# pxy_de/api.py
from __future__ import annotations
from pathlib import Path
from typing import List, Dict, Any

import pandas as pd
from django.conf import settings
from rest_framework.decorators import api_view
from rest_framework.response import Response

from .providers.base import get_provider


def _rel(path: Path, base_dir: Path) -> str:
    """
    Return a clean relative path like 'data/...'
    """
    try:
        return str(path.relative_to(settings.BASE_DIR))
    except Exception:
        # Fallback: show relative to provider base_dir
        try:
            return str(Path("data") / path.relative_to(base_dir))
        except Exception:
            return str(path)


def _probe_csv(path: Path) -> Dict[str, Any]:
    """
    Lightweight readability probe: existence + sample columns (no full read).
    """
    info: Dict[str, Any] = {"exists": path.exists()}
    if not info["exists"]:
        return info
    try:
        sample = pd.read_csv(path, nrows=5)
        info["columns"] = list(sample.columns)
        info["sample_rows"] = int(sample.shape[0])  # up to 5
    except Exception as e:
        info["error"] = f"{type(e).__name__}: {e}"
    return info


@api_view(["GET"])
def de_health(request):
    """
    GET /api/de/health?city=CDMX&business=cafe&indicator=imss_wages_2023

    Reports:
      - provider in use
      - base_dir used by the provider
      - required/missing files (population.csv always; others if params passed)
      - lightweight probes for each checked file (exists, columns, sample_rows)
    """
    provider = get_provider()
    base_dir: Path = getattr(provider, "base_dir", Path(settings.BASE_DIR) / "data")
    checks: List[Dict[str, Any]] = []
    missing: List[str] = []

    city = (request.query_params.get("city") or "").strip()
    business = (request.query_params.get("business") or "").strip()
    indicator = (request.query_params.get("indicator") or "").strip()

    # Always check SAMI population
    pop_path = base_dir / "sami" / "population.csv"
    pop_probe = _probe_csv(pop_path)
    pop_probe["path"] = _rel(pop_path, base_dir)
    checks.append(pop_probe)
    if not pop_probe["exists"]:
        missing.append(pop_probe["path"])

    # Optional: indicator for SAMI
    if indicator:
        ind_path = base_dir / "sami" / f"{indicator}.csv"
        ind_probe = _probe_csv(ind_path)
        ind_probe["path"] = _rel(ind_path, base_dir)
        checks.append(ind_probe)
        if not ind_probe["exists"]:
            missing.append(ind_probe["path"])

    # Optional: Sites (competition / DENUE)
    if city and business:
        denue_path = base_dir / "denue" / f"{city}_{business}.csv"
        denue_probe = _probe_csv(denue_path)
        denue_probe["path"] = _rel(denue_path, base_dir)
        checks.append(denue_probe)
        if not denue_probe["exists"]:
            missing.append(denue_probe["path"])

    # Optional: Sites (demand / population grid)
    if city:
        grid_path = base_dir / "popgrid" / f"{city}_grid.csv"
        grid_probe = _probe_csv(grid_path)
        grid_probe["path"] = _rel(grid_path, base_dir)
        checks.append(grid_probe)
        if not grid_probe["exists"]:
            missing.append(grid_probe["path"])

    ok = len(missing) == 0
    return Response({
        "provider": "csv-data",
        "base_dir": str(base_dir),
        "ok": ok,
        "missing": missing,
        "files": checks,
    })