RE: ODBC SELECT WHERE a IN ('frob') returns where a = '' too
От | Hiroshi Inoue |
---|---|
Тема | RE: ODBC SELECT WHERE a IN ('frob') returns where a = '' too |
Дата | |
Msg-id | 000901bfc48c$7f1dfb00$2801007e@tpf.co.jp обсуждение исходный текст |
Ответ на | Re: ODBC SELECT WHERE a IN ('frob') returns where a = '' too (Adam Haberlach <adam@newsnipple.com>) |
Список | pgsql-bugs |
> -----Original Message----- > From: pgsql-bugs-owner@hub.org [mailto:pgsql-bugs-owner@hub.org]On > Behalf Of Adam Haberlach > > On Tue, May 23, 2000 at 01:33:56AM -0400, Tom Lane wrote: > > Adam Haberlach <adam@newsnipple.com> writes: > > > When I do a query of the form > > > SELECT a FROM tbl WHERE a IN ('frob'); > > > where a is an indexed text field containing 'frob', 'dingus', '', and > > > NULL > > > I get a set that includes rows both where > > > a == 'frob' > > > a == '' > > > > > This happens when I use ODBC from Microsoft Access's > pass-through, but > > > does not seem to happen from the psql > > > command-line. This seems to confuse my coworker who is more familiar > > > with non-Postgres databases then I. > > > > I'm confused too. Best theory I can think of is that Access is > > transforming the query into something different before sending it to the > > backend. That'd be incredibly braindead, but then this is M$ software > > we're talking about. Anyway, the first thing to do is confirm or deny > > that theory by looking at the exact query the backend is getting. > > If you can't get Access to log what it sends, restart the postmaster > > with "-d2" or higher, and see what gets logged... > > > Here is what actually seems to happen. Put your "Oh My GOD that is so > DUMB hats on now." > > /*************/ > query: declare SQL_CUR071E7D04 cursor for SELECT > "rawbebugs"."bbcnum" FROM > "rawbebugs" WHERE ("bestatus" IN ('fixed' ) ) > . . This seems to be the behavior of DAO-Jet with declare/fetch option of psqlodbc driver. Is this really a result of Access's *pass-through* query ? Regards. Hiroshi Inoue Inoue@tpf.co.jp
В списке pgsql-bugs по дате отправления: