sqlobject-discuss — All things SQLObject-related

Re: [SQLObject] SQLObject 1.0.0 problem with Py 2.4

[Christoph Z. <ci...@on...>]

Am 26.05.2011 21:47 schrieb Oleg Broytman:
 >     There is no need for opinions - I'm going to restore Python 2.4
 > compatibility and release SQLObject 1.0.1, just not immediately

Ok, thanks. If you need any help with testing, let me know.

-- Christoph

Re: [SQLObject] SQLObject 1.0.0 problem with Py 2.4

[Oleg B. <ph...@ph...>]

On Thu, May 26, 2011 at 09:20:11PM +0200, Christoph Zwerschke wrote:
> Since this involves discussing which Python versions are supported, 
> maybe it's better to discuss it here so people can voice their opinions.

   There is no need for opinions - I'm going to restore Python 2.4
compatibility and release SQLObject 1.0.1, just not immediately - I've
upgraded to Debian 6.0 and lost some python modules for older versions,
including PySQLite2. As of tests - of course I always run tests with all
supported python versions and with SQLite and PostgreSQL backends; there
are many tests so I developed a set of shell scripts that run tests and
it seems there is a bug in the test system somewhere as the scripts
didn't report the problem.
   Quis custodiet ipsos custodes? Who does fix bugs in test? ;-)

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

Re: [SQLObject] SQLObject 1.0.0 problem with Py 2.4

[Christoph Z. <ci...@on...>]

Am 26.05.2011 18:46 schrieb Oleg Broytman:
> Thank you for spotting this! Where do you prefer to discuss the issue
> - in the mailing list or in the SF tracker?

Since this involves discussing which Python versions are supported, 
maybe it's better to discuss it here so people can voice their opinions.

-- Christoph

Re: [SQLObject] SQLObject 1.0.0 problem with Py 2.4

[Oleg B. <ph...@ph...>]

On Thu, May 26, 2011 at 06:34:55PM +0200, Christoph Zwerschke wrote:
> Just noticed that SQLObject 1.0.0 does not work with Py 2.4 any more 
> which also breaks Py 2.4 compatibility for TurboGears 1.
> 
> Can somebody look into this and either release a bugfix or declare Py 
> 2.4 as officially unsupported?
> 
> I've already created a bug report at
> https://sourceforge.net/tracker/?group_id=74338&atid=540672

   Thank you for spotting this! Where do you prefer to discuss the issue -
in the mailing list or in the SF tracker?

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

[SQLObject] SQLObject 1.0.0 problem with Py 2.4

[Christoph Z. <ci...@on...>]

Just noticed that SQLObject 1.0.0 does not work with Py 2.4 any more 
which also breaks Py 2.4 compatibility for TurboGears 1.

Can somebody look into this and either release a bugfix or declare Py 
2.4 as officially unsupported?

I've already created a bug report at
https://sourceforge.net/tracker/?group_id=74338&atid=540672

-- Christoph

Re: [SQLObject] py3k

[Oleg B. <ph...@ph...>]

On Mon, May 16, 2011 at 09:15:57PM +0200, Daniel Fetchinson wrote:
> > AFAIU sqlite3 is the only py3k-compatible driver now. I've added py-postgresql, MySQL
> > Connector/Python, oursql and pyfirebirdsql to TODO, but not at the very
> > top.
> 
> What about psycopg?
> 
> http://initd.org/psycopg/articles/2011/01/24/psycopg2-porting-python-3-report/

   I haven't heard the news, thank you for the reference.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

Re: [SQLObject] py3k

[Daniel F. <fet...@go...>]

> BTW, what backend are you going to use for testing?

I'll be using sqlite3.

> AFAIU sqlite3 is the only py3k-compatible driver now. I've added py-postgresql, MySQL
> Connector/Python, oursql and pyfirebirdsql to TODO, but not at the very
> top.

What about psycopg?

http://initd.org/psycopg/articles/2011/01/24/psycopg2-porting-python-3-report/

Cheers,
Daniel

-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

Re: [SQLObject] py3k

[Oleg B. <ph...@ph...>]

BTW, what backend are you going to use for testing? AFAIU sqlite3 is the
only py3k-compatible driver now. I've added py-postgresql, MySQL
Connector/Python, oursql and pyfirebirdsql to TODO, but not at the very
top.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

Re: [SQLObject] python 3 compatibility patch

[Oleg B. <ph...@ph...>]

On Sun, May 15, 2011 at 04:10:04PM +0200, Daniel Fetchinson wrote:
> I planning to send a series of patches for moving closer to python 3
> compatibility as far as it is possible with python 2.x
> 
> This first patch changes all 'mydict.has_key(name)' type of checks
> with 'name in mydict'.

   Applied and committed in the revision 4387 in the trunk. Thank you!

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

[SQLObject] python 3 compatibility patch

[Daniel F. <fet...@go...>]

I planning to send a series of patches for moving closer to python 3
compatibility as far as it is possible with python 2.x

This first patch changes all 'mydict.has_key(name)' type of checks
with 'name in mydict'.

Cheers,
Daniel


-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

Re: [SQLObject] py.test version

[Daniel F. <fet...@go...>]

> I'll some the patches soon,

s/some/send/

Cheers,
Daniel


-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

Re: [SQLObject] py.test version

[Daniel F. <fet...@go...>]

>> So basically my question is what versions of the test framework
>> 'py.test' do you use for the tests
>
>    I use py.test from SVN, currently at revision 80343; I think it's
> version 1.0.1 or 1.0.2. I can upgrade to 1.3.3 (the version in Debian 6.0).

I downgraded to py-1.1.1 at which version 'py' and 'py.test' were both
included in 'py'. And now things work with python 2.7 just like with
python 2.6. So I guess there was some non-backward compatible change
in 'py' and/or 'py.test' recently.

In any case the reason I needed python 2.7 is that I can test python 3
related stuff with 2.7.

>> and how do you run them?
>
>    I use a number of shell scripts. Will send them to you.

Thanks!

I'll some the patches soon,
Daniel


-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

Re: [SQLObject] py.test version

[Oleg B. <ph...@ph...>]

On Sun, May 15, 2011 at 01:50:24PM +0200, Daniel Fetchinson wrote:
> So basically my question is what versions of the test framework
> 'py.test' do you use for the tests

   I use py.test from SVN, currently at revision 80343; I think it's
version 1.0.1 or 1.0.2. I can upgrade to 1.3.3 (the version in Debian 6.0).

> and how do you run them?

   I use a number of shell scripts. Will send them to you.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

[SQLObject] py.test version

[Daniel F. <fet...@go...>]

Hi Oleg,

I'm trying to move sqlobject closer to python 3 and plan to submit a
couple of patches. But before doing so I wanted to run the test suite
and ran into a bit of a mess with py.test versions.

With python 2.6 and py 1.1.1 I can run the tests but with python 2.7
and py 1.4.3 and py.test 2.0.3 even 'py.test --help' doesn't work:

[fetchinson@fetch sqlobject]$ py.test --version
Traceback (most recent call last):
  File "/usr/local/bin/py.test", line 8, in <module>
    load_entry_point('pytest==2.0.3', 'console_scripts', 'py.test')()
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/core.py",
line 457, in main
    pluginmanager=_pluginmanager, args=args)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/core.py",
line 412, in __call__
    return self._docall(methods, kwargs)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/core.py",
line 423, in _docall
    res = mc.execute()
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/core.py",
line 341, in execute
    res = method(**kwargs)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 10, in pytest_cmdline_parse
    config.parse(args)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 337, in parse
    self._preparse(args)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 316, in _preparse
    self._setinitialconftest(args)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 295, in _setinitialconftest
    self._conftest.setinitial(args)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 159, in setinitial
    self._path2confmods[None] = self.getconftestmodules(anchor)
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 185, in getconftestmodules
    clist.append(self.importconftest(conftestpath))
  File "/usr/local/lib/python2.7/site-packages/pytest-2.0.3-py2.7.egg/_pytest/config.py",
line 214, in importconftest
    self._conftestpath2mod[conftestpath] = mod = conftestpath.pyimport()
  File "/usr/local/lib/python2.7/site-packages/py-1.4.3-py2.7.egg/py/_path/local.py",
line 529, in pyimport
    mod = __import__(modname, None, None, ['__doc__'])
  File "/home/fetchinson/files/python/svn/sqlobject/sqlobject/conftest.py",
line 34, in <module>
    Option = py.test.config.Option
  File "/usr/local/lib/python2.7/site-packages/py-1.4.3-py2.7.egg/py/_apipkg.py",
line 159, in __getattribute__
    return getattr(getmod(), name)
AttributeError: 'module' object has no attribute 'config'

The sqlobject svn source is in
/home/fetchinson/files/python/svn/sqlobject and I did a

python2.7 setup.py develop

in that directory (similarly with formencode). The installations are
both fine, with python2.7 I can import sqlobject.

With the py and py.test versions I have with python 2.6 there was only
one library, 'py' and this included py.test. But with the newer
versions that I have installed for python 2.7, they are separate,
there is a 'py' and a 'pytest' package with different version numbers.

So basically my question is what versions of the test framework
'py.test' do you use for the tests and how do you run them? I'd prefer
duplicating the exact same setup as what you have and in that way
there will be the least amount of friction between my patches and your
work flow.

Cheers,
Daniel


-- 
Psss, psss, put it down! - http://www.cafepress.com/putitdown

Re: [SQLObject] Custom query results?

[Oleg B. <ph...@ph...>]

On Wed, May 11, 2011 at 01:14:22PM -0600, Pгoм???тн?????s 0x01 wrote:
> I was wondering if it was possible to attach columns from another table

   Not with high-level API (classes/instances); see

http://sqlobject.org/FAQ.html#how-can-i-do-a-left-join

   See also sqlbuilder.Select:

http://sqlobject.org/SQLBuilder.html#select

   but it's lower-level API.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

[SQLObject] Custom query results?

[Pгoмᴇтнᴇυs 0. <pro...@gm...>]

Hi all,

I was wondering if it was possible to attach columns from another table (a
many-to-many relationship table) to results returned from er-- that other
table.

To use an example from the SQLObject documentation:
http://sqlobject.org/FAQ.html#how-can-i-define-my-own-intermediate-table-in-my-many-to-many-relationship

Is it possible to have User.roles return the Role information along with
'active' from the UserRoles class? How might this be done?

Thanks again!

Re: [SQLObject] Registry error? (class X is already in the registry)

[Pгoмᴇтнᴇυs 0. <pro...@gm...>]

Thanks; I'm gonna have to comb through that code later and try to figure out
how to avoid this.

2011/5/10 Oleg Broytman <ph...@ph...>

> On Tue, May 10, 2011 at 03:33:43PM -0600, Pгoм???тн?????s 0x01 wrote:
> > Is there something I can do to catch this? Should I be using some kind of
> > registry or dictionary to keep track of imports? (Would that help?)
>
>    In my experience, problematic import is usually a result of mixed
> absolute and relative import - a submodule A imports submodule X as
> parent.X and submodule B imports X as X; add sys.path manipulations and
> voila - X is imported both as parent.X and top-level X.
>
> Oleg.
> --
>     Oleg Broytman            http://phdru.name/            ph...@ph...
>           Programmers don't die, they just GOSUB without RETURN.
>
>
> ------------------------------------------------------------------------------
> Achieve unprecedented app performance and reliability
> What every C/C++ and Fortran developer should know.
> Learn how Intel has extended the reach of its next-generation tools
> to help boost performance applications - inlcuding clusters.
> http://p.sf.net/sfu/intel-dev2devmay
> _______________________________________________
> sqlobject-discuss mailing list
> sql...@li...
> https://lists.sourceforge.net/lists/listinfo/sqlobject-discuss
>

Re: [SQLObject] Registry error? (class X is already in the registry)

[Oleg B. <ph...@ph...>]

On Tue, May 10, 2011 at 03:33:43PM -0600, Pгoм???тн?????s 0x01 wrote:
> Is there something I can do to catch this? Should I be using some kind of
> registry or dictionary to keep track of imports? (Would that help?)

   In my experience, problematic import is usually a result of mixed
absolute and relative import - a submodule A imports submodule X as
parent.X and submodule B imports X as X; add sys.path manipulations and
voila - X is imported both as parent.X and top-level X.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

Re: [SQLObject] Registry error? (class X is already in the registry)

[Pгoмᴇтнᴇυs 0. <pro...@gm...>]

Is there something I can do to catch this? Should I be using some kind of
registry or dictionary to keep track of imports? (Would that help?)

2011/5/10 Oleg Broytman <ph...@ph...>

> On Tue, May 10, 2011 at 02:18:55PM -0600, Pгoм???тн?????s 0x01 wrote:
> > Is it possible for it to be loading the same class twice?
>
>    Yes, and bugs caused by dual import are, in my opinion, the most
> painful to debug.
>
> Oleg.
> --
>     Oleg Broytman            http://phdru.name/            ph...@ph...
>           Programmers don't die, they just GOSUB without RETURN.
>
>
> ------------------------------------------------------------------------------
> Achieve unprecedented app performance and reliability
> What every C/C++ and Fortran developer should know.
> Learn how Intel has extended the reach of its next-generation tools
> to help boost performance applications - inlcuding clusters.
> http://p.sf.net/sfu/intel-dev2devmay
> _______________________________________________
> sqlobject-discuss mailing list
> sql...@li...
> https://lists.sourceforge.net/lists/listinfo/sqlobject-discuss
>

Re: [SQLObject] Registry error? (class X is already in the registry)

[Oleg B. <ph...@ph...>]

On Tue, May 10, 2011 at 02:18:55PM -0600, Pгoм???тн?????s 0x01 wrote:
> Is it possible for it to be loading the same class twice?

   Yes, and bugs caused by dual import are, in my opinion, the most
painful to debug.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

Re: [SQLObject] Registry error? (class X is already in the registry)

[Pгoмᴇтнᴇυs 0. <pro...@gm...>]

Is it possible for it to be loading the same class twice? (I am using
raw__import__ statements) -- and if so, is there any way to use some kind of
if/else statement to avoid a dual attempt?

The big picture is that I have a uWSGI application "loader.py" which then
loads up the requested *.py file and calls a run function within that file.

So index.py has that run function, and it loads up a View object from
SQLObject and from there  determines what other additional modules to load,
etc...

2011/5/10 Oleg Broytman <ph...@ph...>

> On Tue, May 10, 2011 at 01:08:46PM -0600, Pгoм???тн?????s 0x01 wrote:
> > ValueError: class View is already in the registry (other class is , from
> the
> > module lf.db in /local/lf/db.pyc; attempted new class
>
>    Are you sure you don't have two classes named View? If not I suspect
> somewhere (may be in PHP-Python bridge) there are sys.path manipulations
> and you ended up importing the same module twice as different modules.
> Try to print sys.modules.
>
> Oleg.
> --
>     Oleg Broytman            http://phdru.name/            ph...@ph...
>           Programmers don't die, they just GOSUB without RETURN.
>
>
> ------------------------------------------------------------------------------
> Achieve unprecedented app performance and reliability
> What every C/C++ and Fortran developer should know.
> Learn how Intel has extended the reach of its next-generation tools
> to help boost performance applications - inlcuding clusters.
> http://p.sf.net/sfu/intel-dev2devmay
> _______________________________________________
> sqlobject-discuss mailing list
> sql...@li...
> https://lists.sourceforge.net/lists/listinfo/sqlobject-discuss
>

Re: [SQLObject] Registry error? (class X is already in the registry)

[Oleg B. <ph...@ph...>]

On Tue, May 10, 2011 at 01:08:46PM -0600, Pгoм???тн?????s 0x01 wrote:
> ValueError: class View is already in the registry (other class is , from the
> module lf.db in /local/lf/db.pyc; attempted new class

   Are you sure you don't have two classes named View? If not I suspect
somewhere (may be in PHP-Python bridge) there are sys.path manipulations
and you ended up importing the same module twice as different modules.
Try to print sys.modules.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.

Re: [SQLObject] Registry error? (class X is already in the registry)

[Sean D. <hal...@gm...>]

Im just guessing here, but perhaps you should try naming your class
something other than "View"

On Tue, May 10, 2011 at 12:08 PM, Pгoмᴇтнᴇυs 0x01 <pro...@gm...
> wrote:

> Hi all,
>
> First, thanks for the help and support with all my earlier queries
> (especially Oleg); you've all helped me a lot with this stuff, and I've
> gotten over all my previous problems.
>
> I currently have a new problem (I am using a wrapper to run python scripts
> like a PHP script/cgi script):
>
> Traceback (most recent call last):
>   File "/local/lf/index.py", line 38, in main
>     oContent = Content(sPageName, req);
>   File "/local/lf/content.py", line 75, in __init__
>     self.template, self.module = self.loadTemplate(self.view.template);
>   File "/local/lf/content.py", line 277, in loadTemplate
>     __import__(config.sModulePrefix + sTemplateModule);
>   File "/local/lf/lf/__init__.py", line 62, in
>     import db;
>   File "/local/lf/db.py", line 34, in
>     class View(SQLObject):
>   File
> "/usr/local/lib/python2.6/dist-packages/SQLObject-1.0.0-py2.6.egg/sqlobject/declarative.py",
> line 92, in __new__
>     cls.__classinit__(cls, new_attrs)
>   File
> "/usr/local/lib/python2.6/dist-packages/SQLObject-1.0.0-py2.6.egg/sqlobject/main.py",
> line 818, in __classinit__
>     classregistry.registry(sqlmeta.registry).addClass(cls)
>   File
> "/usr/local/lib/python2.6/dist-packages/SQLObject-1.0.0-py2.6.egg/sqlobject/classregistry.py",
> line 91, in addClass
>     '__file__', '(unknown)')))
> ValueError: class View is already in the registry (other class is , from
> the module lf.db in /local/lf/db.pyc; attempted new class
>
> Any ideas on what I can do to overcome this? I've tried to replicate this
> error in python (interactive mode), but I haven't been able to. What causes
> this error exactly?
>
>
>
> ------------------------------------------------------------------------------
> Achieve unprecedented app performance and reliability
> What every C/C++ and Fortran developer should know.
> Learn how Intel has extended the reach of its next-generation tools
> to help boost performance applications - inlcuding clusters.
> http://p.sf.net/sfu/intel-dev2devmay
> _______________________________________________
> sqlobject-discuss mailing list
> sql...@li...
> https://lists.sourceforge.net/lists/listinfo/sqlobject-discuss
>
>

[SQLObject] Registry error? (class X is already in the registry)

[Pгoмᴇтнᴇυs 0. <pro...@gm...>]

Hi all,

First, thanks for the help and support with all my earlier queries
(especially Oleg); you've all helped me a lot with this stuff, and I've
gotten over all my previous problems.

I currently have a new problem (I am using a wrapper to run python scripts
like a PHP script/cgi script):

Traceback (most recent call last):
  File "/local/lf/index.py", line 38, in main
    oContent = Content(sPageName, req);
  File "/local/lf/content.py", line 75, in __init__
    self.template, self.module = self.loadTemplate(self.view.template);
  File "/local/lf/content.py", line 277, in loadTemplate
    __import__(config.sModulePrefix + sTemplateModule);
  File "/local/lf/lf/__init__.py", line 62, in
    import db;
  File "/local/lf/db.py", line 34, in
    class View(SQLObject):
  File
"/usr/local/lib/python2.6/dist-packages/SQLObject-1.0.0-py2.6.egg/sqlobject/declarative.py",
line 92, in __new__
    cls.__classinit__(cls, new_attrs)
  File
"/usr/local/lib/python2.6/dist-packages/SQLObject-1.0.0-py2.6.egg/sqlobject/main.py",
line 818, in __classinit__
    classregistry.registry(sqlmeta.registry).addClass(cls)
  File
"/usr/local/lib/python2.6/dist-packages/SQLObject-1.0.0-py2.6.egg/sqlobject/classregistry.py",
line 91, in addClass
    '__file__', '(unknown)')))
ValueError: class View is already in the registry (other class is , from the
module lf.db in /local/lf/db.pyc; attempted new class

Any ideas on what I can do to overcome this? I've tried to replicate this
error in python (interactive mode), but I haven't been able to. What causes
this error exactly?

Re: [SQLObject] Is it possible to override the *.by<AltID> method?

[Oleg B. <ph...@ph...>]

On Sat, Apr 23, 2011 at 08:06:50PM -0600, Pгoм???тн?????s 0x01 wrote:
> Another newbie question: is it possible to override the *.by<AltID> method?

   It's hard 'cause .by* methods are created on the fly, not via
inheritance.

> I have a custom encoding for that column (it is in base64), and I can't call
> it naturally unless I can override the method. Writing my own method doesn't
> seem to work -- it doesn't get called at all!

   In this case you should rather create a validator that will encode
and decode your data. Similar to what HashCol does - see the source in
sqlobject/include/hascol.py. Validators work for all methods.

Oleg.
-- 
     Oleg Broytman            http://phdru.name/            ph...@ph...
           Programmers don't die, they just GOSUB without RETURN.