trying to get sub label showed properly

CHEATSHEET.md

@@ -25,11 +25,24 @@ uv run bincio extract --since 2025-01-01        # only files newer than date
 uv run bincio extract --file ride.gpx           # single file → JSON on stdout
 uv run bincio extract --input ~/rides \
                       --output ~/bincio_data    # override config paths
+uv run bincio extract --dev 50                  # dev mode: 50 files → /tmp/bincio_dev/
 ```
 
 Re-extraction is safe — unchanged files are skipped (hash-based dedup).
 To force a full re-extract: `rm -rf ~/bincio_data && uv run bincio extract`
 
+### Dev mode
+
+`--dev N` samples N files evenly across the full file list (spread by date and format)
+and writes to `/tmp/bincio_dev/` so your real data is never touched. Use it for fast
+iteration on UI or pipeline changes:
+
+```bash
+uv run bincio extract --dev 50
+uv run bincio import strava --dev 50     # N most recent Strava activities
+uv run bincio render --serve --data-dir /tmp/bincio_dev
+```
+
 ---
 
 ## Import from Strava
@@ -54,6 +67,7 @@ uv run bincio import strava
 uv run bincio import strava --since 2025-01-01   # explicit date cutoff
 uv run bincio import strava --reauth             # force new OAuth flow
 uv run bincio import strava --output ~/other_dir # override output dir
+uv run bincio import strava --dev 50             # dev mode: 50 most recent → /tmp/bincio_dev/
 ```
 
 Credentials resolution order:

CLAUDE.md

@@ -85,19 +85,22 @@ site/                      Astro project
 ## How to run
 
 ```bash
-# Extract from local files
+# Fast dev loop (50-file sample → /tmp/bincio_dev/, no real data touched)
+uv run bincio extract --dev 50
+uv run bincio import strava --dev 50   # 50 most recent Strava activities
+uv run bincio render --serve --data-dir /tmp/bincio_dev
+
+# Full extract from local files
 cd ~/src/bincio_activity
-uv run bincio extract --input ~/src/cycling_data_davide/activities --output /tmp/bincio_test
+uv run bincio extract                  # uses extract_config.yaml
 
 # Import from Strava (credentials in extract_config.yaml under import.strava)
 uv sync --extra strava
 uv run bincio import strava          # first run opens browser for OAuth
 uv run bincio import strava          # subsequent runs are incremental
 
-# Site dev server
+# Site dev server (render handles symlink + merge automatically)
-cd site
+uv run bincio render --serve
-ln -sf /tmp/bincio_test public/data   # symlink data
-npm run dev
 
 # Edit server (enables drawer + file upload in the site)
 uv run bincio edit --data-dir ~/bincio_data   # port 4041

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():

bincio/extract/parsers/gpx.py

@@ -8,7 +8,7 @@ import gpxpy.gpx
 
 from bincio.extract.models import DataPoint, ParsedActivity
 from bincio.extract.parsers.base import BaseParser
-from bincio.extract.sport import normalise_sport
+from bincio.extract.sport import normalise_sport, normalise_sub_sport
 
 # Known GPX extension namespaces
 _NS_GARMIN = "http://www.garmin.com/xmlschemas/TrackPointExtension/v1"
@@ -41,14 +41,15 @@ class GpxParser(BaseParser):
         if not points:
             raise ValueError(f"No trackpoints found in {path.name}")
 
-        sport = normalise_sport(
+        raw_sport = (gpx.tracks[0].type if gpx.tracks else None) or "cycling"
-            (gpx.tracks[0].type if gpx.tracks else None) or "cycling"
+        sport     = normalise_sport(raw_sport)
-        )
+        sub_sport = normalise_sub_sport(raw_sport)
         started_at = points[0].timestamp
 
         return ParsedActivity(
             points=points,
             sport=sport,
+            sub_sport=sub_sport,
             started_at=started_at,
             source_file=path.name,
             source_hash="",  # set by factory

bincio/extract/parsers/tcx.py

@@ -6,7 +6,7 @@ from pathlib import Path
 from lxml import etree
 
 from bincio.extract.models import DataPoint, ParsedActivity
-from bincio.extract.sport import normalise_sport
+from bincio.extract.sport import normalise_sport, normalise_sub_sport
 
 _NS_HTTP  = {
     "tcx": "http://www.garmin.com/xmlschemas/TrainingCenterDatabase/v2",
@@ -33,7 +33,8 @@ class TcxParser:
         # Use the first activity
         act = activities[0]
         sport_attr = act.get("Sport", "Biking")
-        sport = normalise_sport(sport_attr)
+        sport     = normalise_sport(sport_attr)
+        sub_sport = normalise_sub_sport(sport_attr)
 
         points: list[DataPoint] = []
         for tp in act.findall(".//tcx:Trackpoint", _NS):
@@ -78,6 +79,7 @@ class TcxParser:
         return ParsedActivity(
             points=points,
             sport=sport,
+            sub_sport=sub_sport,
             started_at=points[0].timestamp,
             source_file=path.name,
             source_hash="",

bincio/extract/sport.py

@@ -47,10 +47,13 @@ _MAPPING: dict[str, str] = {
     "skiing": "skiing",
     "cross_country_skiing": "skiing",
     "nordic_skiing": "skiing",
+    "nordic_ski": "skiing",
     "downhill_skiing": "skiing",
     "alpine_skiing": "skiing",
+    "alpine_ski": "skiing",
     "skate_skiing": "skiing",
     "backcountry_skiing": "skiing",
+    "backcountry_ski": "skiing",
     # swimming
     "swimming": "swimming",
     "swim": "swimming",
@@ -58,13 +61,63 @@ _MAPPING: dict[str, str] = {
     "lap_swimming": "swimming",
 }
 
+_SUB_SPORT_MAPPING: dict[str, str] = {
+    # cycling
+    "road_biking":        "road",
+    "road_cycling":       "road",
+    "mountain_biking":    "mountain",
+    "mountain_bike_ride": "mountain",
+    "gravel_cycling":     "gravel",
+    "gravel_ride":        "gravel",
+    "cyclocross":         "gravel",
+    "indoor_cycling":     "indoor",
+    "indoor_ride":        "indoor",
+    "virtual_ride":       "indoor",
+    # running
+    "trail_running":      "trail",
+    "trail_run":          "trail",
+    "treadmill_running":  "indoor",
+    "treadmill":          "indoor",
+    "indoor_run":         "indoor",
+    "virtual_run":        "indoor",
+    "track_run":          "track",
+    # skiing
+    "cross_country_skiing":  "nordic",
+    "nordic_skiing":         "nordic",
+    "nordic_ski":            "nordic",
+    "skate_skiing":          "nordic",
+    "backcountry_skiing":    "nordic",
+    "backcountry_ski":       "nordic",
+    "downhill_skiing":       "alpine",
+    "alpine_skiing":         "alpine",
+    "alpine_ski":            "alpine",
+    # swimming
+    "open_water_swimming":   "open_water",
+    "lap_swimming":          "pool",
+}
+
 BAS_SPORTS = {"cycling", "running", "hiking", "walking", "swimming", "skiing", "other"}
 
 
+def _normalise_key(raw: object) -> str:
+    key = str(raw).strip()
+    # CamelCase → snake_case  ("MountainBikeRide" → "mountain_bike_ride")
+    key = re.sub(r"([A-Z])", r"_\1", key).lower().lstrip("_")
+    key = key.replace(" ", "_").replace("-", "_")
+    return re.sub(r"^\d+", "", key)
+
+
 def normalise_sport(raw: object) -> str:
     if raw is None:
         return "other"
-    key = str(raw).lower().strip().replace(" ", "_").replace("-", "_")
+    return _MAPPING.get(_normalise_key(raw), "other")
-    # Strip leading date-like prefixes e.g. "20231117outdoor_run" → "outdoor_run"
+
-    key = re.sub(r"^\d+", "", key)
+
-    return _MAPPING.get(key, "other")
+def normalise_sub_sport(raw: object) -> str | None:
+    """Infer sub_sport from a raw sport type string (e.g. 'mountain_bike_ride' → 'mountain').
+
+    Returns None when no sub_sport is implied (e.g. plain 'ride', 'run').
+    """
+    if raw is None:
+        return None
+    return _SUB_SPORT_MAPPING.get(_normalise_key(raw))

bincio/import_/cli.py

@@ -29,6 +29,8 @@ def import_group() -> None:
               help="Only import activities after this date (default: incremental from last sync).")
 @click.option("--reauth", is_flag=True, default=False,
               help="Force re-authorization even if valid tokens exist.")
+@click.option("--dev", "dev_sample", default=None, type=int, metavar="N",
+              help="Dev mode: import only the N most recent activities, output to /tmp/bincio_dev/.")
 def strava_cmd(
     client_id:    Optional[str],
     client_secret: Optional[str],
@@ -36,6 +38,7 @@ def strava_cmd(
     config_path:  Optional[str],
     since:        Optional[str],
     reauth:       bool,
+    dev_sample:   Optional[int],
 ) -> None:
     """Import activities from Strava.
 
@@ -90,7 +93,11 @@ def strava_cmd(
             "Add them to extract_config.yaml under import.strava, or pass --client-id/--client-secret."
         )
 
-    out = _resolve_output(output_dir, cfg)
+    if dev_sample is not None:
+        out = Path("/tmp/bincio_dev")
+        console.print(f"[yellow]Dev mode:[/yellow] importing {dev_sample} activities → [cyan]{out}[/cyan]")
+    else:
+        out = _resolve_output(output_dir, cfg)
     console.print(f"Output dir: [cyan]{out}[/cyan]")
 
     if reauth and TOKENS_FILE.exists():
@@ -108,7 +115,7 @@ def strava_cmd(
         except ValueError:
             raise click.BadParameter(f"Expected YYYY-MM-DD, got {since!r}", param_hint="--since")
 
-    strava_sync(client, out, since_dt, console)
+    strava_sync(client, out, since_dt, console, limit=dev_sample)
 
 
 def _load_config(config_path: Optional[str]):

bincio/import_/strava.py

@@ -26,7 +26,7 @@ from rich.console import Console
 from rich.progress import BarColumn, MofNCompleteColumn, Progress, TextColumn, TimeElapsedColumn
 
 from bincio.extract.models import DataPoint, ParsedActivity
-from bincio.extract.sport import normalise_sport
+from bincio.extract.sport import normalise_sport, normalise_sub_sport
 
 STRAVA_AUTH_URL = "https://www.strava.com/oauth/authorize"
 STRAVA_TOKEN_URL = "https://www.strava.com/oauth/token"
@@ -214,7 +214,9 @@ def _strava_to_parsed(act: dict, streams: dict[str, list]) -> ParsedActivity:
     """Build a ParsedActivity from a Strava activity dict + its streams."""
     started_at = datetime.fromisoformat(act["start_date"].replace("Z", "+00:00"))
 
-    sport = normalise_sport(act.get("sport_type") or act.get("type") or "")
+    raw_sport = act.get("sport_type") or act.get("type") or ""
+    sport     = normalise_sport(raw_sport)
+    sub_sport = normalise_sub_sport(raw_sport)
 
     times      = streams.get("time", [])          # seconds since start
     latlngs    = streams.get("latlng", [])         # [[lat, lon], ...]
@@ -244,6 +246,7 @@ def _strava_to_parsed(act: dict, streams: dict[str, list]) -> ParsedActivity:
     return ParsedActivity(
         points      = points,
         sport       = sport,
+        sub_sport   = sub_sport,
         started_at  = started_at,
         source_file = f"strava_{strava_id}",
         source_hash = source_hash,
@@ -287,6 +290,7 @@ def sync(
     output_dir: Path,
     since: datetime | None,
     console: Console,
+    limit: int | None = None,
 ) -> None:
     """Fetch new Strava activities and write BAS JSON files.
 
@@ -323,6 +327,9 @@ def sync(
         f"Found [bold]{len(new_acts)}[/bold] new activities "
         f"([bold]{len(all_acts) - len(new_acts)}[/bold] already imported)."
     )
+    if limit is not None and len(new_acts) > limit:
+        new_acts = new_acts[:limit]
+        console.print(f"[yellow]Dev mode:[/yellow] capped to {limit} activities.")
     if not new_acts:
         console.print("[green]All up to date.[/green]")
         return

site/src/components/ActivityDetail.svelte

@@ -158,8 +158,16 @@
         class="text-xs font-medium px-2 py-0.5 rounded-full"
         style="background:{color}22;color:{color}"
       >
-        {sportIcon(activity.sport)} {sportLabel(activity.sport, activity.sub_sport)}
+        {sportIcon(activity.sport)} {sportLabel(activity.sport)}
       </span>
+      {#if activity.sub_sport && activity.sub_sport !== 'generic'}
+        <span
+          class="text-xs font-medium px-2 py-0.5 rounded-full"
+          style="background:{color}11;color:{color}cc"
+        >
+          {sportLabel(activity.sport, activity.sub_sport).split(' ')[0]}
+        </span>
+      {/if}
       <span class="text-xs text-zinc-500">
         {formatDate(activity.started_at)} · {formatTime(activity.started_at)}
       </span>

site/src/lib/format.ts

@@ -75,10 +75,24 @@ export function sportColor(sport: Sport): string {
   return SPORT_COLORS[sport] ?? '#a78bfa';
 }
 
+const SUB_SPORT_LABELS: Record<string, string> = {
+  road:       'Road',
+  mountain:   'MTB',
+  gravel:     'Gravel',
+  indoor:     'Indoor',
+  trail:      'Trail',
+  track:      'Track',
+  nordic:     'Nordic',
+  alpine:     'Alpine',
+  open_water: 'Open Water',
+  pool:       'Pool',
+};
+
 export function sportLabel(sport: Sport, subSport?: string | null): string {
   const base = sport.charAt(0).toUpperCase() + sport.slice(1);
   if (subSport && subSport !== 'generic') {
-    return `${subSport.charAt(0).toUpperCase() + subSport.slice(1)} ${base}`;
+    const sub = SUB_SPORT_LABELS[subSport] ?? (subSport.charAt(0).toUpperCase() + subSport.slice(1));
+    return `${sub} ${base}`;
   }
   return base;
 }

site/src/lib/types.ts

@@ -1,7 +1,7 @@
 /** TypeScript types mirroring BAS v1.0 schema. */
 
 export type Sport = "cycling" | "running" | "hiking" | "walking" | "swimming" | "skiing" | "other";
-export type SubSport = "road" | "mountain" | "gravel" | "indoor" | "trail" | "track" | "nordic" | null;
+export type SubSport = "road" | "mountain" | "gravel" | "indoor" | "trail" | "track" | "nordic" | "alpine" | "open_water" | "pool" | null;
 export type Privacy = "public" | "blur_start" | "no_gps" | "private";
 
 /** [duration_s, avg_watts] pairs, sorted by duration ascending. */