Обсуждение: [Fwd: Bug#257263: pgadmin3: Size of bit string types displayed incorrectly]

Dear friends,

can one of you point me to the patch which fixes this issue, I didn't
find a trace of it after a cvs log.
If it's not too hard to backport, I'd like to provide a corrected debian
package and close the bug.

Cheers,
Raphaël

Detail for this bug:

-------- Original Message --------
Subject: Bug#257263: pgadmin3: Size of bit string types displayed
incorrectly
From: Raphaël Enrici <blacknoz@club-internet.fr>
Reply-To: Raphaël Enrici <blacknoz@club-internet.fr>, 257263@bugs.debian.org
To: Richard van den Berg <richard@trust-factory.com>,    257263@bugs.debian.org
References: <20040702095622.41BEC1837B9@localhost.net>

Hi Richard,

Richard van den Berg wrote:

>Package: pgadmin3
>Version: 1.0.2-5
>Severity: normal
>
>When displaying the columns of a table containing a bit string data type,
>pgadmin3 displays its size as 4 less than defined. For example create
>this table, and pgadmin3 will show bit(-2) and bit(0) as datatypes:
>create table foo (bar2 bit(2), bar4 bit(4));
>This bug is in pgadmin3 and not the database since the columns function
>as expected. It is easy to understand where the confusion comes from,
>because the atttypmod column of the pg_attribute table gives the size
>of varchars incremented by 4. For bit strings however, the atttypmod
>column gives the actual size.
>
>
It seems this one is solved in CVS, at least in 20040625 snapshot.
I'm gonna try to backport a patch if possible.
Until that, you can try the 20040625 snapshot debian package (please
apt-get remove your actual pgadmin3 package before installing this one
as I didn't get time to take care of upgrade to 1.1.x packages).
Snapshot can be found here:
http://www.pgadmin.org/snapshots/linux/debian/dists/unstable/pgadmin/binary-i386/

We stay in touch,
Regards,
Raphaël

Raphaël Enrici wrote:

> Dear friends,
>
> can one of you point me to the patch which fixes this issue, I didn't
> find a trace of it after a cvs log.
> If it's not too hard to backport, I'd like to provide a corrected
> debian package and close the bug.


Probably pgDatatype.cpp.

Regards,
Andreas

Andreas Pflug wrote:
> Raphaël Enrici wrote:
>
>> can one of you point me to the patch which fixes this issue, I didn't
>> find a trace of it after a cvs log.
>> If it's not too hard to backport, I'd like to provide a corrected
>> debian package and close the bug.
>
> Probably pgDatatype.cpp.

Thanks! That's it (just like if you were knowing the app ;p).

In fact I found all these files:
RCS file: /projects/pgadmin3/src/db/pgConn.cpp,v
(1.38)
RCS file: /projects/pgadmin3/src/db/pgSet.cpp,v
(1.33)
RCS file: /projects/pgadmin3/src/include/pgConn.h,v
(1.18)
RCS file: /projects/pgadmin3/src/include/pgDatatype.h,v
(1.6)
RCS file: /projects/pgadmin3/src/include/pgSet.h,v
(1.23)
RCS file: /projects/pgadmin3/src/schema/pgDatatype.cpp,v
(1.7)
RCS file: /projects/pgadmin3/src/ui/frmEditGrid.cpp,v
(1.40)
RCS file: /projects/pgadmin3/src/ui/frmMaintenance.cpp,v
(1.3)
RCS file: /projects/pgadmin3/CHANGELOG.txt,v
(1.50)


I think I won't backport this one and wait for a proper new stable release.

Regards,
Raphaël