"""Shared utilities — upload validation, embed status tracking, global process registries."""
import json
import asyncio
from pathlib import Path
from fastapi import HTTPException, UploadFile
from config import _EMBED_STATUS_DIR, _MAX_UPLOAD_BYTES

# ── Global process registries (used by routes and shutdown handler) ───────────
running_reports: dict[str, asyncio.subprocess.Process] = {}
running_terraform: dict[str, asyncio.subprocess.Process] = {}


# ── Embed status (file-based) ────────────────────────────────────────────────
def set_embed_status(task_id: str, data: dict):
    (_EMBED_STATUS_DIR / f"{task_id}.json").write_text(json.dumps(data), encoding="utf-8")


def get_embed_status(task_id: str) -> dict | None:
    p = _EMBED_STATUS_DIR / f"{task_id}.json"
    if p.exists():
        try:
            return json.loads(p.read_text(encoding="utf-8"))
        except Exception:
            return None
    return None


def update_embed_status(task_id: str, updates: dict):
    current = get_embed_status(task_id) or {}
    current.update(updates)
    set_embed_status(task_id, current)


# ── File upload validation ────────────────────────────────────────────────────
async def validate_upload(file: UploadFile, allowed_exts: set[str] | None = None, max_bytes: int = _MAX_UPLOAD_BYTES):
    """Validate uploaded file size and extension. Returns file data bytes."""
    data = await file.read()
    if len(data) > max_bytes:
        raise HTTPException(400, f"Arquivo excede o limite de {max_bytes // (1024*1024)}MB")
    if allowed_exts and file.filename:
        ext = Path(file.filename).suffix.lower()
        if ext not in allowed_exts:
            raise HTTPException(400, f"Extensão '{ext}' não permitida. Aceitas: {', '.join(sorted(allowed_exts))}")
    return data


# ── Report file classification ────────────────────────────────────────────────
def classify_report_file(fname: str) -> str:
    """Classify a report file into a category based on its filename."""
    fl = fname.lower()
    if "summary_report" in fl: return "summary"
    if "error_report" in fl or "error" in fl and fl.endswith(".csv"): return "error"
    if fl.startswith("obp_") and "findings" in fl: return "obp_finding"
    if fl.startswith("obp_") and "best_practices" in fl: return "obp_best_practice"
    if fl.startswith("obp_"): return "obp_finding"
    if fl.startswith("raw_data_"): return "raw_data"
    if fl.startswith("cis_"): return "cis_finding"
    if "consolidated_report" in fl: return "consolidated"
    if fl.endswith(".png"): return "diagram"
    return "other"