voice-chatbot 0.1.0__tar.gz

voice_chatbot-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Aapo
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

voice_chatbot-0.1.0/MANIFEST.in

@@ -0,0 +1,3 @@
+include LICENSE
+include README.md
+include requirements.txt

voice_chatbot-0.1.0/PKG-INFO

@@ -0,0 +1,274 @@
+Metadata-Version: 2.1
+Name: voice-chatbot
+Version: 0.1.0
+Summary: Local speech-to-speech voice assistant with PySide6 GUI, CLI, and ROS 2 integration.
+Author-email: Aapo <noreply@example.com>
+Maintainer: Aapo
+Maintainer-email: noreply@example.com
+License: MIT
+Project-URL: Homepage, https://github.com/Aapo2001/python-chatbot
+Project-URL: Documentation, https://docs-site-kappa-coral.vercel.app
+Project-URL: Repository, https://github.com/Aapo2001/python-chatbot
+Project-URL: Issues, https://github.com/Aapo2001/python-chatbot/issues
+Keywords: voice-assistant,chatbot,speech-to-text,text-to-speech,llm
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Multimedia :: Sound/Audio :: Speech
+Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: numpy
+Requires-Dist: sounddevice
+Requires-Dist: huggingface-hub
+Provides-Extra: test
+Requires-Dist: numpy; extra == "test"
+Requires-Dist: pytest<10,>=8; extra == "test"
+Requires-Dist: pytest-cov<8,>=5; extra == "test"
+Provides-Extra: stt
+Requires-Dist: faster-whisper; extra == "stt"
+Provides-Extra: llm
+Requires-Dist: llama-cpp-python; extra == "llm"
+Provides-Extra: tts
+Requires-Dist: coqui-tts[codec]; extra == "tts"
+Provides-Extra: vad
+Requires-Dist: silero-vad; extra == "vad"
+Provides-Extra: gui
+Requires-Dist: PySide6<7,>=6.7; extra == "gui"
+Provides-Extra: all
+Requires-Dist: voice-chatbot[gui,llm,stt,tts,vad]; extra == "all"
+Provides-Extra: dev
+Requires-Dist: voice-chatbot[all]; extra == "dev"
+Requires-Dist: pytest<10,>=9; extra == "dev"
+Requires-Dist: pytest-cov<8,>=5; extra == "dev"
+
+# Voice Chatbot
+
+Local voice chatbot for Windows with a Finnish-first default configuration. The application captures microphone audio, detects speech with Silero VAD, transcribes it with Whisper, generates a reply with a local GGUF LLM, and speaks the reply with Coqui TTS.
+
+The repository's application code now lives under the `voice_chatbot/`
+package. The repo root keeps thin compatibility wrappers (`app.py`,
+`chatbot.py`, `ros_app.py`, `setup_models.py`) so older commands still work.
+
+Primary entry points:
+
+- `python -m voice_chatbot.app`: PySide6 desktop UI for configuring and running the chatbot.
+- `python -m voice_chatbot.chatbot`: terminal-only runner with the same audio pipeline.
+- `python -m voice_chatbot.ros_app`: ROS-connected PySide6 GUI.
+- `voice_chatbot_ros/node.py`: ROS 2 Humble node that exposes the pipeline through ROS topics and a service.
+
+## What The Code Does
+
+Runtime flow:
+
+1. `AudioIO` captures 16 kHz mono microphone audio in fixed-size chunks.
+2. `VoiceActivityDetector` buffers audio until Silero VAD reports speech start and end.
+3. `SpeechToText` transcribes the captured utterance with `pywhispercpp`.
+4. `ChatLLM` sends the user text plus recent conversation history to `llama-cpp-python`.
+5. `TextToSpeech` synthesizes the assistant reply with Coqui TTS.
+6. `AudioIO` plays the generated speech back through the default output device.
+
+The GUI wraps this pipeline in a background `QThread` and exposes model and runtime settings in a sidebar.
+
+## Repository Layout
+
+| Path | Purpose |
+| --- | --- |
+| `voice_chatbot/` | Main Python application package for GUI, CLI, config, audio, VAD, STT, LLM, and TTS code |
+| `app.py`, `chatbot.py`, `ros_app.py`, `setup_models.py` | Thin compatibility entry points that call into `voice_chatbot/` |
+| `install.bat` | Windows setup script for Python packages and CUDA builds |
+| `pixi.toml` | Pixi workspace manifest for the base toolchain and common tasks |
+| `pixi.lock` | Resolved Pixi lockfile for the base environment |
+| `tools/install_python_windows.bat` | Pixi bootstrap script for CUDA-enabled Python packages |
+| `tools/install_python_linux.sh` | Pixi bootstrap script for Linux Python packages |
+| `config.json` | Persisted GUI configuration |
+| `voice_chatbot_ros/` | ROS 2 Humble package and node implementation |
+| `launch/` | ROS 2 launch file |
+| `package.xml`, `setup.py`, `setup.cfg` | ROS 2 `ament_python` package metadata |
+
+More implementation detail is in [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md).
+ROS-specific usage is in [docs/ROS2.md](docs/ROS2.md).
+
+## Environment Assumptions
+
+The codebase is currently optimized for a Windows workstation with local GPU inference:
+
+- `app.py` and `chatbot.py` add CUDA DLL paths from `CUDA_PATH` or `D:\cuda`.
+- `install.bat` installs CUDA-enabled PyTorch, `llama-cpp-python`, and `pywhispercpp`.
+- The default LLM is a GGUF file under `models/`.
+- The default voice assistant language is Finnish (`fi`).
+
+The code can fall back to CPU for some components, but the intended deployment is local GPU acceleration.
+
+## Installation
+
+The recommended setup path is now Pixi. The repository ships with a `pixi.toml` workspace manifest and uses Pixi to install both the base Python toolchain and the ROS 2 Humble packages.
+
+### 1. System prerequisites
+
+- Windows
+- `pixi` or permission to let `install.bat` install it
+- CUDA Toolkit if you want GPU acceleration for PyTorch and `llama-cpp-python`
+- CMake and a working build toolchain for Python packages with native extensions
+- Microphone and speakers/headphones configured as default audio devices
+
+### 2. Create the Pixi environment and install packages
+
+Run:
+
+```powershell
+install.bat
+```
+
+What the script does:
+
+- installs `pixi` if it is missing
+- creates or updates the local Pixi environment from `pixi.toml`
+- runs the Pixi bootstrap task that installs the Python packages needed by the project
+
+The Pixi workspace provides:
+
+- Python 3.11
+- `pip`
+- `cmake`
+- `git`
+- `ninja`
+- `colcon-common-extensions`
+- `setuptools>=69.5,<80`
+- `ros-humble-desktop`
+
+The bootstrap task installs:
+
+- CUDA-enabled `torch`, `torchvision`, `torchaudio`
+- `llama-cpp-python` compiled with `GGML_CUDA=on`
+- `pywhispercpp` compiled with CUDA flags
+- the remaining packages from `requirements.txt`
+
+It also enforces a `setuptools` version that stays compatible with both `sip`
+and ROS 2 Humble's `colcon` editable Python build flow.
+
+`requirements.txt` is still not the full environment by itself. `torch`, `llama-cpp-python`, and `pywhispercpp` are installed separately because they need a custom wheel index or CUDA-specific build flags.
+
+### 3. Direct Pixi workflow
+
+If you do not want to use `install.bat`, the equivalent commands are:
+
+```powershell
+pixi install
+pixi run install-python-deps
+```
+
+`pixi run build` also re-checks the `setuptools` compatibility window before
+invoking `colcon`.
+
+### 4. Download models
+
+Run:
+
+```powershell
+pixi run setup-models
+```
+
+This script:
+
+- checks CUDA visibility in PyTorch
+- initializes the Silero VAD model
+- downloads or validates the configured Whisper model
+- downloads the configured GGUF LLM from Hugging Face if missing
+- initializes the configured Coqui TTS model
+
+## Running The Application
+
+Desktop UI:
+
+```powershell
+pixi run app
+# equivalent: python -m voice_chatbot.app
+```
+
+Terminal mode:
+
+```powershell
+pixi run chatbot
+# equivalent: python -m voice_chatbot.chatbot
+```
+
+ROS 2 Humble node:
+
+```bash
+pixi run ros-run /absolute/path/to/config.json
+```
+
+This follows Pixi's ROS 2 workflow with `robostack-humble` packages installed into the Pixi environment.
+
+## Testing
+
+Run the automated unit test suite with:
+
+```powershell
+pixi run test
+```
+
+The repository also includes a GitHub Actions workflow that runs the same
+pytest suite automatically on every push and pull request on both Windows
+and Linux.
+
+## Configuration
+
+Configuration is defined in `config.py` and can be persisted to `config.json`.
+
+Important settings:
+
+- Audio: `sample_rate`, `channels`, `chunk_samples`
+- VAD: `vad_threshold`, `min_silence_duration_ms`, `speech_pad_ms`, `min_speech_duration_ms`, `vad_pre_buffer_ms`
+- STT: `language`, `whisper_model`, `whisper_n_threads`
+- LLM: `llm_model_path`, `llm_n_gpu_layers`, `llm_n_ctx`, `llm_max_tokens`, `llm_temperature`, `llm_system_prompt`, `max_conversation_turns`
+- TTS: `tts_model`, `tts_gpu`
+- Download metadata: `llm_repo_id`, `llm_filename`
+
+### Configuration behavior to know
+
+- The GUI loads from `config.json` through `Config.load()` and writes the current sidebar values back to disk when you start the worker.
+- The CLI and model setup script also load `config.json` by default, so they now use the same persisted settings as the GUI unless you edit the package code.
+
+## GUI Behavior
+
+The desktop app provides:
+
+- a settings sidebar for language, Whisper, LLM, TTS, and VAD parameters
+- a chat panel for user and assistant messages
+- a system log panel that receives redirected `stdout` and `stderr`
+- start, stop, restart, and clear-chat controls
+
+Behavior details from the current code:
+
+- Settings remain editable while the worker is running.
+- Changing settings during runtime requires `Käynnistä uudelleen` to rebuild the worker with the new values.
+- `Tyhjennä keskustelu` clears the visible chat panel only. It does not clear the LLM conversation history; history resets only when a new `ChatLLM` instance is created, such as after restart.
+
+## Architecture Notes
+
+- The VAD implementation keeps a rolling pre-buffer so the first syllables are not clipped before speech onset is confirmed.
+- After TTS playback, the app clears queued microphone chunks and resets VAD state to reduce the chance of transcribing its own synthesized output.
+- The LLM wrapper stores alternating user and assistant messages and trims the oldest turns once `max_conversation_turns` is exceeded.
+- The GUI runs model loading and the audio loop in `ChatbotWorker`, a `QThread`, so the main UI thread stays responsive.
+
+## Operational Limitations
+
+- There are no automated tests in the repository.
+- Audio device selection is not exposed; capture and playback use the default system devices through `sounddevice`.
+- The code and UI text are partly Finnish and partly English.
+- Model initialization happens synchronously inside the worker or CLI startup path, so startup cost depends on model size.
+- ROS 2 support assumes the Robostack Humble packages and the ML/audio dependencies can coexist in the same Pixi environment.
+
+## Suggested First Run
+
+1. Run `install.bat`.
+2. Run `pixi run setup-models`.
+3. Start `pixi run app`.
+4. Confirm the GGUF model path in the left sidebar.
+5. Click `Käynnistä` and watch the system log for CUDA and model load status.

voice_chatbot-0.1.0/README.md

@@ -0,0 +1,226 @@
+# Voice Chatbot
+
+Local voice chatbot for Windows with a Finnish-first default configuration. The application captures microphone audio, detects speech with Silero VAD, transcribes it with Whisper, generates a reply with a local GGUF LLM, and speaks the reply with Coqui TTS.
+
+The repository's application code now lives under the `voice_chatbot/`
+package. The repo root keeps thin compatibility wrappers (`app.py`,
+`chatbot.py`, `ros_app.py`, `setup_models.py`) so older commands still work.
+
+Primary entry points:
+
+- `python -m voice_chatbot.app`: PySide6 desktop UI for configuring and running the chatbot.
+- `python -m voice_chatbot.chatbot`: terminal-only runner with the same audio pipeline.
+- `python -m voice_chatbot.ros_app`: ROS-connected PySide6 GUI.
+- `voice_chatbot_ros/node.py`: ROS 2 Humble node that exposes the pipeline through ROS topics and a service.
+
+## What The Code Does
+
+Runtime flow:
+
+1. `AudioIO` captures 16 kHz mono microphone audio in fixed-size chunks.
+2. `VoiceActivityDetector` buffers audio until Silero VAD reports speech start and end.
+3. `SpeechToText` transcribes the captured utterance with `pywhispercpp`.
+4. `ChatLLM` sends the user text plus recent conversation history to `llama-cpp-python`.
+5. `TextToSpeech` synthesizes the assistant reply with Coqui TTS.
+6. `AudioIO` plays the generated speech back through the default output device.
+
+The GUI wraps this pipeline in a background `QThread` and exposes model and runtime settings in a sidebar.
+
+## Repository Layout
+
+| Path | Purpose |
+| --- | --- |
+| `voice_chatbot/` | Main Python application package for GUI, CLI, config, audio, VAD, STT, LLM, and TTS code |
+| `app.py`, `chatbot.py`, `ros_app.py`, `setup_models.py` | Thin compatibility entry points that call into `voice_chatbot/` |
+| `install.bat` | Windows setup script for Python packages and CUDA builds |
+| `pixi.toml` | Pixi workspace manifest for the base toolchain and common tasks |
+| `pixi.lock` | Resolved Pixi lockfile for the base environment |
+| `tools/install_python_windows.bat` | Pixi bootstrap script for CUDA-enabled Python packages |
+| `tools/install_python_linux.sh` | Pixi bootstrap script for Linux Python packages |
+| `config.json` | Persisted GUI configuration |
+| `voice_chatbot_ros/` | ROS 2 Humble package and node implementation |
+| `launch/` | ROS 2 launch file |
+| `package.xml`, `setup.py`, `setup.cfg` | ROS 2 `ament_python` package metadata |
+
+More implementation detail is in [docs/ARCHITECTURE.md](docs/ARCHITECTURE.md).
+ROS-specific usage is in [docs/ROS2.md](docs/ROS2.md).
+
+## Environment Assumptions
+
+The codebase is currently optimized for a Windows workstation with local GPU inference:
+
+- `app.py` and `chatbot.py` add CUDA DLL paths from `CUDA_PATH` or `D:\cuda`.
+- `install.bat` installs CUDA-enabled PyTorch, `llama-cpp-python`, and `pywhispercpp`.
+- The default LLM is a GGUF file under `models/`.
+- The default voice assistant language is Finnish (`fi`).
+
+The code can fall back to CPU for some components, but the intended deployment is local GPU acceleration.
+
+## Installation
+
+The recommended setup path is now Pixi. The repository ships with a `pixi.toml` workspace manifest and uses Pixi to install both the base Python toolchain and the ROS 2 Humble packages.
+
+### 1. System prerequisites
+
+- Windows
+- `pixi` or permission to let `install.bat` install it
+- CUDA Toolkit if you want GPU acceleration for PyTorch and `llama-cpp-python`
+- CMake and a working build toolchain for Python packages with native extensions
+- Microphone and speakers/headphones configured as default audio devices
+
+### 2. Create the Pixi environment and install packages
+
+Run:
+
+```powershell
+install.bat
+```
+
+What the script does:
+
+- installs `pixi` if it is missing
+- creates or updates the local Pixi environment from `pixi.toml`
+- runs the Pixi bootstrap task that installs the Python packages needed by the project
+
+The Pixi workspace provides:
+
+- Python 3.11
+- `pip`
+- `cmake`
+- `git`
+- `ninja`
+- `colcon-common-extensions`
+- `setuptools>=69.5,<80`
+- `ros-humble-desktop`
+
+The bootstrap task installs:
+
+- CUDA-enabled `torch`, `torchvision`, `torchaudio`
+- `llama-cpp-python` compiled with `GGML_CUDA=on`
+- `pywhispercpp` compiled with CUDA flags
+- the remaining packages from `requirements.txt`
+
+It also enforces a `setuptools` version that stays compatible with both `sip`
+and ROS 2 Humble's `colcon` editable Python build flow.
+
+`requirements.txt` is still not the full environment by itself. `torch`, `llama-cpp-python`, and `pywhispercpp` are installed separately because they need a custom wheel index or CUDA-specific build flags.
+
+### 3. Direct Pixi workflow
+
+If you do not want to use `install.bat`, the equivalent commands are:
+
+```powershell
+pixi install
+pixi run install-python-deps
+```
+
+`pixi run build` also re-checks the `setuptools` compatibility window before
+invoking `colcon`.
+
+### 4. Download models
+
+Run:
+
+```powershell
+pixi run setup-models
+```
+
+This script:
+
+- checks CUDA visibility in PyTorch
+- initializes the Silero VAD model
+- downloads or validates the configured Whisper model
+- downloads the configured GGUF LLM from Hugging Face if missing
+- initializes the configured Coqui TTS model
+
+## Running The Application
+
+Desktop UI:
+
+```powershell
+pixi run app
+# equivalent: python -m voice_chatbot.app
+```
+
+Terminal mode:
+
+```powershell
+pixi run chatbot
+# equivalent: python -m voice_chatbot.chatbot
+```
+
+ROS 2 Humble node:
+
+```bash
+pixi run ros-run /absolute/path/to/config.json
+```
+
+This follows Pixi's ROS 2 workflow with `robostack-humble` packages installed into the Pixi environment.
+
+## Testing
+
+Run the automated unit test suite with:
+
+```powershell
+pixi run test
+```
+
+The repository also includes a GitHub Actions workflow that runs the same
+pytest suite automatically on every push and pull request on both Windows
+and Linux.
+
+## Configuration
+
+Configuration is defined in `config.py` and can be persisted to `config.json`.
+
+Important settings:
+
+- Audio: `sample_rate`, `channels`, `chunk_samples`
+- VAD: `vad_threshold`, `min_silence_duration_ms`, `speech_pad_ms`, `min_speech_duration_ms`, `vad_pre_buffer_ms`
+- STT: `language`, `whisper_model`, `whisper_n_threads`
+- LLM: `llm_model_path`, `llm_n_gpu_layers`, `llm_n_ctx`, `llm_max_tokens`, `llm_temperature`, `llm_system_prompt`, `max_conversation_turns`
+- TTS: `tts_model`, `tts_gpu`
+- Download metadata: `llm_repo_id`, `llm_filename`
+
+### Configuration behavior to know
+
+- The GUI loads from `config.json` through `Config.load()` and writes the current sidebar values back to disk when you start the worker.
+- The CLI and model setup script also load `config.json` by default, so they now use the same persisted settings as the GUI unless you edit the package code.
+
+## GUI Behavior
+
+The desktop app provides:
+
+- a settings sidebar for language, Whisper, LLM, TTS, and VAD parameters
+- a chat panel for user and assistant messages
+- a system log panel that receives redirected `stdout` and `stderr`
+- start, stop, restart, and clear-chat controls
+
+Behavior details from the current code:
+
+- Settings remain editable while the worker is running.
+- Changing settings during runtime requires `Käynnistä uudelleen` to rebuild the worker with the new values.
+- `Tyhjennä keskustelu` clears the visible chat panel only. It does not clear the LLM conversation history; history resets only when a new `ChatLLM` instance is created, such as after restart.
+
+## Architecture Notes
+
+- The VAD implementation keeps a rolling pre-buffer so the first syllables are not clipped before speech onset is confirmed.
+- After TTS playback, the app clears queued microphone chunks and resets VAD state to reduce the chance of transcribing its own synthesized output.
+- The LLM wrapper stores alternating user and assistant messages and trims the oldest turns once `max_conversation_turns` is exceeded.
+- The GUI runs model loading and the audio loop in `ChatbotWorker`, a `QThread`, so the main UI thread stays responsive.
+
+## Operational Limitations
+
+- There are no automated tests in the repository.
+- Audio device selection is not exposed; capture and playback use the default system devices through `sounddevice`.
+- The code and UI text are partly Finnish and partly English.
+- Model initialization happens synchronously inside the worker or CLI startup path, so startup cost depends on model size.
+- ROS 2 support assumes the Robostack Humble packages and the ML/audio dependencies can coexist in the same Pixi environment.
+
+## Suggested First Run
+
+1. Run `install.bat`.
+2. Run `pixi run setup-models`.
+3. Start `pixi run app`.
+4. Confirm the GGUF model path in the left sidebar.
+5. Click `Käynnistä` and watch the system log for CUDA and model load status.

voice_chatbot-0.1.0/launch/voice_chatbot.launch.py

@@ -0,0 +1,71 @@
+"""
+ROS 2 launch file for the split voice chatbot nodes.
+
+Launches three nodes in the ``/voice_chatbot`` namespace:
+
+- ``voice_stt`` – microphone capture, VAD, Whisper STT
+- ``voice_llm`` – LLM chat inference (LLaMA/GGUF)
+- ``voice_tts`` – Coqui TTS synthesis + audio playback
+
+All three share the same ``config_path`` and ``load_config_file``
+parameters so they read the same ``config.json``.
+
+Usage::
+
+    pixi run ros-launch
+    # or: ros2 launch voice_chatbot_ros voice_chatbot.launch.py
+"""
+
+from launch.actions import DeclareLaunchArgument
+from launch.substitutions import LaunchConfiguration
+from launch_ros.actions import Node
+from launch_ros.parameter_descriptions import ParameterValue
+
+from launch import LaunchDescription
+
+
+def generate_launch_description() -> LaunchDescription:
+    """Build the launch description with shared parameters."""
+    config_path = LaunchConfiguration("config_path")
+    load_config_file = LaunchConfiguration("load_config_file")
+
+    shared_params = [
+        {
+            "config_path": ParameterValue(config_path, value_type=str),
+            "load_config_file": ParameterValue(load_config_file, value_type=bool),
+        }
+    ]
+
+    return LaunchDescription(
+        [
+            DeclareLaunchArgument("config_path", default_value="config.json"),
+            DeclareLaunchArgument("load_config_file", default_value="true"),
+            # STT node: microphone capture, VAD, speech-to-text
+            Node(
+                package="voice_chatbot_ros",
+                executable="voice_stt_node",
+                name="voice_stt",
+                namespace="voice_chatbot",
+                output="screen",
+                parameters=shared_params,
+            ),
+            # LLM node: chat inference
+            Node(
+                package="voice_chatbot_ros",
+                executable="voice_llm_node",
+                name="voice_llm",
+                namespace="voice_chatbot",
+                output="screen",
+                parameters=shared_params,
+            ),
+            # TTS node: text-to-speech synthesis and audio playback
+            Node(
+                package="voice_chatbot_ros",
+                executable="voice_tts_node",
+                name="voice_tts",
+                namespace="voice_chatbot",
+                output="screen",
+                parameters=shared_params,
+            ),
+        ]
+    )

voice_chatbot-0.1.0/package.xml

@@ -0,0 +1,21 @@
+<?xml version="1.0"?>
+<package format="3">
+  <name>voice_chatbot_ros</name>
+  <version>0.1.0</version>
+  <description>ROS 2 Humble integration for the local voice chatbot pipeline.</description>
+
+  <maintainer email="noreply@example.com">OpenAI Codex</maintainer>
+  <license>Proprietary</license>
+
+  <buildtool_depend>ament_python</buildtool_depend>
+
+  <exec_depend>rclpy</exec_depend>
+  <exec_depend>std_msgs</exec_depend>
+  <exec_depend>std_srvs</exec_depend>
+  <exec_depend>launch</exec_depend>
+  <exec_depend>launch_ros</exec_depend>
+
+  <export>
+    <build_type>ament_python</build_type>
+  </export>
+</package>

voice_chatbot-0.1.0/pyproject.toml

@@ -0,0 +1,74 @@
+[build-system]
+requires = ["setuptools>=69.5"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "voice-chatbot"
+version = "0.1.0"
+description = "Local speech-to-speech voice assistant with PySide6 GUI, CLI, and ROS 2 integration."
+readme = "README.md"
+license = {text = "MIT"}
+requires-python = ">=3.11"
+authors = [
+    { name = "Aapo", email = "noreply@example.com" },
+]
+keywords = ["voice-assistant", "chatbot", "speech-to-text", "text-to-speech", "llm"]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Developers",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Topic :: Multimedia :: Sound/Audio :: Speech",
+    "Topic :: Scientific/Engineering :: Artificial Intelligence",
+]
+dependencies = [
+    "numpy",
+    "sounddevice",
+    "huggingface-hub",
+]
+
+[project.optional-dependencies]
+stt = [
+    "faster-whisper",
+]
+llm = [
+    "llama-cpp-python",
+]
+tts = [
+    "coqui-tts[codec]",
+]
+vad = [
+    "silero-vad",
+]
+gui = [
+    "PySide6>=6.7,<7",
+]
+all = [
+    "voice-chatbot[stt,llm,tts,vad,gui]",
+]
+dev = [
+    "voice-chatbot[all]",
+    "pytest>=9,<10",
+    "pytest-cov>=5,<8",
+]
+
+[project.scripts]
+voice-chatbot = "voice_chatbot.chatbot:main"
+voice-chatbot-app = "voice_chatbot.app:main"
+voice-chatbot-setup-models = "voice_chatbot.setup_models:main"
+
+[project.urls]
+Homepage = "https://github.com/Aapo2001/python-chatbot"
+Documentation = "https://docs-site-kappa-coral.vercel.app"
+Repository = "https://github.com/Aapo2001/python-chatbot"
+Issues = "https://github.com/Aapo2001/python-chatbot/issues"
+
+[tool.setuptools.packages.find]
+include = ["voice_chatbot", "voice_chatbot.*"]
+exclude = ["voice_chatbot_ros", "voice_chatbot_ros.*", "tests", "tests.*"]
+
+[tool.pytest.ini_options]
+testpaths = ["tests"]
+addopts = "-ra"