Notes subcommand still uses --input-dir | Generated by AI

Home 2026.06

Here’s the analysis:

=== Conversation Subcommands: –input-dir vs positional file ===

The commit 7f3f368 changed ww conversation generate from --input-dir (batch all JSONs in a dir) to a positional file arg (single file). Here’s the status of all 5 subcommands:

  1. json (json.py) - positional filename arg ✅ No –input-dir
  2. generate (generate.py) - positional file arg ✅ Fixed in 7f3f368
  3. notes (notes.py) - still uses --input-dir ⚠️ See below
  4. to-video (video.py) - positional file (sys.argv) ✅ No –input-dir
  5. to-image (image.py) - positional file (sys.argv) ✅ No –input-dir

=== Only notes still uses –input-dir ===

notes.py line 78-82: parser.add_argument(“–input-dir”, default=”scripts/conversation”, …)

But this is arguably intentional — notes iterates ALL .json files in a directory and converts each to a markdown note. It’s a batch-by-design command.

However, if you want consistency with generate (single file arg), it could be refactored to accept a single file OR a directory. The generate command before 7f3f368 also had the “process all files in dir” batch mode, and you removed it.

=== video.py has a local var named input_dir but it’s NOT a CLI flag ===

video.py line 139: input_dir = os.path.dirname(input_path) or "." This is just extracting the parent directory from the positional file arg. No --input-dir flag exists. ✅ Clean.

=== Summary ===

Only ww conversation notes has the --input-dir pattern. All others use positional file args. Want me to refactor notes to match — accept either a single file or a directory as a positional arg?

Back Donate