|
1 | 1 | # SPDX-FileCopyrightText: Copyright (c) 2026 NVIDIA CORPORATION & AFFILIATES. All rights reserved. |
2 | 2 | # SPDX-License-Identifier: Apache-2.0 |
3 | 3 |
|
| 4 | +from pathlib import Path |
4 | 5 | from unittest.mock import Mock, patch |
5 | 6 |
|
6 | 7 | import pytest |
7 | | -from data_designer.engine.resources.seed_reader import SeedReaderConfigError |
8 | | -from data_designer_nemo.fileset_filesystem_provider import FilesetFileSystemProvider |
| 8 | +from data_designer.engine.resources.seed_reader import ( |
| 9 | + DirectorySeedReader, |
| 10 | + FileContentsSeedReader, |
| 11 | + SeedReaderConfigError, |
| 12 | +) |
| 13 | +from data_designer_nemo.context import LocalDataDesignerContext |
| 14 | +from data_designer_nemo.fileset_filesystem_provider import FilesetFileSystemProvider, HybridFileSystemProvider |
9 | 15 |
|
10 | 16 |
|
11 | 17 | def test_create_context_roots_reader_in_canonical_fileset_ref() -> None: |
@@ -44,3 +50,49 @@ def test_ensure_root_exists_reports_missing_fileset_path() -> None: |
44 | 50 | provider.ensure_root_exists(runtime_path="docs#corpus") |
45 | 51 |
|
46 | 52 | assert fs_class.return_value.exists.call_count == 2 |
| 53 | + |
| 54 | + |
| 55 | +def test_hybrid_routes_existing_local_directory_to_disk(tmp_path: Path) -> None: |
| 56 | + sdk = Mock() |
| 57 | + provider = HybridFileSystemProvider(sdk, workspace="default") |
| 58 | + |
| 59 | + with patch("data_designer_nemo.fileset_filesystem_provider.FilesetFileSystem") as fs_class: |
| 60 | + context = provider.create_context(runtime_path=str(tmp_path)) |
| 61 | + provider.ensure_root_exists(runtime_path=str(tmp_path)) |
| 62 | + |
| 63 | + assert context.root_path == tmp_path.resolve() |
| 64 | + fs_class.assert_not_called() |
| 65 | + |
| 66 | + |
| 67 | +def test_hybrid_routes_non_local_path_to_fileset() -> None: |
| 68 | + sdk = Mock() |
| 69 | + provider = HybridFileSystemProvider(sdk, workspace="default") |
| 70 | + |
| 71 | + with patch("data_designer_nemo.fileset_filesystem_provider.FilesetFileSystem") as fs_class: |
| 72 | + fs_class.return_value.async_impl = True |
| 73 | + fs_class.return_value.asynchronous = False |
| 74 | + context = provider.create_context(runtime_path="docs#corpus") |
| 75 | + |
| 76 | + fs_class.assert_called_once_with(sdk) |
| 77 | + assert str(context.root_path) == "default/docs#corpus" |
| 78 | + |
| 79 | + |
| 80 | +def test_hybrid_ensure_root_exists_validates_fileset_for_non_local_path() -> None: |
| 81 | + sdk = Mock() |
| 82 | + provider = HybridFileSystemProvider(sdk, workspace="default") |
| 83 | + |
| 84 | + with patch("data_designer_nemo.fileset_filesystem_provider.FilesetFileSystem") as fs_class: |
| 85 | + fs_class.return_value.exists.side_effect = [False, True] |
| 86 | + |
| 87 | + with pytest.raises(SeedReaderConfigError, match="Path 'corpus' not found in fileset 'default/docs'"): |
| 88 | + provider.ensure_root_exists(runtime_path="docs#corpus") |
| 89 | + |
| 90 | + assert fs_class.return_value.exists.call_count == 2 |
| 91 | + |
| 92 | + |
| 93 | +def test_local_context_wires_hybrid_provider_into_filesystem_readers() -> None: |
| 94 | + readers = LocalDataDesignerContext(Mock(), "default").get_seed_readers() |
| 95 | + |
| 96 | + fs_readers = [r for r in readers if isinstance(r, DirectorySeedReader | FileContentsSeedReader)] |
| 97 | + assert len(fs_readers) == 2 |
| 98 | + assert all(isinstance(r._fs_provider, HybridFileSystemProvider) for r in fs_readers) |
0 commit comments