General
The key words “MUST”, “MUST NOT”, “SHOULD”, etc. are to be interpreted as described in RFC 2119.
Replication of test data
New test data files within a branch (modules or pipelines) SHOULD NOT replicate existing test data unless absolutely necessary.
Files that can be generated from upstream files SHOULD be derived from existing test data on the test-datasets branch. For example, if a particular bioinformatic index file is needed for a tool, index an existing FASTA file from the test-datasets branch rather than uploading a new index file.
CI tests for nf-core modules, subworkflows, or pipelines are not required to produce meaningful output.
The main goal of nf-core CI tests is to ensure that a given tool executes without errors.
A test may produce nonsense output or find nothing, as long as the tool does not crash or produce an error.
You SHOULD reuse existing test data as much as possible to reduce the size of the test dataset repository.
You SHOULD upload new test data only if there is no other option within the existing test-data archive.
Size of test data
Test data SHOULD be as small as possible.
Test data files MUST NOT exceed the GitHub file size maximum.
Data SHOULD be sub-sampled as aggressively as possible while still allowing the tool to execute successfully.
License of test data
Test data MUST be publicly available and have licenses that allow public reuse.
Documentation of test data
Test data files SHOULD be described in the given branch’s README file.
The README SHOULD include the source of the data, how it was generated, and license information.