The LOP Output Processors API has had several changes for handling references with stage variable expressions:

• beginSave takes a new stage_variables dictionary parameter whichshould be passed to the base class implementation. This dictionarycontains the stage variables set on the root layer of the stagebeing written to disk.
• A new stageVariables method returns the dictionary passed to thebase class implementation of beginSave.
• A new processReferenceExpression method behaves exactly likeprocessReferencePath except it is called when the asset referenceis expressed as a stage variable expression.

The API for USD Output Processors has changed. Specifically, the "beginSave()" method now takes an addition parameter for overrides.

In the case of a Python-based translator, that means code that was:

def beginSave(self, config_node, t):

Should now be:

def beginSave(self, config_node, config_overrides, t):

To aid with the layering of config_node and config_overrides, a helper method has been added to husd.outputprocessor.OutputProcessor

Where you used to call, for example:

outputdir = config_node.parm("outputdir").evalAtTime(t)

You should now call:

outputdir = self.evalConfig("outputdir", config_node, config_overrides, t)

Note that "evalConfig()" also handles the validation that the node and parm exist (i.e., you can remove these safety-checks from your translator)

The API for USD Output Processors has changed. Specifically, the "beginSave()" method now takes an addition parameter for overrides.

In the case of a Python-based translator, that means code that was:

def beginSave(self, config_node, t):

Should now be:

def beginSave(self, config_node, config_overrides, t):

To aid with the layering of config_node and config_overrides, a helper method has been added to husd.outputprocessor.OutputProcessor

Where you used to call, for example:

outputdir = config_node.parm("outputdir").evalAtTime(t)

You should now call:

outputdir = self.evalConfig("outputdir", config_node, config_overrides, t)

Note that "evalConfig()" also handles the validation that the node and parm exist (i.e., you can remove these safety-checks from your translator)

The API for USD Output Processors has changed. Specifically, the "beginSave()" method now takes an addition parameter for overrides.

In the case of a Python-based translator, that means code that was:

def beginSave(self, config_node, t):

Should now be:

def beginSave(self, config_node, config_overrides, t):

To aid with the layering of config_node and config_overrides, a helper method has been added to husd.outputprocessor.OutputProcessor

Where you used to call, for example:

outputdir = config_node.parm("outputdir").evalAtTime(t)

You should now call:

outputdir = self.evalConfig("outputdir", config_node, config_overrides, t)

Note that "evalConfig()" also handles the validation that the node and parm exist (i.e., you can remove these safety-checks from your translator)

The API for USD Output Processors has changed. Specifically, the "beginSave()" method now takes an addition parameter for overrides.

In the case of a Python-based translator, that means code that was:

def beginSave(self, config_node, t):

Should now be:

def beginSave(self, config_node, config_overrides, t):

To aid with the layering of config_node and config_overrides, a helper method has been added to husd.outputprocessor.OutputProcessor

Where you used to call, for example:

outputdir = config_node.parm("outputdir").evalAtTime(t)

You should now call:

outputdir = self.evalConfig("outputdir", config_node, config_overrides, t)

Note that "evalConfig()" also handles the validation that the node and parm exist (i.e., you can remove these safety-checks from your translator)