CODEWITHSUNDEEP
postgresqlfull-text-searchexact-match
Mike.S
Mike.S

Reputation: 47

exact word fulltext search postgresql

Postgresql fulltext search match nearby words, but i want to search the exact word in large content table (thousands of words in hundreds of thousands of lines). My search must be very fast (less than a second). Using like or ilike is to slow( with 200000 words: more than 5 seconds). Does anyone have a query to advise me?

Upvotes: 2

Views: 2390

Answers (1)

Laurenz Albe
Laurenz Albe

Reputation: 247445

You should be able to solve your problem with PostgreSQL full text search if you use the simple dictionary and create an appropriate GIN index:

CREATE TABLE haystack (id serial PRIMARY KEY, string text NOT NULL);
INSERT INTO haystack (string) VALUES ('I am your servant');
INSERT INTO haystack (string) VALUES ('What use is a server without a client?');

CREATE INDEX haystack_fts_ind ON haystack USING gin (to_tsvector('simple', string));

Let's disable sequential scans so the index is used, even though the example table is too small:

SET enable_seqscan=off;

Now only exact matches are found, and no stemming takes place:

SELECT * FROM haystack
WHERE to_tsvector('simple', string) @@ to_tsquery('simple', 'servant');

 id |      string       
----+-------------------
  1 | I am your servant
(1 row)

The index can be used to speed up the query:

EXPLAIN (COSTS off) SELECT * FROM haystack
WHERE to_tsvector('simple', string) @@ to_tsquery('simple', 'servant');

                                        QUERY PLAN                                        
------------------------------------------------------------------------------------------
 Bitmap Heap Scan on haystack
   Recheck Cond: (to_tsvector('simple'::regconfig, string) @@ '''servant'''::tsquery)
   ->  Bitmap Index Scan on haystack_fts_ind
         Index Cond: (to_tsvector('simple'::regconfig, string) @@ '''servant'''::tsquery)
(4 rows)

Upvotes: 5

Related Questions