make pydub an optional import

README.md

@@ -30,6 +30,12 @@ pip install git+https://github.com/pathintegral-institute/markitup.git@main#subd
 uv add git+https://github.com/pathintegral-institute/markitup.git@main#subdirectory=packages/markitup
 ```
 
+To use audio transciption using `pydub`, install `markitup[audio]`:
+```bash
+uv add "git+https://github.com/pathintegral-institute/markitup.git@main#subdirectory=packages/markitup[audio]"
+```
+
+
 ## Usage
 ```python
 from markitup.converter_utils.utils import read_files_to_bytestreams

packages/markitup/pyproject.toml

@@ -34,13 +34,22 @@ dependencies = [
   "pandas",
   "lxml",
   "olefile",
-  "pydub",
+  # "pydub",  <- Removed from here
   "SpeechRecognition",
   "pymupdf>=1.25.5",
   "openpyxl>=3.1.5",
   "xlrd>=2.0.1",
 ]
 
+[project.optional-dependencies]
+audio = [
+  "pydub",
+]
+# Optional: You may want to create an 'all' extra that includes all optional dependencies
+all = [
+  "pydub",
+]
+
 [tool.hatch.version]
 path = "src/markitup/__about__.py"
 
@@ -84,4 +93,4 @@ exclude_lines = [
 ]
 
 [tool.hatch.build.targets.sdist]
-only-include = ["src/markitup"]
+only-include = ["src/markitup"]

packages/markitup/src/markitup/converter_utils/utils.py

@@ -1,9 +1,6 @@
 import os
 from io import BytesIO
-from markitup._schemas import StreamInfo
-import magic
 import speech_recognition as sr
-import pydub
 import io
 from typing import BinaryIO
 
@@ -47,6 +44,7 @@ def transcribe_audio(file_stream: BinaryIO, *, magic_type: str = "audio/mpeg") -
 
     match audio_format:
         case 'mp3':
+            import pydub
             audio_segment = pydub.AudioSegment.from_file(
                 file_stream, format=audio_format)
             audio_source = io.BytesIO()