feat: Strava support — bulk-export .gz/.tcx import fix + .tcx parser; full Strava API OAuth live sync (activities via streams) with Profile connect UI; colour gym/no-GPS sports red

backend/app/services/fit_parser.py

@@ -406,6 +406,143 @@ def parse_gpx_file(filepath: str) -> dict:
     }
 
 
+def parse_tcx_file(filepath: str) -> dict:
+    """Parse a Garmin Training Center XML (.tcx) activity file.
+
+    Strava bulk exports include older device uploads as .tcx (often gzipped).
+    TCX is namespaced XML; we match by local tag name so the parser is robust to
+    the various TCX namespace declarations in the wild. Output mirrors
+    parse_gpx_file so downstream ingestion is identical."""
+    import xml.etree.ElementTree as ET
+
+    def local(tag: str) -> str:
+        return tag.split("}")[-1] if "}" in tag else tag
+
+    def find(el, name):
+        for child in el.iter():
+            if local(child.tag) == name:
+                return child
+        return None
+
+    def findall(el, name):
+        return [c for c in el.iter() if local(c.tag) == name]
+
+    def child_text(el, name):
+        for c in list(el):
+            if local(c.tag) == name and c.text:
+                return c.text.strip()
+        return None
+
+    tree = ET.parse(filepath)
+    root = tree.getroot()
+
+    activity_el = find(root, "Activity")
+    sport_raw = (activity_el.get("Sport") if activity_el is not None else None) or "Other"
+    sport = {"running": "running", "biking": "cycling",
+             "walking": "walking", "hiking": "hiking"}.get(sport_raw.lower(), sport_raw.lower())
+
+    data_points = []
+    for tp in findall(root, "Trackpoint"):
+        ts_str = child_text(tp, "Time")
+        ts = None
+        if ts_str:
+            try:
+                ts = datetime.fromisoformat(ts_str.replace("Z", "+00:00"))
+                if ts.tzinfo is None:
+                    ts = ts.replace(tzinfo=timezone.utc)
+            except ValueError:
+                ts = None
+
+        lat = lng = None
+        pos = next((c for c in list(tp) if local(c.tag) == "Position"), None)
+        if pos is not None:
+            lat = _safe_float(child_text(pos, "LatitudeDegrees"))
+            lng = _safe_float(child_text(pos, "LongitudeDegrees"))
+
+        hr = None
+        hr_el = next((c for c in list(tp) if local(c.tag) == "HeartRateBpm"), None)
+        if hr_el is not None:
+            hr = _safe_float(child_text(hr_el, "Value"))
+
+        # Speed/Watts live in a TPX extension; search descendants by local name.
+        speed = watts = None
+        for ext in findall(tp, "Speed"):
+            speed = _safe_float(ext.text)
+            break
+        for ext in findall(tp, "Watts"):
+            watts = _safe_float(ext.text)
+            break
+
+        data_points.append({
+            "timestamp": ts.isoformat() if ts else None,
+            "latitude": lat, "longitude": lng,
+            "altitude_m": _safe_float(child_text(tp, "AltitudeMeters")),
+            "heart_rate": hr,
+            "cadence": _safe_float(child_text(tp, "Cadence")),
+            "speed_ms": speed,
+            "power": watts,
+            "temperature_c": None,
+            "distance_m": _safe_float(child_text(tp, "DistanceMeters")),
+        })
+
+    coords = [(p["latitude"], p["longitude"]) for p in data_points if p["latitude"] and p["longitude"]]
+    encoded_polyline = polyline_lib.encode(coords) if coords else None
+    bounding_box = _bounding_box(coords)
+
+    # Distance: prefer the cumulative DistanceMeters from the file; fall back to
+    # haversine over GPS points when absent (some TCX trackpoints omit it).
+    dist_vals = [p["distance_m"] for p in data_points if p["distance_m"] is not None]
+    if dist_vals:
+        total_dist = max(dist_vals)
+    else:
+        total_dist = 0.0
+        prev = None
+        for p in data_points:
+            if p["latitude"] and p["longitude"]:
+                if prev:
+                    R = 6371000
+                    phi1, phi2 = math.radians(prev[0]), math.radians(p["latitude"])
+                    dphi = math.radians(p["latitude"] - prev[0])
+                    dlam = math.radians(p["longitude"] - prev[1])
+                    a = math.sin(dphi/2)**2 + math.cos(phi1)*math.cos(phi2)*math.sin(dlam/2)**2
+                    total_dist += 2 * R * math.asin(math.sqrt(a))
+                prev = (p["latitude"], p["longitude"])
+            p["distance_m"] = total_dist
+
+    uphill, downhill = 0.0, 0.0
+    alts = [p["altitude_m"] for p in data_points if p["altitude_m"] is not None]
+    for i in range(1, len(alts)):
+        diff = alts[i] - alts[i-1]
+        if diff > 0: uphill += diff
+        else: downhill += abs(diff)
+
+    hrs = [p["heart_rate"] for p in data_points if p["heart_rate"]]
+    start_time_str = next((p["timestamp"] for p in data_points if p["timestamp"]), None)
+    last_time_str = next((p["timestamp"] for p in reversed(data_points) if p["timestamp"]), None)
+    start_dt = datetime.fromisoformat(start_time_str) if start_time_str else None
+    end_dt = datetime.fromisoformat(last_time_str) if last_time_str else None
+    duration = (end_dt - start_dt).total_seconds() if (start_dt and end_dt) else None
+    avg_speed = (total_dist / duration) if (total_dist and duration) else None
+
+    return {
+        "name": f"{sport.title()} {start_dt.date() if start_dt else ''}".strip(),
+        "sport_type": sport, "start_time": start_time_str,
+        "distance_m": total_dist or None, "duration_s": duration, "moving_time_s": None,
+        "elevation_gain_m": uphill, "elevation_loss_m": downhill,
+        "avg_heart_rate": (sum(hrs) / len(hrs)) if hrs else None,
+        "max_heart_rate": max(hrs) if hrs else None,
+        "avg_cadence": None, "avg_power": None, "normalized_power": None,
+        "avg_speed_ms": avg_speed,
+        "max_speed_ms": None, "avg_temperature_c": None, "calories": None,
+        "training_stress_score": None, "vo2max_estimate": None,
+        "polyline": encoded_polyline, "bounding_box": bounding_box,
+        "active_spans": _active_spans(data_points),
+        "source_type": "tcx",
+        "rejected_reason": _vehicle_reason(sport, avg_speed, total_dist, duration),
+        "data_points": data_points, "laps": [],
+    }
+
+
 def calculate_hr_zones(data_points: list, user_max_hr: float) -> dict:
     if not user_max_hr or user_max_hr < 100:
         return {}

backend/app/services/strava_sync.py

@@ -0,0 +1,290 @@
+"""
+Strava API sync helpers.
+
+OAuth: exchange_code() / refresh_tokens() talk to Strava's token endpoint.
+get_valid_access_token() transparently refreshes an expired access token and
+persists the rotated refresh token.
+
+sync_strava_activities() lists the athlete's activities, fetches per-activity
+streams, builds a fit_parser-shaped dict and hands it to persist_activity()
+(shared with the file-upload path) so Strava activities get identical dedup,
+PR, route and segment handling.
+
+Tokens are Fernet-encrypted with the same SECRET_KEY scheme as Garmin creds.
+"""
+import logging
+from datetime import date, datetime, timedelta, timezone
+from typing import Optional
+
+import httpx
+import polyline as polyline_lib
+
+from app.services.fit_parser import _bounding_box, _active_spans
+from app.services.garmin_connect_sync import _fernet  # reuse the SECRET_KEY-derived Fernet
+
+logger = logging.getLogger(__name__)
+
+STRAVA_AUTHORIZE_URL = "https://www.strava.com/oauth/authorize"
+STRAVA_TOKEN_URL = "https://www.strava.com/oauth/token"
+STRAVA_API_BASE = "https://www.strava.com/api/v3"
+
+# Scope needed to read all activities (including those marked "Only You").
+STRAVA_SCOPE = "read,activity:read_all"
+
+# Like Garmin: incremental syncs only re-scan the last day or two for late edits.
+INCREMENTAL_BUFFER_DAYS = 1
+
+# Strava activity type / sport_type → MileVault internal sport_type. Gym types map
+# onto the internal vocabulary so sportColor() paints them red, GPS types green/etc.
+STRAVA_SPORT_MAP = {
+    "run": "running", "trailrun": "running", "virtualrun": "running",
+    "ride": "cycling", "virtualride": "cycling", "mountainbikeride": "cycling",
+    "gravelride": "cycling", "ebikeride": "cycling", "emountainbikeride": "cycling",
+    "handcycle": "cycling", "velomobile": "cycling",
+    "walk": "walking", "hike": "hiking",
+    "swim": "swimming",
+    "weighttraining": "strength_training", "workout": "training",
+    "crossfit": "hiit", "hiit": "hiit",
+    "elliptical": "fitness_equipment", "stairstepper": "fitness_equipment",
+    "golf": "golf", "yoga": "yoga",
+}
+
+
+def encrypt_token(token: str) -> str:
+    return _fernet().encrypt(token.encode()).decode()
+
+
+def decrypt_token(enc: str) -> str:
+    return _fernet().decrypt(enc.encode()).decode()
+
+
+# ── OAuth ──────────────────────────────────────────────────────────────────────
+
+def _token_request(payload: dict) -> dict:
+    from app.core.config import settings
+    payload = {
+        "client_id": settings.strava_client_id,
+        "client_secret": settings.strava_client_secret,
+        **payload,
+    }
+    with httpx.Client(timeout=30) as client:
+        resp = client.post(STRAVA_TOKEN_URL, data=payload)
+    resp.raise_for_status()
+    return resp.json()
+
+
+def exchange_code(code: str) -> dict:
+    """Exchange an authorization code for tokens. Returns the raw token dict
+    (access_token, refresh_token, expires_at, athlete{...})."""
+    return _token_request({"code": code, "grant_type": "authorization_code"})
+
+
+def refresh_tokens(refresh_token: str) -> dict:
+    """Get a fresh access token (and possibly rotated refresh token)."""
+    return _token_request({"refresh_token": refresh_token, "grant_type": "refresh_token"})
+
+
+def get_valid_access_token(cfg, db) -> str:
+    """Return a usable access token for `cfg`, refreshing and persisting if the
+    current one is expired or about to expire. `cfg` is a StravaConfig row."""
+    now = datetime.now(timezone.utc)
+    expires_at = cfg.expires_at
+    if expires_at is not None and expires_at.tzinfo is None:
+        expires_at = expires_at.replace(tzinfo=timezone.utc)
+
+    if expires_at is None or expires_at <= now + timedelta(seconds=60):
+        tok = refresh_tokens(decrypt_token(cfg.refresh_token_enc))
+        cfg.access_token_enc = encrypt_token(tok["access_token"])
+        cfg.refresh_token_enc = encrypt_token(tok["refresh_token"])
+        cfg.expires_at = datetime.fromtimestamp(tok["expires_at"], tz=timezone.utc)
+        db.commit()
+        return tok["access_token"]
+
+    return decrypt_token(cfg.access_token_enc)
+
+
+# ── Activity sync ──────────────────────────────────────────────────────────────
+
+def _api_get(client: httpx.Client, token: str, path: str, **params):
+    resp = client.get(
+        f"{STRAVA_API_BASE}{path}",
+        headers={"Authorization": f"Bearer {token}"},
+        params=params,
+    )
+    if resp.status_code == 429:
+        raise StravaRateLimited()
+    resp.raise_for_status()
+    return resp.json()
+
+
+class StravaRateLimited(Exception):
+    """Raised when Strava returns HTTP 429 (rate limit exceeded)."""
+
+
+def _streams_to_points(streams: dict, start_dt: datetime) -> list:
+    """Convert Strava's key_by_type streams into fit_parser-shaped data points."""
+    time_s = (streams.get("time") or {}).get("data") or []
+    if not time_s:
+        return []
+    latlng = (streams.get("latlng") or {}).get("data") or []
+    altitude = (streams.get("altitude") or {}).get("data") or []
+    heartrate = (streams.get("heartrate") or {}).get("data") or []
+    cadence = (streams.get("cadence") or {}).get("data") or []
+    watts = (streams.get("watts") or {}).get("data") or []
+    velocity = (streams.get("velocity_smooth") or {}).get("data") or []
+    temp = (streams.get("temp") or {}).get("data") or []
+    distance = (streams.get("distance") or {}).get("data") or []
+
+    def at(arr, i):
+        return arr[i] if i < len(arr) else None
+
+    points = []
+    for i, off in enumerate(time_s):
+        ll = at(latlng, i)
+        lat = ll[0] if ll else None
+        lng = ll[1] if ll else None
+        points.append({
+            "timestamp": (start_dt + timedelta(seconds=off)).isoformat(),
+            "latitude": lat, "longitude": lng,
+            "altitude_m": at(altitude, i),
+            "heart_rate": at(heartrate, i),
+            "cadence": at(cadence, i),
+            "speed_ms": at(velocity, i),
+            "power": at(watts, i),
+            "temperature_c": at(temp, i),
+            "distance_m": at(distance, i),
+        })
+    return points
+
+
+def _build_parsed(summary: dict, points: list) -> dict:
+    """Assemble a fit_parser-shaped dict from a Strava summary + stream points."""
+    raw_type = (summary.get("sport_type") or summary.get("type") or "workout")
+    sport = STRAVA_SPORT_MAP.get(str(raw_type).lower(), str(raw_type).lower())
+
+    start_str = summary.get("start_date")  # UTC ISO, e.g. 2020-01-01T08:00:00Z
+    start_dt = datetime.fromisoformat(start_str.replace("Z", "+00:00")) if start_str else None
+
+    coords = [(p["latitude"], p["longitude"]) for p in points if p["latitude"] and p["longitude"]]
+    encoded = polyline_lib.encode(coords) if coords else (summary.get("map") or {}).get("summary_polyline")
+
+    # Elevation loss from the altitude stream (summary only gives gain).
+    alts = [p["altitude_m"] for p in points if p["altitude_m"] is not None]
+    downhill = sum(max(0.0, alts[i-1] - alts[i]) for i in range(1, len(alts))) if alts else None
+
+    return {
+        "name": summary.get("name") or f"{sport.title()} {start_dt.date() if start_dt else ''}".strip(),
+        "sport_type": sport,
+        "start_time": start_dt.isoformat() if start_dt else None,
+        "distance_m": summary.get("distance"),
+        "duration_s": summary.get("elapsed_time"),
+        "moving_time_s": summary.get("moving_time"),
+        "elevation_gain_m": summary.get("total_elevation_gain"),
+        "elevation_loss_m": downhill,
+        "avg_heart_rate": summary.get("average_heartrate"),
+        "max_heart_rate": summary.get("max_heartrate"),
+        "avg_cadence": summary.get("average_cadence"),
+        "avg_power": summary.get("average_watts"),
+        "normalized_power": summary.get("weighted_average_watts"),
+        "avg_speed_ms": summary.get("average_speed"),
+        "max_speed_ms": summary.get("max_speed"),
+        "active_spans": _active_spans(points),
+        "avg_temperature_c": summary.get("average_temp"),
+        "calories": summary.get("calories"),
+        "training_stress_score": None,
+        "vo2max_estimate": None,
+        "polyline": encoded,
+        "bounding_box": _bounding_box(coords),
+        "source_type": "strava",
+        "rejected_reason": None,
+        "data_points": points,
+        "laps": [],
+    }
+
+
+def sync_strava_activities(cfg, user_id: int, db, lookback_days: int = 30,
+                           status_callback=None) -> int:
+    """List the athlete's activities since the appropriate window, fetch streams
+    for each new one and persist it. Returns the number of activities imported."""
+    import time as _time
+    from app.workers.tasks import persist_activity
+
+    since = cfg.last_sync_at
+    if since is not None and since.tzinfo is None:
+        since = since.replace(tzinfo=timezone.utc)
+
+    if since:
+        after_dt = since - timedelta(days=INCREMENTAL_BUFFER_DAYS)
+    elif lookback_days == -1:
+        after_dt = datetime(2010, 1, 1, tzinfo=timezone.utc)
+    else:
+        after_dt = datetime.now(timezone.utc) - timedelta(days=max(lookback_days, 1))
+    after_epoch = int(after_dt.timestamp())
+
+    token = get_valid_access_token(cfg, db)
+    imported = 0
+
+    with httpx.Client(timeout=60) as client:
+        # 1) Page through activity summaries (newest first within each page).
+        summaries = []
+        page = 1
+        while True:
+            batch = _api_get(client, token, "/athlete/activities",
+                             after=after_epoch, page=page, per_page=100)
+            if not batch:
+                break
+            summaries.extend(batch)
+            if len(batch) < 100:
+                break
+            page += 1
+            _time.sleep(0.3)
+
+        total = len(summaries)
+        if status_callback:
+            status_callback(f"Syncing activities: 0/{total} imported")
+
+        # 2) For each, fetch streams and persist. Oldest-first so PRs accrue in order.
+        for idx, summary in enumerate(sorted(summaries, key=lambda s: s.get("start_date") or "")):
+            sid = str(summary.get("id") or "").strip()
+            if not sid:
+                continue
+
+            points = []
+            try:
+                streams = _api_get(
+                    client, token, f"/activities/{sid}/streams",
+                    keys="time,latlng,altitude,heartrate,cadence,watts,velocity_smooth,temp,distance",
+                    key_by_type="true",
+                )
+                start_str = summary.get("start_date")
+                start_dt = datetime.fromisoformat(start_str.replace("Z", "+00:00")) if start_str else None
+                if start_dt:
+                    points = _streams_to_points(streams, start_dt)
+            except StravaRateLimited:
+                # Out of quota — stop here; the next scheduled sync resumes the rest.
+                if status_callback:
+                    status_callback(f"Rate-limited by Strava — imported {imported}, will resume next sync")
+                logger.warning("Strava rate limit hit for user %s after %d imports", user_id, imported)
+                break
+            except Exception as exc:
+                logger.warning("Failed to fetch streams for Strava activity %s: %s", sid, exc)
+                # Still persist from the summary (map polyline only, no per-point data).
+
+            parsed = _build_parsed(summary, points)
+            if not parsed.get("start_time"):
+                continue
+
+            try:
+                result = persist_activity(db, user_id, parsed, strava_activity_id=sid)
+                if result.get("status") == "ok":
+                    imported += 1
+            except Exception as exc:
+                logger.warning("Failed to persist Strava activity %s: %s", sid, exc)
+                db.rollback()
+
+            if status_callback and (idx % 5 == 0 or idx == total - 1):
+                status_callback(f"Syncing activities: {imported}/{total} imported")
+
+            _time.sleep(0.3)   # be gentle on the rate limit
+
+    return imported