load_audio

class AudioLoader(plugin.BlenderLoader):
    """Load audio in Blender."""

    product_base_types = {"audio"}
    product_types = product_base_types
    representations = {"*"}
    extensions = {"wav"}

    label = "Load Audio"
    icon = "volume-up"
    color = "orange"

    def process_asset(
        self, context: dict, name: str, namespace: Optional[str] = None,
        options: Optional[Dict] = None
    ) -> Optional[List]:
        """
        Arguments:
            name: Use pre-defined name
            namespace: Use pre-defined namespace
            context: Full parenthood of representation to load
            options: Additional settings dictionary
        """
        libpath = self.filepath_from_context(context)
        folder_name = context["folder"]["name"]
        product_name = context["product"]["name"]

        asset_name = plugin.prepare_scene_name(folder_name, product_name)
        unique_number = plugin.get_unique_number(folder_name, product_name)
        group_name = plugin.prepare_scene_name(
            folder_name, product_name, unique_number
        )
        namespace = namespace or f"{folder_name}_{unique_number}"

        asset_group = bpy.data.objects.new(group_name, object_data=None)
        add_to_ayon_container(asset_group)

        # Blender needs the Sequence Editor in the current window, to be able
        # to load the audio. We take one of the areas in the window, save its
        # type, and switch to the Sequence Editor. After loading the audio,
        # we switch back to the previous area.
        window_manager = bpy.context.window_manager
        old_type = window_manager.windows[-1].screen.areas[0].type
        window_manager.windows[-1].screen.areas[0].type = "SEQUENCE_EDITOR"

        # The sequencer scene must be resolved before copying the context,
        # otherwise the override would still contain an unset sequencer scene.
        scene = get_sequencer_scene()

        # Versions without frame range data start at the scene frame start
        frame_start = get_version_frame_start(context)
        if frame_start is None:
            frame_start = scene.frame_start

        # We override the context to load the audio in the sequence editor.
        oc = bpy.context.copy()
        oc["area"] = window_manager.windows[-1].screen.areas[0]

        with bpy.context.temp_override(**oc):
            strip = add_sound_strip(
                scene, filepath=libpath, frame_start=frame_start)
            audio = strip.name

        window_manager.windows[-1].screen.areas[0].type = old_type

        asset_group[AYON_PROPERTY] = {
            "schema": "ayon:container-3.0",
            "id": AYON_CONTAINER_ID,
            "name": name,
            "namespace": namespace or '',
            "loader": str(self.__class__.__name__),
            "representation": context["representation"]["id"],
            "libpath": libpath,
            "asset_name": asset_name,
            "objectName": group_name,
            "audio": audio,
            "project_name": context["project"]["name"],
        }

        objects = []
        self[:] = objects
        return [objects]

    def exec_update(self, container: Dict, context: Dict):
        """Update an audio strip in the sequence editor.

        Arguments:
            container (ayon:container-1.0): Container to update,
                from `host.ls()`.
            representation (ayon:representation-1.0): Representation to
                update, from `host.ls()`.
        """
        repre_entity = context["representation"]
        object_name = container["objectName"]
        asset_group = bpy.data.objects.get(object_name)
        libpath = Path(self.filepath_from_context(context))

        self.log.info(
            "Container: %s\nRepresentation: %s",
            pformat(container, indent=2),
            pformat(repre_entity, indent=2),
        )

        assert asset_group, (
            f"The asset is not loaded: {container['objectName']}"
        )
        assert libpath, (
            "No existing library file found for {container['objectName']}"
        )
        assert libpath.is_file(), (
            f"The file doesn't exist: {libpath}"
        )

        metadata = asset_group.get(AYON_PROPERTY)
        group_libpath = metadata["libpath"]

        normalized_group_libpath = (
            str(Path(bpy.path.abspath(group_libpath)).resolve())
        )
        normalized_libpath = (
            str(Path(bpy.path.abspath(str(libpath))).resolve())
        )
        self.log.debug(
            "normalized_group_libpath:\n  %s\nnormalized_libpath:\n  %s",
            normalized_group_libpath,
            normalized_libpath,
        )
        if normalized_group_libpath == normalized_libpath:
            self.log.info("Library already loaded, not updating...")
            return

        old_audio = container["audio"]

        # Blender needs the Sequence Editor in the current window, to be able
        # to update the audio. We take one of the areas in the window, save its
        # type, and switch to the Sequence Editor. After updating the audio,
        # we switch back to the previous area.
        window_manager = bpy.context.window_manager
        old_type = window_manager.windows[-1].screen.areas[0].type
        window_manager.windows[-1].screen.areas[0].type = "SEQUENCE_EDITOR"

        # The sequencer scene must be resolved before copying the context,
        # otherwise the override would still contain an unset sequencer scene.
        scene = get_sequencer_scene()

        # We override the context to load the audio in the sequence editor.
        oc = bpy.context.copy()
        oc["area"] = window_manager.windows[-1].screen.areas[0]

        with bpy.context.temp_override(**oc):
            old_strip = get_all_strips(scene.sequence_editor)[old_audio]

            # Keep the strip where the user placed it. Moving it to match
            # the new version is left to the user.
            channel = old_strip.channel
            frame_start = int(old_strip.frame_start)

            remove_sound_strip(old_strip)
            new_strip = add_sound_strip(
                scene,
                filepath=str(libpath),
                frame_start=frame_start,
                channel=channel,
            )
            new_audio = new_strip.name

        window_manager.windows[-1].screen.areas[0].type = old_type

        metadata["libpath"] = str(libpath)
        metadata["representation"] = repre_entity["id"]
        metadata["audio"] = new_audio
        metadata["project_name"] = context["project"]["name"]

    def exec_remove(self, container: Dict) -> bool:
        """Remove an audio strip from the sequence editor and the container.

        Arguments:
            container (ayon:container-1.0): Container to remove,
                from `host.ls()`.

        Returns:
            bool: Whether the container was deleted.
        """
        object_name = container["objectName"]
        asset_group = bpy.data.objects.get(object_name)

        if not asset_group:
            return False

        audio = container["audio"]

        # Blender needs the Sequence Editor in the current window, to be able
        # to remove the audio. We take one of the areas in the window, save its
        # type, and switch to the Sequence Editor. After removing the audio,
        # we switch back to the previous area.
        window_manager = bpy.context.window_manager
        old_type = window_manager.windows[-1].screen.areas[0].type
        window_manager.windows[-1].screen.areas[0].type = "SEQUENCE_EDITOR"

        # The sequencer scene must be resolved before copying the context,
        # otherwise the override would still contain an unset sequencer scene.
        scene = get_sequencer_scene()

        # We override the context to load the audio in the sequence editor.
        oc = bpy.context.copy()
        oc["area"] = window_manager.windows[-1].screen.areas[0]

        with bpy.context.temp_override(**oc):
            remove_sound_strip(get_all_strips(scene.sequence_editor)[audio])

        window_manager.windows[-1].screen.areas[0].type = old_type

        bpy.data.objects.remove(asset_group)

        return True