[sql] Get data type of field in select statement in ORACLE

Can I get data types of each column I selected instead of the values, using a select statement?

FOR EXAMPLE:

SELECT a.name, a.surname, b.ordernum 
FROM customer a
JOIN orders b
ON a.id = b.id

and result should be like this

name    | NVARCHAR(100)
surname | NVARCHAR(100)
ordernum| INTEGER

or it can be in row like this, it isn't important:

name           |   surname     |  ordernum
NVARCHAR(100)  | NVARCHAR(100) |   INTEGER

Thanks

I usually create a view and use the DESC command:

CREATE VIEW tmp_view AS 
SELECT 
      a.name
    , a.surname
    , b.ordernum 
FROM customer a
  JOIN orders b
    ON a.id = b.id

Then, the DESC command will show the type of each field.

DESC tmp_view

you can use the DBMS_SQL.DESCRIBE_COLUMNS2

    SET SERVEROUTPUT ON;
DECLARE
    STMT CLOB;
    CUR NUMBER;
    COLCNT NUMBER;
    IDX NUMBER;
    COLDESC DBMS_SQL.DESC_TAB2;
BEGIN
    CUR := DBMS_SQL.OPEN_CURSOR;
    STMT := 'SELECT  object_name , to_char(object_id), created FROM    DBA_OBJECTS where rownum<10';

    SYS.DBMS_SQL.PARSE(CUR, STMT, DBMS_SQL.NATIVE);
    DBMS_SQL.DESCRIBE_COLUMNS2(CUR, COLCNT, COLDESC);
    DBMS_OUTPUT.PUT_LINE('Statement: ' || STMT);
    FOR IDX IN 1 .. COLCNT
    LOOP
        CASE COLDESC(IDX).col_type
        WHEN 2 THEN
            DBMS_OUTPUT.PUT_LINE('#' || TO_CHAR(IDX) || ': NUMBER');
        WHEN 12 THEN
            DBMS_OUTPUT.PUT_LINE('#' || TO_CHAR(IDX) || ': DATE');
        WHEN 180 THEN
            DBMS_OUTPUT.PUT_LINE('#' || TO_CHAR(IDX) || ': TIMESTAMP');
        WHEN 1 THEN
            DBMS_OUTPUT.PUT_LINE('#' || TO_CHAR(IDX) || ': VARCHAR'||':'|| COLDESC(IDX).col_max_len);
        WHEN 9 THEN
            DBMS_OUTPUT.PUT_LINE('#' || TO_CHAR(IDX) || ': VARCHAR2');
        -- Insert more cases if you need them
        ELSE
            DBMS_OUTPUT.PUT_LINE('#' || TO_CHAR(IDX) || ': OTHERS (' || TO_CHAR(COLDESC(IDX).col_type) || ')');
        END CASE;
    END LOOP;
    SYS.DBMS_SQL.CLOSE_CURSOR(CUR);
EXCEPTION 
    WHEN OTHERS THEN
        DBMS_OUTPUT.PUT_LINE(SQLERRM(SQLCODE()) || ': ' || DBMS_UTILITY.FORMAT_ERROR_BACKTRACE);
        SYS.DBMS_SQL.CLOSE_CURSOR(CUR);
END;
/


full example in the below url

https://www.ibm.com/support/knowledgecenter/sk/SSEPGG_9.7.0/com.ibm.db2.luw.sql.rtn.doc/doc/r0055146.html

Also, if you have Toad for Oracle, you can highlight the statement and press CTRL + F9 and you'll get a nice view of column and their datatypes.

I found a not-very-intuitive way to do this by using DUMP()

SELECT DUMP(A.NAME), 
       DUMP(A.surname), 
       DUMP(B.ordernum) 
FROM   customer A 
       JOIN orders B 
         ON A.id = B.id

It will return something like:

'Typ=1 Len=2: 0,48' for each column.

Type=1 means VARCHAR2/NVARCHAR2
Type=2 means NUMBER/FLOAT
Type=12 means DATE, etc.

You can refer to this oracle doc for information Datatype Code
or this for a simple mapping Oracle Type Code Mappings

I came into the same situation. As a workaround, I just created a view (If you have privileges) and described it and dropped it later. :)

If you don't have privileges to create a view in Oracle, a "hack" around it to use MS Access :-(

In MS Access, create a pass through query with your sql (but add where clause to just select 1 record), create a select query from the view (very important), selecting all *, then create a make table from the select query. When this runs it will create a table with one record, all the data types should "match" oracle. i.e. Passthrough --> Select --> MakeTable --> Table

I am sure there are other better ways, but if you have limited tools and privileges this will work.