On 10.01.06 10:42:03, Torsten Marek wrote: > > Couldn't let it lie around ;-) > > > > So here's a small patch which makes empty extends-Elements work and also > > "corrects" the import statements by simply transforming whatever is in > > the header-element to a pythonic module-path. > > > > The latter means that a string like > > > > widgets/mycustomwidget.h > > > > gets transformed into > > > > widgets.mycustomwidget > > > > I think you can't get much better than this, except maybe provide a way > > to override this with a custom module path (like -m > > mycustomwidget.module). However that is beyond my scope for now. > > > > Phil, maybe you can put this into the next snapshot of PyQt4 so I don't > > have to patch myself. > > I'm kind of in a general rush right now since I'm moving back to Germany right > now.
That's kind of what I was thinking as you "vanished" from this list around mid december. > The patch should be okay for now. I only write this to confirm that the python code generated with the patched pyuic4 actually works and a widget using the ui is displaying everything correct. I hope Phil includes it in the next snapshot (it wasn't in tonights). Regarding the "for now": I guess you mean the manual inclusion of QObject as base class when the extends element is empty. That is a ugly hack indeed, but actually I didn't want to dive into the depth of pyuic4 to find out why it needs a QObject or QWidget there. Especially as this base class is not used anywhere, i.e. the ui-class is derived from python "object" and nothing else. Andreas -- You have an unusual understanding of the problems of human relationships. _______________________________________________ PyKDE mailing list [email protected] http://mats.imk.fraunhofer.de/mailman/listinfo/pykde
