25

Is there any other way to search for a certain value in an integer[] column in Postgres?

My currently installed Postgres version does not allow the following statement:

SELECT * FROM table WHERE values *= 10;

Array examples:

'{11043,10859,10860,10710,10860,10877,10895,11251}'
'{11311,10698,10697,10710,10712,10711,10708}'

The statement should return every row where the array contains '10710'.

Erwin Brandstetter
  • 605,456
  • 145
  • 1,078
  • 1,228
jussi
  • 2,166
  • 1
  • 22
  • 37

3 Answers3

59

For equality checks you can simply:

SELECT * FROM tbl WHERE 10 = ANY (values);

Read about ANY/SOME in the manual.
For big tables / big arrays consider array operators with index support, like:

SELECT * FROM tbl WHERE values @> '{10}'::int[];

See:

For integer arrays consider the additional module intarray:

Related:

Erwin Brandstetter
  • 605,456
  • 145
  • 1,078
  • 1,228
7

quickly search will be so, but you should use index gist or gin for intarray type Postgres intarray

 SELECT * FROM table WHERE values @> ARRAY[10];
anydasa
  • 2,616
  • 1
  • 15
  • 13
-2
**Store Integer Array as Strings in Postgresql and Query the Array**    
Finally I could save the integer as string array in one column able to successfully convert into array and query the array using below example.

    CREATE TABLE test
    (
      year character varying,
      id serial NOT NULL,
      category_id character varying,
      CONSTRAINT test_pkey PRIMARY KEY (id)
    )

    Data
    "2005";1;"1,2,3,4"
    "2006";2;"2,3,5,6"
    "2006";3;"4,3,5,6"
    "2007";7;"1,2"


    select distinct(id) from test, (select id as cid, unnest(string_to_array(category_id ,  ',')::integer[]) as cat from test) c where c.cid=test.id and cat in (1,2,3);

    Result:
    2
    1
    3
    7