Aislo/db_management/tools_clone_project.py

"""프로젝트 통째 복제 — **검증용 프로젝트를 창마다 하나씩 뜨는 도구** (2026-09-08).

⚠ **왜 저장소에 두나** — 네 창이 **같은 원본에서 같은 방법으로** 떠야 수치를 견줄 수 있다.
  `tmp/` 는 창마다 따로라 서로 못 봄(2026-09-08 랩탑 두 창이 막힌 자리).

⚠ **원본은 읽기만 한다** — SELECT 뿐이고 원본 행·파일에 쓰지 않는다.
  ⚠ 다만 **복제 중에 원본이 바뀌면 창마다 복제본이 달라진다** — 원본을 얼린 뒤 뜰 것.

무엇을 뜨나
  DB   projects · input_files · processed_point_cloud · surface_models · routes ·
       route_points · route_statistics · longitudinal_sections · cross_sections ·
       project_workflow_stages
       ⚠ 자동증가 id 는 **새로 받고** 참조(외래키)를 새 id 로 다시 이어 붙인다.
  파일  저장소 폴더 통째(robocopy). 경로가 프로젝트 루트 기준 상대라 그대로 쓰인다.
  치환  파일 안에 박힌 **옛 프로젝트 UUID**, 초기값 스냅숏의 **surface_model_id**.

⚠ **DB 만 뜨거나 파일만 떠서는 안 된다** — 수량(B08)은 DB(측점·횡단)를 보고,
  구조물·관 정본은 파일(`structures.json`·`pipe_points.json`)이라 **둘 다** 있어야 화면이 선다.

쓰는 법
    ./venv/Scripts/python.exe db_management/tools_clone_project.py <원본 UUID> "<새 이름>"
"""

import argparse
import asyncio
import json
import subprocess
import sys
import time
from pathlib import Path
from uuid import uuid4

sys.path.append(str(Path(__file__).resolve().parent.parent))

import aiomysql

from common_util.common_util_storage import resolve_stored_project_path
from config.config_db import DB_HOST, DB_NAME, DB_PASSWORD, DB_PORT, DB_USER


def _args() -> argparse.Namespace:
    parser = argparse.ArgumentParser(description="프로젝트를 통째로 복제한다(원본은 읽기만).")
    parser.add_argument("source_id", help="원본 프로젝트 UUID")
    parser.add_argument("new_name", help="새 프로젝트 이름")
    return parser.parse_args()


ARGS = _args()
SOURCE_ID = ARGS.source_id
NEW_NAME = ARGS.new_name


async def _rows(cursor, sql, args):
    await cursor.execute(sql, args)
    return await cursor.fetchall()


async def _insert(cursor, table, row: dict) -> int:
    names = ", ".join(f"`{key}`" for key in row)
    holders = ", ".join(["%s"] * len(row))
    await cursor.execute(f"INSERT INTO `{table}` ({names}) VALUES ({holders})", tuple(row.values()))
    return int(cursor.lastrowid)


async def main() -> None:
    connection = await aiomysql.connect(
        host=DB_HOST,
        port=DB_PORT,
        user=DB_USER,
        password=DB_PASSWORD,
        db=DB_NAME,
        charset="utf8mb4",
    )
    try:
        async with connection.cursor(aiomysql.DictCursor) as cursor:
            source = (await _rows(cursor, "SELECT * FROM projects WHERE id = %s", (SOURCE_ID,)))[0]
            input_files = await _rows(
                cursor, "SELECT * FROM input_files WHERE project_id = %s ORDER BY id", (SOURCE_ID,)
            )
            clouds = await _rows(
                cursor,
                "SELECT * FROM processed_point_cloud WHERE project_id = %s ORDER BY id",
                (SOURCE_ID,),
            )
            models = await _rows(
                cursor,
                "SELECT * FROM surface_models WHERE project_id = %s ORDER BY id",
                (SOURCE_ID,),
            )
            routes = await _rows(
                cursor, "SELECT * FROM routes WHERE project_id = %s ORDER BY id", (SOURCE_ID,)
            )
            stages = await _rows(
                cursor,
                "SELECT * FROM project_workflow_stages WHERE project_id = %s ORDER BY stage_no",
                (SOURCE_ID,),
            )
            route_ids = [route["id"] for route in routes]
            points: list = []
            statistics: list = []
            longitudinal: list = []
            crosses: list = []
            if route_ids:
                holders = ", ".join(["%s"] * len(route_ids))
                points = await _rows(
                    cursor,
                    f"SELECT * FROM route_points WHERE route_id IN ({holders}) ORDER BY id",
                    route_ids,
                )
                statistics = await _rows(
                    cursor,
                    f"SELECT * FROM route_statistics WHERE route_id IN ({holders}) ORDER BY id",
                    route_ids,
                )
                longitudinal = await _rows(
                    cursor,
                    "SELECT * FROM longitudinal_sections"
                    f" WHERE route_id IN ({holders}) ORDER BY id",
                    route_ids,
                )
                crosses = await _rows(
                    cursor,
                    f"SELECT * FROM cross_sections WHERE route_id IN ({holders}) ORDER BY id",
                    route_ids,
                )

        new_id = str(uuid4())
        source_root = Path(resolve_stored_project_path(source["storage_path"]))
        new_storage = "/".join(source["storage_path"].split("/")[:-1] + [new_id])
        new_root = source_root.parent / new_id

        # 1) 파일 먼저 - 실패하면 DB 를 건드리지 않은 채로 끝난다.
        print(f"파일 복사 시작: {source_root.name} -> {new_id}")
        started = time.perf_counter()
        result = subprocess.run(
            [
                "robocopy",
                str(source_root),
                str(new_root),
                "/E",
                "/NFL",
                "/NDL",
                "/NJH",
                "/NJS",
                "/MT:8",
            ],
            capture_output=True,
            text=True,
        )
        if result.returncode >= 8:
            raise RuntimeError(f"robocopy 실패 (코드 {result.returncode})")
        print(f"파일 복사 완료: {time.perf_counter() - started:.0f}초")

        # 2) DB - 한 트랜잭션으로.
        await connection.begin()
        async with connection.cursor(aiomysql.DictCursor) as cursor:
            project = dict(source)
            project["id"] = new_id
            project["name"] = NEW_NAME
            project["storage_path"] = new_storage
            await _insert(cursor, "projects", project)

            file_map: dict[int, int] = {}
            for row in input_files:
                row = dict(row)
                old = row.pop("id")
                row["project_id"] = new_id
                file_map[old] = await _insert(cursor, "input_files", row)

            cloud_map: dict[int, int] = {}
            for row in clouds:
                row = dict(row)
                old = row.pop("id")
                row["project_id"] = new_id
                row["input_file_id"] = file_map.get(row["input_file_id"], row["input_file_id"])
                cloud_map[old] = await _insert(cursor, "processed_point_cloud", row)

            model_map: dict[int, int] = {}
            for row in models:
                row = dict(row)
                old = row.pop("id")
                row["project_id"] = new_id
                row["source_file_id"] = file_map.get(row["source_file_id"], row["source_file_id"])
                row["processed_cloud_id"] = cloud_map.get(
                    row["processed_cloud_id"], row["processed_cloud_id"]
                )
                model_map[old] = await _insert(cursor, "surface_models", row)

            route_map: dict[int, int] = {}
            for row in routes:
                row = dict(row)
                old = row.pop("id")
                row["project_id"] = new_id
                row["surface_model_id"] = model_map.get(
                    row["surface_model_id"], row["surface_model_id"]
                )
                route_map[old] = await _insert(cursor, "routes", row)

            for table, rows in (
                ("route_points", points),
                ("route_statistics", statistics),
                ("longitudinal_sections", longitudinal),
                ("cross_sections", crosses),
            ):
                for row in rows:
                    row = dict(row)
                    row.pop("id")
                    row["route_id"] = route_map.get(row["route_id"], row["route_id"])
                    if "project_id" in row:
                        row["project_id"] = new_id
                    await _insert(cursor, table, row)

            for row in stages:
                row = dict(row)
                row.pop("id")
                row["project_id"] = new_id
                params = row.get("params")
                if params:
                    # 단계 설정에 박힌 옛 id 를 새 id 로 바꾼다(입력 파일·지표면 모델).
                    data = json.loads(params)
                    if data.get("input_file_id") is not None:
                        old_file = int(data["input_file_id"])
                        data["input_file_id"] = str(file_map.get(old_file, old_file))
                    if data.get("surface_model_id") is not None:
                        old_model = int(data["surface_model_id"])
                        data["surface_model_id"] = model_map.get(old_model, old_model)
                    row["params"] = json.dumps(data, ensure_ascii=False)
                await _insert(cursor, "project_workflow_stages", row)
        await connection.commit()
        print(
            f"DB 복제 완료: 입력파일 {len(file_map)} · 지표면 {len(model_map)}"
            f" · 노선 {len(route_map)}"
            f" · 정점 {len(points)} · 횡단 {len(crosses)} · 단계 {len(stages)}"
        )

        # 3) 파일 안에 박힌 옛 프로젝트 id 치환 + 초기값 스냅샷의 지표면 모델 id 교정.
        replaced = []
        for path in new_root.rglob("*.json"):
            try:
                text = path.read_text(encoding="utf-8")
            except (OSError, UnicodeDecodeError):
                continue
            if SOURCE_ID not in text:
                continue
            path.write_text(text.replace(SOURCE_ID, new_id), encoding="utf-8")
            replaced.append(str(path.relative_to(new_root)))
        print("프로젝트 id 치환:", replaced)

        snapshot = new_root / "initial_snapshot" / "db.json"
        if snapshot.is_file():
            dump = json.loads(snapshot.read_text(encoding="utf-8"))
            changed = False
            for route in dump.get("routes", []):
                old_model = route.get("surface_model_id")
                if old_model in model_map:
                    route["surface_model_id"] = model_map[old_model]
                    changed = True
            if changed:
                snapshot.write_text(json.dumps(dump, ensure_ascii=False), encoding="utf-8")
            print("초기값 스냅샷 지표면 id 교정:", changed)

        print(f"\n새 프로젝트: {NEW_NAME}\n  id = {new_id}\n  저장소 = {new_storage}")
    finally:
        connection.close()


asyncio.run(main())