import json
import os
import logging
from datetime import datetime
from typing import Dict, List, Optional

logger = logging.getLogger("SyncAudit")

AUDIT_DIR = os.path.dirname(os.path.abspath(__file__))
AUDIT_FILE = os.path.join(AUDIT_DIR, "sync_log.json")


def _load_log() -> Dict:
    """Load audit log từ file."""
    if os.path.exists(AUDIT_FILE):
        try:
            with open(AUDIT_FILE, "r", encoding="utf-8") as f:
                return json.load(f)
        except json.JSONDecodeError:
            return {"runs": []}
    return {"runs": []}


def _save_log(data: Dict):
    """Lưu audit log ra file."""
    with open(AUDIT_FILE, "w", encoding="utf-8") as f:
        json.dump(data, f, indent=2, ensure_ascii=False)


def start_run() -> str:
    """
    Bắt đầu một sync run mới.
    Returns: run_id
    """
    run_id = datetime.now().strftime("%Y%m%d_%H%M%S")
    data = _load_log()
    
    run = {
        "run_id": run_id,
        "started_at": datetime.now().isoformat(),
        "finished_at": None,
        "status": "running",
        "summary": {"processed": 0, "skipped": 0, "errors": 0},
        "files": [],
        "errors": []
    }
    
    data["runs"].insert(0, run)  # Mới nhất lên đầu
    
    # Giới hạn 50 runs gần nhất
    if len(data["runs"]) > 50:
        data["runs"] = data["runs"][:50]
    
    _save_log(data)
    logger.info(f"Audit: Started run {run_id}")
    return run_id


def log_file(run_id: str, name: str, item_id: str, policy: str, 
             doc_type: str, chunks: int, status: str, detail: str = ""):
    """
    Ghi log cho 1 file đã xử lý.
    
    Args:
        run_id: ID của sync run
        name: Tên file
        item_id: SharePoint item ID
        policy: Processing policy (requires_ocr, skip_ocr, metadata_only, unsupported)
        doc_type: Document type (textual_document, spreadsheet, drawing, binary)
        chunks: Số chunks đã tạo
        status: indexed, skipped, error
        detail: Ghi chú thêm
    """
    data = _load_log()
    
    for run in data["runs"]:
        if run["run_id"] == run_id:
            file_entry = {
                "name": name,
                "item_id": item_id,
                "doc_type": doc_type,
                "policy": policy,
                "chunks": chunks,
                "status": status,
                "detail": detail,
                "timestamp": datetime.now().isoformat()
            }
            run["files"].append(file_entry)
            
            # Cập nhật summary
            if status == "indexed":
                run["summary"]["processed"] += 1
            elif status == "skipped":
                run["summary"]["skipped"] += 1
            elif status == "error":
                run["summary"]["errors"] += 1
                run["errors"].append(f"{name}: {detail}")
            
            break
    
    _save_log(data)


def finish_run(run_id: str, status: str = "completed"):
    """
    Đánh dấu sync run hoàn thành.
    """
    data = _load_log()
    
    for run in data["runs"]:
        if run["run_id"] == run_id:
            run["finished_at"] = datetime.now().isoformat()
            run["status"] = status
            break
    
    _save_log(data)
    logger.info(f"Audit: Finished run {run_id} ({status})")


def get_history(limit: int = 10) -> List[Dict]:
    """
    Lấy lịch sử sync runs.
    """
    data = _load_log()
    return data["runs"][:limit]


def get_run_detail(run_id: str) -> Optional[Dict]:
    """
    Lấy chi tiết của 1 sync run.
    """
    data = _load_log()
    for run in data["runs"]:
        if run["run_id"] == run_id:
            return run
    return None


def get_latest_run() -> Optional[Dict]:
    """
    Lấy sync run gần nhất.
    """
    data = _load_log()
    return data["runs"][0] if data["runs"] else None