← Files Code Ontology CompanionARCHIVED FILE
skills/manage-code-ontology/scripts/companion.py
55.3 KB · Oct 4, 2026 · 12:29 UTC
#!/usr/bin/env python3
"""Maintain a local, versioned code ontology without executing target code.
The companion uses only the Python standard library and the bundled
``code_ontology_core`` analyzer. It stores immutable snapshots, a durable
lineage journal, and a small local registry used by the read-only MCP server.
"""
from __future__ import annotations
import argparse
import copy
import datetime as dt
import hashlib
import json
import math
import os
import platform
import re
import shutil
import stat
import sys
import tempfile
import time
import uuid
from pathlib import Path
from collections import OrderedDict
from typing import Any, Iterable
import code_ontology_core as core
COMPANION_VERSION = "0.6.0"
OLLAMA_MACOS_APP = Path("/Applications/Ollama.app")
WORKSPACE_SCHEMA_VERSION = 1
PROVENANCE_NS = "https://battle-doll.github.io/code-ontology-companion/provenance#"
PROV_NS = "http://www.w3.org/ns/prov#"
EVIDENCE_TYPES = {"observed", "declared", "inferred", "validated", "approved"}
ADAPTER_SUPPORT_STATUSES = {"supported", "partial", "unsupported"}
MAX_ADAPTER_CAPABILITIES = 32
MAX_UNSUPPORTED_RUNTIME_ITEMS = 32
MAX_SNAPSHOT_CACHE_BYTES = 8 * 1024 * 1024
MAX_SNAPSHOT_CACHE_ENTRIES = 2
_SNAPSHOT_CACHE: OrderedDict[str, tuple[int, dict[str, Any], core.SnapshotIndex]] = OrderedDict()
EVENT_KINDS = {
"decision",
"change",
"validation",
"activation",
"observation",
"outcome",
"rollback",
"note",
}
class CompanionError(RuntimeError):
"""Expected, user-actionable failure."""
def _now() -> str:
return dt.datetime.now(dt.timezone.utc).replace(microsecond=0).isoformat()
def _timestamp_id() -> str:
return dt.datetime.now(dt.timezone.utc).strftime("%Y%m%dT%H%M%SZ")
def _json_bytes(value: Any) -> bytes:
return json.dumps(
value,
ensure_ascii=False,
sort_keys=True,
separators=(",", ":"),
).encode("utf-8")
def _is_link_like(path_stat: os.stat_result) -> bool:
attributes = getattr(path_stat, "st_file_attributes", 0)
return stat.S_ISLNK(path_stat.st_mode) or bool(attributes & core.WINDOWS_REPARSE_POINT)
def _read_regular_bytes(
path: Path,
label: str,
maximum: int | None = None,
) -> bytes:
try:
initial_stat = path.lstat()
except FileNotFoundError:
raise CompanionError(f"{label} is missing.")
except OSError as exc:
raise CompanionError(f"{label} is unreadable: {exc}") from exc
if _is_link_like(initial_stat) or not stat.S_ISREG(initial_stat.st_mode):
raise CompanionError(f"{label} must be a regular file.")
if getattr(initial_stat, "st_nlink", 1) != 1:
raise CompanionError(f"{label} may not have multiple filesystem links.")
if maximum is not None and initial_stat.st_size > maximum:
raise CompanionError(f"{label} exceeds the allowed size.")
flags = os.O_RDONLY
flags |= getattr(os, "O_BINARY", 0)
flags |= getattr(os, "O_CLOEXEC", 0)
flags |= getattr(os, "O_NOFOLLOW", 0)
flags |= getattr(os, "O_NONBLOCK", 0)
try:
descriptor = os.open(str(path), flags)
except OSError as exc:
raise CompanionError(f"{label} could not be opened safely: {exc}") from exc
try:
try:
opened_stat = os.fstat(descriptor)
if _is_link_like(opened_stat) or not stat.S_ISREG(opened_stat.st_mode):
raise CompanionError(f"{label} must be a regular file.")
if getattr(opened_stat, "st_nlink", 1) != 1:
raise CompanionError(f"{label} may not have multiple filesystem links.")
if not core._same_file(initial_stat, opened_stat):
raise CompanionError(f"{label} changed before it could be read.")
if core._stable_file_metadata(initial_stat) != core._stable_file_metadata(opened_stat):
raise CompanionError(f"{label} changed before it could be read.")
with os.fdopen(descriptor, "rb", closefd=False) as handle:
content = handle.read(maximum + 1) if maximum is not None else handle.read()
final_stat = os.fstat(descriptor)
except OSError as exc:
raise CompanionError(f"{label} is unreadable: {exc}") from exc
finally:
os.close(descriptor)
if not core._same_file(opened_stat, final_stat):
raise CompanionError(f"{label} changed while it was being read.")
if core._stable_file_metadata(opened_stat) != core._stable_file_metadata(final_stat):
raise CompanionError(f"{label} changed while it was being read.")
if maximum is not None and len(content) > maximum:
raise CompanionError(f"{label} exceeds the allowed size.")
try:
current_stat = path.lstat()
except OSError as exc:
raise CompanionError(f"{label} could not be verified: {exc}") from exc
if _is_link_like(current_stat) or not stat.S_ISREG(current_stat.st_mode):
raise CompanionError(f"{label} must be a regular file.")
if getattr(current_stat, "st_nlink", 1) != 1:
raise CompanionError(f"{label} may not have multiple filesystem links.")
if not core._same_file(final_stat, current_stat):
raise CompanionError(f"{label} changed while it was being read.")
if core._stable_file_metadata(final_stat) != core._stable_file_metadata(current_stat):
raise CompanionError(f"{label} changed while it was being read.")
return content
def _resolve_existing_dir(raw_path: str | Path, label: str) -> Path:
path = Path(raw_path).expanduser()
try:
path_stat = path.lstat()
except OSError:
raise CompanionError(f"{label} is not a readable directory.")
if _is_link_like(path_stat):
raise CompanionError(f"{label} may not be a symbolic link or reparse point.")
if not stat.S_ISDIR(path_stat.st_mode):
raise CompanionError(f"{label} is not a readable directory.")
return path.resolve()
def _resolve_managed_directory(
workspace: Path,
name: str,
*,
create: bool = False,
required: bool = True,
) -> Path | None:
"""Resolve a Companion-owned directory without following links or reparse points."""
if name not in {"snapshots", ".staging"}:
raise CompanionError("Unsupported managed directory.")
workspace_resolved = workspace.resolve()
candidate = workspace_resolved / name
if create:
try:
candidate.mkdir(mode=0o700, parents=False, exist_ok=False)
except FileExistsError:
pass
except OSError as exc:
raise CompanionError(f"Managed directory could not be created: {name}: {exc}") from exc
try:
candidate_stat = candidate.lstat()
except FileNotFoundError:
if required:
raise CompanionError(f"Managed directory is missing: {name}")
return None
except OSError as exc:
raise CompanionError(f"Managed directory is unreadable: {name}: {exc}") from exc
if _is_link_like(candidate_stat) or not stat.S_ISDIR(candidate_stat.st_mode):
raise CompanionError(
f"Managed directory may not be a symbolic link or reparse point: {name}"
)
resolved = candidate.resolve()
if resolved.parent != workspace_resolved or resolved.name != name:
raise CompanionError(f"Managed directory escaped the workspace: {name}")
return resolved
def _resolve_new_workspace(raw_path: str | Path, repo: Path) -> Path:
raw = Path(raw_path).expanduser()
if raw.exists() or raw.is_symlink():
raise CompanionError("Workspace already exists; refusing to replace it.")
parent = raw.parent
parent_resolved = _resolve_existing_dir(parent, "Workspace parent")
workspace = parent_resolved / raw.name
if core._is_relative_to(workspace, repo):
raise CompanionError("Workspace must be outside the target repository.")
if core._is_relative_to(repo, workspace):
raise CompanionError("Workspace may not contain the target repository.")
return workspace
def _atomic_write(path: Path, content: bytes, mode: int = 0o600) -> None:
path.parent.mkdir(parents=True, exist_ok=True)
descriptor, temp_name = tempfile.mkstemp(prefix=f".{path.name}.", dir=str(path.parent))
try:
with os.fdopen(descriptor, "wb") as handle:
handle.write(content)
handle.flush()
os.fsync(handle.fileno())
try:
os.chmod(temp_name, mode)
except OSError:
pass
os.replace(temp_name, path)
finally:
try:
os.unlink(temp_name)
except FileNotFoundError:
pass
def _atomic_json(path: Path, value: Any, mode: int = 0o600) -> None:
_atomic_write(path, json.dumps(value, indent=2, ensure_ascii=False).encode("utf-8") + b"\n", mode)
def _read_json(path: Path, label: str) -> dict[str, Any]:
return _json_object_from_bytes(_read_regular_bytes(path, label), label)
def _json_object_from_bytes(content: bytes, label: str) -> dict[str, Any]:
def reject_constant(value: str) -> None:
raise ValueError(f"non-finite number {value}")
def unique_object(pairs: list[tuple[str, Any]]) -> dict[str, Any]:
value: dict[str, Any] = {}
for key, item in pairs:
if key in value:
raise ValueError(f"duplicate object key {key}")
value[key] = item
return value
try:
value = json.loads(
content.decode("utf-8"),
parse_constant=reject_constant,
object_pairs_hook=unique_object,
)
except (UnicodeDecodeError, json.JSONDecodeError, ValueError, RecursionError) as exc:
raise CompanionError(f"{label} is unreadable: {exc}") from exc
if not isinstance(value, dict):
raise CompanionError(f"{label} must contain a JSON object.")
return value
def _workspace(raw_path: str | Path) -> tuple[Path, dict[str, Any]]:
workspace = _resolve_existing_dir(raw_path, "Workspace")
config = _read_json(workspace / "companion.json", "Workspace configuration")
if config.get("schemaVersion") != WORKSPACE_SCHEMA_VERSION:
raise CompanionError("Unsupported workspace schema version.")
repo = _resolve_existing_dir(config.get("repositoryRoot", ""), "Configured repository")
if core._is_relative_to(workspace, repo) or core._is_relative_to(repo, workspace):
raise CompanionError("Workspace and repository boundaries overlap.")
return workspace, config
def _data_home() -> Path:
override = os.environ.get("CODE_ONTOLOGY_HOME")
if override:
return Path(override).expanduser().resolve()
if os.name == "nt":
base = os.environ.get("LOCALAPPDATA")
if base:
return Path(base) / "CodeOntologyCompanion"
return Path.home() / "AppData" / "Local" / "CodeOntologyCompanion"
if sys.platform == "darwin":
return Path.home() / "Library" / "Application Support" / "CodeOntologyCompanion"
xdg = os.environ.get("XDG_DATA_HOME")
return (Path(xdg).expanduser() if xdg else Path.home() / ".local" / "share") / (
"code-ontology-companion"
)
def _registry_path() -> Path:
return _data_home() / "registry.json"
def _load_registry() -> dict[str, Any]:
path = _registry_path()
if not path.exists():
return {"schemaVersion": 1, "workspaces": []}
value = _read_json(path, "Companion registry")
if value.get("schemaVersion") != 1 or not isinstance(value.get("workspaces"), list):
raise CompanionError("Unsupported companion registry.")
return value
def _register_workspace(workspace: Path, config: dict[str, Any]) -> None:
registry = _load_registry()
items = [
item
for item in registry["workspaces"]
if isinstance(item, dict) and item.get("id") != config["workspaceId"]
]
items.append(
{
"id": config["workspaceId"],
"label": config["repositoryLabel"],
"workspace": str(workspace),
"registeredAt": _now(),
}
)
registry["workspaces"] = sorted(items, key=lambda item: (item["label"].lower(), item["id"]))
_atomic_json(_registry_path(), registry)
def list_workspaces() -> dict[str, Any]:
registry = _load_registry()
available: list[dict[str, Any]] = []
stale: list[dict[str, Any]] = []
for item in registry["workspaces"]:
if not isinstance(item, dict):
continue
path = Path(str(item.get("workspace", ""))).expanduser()
summary = {"id": item.get("id"), "label": item.get("label")}
if path.is_dir() and (path / "companion.json").is_file():
available.append(summary)
else:
stale.append(summary)
return {
"status": "ok",
"workspaces": available,
"staleRegistrations": stale,
"registryContainsAbsolutePaths": True,
}
def resolve_registered_workspace(workspace_id: str) -> Path:
registry = _load_registry()
matches = [
item
for item in registry["workspaces"]
if isinstance(item, dict) and item.get("id") == workspace_id
]
if not matches:
raise CompanionError(f"Unknown workspace id: {workspace_id}")
workspace, _ = _workspace(str(matches[-1].get("workspace", "")))
return workspace
def _manifest(repo: Path) -> dict[str, Any]:
sources, skipped = core.discover_sources(repo)
files: list[dict[str, Any]] = []
digest = hashlib.sha256()
for source in sources:
try:
content = core._safe_read_bytes(source)
relative = source.relative_to(repo).as_posix()
except (OSError, ValueError, core.OntologyError):
raise CompanionError("A source changed or became unreadable during snapshot planning.")
file_hash = hashlib.sha256(content).hexdigest()
item = {
"path": relative,
"language": core.SUPPORTED_SUFFIXES[source.suffix.lower()],
"bytes": len(content),
"sha256": file_hash,
}
files.append(item)
digest.update(_json_bytes(item))
return {
"algorithm": "sha256",
"fingerprint": digest.hexdigest(),
"files": files,
"skipped": dict(sorted(skipped.items())),
}
def _git_revision(repo: Path) -> str | None:
git_dir = repo / ".git"
try:
git_stat = git_dir.lstat()
except OSError:
return None
if _is_link_like(git_stat) or not stat.S_ISDIR(git_stat.st_mode):
return None
head = git_dir / "HEAD"
try:
value = _read_regular_bytes(head, "Git HEAD", 4096).decode("utf-8").strip()
except (CompanionError, UnicodeError):
return None
if value.startswith("ref: "):
ref_name = value[5:].strip()
if (
"\\" in ref_name
or not re.fullmatch(r"refs/(?:heads|tags)/[A-Za-z0-9._/-]+", ref_name)
or any(part in {"", ".", ".."} for part in ref_name.split("/"))
):
return None
try:
git_root = git_dir.resolve()
refs_root = (git_dir / "refs").resolve()
if not core._is_relative_to(refs_root, git_root):
return None
ref_path = (git_dir / ref_name).resolve()
if not core._is_relative_to(ref_path, refs_root):
return None
value = _read_regular_bytes(ref_path, "Git reference", 4096).decode(
"utf-8"
).strip()
except (CompanionError, OSError, UnicodeError):
return None
return value if len(value) == 40 and all(char in "0123456789abcdefABCDEF" for char in value) else None
def _state(workspace: Path) -> dict[str, Any]:
path = workspace / "state.json"
if not path.exists():
return {"schemaVersion": 1, "currentSnapshot": None, "previousSnapshot": None}
value = _read_json(path, "Workspace state")
if value.get("schemaVersion") != 1:
raise CompanionError("Unsupported workspace state.")
return value
def _append_journal(workspace: Path, event: dict[str, Any]) -> None:
path = workspace / "lineage.jsonl"
payload = _json_bytes(event) + b"\n"
flags = os.O_APPEND | os.O_WRONLY
flags |= getattr(os, "O_BINARY", 0)
flags |= getattr(os, "O_CLOEXEC", 0)
flags |= getattr(os, "O_NOFOLLOW", 0)
try:
initial_stat = path.lstat()
except FileNotFoundError:
initial_stat = None
flags |= os.O_CREAT | os.O_EXCL
except OSError as exc:
raise CompanionError(f"Lineage journal is unreadable: {exc}") from exc
if initial_stat is not None and (
_is_link_like(initial_stat) or not stat.S_ISREG(initial_stat.st_mode)
):
raise CompanionError("Lineage journal must be a regular file.")
try:
descriptor = os.open(str(path), flags, 0o600)
except OSError as exc:
raise CompanionError(f"Lineage journal could not be opened safely: {exc}") from exc
try:
opened_stat = os.fstat(descriptor)
if _is_link_like(opened_stat) or not stat.S_ISREG(opened_stat.st_mode):
raise CompanionError("Lineage journal must be a regular file.")
if getattr(opened_stat, "st_nlink", 1) != 1:
raise CompanionError("Lineage journal may not have multiple filesystem links.")
try:
current_stat = path.lstat()
except OSError as exc:
raise CompanionError(f"Lineage journal could not be verified: {exc}") from exc
if _is_link_like(current_stat) or not stat.S_ISREG(current_stat.st_mode):
raise CompanionError("Lineage journal must be a regular file.")
if getattr(current_stat, "st_nlink", 1) != 1:
raise CompanionError("Lineage journal may not have multiple filesystem links.")
if not core._same_file(opened_stat, current_stat):
raise CompanionError("Lineage journal changed before append.")
if initial_stat is not None and not core._same_file(initial_stat, opened_stat):
raise CompanionError("Lineage journal changed before append.")
os.write(descriptor, payload)
os.fsync(descriptor)
finally:
os.close(descriptor)
_render_lineage_turtle(workspace)
def _read_journal(workspace: Path) -> list[dict[str, Any]]:
path = workspace / "lineage.jsonl"
try:
path.lstat()
except FileNotFoundError:
return []
except OSError:
raise CompanionError("Lineage journal is unreadable.")
events: list[dict[str, Any]] = []
try:
content = _read_regular_bytes(path, "Lineage journal").decode("utf-8")
for line_number, line in enumerate(content.splitlines(), 1):
if not line.strip():
continue
value = json.loads(line)
if not isinstance(value, dict):
raise ValueError(f"line {line_number} is not an object")
events.append(value)
except (UnicodeDecodeError, json.JSONDecodeError, ValueError) as exc:
raise CompanionError(f"Lineage journal is invalid: {exc}")
return events
def _ttl_literal(value: Any) -> str:
return core._turtle_literal(value)
def _render_lineage_turtle(workspace: Path) -> None:
lines = [
f"@prefix coc: <{PROVENANCE_NS}> .",
f"@prefix prov: <{PROV_NS}> .",
"@prefix rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#> .",
"@prefix rdfs: <http://www.w3.org/2000/01/rdf-schema#> .",
"@prefix xsd: <http://www.w3.org/2001/XMLSchema#> .",
"",
]
for event in _read_journal(workspace):
event_id = str(event.get("eventId", "")).replace("-", "")
if not event_id:
continue
subject = f"coc:event-{event_id}"
lines.extend(
[
f"{subject} a prov:Activity ;",
f" rdfs:label {_ttl_literal(event.get('summary', ''))} ;",
f" coc:eventKind {_ttl_literal(event.get('kind', ''))} ;",
f" coc:evidenceType {_ttl_literal(event.get('evidenceType', ''))} ;",
f" prov:generatedAtTime {_ttl_literal(event.get('recordedAt', ''))} ;",
]
)
optional_values = [
("coc:subject", event.get("subject")),
("coc:snapshotId", event.get("snapshotId")),
("coc:previousSnapshotId", event.get("previousSnapshotId")),
("coc:repositoryRevision", event.get("repositoryRevision")),
]
present = [(predicate, value) for predicate, value in optional_values if value]
for index, (predicate, value) in enumerate(present):
terminal = " ." if index == len(present) - 1 else " ;"
lines.append(f" {predicate} {_ttl_literal(value)}{terminal}")
if not present:
lines[-1] = lines[-1][:-1] + "."
lines.append("")
_atomic_write(workspace / "lineage.ttl", ("\n".join(lines) + "\n").encode("utf-8"), 0o600)
def _ollama_runtime_indicator() -> bool:
if shutil.which("ollama"):
return True
if sys.platform != "darwin":
return False
try:
metadata = OLLAMA_MACOS_APP.lstat()
except OSError:
return False
return not _is_link_like(metadata) and stat.S_ISDIR(metadata.st_mode)
def doctor(repo_path: str | None = None) -> dict[str, Any]:
commands = {
name: shutil.which(name)
for name in ("git", "java", "ollama", "lms", "docker", "podman")
}
commands["ollama"] = commands["ollama"] or (
"known-macos-app" if _ollama_runtime_indicator() else None
)
result: dict[str, Any] = {
"status": "ok",
"companionVersion": COMPANION_VERSION,
"platform": {"system": platform.system(), "machine": platform.machine()},
"python": {
"version": platform.python_version(),
"compatible": sys.version_info >= (3, 9),
"executableLabel": Path(sys.executable).name,
},
"requiredDependencies": [],
"optionalRuntimesDetected": {
name: bool(path) for name, path in commands.items() if name in {"java", "ollama", "lms", "docker", "podman"}
},
"defaults": {
"graphStore": "immutable JSON snapshots",
"portableExport": "RDF 1.1 Turtle",
"localLlmRequired": False,
"networkRequired": False,
"backgroundServiceInstalled": False,
},
}
if repo_path:
repo = _resolve_existing_dir(repo_path, "Repository")
preflight = core.preflight_document(repo)
result["repository"] = {
"label": preflight["repository_name"],
"sourceFileCount": preflight["source_file_count"],
"supportedLanguages": preflight["supported_languages"],
}
return result
def preflight(repo_path: str) -> dict[str, Any]:
repo = _resolve_existing_dir(repo_path, "Repository")
result = core.preflight_document(repo)
result["companion"] = {
"version": COMPANION_VERSION,
"writesDuringPreflight": False,
"requiresLocalLlm": False,
"requiresGraphDatabase": False,
"nextStep": "Obtain authorization and choose a workspace outside the repository.",
}
return result
def _snapshot_metadata(path: Path) -> dict[str, Any]:
return _read_json(path / "snapshot.json", "Snapshot metadata")
def _snapshot_path(workspace: Path, snapshot_id: str) -> Path:
if not snapshot_id or snapshot_id in {".", ".."} or "/" in snapshot_id or "\\" in snapshot_id:
raise CompanionError("Invalid snapshot id.")
snapshots_root = _resolve_managed_directory(workspace, "snapshots")
assert snapshots_root is not None
path = snapshots_root / snapshot_id
resolved = _resolve_existing_dir(path, "Snapshot")
if resolved.parent != snapshots_root:
raise CompanionError("Snapshot escaped the workspace.")
return resolved
def _resolve_snapshot_alias(workspace: Path, value: str) -> str:
state = _state(workspace)
if value == "current":
snapshot_id = state.get("currentSnapshot")
elif value == "previous":
snapshot_id = state.get("previousSnapshot")
else:
snapshot_id = value
if not snapshot_id:
raise CompanionError(f"Snapshot alias has no value: {value}")
return str(snapshot_id)
def _create_snapshot(
workspace: Path,
config: dict[str, Any],
trigger: str,
planned_manifest: dict[str, Any] | None = None,
) -> dict[str, Any]:
repo = _resolve_existing_dir(config["repositoryRoot"], "Configured repository")
before_manifest = planned_manifest or _manifest(repo)
current_state = _state(workspace)
current_id = current_state.get("currentSnapshot")
previous_index_path: str | None = None
if current_id:
current_path = _snapshot_path(workspace, str(current_id))
previous_index_path = str(current_path / "ontology.json")
current_manifest = _read_json(current_path / "source-manifest.json", "Source manifest")
current_snapshot = _snapshot_metadata(current_path)
if (
current_manifest.get("fingerprint") == before_manifest["fingerprint"]
and current_snapshot.get("analyzerVersion") == core.PLUGIN_VERSION
and current_snapshot.get("companionVersion") == COMPANION_VERSION
):
return {
"status": "no_change",
"workspaceId": config["workspaceId"],
"snapshotId": current_id,
"fingerprint": before_manifest["fingerprint"],
}
run_id = str(uuid.uuid4())
snapshots_root = _resolve_managed_directory(workspace, "snapshots", create=True)
staging_root = _resolve_managed_directory(workspace, ".staging", create=True)
assert snapshots_root is not None
assert staging_root is not None
staging = staging_root / run_id
snapshot_id = f"{_timestamp_id()}-{before_manifest['fingerprint'][:12]}"
final = snapshots_root / snapshot_id
if final.exists():
snapshot_id = f"{snapshot_id}-{run_id[:8]}"
final = snapshots_root / snapshot_id
started_at = _now()
try:
result = core.write_index(repo, staging, authorized=True, overwrite=False)
after_manifest = _manifest(repo)
if after_manifest["fingerprint"] != before_manifest["fingerprint"]:
raise CompanionError(
"Repository changed during analysis; staged artifacts were not promoted. Run sync again."
)
document = _read_json(staging / "ontology.json", "Staged ontology")
document["companion"] = {
"workspaceId": config["workspaceId"],
"snapshotId": snapshot_id,
"sourceFingerprint": before_manifest["fingerprint"],
"evidenceType": "observed",
"companionVersion": COMPANION_VERSION,
}
_atomic_json(staging / "ontology.json", document, 0o600)
core.write_visualization(
str(staging / "ontology.json"),
str(staging / "graph.html"),
max_nodes=core.VISUALIZATION_MAX_VISIBLE_NODES,
overwrite=False,
previous_index_path=previous_index_path,
)
snapshot = {
"schemaVersion": 1,
"snapshotId": snapshot_id,
"workspaceId": config["workspaceId"],
"repositoryLabel": config["repositoryLabel"],
"repositoryRevision": _git_revision(repo),
"sourceFingerprint": before_manifest["fingerprint"],
"createdAt": _now(),
"trigger": trigger,
"analyzerVersion": core.PLUGIN_VERSION,
"companionVersion": COMPANION_VERSION,
"counts": result["statistics"],
}
_atomic_json(staging / "source-manifest.json", after_manifest, 0o600)
_atomic_json(staging / "snapshot.json", snapshot, 0o600)
os.replace(staging, final)
state = {
"schemaVersion": 1,
"currentSnapshot": snapshot_id,
"previousSnapshot": current_id,
"promotedAt": snapshot["createdAt"],
"lastRunId": run_id,
}
_atomic_json(workspace / "state.json", state, 0o600)
_append_journal(
workspace,
{
"eventId": str(uuid.uuid4()),
"kind": "observation",
"evidenceType": "observed",
"summary": f"Promoted ontology snapshot {snapshot_id}",
"subject": config["repositoryLabel"],
"snapshotId": snapshot_id,
"previousSnapshotId": current_id,
"repositoryRevision": snapshot["repositoryRevision"],
"recordedAt": snapshot["createdAt"],
"runId": run_id,
"trigger": trigger,
},
)
return {
"status": "promoted",
"workspaceId": config["workspaceId"],
"snapshotId": snapshot_id,
"previousSnapshotId": current_id,
"counts": result["statistics"],
"sourceFileCount": len(before_manifest["files"]),
"portableRdf": str(final / "ontology.ttl"),
"visualization": str(final / "graph.html"),
"lineage": str(workspace / "lineage.ttl"),
"targetCodeExecuted": False,
"networkAccess": False,
}
except Exception:
if staging.exists():
shutil.rmtree(staging, ignore_errors=True)
raise
finally:
if staging_root.exists():
try:
staging_root.rmdir()
except OSError:
pass
def initialize(
repo_path: str,
workspace_path: str,
authorized: bool,
label: str | None = None,
) -> dict[str, Any]:
if not authorized:
raise CompanionError("Initialization requires --authorized.")
repo = _resolve_existing_dir(repo_path, "Repository")
workspace = _resolve_new_workspace(workspace_path, repo)
planned = _manifest(repo)
if not planned["files"]:
raise CompanionError("No supported Java or Python source files were found.")
workspace.mkdir(mode=0o700, parents=False, exist_ok=False)
config = {
"schemaVersion": WORKSPACE_SCHEMA_VERSION,
"workspaceId": str(uuid.uuid4()),
"repositoryLabel": (label or repo.name).strip() or repo.name,
"repositoryRoot": str(repo),
"createdAt": _now(),
"privacy": {
"storesAbsoluteRepositoryPathLocally": True,
"storesSourceBodies": False,
"storesPerFileSha256Locally": True,
"networkAccess": False,
},
}
try:
_atomic_json(workspace / "companion.json", config, 0o600)
result = _create_snapshot(workspace, config, trigger="initialize", planned_manifest=planned)
_register_workspace(workspace, config)
result["workspace"] = str(workspace)
result["repositoryLabel"] = config["repositoryLabel"]
return result
except Exception:
# Initialization owns the newly-created, previously nonexistent path.
shutil.rmtree(workspace, ignore_errors=True)
raise
def sync(workspace_path: str, trigger: str = "manual") -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
return _create_snapshot(workspace, config, trigger=trigger)
def status(workspace_path: str, check_freshness: bool = True) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
state = _state(workspace)
current_id = state.get("currentSnapshot")
if not current_id:
return {
"status": "partial",
"workspaceId": config["workspaceId"],
"repositoryLabel": config["repositoryLabel"],
"freshness": "partial",
"message": "No promoted snapshot exists.",
}
snapshot_path = _snapshot_path(workspace, str(current_id))
snapshot = _snapshot_metadata(snapshot_path)
document = _read_json(snapshot_path / "ontology.json", "Ontology index")
version_current = (
snapshot.get("analyzerVersion") == core.PLUGIN_VERSION
and snapshot.get("companionVersion") == COMPANION_VERSION
)
freshness = "stale" if not version_current else "unknown"
if check_freshness and version_current:
repo = _resolve_existing_dir(config["repositoryRoot"], "Configured repository")
current_manifest = _manifest(repo)
freshness = (
"current"
if current_manifest["fingerprint"] == snapshot["sourceFingerprint"]
else "stale"
)
return {
"status": "ok",
"workspaceId": config["workspaceId"],
"repositoryLabel": config["repositoryLabel"],
"snapshotId": current_id,
"previousSnapshotId": state.get("previousSnapshot"),
"generatedAt": snapshot.get("createdAt"),
"freshness": freshness,
"snapshotAnalyzerVersion": snapshot.get("analyzerVersion"),
"currentAnalyzerVersion": core.PLUGIN_VERSION,
"snapshotCompanionVersion": snapshot.get("companionVersion"),
"currentCompanionVersion": COMPANION_VERSION,
"evidenceType": "observed",
"counts": snapshot.get("counts", {}),
"quality": _quality_summary(document),
"pipelineStatus": "healthy" if freshness != "stale" else "refresh_required",
"message": (
"Run sync to rebuild with the current analyzer and Companion."
if not version_current
else None
),
"portableRdf": str(snapshot_path / "ontology.ttl"),
"visualization": str(snapshot_path / "graph.html"),
}
def history(workspace_path: str, limit: int = 20) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
snapshots_root = _resolve_managed_directory(
workspace,
"snapshots",
required=False,
)
items: list[dict[str, Any]] = []
if snapshots_root is not None:
for child in sorted(snapshots_root.iterdir(), key=lambda path: path.name, reverse=True):
if len(items) >= limit:
break
try:
child_stat = child.lstat()
except OSError as exc:
raise CompanionError(f"Snapshot entry is unreadable: {child.name}: {exc}") from exc
if _is_link_like(child_stat):
raise CompanionError(
f"Snapshot may not be a symbolic link or reparse point: {child.name}"
)
if not stat.S_ISDIR(child_stat.st_mode):
continue
try:
snapshot = _snapshot_path(workspace, child.name)
metadata = _snapshot_metadata(snapshot)
except CompanionError:
continue
items.append(
{
"snapshotId": metadata.get("snapshotId"),
"createdAt": metadata.get("createdAt"),
"repositoryRevision": metadata.get("repositoryRevision"),
"trigger": metadata.get("trigger"),
"counts": metadata.get("counts", {}),
}
)
return {
"status": "ok",
"workspaceId": config["workspaceId"],
"snapshots": items,
"truncated": len(items) >= limit,
}
def _document_for_snapshot(workspace: Path, snapshot: str = "current") -> tuple[str, dict[str, Any]]:
current_id = _resolve_snapshot_alias(workspace, snapshot)
snapshot = _snapshot_path(workspace, current_id)
return current_id, _read_json(snapshot / "ontology.json", "Ontology index")
def _current_document(workspace: Path) -> tuple[str, dict[str, Any]]:
return _document_for_snapshot(workspace)
def _snapshot_view(workspace: Path, snapshot: str) -> tuple[str, dict[str, Any], core.SnapshotIndex]:
"""Cache parsed indexes only after the existing safe read and exact content hash.
No stat-only trust: replacement, symlink and same-size edits still pass through
the safe reader on every request. Public results are copied before returning.
"""
snapshot_id = _resolve_snapshot_alias(workspace, snapshot)
path = _snapshot_path(workspace, snapshot_id) / "ontology.json"
raw = _read_regular_bytes(path, "Ontology index")
key = hashlib.sha256(raw).hexdigest()
cached = _SNAPSHOT_CACHE.get(key)
if cached is not None:
_SNAPSHOT_CACHE.move_to_end(key)
return snapshot_id, cached[1], cached[2]
document = _json_object_from_bytes(raw, "Ontology index")
index = core.SnapshotIndex(document)
if len(raw) <= MAX_SNAPSHOT_CACHE_BYTES:
while _SNAPSHOT_CACHE and (
len(_SNAPSHOT_CACHE) >= MAX_SNAPSHOT_CACHE_ENTRIES
or sum(item[0] for item in _SNAPSHOT_CACHE.values()) + len(raw) > MAX_SNAPSHOT_CACHE_BYTES
):
_SNAPSHOT_CACHE.popitem(last=False)
_SNAPSHOT_CACHE[key] = (len(raw), document, index)
return snapshot_id, document, index
def _bounded_nonnegative_integer(value: Any, maximum: int) -> int:
if isinstance(value, bool) or not isinstance(value, int):
return 0
return max(0, min(value, maximum))
def _quality_summary(document: dict[str, Any]) -> dict[str, Any]:
"""Return a compact, backward-compatible relationship-quality summary."""
edges = document.get("edges")
edge_count = min(len(edges), core.MAX_GRAPH_EDGES) if isinstance(edges, list) else 0
quality = document.get("quality")
if not isinstance(quality, dict):
return {
"status": "legacy_unknown",
"contractVersion": "legacy_unknown",
"totalEdges": edge_count,
"documentedEdges": 0,
"missingEvidence": edge_count,
"coveragePercent": 0.0,
"adapters": {},
}
relationship = quality.get("relationship_evidence")
if not isinstance(relationship, dict):
relationship = {}
total_edges = _bounded_nonnegative_integer(
relationship.get("total_edges"), core.MAX_GRAPH_EDGES
)
documented_edges = min(
_bounded_nonnegative_integer(
relationship.get("documented_edges"), core.MAX_GRAPH_EDGES
),
total_edges,
)
missing_evidence = min(
_bounded_nonnegative_integer(
relationship.get("missing_evidence"), core.MAX_GRAPH_EDGES
),
total_edges,
)
raw_coverage = relationship.get("coverage_percent")
coverage_percent = 0.0
if isinstance(raw_coverage, (int, float)) and not isinstance(raw_coverage, bool):
numeric_coverage = float(raw_coverage)
if math.isfinite(numeric_coverage):
coverage_percent = round(max(0.0, min(numeric_coverage, 100.0)), 2)
contract_version = quality.get("contract_version")
if not isinstance(contract_version, str) or not contract_version.strip():
contract_version = "unknown"
else:
contract_version = contract_version.strip()[:50]
adapters: dict[str, dict[str, Any]] = {}
raw_adapters = quality.get("adapters")
if isinstance(raw_adapters, dict):
for language in ("Java", "Python"):
adapter = raw_adapters.get(language)
if not isinstance(adapter, dict):
continue
adapter_status = adapter.get("status")
if adapter_status not in ADAPTER_SUPPORT_STATUSES:
continue
raw_capabilities = adapter.get("capabilities")
capabilities = (
{
name: value
for name, value in sorted(raw_capabilities.items())[
:MAX_ADAPTER_CAPABILITIES
]
if isinstance(name, str)
and re.fullmatch(r"[a-z][a-z0-9_.-]{2,79}", name)
and value in ADAPTER_SUPPORT_STATUSES
}
if isinstance(raw_capabilities, dict)
else {}
)
raw_unsupported = adapter.get("unsupported_runtime")
unsupported_runtime = (
[
value
for value in raw_unsupported[:MAX_UNSUPPORTED_RUNTIME_ITEMS]
if isinstance(value, str)
and re.fullmatch(r"[a-z][a-z0-9_.-]{2,79}", value)
]
if isinstance(raw_unsupported, list)
else []
)
adapters[language] = {
"status": adapter_status,
"detected": adapter.get("detected") is True,
"capabilities": capabilities,
"unsupportedRuntime": unsupported_runtime,
}
return {
"status": "documented",
"contractVersion": contract_version,
"totalEdges": total_edges,
"documentedEdges": documented_edges,
"missingEvidence": missing_evidence,
"coveragePercent": coverage_percent,
"adapters": adapters,
}
def query(
workspace_path: str, term: str, limit: int = 20, *, snapshot: str = "current",
offset: int = 0, language: str | None = None, node_type: str | None = None,
path_prefix: str | None = None,
) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
snapshot_id, document, index = _snapshot_view(workspace, snapshot)
result = core.query_document(document, term, limit, offset=offset, language=language,
node_type=node_type, path_prefix=path_prefix, index=index)
result.update(
{
"workspaceId": config["workspaceId"],
"snapshotId": snapshot_id,
"freshness": "snapshot",
"evidenceType": "observed",
"quality": _quality_summary(document),
}
)
return copy.deepcopy(result)
def impact(
workspace_path: str, symbol: str, depth: int = 2, *, snapshot: str = "current",
direction: str = "both", relationships: Iterable[str] | None = None, limit: int | None = None,
) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
snapshot_id, document, index = _snapshot_view(workspace, snapshot)
result = core.impact_document(document, symbol, depth, direction=direction,
relationships=relationships, limit=limit, index=index)
for item in result.get("impact", []):
if isinstance(item, dict) and not isinstance(item.get("evidence"), list):
item["evidence"] = []
result.update(
{
"workspaceId": config["workspaceId"],
"snapshotId": snapshot_id,
"freshness": "snapshot",
"evidenceType": "observed",
"quality": _quality_summary(document),
"interpretation": "possible static impact, not runtime proof",
}
)
return copy.deepcopy(result)
def _edge_key(edge: dict[str, Any]) -> tuple[str, str, str]:
return str(edge.get("source")), str(edge.get("target")), str(edge.get("type"))
def _diff_change_basis(
before_doc: dict[str, Any],
after_doc: dict[str, Any],
before_snapshot: dict[str, Any],
after_snapshot: dict[str, Any],
) -> str:
before_companion = before_doc.get("companion")
after_companion = after_doc.get("companion")
before_fingerprint = (
before_companion.get("sourceFingerprint")
if isinstance(before_companion, dict)
else None
) or before_snapshot.get("sourceFingerprint")
after_fingerprint = (
after_companion.get("sourceFingerprint")
if isinstance(after_companion, dict)
else None
) or after_snapshot.get("sourceFingerprint")
if not isinstance(before_fingerprint, str) or not isinstance(after_fingerprint, str):
return "legacy_unknown"
before_analyzer = before_snapshot.get("analyzerVersion")
after_analyzer = after_snapshot.get("analyzerVersion")
before_companion_version = before_snapshot.get("companionVersion")
after_companion_version = after_snapshot.get("companionVersion")
versions = (
before_analyzer,
after_analyzer,
before_companion_version,
after_companion_version,
)
version_known = all(isinstance(value, str) and value for value in versions)
if before_fingerprint != after_fingerprint:
if version_known and (
before_analyzer != after_analyzer
or before_companion_version != after_companion_version
):
return "mixed"
return "source_change"
if not version_known:
return "legacy_unknown"
if before_analyzer != after_analyzer:
return "analyzer_reinterpretation"
if before_companion_version != after_companion_version:
return "analysis_refresh"
return "no_change"
def diff(
workspace_path: str,
before: str = "previous",
after: str = "current",
limit: int = 100,
) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
before_id = _resolve_snapshot_alias(workspace, before)
after_id = _resolve_snapshot_alias(workspace, after)
before_path = _snapshot_path(workspace, before_id)
after_path = _snapshot_path(workspace, after_id)
before_doc = _read_json(before_path / "ontology.json", "Before ontology")
after_doc = _read_json(after_path / "ontology.json", "After ontology")
before_snapshot = _snapshot_metadata(before_path)
after_snapshot = _snapshot_metadata(after_path)
changes = core.canonical_diff(after_doc, before_doc, limit)
changes.pop("available", None)
changes.pop("basis", None)
changes.update({
"status": "ok",
"workspaceId": config["workspaceId"],
"beforeSnapshotId": before_id,
"afterSnapshotId": after_id,
"changeBasis": _diff_change_basis(before_doc, after_doc, before_snapshot, after_snapshot),
"quality": _quality_summary(after_doc),
"interpretation": "Structural and evidence diff; source changes do not prove behavioral changes.",
})
return changes
def record(
workspace_path: str,
kind: str,
summary: str,
evidence_type: str,
subject: str | None = None,
) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
if kind not in EVENT_KINDS:
raise CompanionError(f"Unsupported event kind: {kind}")
if evidence_type not in EVIDENCE_TYPES:
raise CompanionError(f"Unsupported evidence type: {evidence_type}")
clean_summary = summary.strip()
if not clean_summary or len(clean_summary) > 1000:
raise CompanionError("Summary must contain 1 to 1000 characters.")
clean_subject = (subject or "").strip()
if len(clean_subject) > 300:
raise CompanionError("Subject must contain at most 300 characters.")
state = _state(workspace)
event = {
"eventId": str(uuid.uuid4()),
"kind": kind,
"evidenceType": evidence_type,
"summary": clean_summary,
"subject": clean_subject or None,
"snapshotId": state.get("currentSnapshot"),
"recordedAt": _now(),
"recordedBy": "local-user-or-agent",
}
_append_journal(workspace, event)
return {
"status": "recorded",
"workspaceId": config["workspaceId"],
"eventId": event["eventId"],
"kind": kind,
"evidenceType": evidence_type,
"lineage": str(workspace / "lineage.ttl"),
}
def lineage(
workspace_path: str,
limit: int = 50,
evidence_type: str | None = None,
) -> dict[str, Any]:
workspace, config = _workspace(workspace_path)
if evidence_type is not None and evidence_type not in EVIDENCE_TYPES:
raise CompanionError(f"Unsupported evidence type: {evidence_type}")
events = _read_journal(workspace)
if evidence_type is not None:
events = [event for event in events if event.get("evidenceType") == evidence_type]
selected = list(reversed(events[-limit:]))
public_events = [
{
key: event.get(key)
for key in (
"eventId",
"kind",
"evidenceType",
"summary",
"subject",
"snapshotId",
"previousSnapshotId",
"recordedAt",
)
if event.get(key) is not None
}
for event in selected
]
return {
"status": "ok",
"workspaceId": config["workspaceId"],
"events": public_events,
"truncated": len(events) > limit,
}
def watch(workspace_path: str, interval_seconds: int, max_cycles: int) -> int:
workspace, config = _workspace(workspace_path)
cycles = 0
last_fingerprint: str | None = None
while max_cycles == 0 or cycles < max_cycles:
cycles += 1
repo = _resolve_existing_dir(config["repositoryRoot"], "Configured repository")
planned = _manifest(repo)
if planned["fingerprint"] != last_fingerprint:
result = _create_snapshot(
workspace,
config,
trigger="foreground-watch",
planned_manifest=planned,
)
print(json.dumps(result, ensure_ascii=False), flush=True)
last_fingerprint = planned["fingerprint"]
if max_cycles and cycles >= max_cycles:
break
time.sleep(interval_seconds)
return 0
def _json_print(value: Any) -> None:
print(json.dumps(value, indent=2, ensure_ascii=False))
def build_parser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(
prog="companion.py",
description="Maintain a local, versioned code ontology without executing target code.",
)
subparsers = parser.add_subparsers(dest="command", required=True)
command = subparsers.add_parser("doctor", help="Read-only runtime and capability check.")
command.add_argument("--repo", help="Optional authorized repository to include in the check.")
command = subparsers.add_parser("preflight", help="Read-only repository scan; writes nothing.")
command.add_argument("--repo", required=True)
command = subparsers.add_parser("init", help="Create a local Companion workspace.")
command.add_argument("--repo", required=True)
command.add_argument("--workspace", required=True)
command.add_argument("--label")
command.add_argument("--authorized", action="store_true")
command = subparsers.add_parser("sync", help="Create and atomically promote a changed snapshot.")
command.add_argument("--workspace", required=True)
command.add_argument("--trigger", default="manual")
command = subparsers.add_parser("status", help="Inspect workspace and source freshness.")
command.add_argument("--workspace", required=True)
command.add_argument("--no-freshness-check", action="store_true")
subparsers.add_parser("list", help="List registered Companion workspaces.")
command = subparsers.add_parser("history", help="List immutable ontology snapshots.")
command.add_argument("--workspace", required=True)
command.add_argument("--limit", type=int, default=20, choices=range(1, 201), metavar="1..200")
command = subparsers.add_parser("query", help="Search the current ontology snapshot.")
command.add_argument("--workspace", required=True)
command.add_argument("--term", required=True)
command.add_argument("--snapshot", default="current")
command.add_argument("--offset", type=int, default=0)
command.add_argument("--language")
command.add_argument("--node-type")
command.add_argument("--path-prefix")
command.add_argument("--limit", type=int, default=20, choices=range(1, 201), metavar="1..200")
command = subparsers.add_parser("impact", help="Explore bounded possible static impact.")
command.add_argument("--workspace", required=True)
command.add_argument("--symbol", required=True)
command.add_argument("--snapshot", default="current")
command.add_argument("--direction", choices=("incoming", "outgoing", "both"), default="both")
command.add_argument("--relationship", action="append", choices=sorted(core.EDGE_EVIDENCE_DEFAULTS))
command.add_argument("--limit", type=int, default=200)
command.add_argument("--depth", type=int, default=2, choices=range(1, 6), metavar="1..5")
command = subparsers.add_parser("diff", help="Compare two immutable snapshots.")
command.add_argument("--workspace", required=True)
command.add_argument("--before", default="previous")
command.add_argument("--after", default="current")
command.add_argument("--limit", type=int, default=100, choices=range(1, 501), metavar="1..500")
command = subparsers.add_parser("record", help="Append a declared or validated lineage event.")
command.add_argument("--workspace", required=True)
command.add_argument("--kind", required=True, choices=sorted(EVENT_KINDS))
command.add_argument("--evidence-type", required=True, choices=sorted(EVIDENCE_TYPES))
command.add_argument("--summary", required=True)
command.add_argument("--subject")
command = subparsers.add_parser("lineage", help="List recorded provenance events.")
command.add_argument("--workspace", required=True)
command.add_argument("--limit", type=int, default=50, choices=range(1, 501), metavar="1..500")
command.add_argument("--evidence-type", choices=sorted(EVIDENCE_TYPES))
command = subparsers.add_parser(
"watch",
help="Run an explicit foreground polling loop; never installs a daemon.",
)
command.add_argument("--workspace", required=True)
command.add_argument(
"--interval-seconds",
type=int,
default=10,
choices=range(2, 3601),
metavar="2..3600",
)
command.add_argument(
"--max-cycles",
type=int,
default=0,
choices=range(0, 1000001),
metavar="0..1000000",
help="0 runs until interrupted; positive values make bounded runs.",
)
return parser
def main(argv: list[str] | None = None) -> int:
parser = build_parser()
args = parser.parse_args(argv)
try:
if args.command == "doctor":
_json_print(doctor(args.repo))
elif args.command == "preflight":
_json_print(preflight(args.repo))
elif args.command == "init":
_json_print(initialize(args.repo, args.workspace, args.authorized, args.label))
elif args.command == "sync":
_json_print(sync(args.workspace, args.trigger))
elif args.command == "status":
_json_print(status(args.workspace, check_freshness=not args.no_freshness_check))
elif args.command == "list":
_json_print(list_workspaces())
elif args.command == "history":
_json_print(history(args.workspace, args.limit))
elif args.command == "query":
_json_print(query(args.workspace, args.term, args.limit, snapshot=args.snapshot,
offset=args.offset, language=args.language, node_type=args.node_type,
path_prefix=args.path_prefix))
elif args.command == "impact":
_json_print(impact(args.workspace, args.symbol, args.depth, snapshot=args.snapshot,
direction=args.direction, relationships=args.relationship, limit=args.limit))
elif args.command == "diff":
_json_print(diff(args.workspace, args.before, args.after, args.limit))
elif args.command == "record":
_json_print(
record(
args.workspace,
args.kind,
args.summary,
args.evidence_type,
args.subject,
)
)
elif args.command == "lineage":
_json_print(lineage(args.workspace, args.limit, args.evidence_type))
elif args.command == "watch":
return watch(args.workspace, args.interval_seconds, args.max_cycles)
else:
parser.error(f"Unknown command: {args.command}")
except (CompanionError, core.OntologyError) as exc:
print(json.dumps({"status": "error", "message": str(exc)}, ensure_ascii=False), file=sys.stderr)
return 2
except KeyboardInterrupt:
print(json.dumps({"status": "stopped", "message": "Foreground watcher stopped."}))
return 130
return 0
if __name__ == "__main__":
raise SystemExit(main())
SHA-256: 0226cfb9c6fc30b9229c2dbae7a986b3b5a7163fb0c6e1873e9772d67ebe0726