generate_samplesheet.py emits malformed CSV when FASTQ paths contain commas #1296

Activity

1. CryptGodSon commented on Oct 10, 2026

   

   CryptGodSon

   on Oct 10, 2026

   More actions

   I have a fix for this on my fork: CryptGodSon@88fa205

   _write_samplesheet now writes rows with csv.writer instead of ','.join, keeping \n line endings, so output for ordinary paths is unchanged. Tested with the repro above: the comma path now gives 4 header fields and 4 data fields and passes --validate, and a plain path gives the same file as before.

   External PRs to this repo are auto-closed, so I'm linking it here in case a maintainer wants to pick it up.
2. CryptGodSon commented on Oct 11, 2026

   

   CryptGodSon

   on Oct 11, 2026

   More actions

   I looked into this and the cause is in _write_samplesheet: rows are built with ','.join(...), so any field containing a comma is split into extra columns and the file then fails the script's own --validate.

   A minimal fix is to write the header and rows with csv.writer, keeping \n line endings so output for ordinary paths stays byte-for-byte the same:

   import csv  # moved to module level
   
   with open(output_path, 'w', newline='') as f:
       writer = csv.writer(f, lineterminator='\n')
       writer.writerow(active_columns)
       for row in rows:
           writer.writerow([str(row.get(col, '')) for col in active_columns])

   I opened #1305 with this change before realizing the repo only accepts contributions from Anthropic team members, so it was auto-closed. The full commit is here in case it's useful: CryptGodSon@88fa205