made LF object persistent

so we can change the array setup while keeping animation keyframes for the whole camera rig.
ehedayati · Jun 28, 2017 · 607e64c · 607e64c
1 parent 19cf15c
commit 607e64c
Showing 1 changed file with 54 additions and 13 deletions.
diff --git a/lightfield_simulator.py b/lightfield_simulator.py
@@ -104,29 +104,71 @@ class OBJECT_OT_create_lightfield(bpy.types.Operator):
     bl_options = {'REGISTER'}
 
     def execute(self, context):
-        bpy.ops.scene.delete_lightfield('EXEC_DEFAULT')
+
+
+        LF = bpy.context.scene.LF
+
+        # Create lightfield container, but only if it doesn't exist yet.
+        # If it exists, just clear it.
+        # This is required as deleting the LF-object would also delete
+        # defined keyframes for camera animation.
+        #
+        # For this reason we also only restore the position of the LF-object
+        # from config file if it's non existent.
+        try:
+
+            lightfield = bpy.data.objects[LF.get_lightfield_name()]
+
+            # -> now we are going to clear the LF object
+
+            # save initially selected objects
+            selected_objects = bpy.context.selected_objects
+
+            # delesect everything
+            bpy.ops.object.mode_set(mode='OBJECT')
+            bpy.ops.object.select_all(action='DESELECT')
+
+            # delete frustum, cameras and container
+            for camera in LF.get_lightfield_cameras():
+                camera.select = True
+
+            LF.get_frustum().hide = False
+            LF.get_frustum().hide_select = False
+            LF.get_frustum().select = True
+
+            bpy.ops.object.delete()
+
+            # restore initial state
+            for object in selected_objects:
+                object.select = True
+
+        except KeyError:
+
+            bpy.ops.object.empty_add(type='PLAIN_AXES', view_align=False, location=(0, 0, 0), rotation=(0, 0, 0))
+            lightfield = bpy.context.object
+
+            lightfield.empty_draw_size = 0.4
+            lightfield.name = LF.get_lightfield_name()
+
+            lightfield.location = [LF.center_cam_x, LF.center_cam_y, LF.center_cam_z]
+            lightfield.rotation_euler = [LF.center_cam_rot_x, LF.center_cam_rot_y, LF.center_cam_rot_z]
+
 
         # initialize lightfield elements
         self.set_render_properties()
         cameras = self.create_cameras()
-        frustum = self.create_frustum()
 
-        # create lightfield container
-        LF = bpy.context.scene.LF
-        bpy.ops.object.empty_add(type='PLAIN_AXES', view_align=False, location=(0, 0, 0), rotation=(0, 0, 0))
-        bpy.context.object.empty_draw_size = 0.4
-        bpy.context.object.name = LF.get_lightfield_name()
+        frustum = self.create_frustum()
+        frustum.hide_select = True
 
         # add all cameras and frustum to container
         lightfield_objects = cameras + [frustum]
         for object in lightfield_objects:
-            object.select = True
+            object.parent = lightfield
 
-        bpy.ops.object.parent_set(type='OBJECT', xmirror=False, keep_transform=False)
-        bpy.context.object.location = [LF.center_cam_x, LF.center_cam_y, LF.center_cam_z]
-        bpy.context.object.rotation_euler = [LF.center_cam_rot_x, LF.center_cam_rot_y, LF.center_cam_rot_z]
+        # set the active object to LF container
+        bpy.context.scene.objects.active = lightfield
 
-        LF.get_frustum().hide_select = True
         return {'FINISHED'}
 
     def create_cameras(self):
@@ -383,7 +425,6 @@ def renderFrame(self, tgt_dir = None):
     def render_input_views(self, cameras, scene_key, LF, tgt_dir):
 
 
-
         # create image output node
         image_out_node = bpy.data.scenes[scene_key].node_tree.nodes.new(type='CompositorNodeOutputFile')
         image_out_node.format.file_format = 'PNG'