«

07.01.2011, 16:22, "Jean-Michel Pichavant" <>:
> kost BebiX wrote:
>
>> šYou're absolutely right! Now try to do except Keyerror: raise AttributeError and it will also fail. But why?
>>
>> š07.01.2011, 15:45, "Jean-Michel Pichavant" <>;:
>>> škost BebiX wrote:
>>>> ššHi everyone!
>>>> ššI just saw a bug (?) in bson.dbrefBRef.__getattr__
>>>>
>>>> ššHere's they're code:
>>>> ššššššdef __getattr__(self, key):
>>>> ššššššššššreturn self.__kwargs[key]
>>>>
>>>> ššAnd when you do copy.deepcopy on that object it will raise you KeyError. So here's a small piece of code that reproduces the problem:
>>> šfrom http://docs.python.org/reference/datamodel.html
>>>
>>> šAbout __getattr__
>>> š"This method should return the (computed) attribute value or raise an
>>> šAttributeError
>>> š<http://docs.python.org/library/exceptions.html#exceptions.AttributeError>
>>> šexception."
>>>
>>> šThe code you provided raises a KeyError thus methods such as 'getattr'
>>> šwill fail as they expect an AttributeError exception.
>>>
>>> šJM
>>>
>>> š--
>>> šhttp://mail.python.org/mailman/listinfo/python-list
>
> please don't top post
>
> It fails because you simply did not returned any value (with your class
> A example).
>
> class A(object):
> ššššdef __init__(self):
> ššššššššself.d = {}
> ššššdef __getattr__(self, key):
> šššššššštry:
> šššššššššššš*return* self.d[key]
> ššššššššexcept KeyError:
> ššššššššššššraise AttributeError
>
> works fine with deepcopy
>
> JM
>
> --
> http://mail.python.org/mailman/listinfo/python-list

Sorry for top posting, didn't know about that) I'm quote new to posting to mailing lists.

Well, actually the code you showed doesn't work)

>>> class A(object):
... def __init__(self):
... self.d = {}
... def __getattr__(self, key):
... try:
... return self.d[key]
... except KeyError:
... raise AttributeError
>>> from copy import deepcopy
>>> a = A()
>>> deepcopy(a)
Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored
Exception RuntimeError: 'maximum recursion depth exceeded while calling a Python object' in <type 'exceptions.AttributeError'> ignored
0: <__main__.A object at 0xda0250>


--
jabber:

On Fri, 07 Jan 2011 16:47:55 +0200, kost BebiX wrote:

> Well, actually the code you showed doesn't work)

Actually, it does. It just prints a warning message as well. Look
carefully:


>>>> class A(object):
> .. def __init__(self):
> .. self.d = {}
> .. def __getattr__(self, key):
> .. try:
> .. return self.d[key]
> .. except KeyError:
> .. raise AttributeError
>>>> from copy import deepcopy
>>>> a = A()
>>>> deepcopy(a)
> Exception RuntimeError: 'maximum recursion depth exceeded while calling
> a Python object' in <type 'exceptions.AttributeError'> ignored
> Exception RuntimeError: 'maximum recursion depth exceeded while calling
> a Python object' in <type 'exceptions.AttributeError'> ignored
> 0: <__main__.A object at 0xda0250>


The last thing printed is the deepcopied object.


I've tested the above code in Python versions 2.4 through 3.2 and the
only one that prints that message is 2.6.


--
Steven

07.01.2011, 17:47, "Steven D'Aprano" <steve+>:
> On Fri, 07 Jan 2011 16:47:55 +0200, kost BebiX wrote:
>
>> šWell, actually the code you showed doesn't work)
>
> Actually, it does. It just prints a warning message as well. Look
> carefully:
>
>>>>> šclass A(object):
>> š.. ššššdef __init__(self):
>> š.. ššššššššself.d = {}
>> š.. ššššdef __getattr__(self, key):
>> š.. šššššššštry:
>> š.. ššššššššššššreturn self.d[key]
>> š.. ššššššššexcept KeyError:
>> š.. ššššššššššššraise AttributeError
>>>>> šfrom copy import deepcopy
>>>>> ša = A()
>>>>> šdeepcopy(a)
>> šException RuntimeError: 'maximum recursion depth exceeded while calling
>> ša Python object' in <type 'exceptions.AttributeError'> ignored
>> šException RuntimeError: 'maximum recursion depth exceeded while calling
>> ša Python object' in <type 'exceptions.AttributeError'> ignored
>> š0: <__main__.A object at 0xda0250>
>
> The last thing printed is the deepcopied object.
>
> I've tested the above code in Python versions 2.4 through 3.2 and the
> only one that prints that message is 2.6.
>
> --
> Steven
> --
> http://mail.python.org/mailman/listinfo/python-list

So maybe it should be fixed in 2.6? When I'll have time I'll look at copy.py in different python's. Maybe there's the answer)
Thanks anyway.

On Fri, 07 Jan 2011 23:54:24 +0200, kost BebiX wrote:

> 07.01.2011, 17:47, "Steven D'Aprano"
> <steve+>:
>> On Fri, 07 Jan 2011 16:47:55 +0200, kost BebiX wrote:
>>
>>> Â*Well, actually the code you showed doesn't work)
>>
>> Actually, it does. It just prints a warning message as well. Look
>> carefully:
[...]
>> I've tested the above code in Python versions 2.4 through 3.2 and the
>> only one that prints that message is 2.6.

> So maybe it should be fixed in 2.6? When I'll have time I'll look at
> copy.py in different python's. Maybe there's the answer) Thanks anyway.


Before you spend too much (i.e. any) time trying to fix this, I think
that Python 2.6 is now only accepting security fixes.

http://www.python.org/download/releases/2.6.6/


--
Steven

08.01.2011, 02:20, "Steven D'Aprano" <steve+>:
> On Fri, 07 Jan 2011 23:54:24 +0200, kost BebiX wrote:
>
>> š07.01.2011, 17:47, "Steven D'Aprano"
>> š<steve+>;:
>>> šOn Fri, 07 Jan 2011 16:47:55 +0200, kost BebiX wrote:
>>>> ššWell, actually the code you showed doesn't work)
>>> šActually, it does. It just prints a warning message as well. Look
>>> šcarefully:
>
> [...]
>
>>> šI've tested the above code in Python versions 2.4 through 3.2 and the
>>> šonly one that prints that message is 2.6.
>> šSo maybe it should be fixed in 2.6? When I'll have time I'll look at
>> šcopy.py in different python's. Maybe there's the answer) Thanks anyway.
>
> Before you spend too much (i.e. any) time trying to fix this, I think
> that Python 2.6 is now only accepting security fixes.
>
> http://www.python.org/download/releases/2.6.6/
>
> --
> Steven

Too bad (I mean, it's not too bad, because this thing can break other's 2.6 code that somehow depends on that). Now when I'll install 2.7 I'll have to remember about this bug when coding)