Oracle CHAR

Summary: in this tutorial, you will learn about the Oracle CHAR data type which is a fixed-length character string type.

Introduction to Oracle CHAR data type

The Oracle CHAR data type allows you to store fixed-length character strings. The CHAR data type can store a character string with the size from 1 to 2000 bytes.

To define a CHAR column, you need to specify a string length either in bytes or characters as shown following:

CHAR(length BYTE)
CHAR(length CHAR)

If you don’t explicitly specify BYTE or CHAR followed the length, Oracle uses the BYTE by default.

The default value of length is 1 if you skip it like the following example:

column_name CHAR

When you insert or update a fixed-length character string column, Oracle stores the characters as the fixed-length data. It means that if you store a value whose length is less than the maximum length defined in the column, Oracle pads the spaces to the character string up to the maximum length. In case you insert a value whose length is larger than the column, Oracle returns an error.

Oracle uses blank-padded comparison semantics for comparing CHAR values.

Oracle CHAR examples

Let’s take a look some examples to understand how the CHAR data type works.

A) Space usage example

First, create a new table named t that consists of a CHAR column (x) and VARCHAR2 column (y). The length of each column is 10 bytes.

CREATE TABLE t (
    x CHAR(10),
    y VARCHAR2(10)
);

Second, insert a new row into the t table with the same data for both x and y columns:

INSERT INTO t(x, y )
VALUES('Oracle', 'Oracle');

Third, verify the insert by using the following query:

SELECT
    *
FROM
    t

The following statement retrieves data from the t table:

SELECT
    x,
    DUMP(x),
    y,
    DUMP(y)
FROM
    t;

In this example, we used the DUMP() function to return the detailed information on x and y columns:

The string Oracle takes 6 bytes. However, Oracle padded 4 more spaces on the right of the string to make its length 10 bytes for the x column. It is not the case for the y column because the data type of y column is a variable-length character string (VARCHAR2).

It is more clear if you use the LENGTHB() function to get the number of bytes used by the x and y columns:

SELECT
    LENGTHB(x),
    LENGTHB(y)
FROM
    t;

B) Characters comparison example

The following statements return the same result:

SELECT * FROM t WHERE x = 'Oracle';
SELECT * FROM t WHERE y = 'Oracle';

However, if you use bind variables, the effect is different. Consider the following example:

SQL> variable v varchar2(10)
SQL> exec :v := 'Oracle';

PL/SQL procedure successfully completed.

In this code block, we declared v as a bind variable with the VARCHAR2 data type.

Now, we use v as an input to compare against the the x column:

SQL> select * from t where x = :v;

no rows selected

The statement returned an empty result set.

The following query uses the v variable to compare with the y column:

SQL> select * from t where y = :v;

X          Y
---------- ----------
Oracle     Oracle

It returned a row as expected.

This is because when comparing the string of character types with unequal length, Oracle uses non-blank-padding semantics.

To make it work, you need to use the RTRIM() function to strip spaces from the CHAR data before comparing it with the input string as follows:

SQL> select * from t where rtrim(x) = :v;

X          Y
---------- ----------
Oracle     Oracle

In this tutorial, you have learned about the Oracle CHAR data type and understood the behaviors of the CHAR columns in terms of space usages and character comparisons.