dosubot[bot]
commented
5 months ago To resolve the TypeError: Object of type FileVar is not JSON serializable and ensure the iteration node correctly recognizes an array of files in your workflow, you need to use the jsonable_encoder function to encode the outputs before serializing them to JSON. This function is already imported and used in the _get_next_iteration method.
Here's how you can ensure the iteration node correctly recognizes an array of files and handles the serialization:
- Ensure that the
iteratoris a list of files. - Use the
jsonable_encoderto encode the outputs before returning them. - Handle special values like
FileVarusing thehandle_special_valuesmethod to convert them to a serializable format.
The relevant part of the code that handles the serialization is in the _get_next_iteration method:
def _get_next_iteration(self, variable_pool: VariablePool, state: IterationState) -> NodeRunResult | str:
"""
Get next iteration start node id based on the graph.
:param graph: graph
:return: next node id
"""
# resolve current output
self._resolve_current_output(variable_pool, state)
# move to next iteration
self._next_iteration(variable_pool, state)
node_data = cast(IterationNodeData, self.node_data)
if self._reached_iteration_limit(variable_pool, state):
return NodeRunResult(
status: WorkflowNodeExecutionStatus.SUCCEEDED,
outputs={
'output': jsonable_encoder(self.handle_special_values(state.outputs)) # Ensure outputs are JSON serializable
}
)
return node_data.start_node_idBy using jsonable_encoder and handle_special_values, you ensure that the state.outputs are properly encoded and can be serialized to JSON, thus avoiding the TypeError. The handle_special_values method converts FileVar objects to a dictionary format that is JSON serializable.
Additionally, ensure that the IterationNodeType correctly handles the iterator_selector and output variables. The getNodeOutputVars function should include the output variable for the Iteration block:
case BlockEnum.Iteration: {
res.push([id, 'output'])
break
}This ensures that the iteration node correctly recognizes and processes the array of files.
In the useConfig hook, ensure that the iterator and output variables are correctly set and handled:
const handleRun = useCallback((data: Record<string, any>) => {
const formattedData: Record<string, any> = {}
Object.keys(allVarObject).forEach((key) => {
const [varSectorStr, nodeId] = key.split(DELIMITER)
formattedData[`${nodeId}.${allVarObject[key].inSingleRunPassedKey}`] = data[varSectorStr]
})
formattedData[iteratorInputKey] = data[iteratorInputKey]
doHandleRun(formattedData)
}, [allVarObject, doHandleRun, iteratorInputKey])
const iterator = runInputData[iteratorInputKey]
const setIterator = useCallback((newIterator: string[]) => {
setRunInputData({
...runInputData,
[iteratorInputKey]: newIterator,
})
}, [iteratorInputKey, runInputData, setRunInputData])This ensures that the iteration node correctly recognizes and processes the array of files [1][2][3][4].
To continue talking to Dosu, mention @dosu.
AndyMik90
Self Checks
Dify version
0.6.9
Cloud or Self Hosted
Cloud, Self Hosted (Docker), Self Hosted (Source)
Steps to reproduce
DSL to reproduce:
✔️ Expected Behavior
Output the files in an array so multiple files can be outputted in the end node.
❌ Actual Behavior
TypeError: Object of type FileVar is not JSON serializable and wrong Array.