From: Tom Lane <tgl@sss.pgh.pa.us>
To: Teju Jakkidi vlogs <teja.jakkidi05@gmail.com>
cc: pgsql-admin <pgsql-admin@lists.postgresql.org>
Subject: Re: Query taking long with levenshtein function
In-reply-to: 
 <CAKA2XvZ_zFgfnavkEDMvfLtHzQeH-3qGROfAp+=CnrLF8TTvyQ@mail.gmail.com>
References: 
 <CAKA2XvZ_zFgfnavkEDMvfLtHzQeH-3qGROfAp+=CnrLF8TTvyQ@mail.gmail.com>
Comments: In-reply-to Teju Jakkidi vlogs <teja.jakkidi05@gmail.com>
	message dated "Tue, 28 Feb 2023 15:11:55 -0800"
MIME-Version: 1.0
Content-Type: text/plain; charset="us-ascii"
Content-ID: <1278257.1677627051.1@sss.pgh.pa.us>
Date: Tue, 28 Feb 2023 18:30:51 -0500
Message-ID: <1278258.1677627051@sss.pgh.pa.us>
Archived-At: 
 <https://www.postgresql.org/message-id/1278258.1677627051%40sss.pgh.pa.us>
Precedence: bulk

Teju Jakkidi vlogs <teja.jakkidi05@gmail.com> writes:
> We have a query as below which uses the levenshtein function to compare
> strings.

> SELECT "name", levenshtein("name",'some string') as p FROM table1
> where levenshtein("name",'some string')   <= 2 order by p desc;

> We have a GIST index built on top of this table as below:
> CREATE INDEX gist_idx ON table1 USING GIST("name");

AFAIK, that index is completely useless for this query.  I don't
really see a good way to index it either --- "levenshtein distance
less than X" seems like a not very tractable requirement.  Can
you formulate your matching rules some other way?

			regards, tom lane