| """Tests for data models and IO helpers.""" |
|
|
| from __future__ import annotations |
|
|
| from pathlib import Path |
|
|
| import pytest |
|
|
| from video_vec2wav2_tokenizer.utils.io import ( |
| DatasetEntry, |
| Segment, |
| clip_index_name, |
| iter_video_files, |
| read_jsonl, |
| write_jsonl, |
| ) |
|
|
|
|
| def test_segment_roundtrip() -> None: |
| seg = Segment(start=0.0, end=2.5, text="example sentence") |
| assert seg.duration == 2.5 |
| assert seg.to_dict() == {"start": 0.0, "end": 2.5, "text": "example sentence"} |
| assert Segment.from_dict(seg.to_dict()) == seg |
|
|
|
|
| def test_dataset_entry_jsonl_shape() -> None: |
| entry = DatasetEntry(audio="audio/000001.wav", text="Hello world", duration=2.5) |
| row = entry.to_jsonl_dict() |
| assert row == {"audio": "audio/000001.wav", "text": "Hello world", "duration": 2.5} |
|
|
|
|
| def test_dataset_entry_with_speaker() -> None: |
| entry = DatasetEntry( |
| audio="audio/000001.wav", text="hi", duration=1.0, speaker="speaker_002" |
| ) |
| assert entry.to_jsonl_dict()["speaker"] == "speaker_002" |
|
|
|
|
| def test_clip_index_name() -> None: |
| assert clip_index_name(1) == "000001.wav" |
| assert clip_index_name(42) == "000042.wav" |
|
|
|
|
| def test_jsonl_roundtrip(tmp_path: Path) -> None: |
| rows = [{"a": 1}, {"b": "x"}] |
| path = tmp_path / "out.jsonl" |
| assert write_jsonl(path, rows) == 2 |
| assert list(read_jsonl(path)) == rows |
|
|
|
|
| def test_iter_video_files(tmp_path: Path) -> None: |
| (tmp_path / "a.mp4").write_bytes(b"") |
| (tmp_path / "sub").mkdir() |
| (tmp_path / "sub" / "b.mkv").write_bytes(b"") |
| (tmp_path / "note.txt").write_bytes(b"") |
| found = [p.name for p in iter_video_files(tmp_path)] |
| assert found == ["a.mp4", "b.mkv"] |
|
|
|
|
| def test_iter_video_files_missing(tmp_path: Path) -> None: |
| with pytest.raises(FileNotFoundError): |
| list(iter_video_files(tmp_path / "nope")) |
|
|
