My Experiments with Sphinx

Objective: The main objective of this deliverable was to perform experiments with Sphinx. The experiments involved the study of how Sphinx works and also indexing databases using Sphinx.

Description:

Introduction to Sphinx

Sphinx is an open source full text search server, designed from the ground up with performance, relevance, and integration simplicity in mind. It is written in C++ and works on Linux, Windows, MacOS, Solaris, FreeBSD, and a few other systems. Sphinx allows to either batch index and search data stored in an SQL database or just files quickly and easily [1].

Experiments

In this deliverable, two tasks were accomplished using Sphinx. The two tasks were:

1. Index and search data from the sample database provided by Sphinx.
2. Index and search data from a newly created database.

Firstly, indexing is done on the sample database provided by Sphinx. For this purpose, the following steps are followed.

• Sphinx was downloaded and unzipped into the C:/ drive.
• A data folder was created in the Sphinx folder.
• The configuration file was present in the Sphinx folder.
• All the necessary changes were made in the configuration file to match the path suitable for the system.
• The command prompt was opened.
• Sphinx is installed as a window service using the following command:

C:\Sphinx\bin> C:\Sphinx\bin\searchd --install --config C:\Sphinx\sphinx.conf.in --searchd1 SphinxSearch

• A sample data file example.sql is provided by Sphinx to populate the documents table with a few documents for testing purposes.
• Documents table is present in the test database.
• Import the example.sql file.
• A full-text index of the data was created by running the indexer.
• Search was done on the keyword test using the search utility in the bin folder.

The experiments done can be seen in the following figures:

Figure1: Importing the database and running the indexer

Figure2: Search for keyword using search utility

Secondly, a full-text index was created for my own database and a search is done on few keywords. For this purpose, the following steps were followed:

• A new database testSphinx is created using mysql.
• A table named details is created in the database.
• A .sql file is created for the database.
• The sphinx configuration file is changed according to the newly created database.
• The main changes made were changing the sql database, changing the field names in the query and changing the index names.
• A document id field should be the first field in any new database in order to identify the document.
• Indexing is done on the newly created database.
• Search was done with two keywords: details and tanmayee.

The experiments done can be seen in the following figures:

Figure3: Creation of a new database and .sql file

Figure4: Details table

Figure5: Importing testSphinx database and indexing it

Figure6: Search results for the keyword "details"

Figure7: Search results for the keyword "details"(cont…)

Figure8: Search results for the keyword "tanmayee"

References