postgres-tsearch-bulgarian

This repository contains Bulgarian ispell (affix and dict) and stopword dictionaries for full text search in PostgreSQL.

Credits

ISpell dictionary

The ispell dictionary files (bulgarian.affix and bulgarian.dict) have been created by the bgOffice/БГ Офис project for use in OpenOffice and are licensed under LGPL 3.0.

This repository contains a modified version of those files (minor changes) to make them compatible with the format expected by PostgreSQL. The original ispell files (bulgarian.aff and bulgarian.dic) can be downloaded from http://bgoffice.sourceforge.net/ispell/index.html

Stop words

The stop words list used in this repository (bulgarian.stop) is a modified version of the list published in article "Searching strategies for the Bulgarian language" (the list is in Table A.1) by Prof. Jacques Savoy.

How to install

1. Copy the three files bulgarian.affix, bulgarian.dict and bulgarian.stop to your $SHAREDIR/tsearch_data/ directory (eg. C:\Program Files\PostgreSQL\12\share\tsearch_data). You can determine what your $SHAREDIR is by running pg_config --sharedir.

2. Execute the following SQL script:

   CREATE TEXT SEARCH CONFIGURATION bulgarian (COPY = simple);
   CREATE TEXT SEARCH DICTIONARY bulgarian_ispell (
       TEMPLATE = ispell,
       DictFile = bulgarian, 
       AffFile = bulgarian, 
       StopWords = bulgarian
   );
   CREATE TEXT SEARCH DICTIONARY bulgarian_simple (
       TEMPLATE = pg_catalog.simple,
       STOPWORDS = bulgarian
   );
   ALTER TEXT SEARCH CONFIGURATION bulgarian ALTER MAPPING FOR asciiword, asciihword, hword, hword_part, word WITH bulgarian_ispell, bulgarian_simple;
   

3. Make sure its working by running a full text search query.

   A query like this one:

   SELECT to_tsvector('bulgarian', 'текстовете');
   

   should output only the base of the word (текст):

   `"'текст':1"`