How to use dynamic column names in an UPDATE or SELECT statement in a function?

Question

In PostgreSQL 9.1, PL/pgSQL, given a query:

select fk_list.relname from ...

where relname is of type name (e.g., "table_name").

How do you get the appropriate value for "relname" that can be used directly in an UPDATE statement as:

Update <relname> set ...

within the PL/pgSQL script?

Using quote_ident(r.relname) as:

Update quote_ident(r.relname) Set ...

fails with:

syntax error at or near "(" LINE 55: UPDATE quote_ident(r.relname) ....

The complete code I am working with:

CREATE FUNCTION merge_children_of_icd9 (ocicd9 text, 
                      ocdesc text, ncicd9 text, ncdesc text)
RETURNS void AS $BODY$
DECLARE
  r RECORD;
BEGIN
  FOR r IN
    WITH fk_actions ( code, action ) AS (
      VALUES ('a', 'error'), 
             ('r', 'restrict'), 
             ('c', 'cascade'),
             ('n', 'set null'),
             ('d', 'set default')
    ), fk_list AS (
      SELECT pg_constraint.oid AS fkoid, conrelid, confrelid::regclass AS parentid,
             conname, relname, nspname,
             fk_actions_update.action AS update_action,
             fk_actions_delete.action AS delete_action,
             conkey AS key_cols
      FROM pg_constraint
      JOIN pg_class ON conrelid = pg_class.oid
      JOIN pg_namespace ON pg_class.relnamespace = pg_namespace.oid
      JOIN fk_actions AS fk_actions_update ON confupdtype = fk_actions_update.code
      JOIN fk_actions AS fk_actions_delete ON confdeltype = fk_actions_delete.code
      WHERE contype = 'f'
    ), fk_attributes AS (
      SELECT fkoid, conrelid, attname, attnum
      FROM fk_list
      JOIN pg_attribute ON conrelid = attrelid AND attnum = ANY(key_cols)
      ORDER BY fkoid, attnum
    ), fk_cols_list AS (
      SELECT fkoid, array_agg(attname) AS cols_list
      FROM fk_attributes
      GROUP BY fkoid
    )
    SELECT fk_list.fkoid, fk_list.conrelid, fk_list.parentid, fk_list.conname, 
           fk_list.relname, fk_cols_list.cols_list
    FROM fk_list
    JOIN fk_cols_list USING (fkoid)
    WHERE parentid = 'icd9'::regclass
  LOOP
    RAISE NOTICE 'now in loop. relname is %', quote_ident(r.relname);
    RAISE NOTICE 'cols_list[1] is %', quote_ident(r.cols_list[1]);
    RAISE NOTICE 'cols_list[2] is %', quote_ident(r.cols_list[2]);
    RAISE NOTICE 'now doing update';

    UPDATE quote_ident(r.relname) SET r.cols_list[1] = ncicd9, r.cols_list[2] = ncdesc
    WHERE r.cols_list[1] = ocicd9 AND r.cols_list[2] = ocdesc;

    RAISE NOTICE 'finished update'; 
  END LOOP;     
  RETURN;
END $BODY$ LANGUAGE plpgsql VOLATILE;   

-- select merge_children_of_icd9('', 'aodm type 2', '', 'aodm, type 2'); 

I'm sure this kind of thing is done often, but I can't seem to find anything like it using PostgreSQL. Is there a better way?

Answer 1

In an UPDATE statement in PL/pgSQL, the table name has to be given as a literal. If you want to dynamically set the table name and the columns, you should use the EXECUTE command and paste the query string together:

EXECUTE 'UPDATE ' || quote_ident(r.relname) ||
       ' SET ' || quote_ident(r.cols_list[1]) || ' = $1, ' || 
                  quote_ident(r.cols_list[2]) || ' = $2' ||
       ' WHERE ' || quote_ident(r.cols_list[1]) || ' = $3 AND ' ||
                    quote_ident(r.cols_list[2]) || ' = $4'
USING ncicd9, ncdesc, ocicd9, ocdesc;

The USING clause can only be used for substituting data values, as shown above.

Answer 2

You need dynamic SQL with EXECUTE like @Patrick already provided. However, both your function and the EXECUTE part can be much simpler. In particular, use format() to concatenate longer query strings safely (available since pg 9.1):

CREATE OR REPLACE FUNCTION merge_children_of_icd9 (_ocicd9 text, _ocdesc text
                                                 , _ncicd9 text, _ncdesc text)
  RETURNS void
  LANGUAGE plpgsql AS
$func$
DECLARE
   _sql text;
BEGIN
   FOR _sql IN
      SELECT format('UPDATE %3$s SET %1$I = $3   , %2$I = $4
                     WHERE           %1$I = $1 AND %2$I = $2'
                   , x.cols[1], x.cols[2], x.conrelid::regclass::text)
      FROM  (
         SELECT c.conrelid, array_agg(a.attname ORDER BY a.attnum) AS cols
         FROM   pg_constraint c
         JOIN   pg_attribute  a ON a.attrelid = c.conrelid
                               AND a.attnum = ANY(c.conkey)
         WHERE  c.confrelid = 'icd9'::regclass
         AND    c.contype = 'f'
         GROUP  BY c.oid, c.conrelid
         ORDER  BY c.oid
         ) x
   LOOP
   -- RAISE NOTICE '%', _sql;  -- debug?
      EXECUTE _sql
      USING   _ocicd9, _ocdesc, _ncicd9, _ncdesc;
   END LOOP;
END
$func$;

The function errors out if the FK constraint does not span at least two columns or if the data type of the columns is not compatible with text. May or may not be as intended.

Details on how to pass identifiers safely and execute dynamic SQL:

• Table name as a PostgreSQL function parameter
• Define table and column names as arguments in a plpgsql function?