geointel/backend/app/services/vector_feature_service.py

from __future__ import annotations

from typing import Any
from uuid import UUID

from geoalchemy2.functions import ST_Intersects, ST_MakeEnvelope
from geoalchemy2.shape import from_shape
from geoalchemy2.shape import to_shape
from shapely.geometry import mapping
from shapely.geometry import shape
from shapely.validation import make_valid

from app.core.errors import AppError
from app.models import VectorFeature


class VectorFeatureService:
    @staticmethod
    def _normalize_selection_bbox(bbox: dict[str, Any]) -> dict[str, float | str]:
        try:
            min_x = float(bbox["min_x"])
            min_y = float(bbox["min_y"])
            max_x = float(bbox["max_x"])
            max_y = float(bbox["max_y"])
        except (KeyError, TypeError, ValueError) as exc:
            raise AppError(
                code="INVALID_SELECTION_BBOX",
                message="Selection bbox must include numeric min_x, min_y, max_x and max_y values",
                status_code=400,
            ) from exc

        crs = str(bbox.get("crs") or "EPSG:4326").upper()
        if crs != "EPSG:4326":
            raise AppError(
                code="UNSUPPORTED_SELECTION_CRS",
                message="Map selection currently supports EPSG:4326 bbox coordinates only",
                details={"crs": crs},
                status_code=400,
            )
        if min_x >= max_x or min_y >= max_y:
            raise AppError(
                code="INVALID_SELECTION_BBOX",
                message="Selection bbox must have min_x < max_x and min_y < max_y",
                status_code=400,
            )
        if min_x < -180 or max_x > 180 or min_y < -90 or max_y > 90:
            raise AppError(
                code="INVALID_SELECTION_BBOX",
                message="Selection bbox is outside EPSG:4326 longitude/latitude bounds",
                status_code=400,
            )

        return {"min_x": min_x, "min_y": min_y, "max_x": max_x, "max_y": max_y, "crs": "EPSG:4326"}

    @staticmethod
    def _row_to_geojson_feature(row: VectorFeature) -> dict[str, Any]:
        geometry_value = row.geometry
        try:
            geometry = geometry_value if hasattr(geometry_value, "__geo_interface__") else to_shape(geometry_value)
        except Exception as exc:
            raise AppError(
                code="INVALID_VECTOR_FEATURE_GEOMETRY",
                message="Persisted vector feature geometry could not be converted to GeoJSON",
                details={"vector_feature_id": str(row.id)},
                status_code=500,
            ) from exc

        properties = dict(row.properties_json or {})
        properties.update(
            {
                "vector_feature_id": str(row.id),
                "dataset_id": str(row.dataset_id),
                "source_feature_id": row.source_feature_id,
                "feature_class": row.feature_class,
            }
        )

        return {
            "type": "Feature",
            "id": str(row.id),
            "geometry": mapping(geometry),
            "properties": properties,
        }

    @staticmethod
    def select_features_by_bbox(
        db,
        dataset_id: UUID,
        bbox: dict[str, Any],
        limit: int = 100,
    ) -> dict[str, Any]:
        normalized_bbox = VectorFeatureService._normalize_selection_bbox(bbox)
        safe_limit = max(1, min(int(limit), 1000))

        rows = (
            db.query(VectorFeature)
            .filter(VectorFeature.dataset_id == dataset_id)
            .filter(
                ST_Intersects(
                    VectorFeature.geometry,
                    ST_MakeEnvelope(
                        normalized_bbox["min_x"],
                        normalized_bbox["min_y"],
                        normalized_bbox["max_x"],
                        normalized_bbox["max_y"],
                        4326,
                    ),
                )
            )
            .order_by(VectorFeature.created_at.asc())
            .limit(safe_limit + 1)
            .all()
        )
        truncated = len(rows) > safe_limit
        selected_rows = rows[:safe_limit]
        features = [VectorFeatureService._row_to_geojson_feature(row) for row in selected_rows]

        return {
            "selection_bbox": normalized_bbox,
            "feature_count": len(features),
            "limit": safe_limit,
            "truncated": truncated,
            "geojson": {
                "type": "FeatureCollection",
                "features": features,
            },
        }

    @staticmethod
    def persist_geojson_features(
        db,
        dataset_id: UUID,
        payload: dict[str, Any],
        feature_class: str | None = None,
        *,
        commit: bool = True,
    ) -> list[VectorFeature]:
        features = payload.get("features")
        if payload.get("type") != "FeatureCollection" or not isinstance(features, list):
            raise AppError(code="INVALID_GEOJSON", message="GeoJSON payload must be a FeatureCollection", status_code=400)

        persisted: list[VectorFeature] = []
        for index, feature in enumerate(features):
            if not isinstance(feature, dict):
                raise AppError(code="INVALID_GEOJSON", message=f"Feature {index} must be an object", status_code=400)
            geometry_payload = feature.get("geometry")
            if geometry_payload is None:
                continue
            try:
                geometry = shape(geometry_payload)
            except Exception as exc:
                raise AppError(code="INVALID_GEOJSON", message=f"Invalid feature geometry at index {index}", status_code=400) from exc
            if geometry.is_empty:
                continue
            if not geometry.is_valid:
                geometry = make_valid(geometry)
            if geometry.is_empty or not geometry.is_valid:
                raise AppError(code="INVALID_GEOMETRY", message=f"Invalid feature geometry at index {index}", status_code=400)

            properties = feature.get("properties") if isinstance(feature.get("properties"), dict) else {}
            source_feature_id = feature.get("id")
            if source_feature_id is None:
                source_feature_id = properties.get("id") or properties.get("source_feature_id")

            row = VectorFeature(
                dataset_id=dataset_id,
                feature_class=feature_class,
                source_feature_id=str(source_feature_id) if source_feature_id is not None else None,
                properties_json=properties,
                geometry=from_shape(geometry, srid=4326),
            )
            db.add(row)
            persisted.append(row)

        if commit:
            db.commit()
            for row in persisted:
                db.refresh(row)
        return persisted