init

app/services/geo_pipeline.py

@@ -0,0 +1,211 @@
+"""Geo pipeline — processes raw GPS tracks into zone polygons.
+
+All geo math is done via Shapely (pure Python), so it works
+with both SQLite (dev) and PostgreSQL (prod).
+"""
+
+import json
+import math
+import uuid
+from datetime import datetime, timezone
+
+from rdp import rdp
+from shapely.geometry import Polygon, mapping
+from shapely import wkt
+from sqlmodel import Session
+
+from app.config import (
+    MIN_ZONE_AREA_M2,
+    LOOP_CLOSURE_RADIUS_M,
+    DOUGLAS_PEUCKER_EPSILON_M,
+)
+from app.models.activity import Activity
+from app.models.zone import Zone
+from app.schemas.activity import GpsPoint, ActivityDetail
+
+
+def haversine_m(lat1: float, lon1: float, lat2: float, lon2: float) -> float:
+    """Haversine distance in meters between two lat/lon points."""
+    R = 6_371_000  # Earth radius in meters
+    phi1, phi2 = math.radians(lat1), math.radians(lat2)
+    dphi = math.radians(lat2 - lat1)
+    dlam = math.radians(lon2 - lon1)
+    a = (
+        math.sin(dphi / 2) ** 2
+        + math.cos(phi1) * math.cos(phi2) * math.sin(dlam / 2) ** 2
+    )
+    return 2 * R * math.atan2(math.sqrt(a), math.sqrt(1 - a))
+
+
+def compute_distance_m(points: list[GpsPoint]) -> float:
+    """Total distance in meters along a GPS track."""
+    total = 0.0
+    for i in range(1, len(points)):
+        total += haversine_m(
+            points[i - 1].lat, points[i - 1].lon, points[i].lat, points[i].lon
+        )
+    return total
+
+
+def simplify_track(points: list[GpsPoint], epsilon: float) -> list[GpsPoint]:
+    """Douglas-Peucker simplification on lat/lon coordinates.
+
+    epsilon is in approximate degrees; we convert from meters.
+    ~0.00001 degree ≈ 1.11 m at the equator.
+    """
+    coords = [(p.lat, p.lon) for p in points]
+    eps_deg = epsilon / 111_000  # rough meters->degrees
+    simplified = rdp(coords, epsilon=eps_deg)
+    # Rebuild GpsPoint list (drop timestamps for simplified points)
+    return [GpsPoint(lat=c[0], lon=c[1]) for c in simplified]
+
+
+def is_loop_closed(points: list[GpsPoint], radius_m: float) -> bool:
+    """Check if start and end points are within radius_m meters."""
+    if len(points) < 4:
+        return False
+    return (
+        haversine_m(points[0].lat, points[0].lon, points[-1].lat, points[-1].lon)
+        <= radius_m
+    )
+
+
+def build_polygon(points: list[GpsPoint]) -> Polygon | None:
+    """Build a Shapely Polygon from GPS points.
+
+    Uses (lon, lat) ordering for GeoJSON/WKT compatibility.
+    Closes the ring if needed.
+    """
+    coords = [(p.lon, p.lat) for p in points]
+    if coords[0] != coords[-1]:
+        coords.append(coords[0])
+    if len(coords) < 4:
+        return None
+    poly = Polygon(coords)
+    if not poly.is_valid:
+        poly = poly.buffer(0)  # attempt to fix self-intersections
+    return poly if poly.is_valid and not poly.is_empty else None
+
+
+def polygon_area_m2(poly: Polygon) -> float:
+    """Approximate area in m² from a WGS84 polygon.
+
+    Uses a simple cos(lat) correction. Accurate enough for city-scale zones.
+    """
+    centroid = poly.centroid
+    lat_rad = math.radians(centroid.y)
+    # Degrees to meters conversion factors
+    m_per_deg_lat = 111_320.0
+    m_per_deg_lon = 111_320.0 * math.cos(lat_rad)
+    # Scale polygon to meters and compute area
+    coords = list(poly.exterior.coords)
+    scaled = [(x * m_per_deg_lon, y * m_per_deg_lat) for x, y in coords]
+    return abs(Polygon(scaled).area)
+
+
+def wkt_to_geojson(wkt_str: str) -> dict:
+    """Convert WKT polygon string to GeoJSON dict."""
+    geom = wkt.loads(wkt_str)
+    return mapping(geom)
+
+
+def process_activity(
+    user_id: uuid.UUID,
+    activity_type: str,
+    started_at: datetime,
+    ended_at: datetime,
+    gps_track: list[GpsPoint],
+    session: Session,
+) -> ActivityDetail:
+    """Full geo pipeline: filter -> simplify -> validate loop -> build polygon -> persist.
+
+    Returns an ActivityDetail with zone info if a zone was created.
+    """
+    # 1. Create activity record
+    distance = compute_distance_m(gps_track)
+    activity = Activity(
+        user_id=user_id,
+        type=activity_type,
+        started_at=started_at,
+        ended_at=ended_at,
+        distance_m=distance,
+        raw_gpx=json.dumps([{"lat": p.lat, "lon": p.lon} for p in gps_track]),
+        status="pending",
+    )
+    session.add(activity)
+    session.flush()  # get activity.id
+
+    zone_id = None
+    area_m2 = None
+
+    try:
+        # 2. Filter outliers (skip points with hdop > 4 if present)
+        filtered = [p for p in gps_track if p.hdop is None or p.hdop <= 4]
+        if len(filtered) < 10:
+            activity.status = "failed"
+            session.commit()
+            return _to_detail(activity, None, None)
+
+        # 3. Simplify
+        simplified = simplify_track(filtered, DOUGLAS_PEUCKER_EPSILON_M)
+
+        # 4. Check loop closure
+        if not is_loop_closed(simplified, LOOP_CLOSURE_RADIUS_M):
+            activity.status = "failed"
+            session.commit()
+            return _to_detail(activity, None, None)
+
+        # 5. Build polygon
+        poly = build_polygon(simplified)
+        if poly is None:
+            activity.status = "failed"
+            session.commit()
+            return _to_detail(activity, None, None)
+
+        # 6. Validate area
+        area_m2 = polygon_area_m2(poly)
+        if area_m2 < MIN_ZONE_AREA_M2:
+            activity.status = "failed"
+            session.commit()
+            return _to_detail(activity, None, None)
+
+        # 7. Persist zone
+        zone = Zone(
+            owner_id=user_id,
+            activity_id=activity.id,
+            polygon_wkt=poly.wkt,
+            area_m2=area_m2,
+        )
+        session.add(zone)
+        activity.status = "completed"
+        session.commit()
+        session.refresh(zone)
+        session.refresh(activity)
+
+        zone_id = zone.id
+        area_m2 = zone.area_m2
+
+    except Exception:
+        activity.status = "failed"
+        session.commit()
+
+    return _to_detail(activity, zone_id, area_m2)
+
+
+def _to_detail(
+    activity: Activity,
+    zone_id: uuid.UUID | None,
+    area_m2: float | None,
+) -> ActivityDetail:
+    return ActivityDetail(
+        id=activity.id,
+        user_id=activity.user_id,
+        type=activity.type,
+        started_at=activity.started_at,
+        ended_at=activity.ended_at,
+        distance_m=activity.distance_m,
+        status=activity.status,
+        created_at=activity.created_at,
+        zone_id=zone_id,
+        area_m2=area_m2,
+    )