We'd also like to be able to collect provenance (mainly checksums) on output files. We could approach this in two ways:
Change the InputFileProvenanceRecord into a more generic FileProvenanceRecord. This would involve changing the names of the input_filename and input_path attributes to filename and file_path.
Add a separate OutputFileProvenanceRecord with attributes output_filename (and possibly output_path)
We currently have an
InputFileProvenanceRecord
type, which is intended to support details about input files:https://github.com/BCCDC-PHL/pipeline-provenance-schema/blob/58cda2e938d72c193151e4b4b1ed62967cb2f71c/schema/pipeline-provenance.json#L218-L262
We'd also like to be able to collect provenance (mainly checksums) on output files. We could approach this in two ways:
InputFileProvenanceRecord
into a more genericFileProvenanceRecord
. This would involve changing the names of theinput_filename
andinput_path
attributes tofilename
andfile_path
.OutputFileProvenanceRecord
with attributesoutput_filename
(and possiblyoutput_path
)