For stub tests in modules is there a recommendation on testing process where the exact set of output files generated varies wildly based on the parameters input? Specifically I'm working on adding stub tests for

cnvkit/batch

. https://github.com/nf-core/modules/pull/8618. In cnvkit, this command is essentially a full pipeline on it's own that handles a wide range of inputs and generates very different sets of output files. https://cnvkit.readthedocs.io/en/stable/pipeline.html I'm wondering if I should be completely re-creating the logic that determines which files are generated for each different set of inputs provided or if I can just touch all the potential output files and call it a day.

While the former is a nice to have, the latter is fine and what most people do

Yeah, don't spend hours to deal with every possibility

Thank you! PR is ready for review: https://github.com/nf-core/modules/pull/8618 Sorry for the delay @Simon Pearce, it's been a busy summer!

No worries