Rough summary of fixes/changes:
- Blender Py API: GameLogic -> bge.logic
- Blender Py API: Implemented missing KX_PYATTRIBUTE_TODOs and -DUMMYs.
- Fix for [#22924] KX_PolygonMaterial.diffuse does not return expected list[r,g,b]
- Py API: Renaming _owner attribute of mathutils classes to owner.
- Fix some minor errors in mathutils and blf.
- Enabling game engine autoplay again based on a patch by Dalai:
* The biggest 3D view in the open scene is used, if there is none, blender opens the file normally and raises an error.
* The 3D view are is made fullscreen.
* Quad view, header, properties and toolbox panel are all hidden to get the maximum view.
* If the game engine full screen setting is set, the game starts in fullscreen.
- Fix for ipo conversion on file transition in the game engine.
The BGE was getting the namespace dict directly from __main__ which conflicts
with my recent fix to get the pickle module working which to overwrote the __main__ module on script execution.
Simple fix is to have the BGE and Blender use the same method of getting namespaces.
Renamed CreateGlobalDictionary() to bpy_namespace_dict_new() and moved into bpy_internal_import.c
pickle still wont work in the BGE since we make a copy of __main__ namespace but for speed would rather not have to replace the __main__ module many times per second.
BGE Py Controllers were effectively doing this...
"a.b.c" --> "__import__('a').b.c()"
This was annoying because it meant module 'a' would need to import 'b' explicitly.
Now use import like this.
"a.b.c" --> "__import__("a.b").c()"
Note that this has the slight disadvantage that these need to be modules, where as before they could be collections of functions in a class instance for eg. So its possible this breaks existing files but dont think anyone used this since its a fairly obscure use case.
patch from Mitchell Stokes (moguri)
simple use case
scene.post_draw = [pyOpenGLFunc]
this only needs to be set once, then the funcion runs each redraw.
note, this patch also changes how python scripts run (not modules): Dont clear the namespace after running a script, since functions still use the namespace, BGE API is now better when dealing with stale data.
made some changes to this patch.
- assigning a list didnt decrement the existing list.
- initialize as NULL rather then a blank list
- dont use string comparisons for the callbacks, pass the python list to use instead.
- dont check the list items are callable. python will display an error if they are not.
- use python list macros that dont do any type checking sine blender does this when assigning the list
---- from tracker, edited since an updated patch changes some things.
Here is a patch to be able to draw to the screen with BGE Python. This will be very handy for GUI stuff. This patch
works by having the user register a callback in the scene. Two options are available KX_Scene.pre_draw
and KX_Scene.post_draw. The difference between these is when Python draws to the screen (before or after the BGE).
Each can take a list of functions. Here is an example that draws a blue semi-transparent
Adding a UI to set the type on startup can be added easily.
# ----
class myPlayer(GameTypes.KX_GameObject):
def die(self):
# ... do stuff ...
self.endObject()
# make an instance
player = myPlayer(gameOb) # gameOb is made invalid now.
player.die()
# ----
One limitation (which could also be an advantage), is making the subclass instance will return that subclass everywhere, you cant have 2 different subclasses of the same BGE data at once.
Remove the last of the odd C++/python wrapper code from http://www.python.org/doc/PyCPP.html (~1998)
* Use python subclasses rather then having fake subclassing through get/set attributes calling parent types.
* PyObject getset arrays are created while initializing the types, converted from our own attribute arrays. This way python deals with subclasses and we dont have to define getattro or setattro functions for each type.
* GameObjects and Scenes no longer have attribute access to properties. only dictionary style access - ob['prop']
* remove each class's get/set/dir functions.
* remove isA() methods, can use PyObject_TypeCheck() in C and issubclass() in python.
* remove Parents[] array for each C++ class, was only used for isA() and wasnt correct in quite a few cases.
* remove PyTypeObject that was being passed as the last argument to each class (the parent classes too).
TODO -
* Light and VertexProxy need to be converted to using attributes.
* memory for getset arrays is never freed, not that bad since its will only allocates once.
svn merge https://svn.blender.org/svnroot/bf-blender/trunk/blender -r19820:HEAD
Notes:
* Game and sequencer RNA, and sequencer header are now out of date
a bit after changes in trunk.
* I didn't know how to port these bugfixes, most likely they are
not needed anymore.
* Fix "duplicate strip" always increase the user count for ipo.
* IPO pinning on sequencer strips was lost during Undo.
- More verbose error messages.
- BL_Shader wasnt setting error messages on some errors
- FilterNormal depth attribute was checking for float which is bad because scripts often expect ints assigned to float attributes.
- Added a check to PyVecTo for a tuple rather then always using a generic python sequence. On my system this is over 2x faster with an optmized build.
- comments to PyObjectPlus.h
- remove unused/commented junk.
- renamed PyDestructor to py_base_dealloc for consistency
- all the PyTypeObject's were still using the sizeof() their class, can use sizeof(PyObjectPlus_Proxy) now which is smaller too.
This changes how the BGE classes and Python work together, which hasnt changed since blender went opensource.
The main difference is PyObjectPlus - the base class for most game engine classes, no longer inherit from PyObject, and cannot be cast to a PyObject.
This has the advantage that the BGE does not have to keep 2 reference counts valid for C++ and Python.
Previously C++ classes would never be freed while python held a reference, however this reference could be problematic eg: a GameObject that isnt in a scene anymore should not be used by python, doing so could even crash blender in some cases.
Instead PyObjectPlus has a member "PyObject *m_proxy" which is lazily initialized when python needs it. m_proxy reference counts are managed by python, though it should never be freed while the C++ class exists since it holds a reference to avoid making and freeing it all the time.
When the C++ class is free'd it sets the m_proxy reference to NULL, If python accesses this variable it will raise a RuntimeError, (check the isValid attribute to see if its valid without raising an error).
- This replaces the m_zombie bool and IsZombie() tests added recently.
In python return values that used to be..
return value->AddRef();
Are now
return value->GetProxy();
or...
return value->NewProxy(true); // true means python owns this C++ value which will be deleted when the PyObject is freed
- The armature weakref list was being incref'd twice then decrefed twice (incref and decref were used incorrectly), now only once. My 'fix' broke this.
- In bpy_pydriver_create_dict the 2 refs added from running PyDict_SetItemString twice were undone when clearing the dictionary (added comment)
- changed Py_XDECREF to Py_DECREF int BPY_pyconstraint_update and BPY_pyconstraint_target, Py_XDECREF checs for NULL value which would have crashed blender before it got to Py_XDECREF anyway.
- after every error is reported (PyErr_Print), remove sys.last_traceback and clear the error, I found this fixed certain crashes (usually when starting the game engine or exiting blender), so best do this all the time.
- header_text.c, CcdPhysicsEnvironment.cpp, KX_CameraActuator.cpp - remove some warnings.
Use each types dictionary to store attributes PyAttributeDef's so it uses pythons hash lookup (which it was already doing for methods) rather then doing a string lookup on the array each time.
This also means attributes can be found in the type without having to do a dir() on the instance.
- Initialize python types with PyType_Ready, which adds methods to the type dictionary.
- use Pythons get/setattro (uses a python string for the attribute rather then char*). Using basic C strings seems nice but internally python converts them to python strings and discards them for most functions that accept char arrays.
- Method lookups use the PyTypes dictionary (should be faster then Py_FindMethod)
- Renamed __getattr -> py_base_getattro, _getattr -> py_getattro, __repr -> py_base_repr, py_delattro, py_getattro_self etc.
From here is possible to put all the parent classes methods into each python types dictionary to avoid nested lookups (api has 4 levels of lookups in some places), tested this but its not ready yet.
Simple tests for getting a method within a loop show this to be between 0.5 and 3.2x faster then using Py_FindMethod()
Added the method into the PyType so python knows about the methods (its supposed to work this way).
This means in the future the api can use PyType_Ready() to store the methods in the types dictionary.
Python3 removes Py_FindMethod and we should not be using it anyway since its not that efficient.
Python dir(ob) for game types now includes attributes names,
* Use "__dict__" rather then "__methods__" attribute to be Python 3.0 compatible
* Added _getattr_dict() for getting the method and attribute names from a PyObject, rather then building it in the macro.
* Added place holder *::Attribute array, needed for the _getattr_up macro.
* Made GameLogic.addActiveActuator(actu, bool) to raise an error if the actuator is not in the list. Before it would allow any value as the actuator and fail silently (makes debugging scripts more difficult).
* Allow the actuator to be a string which is convenient if you dont want to change the settings of the actuator.
* Added activate/deactivate functions to the controller, this is more logical since the GameLogic.addActiveActuator() function is running through the controller anyway.
GameLogic.addActiveActuator(controller.getActuator("SomeAct"), True)
...can be replaced with...
controller.activate("SomeAct")
Use 'const char *' rather then the C++ 'STR_String' type for the attribute identifier of python attributes.
Each attribute and method access from python was allocating and freeing the string.
A simple test with getting an attribute a loop shows this speeds up attribute lookups a bit over 2x.
The principle is to replace most get/set methods of logic bricks by direct property access.
To make porting of game code easier, the properties have usually the same type and use than
the return values/parameters of the get/set methods.
More details on http://wiki.blender.org/index.php/GameEngineDev/Python_API_Clean_Up
Old methods are still available but will produce deprecation warnings on the console:
"<method> is deprecated, use the <property> property instead"
You can avoid these messages by turning on the "Ignore deprecation warnings" option in Game menu.
PyDoc is updated to include the new properties and display a deprecation warning
for the get/set methods that are being deprecated.
This is an interesting bug since it is likely the cause of many other suspicious python crashes in blender.
sys.last_traceback would store references to PyObjects at the point of the crash.
it would only free these when sys.last_traceback was set again or on exit.
This caused many crashes in the BGE while testing since python would end up freeing invalid game objects -
When running scripts with errors, Blender would crash every 2-5 runs - in my test just now it crashed after 4 trys.
It could also segfault blender, when (for eg) you run a script that has objects referenced. then load a new file and run another script that raises an error.
In this case all the invalid Blender-Object's user counts would be decremented, even though none of the pointers were still valid.
I'm getting this error now:
GPG_Application.cpp: In member function 'void GPG_Application::stopEngine()':
/System/Library/Frameworks/Python.framework/Versions/2.3/include/python2.3/marshal.h:12: error: too many arguments to function 'PyObject* PyMarshal_WriteObjectToString(PyObject*)'
GPG_Application.cpp:720: error: at this point in file
Are we offically not supporint older versions of python now? :)
Kent