Aislo/B04_wf1_Surface/B04_wf1_Surface_Repository.py

"""B04 지표면 분석 결과의 aiomysql Raw SQL 접근.

processed_point_cloud(변환 포인트클라우드), surface_models(지표면 모델),
terrain_layers(지형 레이어) 테이블에 메타데이터와 상대 경로를 기록한다.
공간 데이터는 MariaDB JSON 컬럼에 GeoJSON 문자열로 저장한다.
"""

import json
from pathlib import PurePosixPath
from typing import Any
from uuid import UUID

import aiomysql

_STAGE_ROOT = "B04_wf1_Surface"


def _validate_stage_path(relative_path: str) -> str:
    """B04_wf1_Surface 아래의 안전한 상대 경로인지 검증하고 posix 문자열로 반환한다."""
    normalized = PurePosixPath(relative_path.replace("\\", "/"))
    if normalized.is_absolute() or ".." in normalized.parts:
        raise ValueError("DB에는 프로젝트 루트 기준 상대 경로만 저장할 수 있습니다.")
    if not normalized.parts or normalized.parts[0] != _STAGE_ROOT:
        raise ValueError(f"B04 산출물 경로는 {_STAGE_ROOT} 아래여야 합니다.")
    return normalized.as_posix()


async def create_processed_point_cloud(
    connection: aiomysql.Connection,
    *,
    input_file_id: int,
    project_id: UUID,
    process_type: str,
    processed_file_path: str | None,
    converted_format: str | None,
    converted_file_path: str | None,
    point_count: int | None,
    bounds: dict[str, Any] | None,
    statistics: dict[str, Any] | None,
    classification_summary: dict[str, Any] | None,
    processing_params: dict[str, Any] | None,
    status: str = "COMPLETE",
) -> int:
    """변환 포인트클라우드 메타데이터를 저장하고 생성된 ID를 반환한다."""
    processed_rel = _validate_stage_path(processed_file_path) if processed_file_path else None
    converted_rel = _validate_stage_path(converted_file_path) if converted_file_path else None
    stats = statistics or {}

    async with connection.cursor() as cursor:
        await cursor.execute(
            """
            INSERT INTO processed_point_cloud (
                input_file_id, project_id, process_type,
                processed_file_path, converted_format, converted_file_path,
                point_count, min_z, max_z, mean_z,
                x_min, x_max, y_min, y_max, density_per_sqm,
                classification_summary, processing_params, status
            )
            VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
            """,
            (
                input_file_id,
                str(project_id),
                process_type,
                processed_rel,
                converted_format,
                converted_rel,
                point_count,
                stats.get("min_z"),
                stats.get("max_z"),
                stats.get("mean_z"),
                (bounds or {}).get("x_min"),
                (bounds or {}).get("x_max"),
                (bounds or {}).get("y_min"),
                (bounds or {}).get("y_max"),
                stats.get("density_per_sqm"),
                json.dumps(classification_summary, ensure_ascii=False)
                if classification_summary is not None
                else None,
                json.dumps(processing_params, ensure_ascii=False)
                if processing_params is not None
                else None,
                status,
            ),
        )
        new_id = cursor.lastrowid
    if not new_id:
        raise RuntimeError("processed_point_cloud 레코드 생성 결과에 ID가 없습니다.")
    return int(new_id)


async def create_surface_model(
    connection: aiomysql.Connection,
    *,
    project_id: UUID,
    model_type: str,
    source_file_id: int | None,
    processed_cloud_id: int | None,
    crs_epsg: int | None,
    resolution_m: float | None,
    model_file_path: str | None,
    generation_params: dict[str, Any] | None,
    status: str = "COMPLETE",
) -> int:
    """지표면 모델 메타데이터를 저장하고 생성된 ID를 반환한다."""
    model_rel = _validate_stage_path(model_file_path) if model_file_path else None

    async with connection.cursor() as cursor:
        await cursor.execute(
            """
            INSERT INTO surface_models (
                project_id, model_type, source_file_id, processed_cloud_id,
                status, crs_epsg, resolution_m, model_file_path,
                generation_params, completed_at
            )
            VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, CURRENT_TIMESTAMP)
            """,
            (
                str(project_id),
                model_type,
                source_file_id,
                processed_cloud_id,
                status,
                crs_epsg,
                resolution_m,
                model_rel,
                json.dumps(generation_params, ensure_ascii=False)
                if generation_params is not None
                else None,
            ),
        )
        new_id = cursor.lastrowid
    if not new_id:
        raise RuntimeError("surface_models 레코드 생성 결과에 ID가 없습니다.")
    return int(new_id)


async def create_terrain_layer(
    connection: aiomysql.Connection,
    *,
    surface_model_id: int,
    layer_name: str,
    geometry_type: str,
    layer_file_path: str | None,
    file_format: str | None,
    file_size_mb: float | None,
    statistics: dict[str, Any] | None,
) -> int:
    """지형 레이어 메타데이터를 저장하고 생성된 ID를 반환한다."""
    layer_rel = _validate_stage_path(layer_file_path) if layer_file_path else None

    async with connection.cursor() as cursor:
        await cursor.execute(
            """
            INSERT INTO terrain_layers (
                surface_model_id, layer_name, geometry_type,
                layer_file_path, file_format, file_size_mb, statistics
            )
            VALUES (%s, %s, %s, %s, %s, %s, %s)
            """,
            (
                surface_model_id,
                layer_name,
                geometry_type,
                layer_rel,
                file_format,
                file_size_mb,
                json.dumps(statistics, ensure_ascii=False) if statistics is not None else None,
            ),
        )
        new_id = cursor.lastrowid
    if not new_id:
        raise RuntimeError("terrain_layers 레코드 생성 결과에 ID가 없습니다.")
    return int(new_id)


async def get_input_file(
    connection: aiomysql.Connection, project_id: UUID, input_file_id: int
) -> dict[str, Any]:
    """프로젝트의 특정 입력 파일 경로·좌표계를 조회한다."""
    async with connection.cursor() as cursor:
        await cursor.execute(
            """
            SELECT id, file_type, raw_file_path, crs_epsg
            FROM input_files
            WHERE id = %s AND project_id = %s
            """,
            (input_file_id, str(project_id)),
        )
        row = await cursor.fetchone()
    if not row:
        raise LookupError("입력 파일을 찾을 수 없습니다.")
    return {
        "id": int(row[0]),
        "file_type": row[1],
        "raw_file_path": row[2],
        "crs_epsg": row[3],
    }


async def list_project_point_cloud_inputs(
    connection: aiomysql.Connection, project_id: UUID
) -> list[dict[str, Any]]:
    """프로젝트의 LAS/LAZ 원본 입력 파일 목록을 최신순으로 조회한다."""
    async with connection.cursor() as cursor:
        await cursor.execute(
            """
            SELECT id, file_type, original_filename, raw_file_path, file_size_mb,
                   crs_epsg, status, upload_at
            FROM input_files
            WHERE project_id = %s AND file_type IN ('las', 'laz')
            ORDER BY upload_at DESC, id DESC
            """,
            (str(project_id),),
        )
        rows = await cursor.fetchall()

    return [
        {
            "id": int(row[0]),
            "file_type": row[1],
            "original_filename": row[2],
            "raw_file_path": row[3],
            "file_size_mb": float(row[4]) if row[4] is not None else None,
            "crs_epsg": row[5],
            "status": row[6],
            "created_at": row[7].isoformat() if row[7] else None,
        }
        for row in rows
    ]


async def list_surface_models(
    connection: aiomysql.Connection, project_id: UUID
) -> list[dict[str, Any]]:
    """프로젝트의 지표면 모델 목록을 최신순으로 조회한다."""
    async with connection.cursor() as cursor:
        await cursor.execute(
            """
            SELECT id, model_type, status, resolution_m, model_file_path, created_at
            FROM surface_models
            WHERE project_id = %s
            ORDER BY created_at DESC
            """,
            (str(project_id),),
        )
        rows = await cursor.fetchall()

    return [
        {
            "id": int(row[0]),
            "model_type": row[1],
            "status": row[2],
            "resolution_m": row[3],
            "model_file_path": row[4],
            "created_at": row[5].isoformat() if row[5] else None,
        }
        for row in rows
    ]