Oracle - Why does the leading zero of a number dis

问题:

In Oracle, when converting a number with a leading zero to a character, why does the leading number disappear? Is this logic Oracle specific, or specific to SQL?

Example:

SELECT TO_CHAR(0.56) FROM DUAL;
/* Result = .56 */

回答1:

It's the default formatting that Oracle provides. If you want leading zeros on output, you'll need to explicitly provide the format. Use:

SELECT TO_CHAR(0.56,'0.99') FROM DUAL;

or even:

SELECT TO_CHAR(.56,'0.99') FROM DUAL;

The same is true for trailing zeros:

SQL> SELECT TO_CHAR(.56,'0.990') val FROM DUAL;

VAL
------
 0.560

The general form of the TO_CHAR conversion function is:

TO_CHAR(number, format)

回答2:

I was looking for a way to format numbers without leading or trailing spaces, periods, zeros (except one leading zero for numbers less than 1 that should be present).

This is frustrating that such most usual formatting can't be easily achieved in Oracle.

Even Tom Kyte only suggested long complicated workaround like this:

case when trunc(x)=x
    then to_char(x, 'FM999999999999999999')
    else to_char(x, 'FM999999999999999.99')
end x

But I was able to find shorter solution that mentions the value only once:

rtrim(to_char(x, 'FM999999999999990.99'), '.')

This works as expected for all possible values:

select 
    to_char(num, 'FM99.99') wrong_leading_period,
    to_char(num, 'FM90.99') wrong_trailing_period,
    rtrim(to_char(num, 'FM90.99'), '.') correct
from (
  select num from (select 0.25 c1, 0.1 c2, 1.2 c3, 13 c4, -70 c5 from dual)
  unpivot (num for dummy in (c1, c2, c3, c4, c5))
) sampledata;

    | WRONG_LEADING_PERIOD | WRONG_TRAILING_PERIOD | CORRECT |
    |----------------------|-----------------------|---------|
    |                  .25 |                  0.25 |    0.25 |
    |                   .1 |                   0.1 |     0.1 |
    |                  1.2 |                   1.2 |     1.2 |
    |                  13. |                   13. |      13 |
    |                 -70. |                  -70. |     -70 |

Still looking for even shorter solution.

There is a shortening approarch with custom helper function:

create or replace function str(num in number) return varchar2
as
begin
    return rtrim(to_char(num, 'FM999999999999990.99'), '.');
end;

But custom pl/sql functions have significant performace overhead that is not suitable for heavy queries.

回答3:

Seems like the only way to get decimal in a pretty (for me) form requires some ridiculous code.

The only solution I got so far:

CASE WHEN xy>0 and xy<1 then '0' || to_char(xy) else to_char(xy)

xy is a decimial.

xy             query result
0.8            0.8  --not sth like .80
10             10  --not sth like 10.00

please answer this question if there is a real format option for this.

回答4:

That only works for numbers less than 1.

select to_char(12.34, '0D99') from dual;
-- Result: #####

This won't work.

You could do something like this but this results in leading whitespaces:

select to_char(12.34, '999990D99') from dual;
-- Result: '     12,34'

Ultimately, you could add a TRIM to get rid of the whitespaces again but I wouldn't consider that a proper solution either...

select trim(to_char(12.34, '999990D99')) from dual;
-- Result: 12,34

Again, this will only work for numbers with 6 digits max.

Edit: I wanted to add this as a comment on DCookie's suggestion but I can't.

回答5:

Try this to avoid to_char limitations:

SELECT 
regexp_replace(regexp_replace(n,'^-\'||s,'-0'||s),'^\'||s,'0'||s)
FROM (SELECT -0.89 n,RTrim(1/2,5) s FROM dual);