#!/usr/bin/env python3
"""Validate that the public Smoke24 project boundary stays clean."""

from __future__ import annotations

import csv
import json
from pathlib import Path


ROOT = Path(__file__).resolve().parents[1]
PUBLIC_ROOTS = [
    ROOT / "data/public",
    ROOT / "reports/public",
    ROOT / "dashboard",
    ROOT / "scripts",
    ROOT / "docs",
]
BANNED_TEXT = (
    "BOBSHELL_" + "API_KEY",
    "HUGGING_FACE_HUB_" + "TOKEN",
    "HF_" + "TOKEN",
    "OPENAI_" + "API_KEY",
    "ANTHROPIC_" + "API_KEY",
    "internal/" + "ibm-bob",
    "ornith-1.0-35b-" + "nvfp4",
)
BANNED_COLUMNS = {"api_" + "base", "client_model_" + "arg"}


def read_csv(path: Path) -> list[dict[str, str]]:
    with path.open(newline="", encoding="utf-8") as handle:
        return list(csv.DictReader(handle))


def validate_text() -> list[str]:
    errors: list[str] = []
    for root in PUBLIC_ROOTS:
        for path in root.rglob("*"):
            if "__pycache__" in path.parts:
                continue
            if not path.is_file() or path.suffix.lower() in {".png", ".jpg", ".jpeg", ".webp", ".pyc"}:
                continue
            text = path.read_text(encoding="utf-8", errors="ignore")
            lowered = text.lower()
            for needle in BANNED_TEXT:
                if needle.lower() in lowered:
                    errors.append(f"{path.relative_to(ROOT)} contains banned marker {needle}")
    return errors


def validate_csvs() -> list[str]:
    errors: list[str] = []
    for path in (ROOT / "data/public").rglob("*.csv"):
        rows = read_csv(path)
        if not rows:
            errors.append(f"{path.relative_to(ROOT)} is empty")
            continue
        fields = set(rows[0])
        leaked = sorted(fields & BANNED_COLUMNS)
        if leaked:
            errors.append(f"{path.relative_to(ROOT)} contains private endpoint columns: {', '.join(leaked)}")
        for row in rows:
            model_id = (row.get("model_id") or row.get("canonical_model_id") or "").lower()
            if model_id.startswith("internal/") or "nvfp4" in model_id:
                errors.append(f"{path.relative_to(ROOT)} contains non-public model id {model_id}")
    return errors


def validate_dashboard_data() -> list[str]:
    errors: list[str] = []
    data_path = ROOT / "site/assets/data/report-data.json"
    if not data_path.exists():
        return errors
    data = json.loads(data_path.read_text(encoding="utf-8"))
    for row in data.get("models", []):
        model_id = str(row.get("model_id") or "").lower()
        if model_id.startswith("internal/") or "nvfp4" in model_id:
            errors.append(f"site dashboard contains non-public model id {model_id}")
    return errors


def main() -> None:
    errors = validate_text() + validate_csvs() + validate_dashboard_data()
    if errors:
        raise SystemExit("Public validation failed:\n" + "\n".join(errors))
    print("public data boundary ok")


if __name__ == "__main__":
    main()