ECPG Preprocessor throws Syntax Error [Devel Repository]

Hi,

Yesterday I pull source from PG-Devel repository (Head) and was playing
around with ECPG. I got following two issues while parsing my script files
through ECPG preprocessor which I think are potential bugs;

1- I get this error message from preprocessor,

*ERROR: cursor "m_name_cursor" already defined *

Below is the code snippet, error message is thrown on the last line of this
snippet;

...
...
EXEC SQL DECLARE m_name_cursor CURSOR FOR SELECT name,id FROM m_table3;
EXEC SQL OPEN m_name_cursor;
EXEC SQL CLOSE m_name_cursor;

/* Declaring cursor with the same name */
EXEC SQL DECLARE m_name_cursor CURSOR FOR SELECT name,id FROM m_table4;

2- Preprocessor throws this error;

*ERROR: syntax error at or near "-"*

Below is the code snippet, the error is thrown on the last line of the
snippet;

...
...
const char *stmt1 = "SELECT abs(?)";
EXEC SQL PREPARE mystmt1 FROM :stmt1;
EXEC SQL EXECUTE mystmt1 INTO :result USING -.55566;

Regards,
Ahmed

On Thu, Dec 11, 2008 at 10:09:40PM +0500, Ahmed Shinwari wrote:
> through ECPG preprocessor which I think are potential bugs;
>
> 1- I get this error message from preprocessor,
>
> *ERROR: cursor "m_name_cursor" already defined *
>
> Below is the code snippet, error message is thrown on the last line of this
> snippet;
>
> EXEC SQL DECLARE m_name_cursor CURSOR FOR SELECT name,id FROM m_table3;
> EXEC SQL OPEN m_name_cursor;
> EXEC SQL CLOSE m_name_cursor;
>
> /* Declaring cursor with the same name */
> EXEC SQL DECLARE m_name_cursor CURSOR FOR SELECT name,id FROM m_table4;

I might be missing something here, but where do you see the bug? The cursor is
indeed already defined.

> 2- Preprocessor throws this error;
>
> *ERROR: syntax error at or near "-"*
> ...

Fixed in CVS. Thanks for the report.

Michael
--
Michael Meskes
Michael at Fam-Meskes dot De, Michael at Meskes dot (De|Com|Net|Org)
Michael at BorussiaFan dot De, Meskes at (Debian|Postgresql) dot Org
ICQ: 179140304, AIM/Yahoo: michaelmeskes, Jabber: meskes(at)jabber(dot)org
Go VfL Borussia! Go SF 49ers! Use Debian GNU/Linux! Use PostgreSQL!

>I might be missing something here, but where do you see the bug? The cursor is
>indeed already defined.

What i understand here is,we should be able to DECLARE cursor again
with the same name if that is CLOSE already.

--- On Fri, 12/12/08, Michael Meskes <meskes(at)postgresql(dot)org> wrote:
From: Michael Meskes <meskes(at)postgresql(dot)org>
Subject: Re: [BUGS] ECPG Preprocessor throws Syntax Error [Devel Repository]
To: "Ahmed Shinwari" <ahmed(dot)shibwari(at)gmail(dot)com>
Cc: pgsql-bugs(at)postgresql(dot)org
Date: Friday, December 12, 2008, 4:26 AM

On Thu, Dec 11, 2008 at 10:09:40PM +0500, Ahmed Shinwari wrote:
> through ECPG preprocessor which I think are potential bugs;
>
> 1- I get this error message from preprocessor,
>
> *ERROR: cursor "m_name_cursor" already defined *
>
> Below is the code snippet, error message is thrown on the last line of
this
> snippet;
>
> EXEC SQL DECLARE m_name_cursor CURSOR FOR SELECT name,id FROM m_table3;
> EXEC SQL OPEN m_name_cursor;
> EXEC SQL CLOSE m_name_cursor;
>
> /* Declaring cursor with the same name */
> EXEC SQL DECLARE m_name_cursor CURSOR FOR SELECT name,id FROM m_table4;

I might be missing something here, but where do you see the bug? The cursor is
indeed already defined.

> 2- Preprocessor throws this error;
>
> *ERROR: syntax error at or near "-"*
> ...

Fixed in CVS. Thanks for the report.

Michael
--
Michael Meskes
Michael at Fam-Meskes dot De, Michael at Meskes dot (De|Com|Net|Org)
Michael at BorussiaFan dot De, Meskes at (Debian|Postgresql) dot Org
ICQ: 179140304, AIM/Yahoo: michaelmeskes, Jabber: meskes(at)jabber(dot)org
Go VfL Borussia! Go SF 49ers! Use Debian GNU/Linux! Use PostgreSQL!

Zahid Khan <khanzahid235(at)yahoo(dot)com> writes:
>> I might be missing something here, but where do you see the bug? The cursor is
>> indeed already defined.

> What i understand here is,we should be able to DECLARE cursor again
> with the same name if that is CLOSE already.

No, CLOSE reverses the effect of OPEN, not of DECLARE.

regards, tom lane