Olympus/backend/hermes/app/storage/local.py
2026-07-03 18:11:58 +02:00

141 lines
5.1 KiB
Python

from pathlib import Path
import hashlib
from app.storage.base import StorageProvider
from app.storage.exceptions import StorageFileNotFoundError, StorageValidationError
from app.storage.schemas import FileMetadata, StoredFileMetadata
class LocalDiskStorageProvider(StorageProvider):
def __init__(
self,
*,
base_path: str,
legacy_base_paths: list[str] | None = None,
) -> None:
self.base_path = Path(base_path).expanduser().resolve()
self.legacy_base_paths = [
Path(path).expanduser().resolve()
for path in legacy_base_paths or []
if path
]
self._ensure_layout()
def save_file(
self,
*,
namespace: str,
content: bytes,
original_filename: str,
stored_filename: str,
mime_type: str,
) -> StoredFileMetadata:
if Path(stored_filename).name != stored_filename or ".." in Path(stored_filename).parts:
raise StorageValidationError("Ungültiger gespeicherter Dateiname")
namespace_path = self._safe_namespace_path(namespace)
namespace_path.mkdir(parents=True, exist_ok=True)
absolute_path = self._safe_child(namespace_path / stored_filename)
absolute_path.write_bytes(content)
relative_path = absolute_path.relative_to(self.base_path).as_posix()
checksum = self._checksum(absolute_path)
return StoredFileMetadata(
storage_key=relative_path,
original_filename=original_filename,
stored_filename=stored_filename,
relative_path=relative_path,
absolute_path=absolute_path,
mime_type=mime_type,
size=len(content),
checksum_sha256=checksum,
)
def open_file(self, storage_key: str) -> Path:
path = self._resolve_storage_key(storage_key)
if not path.exists() or not path.is_file():
raise StorageFileNotFoundError("Datei nicht gefunden")
return path
def delete_file(self, storage_key: str) -> None:
path = self._resolve_storage_key(storage_key)
if not path.exists():
return
if not path.is_file():
raise StorageValidationError("Storage-Key verweist nicht auf eine Datei")
path.unlink()
def file_exists(self, storage_key: str) -> bool:
try:
path = self._resolve_storage_key(storage_key)
except StorageValidationError:
return False
return path.exists() and path.is_file()
def get_file_metadata(self, storage_key: str) -> FileMetadata:
path = self.open_file(storage_key)
return FileMetadata(
storage_key=storage_key,
relative_path=self._relative_storage_key(path),
absolute_path=path,
size=path.stat().st_size,
checksum_sha256=self._checksum(path),
)
def _ensure_layout(self) -> None:
for relative_path in [
"knowledge/documents",
"knowledge/thumbnails",
"customers",
"projects",
"tickets",
"imports",
"temp",
]:
(self.base_path / relative_path).mkdir(parents=True, exist_ok=True)
def _safe_namespace_path(self, namespace: str) -> Path:
if namespace.startswith("/") or ".." in Path(namespace).parts:
raise StorageValidationError("Ungültiger Storage-Namespace")
return self._safe_child(self.base_path / namespace)
def _safe_child(self, path: Path) -> Path:
resolved = path.resolve()
if self.base_path != resolved and self.base_path not in resolved.parents:
raise StorageValidationError("Ungültiger Storage-Pfad")
return resolved
def _resolve_storage_key(self, storage_key: str) -> Path:
raw_path = Path(storage_key)
candidate_paths: list[Path] = []
if raw_path.is_absolute():
candidate_paths.append(raw_path)
else:
candidate_paths.append(self.base_path / raw_path)
candidate_paths.extend(legacy_base / raw_path for legacy_base in self.legacy_base_paths)
for candidate in candidate_paths:
resolved = candidate.expanduser().resolve()
if self._is_allowed_path(resolved):
return resolved
raise StorageValidationError("Ungültiger Storage-Key")
def _is_allowed_path(self, path: Path) -> bool:
roots = [self.base_path, *self.legacy_base_paths]
return any(root == path or root in path.parents for root in roots)
def _relative_storage_key(self, path: Path) -> str:
for root in [self.base_path, *self.legacy_base_paths]:
try:
return path.relative_to(root).as_posix()
except ValueError:
continue
return path.name
@staticmethod
def _checksum(path: Path) -> str:
digest = hashlib.sha256()
with path.open("rb") as handle:
for chunk in iter(lambda: handle.read(1024 * 1024), b""):
digest.update(chunk)
return digest.hexdigest()