Metadata-Version: 2.4
Name: speakinpy
Version: 0.1.2
Summary: A Python module for recorded audio transcription via Gemini API
Requires-Python: >=3.11
Description-Content-Type: text/markdown
License-File: LICENSE
Requires-Dist: google-genai>=2.33.0
Dynamic: license-file

# SpeakINPy
A python module for transcribing audio files to text via Google GenAI

## Installation
```bash
pip install SpeakINPy
```
or via UV:
```bash
uv add SpeakINPy
```

## Usage
```python
from speakinpy import transcribe
import os

GEMINI_API_KEY = os.environ.get("GEMINI_API_KEY")
audio_path = "path/to/audio.m4a"

text = transcribe(audio_path, GEMINI_API_KEY)
print(text)
```

## Release / PyPI publishing

1. Build artifacts locally:
   ```bash
   python -m pip install --upgrade build twine
   python -m build
   python -m twine check dist/*
   ```
2. Validate artifacts:
   - inspect `dist/*.whl` and `dist/*.tar.gz`
   - install the wheel in a clean virtual environment (`pip install --no-deps dist/*.whl`)
3. Configure Trusted Publishing in both PyPI and TestPyPI for this repository and workflow:
   - workflow: `.github/workflows/publish.yml`
   - environment: `pypi` (production) and optional `testpypi`
4. Create and publish a GitHub release/tag from `main`.
   - release publication triggers automatic PyPI publish via OIDC (no API token required)
   - optional manual `workflow_dispatch` can publish to TestPyPI

The `examples/` directory (including sample audio files) is intentionally excluded from published distributions.
