#!/usr/bin/env python3
"""Write results/manifest.json: every result artefact with size, SHA-256, producing script and inputs."""
import hashlib
import json
import subprocess
import time
from pathlib import Path

ROOT = Path(__file__).resolve().parents[1]
R = ROOT / "results"
PRODUCERS = {
    "tables/frame_exclusions.csv": "src/ingestion/build_frame.py", "tables/robots_exclusion_balance.csv": "src/ingestion/build_frame.py",
    "raw/frame_summary.json": "src/ingestion/build_frame.py", "raw/sample_summary.json": "src/ingestion/sample.py",
    "raw/exp_agreement.json": "scripts/06_exp_agreement.py", "raw/exp_cost.json": "scripts/07_exp_cost.py",
    "raw/aggregate_summary.json": "scripts/05_aggregate.py", "raw/api_usage.jsonl": "src/extraction/client.py",
    "raw/image_hosts.json": "ad hoc host census (research_log 2026-10-20)", "raw/robots_check.txt": "curl robots.txt census (research_log)",
}


def sha(p: Path) -> str:
    h = hashlib.sha256()
    with open(p, "rb") as fh:
        for chunk in iter(lambda: fh.read(1 << 20), b""):
            h.update(chunk)
    return h.hexdigest()


def main():
    items = []
    for p in sorted(R.rglob("*")):
        if p.is_dir() or p.name == "manifest.json" or ".DS_Store" in p.name:
            continue
        rel = str(p.relative_to(R))
        prod = PRODUCERS.get(rel)
        if prod is None:
            if rel.startswith("tables/exp1") and rel.startswith(("tables/exp1_", "tables/exp2", "tables/exp3", "tables/exp4", "tables/exp9", "tables/exp12", "tables/exp16")):
                prod = "scripts/06_exp_agreement.py"
            elif rel.startswith("tables/exp"):
                prod = "scripts/07_exp_cost.py" if any(rel.startswith(f"tables/exp{k}") for k in ("5", "6", "7", "10", "13", "14", "15")) else "scripts/06_exp_agreement.py"
            elif rel.startswith("tables/tex/"):
                prod = "src/visualization/tables.py"
            elif rel.startswith("figures/"):
                prod = "src/visualization/figures.py"
            elif rel.startswith("raw/batches/"):
                prod = "scripts/04_run_batches.py (Anthropic Message Batches API)"
        items.append({"path": rel, "bytes": p.stat().st_size, "sha256": sha(p), "producer": prod, "modified": time.strftime("%Y-%m-%dT%H:%M:%S", time.localtime(p.stat().st_mtime))})
    try:
        commit = subprocess.run(["git", "rev-parse", "HEAD"], cwd=ROOT, capture_output=True, text=True).stdout.strip()
    except Exception:
        commit = None
    manifest = {"generated": time.strftime("%Y-%m-%dT%H:%M:%S"), "git_commit": commit or "uncommitted",
                "schema_version": json.load(open(R / "raw" / "batches" / "run_meta.json"))["schema_version"],
                "prompt_version": json.load(open(R / "raw" / "batches" / "run_meta.json"))["prompt_version"],
                "n_items": len(items), "items": items}
    json.dump(manifest, open(R / "manifest.json", "w"), indent=1)
    print(len(items), "artefacts listed")


if __name__ == "__main__":
    main()