attlen weirdness?

Quoting The Hermit Hacker (scrappy@hub.org):

Bruce? Guys? Any word on this one? I *thought* I remembered reading
soething about the 'attlen' becoming -1, but just checked the mailing list
archives and couldn't find anything :(

It appears as if atttypmod now has the field length + 4 ???

Julie

--
[ Julia Anne Case ] [ Ships are safe inside the harbor, ]
[Programmer at large] [ but is that what ships are really for. ]
[ Admining Linux ] [ To thine own self be true. ]
[ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]

Quoting Bruce Momjian (maillist@candle.pha.pa.us):

I assume you are asking of attlen for char() and varchar() have the max
length. The answer is now NO. They are varlena structures like text,
and so have a attlen of -1. Atttypmod for that pg_attribute row now has
the defined length.

Even for a char(4) field? ok, if attlen doesn't give the
length of a field, how do I find it?

Julie

--
[ Julia Anne Case ] [ Ships are safe inside the harbor, ]
[Programmer at large] [ but is that what ships are really for. ]
[ Admining Linux ] [ To thine own self be true. ]
[ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]

Quoting Bruce Momjian (maillist@candle.pha.pa.us):

pg_attribute.atttypmod.

this field appears to be field size + 4
in a char(1024) field it returns 1028
in a char(10) it returns 14

is it padded?

Julie

--
[ Julia Anne Case ] [ Ships are safe inside the harbor, ]
[Programmer at large] [ but is that what ships are really for. ]
[ Admining Linux ] [ To thine own self be true. ]
[ Windows/WindowsNT ] [ Fair is where you take your cows to be judged. ]

Bruce? Guys? Any word on this one? I *thought* I remembered reading
soething about the 'attlen' becoming -1, but just checked the mailing list
archives and couldn't find anything :(

Thanks...

On Tue, 10 Mar 1998, Julia A.Case wrote:

Quoting The Hermit Hacker (scrappy@hub.org):

Bruce? Guys? Any word on this one? I *thought* I remembered reading
soething about the 'attlen' becoming -1, but just checked the mailing list
archives and couldn't find anything :(

It appears as if atttypmod now has the field length + 4 ???

I am the atttypmod guy, but there is not enough context to answer this
question.

-- 
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)

Quoting The Hermit Hacker (scrappy@hub.org):

Bruce? Guys? Any word on this one? I *thought* I remembered reading
soething about the 'attlen' becoming -1, but just checked the mailing list
archives and couldn't find anything :(

It appears as if atttypmod now has the field length + 4 ???

I assume you are asking of attlen for char() and varchar() have the max
length. The answer is now NO. They are varlena structures like text,
and so have a attlen of -1. Atttypmod for that pg_attribute row now has
the defined length.

In the old days, pre 6.3, pg_type.typlen != pg_attribute.attlen for
char() and varchar(), causing all sorts of type-conditional wierness in
the backend for these types. With the new code, pg_type.typlen =
pg_attribute.attlen all the time.

-- 
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)

Quoting Bruce Momjian (maillist@candle.pha.pa.us):

I assume you are asking of attlen for char() and varchar() have the max
length. The answer is now NO. They are varlena structures like text,
and so have a attlen of -1. Atttypmod for that pg_attribute row now has
the defined length.

Even for a char(4) field? ok, if attlen doesn't give the
length of a field, how do I find it?

pg_attribute.atttypmod.

-- 
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)

Quoting Bruce Momjian (maillist@candle.pha.pa.us):

pg_attribute.atttypmod.

this field appears to be field size + 4
in a char(1024) field it returns 1028
in a char(10) it returns 14

is it padded?

Yes. All varlena sizes store the VARHDRSZ, variable header size of 4,
in the length field.

-- 
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)

Quoting Bruce Momjian (maillist@candle.pha.pa.us):

I assume you are asking of attlen for char() and varchar() have the max
length. The answer is now NO. They are varlena structures like text,
and so have a attlen of -1. Atttypmod for that pg_attribute row now has
the defined length.

Even for a char(4) field? ok, if attlen doesn't give the
length of a field, how do I find it?

Right now: yes, but I would like to see the char() type changed to a fixed length type (which it really is)
then it would have the length in attlen. Right now the length for char and varchar is in atttypmod.

Andreas

Quoting Bruce Momjian (maillist@candle.pha.pa.us):

I assume you are asking of attlen for char() and varchar() have the max
length. The answer is now NO. They are varlena structures like text,
and so have a attlen of -1. Atttypmod for that pg_attribute row now has
the defined length.

Even for a char(4) field? ok, if attlen doesn't give the
length of a field, how do I find it?

Right now: yes, but I would like to see the char() type changed to a fixed length type (which it really is)
then it would have the length in attlen. Right now the length for char and varchar is in atttypmod.

Again, we can do this, but we then have to have atttypmod available in
all places where we check for length. I will add it to the TODO list.

-- 
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)

Right now: yes, but I would like to see the char() type changed to a fixed length type (which it really is)
then it would have the length in attlen. Right now the length for char and varchar is in atttypmod.

Again, we can do this, but we then have to have atttypmod available in
all places where we check for length. I will add it to the TODO list.

I am starting to see the problem, only atttypmod is passed around not attlen.
Therefore even if we do this, the char() length of a column still has to be in atttypmod.
So it will stay the same for Julie, whether it is taken out or not. But since it is really
hard for a generic application like ODBC, maybe we could change atttypmod to not
include the VARHDRSZ, most of the time it is needed without VARHDRSZ anyway ?
I am starting to think that this would be more important than to remove the 4 bytes.

Andreas

Right now: yes, but I would like to see the char() type changed to a fixed length type (which it really is)
then it would have the length in attlen. Right now the length for char and varchar is in atttypmod.

Again, we can do this, but we then have to have atttypmod available in
all places where we check for length. I will add it to the TODO list.

I am starting to see the problem, only atttypmod is passed around not attlen.
Therefore even if we do this, the char() length of a column still has to be in atttypmod.
So it will stay the same for Julie, whether it is taken out or not. But since it is really
hard for a generic application like ODBC, maybe we could change atttypmod to not
include the VARHDRSZ, most of the time it is needed without VARHDRSZ anyway ?
I am starting to think that this would be more important than to remove the 4 bytes.

Those four bytes have always been there. Check psql or pg_dump. See
them subtracting VARHDRSZ all the time.

-- 
Bruce Momjian                          |  830 Blythe Avenue
maillist@candle.pha.pa.us              |  Drexel Hill, Pennsylvania 19026
  +  If your life is a hard drive,     |  (610) 353-9879(w)
  +  Christ can be your backup.        |  (610) 853-3000(h)