-ruby - code4lib · -ruby the best open source search engine + ruby rubyconf 2007 presented by:...

19
-ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher

Upload: others

Post on 14-Jun-2020

7 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

-rubythe best open source search engine + ruby

rubyconf 2007Presented by: Erik Hatcher

Page 2: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Solr

• Search server

• Enterprise scale (100M+ documents), very fast

• Open source: Apache Software License

• Java webapp, built on Lucene

• Features: caching, replication, faceting, highlighting, spell checking, admin interface, more...

• Very active community, evolving continuously

Page 3: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Solr Powered

• CNET

• Internet Archive

• Netflix

• Smithsonian

• digg

• AOL: sports and music channels

• more every day...

Page 4: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Lucene

• Java search engine library

• Created by Doug Cutting, renowned search engine expert

• Powers Technorati, IBM OmniFind Yahoo Edition, JIRA, Krugle, Nabble, Simpy, jGuru, Monster, Wikipedia, and many many more

Page 5: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Lucene

• Index Structure

• Documents

• Fields

• Terms

• Relevance:

Page 6: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Solr 101: add/updatePOST /solr/update

Page 7: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Solr 101: searchingGET /solr/select?q=ipod+AND+video&indent=on

Page 8: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Solr 101: searchingGET /solr/select?q=ipod+AND+video&indent=on&wt=ruby

Page 9: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

solr-ruby

• Ruby DSL

• HTTP communication to Solr

Page 10: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

solr-ruby powered

• acts_as_solr

• Flare

• Blacklight

• Collex

• ... your app!?

Page 11: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Mapper• Quack: #each

• Moo: #[]

Page 12: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

acts_as_solr

Page 13: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

acts_as_solr_“lite”

Page 14: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Flare

• Expose Solr through a Rails-based UI

• Constraint management

• Facet visualization

• Ajax suggest

• SIMILE integration: Timeline and Exhibit

• Status: nice for demos; needs work

Page 15: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

“ruby” on “rails”

Page 16: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Future

• Solr introspection

• schema/config savvy

• Support upcoming query components

• Fold in acts_as_solr

• Custom Solr response writer for full data type compatibility

• dates, ordered Hash

Page 17: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Help!• DSL/API guidance

• ActiveRecord integration

• Documentation

• Currently lame, I know. Sorry!

• Apache is new to Ruby

• gem server?

• Flare design

Page 18: -ruby - Code4Lib · -ruby the best open source search engine + ruby rubyconf 2007 Presented by: Erik Hatcher. Solr • Search server • Enterprise scale (100M+ documents), very fast

Coming soon...

• Contributed: solr-ruby, acts_as_solr, and Flare recipes