Wikipedia Search for Groonga family

It is a sample search system for Groonga family such as Groonga, Droonga, Mroonga and PGroonga. It searches Wikipedia data.

You can use the search system for the following proposes:

• Understanding how to implement Droonga system
• Running benchmark for PGroonga versus pg_bigm

Usage

For understanding how to implement Droonga system

TODO

For running benchmark for PGroonga versus pg_bigm

You need to set up clean CentOS environment. Run the following command lines to run benchmark for PGroonga versus pg_bigm:

% sudo yum install -y git
% git clone https://github.com/groonga/wikipedia-search.git
% cd wikipedia-search
% benchmark/centos6/pgroonga.sh |& tee benchmark.log

The benchmark script does the followings:

• Installs PostgreSQL.
• Installs PGroonga.
• Installs pg_bigm.
• Downloads Japanese Wikipedia data.
• Loads the Japanese Wikipedia data into PostgreSQL. (measured)
• Creates index by PGroonga. (measured)
• Creates index by pg_bigm. (measured)
• Search by some queries by PGroonga. (measured)
• Search by some queries by pg_bigm. (measured)

About chef cookbooks

If you use Ubuntu 12.04LTS, you need to install depending library via apt:

% sudo apt-get install libgecode-dev

Then, install "berks":

% cd chef
% bundle install --path vendor/bundle
% bundle exec berks

License

CC0 1.0 Universal. See LICENSE.txt for details.