cx-oracle-users — Primary discussion list for users (and developers) of cx_Oracle

[cx-oracle-users] build/install on intel os x?

[Steve S. <ssw...@ve...>]

Has anyone had success building on an Intel Mac?  I've had a notable  
lack of success building cx_oracle 4.1.2 on OS X 10.4.5 for Intel.   
I'm using Oracle's 10.1.0.3 instant client and gcc 4.0.1.  I haven't  
seen a full 10g client for OS X, just the instant client.  The  
database server itself is an Ubuntu machine running Oracle 10g  
express edition (10.2.x.x.).

The Python version is 2.4.1; the machine came with 2.3, and I  
installed the latest MacPython distribution.  I assume that's a Power  
version that is running under Rosetta; I don't think a universal  
binary version has yet been released.

The interpreter reports the build used gcc 3.3.  Is that part of the  
problem?

I'm very new to Macs and fairly new to Python, and the whole Intel  
thing seems to throw in some more fun and games, so I'm not sure if  
I'm just doing something stupid or if this combination of  
technologies is just a no-go for now.  I could post the vast list of  
build errors I'm seeing, if I find out that this should even be  
possible.

I've searched for a while without seeing anything that can help, so  
if anyone can throw some hints and tips my way I would appreciate  
it.  Thanks.

-------------------------------
Steve Swartzlander
ssw...@ve...

Re: [cx-oracle-users] RE: gcc296 error when building cx_Oracle

[Anthony T. <ant...@gm...>]

Yes. I can (and already have) made changes to this to use the compiler
directives __DATE__ and __TIME__ that are available in gcc and the
Microsoft compilers. I have no idea if there are any compilers that
don't support those directives. Any thoughts on that? As you've
noticed, the shells are too inconsistent to pass this on the command
line reliably.

On 3/31/06, Henning von Bargen <H.v...@t-...> wrote:
> > Date: Thu, 30 Mar 2006 11:22:05 -0800
> > From: "Molina, Gerardo" <Ger...@sc...>
> >
> > I'm getting the following error when trying to build
> > cx_Oracle 4.1.2 on Red Hat linux:
> >
> > Linux s0051cdc 2.4.21-27.ELsmp #1 SMP Wed Dec 1 21:59:02 EST 2004 i686
> > s0051cdc:SQLNET(/opt/ORACLE/TAR/python/cx_Oracle-4.1.2)->pytho
> > n setup.py
> > build
> > running build
> > running build_ext
> > building 'cx_Oracle' extension
> > gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
> > -Wstrict-prototypes -fPIC
> > -I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/demo
> > -I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/public
> > -I/opt/ORACLE/app/oracle/product/9.2.0/network/public
> > -I/opt/ORACLE/app/oracle/product/9.2.0/sdk/include
> > -I/opt/ORACLE/python/include/python2.4 -c cx_Oracle.c -o
> > build/temp.linux-i686-2.4/cx_Oracle.o -DBUILD_TIME=3D3D"March 30, 2006
> > 14:17:56"
> > gcc2.96 will be used ...
> >
> > gcc296: cannot specify -o with -c or -S and multiple compilations
> > error: command 'gcc' failed with exit status 1
> >
> > Has anyone seen this before? and have a fix?
> >
> > Thanks in advance,
> > Gerardo
>
> Gerardo, I had a similar problem when I tried to build cx_Oracle
> as a static extension for python on HP/UX.
>
> In my case it was a problem with the quotation marks in the define
> directive for BUILD_TIME - these scrambled the invoked shell somehow.
>
> Here's my workaround: Don't use a DEFINE for BUILD_TIME at all
>  and instead hard-code it (I know that's ugly, but it works at least).
>
> - Remove any occurences for BUILD_TIME in the build scripts.
> - Modify the file cx_Oracle.c (line 421):
> OLD:            PyString_FromString(BUILD_TIME)) < 0)
> NEW:            PyString_FromString("February 03, 2006 15:32:22")) < 0)
>
> P.S. Anthony, is itpossible for you to solve this in a more general way?
>  Or (perhaps even better): Let the Python Build-Gurus change the way
>  define directives in "makesetup.local" and "makesetup" are handled?
>  It really seems to be a problem with the shell invoked.
>
>
>  It seems to be a problem everytime
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting langua=
ge
> that extends applications into web and mobile media. Attend the live webc=
ast
> and join the prime developer group breaking into this new coding territor=
y!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=110944&bid$1720&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

RE: [cx-oracle-users] RE: gcc296 error when building cx_Oracle

[Molina, G. <Ger...@sc...>]

Henning,

Thanks for the reply.

I was able to build cx_Oracle 4.1.2 by using gcc 3.2 instead of gcc
2.96.

Gerardo=20

-----Original Message-----
From: cx-...@li...
[mailto:cx-...@li...] On Behalf Of
Henning von Bargen
Sent: Friday, March 31, 2006 12:43 AM
To: cx-...@li...
Subject: [cx-oracle-users] RE: gcc296 error when building cx_Oracle

> Date: Thu, 30 Mar 2006 11:22:05 -0800
> From: "Molina, Gerardo" <Ger...@sc...>
>
> I'm getting the following error when trying to build cx_Oracle 4.1.2=20
> on Red Hat linux:
>=20
> Linux s0051cdc 2.4.21-27.ELsmp #1 SMP Wed Dec 1 21:59:02 EST 2004 i686

> s0051cdc:SQLNET(/opt/ORACLE/TAR/python/cx_Oracle-4.1.2)->pytho
> n setup.py
> build
> running build
> running build_ext
> building 'cx_Oracle' extension
> gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall=20
> -Wstrict-prototypes -fPIC=20
> -I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/demo
> -I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/public
> -I/opt/ORACLE/app/oracle/product/9.2.0/network/public
> -I/opt/ORACLE/app/oracle/product/9.2.0/sdk/include
> -I/opt/ORACLE/python/include/python2.4 -c cx_Oracle.c -o=20
> build/temp.linux-i686-2.4/cx_Oracle.o -DBUILD_TIME=3D3D"March 30, 2006 =

> 14:17:56"
> gcc2.96 will be used ...
>
> gcc296: cannot specify -o with -c or -S and multiple compilations
> error: command 'gcc' failed with exit status 1
>=20
> Has anyone seen this before? and have a fix?
>
> Thanks in advance,
> Gerardo

Gerardo, I had a similar problem when I tried to build cx_Oracle as a
static extension for python on HP/UX.

In my case it was a problem with the quotation marks in the define
directive for BUILD_TIME - these scrambled the invoked shell somehow.

Here's my workaround: Don't use a DEFINE for BUILD_TIME at all  and
instead hard-code it (I know that's ugly, but it works at least).

- Remove any occurences for BUILD_TIME in the build scripts.
- Modify the file cx_Oracle.c (line 421):
OLD:            PyString_FromString(BUILD_TIME)) < 0)
NEW:            PyString_FromString("February 03, 2006 15:32:22")) < 0)

If you understand German, I can send you a Word-Document by private
mail.

HTH

Henning

P.S. Anthony, is itpossible for you to solve this in a more general way?
 Or (perhaps even better): Let the Python Build-Gurus change the way
define directives in "makesetup.local" and "makesetup" are handled?
 It really seems to be a problem with the shell invoked.

=20
 It seems to be a problem everytime


-------------------------------------------------------
This SF.Net email is sponsored by xPML, a groundbreaking scripting
language that extends applications into web and mobile media. Attend the
live webcast and join the prime developer group breaking into this new
coding territory!
http://sel.as-us.falkag.net/sel?cmd=3Dk&kid=110944&bid$1720&dat=121642
_______________________________________________
cx-oracle-users mailing list
cx-...@li...
https://lists.sourceforge.net/lists/listinfo/cx-oracle-users

[cx-oracle-users] RE: gcc296 error when building cx_Oracle

[Henning v. B. <H.v...@t-...>]

> Date: Thu, 30 Mar 2006 11:22:05 -0800
> From: "Molina, Gerardo" <Ger...@sc...>
>
> I'm getting the following error when trying to build=20
> cx_Oracle 4.1.2 on Red Hat linux:
>=20
> Linux s0051cdc 2.4.21-27.ELsmp #1 SMP Wed Dec 1 21:59:02 EST 2004 i686
> s0051cdc:SQLNET(/opt/ORACLE/TAR/python/cx_Oracle-4.1.2)->pytho
> n setup.py
> build
> running build
> running build_ext
> building 'cx_Oracle' extension
> gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
> -Wstrict-prototypes -fPIC
> -I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/demo
> -I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/public
> -I/opt/ORACLE/app/oracle/product/9.2.0/network/public
> -I/opt/ORACLE/app/oracle/product/9.2.0/sdk/include
> -I/opt/ORACLE/python/include/python2.4 -c cx_Oracle.c -o
> build/temp.linux-i686-2.4/cx_Oracle.o -DBUILD_TIME=3D3D"March 30, 2006
> 14:17:56"
> gcc2.96 will be used ...
>
> gcc296: cannot specify -o with -c or -S and multiple compilations
> error: command 'gcc' failed with exit status 1
>=20
> Has anyone seen this before? and have a fix?
>
> Thanks in advance,
> Gerardo

Gerardo, I had a similar problem when I tried to build cx_Oracle
as a static extension for python on HP/UX.

In my case it was a problem with the quotation marks in the define
directive for BUILD_TIME - these scrambled the invoked shell somehow.

Here's my workaround: Don't use a DEFINE for BUILD_TIME at all
 and instead hard-code it (I know that's ugly, but it works at least).

- Remove any occurences for BUILD_TIME in the build scripts.
- Modify the file cx_Oracle.c (line 421):
OLD:            PyString_FromString(BUILD_TIME)) < 0)
NEW:            PyString_FromString("February 03, 2006 15:32:22")) < 0)

If you understand German, I can send you a Word-Document by private =
mail.

HTH

Henning

P.S. Anthony, is itpossible for you to solve this in a more general way?
 Or (perhaps even better): Let the Python Build-Gurus change the way
 define directives in "makesetup.local" and "makesetup" are handled?
 It really seems to be a problem with the shell invoked.

=20
 It seems to be a problem everytime

[cx-oracle-users] gcc296 error when building cx_Oracle

[Molina, G. <Ger...@sc...>]

I'm getting the following error when trying to build cx_Oracle 4.1.2 on
Red Hat linux:
=20
Linux s0051cdc 2.4.21-27.ELsmp #1 SMP Wed Dec 1 21:59:02 EST 2004 i686
s0051cdc:SQLNET(/opt/ORACLE/TAR/python/cx_Oracle-4.1.2)->python setup.py
build
running build
running build_ext
building 'cx_Oracle' extension
gcc -pthread -fno-strict-aliasing -DNDEBUG -g -O3 -Wall
-Wstrict-prototypes -fPIC
-I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/demo
-I/opt/ORACLE/app/oracle/product/9.2.0/rdbms/public
-I/opt/ORACLE/app/oracle/product/9.2.0/network/public
-I/opt/ORACLE/app/oracle/product/9.2.0/sdk/include
-I/opt/ORACLE/python/include/python2.4 -c cx_Oracle.c -o
build/temp.linux-i686-2.4/cx_Oracle.o -DBUILD_TIME=3D"March 30, 2006
14:17:56"
gcc2.96 will be used ...
=20
gcc296: cannot specify -o with -c or -S and multiple compilations
error: command 'gcc' failed with exit status 1

Has anyone seen this before? and have a fix?
=20
Thanks in advance,
Gerardo

Re: [cx-oracle-users] cx_oracle, returning insert id

[kellie h. <kel...@ya...>]

It works! Great! Thank you.

Kellie

--- Anthony Tuininga <ant...@gm...> wrote:

> Sure. Its quite straightforward but not covered by the DB API.
> 
> # this creates the bind variable for use by Oracle
> idVar = cursor.var(cx_Oracle.NUMBER)
> 
> # execute the statement exactly as you normally would,
> # binding the variable you just created above
> cursor.execute("""
>             insert into SomeTable (
>               SomeOtherColumn
>             ) values (
>                :someValue
>             ) returning Id
>               into :id""",
>             id = idVar,
>             someValue = "SomeValue")
> 
> # get the value after the statement is executed
> id = idVar.getvalue()
> 
> Hope that explains things.
> 
> On 3/22/06, kellie hobbs <kel...@ya...> wrote:
> > Hello,
> >
> > I am trying to get back the id of a record after inserting it into
> the
> > table. My Oracle database has a trigger that automatically creates
> the
> > id on insert. The PL/SQL statement that works for me is:
> > SQL> declare id int;
> >   2  begin
> >   3  insert into table (field) values ('foo') returning tableid
> into
> > id;
> >   4  dbms_output.put_line(id);
> >   5  end;
> >   6  /
> >
> > How can I run this through cx_oracle and get the id stored in a
> > variable so I can use it?
> >
> > Many thanks.
> >
> > Kellie Hobbs, UC Berkeley
> >
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam?  Yahoo! Mail has the best spam protection around
> > http://mail.yahoo.com
> >
> >
> > -------------------------------------------------------
> > This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> > that extends applications into web and mobile media. Attend the
> live webcast
> > and join the prime developer group breaking into this new coding
> territory!
> >
>
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=110944&bid=241720&dat=121642
> > _______________________________________________
> > cx-oracle-users mailing list
> > cx-...@li...
> > https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
> >
> 
> 
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting
> language
> that extends applications into web and mobile media. Attend the live
> webcast
> and join the prime developer group breaking into this new coding
> territory!
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid0944&bid$1720&dat1642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
> 


__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [cx-oracle-users] cx_oracle, returning insert id

[Anthony T. <ant...@gm...>]

Sure. Its quite straightforward but not covered by the DB API.

# this creates the bind variable for use by Oracle
idVar =3D cursor.var(cx_Oracle.NUMBER)

# execute the statement exactly as you normally would,
# binding the variable you just created above
cursor.execute("""
            insert into SomeTable (
              SomeOtherColumn
            ) values (
               :someValue
            ) returning Id
              into :id""",
            id =3D idVar,
            someValue =3D "SomeValue")

# get the value after the statement is executed
id =3D idVar.getvalue()

Hope that explains things.

On 3/22/06, kellie hobbs <kel...@ya...> wrote:
> Hello,
>
> I am trying to get back the id of a record after inserting it into the
> table. My Oracle database has a trigger that automatically creates the
> id on insert. The PL/SQL statement that works for me is:
> SQL> declare id int;
>   2  begin
>   3  insert into table (field) values ('foo') returning tableid into
> id;
>   4  dbms_output.put_line(id);
>   5  end;
>   6  /
>
> How can I run this through cx_oracle and get the id stored in a
> variable so I can use it?
>
> Many thanks.
>
> Kellie Hobbs, UC Berkeley
>
> __________________________________________________
> Do You Yahoo!?
> Tired of spam?  Yahoo! Mail has the best spam protection around
> http://mail.yahoo.com
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting langua=
ge
> that extends applications into web and mobile media. Attend the live webc=
ast
> and join the prime developer group breaking into this new coding territor=
y!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D110944&bid=3D241720&dat=
=3D121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

[cx-oracle-users] cx_oracle, returning insert id

[kellie h. <kel...@ya...>]

Hello,

I am trying to get back the id of a record after inserting it into the
table. My Oracle database has a trigger that automatically creates the
id on insert. The PL/SQL statement that works for me is:
SQL> declare id int;
  2  begin
  3  insert into table (field) values ('foo') returning tableid into
id;
  4  dbms_output.put_line(id);
  5  end;
  6  /

How can I run this through cx_oracle and get the id stored in a
variable so I can use it?

Many thanks.

Kellie Hobbs, UC Berkeley

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

Re: [cx-oracle-users] SEGV

[Lars I. <la...@ib...>]

Dear Anthony,

> Just the obvious question (which I suspect the answer to is yes but
> just to be sure). Are you using the "threaded = True" parameter to
> your connection statements?

We are using SessionPools, and we do supply threaded=True, yes.

> Another user has noticed an intermittent segfault due to faulty module
> initialization. I have fixed the problem but that code is unreleased
> as of yet. I can give you the source or binary for that if you want to
> try that.

That would be good, and source would be best. I'd like to compile 
against the Oracle version I have installed.

I couldn't reproduce the problem with 4.1.2 yet, but that doesn't mean 
much - my current tests hasn't run nearly long enough yet.

- Lars

Re: [cx-oracle-users] SEGV

[Anthony T. <ant...@gm...>]

Just the obvious question (which I suspect the answer to is yes but
just to be sure). Are you using the "threaded =3D True" parameter to
your connection statements?

Another user has noticed an intermittent segfault due to faulty module
initialization. I have fixed the problem but that code is unreleased
as of yet. I can give you the source or binary for that if you want to
try that.

On 3/20/06, Lars Immisch <la...@ib...> wrote:
> Dear all,
>
> I am getting a SEGV under load. This was generated using
> cx_Oracle-4.1-b1 on Linux, compiled against the full Oracle client 10.1
> on Linux (Debian sarge).
>
> The test code is multithreaded. The problem occurred after several hours
> of load.
>
> I have updated to cx_Oracle 4.1.2, compiled with the 10.2 instantclient
> SDK and am waiting whether the problem raises its ugly head again.
>
> In the meantime, I'll appreciate any ideas or suggestions.
>
> #0  0x40ce95c0 in kgesev ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> (gdb) bt
> #0  0x40ce95c0 in kgesev ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> #1  0x40ce93e7 in kgesec0 ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> #2  0x40a32690 in kpmterm ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> #3  0x408c8226 in kpufhndl0 ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> #4  0x408c79de in kpufhndl ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> #5  0x409999bc in OCIHandleFree ()
>    from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
> #6  0x40708464 in initcx_Oracle ()
>    from /usr/lib/python2.4/site-packages/cx_Oracle.so
> #7  0x407086cd in initcx_Oracle ()
>    from /usr/lib/python2.4/site-packages/cx_Oracle.so
> #8  0x407062f1 in ExternalLobVar_New ()
>    from /usr/lib/python2.4/site-packages/cx_Oracle.so
> #9  0x0808bf25 in _PyObject_SlotCompare ()
> #10 0x0805c449 in PyObject_Call ()
> #11 0x40707476 in ExternalLobVar_New ()
>    from /usr/lib/python2.4/site-packages/cx_Oracle.so
> #12 0x080af502 in PyEval_CallObjectWithKeywords ()
> #13 0x080ad9b7 in PyEval_EvalFrame ()
> #14 0x080af8cd in PyEval_CallObjectWithKeywords ()
> #15 0x080af464 in PyEval_CallObjectWithKeywords ()
> #16 0x080ad9b7 in PyEval_EvalFrame ()
> #17 0x080af8cd in PyEval_CallObjectWithKeywords ()
> #18 0x080af464 in PyEval_CallObjectWithKeywords ()
> #19 0x080ad9b7 in PyEval_EvalFrame ()
> #20 0x080af8cd in PyEval_CallObjectWithKeywords ()
> #21 0x080af464 in PyEval_CallObjectWithKeywords ()
> #22 0x080ad9b7 in PyEval_EvalFrame ()
> #23 0x080ae42e in PyEval_EvalCodeEx ()
> #24 0x08106d47 in PyStaticMethod_New ()
> #25 0x0805c449 in PyObject_Call ()
> #26 0x080af2f7 in PyEval_CallObjectWithKeywords ()
> #27 0x0805c40e in PyObject_CallObject ()
> #28 0x402de5f7 in app_thread (evt=3D0xa9c5890c) at sai_wrap.c:3465
> #29 0x40457de3 in appProcWrapper (pPtr=3D0xa9c588a8) at utils.c:74
> #30 0x40022e51 in pthread_start_thread () from /lib/libpthread.so.0
> #31 0x4016e92a in clone () from /lib/libc.so.
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting langua=
ge
> that extends applications into web and mobile media. Attend the live webc=
ast
> and join the prime developer group breaking into this new coding territor=
y!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D110944&bid=3D241720&dat=
=3D121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

[cx-oracle-users] SEGV

[Lars I. <la...@ib...>]

Dear all,

I am getting a SEGV under load. This was generated using
cx_Oracle-4.1-b1 on Linux, compiled against the full Oracle client 10.1
on Linux (Debian sarge).

The test code is multithreaded. The problem occurred after several hours
of load.

I have updated to cx_Oracle 4.1.2, compiled with the 10.2 instantclient
SDK and am waiting whether the problem raises its ugly head again.

In the meantime, I'll appreciate any ideas or suggestions.

#0  0x40ce95c0 in kgesev ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
(gdb) bt
#0  0x40ce95c0 in kgesev ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
#1  0x40ce93e7 in kgesec0 ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
#2  0x40a32690 in kpmterm ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
#3  0x408c8226 in kpufhndl0 ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
#4  0x408c79de in kpufhndl ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
#5  0x409999bc in OCIHandleFree ()
   from /opt/oracle/instantclient_10_2/libclntsh.so.10.1
#6  0x40708464 in initcx_Oracle ()
   from /usr/lib/python2.4/site-packages/cx_Oracle.so
#7  0x407086cd in initcx_Oracle ()
   from /usr/lib/python2.4/site-packages/cx_Oracle.so
#8  0x407062f1 in ExternalLobVar_New ()
   from /usr/lib/python2.4/site-packages/cx_Oracle.so
#9  0x0808bf25 in _PyObject_SlotCompare ()
#10 0x0805c449 in PyObject_Call ()
#11 0x40707476 in ExternalLobVar_New ()
   from /usr/lib/python2.4/site-packages/cx_Oracle.so
#12 0x080af502 in PyEval_CallObjectWithKeywords ()
#13 0x080ad9b7 in PyEval_EvalFrame ()
#14 0x080af8cd in PyEval_CallObjectWithKeywords ()
#15 0x080af464 in PyEval_CallObjectWithKeywords ()
#16 0x080ad9b7 in PyEval_EvalFrame ()
#17 0x080af8cd in PyEval_CallObjectWithKeywords ()
#18 0x080af464 in PyEval_CallObjectWithKeywords ()
#19 0x080ad9b7 in PyEval_EvalFrame ()
#20 0x080af8cd in PyEval_CallObjectWithKeywords ()
#21 0x080af464 in PyEval_CallObjectWithKeywords ()
#22 0x080ad9b7 in PyEval_EvalFrame ()
#23 0x080ae42e in PyEval_EvalCodeEx ()
#24 0x08106d47 in PyStaticMethod_New ()
#25 0x0805c449 in PyObject_Call ()
#26 0x080af2f7 in PyEval_CallObjectWithKeywords ()
#27 0x0805c40e in PyObject_CallObject ()
#28 0x402de5f7 in app_thread (evt=0xa9c5890c) at sai_wrap.c:3465
#29 0x40457de3 in appProcWrapper (pPtr=0xa9c588a8) at utils.c:74
#30 0x40022e51 in pthread_start_thread () from /lib/libpthread.so.0
#31 0x4016e92a in clone () from /lib/libc.so.

Re: [cx-oracle-users] bug in Cursor_Execute when keywordArgs is empty

[Anthony T. <ant...@gm...>]

I have integrated the following patch (the line numbers won't match
yours because I have other changes that I have already made) and added
a test suite to verify that it works properly.

--- Cursor.c    10 Mar 2006 16:38:55 -0000      1.95
+++ Cursor.c    13 Mar 2006 15:26:46 -0000
@@ -1317,9 +1317,13 @@
         return NULL;
     }
     if (executeArgs && keywordArgs) {
-        PyErr_SetString(g_InterfaceErrorException,
-                "expecting argument or keyword arguments, not both");
-        return NULL;
+        if (PyDict_Size(keywordArgs) =3D=3D 0)
+            keywordArgs =3D NULL;
+        else {
+            PyErr_SetString(g_InterfaceErrorException,
+                    "expecting argument or keyword arguments, not both");
+            return NULL;
+        }
     }
     if (keywordArgs)
         executeArgs =3D keywordArgs;

On 3/11/06, Fabien Grumelard <fab...@gr...> wrote:
> Hello,
>
> I use cx_Oracle 4.1.2 with DBUtils. There was an error when I wanted to
> execute a query with a sequence parameter. For example, where 'cursor' is
> a cursor from a PooledDB connection:
> cursor.execute("select * from foo where login=3D:1", (['fabix']))
> produced a cx_Oracle.InterfaceError in
> /usr/lib/python2.3/site-packages/DBUtils/SolidDB.py in
> tough_method(*args, **kwargs) line 243:
> --> r =3D method(*args, **kwargs) # try to execute
> I found the problem is passing kwargs whereas it's empty.
> I changed Cursor.c to solve it (from line 1276) :
> ----------
>     if (executeArgs && keywordArgs) {
>         if (PyDict_Size(keywordArgs) > 0) {
>                 PyErr_SetString(g_InterfaceErrorException,
>                         "expecting argument or keyword arguments, not bot=
h");
>                 return NULL;
>         }
>     }
>     if (!executeArgs && keywordArgs)
>         executeArgs =3D keywordArgs;
> ----------
>
> Maybe someone else can be interested by this "patch".
>
> Regards,
> Fabien Grumelard
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by xPML, a groundbreaking scripting langua=
ge
> that extends applications into web and mobile media. Attend the live webc=
ast
> and join the prime developer group breaking into this new coding territor=
y!
> http://sel.as-us.falkag.net/sel?cmd=3Dlnk&kid=3D110944&bid=3D241720&dat=
=3D121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

[cx-oracle-users] bug in Cursor_Execute when keywordArgs is empty

[Fabien G. <fab...@gr...>]

Hello,

I use cx_Oracle 4.1.2 with DBUtils. There was an error when I wanted to
execute a query with a sequence parameter. For example, where 'cursor' is
a cursor from a PooledDB connection:
cursor.execute("select * from foo where login=:1", (['fabix']))
produced a cx_Oracle.InterfaceError in
/usr/lib/python2.3/site-packages/DBUtils/SolidDB.py in
tough_method(*args, **kwargs) line 243:
--> r = method(*args, **kwargs) # try to execute
I found the problem is passing kwargs whereas it's empty.
I changed Cursor.c to solve it (from line 1276) :
----------
    if (executeArgs && keywordArgs) {
	if (PyDict_Size(keywordArgs) > 0) {
		PyErr_SetString(g_InterfaceErrorException,
			"expecting argument or keyword arguments, not both");
		return NULL;
	}
    }
    if (!executeArgs && keywordArgs)
	executeArgs = keywordArgs;
----------

Maybe someone else can be interested by this "patch".

Regards,
Fabien Grumelard

Re: [cx-oracle-users] Oracle instant client / Solaris / connecting to 9i database

[Anthony T. <ant...@gm...>]

Hmm, sounds like maybe you have the wrong module? When you built
cx_Oracle there is no way it would have linked properly without the
libclntsh.so.9.0 shared library. You want to make sure that when it
links it links against the Oracle 10g client library.

On your other issue, this piece of information may (or may not) help.
From Python 2.4 onwards, the type of date returned by cx_Oracle is
datetime.datetime whereas before Python 2.4 it returned
cx_Oracle.Timestamp. Did you switch Python versions at the same time?

On 2/22/06, Hubert B Hickman <hhi...@un...> wrote:
>
> I have compiled cx_oracle 4.1.2 for Solaris 8 using the Oracle 10g instan=
t
> client.  I had to do a bit of changing in the setup.py file to specify th=
e
> correct include and library paths.
>
> When I try to connect to an Oracle 9i database, I get the following error=
:
>
>   File
> "/export/home0/hhickman_testing/work/ResultsByOrder/ResultsDb.py",
> line 9, in ?
>     import cx_Oracle
> ImportError: ld.so.1: python: fatal: libclntsh.so.9.0: open failed: No su=
ch
> file or directory
>
> If I do a symbolic link to link the 10.0 libclntsh to the 9.0 version, th=
en
> my program (mostly) works, but I eventually get to the following problem.
> TypeError: can't compare datetime.datetime to cx_Oracle.Timestamp
>
> This same program works
> I can download and install the correct 9i client and build cx_Oracle that
> way, but it would be nice to get this going with the instant client!
>
> Thanks!
>
> Hubert B. Hickman
>
>  Email : hhi...@un...

[cx-oracle-users] Oracle instant client / Solaris / connecting to 9i database

[Hubert B H. <hhi...@un...>]

I have compiled cx_oracle 4.1.2 for Solaris 8 using the Oracle 10g instant 
client.  I had to do a bit of changing in the setup.py file to specify the 
correct include and library paths. 

When I try to connect to an Oracle 9i database, I get the following error:

  File "/export/home0/hhickman_testing/work/ResultsByOrder/ResultsDb.py", 
line 9, in ?
    import cx_Oracle
ImportError: ld.so.1: python: fatal: libclntsh.so.9.0: open failed: No 
such file or directory

If I do a symbolic link to link the 10.0 libclntsh to the 9.0 version, 
then my program (mostly) works, but I eventually get to the following 
problem.
TypeError: can't compare datetime.datetime to cx_Oracle.Timestamp

This same program works 
I can download and install the correct 9i client and build cx_Oracle that 
way, but it would be nice to get this going with the instant client!

Thanks!

Hubert B. Hickman

Email : hhi...@un...

Re: [cx-oracle-users] 10g oci.dll on 8.1.7

[JP G. <jpg...@gm...>]

Thanks for your help! The 9i client install is working nicely with
cx_Oracle and our server.

Cheers!
JP

On 2/22/06, Amaury Forgeotdarc <Ama...@gl...> wrote:
> JP Glutting wrote:
> > Ok, I am back with another attempt to use cx_Oracle. I have downloaded
> > the Oracle              Database 10g Express Edition from the website.
> > So, I have a match between the cx_Oracle module (I
> > uninstalled the old one and installed the 10g version) and
> > the oci.dll library. cx_Oracle imports fine.
> >
> > Now, I am fiddling around with the connection, but it is not
> > clear that I can connect to an Oracle 8.1.7 database server
> > with a 10g client library. I am hoping it is backward
> > compatible (the earliest version available on the Oracle site
> > is 9i, so I can't get a matching version).
>
> In principle, Oracle versions try to be interoperable.
> In practice, a jump of two versions can have problems.
>
> > Am I out of luck? I am getting errors when trying to connect,
> > but before I try to work out the correct protocol, can
> > someone tell me if this is even possible?
>
> On Oracle support site, I found this statement:
> """
> An ORA-3134 error is incorrectly reported if a 10g client tries to
> connect to an 8.1.7.3 or lower server
> """
> Do you get this error?
>
> You may try to use Oracle 9 client instead.
> We have successfully used versions 9.2.0.4 and up to connect to various
> databases: 8i, 9i, 10g.
>
> Hope this helps,
>
> --
> Amaury
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=103432&bid#0486&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

Re: [cx-oracle-users] 10g oci.dll on 8.1.7

[JP G. <jpg...@gm...>]

And the error I got was 03114 Unexpected end of page error. But I will
try the version that is closer to what we have anyway.

Cheers,
JP

On 2/22/06, Amaury Forgeotdarc <Ama...@gl...> wrote:
> JP Glutting wrote:
> > Ok, I am back with another attempt to use cx_Oracle. I have downloaded
> > the Oracle              Database 10g Express Edition from the website.
> > So, I have a match between the cx_Oracle module (I
> > uninstalled the old one and installed the 10g version) and
> > the oci.dll library. cx_Oracle imports fine.
> >
> > Now, I am fiddling around with the connection, but it is not
> > clear that I can connect to an Oracle 8.1.7 database server
> > with a 10g client library. I am hoping it is backward
> > compatible (the earliest version available on the Oracle site
> > is 9i, so I can't get a matching version).
>
> In principle, Oracle versions try to be interoperable.
> In practice, a jump of two versions can have problems.
>
> > Am I out of luck? I am getting errors when trying to connect,
> > but before I try to work out the correct protocol, can
> > someone tell me if this is even possible?
>
> On Oracle support site, I found this statement:
> """
> An ORA-3134 error is incorrectly reported if a 10g client tries to
> connect to an 8.1.7.3 or lower server
> """
> Do you get this error?
>
> You may try to use Oracle 9 client instead.
> We have successfully used versions 9.2.0.4 and up to connect to various
> databases: 8i, 9i, 10g.
>
> Hope this helps,
>
> --
> Amaury
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=103432&bid#0486&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

Re: [cx-oracle-users] 10g oci.dll on 8.1.7

[JP G. <jpg...@gm...>]

That helps a lot, thank you! I will uninstall and try version 9i.

Thanks again,
JP

On 2/22/06, Amaury Forgeotdarc <Ama...@gl...> wrote:
> JP Glutting wrote:
> > Ok, I am back with another attempt to use cx_Oracle. I have downloaded
> > the Oracle              Database 10g Express Edition from the website.
> > So, I have a match between the cx_Oracle module (I
> > uninstalled the old one and installed the 10g version) and
> > the oci.dll library. cx_Oracle imports fine.
> >
> > Now, I am fiddling around with the connection, but it is not
> > clear that I can connect to an Oracle 8.1.7 database server
> > with a 10g client library. I am hoping it is backward
> > compatible (the earliest version available on the Oracle site
> > is 9i, so I can't get a matching version).
>
> In principle, Oracle versions try to be interoperable.
> In practice, a jump of two versions can have problems.
>
> > Am I out of luck? I am getting errors when trying to connect,
> > but before I try to work out the correct protocol, can
> > someone tell me if this is even possible?
>
> On Oracle support site, I found this statement:
> """
> An ORA-3134 error is incorrectly reported if a 10g client tries to
> connect to an 8.1.7.3 or lower server
> """
> Do you get this error?
>
> You may try to use Oracle 9 client instead.
> We have successfully used versions 9.2.0.4 and up to connect to various
> databases: 8i, 9i, 10g.
>
> Hope this helps,
>
> --
> Amaury
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=103432&bid#0486&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

RE: [cx-oracle-users] 10g oci.dll on 8.1.7

[Amaury F. <Ama...@gl...>]

JP Glutting wrote:
> Ok, I am back with another attempt to use cx_Oracle. I have downloaded
> the Oracle              Database 10g Express Edition from the website.
> So, I have a match between the cx_Oracle module (I=20
> uninstalled the old one and installed the 10g version) and=20
> the oci.dll library. cx_Oracle imports fine.
>=20
> Now, I am fiddling around with the connection, but it is not=20
> clear that I can connect to an Oracle 8.1.7 database server=20
> with a 10g client library. I am hoping it is backward=20
> compatible (the earliest version available on the Oracle site=20
> is 9i, so I can't get a matching version).

In principle, Oracle versions try to be interoperable.
In practice, a jump of two versions can have problems.

> Am I out of luck? I am getting errors when trying to connect,=20
> but before I try to work out the correct protocol, can=20
> someone tell me if this is even possible?

On Oracle support site, I found this statement:
"""
An ORA-3134 error is incorrectly reported if a 10g client tries to
connect to an 8.1.7.3 or lower server
"""
Do you get this error?

You may try to use Oracle 9 client instead.
We have successfully used versions 9.2.0.4 and up to connect to various
databases: 8i, 9i, 10g.

Hope this helps,

--=20
Amaury

[cx-oracle-users] 10g oci.dll on 8.1.7

[JP G. <jpg...@gm...>]

Ok, I am back with another attempt to use cx_Oracle. I have downloaded
the Oracle              Database 10g Express Edition from the website.
So, I have a match between the cx_Oracle module (I uninstalled the old
one and installed the 10g version) and the oci.dll library. cx_Oracle
imports fine.

Now, I am fiddling around with the connection, but it is not clear
that I can connect to an Oracle 8.1.7 database server with a 10g
client library. I am hoping it is backward compatible (the earliest
version available on the Oracle site is 9i, so I can't get a matching
version).

Am I out of luck? I am getting errors when trying to connect, but
before I try to work out the correct protocol, can someone tell me if
this is even possible?

Many thanks!

JP

RE: [cx-oracle-users] Calling PL/SQL package.procedure(...)

[Orr, S. <so...@ri...>]

PERFECT! And nice explanation.

Thanx!

-----Original Message-----
From: cx-...@li...
[mailto:cx-...@li...] On Behalf Of
Anthony Tuininga
Sent: Tuesday, February 14, 2006 8:11 AM
To: cx-...@li...
Subject: Re: [cx-oracle-users] Calling PL/SQL package.procedure(...)

For procedures perform the following:

cursor.callproc("MyPackage.MyProcedure", (arg1, arg2, arg3))

The example you give is a little more complicated, though. First of
all, booleans cannot be bound by the OCI, only within PL/SQL, so your
example of passing block_sample as true cannot be done using
callproc(). Second, referencing package constants cannot be done
directly since Python doesn't know anything about that -- to do that
with callproc you'd have to first grab the constant using an anonymous
PL/SQL block. Third, the method of calling is positional only, not
named as you have indicated.

So, for your case you need to call an anonymous PL/SQL block as in the
following:

cursor.execute("""
begin
  dbms_stats.gather_schema_stats(:name, dbms_stats.auto_sample_size,
      block_sample =3D> true, options =3D> :options);
end;""",
name =3D 'name',
options =3D 'gather stale')

Hope that helps further your understanding.

On 2/13/06, Orr, Steve <so...@ri...> wrote:
> How do I call an Oracle supplied PL/SQL package.procedure?
>
> Specifically, from SQL*Plus I do this:
> dbms_stats.gather_schema_stats('name', dbms_stats.auto_sample_size, -
> block_sample =3D> true, options =3D> 'gather stale');
>
> How can I call this using cx_Oracle? I see the cursor.callproc(...)
> function but no example of how to use it.
>
> Help! and AtDhVaAnNkCsE !!!
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log
files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD
SPLUNK!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=103432&bid#0486&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>


-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log
files
for problems?  Stop!  Download the new AJAX search engine that makes
searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=3Dk&kid=103432&bid#0486&dat=121642
_______________________________________________
cx-oracle-users mailing list
cx-...@li...
https://lists.sourceforge.net/lists/listinfo/cx-oracle-users

Re: [cx-oracle-users] Calling PL/SQL package.procedure(...)

[Anthony T. <ant...@gm...>]

For procedures perform the following:

cursor.callproc("MyPackage.MyProcedure", (arg1, arg2, arg3))

The example you give is a little more complicated, though. First of
all, booleans cannot be bound by the OCI, only within PL/SQL, so your
example of passing block_sample as true cannot be done using
callproc(). Second, referencing package constants cannot be done
directly since Python doesn't know anything about that -- to do that
with callproc you'd have to first grab the constant using an anonymous
PL/SQL block. Third, the method of calling is positional only, not
named as you have indicated.

So, for your case you need to call an anonymous PL/SQL block as in the
following:

cursor.execute("""
begin
  dbms_stats.gather_schema_stats(:name, dbms_stats.auto_sample_size,
      block_sample =3D> true, options =3D> :options);
end;""",
name =3D 'name',
options =3D 'gather stale')

Hope that helps further your understanding.

On 2/13/06, Orr, Steve <so...@ri...> wrote:
> How do I call an Oracle supplied PL/SQL package.procedure?
>
> Specifically, from SQL*Plus I do this:
> dbms_stats.gather_schema_stats('name', dbms_stats.auto_sample_size, -
> block_sample =3D> true, options =3D> 'gather stale');
>
> How can I call this using cx_Oracle? I see the cursor.callproc(...)
> function but no example of how to use it.
>
> Help! and AtDhVaAnNkCsE !!!
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log fi=
les
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=103432&bid#0486&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

[cx-oracle-users] Calling PL/SQL package.procedure(...)

[Orr, S. <so...@ri...>]

How do I call an Oracle supplied PL/SQL package.procedure?

Specifically, from SQL*Plus I do this:=20
dbms_stats.gather_schema_stats('name', dbms_stats.auto_sample_size, -
block_sample =3D> true, options =3D> 'gather stale');

How can I call this using cx_Oracle? I see the cursor.callproc(...)
function but no example of how to use it.

Help! and AtDhVaAnNkCsE !!!

Re: [cx-oracle-users] Re: DLL Error - python 2.4, Oracle 8.1.7, cx_Oracle

[JP G. <jpg...@gm...>]

Thanks Henning. I ran the script right after you sent it. I have the right
version of cx_Oracle, but the ori.dll is old. It is either form a 6i
installation of Oracle Report (which is doubtful, but I have it here on thi=
s
computer) or an 8.0 version of SQL Plus.

Either way, it is not going to work. I called the guy who set me up with
server access to ask if they have a newer version they can let me install.

Thanks for your help!

JP

On 2/9/06, Henning von Bargen <H.v...@t-...> wrote:
>
> Please run the script I suggested, just to be sure.
>
> It is almost certainly an unclean Oracle environment
> which is causing the error.
>
> And remember that Oracle uses the Windows registry
> AND the environment variables!
>
> Are you sure you're using a cx_Oracle version compiled for Oracle 8.1.7?
>
> Is the ImportError the only error message or is there a
> Windows error message box, too?
>
> Henning
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: Splunk Inc. Do you grep through log
> files
> for problems?  Stop!  Download the new AJAX search engine that makes
> searching your log files as easy as surfing the  web.  DOWNLOAD SPLUNK!
> http://sel.as-us.falkag.net/sel?cmdlnk&kid=103432&bid#0486&dat=121642
> _______________________________________________
> cx-oracle-users mailing list
> cx-...@li...
> https://lists.sourceforge.net/lists/listinfo/cx-oracle-users
>

[cx-oracle-users] Re: DLL Error - python 2.4, Oracle 8.1.7, cx_Oracle

[Henning v. B. <H.v...@t-...>]

Please run the script I suggested, just to be sure.

It is almost certainly an unclean Oracle environment
which is causing the error.

And remember that Oracle uses the Windows registry
AND the environment variables!

Are you sure you're using a cx_Oracle version compiled for Oracle 8.1.7?

Is the ImportError the only error message or is there a
Windows error message box, too?

Henning