#!/usr/bin/env python3
"""Recover only complete, scope-preserving behavior contracts from valid receipts."""
import copy
import hashlib
import json
from pathlib import Path
from wbv2_autopilot import ROOT, contained, read_json, write_json

KINDS = {"positive", "negative", "undo_redo", "save_reopen", "determinism"}


def candidate(worker, contract, root=ROOT):
    try:
        value = read_json(worker)
        proof = read_json(worker.parent / "verified-receipt.json")
        required = set(contract["result"]["sourceUnits"])
        if (proof.get("status") != "verified" or proof.get("job") != contract["job"]
                or proof.get("queueTask") != contract["task"]
                or not required.issubset(value.get("sourceUnits", []))
                or not required.issubset(proof.get("sourceUnits", []))):
            return None
        cases = value.get("cases", [])
        if not KINDS.issubset({case["kind"] for case in cases}):
            return None
        artifacts = {item["path"]: item["sha256"] for item in proof.get("artifacts", [])}
        inputs = proof.get("inputs", {})
        if not artifacts or not inputs:
            return None
        for relative, digest in {**inputs, **artifacts}.items():
            path = contained(root, relative)
            if (not path.is_file() or hashlib.sha256(path.read_bytes()).hexdigest() != digest):
                return None
        for case in cases:
            receipt_path = contained(root, case["receipt"])
            if case["receipt"] not in artifacts:
                return None
            receipt = read_json(receipt_path)
            if (receipt.get("status") != "PASS"
                    or receipt.get("runtime") not in {"CHROMIUM_WEBENGINE", "BLENDER_NATIVE"}
                    or receipt.get("caseId") != case["id"]
                    or receipt.get("kind") != case["kind"]
                    or set(receipt.get("unitIds", [])) != set(case["unitIds"])):
                return None
        result = copy.deepcopy(value)
        result.pop("contractKind", None)
        result.pop("mandatoryCacheInputs", None)
        for case in result["cases"]:
            case["receipt"] = "unallocated/" + case["id"] + ".json"
        return result
    except (OSError, ValueError, KeyError, TypeError):
        return None


def main():
    recovered, unavailable = [], []
    for path in sorted((ROOT / "my-tools/local-workflows").glob("*.json")):
        contract = read_json(path)
        if contract.get("result", {}).get("contractKind") != "compilation_only":
            continue
        job = contract["job"]
        candidates = []
        for worker in (ROOT / "tests/golden/WBV2/autopilot/runs").glob(job + "-*/worker-result.json"):
            result = candidate(worker, contract)
            if result:
                candidates.append((worker.stat().st_mtime_ns, result))
        if not candidates:
            unavailable.append(job)
            continue
        contract["result"] = max(candidates, key=lambda item: item[0])[1]
        contract["acceptance"] = "pending fresh execution; historical receipts are not new results"
        write_json(path, contract)
        recovered.append(job)
    print(json.dumps({"recovered": recovered, "count": len(recovered), "unavailable": unavailable}))


if __name__ == "__main__":
    main()