Skip to content

armgit / dspy

Public Imported from GitHub
0

DSPy: The framework for programming—not prompting—language models

Python

fix(audio): Normalize 'x-wav' audio format to 'wav' (#9017)

akshatvishub3c6350cce45
diff --git a/dspy/adapters/types/audio.py b/dspy/adapters/types/audio.py
index f39d60bda..0ceb734b7 100644
--- a/dspy/adapters/types/audio.py
+++ b/dspy/adapters/types/audio.py
@@ -17,6 +17,11 @@ except ImportError:
SF_AVAILABLE = False
+def _normalize_audio_format(audio_format: str) -> str:
+ """Removes 'x-' prefixes from audio format strings."""
+ return audio_format.removeprefix("x-")
+
+
class Audio(Type):
data: str
audio_format: str
@@ -61,6 +66,9 @@ class Audio(Type):
if not mime_type.startswith("audio/"):
raise ValueError(f"Unsupported MIME type for audio: {mime_type}")
audio_format = mime_type.split("/")[1]
+
+ audio_format = _normalize_audio_format(audio_format)
+
encoded_data = base64.b64encode(response.content).decode("utf-8")
return cls(data=encoded_data, audio_format=audio_format)
@@ -80,6 +88,9 @@ class Audio(Type):
file_data = file.read()
audio_format = mime_type.split("/")[1]
+
+ audio_format = _normalize_audio_format(audio_format)
+
encoded_data = base64.b64encode(file_data).decode("utf-8")
return cls(data=encoded_data, audio_format=audio_format)
@@ -126,6 +137,9 @@ def encode_audio(audio: Union[str, bytes, dict, "Audio", Any], sampling_rate: in
header, b64data = audio.split(",", 1)
mime = header.split(";")[0].split(":")[1]
audio_format = mime.split("/")[1]
+
+ audio_format = _normalize_audio_format(audio_format)
+
return {"data": b64data, "audio_format": audio_format}
except Exception as e:
raise ValueError(f"Malformed audio data URI: {e}")
diff --git a/tests/adapters/test_audio.py b/tests/adapters/test_audio.py
new file mode 100644
index 000000000..b7597fbd2
--- /dev/null
+++ b/tests/adapters/test_audio.py
@@ -0,0 +1,32 @@
+import pytest
+
+from dspy.adapters.types.audio import _normalize_audio_format
+
+
+ "input_format, expected_format",
+ [
+ # Case 1: Standard format (no change)
+ ("wav", "wav"),
+ ("mp3", "mp3"),
+
+ # Case 2: The 'x-' prefix
+ ("x-wav", "wav"),
+ ("x-mp3", "mp3"),
+ ("x-flac", "flac"),
+
+ # Case 3: The edge case
+ ("my-x-format", "my-x-format"),
+ ("x-my-format", "my-format"),
+
+ # Case 4: Empty string and edge cases
+ ("", ""),
+ ("x-", ""),
+ ],
+)
+def test_normalize_audio_format(input_format, expected_format):
+ """
+ Tests that the _normalize_audio_format helper correctly removes 'x-' prefixes.
+ This single test covers the logic for from_url, from_file, and encode_audio.
+ """
+ assert _normalize_audio_format(input_format) == expected_format