trying to get sub label showed properly

bincio/extract/cli.py

@@ -113,6 +113,8 @@ def _process_file(path: Path) -> dict:
               help="Only process files modified after this date.")
 @click.option("--workers", default=None, type=int,
               help="Parallel worker processes (default: CPU count).")
+@click.option("--dev", "dev_sample", default=None, type=int, metavar="N",
+              help="Dev mode: sample N files evenly across the full list, output to /tmp/bincio_dev/.")
 def extract(
     config_path: Optional[str],
     input_dir: Optional[str],
@@ -120,6 +122,7 @@ def extract(
     single_file: Optional[str],
     since: Optional[str],
     workers: Optional[int],
+    dev_sample: Optional[int],
 ) -> None:
     """Parse GPX/FIT/TCX files and write BAS JSON data store."""
 
@@ -128,13 +131,25 @@ def extract(
         return
 
     cfg = _resolve_config(config_path, input_dir, output_dir)
+
+    if dev_sample is not None:
+        cfg.output_dir = Path("/tmp/bincio_dev")
+        cfg.incremental = False
+        console.print(f"[yellow]Dev mode:[/yellow] sampling {dev_sample} files → [cyan]{cfg.output_dir}[/cyan]")
+
     cfg.output_dir.mkdir(parents=True, exist_ok=True)
 
     files = _collect_files(cfg, since)
     if not files:
         console.print("[yellow]No supported files found.[/yellow]")
         return
-    console.print(f"Found [bold]{len(files)}[/bold] activity files.")
+
+    if dev_sample is not None:
+        total = len(files)
+        files = _sample_diverse(files, dev_sample)
+        console.print(f"Sampled [bold]{len(files)}[/bold] files from {total} total.")
+    else:
+        console.print(f"Found [bold]{len(files)}[/bold] activity files.")
 
     # Build strava lookup once (serialised dict, sent to workers via initializer)
     strava_lookup: dict = {}
@@ -314,6 +329,15 @@ def _load_existing_summaries(output_dir: Path) -> list[dict]:
         return []
 
 
+def _sample_diverse(files: list[Path], n: int) -> list[Path]:
+    """Return n files sampled evenly across the sorted list for date/format diversity."""
+    if len(files) <= n:
+        return files
+    files = sorted(files)
+    step = len(files) / n
+    return [files[int(i * step)] for i in range(n)]
+
+
 def _patch_duplicate_of(output_dir: Path, activity_id: str, canonical_id: str) -> None:
     p = output_dir / "activities" / f"{activity_id}.json"
     if not p.exists():