Apply Ruff formatting to audio path changes

Signed-off-by: Dongji Gao <dongjig@nvidia.com>

Author: DongjiGao

nemo_skills/dataset/librispeech-pc/prepare.py

@@ -227,7 +227,9 @@ def main():
     download_manifests(data_dir)
 
     splits = ["test-clean", "test-other"] if args.split == "all" else [args.split]
-    total = sum(process_split(split, data_dir, audio_dir, not args.no_audio, audio_root=audio_root) for split in splits)
+    total = sum(
+        process_split(split, data_dir, audio_dir, not args.no_audio, audio_root=audio_root) for split in splits
+    )
 
     print(f"\n✓ Complete: {total} samples")
 

nemo_skills/dataset/numb3rs/prepare.py

@@ -123,7 +123,9 @@ def save_audio_and_format_entry(
         audio_file_path = audio_dir / audio_filename
         sf.write(str(audio_file_path), audio_array, sampling_rate)
 
-    audio_filepath = build_container_audio_path("numb3rs", "Numb3rs", category, audio_filename, audio_prefix=audio_root)
+    audio_filepath = build_container_audio_path(
+        "numb3rs", "Numb3rs", category, audio_filename, audio_prefix=audio_root
+    )
 
     # Build audio metadata (to be embedded in messages later)
     audio_metadata = {
@@ -288,9 +290,7 @@ def main():
     # Process each category
     total_samples = 0
     for category in categories_to_prepare:
-        total_samples += prepare_category(
-            category, dataset, output_dir, with_audio=with_audio, audio_root=audio_root
-        )
+        total_samples += prepare_category(category, dataset, output_dir, with_audio=with_audio, audio_root=audio_root)
 
     # Combine all category variant files into test variant files
     print("\nCreating combined test files for each variant...")

tests/test_audio_path_prefix.py

@@ -110,10 +110,7 @@ def test_asr_leaderboard_audio_prefix(monkeypatch, tmp_path):
         audio_root="/data",
     )
 
-    assert (
-        formatted["messages"][1]["audio"]["path"]
-        == "/data/asr-leaderboard/data/librispeech_clean/sample_001.flac"
-    )
+    assert formatted["messages"][1]["audio"]["path"] == "/data/asr-leaderboard/data/librispeech_clean/sample_001.flac"
     _assert_clean_audio_paths(formatted, "/data/asr-leaderboard/")
 
 
@@ -136,9 +133,7 @@ def test_asr_leaderboard_end_to_end(monkeypatch, tmp_path):
     ]
     monkeypatch.setattr(prepare, "load_dataset", lambda *args, **kwargs: fake_dataset)
 
-    count = prepare.prepare_dataset(
-        "librispeech_clean", tmp_path, with_audio=False, audio_root="/data"
-    )
+    count = prepare.prepare_dataset("librispeech_clean", tmp_path, with_audio=False, audio_root="/data")
 
     assert count == 2
     _assert_clean_audio_paths(tmp_path / "librispeech_clean.jsonl", "/data/asr-leaderboard/")