llm/scripts/check_1c_patch_workspace_integrity.py

#!/usr/bin/env python3
"""Check integrity of a 1C patch workspace before diff/package/apply steps."""

from __future__ import annotations

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


def load_json(path: Path) -> dict[str, Any]:
    return json.loads(path.read_text(encoding="utf-8-sig"))


def sha256_file(path: Path) -> str:
    digest = hashlib.sha256()
    with path.open("rb") as stream:
        for chunk in iter(lambda: stream.read(1024 * 1024), b""):
            digest.update(chunk)
    return digest.hexdigest()


def issue(severity: str, code: str, message: str, *, path: str | None = None, record: dict[str, Any] | None = None) -> dict[str, Any]:
    result = {"severity": severity, "code": code, "message": message}
    if path:
        result["path"] = path
    if record:
        result["record"] = record
    return result


def is_relative_safe(relative_path: str) -> bool:
    path = Path(relative_path)
    return not path.is_absolute() and ".." not in path.parts


def expected_paths(workspace: Path, relative_path: str) -> tuple[Path, Path]:
    rel = Path(relative_path)
    return workspace / "original" / rel, workspace / "working" / rel


def all_files(root: Path) -> list[Path]:
    if not root.exists():
        return []
    return sorted(path for path in root.rglob("*") if path.is_file())


def rel_set(root: Path) -> set[str]:
    result = set()
    for path in all_files(root):
        result.add(str(path.relative_to(root)).replace("\\", "/"))
    return result


def check_workspace(workspace: Path) -> dict[str, Any]:
    manifest_path = workspace / "manifest.json"
    findings = []
    if not manifest_path.exists():
        findings.append(issue("error", "missing_manifest", "Workspace manifest.json is missing.", path=str(manifest_path)))
        return result(workspace, findings, [])

    manifest = load_json(manifest_path)
    records = manifest.get("files") or []
    manifest_relatives = set()
    file_checks = []
    for record in records:
        relative = str(record.get("relative_path") or "")
        manifest_relatives.add(relative)
        if not relative or not is_relative_safe(relative):
            findings.append(issue("error", "unsafe_relative_path", f"Unsafe relative path in manifest: {relative}", record=record))
            continue
        original, working = expected_paths(workspace, relative)
        check = {
            "relative_path": relative,
            "original_path": str(original),
            "working_path": str(working),
            "expected_sha256": record.get("sha256"),
            "original_exists": original.exists(),
            "working_exists": working.exists(),
        }
        if not original.exists():
            findings.append(issue("error", "missing_original_file", "Original file is missing.", path=str(original), record=record))
        else:
            actual = sha256_file(original)
            check["original_sha256"] = actual
            if record.get("sha256") and actual != record.get("sha256"):
                findings.append(issue("error", "original_hash_mismatch", "Original file hash differs from manifest; original/ must stay immutable.", path=str(original), record=record))
        if not working.exists():
            findings.append(issue("error", "missing_working_file", "Working file is missing.", path=str(working), record=record))
        else:
            check["working_sha256"] = sha256_file(working)
        file_checks.append(check)

    original_extra = sorted(rel_set(workspace / "original") - manifest_relatives)
    working_extra = sorted(rel_set(workspace / "working") - manifest_relatives)
    for relative in original_extra:
        findings.append(issue("error", "extra_original_file", "Unexpected file under original/.", path=str(workspace / "original" / Path(relative))))
    for relative in working_extra:
        findings.append(issue("warning", "extra_working_file", "Unexpected file under working/; future packaging must explicitly include or reject it.", path=str(workspace / "working" / Path(relative))))

    if not (workspace / "proposal.json").exists():
        findings.append(issue("warning", "missing_proposal_copy", "proposal.json is missing from workspace.", path=str(workspace / "proposal.json")))
    if not (workspace / "safety.json").exists():
        findings.append(issue("warning", "missing_safety_copy", "safety.json is missing from workspace.", path=str(workspace / "safety.json")))

    return result(workspace, findings, file_checks, manifest=manifest)


def result(workspace: Path, findings: list[dict[str, Any]], file_checks: list[dict[str, Any]], manifest: dict[str, Any] | None = None) -> dict[str, Any]:
    errors = [row for row in findings if row.get("severity") == "error"]
    warnings = [row for row in findings if row.get("severity") == "warning"]
    return {
        "schema": "onec_patch_workspace_integrity.v1",
        "workspace": str(workspace),
        "manifest_schema": (manifest or {}).get("schema"),
        "passed": not errors,
        "findings": findings,
        "file_checks": file_checks,
        "counts": {
            "files": len(file_checks),
            "errors": len(errors),
            "warnings": len(warnings),
        },
    }


def main() -> int:
    parser = argparse.ArgumentParser(description="Check 1C patch workspace integrity.")
    parser.add_argument("--workspace", type=Path, required=True)
    parser.add_argument("--output", type=Path)
    args = parser.parse_args()

    check = check_workspace(args.workspace)
    output = json.dumps(check, ensure_ascii=False, indent=2) + "\n"
    if args.output:
        args.output.parent.mkdir(parents=True, exist_ok=True)
        args.output.write_text(output, encoding="utf-8")
    print(json.dumps({"output": str(args.output) if args.output else None, "passed": check["passed"], "counts": check["counts"]}, ensure_ascii=False))
    return 0 if check["passed"] else 2


if __name__ == "__main__":
    raise SystemExit(main())