feat: Refactor audio utility functions and update sound services to use shared methods

app/utils/audio.py

@@ -0,0 +1,35 @@
+"""Audio file utility functions shared across audio processing services."""
+
+import hashlib
+from pathlib import Path
+
+import ffmpeg  # type: ignore[import-untyped]
+
+from app.core.logging import get_logger
+
+logger = get_logger(__name__)
+
+
+def get_file_hash(file_path: Path) -> str:
+    """Calculate SHA-256 hash of a file."""
+    hash_sha256 = hashlib.sha256()
+    with file_path.open("rb") as f:
+        for chunk in iter(lambda: f.read(4096), b""):
+            hash_sha256.update(chunk)
+    return hash_sha256.hexdigest()
+
+
+def get_file_size(file_path: Path) -> int:
+    """Get file size in bytes."""
+    return file_path.stat().st_size
+
+
+def get_audio_duration(file_path: Path) -> int:
+    """Get audio duration in milliseconds using ffmpeg."""
+    try:
+        probe = ffmpeg.probe(str(file_path))
+        duration = float(probe["format"]["duration"])
+        return int(duration * 1000)  # Convert to milliseconds
+    except Exception as e:
+        logger.warning("Failed to get duration for %s: %s", file_path, e)
+        return 0