Re: ImportError with libpq.so.2
"Billy G. Allie" <[email protected]>
| Newsgroups | gmane.comp.python.db.pypgsql.user |
|---|---|
| Message-ID | <[email protected]> |
"Billy G. Allie" wrote: > "Todd Reed" wrote: > > I ended up installing pyPgSQL 2.3 from the source distribution, and this > > did resolve the ImportError. However, there is an incompatibility between > > pyPgSQL 2.3 and PostgreSQL 7.3.2 (and possibly other 7.3.x releases). The > > system table pg_type used to have a column called "typprtlen" which is no > > longer present in PostgreSQL 7.3.2. pyPgSQL makes a query on pg_type that > > attempts to access this non-existent column to obtain information for the > > "description" field of cursor objects. I ended up modifying the query to > > always return -1; I think "typprtlen" was used to find the display width > > for the column. > > > > The 2.3 release does have the typprtlen problem. The CVS source contains a > fix for that problem as well as several others. The 2.3 release does contain the fix for the typprtlen problem. The change in the current CVS was to have display length in the cursor.description object return None instead of -1. > I would recommend getting > the latest source from CVS for the latest bug fixes. I still recomend getting the latest source from CVS, though. :-) -- ____ | Billy G. Allie | Domain....: [email protected] | /| | 7436 Hartwell | MSN.......: [email protected] |-/-|----- | Dearborn, MI 48126| |/ |LLIE | (313) 582-1540 |
signature.asc
(application/pgp-signature, 1.7 KB)
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Content-Type: text/plain; charset=us-ascii "Billy G. Allie" wrote: > "Todd Reed" wrote: > > I ended up installing pyPgSQL 2.3 from the source distribution, and this > > did resolve the ImportError. However, there is an incompatibility between > > pyPgSQL 2.3 and PostgreSQL 7.3.2 (and possibly other 7.3.x releases). The > > system table pg_type used to have a column called "typprtlen" which is no > > longer present in PostgreSQL 7.3.2. pyPgSQL makes a query on pg_type that > > attempts to access this non-existent column to obtain information for the > > "description" field of cursor objects. I ended up modifying the query to > > always return -1; I think "typprtlen" was used to find the display width > > for the column. > > > > The 2.3 release does have the typprtlen problem. The CVS source contains a > fix for that problem as well as several others. The 2.3 release does contain the fix for the typprtlen problem. The change in the current CVS was to have display length in the cursor.description object return None instead of -1. > I would recommend getting > the latest source from CVS for the latest bug fixes. I still recomend getting the latest source from CVS, though. :-) - -- ____ | Billy G. Allie | Domain....: [email protected] | /| | 7436 Hartwell | MSN.......: [email protected] |-/-|----- | Dearborn, MI 48126| |/ |LLIE | (313) 582-1540 | -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.0.2 (UnixWare) Comment: Exmh version 2.2 06/23/2000 iD8DBQE+ZZ0LnmIkMXoVVdURAocLAJ49vUJ8tVPZAS7521lM/QplqZX3wwCfSVum K14npltXrmTzHRmQ7iRAJgI= =os7F -----END PGP SIGNATURE-----