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/api/strava_sync.py

@@ -0,0 +1,271 @@
+from fastapi import APIRouter, Depends, HTTPException
+from fastapi.responses import RedirectResponse
+from sqlalchemy.ext.asyncio import AsyncSession
+from sqlalchemy import select
+from pydantic import BaseModel
+from typing import Optional
+from datetime import datetime, timezone, timedelta
+from urllib.parse import urlencode
+
+from jose import jwt, JWTError
+
+from app.core.database import get_db
+from app.core.security import get_current_user, create_access_token
+from app.core.config import settings
+from app.models.user import User, StravaConfig
+from app.services.strava_sync import (
+    STRAVA_AUTHORIZE_URL, STRAVA_SCOPE, exchange_code, encrypt_token,
+)
+
+router = APIRouter()
+
+STATE_PURPOSE = "strava-link"
+
+
+def _redis_client():
+    import redis as redis_lib
+    return redis_lib.Redis.from_url(settings.redis_url)
+
+
+def sync_task_key(user_id: int) -> str:
+    return f"strava_sync_task:{user_id}"
+
+
+def sync_cancel_key(user_id: int) -> str:
+    return f"strava_sync_cancel:{user_id}"
+
+
+def _redirect_uri() -> str:
+    return f"{settings.base_url.rstrip('/')}/api/strava-sync/callback"
+
+
+class StravaConfigIn(BaseModel):
+    sync_enabled: bool = True
+    sync_lookback_days: int = 30
+
+
+class StravaConfigOut(BaseModel):
+    connected: bool
+    athlete_name: Optional[str] = None
+    sync_enabled: bool = False
+    sync_lookback_days: int = 30
+    sync_interval_minutes: int = settings.garmin_sync_interval_minutes
+    last_sync_at: Optional[datetime] = None
+    last_sync_status: Optional[str] = None
+    configured: bool = True   # whether the server has Strava API credentials at all
+
+
+def _out(cfg: Optional[StravaConfig]) -> StravaConfigOut:
+    configured = bool(settings.strava_client_id and settings.strava_client_secret)
+    if not cfg:
+        return StravaConfigOut(
+            connected=False, sync_enabled=False, sync_lookback_days=30,
+            sync_interval_minutes=settings.garmin_sync_interval_minutes,
+            configured=configured,
+        )
+    return StravaConfigOut(
+        connected=True,
+        athlete_name=cfg.athlete_name,
+        sync_enabled=cfg.sync_enabled,
+        sync_lookback_days=cfg.sync_lookback_days if cfg.sync_lookback_days is not None else 30,
+        sync_interval_minutes=settings.garmin_sync_interval_minutes,
+        last_sync_at=cfg.last_sync_at,
+        last_sync_status=cfg.last_sync_status,
+        configured=configured,
+    )
+
+
+@router.get("/config", response_model=StravaConfigOut)
+async def get_config(
+    db: AsyncSession = Depends(get_db),
+    current_user: User = Depends(get_current_user),
+):
+    cfg = (await db.execute(
+        select(StravaConfig).where(StravaConfig.user_id == current_user.id)
+    )).scalar_one_or_none()
+    return _out(cfg)
+
+
+@router.get("/authorize")
+async def authorize(current_user: User = Depends(get_current_user)):
+    """Return the Strava OAuth URL to redirect the browser to. A signed `state`
+    carries the user id through the callback (which has no JWT header)."""
+    if not (settings.strava_client_id and settings.strava_client_secret):
+        raise HTTPException(status_code=400, detail="Strava API is not configured on this server")
+    state = create_access_token(
+        {"sub": str(current_user.id), "purpose": STATE_PURPOSE},
+        expires_delta=timedelta(minutes=10),
+    )
+    params = {
+        "client_id": settings.strava_client_id,
+        "response_type": "code",
+        "redirect_uri": _redirect_uri(),
+        "approval_prompt": "auto",
+        "scope": STRAVA_SCOPE,
+        "state": state,
+    }
+    return {"url": f"{STRAVA_AUTHORIZE_URL}?{urlencode(params)}"}
+
+
+@router.get("/callback")
+async def callback(
+    code: Optional[str] = None,
+    state: Optional[str] = None,
+    error: Optional[str] = None,
+    scope: Optional[str] = None,
+    db: AsyncSession = Depends(get_db),
+):
+    """Strava redirects here after the user authorizes. Exchange the code, store
+    tokens, then bounce back to the SPA profile page."""
+    profile_url = f"{settings.base_url.rstrip('/')}/profile"
+
+    if error or not code:
+        return RedirectResponse(f"{profile_url}?strava=error")
+
+    # Validate state → user id
+    try:
+        payload = jwt.decode(state or "", settings.secret_key, algorithms=[settings.algorithm])
+        if payload.get("purpose") != STATE_PURPOSE:
+            raise ValueError("bad purpose")
+        user_id = int(payload["sub"])
+    except (JWTError, KeyError, TypeError, ValueError):
+        return RedirectResponse(f"{profile_url}?strava=error")
+
+    # Require activity:read_all so private activities sync too.
+    if scope and "activity:read_all" not in scope:
+        return RedirectResponse(f"{profile_url}?strava=scope")
+
+    try:
+        tok = exchange_code(code)
+    except Exception:
+        return RedirectResponse(f"{profile_url}?strava=error")
+
+    athlete = tok.get("athlete") or {}
+    athlete_name = " ".join(
+        x for x in [athlete.get("firstname"), athlete.get("lastname")] if x
+    ).strip() or (athlete.get("username") or None)
+
+    cfg = (await db.execute(
+        select(StravaConfig).where(StravaConfig.user_id == user_id)
+    )).scalar_one_or_none()
+
+    expires_at = datetime.fromtimestamp(tok["expires_at"], tz=timezone.utc)
+    if cfg:
+        cfg.access_token_enc = encrypt_token(tok["access_token"])
+        cfg.refresh_token_enc = encrypt_token(tok["refresh_token"])
+        cfg.expires_at = expires_at
+        cfg.athlete_id = str(athlete.get("id") or "") or cfg.athlete_id
+        cfg.athlete_name = athlete_name or cfg.athlete_name
+        cfg.last_sync_status = "Connected"
+    else:
+        cfg = StravaConfig(
+            user_id=user_id,
+            athlete_id=str(athlete.get("id") or "") or None,
+            athlete_name=athlete_name,
+            access_token_enc=encrypt_token(tok["access_token"]),
+            refresh_token_enc=encrypt_token(tok["refresh_token"]),
+            expires_at=expires_at,
+            sync_enabled=True,
+            sync_lookback_days=30,
+            last_sync_status="Connected",
+        )
+        db.add(cfg)
+    await db.commit()
+
+    # Kick off an initial sync immediately.
+    try:
+        from app.workers.tasks import sync_strava_user
+        sync_strava_user.delay(user_id)
+    except Exception:
+        pass
+
+    return RedirectResponse(f"{profile_url}?strava=connected")
+
+
+@router.put("/config", response_model=StravaConfigOut)
+async def save_config(
+    body: StravaConfigIn,
+    db: AsyncSession = Depends(get_db),
+    current_user: User = Depends(get_current_user),
+):
+    cfg = (await db.execute(
+        select(StravaConfig).where(StravaConfig.user_id == current_user.id)
+    )).scalar_one_or_none()
+    if not cfg:
+        raise HTTPException(status_code=400, detail="Strava is not connected")
+
+    # Asking for more history than before → reset last_sync_at so the next sync
+    # backfills the wider window (mirrors the Garmin behaviour).
+    old = cfg.sync_lookback_days if cfg.sync_lookback_days is not None else 30
+    new = body.sync_lookback_days
+    wants_more = (new != old) and (new == -1 or (old != -1 and new > old))
+    if wants_more:
+        cfg.last_sync_at = None
+        cfg.last_sync_status = "Lookback increased — backfill on next sync"
+
+    cfg.sync_enabled = body.sync_enabled
+    cfg.sync_lookback_days = body.sync_lookback_days
+    await db.commit()
+    await db.refresh(cfg)
+    return _out(cfg)
+
+
+@router.delete("/config")
+async def delete_config(
+    db: AsyncSession = Depends(get_db),
+    current_user: User = Depends(get_current_user),
+):
+    cfg = (await db.execute(
+        select(StravaConfig).where(StravaConfig.user_id == current_user.id)
+    )).scalar_one_or_none()
+    if cfg:
+        await db.delete(cfg)
+        await db.commit()
+    return {"status": "ok"}
+
+
+@router.post("/trigger")
+async def trigger_sync(
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    cfg = (await db.execute(
+        select(StravaConfig).where(StravaConfig.user_id == current_user.id)
+    )).scalar_one_or_none()
+    if not cfg or not cfg.sync_enabled:
+        raise HTTPException(status_code=400, detail="Strava sync is not configured or disabled")
+
+    from app.workers.tasks import sync_strava_user
+    task = sync_strava_user.delay(current_user.id)
+    try:
+        r = _redis_client()
+        r.delete(sync_cancel_key(current_user.id))
+        r.set(sync_task_key(current_user.id), task.id, ex=3600)
+    except Exception:
+        pass
+    return {"task_id": task.id, "status": "queued"}
+
+
+@router.post("/cancel")
+async def cancel_sync(
+    current_user: User = Depends(get_current_user),
+    db: AsyncSession = Depends(get_db),
+):
+    from app.workers.tasks import celery_app
+    try:
+        r = _redis_client()
+        r.set(sync_cancel_key(current_user.id), "1", ex=3600)
+        task_id = r.get(sync_task_key(current_user.id))
+        if task_id:
+            tid = task_id.decode() if isinstance(task_id, (bytes, bytearray)) else task_id
+            celery_app.control.revoke(tid, terminate=False)
+    except Exception:
+        pass
+
+    cfg = (await db.execute(
+        select(StravaConfig).where(StravaConfig.user_id == current_user.id)
+    )).scalar_one_or_none()
+    if cfg:
+        cfg.last_sync_status = "Cancelling…"
+        await db.commit()
+    return {"status": "cancelling"}

backend/app/api/upload.py

@@ -74,6 +74,33 @@ def _safe_extract(zf: zipfile.ZipFile, dest_dir: Path) -> list[Path]:
     return extracted
 
 
+def _gunzip(path: Path) -> Path | None:
+    """Decompress a .gz member to a sibling file without the .gz suffix,
+    enforcing the same uncompressed-size cap. Returns the new path, or None on
+    failure. The .gz is removed once expanded."""
+    import gzip
+    out_path = path.with_suffix("")  # strips the trailing .gz
+    try:
+        total = 0
+        with gzip.open(path, "rb") as src, open(out_path, "wb") as out:
+            while True:
+                chunk = src.read(_CHUNK)
+                if not chunk:
+                    break
+                total += len(chunk)
+                if total > MAX_EXTRACT_SIZE:
+                    out.close()
+                    out_path.unlink(missing_ok=True)
+                    return None
+                out.write(chunk)
+    except (OSError, EOFError):
+        out_path.unlink(missing_ok=True)
+        return None
+    finally:
+        path.unlink(missing_ok=True)
+    return out_path
+
+
 @router.post("/activity")
 async def upload_activity(
     file: UploadFile = File(...),
@@ -81,10 +108,10 @@ async def upload_activity(
     db: AsyncSession = Depends(get_db),
     current_user: User = Depends(get_current_user),
 ):
-    """Upload a single .fit or .gpx activity file."""
+    """Upload a single .fit, .gpx or .tcx activity file."""
     suffix = Path(file.filename).suffix.lower()
-    if suffix not in {".fit", ".gpx"}:
-        raise HTTPException(status_code=400, detail="Only .fit and .gpx files are supported")
+    if suffix not in {".fit", ".gpx", ".tcx"}:
+        raise HTTPException(status_code=400, detail="Only .fit, .gpx and .tcx files are supported")
 
     dest_dir = Path(settings.file_store_path) / str(current_user.id) / "activities"
     dest = save_upload(file, dest_dir)
@@ -184,15 +211,24 @@ async def upload_strava_export(
         raise HTTPException(status_code=400, detail="Uploaded file is not a valid ZIP archive")
 
     for path in extracted:
+        # Strava compresses most exported activities as <id>.fit.gz / .gpx.gz /
+        # .tcx.gz. Decompress those to their underlying file first, otherwise the
+        # gzipped majority of the archive is silently skipped.
+        if path.suffix.lower() == ".gz":
+            inner = _gunzip(path)
+            if inner is None:
+                continue
+            path = inner
+
         suffix = path.suffix.lower()
-        if suffix in (".fit", ".gpx"):
+        if suffix in (".fit", ".gpx", ".tcx"):
             task = process_activity_file.delay(str(path), current_user.id, suffix[1:])
             task_ids.append(task.id)
 
     if not task_ids:
         raise HTTPException(
             status_code=400,
-            detail="No activity files (.fit or .gpx) found in this Strava archive",
+            detail="No activity files (.fit, .gpx or .tcx) found in this Strava archive",
         )
 
     return {

backend/app/core/config.py

@@ -24,6 +24,10 @@ class Settings(BaseSettings):
     pocketid_allowed_group: Optional[str] = Field(None, env="POCKETID_ALLOWED_GROUP")
     # Garmin Connect — how often the beat scheduler runs the automatic sync
     garmin_sync_interval_minutes: int = Field(30, env="GARMIN_SYNC_INTERVAL_MINUTES")
+    # Strava API (optional) — register an app at https://www.strava.com/settings/api.
+    # The Authorization Callback Domain there must match BASE_URL's host.
+    strava_client_id: Optional[str] = Field(None, env="STRAVA_CLIENT_ID")
+    strava_client_secret: Optional[str] = Field(None, env="STRAVA_CLIENT_SECRET")
     # Files
     file_store_path: str = Field("/data/files", env="FILE_STORE_PATH")
     # Environment

backend/app/main.py

@@ -6,7 +6,7 @@ import asyncio
 
 from app.core.database import engine, AsyncSessionLocal, Base
 from app.core.config import settings
-from app.api import auth, activities, routes, health, records, upload, profile, garmin_sync, users, segments
+from app.api import auth, activities, routes, health, records, upload, profile, garmin_sync, strava_sync, users, segments
 
 
 async def init_db():
@@ -260,6 +260,7 @@ app.include_router(records.router, prefix="/api/records", tags=["records"])
 app.include_router(upload.router, prefix="/api/upload", tags=["upload"])
 app.include_router(profile.router, prefix="/api/profile", tags=["profile"])
 app.include_router(garmin_sync.router, prefix="/api/garmin-sync", tags=["garmin-sync"])
+app.include_router(strava_sync.router, prefix="/api/strava-sync", tags=["strava-sync"])
 app.include_router(users.router, prefix="/api/users", tags=["users"])
 app.include_router(segments.router, prefix="/api/segments", tags=["segments"])
 

backend/app/models/user.py

@@ -45,6 +45,7 @@ class User(Base):
     named_routes = relationship("NamedRoute", back_populates="user", cascade="all, delete-orphan")
     weight_logs = relationship("WeightLog", back_populates="user", cascade="all, delete-orphan")
     garmin_connect_config = relationship("GarminConnectConfig", back_populates="user", uselist=False, cascade="all, delete-orphan")
+    strava_config = relationship("StravaConfig", back_populates="user", uselist=False, cascade="all, delete-orphan")
 
 
 class GarminConnectConfig(Base):
@@ -67,6 +68,27 @@ class GarminConnectConfig(Base):
     user = relationship("User", back_populates="garmin_connect_config")
 
 
+class StravaConfig(Base):
+    """Per-user Strava OAuth tokens and sync state. Tokens are Fernet-encrypted
+    (same SECRET_KEY scheme as Garmin credentials)."""
+    __tablename__ = "strava_configs"
+
+    id = Column(Integer, primary_key=True)
+    user_id = Column(Integer, ForeignKey("users.id"), nullable=False, unique=True, index=True)
+    athlete_id = Column(String(64), nullable=True)
+    athlete_name = Column(String(256), nullable=True)
+    access_token_enc = Column(String(512), nullable=False)
+    refresh_token_enc = Column(String(512), nullable=False)
+    expires_at = Column(DateTime(timezone=True), nullable=True)   # access-token expiry
+    sync_enabled = Column(Boolean, default=True)
+    sync_lookback_days = Column(Integer, default=30)   # -1 = all-time, first sync only
+    last_sync_at = Column(DateTime(timezone=True), nullable=True)
+    last_sync_status = Column(String(512), nullable=True)
+    created_at = Column(DateTime(timezone=True), default=now_utc)
+
+    user = relationship("User", back_populates="strava_config")
+
+
 class WeightLog(Base):
     """Manual weight entries separate from health_metrics for easy tracking."""
     __tablename__ = "weight_logs"

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

backend/app/workers/tasks.py

@@ -28,6 +28,12 @@ celery_app.conf.update(
             # Interval is configurable via GARMIN_SYNC_INTERVAL_MINUTES (default 30 min)
             "schedule": float(settings.garmin_sync_interval_minutes * 60),
         },
+        "sync-strava": {
+            "task": "sync_all_strava",
+            # Shares the Garmin cadence setting; Strava's rate limits make a
+            # tighter interval unwise anyway.
+            "schedule": float(settings.garmin_sync_interval_minutes * 60),
+        },
     },
 )
 
@@ -94,7 +100,7 @@ def process_activity_file(self, file_path: str, user_id: int, source_type: str,
         parse_wellness_fit.delay(file_path, user_id)
         return {"status": "routed_to_wellness", "file": file_path}
 
-    from app.services.fit_parser import parse_fit_file, parse_gpx_file, calculate_hr_zones
+    from app.services.fit_parser import parse_fit_file, parse_gpx_file, parse_tcx_file, calculate_hr_zones
     from app.core.database import SyncSessionLocal
     from app.models.user import Activity, ActivityDataPoint, ActivityLap
     from sqlalchemy import select, func
@@ -105,6 +111,8 @@ def process_activity_file(self, file_path: str, user_id: int, source_type: str,
     try:
         if source_type == "fit" or file_path.endswith(".fit"):
             parsed = parse_fit_file(file_path)
+        elif source_type == "tcx" or file_path.endswith(".tcx"):
+            parsed = parse_tcx_file(file_path)
         else:
             parsed = parse_gpx_file(file_path)
     except Exception as e:
@@ -122,123 +130,155 @@ def process_activity_file(self, file_path: str, user_id: int, source_type: str,
         return {"status": "skipped", "reason": parsed["rejected_reason"], "file": file_path}
 
     with SyncSessionLocal() as db:
-        start_time = datetime.fromisoformat(parsed["start_time"])
+        return persist_activity(
+            db, user_id, parsed,
+            source_file=file_path,
+            garmin_activity_id=garmin_activity_id,
+        )
 
-        # Deduplicate: same user + sport_type + start_time within ±60s
-        from datetime import timedelta
+
+def persist_activity(db, user_id: int, parsed: dict, *, source_file: str = None,
+                     garmin_activity_id: str = None, strava_activity_id: str = None) -> dict:
+    """Insert a parsed activity (+ data points, laps) and dispatch the
+    PR/route/segment follow-up tasks. Shared by the file-upload path
+    (process_activity_file) and the Strava API sync, so both get identical
+    dedup, HR-zone and downstream behaviour. `parsed` is the fit_parser-shaped
+    dict. Returns {"activity_id", "status"}."""
+    from app.models.user import Activity, ActivityDataPoint, ActivityLap, User as UserModel
+    from app.services.fit_parser import calculate_hr_zones
+    from sqlalchemy import select
+    from datetime import datetime, timedelta
+
+    start_time = datetime.fromisoformat(parsed["start_time"])
+
+    # Fast-path dedup for re-syncs: same external id already imported.
+    if strava_activity_id:
         existing = db.execute(
-            select(Activity).where(
-                Activity.user_id == user_id,
-                Activity.sport_type == parsed["sport_type"],
-                Activity.start_time >= start_time - timedelta(seconds=60),
-                Activity.start_time <= start_time + timedelta(seconds=60),
-            )
-        ).scalars().first()
-
+            select(Activity).where(Activity.strava_activity_id == str(strava_activity_id))
+        ).scalar_one_or_none()
         if existing:
-            # Stamp garmin_activity_id if this came from a Garmin Connect sync
-            # so future syncs skip the fast-path dedup and don't re-download.
-            if garmin_activity_id and not existing.garmin_activity_id:
-                existing.garmin_activity_id = garmin_activity_id
-                db.commit()
             return {"activity_id": existing.id, "status": "duplicate"}
 
-        # Get user max HR for zone calculation
-        from app.models.user import User as UserModel
-        user_obj = db.execute(
-            select(UserModel).where(UserModel.id == user_id)
-        ).scalar_one_or_none()
-
-        user_max_hr = None
-        if user_obj:
-            user_max_hr = user_obj.max_heart_rate
-            if not user_max_hr and user_obj.birth_year:
-                from datetime import date as _date
-                age = _date.today().year - user_obj.birth_year
-                user_max_hr = 220 - age
-        if not user_max_hr:
-            user_max_hr = parsed.get("max_heart_rate") or 190
-
-        hr_zones = calculate_hr_zones(parsed.get("data_points", []), user_max_hr)
-
-        activity = Activity(
-            user_id=user_id,
-            name=parsed["name"],
-            sport_type=parsed["sport_type"],
-            garmin_activity_id=garmin_activity_id,
-            start_time=start_time,
-            distance_m=parsed.get("distance_m"),
-            duration_s=parsed.get("duration_s"),
-            moving_time_s=parsed.get("moving_time_s"),
-            elevation_gain_m=parsed.get("elevation_gain_m"),
-            elevation_loss_m=parsed.get("elevation_loss_m"),
-            avg_heart_rate=parsed.get("avg_heart_rate"),
-            max_heart_rate=parsed.get("max_heart_rate"),
-            avg_cadence=parsed.get("avg_cadence"),
-            avg_power=parsed.get("avg_power"),
-            normalized_power=parsed.get("normalized_power"),
-            avg_speed_ms=parsed.get("avg_speed_ms"),
-            max_speed_ms=parsed.get("max_speed_ms"),
-            active_spans=parsed.get("active_spans"),
-            avg_temperature_c=parsed.get("avg_temperature_c"),
-            calories=parsed.get("calories"),
-            training_stress_score=parsed.get("training_stress_score"),
-            polyline=parsed.get("polyline"),
-            bounding_box=parsed.get("bounding_box"),
-            source_file=file_path,
-            source_type=parsed.get("source_type"),
-            hr_zones=hr_zones,
+    # Deduplicate across sources: same user + sport_type + start_time within ±60s.
+    # This also collapses an activity synced from both Garmin and Strava into one.
+    existing = db.execute(
+        select(Activity).where(
+            Activity.user_id == user_id,
+            Activity.sport_type == parsed["sport_type"],
+            Activity.start_time >= start_time - timedelta(seconds=60),
+            Activity.start_time <= start_time + timedelta(seconds=60),
         )
-        db.add(activity)
-        db.flush()
+    ).scalars().first()
 
-        seen = set()
-        batch = []
-        for p in parsed.get("data_points", []):
-            if not p.get("timestamp"):
-                continue
-            ts = datetime.fromisoformat(p["timestamp"]) if isinstance(p["timestamp"], str) else p["timestamp"]
-            key = (activity.id, ts)
-            if key in seen:
-                continue
-            seen.add(key)
-            batch.append(ActivityDataPoint(
-                activity_id=activity.id,
-                timestamp=ts,
-                latitude=p.get("latitude"),
-                longitude=p.get("longitude"),
-                altitude_m=p.get("altitude_m"),
-                heart_rate=p.get("heart_rate"),
-                cadence=p.get("cadence"),
-                speed_ms=p.get("speed_ms"),
-                power=p.get("power"),
-                temperature_c=p.get("temperature_c"),
-                distance_m=p.get("distance_m"),
-            ))
-            if len(batch) >= 500:
-                db.add_all(batch)
-                db.flush()
-                batch = []
-        if batch:
+    if existing:
+        # Stamp the external id so future syncs skip straight to the fast path.
+        stamped = False
+        if garmin_activity_id and not existing.garmin_activity_id:
+            existing.garmin_activity_id = garmin_activity_id
+            stamped = True
+        if strava_activity_id and not existing.strava_activity_id:
+            existing.strava_activity_id = str(strava_activity_id)
+            stamped = True
+        if stamped:
+            db.commit()
+        return {"activity_id": existing.id, "status": "duplicate"}
+
+    # Get user max HR for zone calculation
+    user_obj = db.execute(
+        select(UserModel).where(UserModel.id == user_id)
+    ).scalar_one_or_none()
+
+    user_max_hr = None
+    if user_obj:
+        user_max_hr = user_obj.max_heart_rate
+        if not user_max_hr and user_obj.birth_year:
+            from datetime import date as _date
+            age = _date.today().year - user_obj.birth_year
+            user_max_hr = 220 - age
+    if not user_max_hr:
+        user_max_hr = parsed.get("max_heart_rate") or 190
+
+    hr_zones = calculate_hr_zones(parsed.get("data_points", []), user_max_hr)
+
+    activity = Activity(
+        user_id=user_id,
+        name=parsed["name"],
+        sport_type=parsed["sport_type"],
+        garmin_activity_id=garmin_activity_id,
+        strava_activity_id=str(strava_activity_id) if strava_activity_id else None,
+        start_time=start_time,
+        distance_m=parsed.get("distance_m"),
+        duration_s=parsed.get("duration_s"),
+        moving_time_s=parsed.get("moving_time_s"),
+        elevation_gain_m=parsed.get("elevation_gain_m"),
+        elevation_loss_m=parsed.get("elevation_loss_m"),
+        avg_heart_rate=parsed.get("avg_heart_rate"),
+        max_heart_rate=parsed.get("max_heart_rate"),
+        avg_cadence=parsed.get("avg_cadence"),
+        avg_power=parsed.get("avg_power"),
+        normalized_power=parsed.get("normalized_power"),
+        avg_speed_ms=parsed.get("avg_speed_ms"),
+        max_speed_ms=parsed.get("max_speed_ms"),
+        active_spans=parsed.get("active_spans"),
+        avg_temperature_c=parsed.get("avg_temperature_c"),
+        calories=parsed.get("calories"),
+        training_stress_score=parsed.get("training_stress_score"),
+        polyline=parsed.get("polyline"),
+        bounding_box=parsed.get("bounding_box"),
+        source_file=source_file,
+        source_type=parsed.get("source_type"),
+        hr_zones=hr_zones,
+    )
+    db.add(activity)
+    db.flush()
+
+    seen = set()
+    batch = []
+    for p in parsed.get("data_points", []):
+        if not p.get("timestamp"):
+            continue
+        ts = datetime.fromisoformat(p["timestamp"]) if isinstance(p["timestamp"], str) else p["timestamp"]
+        key = (activity.id, ts)
+        if key in seen:
+            continue
+        seen.add(key)
+        batch.append(ActivityDataPoint(
+            activity_id=activity.id,
+            timestamp=ts,
+            latitude=p.get("latitude"),
+            longitude=p.get("longitude"),
+            altitude_m=p.get("altitude_m"),
+            heart_rate=p.get("heart_rate"),
+            cadence=p.get("cadence"),
+            speed_ms=p.get("speed_ms"),
+            power=p.get("power"),
+            temperature_c=p.get("temperature_c"),
+            distance_m=p.get("distance_m"),
+        ))
+        if len(batch) >= 500:
             db.add_all(batch)
             db.flush()
+            batch = []
+    if batch:
+        db.add_all(batch)
+        db.flush()
 
-        for lap in parsed.get("laps", []):
-            ls = datetime.fromisoformat(lap["start_time"]) if lap.get("start_time") else None
-            db.add(ActivityLap(
-                activity_id=activity.id,
-                lap_number=lap["lap_number"],
-                start_time=ls,
-                duration_s=lap.get("duration_s"),
-                distance_m=lap.get("distance_m"),
-                avg_heart_rate=lap.get("avg_heart_rate"),
-                avg_cadence=lap.get("avg_cadence"),
-                avg_speed_ms=lap.get("avg_speed_ms"),
-                avg_power=lap.get("avg_power"),
-            ))
+    for lap in parsed.get("laps", []):
+        ls = datetime.fromisoformat(lap["start_time"]) if lap.get("start_time") else None
+        db.add(ActivityLap(
+            activity_id=activity.id,
+            lap_number=lap["lap_number"],
+            start_time=ls,
+            duration_s=lap.get("duration_s"),
+            distance_m=lap.get("distance_m"),
+            avg_heart_rate=lap.get("avg_heart_rate"),
+            avg_cadence=lap.get("avg_cadence"),
+            avg_speed_ms=lap.get("avg_speed_ms"),
+            avg_power=lap.get("avg_power"),
+        ))
 
-        db.commit()
-        activity_id = activity.id
+    db.commit()
+    activity_id = activity.id
 
     compute_personal_records.delay(activity_id, user_id, parsed)
     if parsed.get("sport_type") in ("running", "cycling", "hiking", "walking"):
@@ -841,6 +881,96 @@ def sync_all_garmin_connect():
     return {"dispatched": len(user_ids)}
 
 
+@celery_app.task(name="sync_strava_user")
+def sync_strava_user(user_id: int):
+    """Sync activities from Strava for one user via the Strava API."""
+    from app.services.strava_sync import sync_strava_activities
+    from app.core.database import SyncSessionLocal
+    from app.models.user import StravaConfig
+    from app.core.config import settings
+    from sqlalchemy import select
+    from datetime import datetime, timezone
+
+    cancel_key = f"strava_sync_cancel:{user_id}"
+    try:
+        import redis as redis_lib
+        _redis = redis_lib.Redis.from_url(settings.redis_url)
+    except Exception:
+        _redis = None
+
+    def _cancelled():
+        try:
+            return bool(_redis and _redis.exists(cancel_key))
+        except Exception:
+            return False
+
+    with SyncSessionLocal() as db:
+        cfg = db.execute(
+            select(StravaConfig).where(StravaConfig.user_id == user_id)
+        ).scalar_one_or_none()
+        if not cfg or not cfg.sync_enabled:
+            return {"status": "skipped"}
+
+        lookback = cfg.sync_lookback_days if cfg.sync_lookback_days is not None else 30
+
+        cfg.last_sync_status = "Connecting to Strava..."
+        db.commit()
+
+        def _set_status(text):
+            if _cancelled():
+                raise SyncCancelled()
+            cfg.last_sync_status = text
+            db.commit()
+
+        try:
+            imported = sync_strava_activities(
+                cfg, user_id, db, lookback_days=lookback, status_callback=_set_status,
+            )
+        except SyncCancelled:
+            db.rollback()
+            cfg.last_sync_at = datetime.now(timezone.utc)
+            cfg.last_sync_status = "Cancelled"
+            db.commit()
+            try:
+                if _redis:
+                    _redis.delete(cancel_key)
+            except Exception:
+                pass
+            return {"status": "cancelled"}
+        except Exception as exc:
+            db.rollback()
+            cfg.last_sync_at = datetime.now(timezone.utc)
+            msg = str(exc)
+            cfg.last_sync_status = f"Auth error: {msg}" if "401" in msg or "auth" in msg.lower() else f"Error: {msg}"
+            db.commit()
+            return {"status": "error", "error": msg}
+
+        cfg.last_sync_at = datetime.now(timezone.utc)
+        cfg.last_sync_status = f"OK — {imported} activities imported"
+        db.commit()
+
+    return {"status": "ok", "activities_imported": imported}
+
+
+@celery_app.task(name="sync_all_strava")
+def sync_all_strava():
+    """Beat task: dispatch a per-user Strava sync for all enabled configs."""
+    from app.core.database import SyncSessionLocal
+    from app.models.user import StravaConfig
+    from sqlalchemy import select
+
+    with SyncSessionLocal() as db:
+        configs = db.execute(
+            select(StravaConfig).where(StravaConfig.sync_enabled == True)
+        ).scalars().all()
+        user_ids = [c.user_id for c in configs]
+
+    for uid in user_ids:
+        sync_strava_user.delay(uid)
+
+    return {"dispatched": len(user_ids)}
+
+
 @celery_app.task(name="recalculate_hr_zones_for_user")
 def recalculate_hr_zones_for_user(user_id: int, new_max_hr: float):
     """Recalculate hr_zones for all of a user's activities using a new max HR."""

docker-compose.deploy.yml

@@ -58,6 +58,9 @@ services:
       POCKETID_ISSUER: ${POCKETID_ISSUER:-}
       POCKETID_CLIENT_ID: ${POCKETID_CLIENT_ID:-}
       POCKETID_CLIENT_SECRET: ${POCKETID_CLIENT_SECRET:-}
+      STRAVA_CLIENT_ID: ${STRAVA_CLIENT_ID:-}
+      STRAVA_CLIENT_SECRET: ${STRAVA_CLIENT_SECRET:-}
+      BASE_URL: ${BASE_URL:-https://milevault.jarrett.eu}
       FILE_STORE_PATH: /data/files
       ENVIRONMENT: production
     volumes:
@@ -82,6 +85,8 @@ services:
       DATABASE_URL: postgresql+asyncpg://${DB_USER:-milevault}:${DB_PASSWORD:-milevault}@db:5432/milevault
       REDIS_URL: redis://:${REDIS_PASSWORD:-milevault}@redis:6379/0
       SECRET_KEY: ${SECRET_KEY:-changeme_run_openssl_rand_hex_32}
+      STRAVA_CLIENT_ID: ${STRAVA_CLIENT_ID:-}
+      STRAVA_CLIENT_SECRET: ${STRAVA_CLIENT_SECRET:-}
       FILE_STORE_PATH: /data/files
     volumes:
       - file_data:/data/files
@@ -100,6 +105,8 @@ services:
       DATABASE_URL: postgresql+asyncpg://${DB_USER:-milevault}:${DB_PASSWORD:-milevault}@db:5432/milevault
       REDIS_URL: redis://:${REDIS_PASSWORD:-milevault}@redis:6379/0
       SECRET_KEY: ${SECRET_KEY:-changeme_run_openssl_rand_hex_32}
+      STRAVA_CLIENT_ID: ${STRAVA_CLIENT_ID:-}
+      STRAVA_CLIENT_SECRET: ${STRAVA_CLIENT_SECRET:-}
       FILE_STORE_PATH: /data/files
     volumes:
       - file_data:/data/files

docker-compose.yml

@@ -48,6 +48,9 @@ services:
       POCKETID_ISSUER: ${POCKETID_ISSUER:-}
       POCKETID_CLIENT_ID: ${POCKETID_CLIENT_ID:-}
       POCKETID_CLIENT_SECRET: ${POCKETID_CLIENT_SECRET:-}
+      STRAVA_CLIENT_ID: ${STRAVA_CLIENT_ID:-}
+      STRAVA_CLIENT_SECRET: ${STRAVA_CLIENT_SECRET:-}
+      BASE_URL: ${BASE_URL:-https://milevault.jarrett.eu}
       FILE_STORE_PATH: /data/files
       ENVIRONMENT: ${ENVIRONMENT:-production}
     volumes:
@@ -74,6 +77,8 @@ services:
       DATABASE_URL: postgresql+asyncpg://${DB_USER:-milevault}:${DB_PASSWORD:-milevault}@db:5432/milevault
       REDIS_URL: redis://:${REDIS_PASSWORD:-milevault}@redis:6379/0
       SECRET_KEY: ${SECRET_KEY:-changeme_please_set_in_env_file_32chars}
+      STRAVA_CLIENT_ID: ${STRAVA_CLIENT_ID:-}
+      STRAVA_CLIENT_SECRET: ${STRAVA_CLIENT_SECRET:-}
       FILE_STORE_PATH: /data/files
     volumes:
       - ./file_data:/data/files
@@ -94,6 +99,8 @@ services:
       DATABASE_URL: postgresql+asyncpg://${DB_USER:-milevault}:${DB_PASSWORD:-milevault}@db:5432/milevault
       REDIS_URL: redis://:${REDIS_PASSWORD:-milevault}@redis:6379/0
       SECRET_KEY: ${SECRET_KEY:-changeme_please_set_in_env_file_32chars}
+      STRAVA_CLIENT_ID: ${STRAVA_CLIENT_ID:-}
+      STRAVA_CLIENT_SECRET: ${STRAVA_CLIENT_SECRET:-}
       FILE_STORE_PATH: /data/files
     volumes:
       - ./file_data:/data/files

frontend/src/pages/ProfilePage.jsx

@@ -175,6 +175,65 @@ export default function ProfilePage() {
       setGcForm({ email: '', password: '', sync_enabled: true, sync_activities: true, sync_wellness: true, sync_lookback_days: '30' })
     },
   })
+  // Strava sync
+  const { data: stravaConfig, refetch: refetchStrava } = useQuery({
+    queryKey: ['strava-config'],
+    queryFn: () => api.get('/strava-sync/config').then(r => r.data),
+    // Poll while a sync is running so the status text stays live.
+    refetchInterval: q => {
+      const s = q.state.data?.last_sync_status || ''
+      const running = s && !/^(OK|Error|Auth error|Connected|Cancelled|Lookback)/.test(s)
+      return running ? 3000 : false
+    },
+  })
+  const [stForm, setStForm] = useState({ sync_enabled: true, sync_lookback_days: '30' })
+  const [stSaved, setStSaved] = useState(false)
+  const stFormLoaded = useRef(false)
+  useEffect(() => {
+    if (stravaConfig?.connected && !stFormLoaded.current) {
+      stFormLoaded.current = true
+      setStForm({
+        sync_enabled: stravaConfig.sync_enabled,
+        sync_lookback_days: String(stravaConfig.sync_lookback_days ?? 30),
+      })
+    } else if (!stravaConfig?.connected) {
+      stFormLoaded.current = false
+    }
+  }, [stravaConfig])
+  // OAuth return banner (?strava=connected|error|scope), then strip the query.
+  const [stReturn, setStReturn] = useState('')
+  useEffect(() => {
+    const p = new URLSearchParams(window.location.search)
+    const s = p.get('strava')
+    if (s) {
+      setStReturn(s)
+      refetchStrava()
+      p.delete('strava')
+      const qs = p.toString()
+      window.history.replaceState({}, '', window.location.pathname + (qs ? `?${qs}` : ''))
+    }
+  }, [])
+  const connectStrava = useMutation({
+    mutationFn: () => api.get('/strava-sync/authorize').then(r => r.data),
+    onSuccess: d => { if (d?.url) window.location.href = d.url },
+  })
+  const saveStrava = useMutation({
+    mutationFn: data => api.put('/strava-sync/config', data).then(r => r.data),
+    onSuccess: () => { refetchStrava(); setStSaved(true); setTimeout(() => setStSaved(false), 3000) },
+  })
+  const triggerStrava = useMutation({
+    mutationFn: () => api.post('/strava-sync/trigger'),
+    onSuccess: () => setTimeout(refetchStrava, 1000),
+  })
+  const deleteStrava = useMutation({
+    mutationFn: () => api.delete('/strava-sync/config'),
+    onSuccess: () => { refetchStrava(); setStForm({ sync_enabled: true, sync_lookback_days: '30' }) },
+  })
+  const stravaSyncing = (() => {
+    const s = stravaConfig?.last_sync_status || ''
+    return !!s && !/^(OK|Error|Auth error|Connected|Cancelled|Lookback)/.test(s)
+  })()
+
   // PocketID config
   const [pidForm, setPidForm] = useState({ issuer: '', client_id: '', client_secret: '', allowed_group: '' })
   const [pidSaved, setPidSaved] = useState(false)
@@ -445,6 +504,94 @@ export default function ProfilePage() {
         })()}
       </Section>
 
+      {/* Strava Sync */}
+      <Section title="🟠 Strava Sync">
+        <p className="text-xs text-gray-500">
+          Connect your Strava account to automatically import activities {formatSyncInterval(stravaConfig?.sync_interval_minutes)}.
+          Works for anything that ends up on Strava — Apple Watch, the Strava phone app, or another GPS watch.
+        </p>
+
+        {stReturn === 'connected' && (
+          <p className="text-xs text-green-400">✓ Strava connected — your first sync has started.</p>
+        )}
+        {stReturn === 'error' && (
+          <p className="text-xs text-red-400">Strava connection failed or was cancelled. Please try again.</p>
+        )}
+        {stReturn === 'scope' && (
+          <p className="text-xs text-yellow-400">Please tick “View data about your activities” when authorizing, so private activities can sync.</p>
+        )}
+
+        {!stravaConfig?.configured && (
+          <p className="text-xs text-yellow-400">
+            Strava API credentials aren’t set on the server. An admin must register an app at
+            strava.com/settings/api and set <code className="text-gray-300">STRAVA_CLIENT_ID</code> /
+            <code className="text-gray-300"> STRAVA_CLIENT_SECRET</code> (callback domain = this site’s domain).
+          </p>
+        )}
+
+        {stravaConfig?.connected ? (
+          <>
+            <div className="flex items-center justify-between bg-orange-900/20 border border-orange-800/40 rounded-lg px-3 py-2 text-xs flex-wrap gap-2">
+              <span className="text-orange-300">✓ Connected{stravaConfig.athlete_name ? ` as ${stravaConfig.athlete_name}` : ''}</span>
+              <div className="flex items-center gap-3 flex-wrap">
+                {stravaConfig.last_sync_at && (
+                  <span className="text-gray-500">
+                    Last sync: {new Date(stravaConfig.last_sync_at).toLocaleString('en-GB', { day: 'numeric', month: 'short', hour: '2-digit', minute: '2-digit' })}
+                  </span>
+                )}
+                {stravaConfig.last_sync_status && (
+                  <span className={stravaConfig.last_sync_status.startsWith('OK') ? 'text-green-400' : /error/i.test(stravaConfig.last_sync_status) ? 'text-red-400' : 'text-yellow-400'}>
+                    {stravaConfig.last_sync_status}
+                  </span>
+                )}
+              </div>
+            </div>
+
+            <label className="flex items-center gap-2 cursor-pointer pt-1">
+              <input type="checkbox" checked={stForm.sync_enabled}
+                onChange={e => setStForm(f => ({ ...f, sync_enabled: e.target.checked }))}
+                className="w-4 h-4 accent-orange-500" />
+              <span className="text-sm text-gray-300">Enable automatic sync ({formatSyncInterval(stravaConfig?.sync_interval_minutes)})</span>
+            </label>
+
+            <Field label="Sync lookback days" hint="How far back to pull on the first sync (-1 = all history). After that, scheduled syncs only refresh the last day or two. Large backfills may hit Strava's rate limits and resume on the next sync.">
+              <Input type="number" value={stForm.sync_lookback_days} min={-1}
+                onChange={e => setStForm(f => ({ ...f, sync_lookback_days: e.target.value }))} />
+            </Field>
+
+            <div className="flex items-center gap-3 flex-wrap pt-1">
+              <SaveButton
+                onClick={() => saveStrava.mutate({
+                  sync_enabled: stForm.sync_enabled,
+                  sync_lookback_days: parseInt(stForm.sync_lookback_days, 10) || 30,
+                })}
+                loading={saveStrava.isPending}
+                saved={stSaved}
+                label="Update"
+              />
+              <button
+                onClick={() => triggerStrava.mutate()}
+                disabled={stravaSyncing || triggerStrava.isPending}
+                className="bg-gray-700 hover:bg-gray-600 disabled:opacity-50 text-white text-sm font-medium px-4 py-2 rounded-lg transition-colors">
+                {stravaSyncing ? 'Syncing…' : '↻ Sync now'}
+              </button>
+              <button
+                onClick={() => { if (confirm('Disconnect Strava?')) deleteStrava.mutate() }}
+                className="text-red-400 hover:text-red-300 text-sm transition-colors">
+                Disconnect
+              </button>
+            </div>
+          </>
+        ) : (
+          <button
+            onClick={() => connectStrava.mutate()}
+            disabled={!stravaConfig?.configured || connectStrava.isPending}
+            className="inline-flex items-center gap-2 bg-[#FC4C02] hover:bg-[#e34402] disabled:opacity-50 text-white text-sm font-medium px-4 py-2 rounded-lg transition-colors">
+            {connectStrava.isPending ? 'Redirecting…' : 'Connect with Strava'}
+          </button>
+        )}
+      </Section>
+
       {/* PocketID — admin only */}
       {user?.is_admin && (
         <Section title="🔑 PocketID Passkey Authentication (Admin)">

frontend/src/pages/UploadPage.jsx

@@ -169,20 +169,21 @@ export default function UploadPage() {
             <li>Click "Request Your Archive"</li>
             <li>Download and upload the ZIP file below</li>
           </ol>
+          <p className="text-gray-500 text-xs mt-2">Handles the gzipped <code>.fit.gz</code>/<code>.gpx.gz</code>/<code>.tcx.gz</code> files Strava puts in the archive. For ongoing sync, connect Strava on the Profile page instead.</p>
         </div>
       </div>
 
       <div className="grid grid-cols-1 lg:grid-cols-2 gap-5">
-        {/* Single FIT/GPX */}
+        {/* Single FIT/GPX/TCX */}
         <UploadZone
           title="Single activity"
-          description="Upload a .fit or .gpx file"
+          description="Upload a .fit, .gpx or .tcx file"
           icon="🏃"
           endpoint="/upload/activity"
           accept={{
             'application/octet-stream': ['.fit'],
             'application/gpx+xml': ['.gpx'],
-            'text/xml': ['.gpx'],
+            'text/xml': ['.gpx', '.tcx'],
           }}
         />