Scripted objects saving attributes/ru

From FreeCAD Documentation
This page is a translated version of the page Scripted objects saving attributes and the translation is 4% complete.
Other languages:

Введение

Scripted objects are rebuilt every time a FCStd document is opened. To do this the document keeps a reference to the module and Python class that were used to create the object, along with its properties.

Attributes of the class used to create the object can also be saved, that is, "serialized". This can be further controlled by the dumps and loads methods of the class.

Saving all attributes

By default, attributes saved in an object class are those from the __dict__ dictionary of the class.

# various_states.py
class VariousStates:
    def __init__(self, obj):
        obj.addProperty("App::PropertyLength", "Length")
        obj.addProperty("App::PropertyArea", "Area")
        obj.Length = 15
        obj.Area = 300
        obj.Proxy = self

        Type = dict()
        Type["Version"] = "Custom"
        Type["Release"] = "production"
        self.Type = Type
        self.Type = "Custom"
        self.ver = "0.18"
        self.color = (0, 0, 1)
        self.width = 2.5

    def execute(self, obj):
        pass

An object can be created using this class, and it can be saved to my_document.FCstd. If no particular viewprovider is assigned to the new object, its proxy class is simply set to a value different from None, in this case, to 1.

import FreeCAD as App
import various_states

doc = App.newDocument()
doc.FileName = "my_document.FCStd"

obj = doc.addObject("Part::FeaturePython", "Custom")
various_states.VariousStates(obj)

if App.GuiUp:
    obj.ViewObject.Proxy = 1

doc.recompute()
doc.save()

When we re-open the file we can inspect the dictionary of the object's class.

>>> obj = App.ActiveDocument.Custom
>>> print(obj.Proxy)
<various_states.VariousStates object at 0x7f0a899bde10>
>>> print(obj.Proxy.__dict__)
{'Type': {'Version': 'Custom', 'Release': 'production'}, 'ver': '0.18', 'color': [0, 0, 1], 'width': 2.5}

We see that all attributes that start with self in the class were saved. These can be of different types, including string, list, float, and dictionary. The original tuple for self.color was converted to a list, but otherwise all attributes were loaded the same.

Saving specific attributes

We can define a class similar to the first one, that implements specific attributes to save.

# various_states.py
class CustomStates:
    def __init__(self, obj):
        obj.addProperty("App::PropertyLength", "Length")
        obj.addProperty("App::PropertyArea", "Area")
        obj.Length = 15
        obj.Area = 300
        obj.Proxy = self

        Type = dict()
        Type["Version"] = "Custom"
        Type["Release"] = "production"
        self.Type = Type
        self.ver = "0.18"
        self.color = (0, 0, 1)
        self.width = 2.5

    def execute(self, obj):
        pass

    def dumps(self):
        return self.color, self.width

    def loads(self, state):
        self.color = state[0]
        self.width = state[1]

The return value of dumps is the object that will be serialized. This can be a single value, or a tuple of values. When the object is restored, the class calls the loads method, passing the state variable with the serialized content. In this case state is a tuple that is unpacked into the respective variables to reconstruct the "state" that original existed.

state = (self.color, self.width)
state = ((0, 0, 1), 2.5)

We can create an object with this class, and save the document, just like in the previous example. When we re-open the file we can inspect the dictionary of the object's class.

>>> obj2 = App.ActiveDocument.Custom2
>>> print(obj2.Proxy)
<various_states.CustomStates object at 0x7fb399496630>
>>> print(obj2.Proxy.__dict__)
{'color': [0, 0, 1], 'width': 2.5}

The original tuple for self.color was converted to a list, but otherwise the information was recovered fine. Instead of restoring all attributes like in the previous case, only the attributes that we specified in dumps and loads were restored.

Usage

Identifying the type

Normally, scripted objects should use properties to store information, as these are automatically restored when the document is opened.

However, sometimes the class restore internal information which is not intended to be modified, but which is helpful to inspect.

For example, most objects of the Draft Workbench set up a Type attribute that can be used to determine the type of object that is under use.

class DraftObject:
    def __init__(self, obj, _type):
        self.Type = _type

    def dumps(self):
        return self.Type

    def loads(self, state):
        if state:
            self.Type = state

All objects have a TypeId property, but for scripted objects this property is not useful, as it always refers to the parent C++ class, for example, Part::Part2DObjectPython or Part::FeaturePython. Therefore, having this additional Proxy.Type attribute in the class is useful to treat each object in a particular way.

Migrating the object

Version information can be stored inside the class in order to verify the origin of an object.

class CustomObject:
    def __init__(self, obj, _type):
        self.Type = _type
        self.version = "0.18"

    def dumps(self):
        return self.Type, self.version

    def loads(self, state):
        if state:
            self.Type = state[0]
            self.version = state[1]

If the structure of the class changes, that is, if its properties or methods change, are renamed, or are removed, we could test the version attribute in order to migrate the older object to a new set of properties or to a new class. This can be done by implementing the onDocumentRestored method, as explained in Scripted objects migration.

class CustomObject:
    def onDocumentRestored(self, obj):
        if hasattr(obj.Proxy, "version") and obj.Proxy.version:
            if obj.Proxy.version == "0.18":
                self.migrate_from_018(obj)

    def migrate_from_018(self, obj):
        ...

Links