fix activities' types

bincio/extract/cli.py

@@ -60,7 +60,7 @@ def _process_file(path: Path) -> dict:
     if activity.source_hash in _known_hashes:
         return {"status": "duplicate"}
 
-    # Enrich from Strava CSV
+    # Enrich from Strava CSV (CSV is authoritative for sport on Strava activities)
     row = _strava_lookup.get(activity.source_file)
     if row:
         if not activity.title:
@@ -69,6 +69,10 @@ def _process_file(path: Path) -> dict:
             activity.description = row.get("Activity Description", "").strip() or None
         if not activity.strava_id:
             activity.strava_id = row.get("Activity ID", "").strip() or None
+        csv_type = row.get("Activity Type", "").strip()
+        if csv_type:
+            from bincio.extract.sport import normalise_sport
+            activity.sport = normalise_sport(csv_type)
 
     try:
         metrics = compute(activity)

bincio/extract/parsers/fit.py

@@ -16,7 +16,7 @@ class FitParser:
 
         points: list[DataPoint] = []
         laps: list[LapData] = []
-        sport: str = "cycling"
+        sport: str = "other"
         sub_sport: str | None = None
         device: str | None = None
 
@@ -26,9 +26,19 @@ class FitParser:
                     continue
 
                 if frame.name == "sport":
-                    sport = normalise_sport(_get(frame, "sport", "cycling"))
+                    sport = normalise_sport(_get(frame, "sport"))
                     sub_sport = _normalise_sub_sport(_get(frame, "sub_sport"))
 
+                elif frame.name == "session":
+                    # Karoo and Strava-generated FIT files store sport here
+                    # instead of (or in addition to) a separate 'sport' message.
+                    # Only use session sport if no 'sport' frame was seen yet.
+                    if sport == "other":
+                        raw_sport = _get(frame, "sport")
+                        if raw_sport is not None:
+                            sport = normalise_sport(raw_sport)
+                            sub_sport = _normalise_sub_sport(_get(frame, "sub_sport"))
+
                 elif frame.name == "device_info":
                     mfr = _get(frame, "manufacturer")
                     prod = _get(frame, "product_name") or _get(frame, "garmin_product")

bincio/extract/sport.py

@@ -1,33 +1,53 @@
 """Sport name normalisation."""
 
+import re
+
 _MAPPING: dict[str, str] = {
-    # cycling variants
+    # cycling variants (FIT enums, GPX types, Strava API/CSV types)
     "cycling": "cycling",
     "biking": "cycling",
     "bike": "cycling",
+    "ride": "cycling",
     "road_biking": "cycling",
+    "road_cycling": "cycling",
     "mountain_biking": "cycling",
+    "mountain_bike_ride": "cycling",
     "gravel_cycling": "cycling",
+    "gravel_ride": "cycling",
     "cyclocross": "cycling",
     "indoor_cycling": "cycling",
+    "indoor_ride": "cycling",
     "virtual_ride": "cycling",
-    "e-biking": "cycling",
+    "e_biking": "cycling",
+    "ebikeride": "cycling",
+    "e_bike_ride": "cycling",
+    "ebike_ride": "cycling",
+    "handcycle": "cycling",
+    "velomobile": "cycling",
     # running
     "running": "running",
     "run": "running",
     "trail_running": "running",
+    "trail_run": "running",
     "treadmill_running": "running",
+    "treadmill": "running",
     "virtual_run": "running",
+    "outdoor_run": "running",
+    "indoor_run": "running",
+    "track_run": "running",
     # hiking
     "hiking": "hiking",
     "hike": "hiking",
+    "outdoor_hike": "hiking",
     # walking
     "walking": "walking",
     "walk": "walking",
+    "outdoor_walk": "walking",
     # swimming
     "swimming": "swimming",
     "swim": "swimming",
     "open_water_swimming": "swimming",
+    "lap_swimming": "swimming",
 }
 
 BAS_SPORTS = {"cycling", "running", "hiking", "walking", "swimming", "other"}
@@ -36,5 +56,7 @@ BAS_SPORTS = {"cycling", "running", "hiking", "walking", "swimming", "other"}
 def normalise_sport(raw: object) -> str:
     if raw is None:
         return "other"
-    key = str(raw).lower().strip().replace(" ", "_")
+    key = str(raw).lower().strip().replace(" ", "_").replace("-", "_")
+    # Strip leading date-like prefixes e.g. "20231117outdoor_run" → "outdoor_run"
+    key = re.sub(r"^\d+", "", key)
     return _MAPPING.get(key, "other")