Continuously monitor a directory for data file changes.

Usage

goldencheck watch data/                        # poll every 60s (default)
goldencheck watch data/ --interval 30          # poll every 30s
goldencheck watch data/ --pattern "*.csv"      # only CSV files
goldencheck watch data/ --exit-on error        # CI: fail on first error
goldencheck watch data/ --json                 # JSON output per scan

Behavior

1. Initial scan — scans all matching files on startup
2. Poll loop — checks file modification times every N seconds
3. Re-scan — only re-scans files whose mtime changed
4. Ctrl+C — graceful shutdown, returns last scan’s exit code

CI Mode

Use --exit-on for pipelines that should fail fast:

goldencheck watch data/ --exit-on error --interval 10

Exits with code 1 as soon as any scan produces an error.

Output

[14:30:15] Watching data/ (*.csv, *.parquet) — polling every 30s
[14:30:15] Scanned orders.csv — 3 errors, 5 warnings (B)
[14:30:15] Scanned customers.csv — 0 errors, 1 warning (A)
[14:31:45] orders.csv changed — re-scanning...
[14:31:46] Scanned orders.csv — 2 errors, 5 warnings (B)