219

How can I get a list of column names and datatypes of a table in PostgreSQL using a query?

1
  • plsql \gdesc is also an option. The neat feature is it could describe entire queries(useful for complex adhoc ones) but it works for simplified scenario SELECT * FROM tab_name \gdesc as well. Commented Nov 21, 2020 at 10:02

16 Answers 16

299
SELECT
    column_name,
    data_type
FROM
    information_schema.columns
WHERE
    table_name = 'table_name';

with the above query you can retrieve columns and its datatype.

5
  • 9
    That won't give the right answer for user-defined types (e.g., Geometry and Geography columns created by ogr2ogr, which are of the form geometry(Geometry,[SRID])).
    – GT.
    Commented Jul 13, 2016 at 4:51
  • 4
    One might also use table_catalog = 'my_database' and table_schema = 'my_schema' in order to get only columns from a specific table of a specific schema of a specific database. Commented Mar 24, 2017 at 16:46
  • 3
    May I suggest to everyone, if you are looking to build on this code. Use the pg_catalog and not the information_schema. The information_schema has some pretty easy and universal SQL, however it is slower because it is higher level. Commented Dec 22, 2020 at 14:51
  • @DanielL.VanDenBosch however, pg_catalog is subject to breaking changes between postgres versions.
    – OrangeDog
    Commented Jul 6, 2023 at 16:21
  • 2
    @MarcoMannes or table_schema = current_schema(). For postgres the information_schema only contains data for the current database, so you don't have to worry about that.
    – OrangeDog
    Commented Jul 6, 2023 at 16:34
249

Open psql command line and type :

\d+ table_name
5
  • 4
    I don't understand why this isn't the most upvoted answer.
    – mjezzi
    Commented Apr 12, 2018 at 20:18
  • 46
    This is incomplete as OP may want to do this programmatically in SQL code, not just via psql.
    – Luke
    Commented Aug 7, 2018 at 10:42
  • 4
    Postgres does it programmatically, so just start postgres with the '-E' flag: psql -E and for every backslash command the respective SQL will be displayed before the result of the command.
    – karatedog
    Commented Aug 11, 2020 at 7:38
  • 5
    Presuming OP has access to psql makes this answer a little out of scope. Having postgres does not presume proficiency or the ability to access psql. Commented Feb 18, 2021 at 4:17
  • I was initially dismissive of this answer, but surprisingly, this worked out the best for me when creating a dblink. I just copied and pasted the output into a spreadsheet, split by | got the first two columns, added a comma column and a ); at the end and we have a dblink query.
    – nurettin
    Commented Mar 18, 2022 at 8:54
37
SELECT
        a.attname as "Column",
        pg_catalog.format_type(a.atttypid, a.atttypmod) as "Datatype"
    FROM
        pg_catalog.pg_attribute a
    WHERE
        a.attnum > 0
        AND NOT a.attisdropped
        AND a.attrelid = (
            SELECT c.oid
            FROM pg_catalog.pg_class c
                LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace
            WHERE c.relname ~ '^(hello world)$'
                AND pg_catalog.pg_table_is_visible(c.oid)
        );

Change the hello world with your table name

More info on it : http://www.postgresql.org/docs/9.3/static/catalog-pg-attribute.html

1
  • 4
    Works, but why are you using c.relname ~ '^(hello world)$ instead of simply c.relname = 'hello world'?
    – Thomas
    Commented Apr 1, 2019 at 12:26
27

Don't forget to add the schema name in case you have multiple schemas with the same table names.

SELECT column_name, data_type 
FROM information_schema.columns
WHERE table_name = 'your_table_name' AND table_schema = 'your_schema_name';

or using psql:

\d+ your_schema_name.your_table_name
27

A version that supports finding the column names and types of a table in a specific schema, and uses JOINs without any subqueries

SELECT
    pg_attribute.attname AS column_name,
    pg_catalog.format_type(pg_attribute.atttypid, pg_attribute.atttypmod) AS data_type
FROM
    pg_catalog.pg_attribute
INNER JOIN
    pg_catalog.pg_class ON pg_class.oid = pg_attribute.attrelid
INNER JOIN
    pg_catalog.pg_namespace ON pg_namespace.oid = pg_class.relnamespace
WHERE
    pg_attribute.attnum > 0
    AND NOT pg_attribute.attisdropped
    AND pg_namespace.nspname = 'my_schema'
    AND pg_class.relname = 'my_table'
ORDER BY
    attnum ASC;
1
  • 4
    This is my favorite answer because it accomplishes 2 things. It uses the pg_catalog and it forces you specify the schema. If you are cramming everything thing in the public schema, I personally believe that is a bad strategy. As your project grows, it will be difficult to keep things organized. IMHO Commented Dec 22, 2020 at 15:23
9

Updated Pratik answer to support more schemas and nullables:

SELECT
    "pg_attribute".attname                                                    as "Column",
    pg_catalog.format_type("pg_attribute".atttypid, "pg_attribute".atttypmod) as "Datatype",

    not("pg_attribute".attnotnull) AS "Nullable"
FROM
    pg_catalog.pg_attribute "pg_attribute"
WHERE
    "pg_attribute".attnum > 0
    AND NOT "pg_attribute".attisdropped
    AND "pg_attribute".attrelid = (
        SELECT "pg_class".oid
        FROM pg_catalog.pg_class "pg_class"
            LEFT JOIN pg_catalog.pg_namespace "pg_namespace" ON "pg_namespace".oid = "pg_class".relnamespace
        WHERE
            "pg_namespace".nspname = 'schema'
            AND "pg_class".relname = 'table'
    );
7
SELECT column_name,data_type 
FROM information_schema.columns 
WHERE
table_name = 'your_table_name' 
AND table_catalog = 'your_database_name' 
AND table_schema = 'your_schema_name';
1
  • While this code may answer the question, providing additional context regarding how and/or why it solves the problem would improve the answer's long-term value. Commented Nov 21, 2020 at 10:16
4
 --how to get a list column names and datatypes of a table in PostgreSQL?


    SELECT DISTINCT
        ROW_NUMBER () OVER (ORDER BY pgc.relname , a.attnum) as rowid , 
        pgc.relname as table_name ,
        a.attnum as attr,
        a.attname as name,
        format_type(a.atttypid, a.atttypmod) as typ,
        a.attnotnull as notnull, 
        com.description as comment,
        coalesce(i.indisprimary,false) as primary_key,
        def.adsrc as default
    FROM pg_attribute a 
    JOIN pg_class pgc ON pgc.oid = a.attrelid
    LEFT JOIN pg_index i ON 
        (pgc.oid = i.indrelid AND i.indkey[0] = a.attnum)
    LEFT JOIN pg_description com on 
        (pgc.oid = com.objoid AND a.attnum = com.objsubid)
    LEFT JOIN pg_attrdef def ON 
        (a.attrelid = def.adrelid AND a.attnum = def.adnum)
    LEFT JOIN pg_catalog.pg_namespace n ON n.oid = pgc.relnamespace

    WHERE 1=1 
        AND pgc.relkind IN ('r','')
        AND n.nspname <> 'pg_catalog'
        AND n.nspname <> 'information_schema'
        AND n.nspname !~ '^pg_toast'

    AND a.attnum > 0 AND pgc.oid = a.attrelid
    AND pg_table_is_visible(pgc.oid)
    AND NOT a.attisdropped
    ORDER BY rowid
    ;
1
  • 1
    Please include context with your code. Also, how do you expand this to other schemas? Commented Dec 22, 2020 at 15:31
3

To make this topic 'more complete'.

I required the column names and data types on a SELECT statement (not a table).

If you want to do this on a SELECT statement instead of an actual existing table, you can do the following:

DROP TABLE IF EXISTS abc;
CREATE TEMPORARY TABLE abc AS
-- your select statement here!
SELECT 
    *
FROM foo
-- end your select statement
;

select column_name, data_type 
from information_schema.columns 
where table_name = 'abc';
DROP IF EXISTS abc;

Short explanation, it makes a (temp) table of your select statement, which you can 'call' upon via the query provided by (among others) @a_horse_with_no_name and @selva.

Hope this helps.

3

I was looking for a way to get column names with data types without PSQL cl, directly from pgAdmin 4, and found a workaround. Adding one more option:

right-click desired database > generate ERD(Beta) > Generate SQL(or Alt+Ctrl+S) and pgAdmin 4 will open Query Editor where you can find all tables with column names and data types: Query Editor

0
2

without mentioning schema also you can get the required details Try this query->

select column_name,data_type from information_schema.columns where table_name = 'table_name';

1
query = "SELECT column_name FROM information_schema.columns WHERE table_schema = 'schema_name' AND table_name = 'table_name'"
qry = execute(query)
print(qry)
3
  • This does not provide an answer to the question. To critique or request clarification from an author, leave a comment below their post. - From Review Commented Jul 10, 2023 at 21:47
  • 1
    @XMehdi01: Why doesn’t it answer the question? It looks like an answer to me. Is this due to a technical issue? E.g., do you believe the answer is incorrect? Because, otherwise, it’s still proposing a solution. Commented Jul 23, 2023 at 0:49
  • 1
    That said, @kamran-kausar, this would be a much more valuable answer if you explained how your approach is different from the validated answers, under what circumstances your approach might be preferred, and/or why you think the previous answers aren’t sufficient. Can you kindly edit your answer to offer an explanation? Commented Jul 23, 2023 at 0:56
0

Below will list all the distinct data types of all the table in the provided schema name.

\copy (select distinct data_type, column_name from information_schema.columns where table_name in (SELECT tablename FROM pg_catalog.pg_tables WHERE schemaname != 'pg_catalog' AND schemaname != 'information_schema' and schemaname = '<Your schema name>')) to 'datatypes.csv'  delimiter as ',' CSV header
0

create a new function to get the table info

CREATE FUNCTION xdesc(in t varchar) RETURNS table(column_name varchar, 
data_type varchar) AS $$
  SELECT column_name, data_type 
  FROM information_schema.columns
WHERE table_name = $1 
$$ LANGUAGE SQL;


select xdesc('rs_rail_job_index')
0

I found @Pratik's answer to be the thing, but it didn't work for me for some reason. I've rewritten it in a more consise, if I may, manner. It's easier to understand, it supports schemas and actually works for me (postgres 15). No joins.

SELECT
    attname AS colname,
    pg_catalog.format_type(atttypid, atttypmod) AS coltype
FROM
    pg_catalog.pg_attribute
WHERE
    attnum > 0 AND
    NOT attisdropped AND
    attrelid = 'your table name'::regclass
ORDER BY
    attnum ASC
;

P.S. It also supports composite types (which I actually needed very much). It would return a string like this: schemaname.tablename.

0

Here's a version that just queries one table by leveraging the slightly magical ::regclass cast (which I have found is often more performant that joining on pg_class and pg_namespace), has no subqueries, and avoids information_schema (which I have found is often slower, so just as rule, I would avoid if you can)

SELECT
    pg_attribute.attname AS column_name,
    pg_catalog.format_type(pg_attribute.atttypid, pg_attribute.atttypmod) AS data_type
FROM
    pg_catalog.pg_attribute
WHERE
    pg_attribute.attrelid = 'my_schema.my_table'::regclass
    AND attnum > 0 
    AND NOT pg_attribute.attisdropped
ORDER BY
    attnum ASC;

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.