Yeah, something is definitely wrong. It does appear to be returning a
string from _windows_.new_Printout() in _windows.py (if you print
type(newobj), for example). I guess in _wrap_new_Printout() in
_windows_wrap.cpp 'result' is actually a wxPyPrintout* (C's typing
system requires it), and it's not jumping to the "fail:" label, which
means wxPyMake_wxObject(result,1) is turning a wxPyPrintout* into a
Python string object. (Does that make any sense?) But it uses
wxPyMake_wxObject(), and everything uses that, so it can't be *that*
broken, right?
So I suspect there's a problem with the wxPyPrintout object itself, but
that's no surprise. And now I'm deeper into the wx-Py interface than I
know what I'm doing (so far).
I looked at the diffs (on ViewCVS) and this function
(_wrap_new_Printout()) of src/gtk/_windows_wrap.cpp has only changed
trivially (using true instead of True) since the 2.4 days. So I think
it's safe to say that's not the problem.
I don't know what's going on. The further I dig, the more complex (and
undocumented!) it becomes.
There don't seem to be any wxPython binaries in Debian with debugging
info, and I'm not up for building my own wxPython binaries tonight :-),
so I'm calling it a night.
If it helps, a really simple way to demonstrate this problem is:
import wxversion
wxversion.select('2.5.3')
import wx
wx.PySimpleApp()
wx.Printout()
Result:
Traceback (most recent call last):
File "<stdin>", line 1, in ?
File
"/usr/lib/python2.3/site-packages/wx-2.5.3-gtk2-unicode/wx/_windows.py",
line 4363, in __init__
self.this = newobj.this
AttributeError: 'str' object has no attribute 'this'
WX2.4 doesn't require a wx.App object, so the equivalent there is
simply:
>>> import wx
>>> wx.Printout()
<wxPython.printfw.wxPrintout instance; proxy of C++ wxPrintout instance
at _82431f8_wxPyPrintout_p>
My best guess is that this isn't Debian-specific, but may be GTK
+-specific. Can the folks upstream offer any help with this?
It would be nice to be able to print. :-)
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
