Currently, using echo to write possibly user-supplied filenames and sample ID fields might introduce a vulnerability to code injection. We should consider moving generate_summary to be after the IRIDA JSON output is produced and have it launch a Python process that reads the final JSON file, handles the data, and then writes the summary.txt.gz file.
Currently, using
echoto write possibly user-supplied filenames and sample ID fields might introduce a vulnerability to code injection. We should consider movinggenerate_summaryto be after the IRIDA JSON output is produced and have it launch a Python process that reads the final JSON file, handles the data, and then writes thesummary.txt.gzfile.