#!/usr/bin/env python3
"""Validate the static judged bundle without running any paper code."""

from __future__ import annotations

import json
import re
from pathlib import Path


ROOT = Path(__file__).resolve().parents[1]
PAGES = ROOT / "pages"
OUTPUT = ROOT / "outputs" / "bundle_validation.json"


def main() -> None:
    claim_pages = sorted(PAGES.glob("claim-*/page.md"))
    judged_pages = sorted(PAGES.glob("**/*.md"))
    page_records = []
    verdicts = []
    for path in judged_pages:
        text = path.read_text(encoding="utf-8")
        page_records.append({
            "path": str(path.relative_to(ROOT)),
            "bytes": len(path.read_bytes()),
            "characters": len(text),
            "nonempty": bool(text.strip()),
        })
        matches = re.findall(r"^## Verdict: \*\*(VERIFIED|FALSIFIED)\*\*$", text, re.MULTILINE)
        if path in claim_pages:
            verdicts.append(matches[0] if len(matches) == 1 else "MISSING_OR_MULTIPLE")

    audit = json.loads((ROOT / "outputs" / "audit_results.json").read_text(encoding="utf-8"))
    audit_verdicts = [item.get("verdict") for item in audit.get("claims", [])]
    all_page_bytes = sum(item["bytes"] for item in page_records)
    all_page_characters = sum(item["characters"] for item in page_records)
    tags_text = (ROOT / "README.md").read_text(encoding="utf-8")
    result = {
        "root": str(ROOT),
        "claim_page_count": len(claim_pages),
        "claim_page_paths": [str(p.relative_to(ROOT)) for p in claim_pages],
        "claim_page_verdicts": verdicts,
        "audit_claim_count": len(audit.get("claims", [])),
        "audit_verdicts": audit_verdicts,
        "judged_page_file_count": len(judged_pages),
        "judged_page_bytes": all_page_bytes,
        "judged_page_characters": all_page_characters,
        "minimum_bytes_met": all_page_bytes >= 5000,
        "character_cap_met": all_page_characters <= 120000,
        "all_judged_pages_nonempty": all(item["nonempty"] for item in page_records),
        "static_tags_present": "icml2026-repro" in tags_text and "paper-JnuwpwbZ8D" in tags_text,
        "all_claim_verdicts_conclusive": (
            len(claim_pages) == 6
            and verdicts == ["FALSIFIED"] * 6
        ),
        "page_records": page_records,
    }
    result["pass"] = all((
        result["claim_page_count"] == 6,
        result["audit_claim_count"] == 6,
        result["all_claim_verdicts_conclusive"],
        result["all_judged_pages_nonempty"],
        result["minimum_bytes_met"],
        result["character_cap_met"],
        result["static_tags_present"],
    ))
    OUTPUT.write_text(json.dumps(result, indent=2, sort_keys=True) + "\n", encoding="utf-8")
    print(json.dumps({
        "output": str(OUTPUT),
        "pass": result["pass"],
        "judged_page_file_count": result["judged_page_file_count"],
        "judged_page_bytes": result["judged_page_bytes"],
        "judged_page_characters": result["judged_page_characters"],
    }, sort_keys=True))
    if not result["pass"]:
        raise SystemExit(1)


if __name__ == "__main__":
    main()