Scripting

Script step

Panda Utils allows inserting arbitrary steps at any point of Asset Pipeline. These steps have to be written using Python scripts. The script must have a single function run which takes one parameter ctx, for example:

def run(ctx):
    ctx.cache_eggs()
    for tree in ctx.eggs.values():
        nodes_for_removal = (
            tree.findall("Material")
            + tree.findall("MRef")
            + tree.findall("Normal")
            + [scalar for scalar in tree.findall("Scalar") if scalar.node_name == "uv-name"]
        )
        tree.remove_nodes(set(nodes_for_removal))

        uvs = tree.findall("UV")
        for uv in uvs:
            uv.node_name = None
        darts = tree.findall("Dart")
        for dart in darts:
            dart.node_value = "1"

The ctx is an instance of Pipeline Context. It has a number of attributes, but the following are the most useful:

model_name - the name of the model being worked on.
eggs - a dictionary. See cache_eggs below.
files - all files in the current directory, in sorted order. Note that this will update in real time as you add or remove files, so be careful.
cache_eggs() - loads all egg files in folder into eggs dictionary, which will have the filenames as keys and the syntax trees as values (see Egg Trees). This is useful if the script does operations on the syntax tree.
uncache_eggs() - saves all egg syntax trees back into the files. This is useful if the script operates on egg files directly (egg-trans, egg-optchar etc.)
putil_ctx - a context that can be used to call the Utils-Core functions (i.e. run_panda).
model_config - a dictionary indexed by step names with the raw model configs from the YAML file.

Generally you don’t need to run uncache_eggs at the end of working with trees, or run cache_eggs at the end of not working with trees, the other steps have a convention to run this only at the start.

These scripts should be placed in the scripts folder.

Arguments

This step uses any number of parameters, the first parameter is required.

script_file - the name of a file to run. It is expected to be located in scripts/{script_file}.py. Folder structures are also supported with dot separators (i.e. folder.script).
parameters - variadic. Will pass the parameters to the script.

Script also allows parametrized launch loading script data from model-config. This does not allow setting parameters directly. For that, [] or {} must be appended to script_file. The mechanics of this will be the same as usual, the key will be read from script/{script_file}.

Examples

script:magic1
script:parametrized:group_name
script:something[]

Blender Script step

Panda Utils allows inserting arbitrary Blender scripts at any point of Asset Pipeline. It is recommended to only use such steps before yabee or blend2bam for obvious reasons. These steps have to be written using Python scripts using the Blender’s Python API. The script should be runnable itself, and it will run with the Blend file already loaded, for example:

import os

import bpy


def update_texture_paths():
    for mat in bpy.data.materials:
        if mat.use_nodes:
            for node in mat.node_tree.nodes:
                if node.type == "TEX_IMAGE":
                    img = node.image
                    if img is not None:
                        img.filepath = img.filepath.replace(os.path.sep, "/").split("/")[-1]


update_texture_paths()
bpy.ops.wm.save_as_mainfile(filepath=bpy.data.filepath)

These scripts should be placed in the bscripts folder to separate them from normal scripts.

Arguments

This step requires one parameter.

script_file - the name of a file to run. It is expected to be located in bscripts/{script_file}.py. Folder structures are not supported at the time.

Examples

bscript:magic2