FILTER SQL database

This repository contains scripts for converting the FILTER corpus in CSV format (provided in filter-data) to a MariaDB database, which is needed by tools such as Runoregi and Visualizations.

The computation is done in two stages. First, prepare the tables as text files in the data/ subdirectory by running:

DATA_DIR=/path/to/filter-data make all

Then, export the tables to the database. For exporting all tables, you can run:

DB_NAME=... make dbexport

The MySQL access needs to be configured using environment variables:

• DB_NAME -- name of the database (must already exist)
• MYSQL_GROUP_SUFFIX -- the section in my.cnf containing database access data (preferred way of passing them)

Alternatively, you can run the export manually table by table by executing the SQL files in the directory tables/ using your SQL client.

Note that the frequency tables v_clust_freq, p_clust_freq and word_freq need to be calculated at the end, i.e. after the respective source tables v_clust, p_clust, word_occ and verse_poem have been exported.