[Python-projects] PyLint determines class "newness" incorrectly

[Sylvain Thénault]

Hi,

On Thursday 17 August à 15:49, skip at pobox.com wrote:
> Given this little program:
> 
>     import pygtk
>     pygtk.require("2.6")
>     import gobject
> 
>     class A(gobject.GObject):
>         def __init__(self, val):
>             gobject.GObject.__init__(self)
>             self._val = val
> 
>         def _get_val(self):
>             print "get"
>             return self._val
> 
>         def _set_val(self, val):
>             print "set"
>             self._val = val
> 
>         val = property(_get_val, _set_val)
> 
>     if __name__ == "__main__":
>         print gobject.GObject.__bases__
>         a = A(7)
>         print a.val
>         a.val = 6
>         print a.val
> 
> When I run it, I get this output:
> 
>     (<type 'object'>,)
>     get
>     7
>     set
>     get
>     6
> 
> However, PyLint reports:
> 
>     gobject_subclass.py:8: [W, A.__init__] __init__ method from a non direct base class <logilab.astng.Yes object at 0x8253d4c> is called
>     gobject_subclass.py:19: [W, A] Use of "property" on an old style class
> 
> It clearly has messed something up about the lineage of the A class.  One,
> the logilab.astng.Yes class is not involved with the A class.  (I believe I
> reported this before.)  Two, A is a new-style class (because gobject.GObject
> is, as the program output demonstrates).

thanks for this valuable report, I've added it to our tracker.
-- 
Sylvain Thénault                               LOGILAB, Paris (France)
Formations Python, Zope, Plone, Debian:  http://www.logilab.fr/formations
Développement logiciel sur mesure:       http://www.logilab.fr/services
Python et calcul scientifique:           http://www.logilab.fr/science