[Bug]: Script transform with python fails when no data is present in the stream

[harrisward]

Apache Hop version?

2.6.0-SNAPSHOT (2023-08-17 07.29.06)

Java version?

openjdk version "11.0.16.1" 2022-08-12 LTS

Operating system

Windows

What happened?

When you create a python step and NO data can be detected in the stream to the python step the step will throw an error that it cannot find the input fields which is not the case as the field exists on the input stream it just that there is no data coming into the step

python-bug.mp4

Issue Priority

Priority: 2

Issue Component

Component: Transforms

[mattcasters]

Well, I'm not sure we can actually fix this. We want to have the Script transform being capable of generating rows, as shown in the integration tests for this transform. On the flip side of that, we can't just invent values or not run the script when there is no input. So that leaves us with proper handling of the scenario, checking for undefined or indeed proper exception handling.

[mattcasters]

Perhaps we can create a new option in the Script transform to not process rows when there is no input to cover this scenario.

[Akira74]

When you create a python step in a pipeline (pipeline_sub), that will receive data from "Get rows from result" and use execute the pipeline with the Pipeline executor from e.g. pipeline_main it will always fail, even if it will receive data. All of the Python Script transformations are failing in the sub pipeline, even if they receive data within the pipeline later. I always get the error, that the Field is not there.