scripts/prove_local_readiness.py

from __future__ import annotations

import argparse
import json
import sys
from pathlib import Path
from typing import Any

ROOT_DIR = Path(__file__).resolve().parent.parent
if str(ROOT_DIR) not in sys.path:
    sys.path.insert(0, str(ROOT_DIR))

from scripts import audit_goal_readiness, check_deployment_readiness, check_research_sources, refresh_research_evidence, research_watchlist

def prove_local_readiness(
    out: Path = ROOT_DIR / "outputs" / "local-readiness.json",
    check_representative_links: bool = False,
    check_key_links: bool = False,
    check_hf_metadata: bool = False,
    hf_metadata_report: Path | None = None,
    verify_audio_pipeline: bool = False,
    voice_id: str = "espeak-ar-clear",
    audio_format: str = "wav",
    max_speech_chars: int = 240,
    refresh_research: bool = False,
    research_refresh_report: Path | None = None,
) -> dict[str, Any]:
    out.parent.mkdir(parents=True, exist_ok=True)

    research_refresh: dict[str, Any] | None = None
    if refresh_research:
        research_refresh_report = research_refresh_report or out.parent / "research-refresh.json"
        research_refresh = refresh_research_evidence.refresh_research_evidence(report_out=research_refresh_report)

    if check_hf_metadata and hf_metadata_report is None:
        hf_metadata_report = out.parent / "hf-model-metadata.md"
    research_checks = check_research_sources.collect_command_checks(
        check_key_links=check_key_links,
        check_representative_links=check_representative_links,
        check_hf_metadata=check_hf_metadata,
        write_hf_metadata_report=hf_metadata_report if check_hf_metadata else None,
    )
    research = check_research_sources.summarize(research_checks)
    license_policy_violations = research_watchlist.license_policy_violations(research_watchlist.CANDIDATES)
    license_policy = {
        "ready": not license_policy_violations,
        "violations": license_policy_violations,
    }

    deployment_checks = check_deployment_readiness.collect_checks()
    deployment = check_deployment_readiness.summarize(deployment_checks)

    audit_checks = audit_goal_readiness.collect_checks()
    audit = audit_goal_readiness.summarize(audit_checks)

    pipeline: dict[str, Any] | None = None
    if verify_audio_pipeline:
        try:
            from scripts.verify_pipeline import create_sample_pdf, verify_pipeline

            sample_pdf = create_sample_pdf(out.parent / "local-readiness-sample.pdf")
            audio_out = out.parent / f"local-readiness-audio.{audio_format}"
            pipeline = {
                "ok": True,
                "result": verify_pipeline(
                    sample_pdf,
                    voice_id,
                    audio_out,
                    audio_format=audio_format,
                    max_speech_chars=max_speech_chars,
                ),
            }
        except Exception as exc:
            pipeline = {"ok": False, "error": str(exc)}

    ready = bool(
        research["ready"]
        and license_policy["ready"]
        and deployment["ready"]
        and audit["ready"]
        and (research_refresh is None or research_refresh["ready"])
        and (pipeline is None or pipeline["ok"])
    )
    result = {
        "ready": ready,
        "complete": bool(audit["complete"]),
        "notes": [
            "This proves local/repo readiness. Final completion still requires deployed Vercel site and worker reports.",
            "Run scripts/prove_live_deployment.py after the Vercel site and worker are live.",
        ],
        "research": research,
        "researchRefresh": research_refresh,
        "licensePolicy": license_policy,
        "deployment": deployment,
        "audit": audit,
        "pipeline": pipeline,
        "hfMetadataReport": str(hf_metadata_report) if hf_metadata_report else None,
    }
    out.write_text(json.dumps(result, ensure_ascii=False, indent=2), encoding="utf-8")
    result["report"] = str(out)
    return result


def print_summary(result: dict[str, Any]) -> None:
    print(f"Local readiness: {'ready' if result['ready'] else 'not ready'}")
    print(f"Goal complete: {'yes' if result['complete'] else 'no, live site and worker reports still required'}")
    print(f"Report: {result['report']}")
    print()
    for section in ["research", "licensePolicy", "deployment", "audit"]:
        summary = result[section]
        counts = summary.get("counts", {})
        print(f"{section}: ready={summary.get('ready')} counts={counts}")
    if result.get("researchRefresh"):
        refresh = result["researchRefresh"]
        source_counts = refresh.get("sourceSummary", {}).get("counts", {})
        live_counts = refresh.get("liveHfMetadata", {}).get("counts", {})
        print(
            "researchRefresh: "
            f"ready={refresh.get('ready')} source={source_counts} "
            f"liveHf={live_counts} report={refresh.get('report')}"
        )
    pipeline = result.get("pipeline")
    if pipeline is not None:
        print(f"pipeline: ok={pipeline.get('ok')}")
        if not pipeline.get("ok"):
            print(f"pipeline error: {pipeline.get('error')}")
    if result.get("hfMetadataReport"):
        print(f"Hugging Face metadata report: {result['hfMetadataReport']}")


def main() -> None:
    parser = argparse.ArgumentParser(description="Write a local readiness proof report for the Arabic audio reader.")
    parser.add_argument("--out", type=Path, default=ROOT_DIR / "outputs" / "local-readiness.json")
    parser.add_argument("--check-representative-links", action="store_true", help="Fetch representative research links.")
    parser.add_argument("--check-key-links", action="store_true", help="Fetch exact key OCR/TTS/hosting source links.")
    parser.add_argument("--check-hf-metadata", action="store_true", help="Fetch Hugging Face model metadata.")
    parser.add_argument(
        "--hf-metadata-report",
        type=Path,
        help="Markdown output path for --check-hf-metadata. Defaults beside the readiness report.",
    )
    parser.add_argument("--verify-audio-pipeline", action="store_true", help="Run a short local PDF-to-audio smoke test.")
    parser.add_argument("--voice-id", default="espeak-ar-clear", help="Voice id for --verify-audio-pipeline.")
    parser.add_argument("--format", choices=["wav", "mp3"], default="wav", help="Audio format for --verify-audio-pipeline.")
    parser.add_argument("--max-speech-chars", type=int, default=240, help="Short smoke-test character limit.")
    parser.add_argument("--refresh-research", action="store_true", help="Refresh research watchlist/recommendation/metadata reports before auditing local readiness.")
    parser.add_argument("--research-refresh-report", type=Path, help="JSON report path for --refresh-research. Defaults beside the readiness report.")
    parser.add_argument("--json", action="store_true", help="Print JSON.")
    args = parser.parse_args()

    result = prove_local_readiness(
        out=args.out,
        check_representative_links=args.check_representative_links,
        check_key_links=args.check_key_links,
        check_hf_metadata=args.check_hf_metadata,
        hf_metadata_report=args.hf_metadata_report,
        verify_audio_pipeline=args.verify_audio_pipeline,
        voice_id=args.voice_id,
        audio_format=args.format,
        max_speech_chars=args.max_speech_chars,
        refresh_research=args.refresh_research,
        research_refresh_report=args.research_refresh_report,
    )
    if args.json:
        print(json.dumps(result, ensure_ascii=False, indent=2))
    else:
        print_summary(result)
    if not result["ready"]:
        raise SystemExit(1)


if __name__ == "__main__":
    main()