1

The default unaccent module has no or little support for Greek Letters using the unaccent module as seen in following links:

Also this is my problem in my case as well. So I wanted to load custom unaccent rules as seen in https://www.postgresql.org/docs/current/unaccent.html

So I wanted to know:

  • In the example CREATE TEXT SEARCH CONFIGURATION fr ( COPY = french ); where the ( COPY = french ) is located?

  • Can be the copy statement ( COPY = french ) use data from the STDIN with my own rules?

  • Also can I use a dedicated database table in order to load my own unaccent rules as well?

The database is hosted into an Amazon RDS instance.

1

In the example CREATE TEXT SEARCH CONFIGURATION fr ( COPY = french ); where the ( COPY = french ) is located?

Text search configurations are stored in pg_catalog.pg_ts_config.

They can also be displayed with \dF[+] in psql.

Can be the copy statement ( COPY = french ) use data from the STDIN with my own rules?

No. The COPY clause in CREATE TEXT SEARCH CONFIGURATION clones a text configuration and has nothing to do (besides the name) with the SQL COPY command meant to import data from the outside of the database.

Also can I use a dedicated database table in order to load my own unaccent rules as well?

No.

The latest version of PostgreSQL (12) comes with unaccent rules that correctly support Greek (as far as I understand) solving the issue #15347 mentioned in the question. For on-premise PostgreSQL instances, DBAs can get the latest unaccent.rules and copy it into an older PostgreSQL version, and making sure that the indexes that depend on unaccent get reindexed. With instances managed by a third party you would have to ask them.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.