE_URL: defaults to

  • LWS_API_URL: defaults to #{LWS_CORE_URL}/api, this is the base URL used for all LWS core (non-user) api, such as getting a list of available collections

  • LWS_SOLR_URL: defaults to #{LWS_CORE_URL}/solr (collections are accessed from this Solr base url, such as /collection1/lucid?q=:...)

For example, LWS developers may run the LWS (lwe-ui) admin UI application using the default Rails port of 3000, so lw_flare can be run like this:

LWS_USER_API_BASE= jruby -S rails s -p 8080