Re: BUG #1862: ECPG Connect, host variable trailing blanks - Mailing list pgsql-bugs

From Jim.Gray@Bull.com
Subject Re: BUG #1862: ECPG Connect, host variable trailing blanks
Date
Msg-id OF00CEF13E.9CE06130-ON0725707A.00669652-0725707A.0067FDC0@us-phx1.az05.bull.com
Whole thread Raw
In response to Re: BUG #1862: ECPG Connect, host variable trailing blanks  (Michael Meskes <meskes@postgresql.org>)
List pgsql-bugs
We are working in a rather complex environment, where COBOL/SQL on a
legacy
mainframe has its SQL portions sent via a high speed link to a Unix box.

The SQL is translated to a C/SQL environment in the process and the
Postgres
ECPG preprocessor is used on the Unix side for interfacing with Postgres.

We are already doing a translation from trailing blank Cobol strings
to C null terminated strings in this process for the connection host
variables.

The reasons for requesting a Postgres option to strip trailing blanks in
this case are:
1) we are not the only project to run into the problem of Postgres being
oriented towards C strings
2) other RDBMS systems use the expected trimming of Cobol trailing blanks
3) when porting Oracle Cobol applications to Postgres, this option would
be helpful
4) it might be better to strip the trailing blanks in the application if
only one such application were involved,
     but when there are several such applications, centralizing the
behavior in a Postgres option is preferrable







Michael Meskes <meskes@postgresql.org>
09/12/2005 04:06 AM

To
Jim.Gray@Bull.com
cc
Tom Lane <tgl@sss.pgh.pa.us>, Ken.Rosensteel@Bull.com, Michael Fuhr
<mike@fuhr.org>, pgsql-bugs@postgresql.org
Subject
Re: [BUGS] BUG #1862: ECPG Connect, host variable trailing blanks






On Fri, Sep 09, 2005 at 11:02:31AM -0700, Jim.Gray@Bull.com wrote:
> This would allow current behavior to be supported by ECPG and the
Postgres
> server,
> but allow applications that need the behavior (like a Cobol/SQL
> preprocessor for Postgres),
> to make use of it.

Which preprocessor are you using? I wonder if such a logic should better
be implemented in the Cobol-C transition instead.

Michael
--
Michael Meskes
Email: Michael at Fam-Meskes dot De, Michael at Meskes dot
(De|Com|Net|Org)
ICQ: 179140304, AIM/Yahoo: michaelmeskes, Jabber: meskes@jabber.org
Go SF 49ers! Go Rhein Fire! Use Debian GNU/Linux! Use PostgreSQL!

pgsql-bugs by date:

Previous
From: Michael Meskes
Date:
Subject: Re: BUG #1862: ECPG Connect, host variable trailing blanks
Next
From: Tom Lane
Date:
Subject: Re: Race-condition with failed block-write?