Обсуждение: split_part bug

PostgreSQL 8.0b4 as released.

select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',1);
select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',2);
select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',3);
select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',4);

should return a b c, and d
returns a, null, null, null

select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',2);
select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',3);
select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
z',' ',4);

should return b c, and d
returns b, c, null

... John

John Hansen <john@geeknet.com.au> writes:
> PostgreSQL 8.0b4 as released.
> select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
> z',' ',1);
> select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
> z',' ',2);
> select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
> z',' ',3);
> select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y
> z',' ',4);

> should return a b c, and d
> returns a, null, null, null

Works fine for me.  What encoding/locale are you using?
        regards, tom lane

> Works fine for me.  What encoding/locale are you using?

unicode / c

John Hansen <john@geeknet.com.au> writes:
>> Works fine for me.  What encoding/locale are you using?

> unicode / c

[ shrug... ]  Works fine for me in unicode, too.

u=# select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y z',' ',1);split_part
------------a
(1 row)

u=# select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y z',' ',2);split_part
------------b
(1 row)

u=# select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y z',' ',3);split_part
------------c
(1 row)

u=# select split_part('a b c d e f g h i j k l m n o p q r s t u v w x y z',' ',4);split_part
------------d
(1 row)

        regards, tom lane

> [ shrug... ]  Works fine for me in unicode, too.

never mind me,. I broke it.....

seems my assumption that UCS2 == UTF16 was way off....

... john