RecursionError: maximum recursion depth exceeded while calling a Python object

[poautran]

In GitLab by @woutdenolf on Oct 16, 2025, 19:44 GMT+2:

import os
from ewoks import execute_graph
from ewokscore.task import Task


class InitIntegrator(Task, input_names=["params_file"], output_names=["integrator"]):
    def run(self):
        self.outputs.integrator = f"Integrator({self.inputs.params_file})"


class FindImages(
    Task,
    input_names=["image_directory", "integrator"],
    optional_input_names=["image_index"],
    output_names=["image_path", "next_image_index", "finished", "integrator"],
):
    def run(self):
        image_index = self.get_input_value("image_index", 0)
        image_directory = self.inputs.image_directory
        image_paths = [
            os.path.join(image_directory, f"data{i}.tiff") for i in range(100)
        ]
        self.outputs.image_path = image_paths[image_index]
        print(f"Emit {self.outputs.image_path}")
        self.outputs.next_image_index = image_index + 1
        self.outputs.finished = self.outputs.next_image_index == len(image_paths)
        self.outputs.integrator = self.inputs.integrator


class IntegrateImage(Task, input_names=["integrator", "image_path"]):
    def run(self):
        print(f"Integrate {self.inputs.image_path} with {self.inputs.integrator}")


workflow = {
    "nodes": [
        {
            "id": "init_integrator",
            "task_identifier": "__main__.InitIntegrator",
            "task_type": "class",
        },
        {
            "id": "find_images",
            "task_identifier": "__main__.FindImages",
            "task_type": "class",
        },
        {
            "id": "integrate_image",
            "task_identifier": "__main__.IntegrateImage",
            "task_type": "class",
        },
    ],
    "links": [
        {
            "source": "init_integrator",
            "target": "find_images",
            "data_mapping": [
                {"source_output": "integrator", "target_input": "integrator"}
            ],
        },
        {
            "source": "find_images",
            "target": "find_images",
            "data_mapping": [
                {"source_output": "next_image_index", "target_input": "image_index"},
            ],
            "conditions": [{"source_output": "finished", "value": False}],
        },
        {
            "source": "find_images",
            "target": "integrate_image",
            "data_mapping": [
                {"source_output": "image_path", "target_input": "image_path"},
                {"source_output": "integrator", "target_input": "integrator"},
            ],
        },
    ],
    "graph": {"id": "wf_pyFAI_1D_SAXSWAXS"},
}

if __name__ == "__main__":
    inputs = [
        {
            "id": "init_integrator",
            "name": "params_file",
            "value": "/path/to/calib.poni",
        },
        {"id": "find_images", "name": "image_directory", "value": "/path/to/data"},
    ]
    result = execute_graph(workflow, inputs=inputs, engine="ppf", pool_type="thread")

Migrated from GitLab: https://gitlab.esrf.fr/workflow/ewoks/ewoksppf/-/issues/22

[poautran]

In GitLab by @woutdenolf on Oct 16, 2025, 19:47 GMT+2:

Even fixing the pformat error, this will give a pool closed error

    result = execute_graph(
        workflow,
        inputs=inputs,
        engine="ppf",
        pool_type="thread",
        scaling_workers=True,
        max_workers=2,
    )