Merge pull request #166 from tiqi-group/fix/logging_settings

fix: removes basic configuration of logging system in task module

.github/workflows/publish-to-pypi.yaml

@@ -70,9 +70,9 @@ jobs:
         name: python-package-distributions
         path: dist/
     - name: Sign the dists with Sigstore
-      uses: sigstore/gh-action-sigstore-python@v1.2.3
+      uses: sigstore/gh-action-sigstore-python@v3.0.0
       with:
-        inputs: >-
+        inputs: |
           ./dist/*.tar.gz
           ./dist/*.whl
     - name: Upload artifact signatures to GitHub Release
@@ -85,27 +85,3 @@ jobs:
         gh release upload
         '${{ github.ref_name }}' dist/**
         --repo '${{ github.repository }}'
-
-  # publish-to-testpypi:
-  #   name: Publish Python 🐍 distribution 📦 to TestPyPI
-  #   needs:
-  #   - build
-  #   runs-on: ubuntu-latest
-  #
-  #   environment:
-  #     name: testpypi
-  #     url: https://test.pypi.org/p/pydase
-  #
-  #   permissions:
-  #     id-token: write  # IMPORTANT: mandatory for trusted publishing
-  #
-  #   steps:
-  #   - name: Download all the dists
-  #     uses: actions/download-artifact@v3
-  #     with:
-  #       name: python-package-distributions
-  #       path: dist/
-  #   - name: Publish distribution 📦 to TestPyPI
-  #     uses: pypa/gh-action-pypi-publish@release/v1
-  #     with:
-  #       repository-url: https://test.pypi.org/legacy/

.github/workflows/python-package.yml

@@ -20,9 +20,6 @@ jobs:
 
     steps:
     - uses: actions/checkout@v4
-    - uses: chartboost/ruff-action@v1
-      with:
-        src: "./src"
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v4
       with:
@@ -32,6 +29,12 @@ jobs:
         python -m pip install --upgrade pip
         python -m pip install poetry
         poetry install --with dev
+    - name: Check with ruff
+      run: |
+        poetry run ruff check src
+    - name: Check formatting with ruff
+      run: |
+        poetry run ruff format --check src
     - name: Test with pytest
       run: |
         poetry run pytest

LICENSE

@@ -1,6 +1,4 @@
-MIT License
+Copyright (c) 2023-2024 Mose Müller <mosemueller@gmail.com>
-
-Copyright (c) 2023 Mose Müller <mosemueller@gmail.com>
 
 Permission is hereby granted, free of charge, to any person obtaining a copy
 of this software and associated documentation files (the "Software"), to deal

README.md

@@ -1,62 +1,34 @@
+<!--introduction-start-->
 # pydase <!-- omit from toc -->
 
-[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
+[![Version](https://img.shields.io/pypi/v/pydase?style=flat)](https://pypi.org/project/pydase/)
-[![Documentation Status](https://readthedocs.org/projects/pydase/badge/?version=latest)](https://pydase.readthedocs.io/en/latest/?badge=latest)
+[![Python Versions](https://img.shields.io/pypi/pyversions/pydase)](https://pypi.org/project/pydase/)
+[![Documentation Status](https://readthedocs.org/projects/pydase/badge/?version=stable)](https://pydase.readthedocs.io/en/stable/)
+[![License: MIT](https://img.shields.io/github/license/tiqi-group/pydase)][License]
 
-`pydase` is a Python library designed to streamline the creation of services that interface with devices and data. It offers a unified API, simplifying the process of data querying and device interaction. Whether you're managing lab sensors, network devices, or any abstract data entity, `pydase` facilitates rapid service development and deployment.
+`pydase` is a Python library that simplifies the creation of remote control interfaces for Python objects. It exposes the public attributes of a user-defined class via a [Socket.IO](https://python-socketio.readthedocs.io/en/stable/) web server, ensuring they are always in sync with the service state. You can interact with these attributes using an RPC client, a RESTful API, or a web browser. The web browser frontend is auto-generated, displaying components that correspond to each public attribute of the class for direct interaction.
+`pydase` implements an [observer pattern][Observer Pattern] to provide the real-time updates, ensuring that changes to the class attributes are reflected across all clients.
 
-- [Features](#features)
+Whether you're managing lab sensors, network devices, or any abstract data entity, `pydase` facilitates service development and deployment.
-- [Installation](#installation)
-- [Usage](#usage)
-  - [Defining a DataService](#defining-a-dataservice)
-  - [Running the Server](#running-the-server)
-  - [Accessing the Web Interface](#accessing-the-web-interface)
-  - [Connecting to the Service via Python Client](#connecting-to-the-service-via-python-client)
-    - [Tab Completion Support](#tab-completion-support)
-    - [Integration within Another Service](#integration-within-another-service)
-  - [RESTful API](#restful-api)
-- [Understanding the Component System](#understanding-the-component-system)
-  - [Built-in Type and Enum Components](#built-in-type-and-enum-components)
-  - [Method Components](#method-components)
-  - [DataService Instances (Nested Classes)](#dataservice-instances-nested-classes)
-  - [Custom Components (`pydase.components`)](#custom-components-pydasecomponents)
-    - [`DeviceConnection`](#deviceconnection)
-      - [Customizing Connection Logic](#customizing-connection-logic)
-      - [Reconnection Interval](#reconnection-interval)
-    - [`Image`](#image)
-    - [`NumberSlider`](#numberslider)
-    - [`ColouredEnum`](#colouredenum)
-    - [Extending with New Components](#extending-with-new-components)
-- [Understanding Service Persistence](#understanding-service-persistence)
-  - [Controlling Property State Loading with `@load_state`](#controlling-property-state-loading-with-load_state)
-- [Understanding Tasks in pydase](#understanding-tasks-in-pydase)
-- [Understanding Units in pydase](#understanding-units-in-pydase)
-- [Using `validate_set` to Validate Property Setters](#using-validate_set-to-validate-property-setters)
-- [Configuring pydase via Environment Variables](#configuring-pydase-via-environment-variables)
-- [Customizing the Web Interface](#customizing-the-web-interface)
-- [Logging in pydase](#logging-in-pydase)
-  - [Changing the Log Level](#changing-the-log-level)
-- [Documentation](#documentation)
-- [Contributing](#contributing)
-- [License](#license)
 
 ## Features
 
 <!-- no toc -->
-- [Simple service definition through class-based interface](#defining-a-dataService)
+- [Simple service definition through class-based interface][Defining DataService]
-- [Integrated web interface for interactive access and control of your service](#accessing-the-web-interface)
+- [Auto-generated web interface for interactive access and control of your service][Web Interface Access]
-- [Support for programmatic control and interaction with your service](#connecting-to-the-service-via-python-client)
+- [Python RPC client][Short RPC Client]
-- [Component system bridging Python backend with frontend visual representation](#understanding-the-component-system)
+- [Customizable web interface][Customizing Web Interface]
-- [Customizable styling for the web interface](#customizing-web-interface-style)
+- [Saving and restoring the service state][Service Persistence]
-- [Saving and restoring the service state for service persistence](#understanding-service-persistence)
+- [Automated task management with built-in start/stop controls and optional autostart][Task Management]
-- [Automated task management with built-in start/stop controls and optional autostart](#understanding-tasks-in-pydase)
+- [Support for units][Units]
-- [Support for units](#understanding-units-in-pydase)
+- [Validating Property Setters][Property Validation]
-- [Validating Property Setters](#using-validate_set-to-validate-property-setters)
+
-<!-- Support for additional servers for specific use-cases -->
+<!--introduction-end-->
+
+<!--getting-started-start-->
 
 ## Installation
 
-<!--installation-start-->
 
 Install `pydase` using [`poetry`](https://python-poetry.org/):
 
@@ -70,26 +42,27 @@ or `pip`:
 pip install pydase
 ```
 
-<!--installation-end-->
 
 ## Usage
 
-<!--usage-start-->
 
-Using `pydase` involves three main steps: defining a `DataService` subclass, running the server, and then connecting to the service either programmatically using `pydase.Client` or through the web interface.
+Using `pydase` involves three main steps: defining a `pydase.DataService` subclass, running the server, and then connecting to the service either programmatically using `pydase.Client` or through the web interface.
 
 ### Defining a DataService
 
-To use pydase, you'll first need to create a class that inherits from `DataService`. This class represents your custom data service, which will be exposed via a web server. Your class can implement class / instance attributes and synchronous and asynchronous tasks.
+To use `pydase`, you'll first need to create a class that inherits from `pydase.DataService`.
+This class represents your custom service, which will be exposed via a web server.<br>
+Your class can implement synchronous and asynchronous methods, some [built-in types](https://docs.python.org/3/library/stdtypes.html) (like `int`, `float`, `str`, `bool`, `list` or `dict`) and [other components][Custom Components] as attributes.
+For more information, please refer to the [components guide][Components].
 
 Here's an example:
 
 ```python
-from pydase import DataService, Server
+import pydase
 from pydase.utils.decorators import frontend
 
 
-class Device(DataService):
+class Device(pydase.DataService):
     _current = 0.0
     _voltage = 0.0
     _power = False
@@ -132,26 +105,30 @@ class Device(DataService):
 
 if __name__ == "__main__":
     service = Device()
-    Server(service).run()
+    pydase.Server(service=service, web_port=8001).run()
 ```
 
-In the above example, we define a Device class that extends DataService. We define a few properties (current, voltage, power) and their getter and setter methods.
+In the above example, we define a `Device` class that inherits from `pydase.DataService`. 
+We define a few properties (current, voltage, power) and their getter and setter methods.
 
 ### Running the Server
 
-Once your DataService is defined, you can create an instance of it and run the server:
+Once your service class is defined, you can create an instance of it and run the server:
 
 ```python
-from pydase import Server
+import pydase
 
 # ... defining the Device class ...
 
 if __name__ == "__main__":
     service = Device()
-    Server(service).run()
+    pydase.Server(service=service, web_port=8001).run()
 ```
 
-This will start the server, making your Device service accessible on [http://localhost:8001](http://localhost:8001).
+This will start the server, making your `Device` service accessible on 
+[http://localhost:8001](http://localhost:8001). The port number for the web server can 
+be customised in the server constructor or through environment variables and defaults 
+to `8001`.
 
 ### Accessing the Web Interface
 
@@ -170,7 +147,7 @@ import pydase
 
 # Replace the hostname and port with the IP address and the port of the machine where 
 # the service is running, respectively
-client_proxy = pydase.Client(url="ws://<ip_addr>:<service_port>").proxy
+client_proxy = pydase.Client(url="ws://<ip_addr>:<web_port>").proxy
 # client_proxy = pydase.Client(url="wss://your-domain.ch").proxy  # if your service uses ssl-encryption
 
 # After the connection, interact with the service attributes as if they were local
@@ -183,7 +160,7 @@ The proxy acts as a local representative of the remote service, enabling straigh
 
 The proxy class dynamically synchronizes with the server's exposed attributes. This synchronization allows the proxy to be automatically updated with any attributes or methods that the server exposes, essentially mirroring the server's API. This dynamic updating enables users to interact with the remote service as if they were working with a local object.
 
-The RPC client also supports tab completion support in the interpreter, can be used as a context manager and integrates very well with other pydase services. For more information, please refer to the [documentation](https://pydase.readthedocs.io/en/latest/user-guide/interaction/main/#python-client).
+The RPC client also supports tab completion support in the interpreter, can be used as a context manager and integrates very well with other pydase services. For more information, please refer to the [documentation][Python RPC Client].
 
 ### RESTful API
 The `pydase` RESTful API allows for standard HTTP-based interactions and provides access to various functionalities through specific routes. 
@@ -196,644 +173,14 @@ import json
 import requests
 
 response = requests.get(
-    "http://<hostname>:<port>/api/v1/get_value?access_path=<full_access_path>"
+    "http://<hostname>:<web_port>/api/v1/get_value?access_path=<full_access_path>"
 )
 serialized_value = json.loads(response.text)
 ```
 
-For more information, see [here](https://pydase.readthedocs.io/en/stable/user-guide/interaction/main/#restful-api).
+For more information, see [here][RESTful API].
 
-<!--usage-end-->
+<!--getting-started-end-->
-
-## Understanding the Component System
-
-<!-- Component User Guide Start -->
-
-In `pydase`, components are fundamental building blocks that bridge the Python backend logic with frontend visual representation and interactions. This system can be understood based on the following categories:
-
-### Built-in Type and Enum Components
-
-`pydase` automatically maps standard Python data types to their corresponding frontend components:
-
-- `str`: Translated into a `StringComponent` on the frontend.
-- `int` and `float`: Manifested as the `NumberComponent`.
-- `bool`: Rendered as a `ButtonComponent`.
-- `list`: Each item displayed individually, named after the list attribute and its index.
-- `dict`: Each key-value pair displayed individually, named after the dictionary attribute and its key. **Note** that the dictionary keys must be strings.
-- `enum.Enum`: Presented as an `EnumComponent`, facilitating dropdown selection.
-
-### Method Components
-Within the `DataService` class of `pydase`, only methods devoid of arguments can be represented in the frontend, classified into two distinct categories
-
-1. [**Tasks**](#understanding-tasks-in-pydase): Argument-free asynchronous functions, identified within `pydase` as tasks, are inherently designed for frontend interaction. These tasks are automatically rendered with a start/stop button, allowing users to initiate or halt the task execution directly from the web interface. 
-2. **Synchronous Methods with `@frontend` Decorator**: Synchronous methods without arguments can also be presented in the frontend. For this, they have to be decorated with the `@frontend` decorator.
-
-```python
-import pydase
-import pydase.components
-import pydase.units as u
-from pydase.utils.decorators import frontend
-
-
-class MyService(pydase.DataService):
-    @frontend
-    def exposed_method(self) -> None:
-        ...
-
-    async def my_task(self) -> None:
-        while True:
-            # ...
-```
-
-![Method Components](docs/images/method_components.png)
-
-You can still define synchronous tasks with arguments and call them using a python client. However, decorating them with the `@frontend` decorator will raise a `FunctionDefinitionError`. Defining a task with arguments will raise a `TaskDefinitionError`.
-I decided against supporting function arguments for functions rendered in the frontend due to the following reasons:
-
-1. Feature Request Pitfall: supporting function arguments create a bottomless pit of feature requests. As users encounter the limitations of supported types, demands for extending support to more complex types would grow.
-2. Complexity in Supported Argument Types: while simple types like `int`, `float`, `bool` and `str` could be easily supported, more complicated types are not (representation, (de-)serialization).
-
-### DataService Instances (Nested Classes)
-
-Nested `DataService` instances offer an organized hierarchy for components, enabling richer applications. Each nested class might have its own attributes and methods, each mapped to a frontend component.
-
-Here is an example:
-
-```python
-from pydase import DataService, Server
-
-
-class Channel(DataService):
-    def __init__(self, channel_id: int) -> None:
-        super().__init__()
-        self._channel_id = channel_id
-        self._current = 0.0
-
-    @property
-    def current(self) -> float:
-        # run code to get current
-        result = self._current
-        return result
-
-    @current.setter
-    def current(self, value: float) -> None:
-        # run code to set current
-        self._current = value
-
-
-class Device(DataService):
-    def __init__(self) -> None:
-        super().__init__()
-        self.channels = [Channel(i) for i in range(2)]
-
-
-if __name__ == "__main__":
-    service = Device()
-    Server(service).run()
-```
-
-![Nested Classes App](docs/images/Nested_Class_App.png)
-
-**Note** that defining classes within `DataService` classes is not supported (see [this issue](https://github.com/tiqi-group/pydase/issues/16)).
-
-### Custom Components (`pydase.components`)
-
-The custom components in `pydase` have two main parts:
-
-- A **Python Component Class** in the backend, implementing the logic needed to set, update, and manage the component's state and data.
-- A **Frontend React Component** that renders and manages user interaction in the browser.
-
-Below are the components available in the `pydase.components` module, accompanied by their Python usage:
-
-#### `DeviceConnection`
-
-The `DeviceConnection` component acts as a base class within the `pydase` framework for managing device connections. It provides a structured approach to handle connections by offering a customizable `connect` method and a `connected` property. This setup facilitates the implementation of automatic reconnection logic, which periodically attempts reconnection whenever the connection is lost.
-
-In the frontend, this class abstracts away the direct interaction with the `connect` method and the `connected` property. Instead, it showcases user-defined attributes, methods, and properties. When the `connected` status is `False`, the frontend displays an overlay that prompts manual reconnection through the `connect()` method. Successful reconnection removes the overlay.
-
-```python
-import pydase.components
-import pydase.units as u
-
-
-class Device(pydase.components.DeviceConnection):
-    def __init__(self) -> None:
-        super().__init__()
-        self._voltage = 10 * u.units.V
-
-    def connect(self) -> None:
-        if not self._connected:
-            self._connected = True
-
-    @property
-    def voltage(self) -> float:
-        return self._voltage
-
-
-class MyService(pydase.DataService):
-    def __init__(self) -> None:
-        super().__init__()
-        self.device = Device()
-
-
-if __name__ == "__main__":
-    service_instance = MyService()
-    pydase.Server(service_instance).run()
-```
-
-![DeviceConnection Component](docs/images/DeviceConnection_component.png)
-
-##### Customizing Connection Logic
-
-Users are encouraged to primarily override the `connect` method to tailor the connection process to their specific device. This method should adjust the `self._connected` attribute based on the outcome of the connection attempt:
-
-```python
-import pydase.components
-
-
-class MyDeviceConnection(pydase.components.DeviceConnection):
-    def __init__(self) -> None:
-        super().__init__()
-        # Add any necessary initialization code here
-
-    def connect(self) -> None:
-        # Implement device-specific connection logic here
-        # Update self._connected to `True` if the connection is successful,
-        # or `False` if unsuccessful
-        ...
-```
-
-Moreover, if the connection status requires additional logic, users can override the `connected` property:
-
-```python
-import pydase.components
-
-class MyDeviceConnection(pydase.components.DeviceConnection):
-    def __init__(self) -> None:
-        super().__init__()
-        # Add any necessary initialization code here
-
-    def connect(self) -> None:
-        # Implement device-specific connection logic here
-        # Ensure self._connected reflects the connection status accurately
-        ...
-
-    @property
-    def connected(self) -> bool:
-        # Implement custom logic to accurately report connection status
-        return self._connected
-```
-
-##### Reconnection Interval
-
-The `DeviceConnection` component automatically executes a task that checks for device availability at a default interval of 10 seconds. This interval is adjustable by modifying the `_reconnection_wait_time` attribute on the class instance.
-
-#### `Image`
-
-This component provides a versatile interface for displaying images within the application. Users can update and manage images from various sources, including local paths, URLs, and even matplotlib figures.
-
-The component offers methods to load images seamlessly, ensuring that visual content is easily integrated and displayed within the data service.
-
-```python
-import matplotlib.pyplot as plt
-import numpy as np
-import pydase
-from pydase.components.image import Image
-
-
-class MyDataService(pydase.DataService):
-    my_image = Image()
-
-
-if __name__ == "__main__":
-    service = MyDataService()
-    # loading from local path
-    service.my_image.load_from_path("/your/image/path/")
-
-    # loading from a URL
-    service.my_image.load_from_url("https://cataas.com/cat")
-
-    # loading a matplotlib figure
-    fig = plt.figure()
-    x = np.linspace(0, 2 * np.pi)
-    plt.plot(x, np.sin(x))
-    plt.grid()
-    service.my_image.load_from_matplotlib_figure(fig)
-
-    pydase.Server(service).run()
-```
-
-![Image Component](docs/images/Image_component.png)
-
-#### `NumberSlider`
-
-The `NumberSlider` component in the `pydase` package provides an interactive slider interface for adjusting numerical values on the frontend. It is designed to support both numbers and quantities and ensures that values adjusted on the frontend are synchronized with the backend.
-
-To utilize the `NumberSlider`, users should implement a class that derives from `NumberSlider`. This class can then define the initial values, minimum and maximum limits, step sizes, and additional logic as needed.
-
-Here's an example of how to implement and use a custom slider:
-
-```python
-import pydase
-import pydase.components
-
-
-class MySlider(pydase.components.NumberSlider):
-    def __init__(
-        self,
-        value: float = 0.0,
-        min_: float = 0.0,
-        max_: float = 100.0,
-        step_size: float = 1.0,
-    ) -> None:
-        super().__init__(value, min_, max_, step_size)
-
-    @property
-    def min(self) -> float:
-        return self._min
-
-    @min.setter
-    def min(self, value: float) -> None:
-        self._min = value
-
-    @property
-    def max(self) -> float:
-        return self._max
-
-    @max.setter
-    def max(self, value: float) -> None:
-        self._max = value
-
-    @property
-    def step_size(self) -> float:
-        return self._step_size
-
-    @step_size.setter
-    def step_size(self, value: float) -> None:
-        self._step_size = value
-
-    @property
-    def value(self) -> float:
-        """Slider value."""
-        return self._value
-
-    @value.setter
-    def value(self, value: float) -> None:
-        if value < self._min or value > self._max:
-            raise ValueError("Value is either below allowed min or above max value.")
-
-        self._value = value
-
-
-class MyService(pydase.DataService):
-    def __init__(self) -> None:
-        super().__init__()
-        self.voltage = MySlider()
-
-
-if __name__ == "__main__":
-    service_instance = MyService()
-    service_instance.voltage.value = 5
-    print(service_instance.voltage.value)  # Output: 5
-    pydase.Server(service_instance).run()
-```
-
-In this example, `MySlider` overrides the `min`, `max`, `step_size`, and `value` properties. Users can make any of these properties read-only by omitting the corresponding setter method.
-
-![Slider Component](docs/images/Slider_component.png)
-
-- Accessing parent class resources in `NumberSlider`
-
-    In scenarios where you need the slider component to interact with or access resources from its parent class, you can achieve this by passing a callback function to it. This method avoids directly passing the entire parent class instance (`self`) and offers a more encapsulated approach. The callback function can be designed to utilize specific attributes or methods of the parent class, allowing the slider to perform actions or retrieve data in response to slider events.
-
-    Here's an illustrative example:
-
-    ```python
-    from collections.abc import Callable
-
-    import pydase
-    import pydase.components
-
-
-    class MySlider(pydase.components.NumberSlider):
-        def __init__(
-            self,
-            value: float,
-            on_change: Callable[[float], None],
-        ) -> None:
-            super().__init__(value=value)
-            self._on_change = on_change
-
-        # ... other properties ...
-
-        @property
-        def value(self) -> float:
-            return self._value
-
-        @value.setter
-        def value(self, new_value: float) -> None:
-            if new_value < self._min or new_value > self._max:
-                raise ValueError("Value is either below allowed min or above max value.")
-            self._value = new_value
-            self._on_change(new_value)
-
-
-    class MyService(pydase.DataService):
-        def __init__(self) -> None:
-            self.voltage = MySlider(
-                5,
-                on_change=self.handle_voltage_change,
-            )
-
-        def handle_voltage_change(self, new_voltage: float) -> None:
-            print(f"Voltage changed to: {new_voltage}")
-            # Additional logic here
-
-    if __name__ == "__main__":
-       service_instance = MyService()
-       my_service.voltage.value = 7  # Output: "Voltage changed to: 7"
-       pydase.Server(service_instance).run()
-    ```
-
-- Incorporating units in `NumberSlider`
-
-    The `NumberSlider` is capable of [displaying units](#understanding-units-in-pydase) alongside values, enhancing its usability in contexts where unit representation is crucial. When utilizing `pydase.units`, you can specify units for the slider's value, allowing the component to reflect these units in the frontend.
-
-    Here's how to implement a `NumberSlider` with unit display:
-
-    ```python
-    import pydase
-    import pydase.components
-    import pydase.units as u
-
-    class MySlider(pydase.components.NumberSlider):
-        def __init__(
-            self,
-            value: u.Quantity = 0.0 * u.units.V,
-        ) -> None:
-            super().__init__(value)
-
-        @property
-        def value(self) -> u.Quantity:
-            return self._value
-
-        @value.setter
-        def value(self, value: u.Quantity) -> None:
-            if value.m < self._min or value.m > self._max:
-                raise ValueError("Value is either below allowed min or above max value.")
-            self._value = value
-
-    class MyService(pydase.DataService):
-        def __init__(self) -> None:
-            super().__init__()
-            self.voltage = MySlider()
-
-    if __name__ == "__main__":
-        service_instance = MyService()
-        service_instance.voltage.value = 5 * u.units.V
-        print(service_instance.voltage.value)  # Output: 5 V
-        pydase.Server(service_instance).run()
-    ```
-
-#### `ColouredEnum`
-
-This component provides a way to visually represent different states or categories in a data service using colour-coded options. It behaves similarly to a standard `Enum`, but the values encode colours in a format understood by CSS. The colours can be defined using various methods like Hexadecimal, RGB, HSL, and more.
-
-If the property associated with the `ColouredEnum` has a setter function, the keys of the enum will be rendered as a dropdown menu, allowing users to interact and select different options. Without a setter function, the selected key will simply be displayed as a coloured box with text inside, serving as a visual indicator.
-
-```python
-import pydase
-import pydase.components as pyc
-
-
-class MyStatus(pyc.ColouredEnum):
-    PENDING = "#FFA500"  # Hexadecimal colour (Orange)
-    RUNNING = "#0000FF80"  # Hexadecimal colour with transparency (Blue)
-    PAUSED = "rgb(169, 169, 169)"  # RGB colour (Dark Gray)
-    RETRYING = "rgba(255, 255, 0, 0.3)"  # RGB colour with transparency (Yellow)
-    COMPLETED = "hsl(120, 100%, 50%)"  # HSL colour (Green)
-    FAILED = "hsla(0, 100%, 50%, 0.7)"  # HSL colour with transparency (Red)
-    CANCELLED = "SlateGray"  # Cross-browser colour name (Slate Gray)
-
-
-class StatusTest(pydase.DataService):
-    _status = MyStatus.RUNNING
-
-    @property
-    def status(self) -> MyStatus:
-        return self._status
-
-    @status.setter
-    def status(self, value: MyStatus) -> None:
-        # do something ...
-        self._status = value
-
-# Modifying or accessing the status value:
-my_service = StatusExample()
-my_service.status = MyStatus.FAILED
-```
-
-![ColouredEnum Component](docs/images/ColouredEnum_component.png)
-
-**Note** that each enumeration name and value must be unique.
-This means that you should use different colour formats when you want to use a colour multiple times.
-
-#### Extending with New Components
-
-Users can also extend the library by creating custom components. This involves defining the behavior on the Python backend and the visual representation on the frontend. For those looking to introduce new components, the [guide on adding components](https://pydase.readthedocs.io/en/latest/dev-guide/Adding_Components/) provides detailed steps on achieving this.
-
-<!-- Component User Guide End -->
-
-## Understanding Service Persistence
-
-`pydase` allows you to easily persist the state of your service by saving it to a file. This is especially useful when you want to maintain the service's state across different runs.
-
-To save the state of your service, pass a `filename` keyword argument to the constructor of the `pydase.Server` class. If the file specified by `filename` does not exist, the state manager will create this file and store its state in it when the service is shut down. If the file already exists, the state manager will load the state from this file, setting the values of its attributes to the values stored in the file.
-
-Here's an example:
-
-```python
-from pydase import DataService, Server
-
-class Device(DataService):
-    # ... defining the Device class ...
-
-
-if __name__ == "__main__":
-    service = Device()
-    Server(service, filename="device_state.json").run()
-```
-
-In this example, the state of the `Device` service will be saved to `device_state.json` when the service is shut down. If `device_state.json` exists when the server is started, the state manager will restore the state of the service from this file.
-
-### Controlling Property State Loading with `@load_state`
-
-By default, the state manager only restores values for public attributes of your service. If you have properties that you want to control the loading for, you can use the `@load_state` decorator on your property setters. This indicates to the state manager that the value of the property should be loaded from the state file.
-
-Here is how you can apply the `@load_state` decorator:
-
-```python
-from pydase import DataService
-from pydase.data_service.state_manager import load_state
-
-class Device(DataService):
-    _name = "Default Device Name"
-
-    @property
-    def name(self) -> str:
-        return self._name
-
-    @name.setter
-    @load_state
-    def name(self, value: str) -> None:
-        self._name = value
-```
-
-With the `@load_state` decorator applied to the `name` property setter, the state manager will load and apply the `name` property's value from the file storing the state upon server startup, assuming it exists.
-
-Note: If the service class structure has changed since the last time its state was saved, only the attributes and properties decorated with `@load_state` that have remained the same will be restored from the settings file.
-
-## Understanding Tasks in pydase
-
-In `pydase`, a task is defined as an asynchronous function without arguments contained in a class that inherits from `DataService`. These tasks usually contain a while loop and are designed to carry out periodic functions.
-
-For example, a task might be used to periodically read sensor data, update a database, or perform any other recurring job. One core feature of `pydase` is its ability to automatically generate start and stop functions for these tasks. This allows you to control task execution via both the frontend and python clients, giving you flexible and powerful control over your service's operation.
-
-Another powerful feature of `pydase` is its ability to automatically start tasks upon initialization of the service. By specifying the tasks and their arguments in the `_autostart_tasks` dictionary in your service class's `__init__` method, `pydase` will automatically start these tasks when the server is started. Here's an example:
-
-```python
-from pydase import DataService, Server
-
-class SensorService(DataService):
-    def __init__(self):
-        super().__init__()
-        self.readout_frequency = 1.0
-        self._autostart_tasks["read_sensor_data"] = ()
-
-    def _process_data(self, data: ...) -> None:
-        ...
-
-    def _read_from_sensor(self) -> Any:
-        ...
-
-    async def read_sensor_data(self):
-        while True:
-            data = self._read_from_sensor()
-            self._process_data(data)  # Process the data as needed
-            await asyncio.sleep(self.readout_frequency)
-
-
-if __name__ == "__main__":
-    service = SensorService()
-    Server(service).run()
-```
-
-In this example, `read_sensor_data` is a task that continuously reads data from a sensor. By adding it to the `_autostart_tasks` dictionary, it will automatically start running when `Server(service).run()` is executed.
-As with all tasks, `pydase` will generate `start_read_sensor_data` and `stop_read_sensor_data` methods, which can be called to manually start and stop the data reading task. The readout frequency can be updated using the `readout_frequency` attribute.
-
-## Understanding Units in pydase
-
-`pydase` integrates with the [`pint`](https://pint.readthedocs.io/en/stable/) package to allow you to work with physical quantities within your service. This enables you to define attributes with units, making your service more expressive and ensuring consistency in the handling of physical quantities.
-
-You can define quantities in your `DataService` subclass using `pydase`'s `units` functionality.
-
-Here's an example:
-
-```python
-from typing import Any
-
-import pydase.units as u
-from pydase import DataService, Server
-
-
-class ServiceClass(DataService):
-    voltage = 1.0 * u.units.V
-    _current: u.Quantity = 1.0 * u.units.mA
-
-    @property
-    def current(self) -> u.Quantity:
-        return self._current
-
-    @current.setter
-    def current(self, value: u.Quantity) -> None:
-        self._current = value
-
-
-if __name__ == "__main__":
-    service = ServiceClass()
-
-    service.voltage = 10.0 * u.units.V
-    service.current = 1.5 * u.units.mA
-
-    Server(service).run()
-```
-
-In the frontend, quantities are rendered as floats, with the unit displayed as additional text. This allows you to maintain a clear and consistent representation of physical quantities across both the backend and frontend of your service.
-![Web interface with rendered units](./docs/images/Units_App.png)
-
-Should you need to access the magnitude or the unit of a quantity, you can use the `.m` attribute or the `.u` attribute of the variable, respectively. For example, this could be necessary to set the periodicity of a task:
-
-```python
-import asyncio
-from pydase import DataService, Server
-import pydase.units as u
-
-
-class ServiceClass(DataService):
-    readout_wait_time = 1.0 * u.units.ms
-
-    async def read_sensor_data(self):
-        while True:
-            print("Reading out sensor ...")
-            await asyncio.sleep(self.readout_wait_time.to("s").m)
-
-
-if __name__ == "__main__":
-    service = ServiceClass()
-
-    Server(service).run()
-```
-
-For more information about what you can do with the units, please consult the documentation of [`pint`](https://pint.readthedocs.io/en/stable/).
-
-## Using `validate_set` to Validate Property Setters
-
-The `validate_set` decorator ensures that a property setter reads back the set value using the property getter and checks it against the desired value.
-This decorator can be used to validate that a parameter has been correctly set on a device within a specified precision and timeout.
-
-The decorator takes two keyword arguments: `timeout` and `precision`. The `timeout` argument specifies the maximum time (in seconds) to wait for the value to be within the precision boundary.
-If the value is not within the precision boundary after this time, an exception is raised.
-The `precision` argument defines the acceptable deviation from the desired value.
-If `precision` is `None`, the value must be exact.
-For example, if `precision` is set to `1e-5`, the value read from the device must be within ±0.00001 of the desired value.
-
-Here’s how to use the `validate_set` decorator in a `DataService` class:
-
-```python
-import pydase
-from pydase.observer_pattern.observable.decorators import validate_set
-
-
-class Service(pydase.DataService):
-    def __init__(self) -> None:
-        super().__init__()
-        self._device = RemoteDevice()  # dummy class
-
-    @property
-    def value(self) -> float:
-        # Implement how to get the value from the remote device...
-        return self._device.value
-
-    @value.setter
-    @validate_set(timeout=1.0, precision=1e-5)
-    def value(self, value: float) -> None:
-        # Implement how to set the value on the remote device...
-        self._device.value = value
-
-
-if __name__ == "__main__":
-    pydase.Server(Service()).run()
-```
 
 ## Configuring pydase via Environment Variables
 
@@ -923,12 +270,27 @@ You have two primary ways to adjust the log levels in `pydase`:
 
 ## Documentation
 
-The full documentation provides more detailed information about `pydase`, including advanced usage examples, API references, and tips for troubleshooting common issues. See the [full documentation](https://pydase.readthedocs.io/en/latest/) for more information.
+The full documentation provides more detailed information about `pydase`, including advanced usage examples, API references, and tips for troubleshooting common issues. See the [full documentation](https://pydase.readthedocs.io/en/stable/) for more information.
 
 ## Contributing
 
-We welcome contributions! Please see [contributing.md](https://pydase.readthedocs.io/en/latest/about/contributing/) for details on how to contribute.
+We welcome contributions! Please see [contributing.md](https://pydase.readthedocs.io/en/stable/about/contributing/) for details on how to contribute.
 
 ## License
 
-`pydase` is licensed under the [MIT License](https://github.com/tiqi-group/pydase/blob/main/LICENSE).
+`pydase` is licensed under the [MIT License][License].
+
+[License]: ./LICENSE
+[Observer Pattern]: https://pydase.readthedocs.io/en/docs/dev-guide/Observer_Pattern_Implementation/
+[Service Persistence]: https://pydase.readthedocs.io/en/stable/user-guide/Service_Persistence
+[Defining DataService]: #defining-a-dataService
+[Web Interface Access]: #accessing-the-web-interface
+[Short RPC Client]: #connecting-to-the-service-via-python-rpc-client
+[Customizing Web Interface]: #customizing-the-web-interface
+[Task Management]: https://pydase.readthedocs.io/en/stable/user-guide/Tasks/
+[Units]: https://pydase.readthedocs.io/en/stable/user-guide/Understanding-Units/
+[Property Validation]: https://pydase.readthedocs.io/en/stable/user-guide/Validating-Property-Setters/
+[Custom Components]: https://pydase.readthedocs.io/en/stable/user-guide/Components/#custom-components-pydasecomponents
+[Components]: https://pydase.readthedocs.io/en/stable/user-guide/Components/
+[RESTful API]: https://pydase.readthedocs.io/en/stable/user-guide/interaction/main/#restful-api
+[Python RPC Client]: https://pydase.readthedocs.io/en/stable/user-guide/interaction/main/#python-rpc-client

docs/dev-guide/Observer_Pattern_Implementation.md

@@ -2,7 +2,7 @@
 
 ## Overview
 
-The Observer Pattern is a fundamental design pattern in the `pydase` package, serving as the central communication mechanism for state updates to clients connected to a service.
+The [Observer Pattern](https://en.wikipedia.org/wiki/Observer_pattern) is a fundamental design pattern in the `pydase` package, serving as the central communication mechanism for state updates to clients connected to a service.
 
 ## How it Works
 

docs/dev-guide/api.md

@@ -0,0 +1,45 @@
+::: pydase.data_service
+    handler: python
+
+::: pydase.server.server
+    handler: python
+
+::: pydase.server.web_server
+    handler: python
+
+::: pydase.client
+    handler: python
+
+::: pydase.components
+    handler: python
+
+::: pydase.task
+    handler: python
+    options:
+      inherited_members: false
+      show_submodules: true
+
+::: pydase.utils.serialization.serializer
+    handler: python
+
+::: pydase.utils.serialization.deserializer
+    handler: python
+    options:
+      show_root_heading: true
+      show_root_toc_entry: false
+      show_symbol_type_heading: true
+      show_symbol_type_toc: true
+
+::: pydase.utils.serialization.types
+    handler: python
+
+::: pydase.utils.decorators
+    handler: python
+    options:
+      filters: ["!render_in_frontend"]
+
+::: pydase.units
+    handler: python
+
+::: pydase.config
+    handler: python

docs/getting-started.md

@@ -1,14 +1,11 @@
 # Getting Started
-## Installation
 {%
    include-markdown "../README.md"
-   start="<!--installation-start-->"
+   start="<!--getting-started-start-->"
-   end="<!--installation-end-->"
+   end="<!--getting-started-end-->"
 %}
 
-## Usage
+[RESTful API]: ./user-guide/interaction/README.md#restful-api
-{%
+[Python RPC Client]: ./user-guide/interaction/README.md#python-rpc-client
-   include-markdown "../README.md"
+[Custom Components]: ./user-guide/Components.md#custom-components-pydasecomponents
-   start="<!--usage-start-->"
+[Components]: ./user-guide/Components.md
-   end="<!--usage-end-->"
-%}

docs/index.md

@@ -1 +1,16 @@
-{% include-markdown "../README.md" %}
+{%
+   include-markdown "../README.md"
+   start="<!--introduction-start-->"
+   end="<!--introduction-end-->"
+%}
+
+[License]: ./about/license.md
+[Observer Pattern]: ./dev-guide/Observer_Pattern_Implementation.md
+[Service Persistence]: ./user-guide/Service_Persistence.md
+[Defining DataService]: ./getting-started.md#defining-a-dataservice
+[Web Interface Access]: ./getting-started.md#accessing-the-web-interface
+[Short RPC Client]: ./getting-started.md#connecting-to-the-service-via-python-rpc-client
+[Customizing Web Interface]: ./user-guide/interaction/README.md#customization-options
+[Task Management]: ./user-guide/Tasks.md
+[Units]: ./user-guide/Understanding-Units.md
+[Property Validation]: ./user-guide/Validating-Property-Setters.md

docs/requirements.txt

@@ -5,6 +5,7 @@ charset-normalizer==3.3.2 ; python_version >= "3.10" and python_version < "4.0"
 click==8.1.7 ; python_version >= "3.10" and python_version < "4.0"
 colorama==0.4.6 ; python_version >= "3.10" and python_version < "4.0"
 ghp-import==2.1.0 ; python_version >= "3.10" and python_version < "4.0"
+griffe==1.1.0 ; python_version >= "3.10" and python_version < "4.0"
 idna==3.7 ; python_version >= "3.10" and python_version < "4.0"
 jinja2==3.1.4 ; python_version >= "3.10" and python_version < "4.0"
 markdown==3.6 ; python_version >= "3.10" and python_version < "4.0"
@@ -14,10 +15,12 @@ mkdocs-autorefs==1.0.1 ; python_version >= "3.10" and python_version < "4.0"
 mkdocs-get-deps==0.2.0 ; python_version >= "3.10" and python_version < "4.0"
 mkdocs-include-markdown-plugin==3.9.1 ; python_version >= "3.10" and python_version < "4.0"
 mkdocs-material-extensions==1.3.1 ; python_version >= "3.10" and python_version < "4.0"
-mkdocs-material==9.5.30 ; python_version >= "3.10" and python_version < "4.0"
+mkdocs-material==9.5.31 ; python_version >= "3.10" and python_version < "4.0"
 mkdocs-swagger-ui-tag==0.6.10 ; python_version >= "3.10" and python_version < "4.0"
 mkdocs==1.6.0 ; python_version >= "3.10" and python_version < "4.0"
-mkdocstrings==0.22.0 ; python_version >= "3.10" and python_version < "4.0"
+mkdocstrings-python==1.10.8 ; python_version >= "3.10" and python_version < "4.0"
+mkdocstrings==0.25.2 ; python_version >= "3.10" and python_version < "4.0"
+mkdocstrings[python]==0.25.2 ; python_version >= "3.10" and python_version < "4.0"
 packaging==24.1 ; python_version >= "3.10" and python_version < "4.0"
 paginate==0.5.6 ; python_version >= "3.10" and python_version < "4.0"
 pathspec==0.12.1 ; python_version >= "3.10" and python_version < "4.0"

docs/user-guide/Components.md

@@ -1,6 +1,435 @@
 # Components Guide
-{%
+
-   include-markdown "../../README.md"
+In `pydase`, components are fundamental building blocks that bridge the Python backend logic with frontend visual representation and interactions. This system can be understood based on the following categories:
-   start="<!-- Component User Guide Start -->"
+
-   end="<!-- Component User Guide End -->"
+## Built-in Type and Enum Components
-%}
+
+`pydase` automatically maps standard Python data types to their corresponding frontend components:
+
+- `str`: Translated into a `StringComponent` on the frontend.
+- `int` and `float`: Manifested as the `NumberComponent`.
+- `bool`: Rendered as a `ButtonComponent`.
+- `list`: Each item displayed individually, named after the list attribute and its index.
+- `dict`: Each key-value pair displayed individually, named after the dictionary attribute and its key. **Note** that the dictionary keys must be strings.
+- `enum.Enum`: Presented as an `EnumComponent`, facilitating dropdown selection.
+
+## Method Components
+Within the `DataService` class of `pydase`, only methods devoid of arguments can be represented in the frontend, classified into two distinct categories
+
+1. [**Tasks**](./Tasks.md): Argument-free asynchronous functions, identified within `pydase` as tasks, are inherently designed for frontend interaction. These tasks are automatically rendered with a start/stop button, allowing users to initiate or halt the task execution directly from the web interface. 
+2. **Synchronous Methods with `@frontend` Decorator**: Synchronous methods without arguments can also be presented in the frontend. For this, they have to be decorated with the `@frontend` decorator.
+
+```python
+import pydase
+import pydase.components
+import pydase.units as u
+from pydase.utils.decorators import frontend
+
+
+class MyService(pydase.DataService):
+    @frontend
+    def exposed_method(self) -> None:
+        ...
+
+    async def my_task(self) -> None:
+        while True:
+            # ...
+```
+
+![Method Components](../images/method_components.png)
+
+You can still define synchronous tasks with arguments and call them using a python client. However, decorating them with the `@frontend` decorator will raise a `FunctionDefinitionError`. Defining a task with arguments will raise a `TaskDefinitionError`.
+I decided against supporting function arguments for functions rendered in the frontend due to the following reasons:
+
+1. Feature Request Pitfall: supporting function arguments create a bottomless pit of feature requests. As users encounter the limitations of supported types, demands for extending support to more complex types would grow.
+2. Complexity in Supported Argument Types: while simple types like `int`, `float`, `bool` and `str` could be easily supported, more complicated types are not (representation, (de-)serialization).
+
+## DataService Instances (Nested Classes)
+
+Nested `DataService` instances offer an organized hierarchy for components, enabling richer applications. Each nested class might have its own attributes and methods, each mapped to a frontend component.
+
+Here is an example:
+
+```python
+from pydase import DataService, Server
+
+
+class Channel(DataService):
+    def __init__(self, channel_id: int) -> None:
+        super().__init__()
+        self._channel_id = channel_id
+        self._current = 0.0
+
+    @property
+    def current(self) -> float:
+        # run code to get current
+        result = self._current
+        return result
+
+    @current.setter
+    def current(self, value: float) -> None:
+        # run code to set current
+        self._current = value
+
+
+class Device(DataService):
+    def __init__(self) -> None:
+        super().__init__()
+        self.channels = [Channel(i) for i in range(2)]
+
+
+if __name__ == "__main__":
+    service = Device()
+    Server(service).run()
+```
+
+![Nested Classes App](../images/Nested_Class_App.png)
+
+**Note** that defining classes within `DataService` classes is not supported (see [this issue](https://github.com/tiqi-group/pydase/issues/16)).
+
+## Custom Components (`pydase.components`)
+
+The custom components in `pydase` have two main parts:
+
+- A **Python Component Class** in the backend, implementing the logic needed to set, update, and manage the component's state and data.
+- A **Frontend React Component** that renders and manages user interaction in the browser.
+
+Below are the components available in the `pydase.components` module, accompanied by their Python usage:
+
+### `DeviceConnection`
+
+The `DeviceConnection` component acts as a base class within the `pydase` framework for managing device connections. It provides a structured approach to handle connections by offering a customizable `connect` method and a `connected` property. This setup facilitates the implementation of automatic reconnection logic, which periodically attempts reconnection whenever the connection is lost.
+
+In the frontend, this class abstracts away the direct interaction with the `connect` method and the `connected` property. Instead, it showcases user-defined attributes, methods, and properties. When the `connected` status is `False`, the frontend displays an overlay that prompts manual reconnection through the `connect()` method. Successful reconnection removes the overlay.
+
+```python
+import pydase.components
+import pydase.units as u
+
+
+class Device(pydase.components.DeviceConnection):
+    def __init__(self) -> None:
+        super().__init__()
+        self._voltage = 10 * u.units.V
+
+    def connect(self) -> None:
+        if not self._connected:
+            self._connected = True
+
+    @property
+    def voltage(self) -> float:
+        return self._voltage
+
+
+class MyService(pydase.DataService):
+    def __init__(self) -> None:
+        super().__init__()
+        self.device = Device()
+
+
+if __name__ == "__main__":
+    service_instance = MyService()
+    pydase.Server(service_instance).run()
+```
+
+![DeviceConnection Component](../images/DeviceConnection_component.png)
+
+#### Customizing Connection Logic
+
+Users are encouraged to primarily override the `connect` method to tailor the connection process to their specific device. This method should adjust the `self._connected` attribute based on the outcome of the connection attempt:
+
+```python
+import pydase.components
+
+
+class MyDeviceConnection(pydase.components.DeviceConnection):
+    def __init__(self) -> None:
+        super().__init__()
+        # Add any necessary initialization code here
+
+    def connect(self) -> None:
+        # Implement device-specific connection logic here
+        # Update self._connected to `True` if the connection is successful,
+        # or `False` if unsuccessful
+        ...
+```
+
+Moreover, if the connection status requires additional logic, users can override the `connected` property:
+
+```python
+import pydase.components
+
+class MyDeviceConnection(pydase.components.DeviceConnection):
+    def __init__(self) -> None:
+        super().__init__()
+        # Add any necessary initialization code here
+
+    def connect(self) -> None:
+        # Implement device-specific connection logic here
+        # Ensure self._connected reflects the connection status accurately
+        ...
+
+    @property
+    def connected(self) -> bool:
+        # Implement custom logic to accurately report connection status
+        return self._connected
+```
+
+#### Reconnection Interval
+
+The `DeviceConnection` component automatically executes a task that checks for device availability at a default interval of 10 seconds. This interval is adjustable by modifying the `_reconnection_wait_time` attribute on the class instance.
+
+### `Image`
+
+This component provides a versatile interface for displaying images within the application. Users can update and manage images from various sources, including local paths, URLs, and even matplotlib figures.
+
+The component offers methods to load images seamlessly, ensuring that visual content is easily integrated and displayed within the data service.
+
+```python
+import matplotlib.pyplot as plt
+import numpy as np
+import pydase
+from pydase.components.image import Image
+
+
+class MyDataService(pydase.DataService):
+    my_image = Image()
+
+
+if __name__ == "__main__":
+    service = MyDataService()
+    # loading from local path
+    service.my_image.load_from_path("/your/image/path/")
+
+    # loading from a URL
+    service.my_image.load_from_url("https://cataas.com/cat")
+
+    # loading a matplotlib figure
+    fig = plt.figure()
+    x = np.linspace(0, 2 * np.pi)
+    plt.plot(x, np.sin(x))
+    plt.grid()
+    service.my_image.load_from_matplotlib_figure(fig)
+
+    pydase.Server(service).run()
+```
+
+![Image Component](../images/Image_component.png)
+
+### `NumberSlider`
+
+The `NumberSlider` component in the `pydase` package provides an interactive slider interface for adjusting numerical values on the frontend. It is designed to support both numbers and quantities and ensures that values adjusted on the frontend are synchronized with the backend.
+
+To utilize the `NumberSlider`, users should implement a class that derives from `NumberSlider`. This class can then define the initial values, minimum and maximum limits, step sizes, and additional logic as needed.
+
+Here's an example of how to implement and use a custom slider:
+
+```python
+import pydase
+import pydase.components
+
+
+class MySlider(pydase.components.NumberSlider):
+    def __init__(
+        self,
+        value: float = 0.0,
+        min_: float = 0.0,
+        max_: float = 100.0,
+        step_size: float = 1.0,
+    ) -> None:
+        super().__init__(value, min_, max_, step_size)
+
+    @property
+    def min(self) -> float:
+        return self._min
+
+    @min.setter
+    def min(self, value: float) -> None:
+        self._min = value
+
+    @property
+    def max(self) -> float:
+        return self._max
+
+    @max.setter
+    def max(self, value: float) -> None:
+        self._max = value
+
+    @property
+    def step_size(self) -> float:
+        return self._step_size
+
+    @step_size.setter
+    def step_size(self, value: float) -> None:
+        self._step_size = value
+
+    @property
+    def value(self) -> float:
+        """Slider value."""
+        return self._value
+
+    @value.setter
+    def value(self, value: float) -> None:
+        if value < self._min or value > self._max:
+            raise ValueError("Value is either below allowed min or above max value.")
+
+        self._value = value
+
+
+class MyService(pydase.DataService):
+    def __init__(self) -> None:
+        super().__init__()
+        self.voltage = MySlider()
+
+
+if __name__ == "__main__":
+    service_instance = MyService()
+    service_instance.voltage.value = 5
+    print(service_instance.voltage.value)  # Output: 5
+    pydase.Server(service_instance).run()
+```
+
+In this example, `MySlider` overrides the `min`, `max`, `step_size`, and `value` properties. Users can make any of these properties read-only by omitting the corresponding setter method.
+
+![Slider Component](../images/Slider_component.png)
+
+- Accessing parent class resources in `NumberSlider`
+
+    In scenarios where you need the slider component to interact with or access resources from its parent class, you can achieve this by passing a callback function to it. This method avoids directly passing the entire parent class instance (`self`) and offers a more encapsulated approach. The callback function can be designed to utilize specific attributes or methods of the parent class, allowing the slider to perform actions or retrieve data in response to slider events.
+
+    Here's an illustrative example:
+
+    ```python
+    from collections.abc import Callable
+
+    import pydase
+    import pydase.components
+
+
+    class MySlider(pydase.components.NumberSlider):
+        def __init__(
+            self,
+            value: float,
+            on_change: Callable[[float], None],
+        ) -> None:
+            super().__init__(value=value)
+            self._on_change = on_change
+
+        # ... other properties ...
+
+        @property
+        def value(self) -> float:
+            return self._value
+
+        @value.setter
+        def value(self, new_value: float) -> None:
+            if new_value < self._min or new_value > self._max:
+                raise ValueError("Value is either below allowed min or above max value.")
+            self._value = new_value
+            self._on_change(new_value)
+
+
+    class MyService(pydase.DataService):
+        def __init__(self) -> None:
+            self.voltage = MySlider(
+                5,
+                on_change=self.handle_voltage_change,
+            )
+
+        def handle_voltage_change(self, new_voltage: float) -> None:
+            print(f"Voltage changed to: {new_voltage}")
+            # Additional logic here
+
+    if __name__ == "__main__":
+       service_instance = MyService()
+       my_service.voltage.value = 7  # Output: "Voltage changed to: 7"
+       pydase.Server(service_instance).run()
+    ```
+
+- Incorporating units in `NumberSlider`
+
+    The `NumberSlider` is capable of [displaying units](./Understanding-Units.md) alongside values, enhancing its usability in contexts where unit representation is crucial. When utilizing `pydase.units`, you can specify units for the slider's value, allowing the component to reflect these units in the frontend.
+
+    Here's how to implement a `NumberSlider` with unit display:
+
+    ```python
+    import pydase
+    import pydase.components
+    import pydase.units as u
+
+    class MySlider(pydase.components.NumberSlider):
+        def __init__(
+            self,
+            value: u.Quantity = 0.0 * u.units.V,
+        ) -> None:
+            super().__init__(value)
+
+        @property
+        def value(self) -> u.Quantity:
+            return self._value
+
+        @value.setter
+        def value(self, value: u.Quantity) -> None:
+            if value.m < self._min or value.m > self._max:
+                raise ValueError("Value is either below allowed min or above max value.")
+            self._value = value
+
+    class MyService(pydase.DataService):
+        def __init__(self) -> None:
+            super().__init__()
+            self.voltage = MySlider()
+
+    if __name__ == "__main__":
+        service_instance = MyService()
+        service_instance.voltage.value = 5 * u.units.V
+        print(service_instance.voltage.value)  # Output: 5 V
+        pydase.Server(service_instance).run()
+    ```
+
+### `ColouredEnum`
+
+This component provides a way to visually represent different states or categories in a data service using colour-coded options. It behaves similarly to a standard `Enum`, but the values encode colours in a format understood by CSS. The colours can be defined using various methods like Hexadecimal, RGB, HSL, and more.
+
+If the property associated with the `ColouredEnum` has a setter function, the keys of the enum will be rendered as a dropdown menu, allowing users to interact and select different options. Without a setter function, the selected key will simply be displayed as a coloured box with text inside, serving as a visual indicator.
+
+```python
+import pydase
+import pydase.components as pyc
+
+
+class MyStatus(pyc.ColouredEnum):
+    PENDING = "#FFA500"  # Hexadecimal colour (Orange)
+    RUNNING = "#0000FF80"  # Hexadecimal colour with transparency (Blue)
+    PAUSED = "rgb(169, 169, 169)"  # RGB colour (Dark Gray)
+    RETRYING = "rgba(255, 255, 0, 0.3)"  # RGB colour with transparency (Yellow)
+    COMPLETED = "hsl(120, 100%, 50%)"  # HSL colour (Green)
+    FAILED = "hsla(0, 100%, 50%, 0.7)"  # HSL colour with transparency (Red)
+    CANCELLED = "SlateGray"  # Cross-browser colour name (Slate Gray)
+
+
+class StatusTest(pydase.DataService):
+    _status = MyStatus.RUNNING
+
+    @property
+    def status(self) -> MyStatus:
+        return self._status
+
+    @status.setter
+    def status(self, value: MyStatus) -> None:
+        # do something ...
+        self._status = value
+
+# Modifying or accessing the status value:
+my_service = StatusExample()
+my_service.status = MyStatus.FAILED
+```
+
+![ColouredEnum Component](../images/ColouredEnum_component.png)
+
+**Note** that each enumeration name and value must be unique.
+This means that you should use different colour formats when you want to use a colour multiple times.
+
+### Extending with New Components
+
+Users can also extend the library by creating custom components. This involves defining the behavior on the Python backend and the visual representation on the frontend. For those looking to introduce new components, the [guide on adding components](https://pydase.readthedocs.io/en/latest/dev-guide/Adding_Components/) provides detailed steps on achieving this.
+

docs/user-guide/Service_Persistence.md

@@ -0,0 +1,49 @@
+# Understanding Service Persistence
+
+`pydase` allows you to easily persist the state of your service by saving it to a file. This is especially useful when you want to maintain the service's state across different runs.
+
+To save the state of your service, pass a `filename` keyword argument to the constructor of the `pydase.Server` class. If the file specified by `filename` does not exist, the state manager will create this file and store its state in it when the service is shut down. If the file already exists, the state manager will load the state from this file, setting the values of its attributes to the values stored in the file.
+
+Here's an example:
+
+```python
+import pydase
+
+class Device(pydase.DataService):
+    # ... defining the Device class ...
+
+
+if __name__ == "__main__":
+    service = Device()
+    pydase.Server(service=service, filename="device_state.json").run()
+```
+
+In this example, the state of the `Device` service will be saved to `device_state.json` when the service is shut down. If `device_state.json` exists when the server is started, the state manager will restore the state of the service from this file.
+
+## Controlling Property State Loading with `@load_state`
+
+By default, the state manager only restores values for public attributes of your service. If you have properties that you want to control the loading for, you can use the `@load_state` decorator on your property setters. This indicates to the state manager that the value of the property should be loaded from the state file.
+
+Here is how you can apply the `@load_state` decorator:
+
+```python
+import pydase
+from pydase.data_service.state_manager import load_state
+
+class Device(pydase.DataService):
+    _name = "Default Device Name"
+
+    @property
+    def name(self) -> str:
+        return self._name
+
+    @name.setter
+    @load_state
+    def name(self, value: str) -> None:
+        self._name = value
+```
+
+With the `@load_state` decorator applied to the `name` property setter, the state manager will load and apply the `name` property's value from the file storing the state upon server startup, assuming it exists.
+
+Note: If the service class structure has changed since the last time its state was saved, only the attributes and properties decorated with `@load_state` that have remained the same will be restored from the settings file.
+

docs/user-guide/Tasks.md

@@ -0,0 +1,38 @@
+# Understanding Tasks
+
+In `pydase`, a task is defined as an asynchronous function without arguments that is decorated with the `@task` decorator and contained in a class that inherits from `pydase.DataService`. These tasks usually contain a while loop and are designed to carry out periodic functions. For example, a task might be used to periodically read sensor data, update a database, or perform any other recurring job.
+
+`pydase` allows you to control task execution via both the frontend and Python clients and can automatically start tasks upon initialization of the service. By using the `@task` decorator with the `autostart=True` argument in your service class, `pydase` will automatically start these tasks when the server is started. Here's an example:
+
+```python
+import pydase
+from pydase.task.decorator import task
+
+
+class SensorService(pydase.DataService):
+    def __init__(self):
+        super().__init__()
+        self.readout_frequency = 1.0
+
+    def _process_data(self, data: ...) -> None:
+        ...
+
+    def _read_from_sensor(self) -> Any:
+        ...
+
+    @task(autostart=True)
+    async def read_sensor_data(self):
+        while True:
+            data = self._read_from_sensor()
+            self._process_data(data)  # Process the data as needed
+            await asyncio.sleep(self.readout_frequency)
+
+
+if __name__ == "__main__":
+    service = SensorService()
+    pydase.Server(service=service).run()
+```
+
+In this example, `read_sensor_data` is a task that continuously reads data from a sensor. By decorating it with `@task(autostart=True)`, it will automatically start running when `pydase.Server(service).run()` is executed.
+
+The `@task` decorator replaces the function with a task object that has `start()` and `stop()` methods. This means you can control the task execution directly using these methods. For instance, you can manually start or stop the task by calling `service.read_sensor_data.start()` and `service.read_sensor_data.stop()`, respectively.

docs/user-guide/Understanding-Units.md

@@ -0,0 +1,64 @@
+# Understanding Units
+
+`pydase` integrates with the [`pint`](https://pint.readthedocs.io/en/stable/) package to allow you to work with physical quantities within your service. This enables you to define attributes with units, making your service more expressive and ensuring consistency in the handling of physical quantities.
+
+You can define quantities in your `pydase.DataService` subclass using the `pydase.units` module.
+Here's an example:
+
+```python
+from typing import Any
+
+import pydase
+import pydase.units as u
+
+
+class ServiceClass(pydase.DataService):
+    voltage = 1.0 * u.units.V
+    _current: u.Quantity = 1.0 * u.units.mA
+
+    @property
+    def current(self) -> u.Quantity:
+        return self._current
+
+    @current.setter
+    def current(self, value: u.Quantity) -> None:
+        self._current = value
+
+
+if __name__ == "__main__":
+    service = ServiceClass()
+
+    service.voltage = 10.0 * u.units.V
+    service.current = 1.5 * u.units.mA
+
+    pydase.Server(service=service).run()
+```
+
+In the frontend, quantities are rendered as floats, with the unit displayed as additional text. This allows you to maintain a clear and consistent representation of physical quantities across both the backend and frontend of your service.
+![Web interface with rendered units](../images/Units_App.png)
+
+Should you need to access the magnitude or the unit of a quantity, you can use the `.m` attribute or the `.u` attribute of the variable, respectively. For example, this could be necessary to set the periodicity of a task:
+
+```python
+import asyncio
+import pydase
+import pydase.units as u
+
+
+class ServiceClass(pydase.DataService):
+    readout_wait_time = 1.0 * u.units.ms
+
+    async def read_sensor_data(self):
+        while True:
+            print("Reading out sensor ...")
+            await asyncio.sleep(self.readout_wait_time.to("s").m)
+
+
+if __name__ == "__main__":
+    service = ServiceClass()
+
+    pydase.Server(service=service).run()
+```
+
+For more information about what you can do with the units, please consult the documentation of [`pint`](https://pint.readthedocs.io/en/stable/).
+

docs/user-guide/Validating-Property-Setters.md

@@ -0,0 +1,38 @@
+# Using `validate_set` to Validate Property Setters
+
+The `validate_set` decorator ensures that a property setter reads back the set value using the property getter and checks it against the desired value.
+This decorator can be used to validate that a parameter has been correctly set on a device within a specified precision and timeout.
+
+The decorator takes two keyword arguments: `timeout` and `precision`. The `timeout` argument specifies the maximum time (in seconds) to wait for the value to be within the precision boundary.
+If the value is not within the precision boundary after this time, an exception is raised.
+The `precision` argument defines the acceptable deviation from the desired value.
+If `precision` is `None`, the value must be exact.
+For example, if `precision` is set to `1e-5`, the value read from the device must be within ±0.00001 of the desired value.
+
+Here’s how to use the `validate_set` decorator in a `DataService` class:
+
+```python
+import pydase
+from pydase.observer_pattern.observable.decorators import validate_set
+
+
+class Service(pydase.DataService):
+    def __init__(self) -> None:
+        super().__init__()
+        self._device = RemoteDevice()  # dummy class
+
+    @property
+    def value(self) -> float:
+        # Implement how to get the value from the remote device...
+        return self._device.value
+
+    @value.setter
+    @validate_set(timeout=1.0, precision=1e-5)
+    def value(self, value: float) -> None:
+        # Implement how to set the value on the remote device...
+        self._device.value = value
+
+
+if __name__ == "__main__":
+    pydase.Server(service=Service()).run()
+```

frontend/src/components/GenericComponent.tsx

@@ -4,7 +4,6 @@ import { NumberComponent, NumberObject } from "./NumberComponent";
 import { SliderComponent } from "./SliderComponent";
 import { EnumComponent } from "./EnumComponent";
 import { MethodComponent } from "./MethodComponent";
-import { AsyncMethodComponent } from "./AsyncMethodComponent";
 import { StringComponent } from "./StringComponent";
 import { ListComponent } from "./ListComponent";
 import { DataServiceComponent, DataServiceJSON } from "./DataServiceComponent";
@@ -17,6 +16,7 @@ import { updateValue } from "../socket";
 import { DictComponent } from "./DictComponent";
 import { parseFullAccessPath } from "../utils/stateUtils";
 import { SerializedEnum, SerializedObject } from "../types/SerializedObject";
+import { TaskComponent, TaskStatus } from "./TaskComponent";
 
 interface GenericComponentProps {
   attribute: SerializedObject;
@@ -144,30 +144,16 @@ export const GenericComponent = React.memo(
         />
       );
     } else if (attribute.type === "method") {
-      if (!attribute.async) {
+      return (
-        return (
+        <MethodComponent
-          <MethodComponent
+          fullAccessPath={fullAccessPath}
-            fullAccessPath={fullAccessPath}
+          docString={attribute.doc}
-            docString={attribute.doc}
+          addNotification={addNotification}
-            addNotification={addNotification}
+          displayName={displayName}
-            displayName={displayName}
+          id={id}
-            id={id}
+          render={attribute.frontend_render}
-            render={attribute.frontend_render}
+        />
-          />
+      );
-        );
-      } else {
-        return (
-          <AsyncMethodComponent
-            fullAccessPath={fullAccessPath}
-            docString={attribute.doc}
-            value={attribute.value as "RUNNING" | null}
-            addNotification={addNotification}
-            displayName={displayName}
-            id={id}
-            render={attribute.frontend_render}
-          />
-        );
-      }
     } else if (attribute.type === "str") {
       return (
         <StringComponent
@@ -182,6 +168,17 @@ export const GenericComponent = React.memo(
           id={id}
         />
       );
+    } else if (attribute.type == "Task") {
+      return (
+        <TaskComponent
+          fullAccessPath={fullAccessPath}
+          docString={attribute.doc}
+          status={attribute.value["status"].value as TaskStatus}
+          addNotification={addNotification}
+          displayName={displayName}
+          id={id}
+        />
+      );
     } else if (attribute.type === "DataService") {
       return (
         <DataServiceComponent

frontend/src/components/NumberComponent.tsx

@@ -132,6 +132,8 @@ const handleNumericKey = (
   selectionStart: number,
   selectionEnd: number,
 ) => {
+  let newValue = value;
+
   // Check if a number key or a decimal point key is pressed
   if (key === "." && value.includes(".")) {
     // Check if value already contains a decimal. If so, ignore input.
@@ -139,14 +141,34 @@ const handleNumericKey = (
     return { value, selectionStart };
   }
 
-  let newValue = value;
+  // Handle minus sign input
+  if (key === "-") {
+    if (selectionStart === 0 && selectionEnd > selectionStart) {
+      // Replace selection with minus if selection starts at 0
+      newValue = "-" + value.slice(selectionEnd);
+      selectionStart = 1;
+    } else if (selectionStart === 0 && !value.startsWith("-")) {
+      // Add minus at the beginning if it doesn't exist
+      newValue = "-" + value;
+      selectionStart = 1;
+    } else if (
+      (selectionStart === 0 || selectionStart === 1) &&
+      value.startsWith("-")
+    ) {
+      // Remove minus if it exists
+      newValue = value.slice(1);
+      selectionStart = 0;
+    }
+
+    return { value: newValue, selectionStart };
+  }
 
   // Add the new key at the cursor's position
   if (selectionEnd > selectionStart) {
     // If there is a selection, replace it with the key
     newValue = value.slice(0, selectionStart) + key + value.slice(selectionEnd);
   } else {
-    // otherwise, append the key after the selection start
+    // Otherwise, insert the key at the cursor position
     newValue = value.slice(0, selectionStart) + key + value.slice(selectionStart);
   }
 
@@ -201,17 +223,7 @@ export const NumberComponent = React.memo((props: NumberComponentProps) => {
       // Select everything when pressing Ctrl + a
       inputTarget.setSelectionRange(0, value.length);
       return;
-    } else if (key === "-") {
+    } else if ((key >= "0" && key <= "9") || key === "-") {
-      if (selectionStart === 0 && !value.startsWith("-")) {
-        newValue = "-" + value;
-        selectionStart++;
-      } else if (value.startsWith("-") && selectionStart === 1) {
-        newValue = value.substring(1); // remove minus sign
-        selectionStart--;
-      } else {
-        return; // Ignore "-" pressed in other positions
-      }
-    } else if (key >= "0" && key <= "9") {
       // Check if a number key or a decimal point key is pressed
       ({ value: newValue, selectionStart } = handleNumericKey(
         key,

frontend/src/components/TaskComponent.tsx

@@ -5,67 +5,51 @@ import { DocStringComponent } from "./DocStringComponent";
 import { LevelName } from "./NotificationsComponent";
 import useRenderCount from "../hooks/useRenderCount";
 
-interface AsyncMethodProps {
+export type TaskStatus = "RUNNING" | "NOT_RUNNING";
+
+interface TaskProps {
   fullAccessPath: string;
-  value: "RUNNING" | null;
   docString: string | null;
-  hideOutput?: boolean;
+  status: TaskStatus;
   addNotification: (message: string, levelname?: LevelName) => void;
   displayName: string;
   id: string;
-  render: boolean;
 }
 
-export const AsyncMethodComponent = React.memo((props: AsyncMethodProps) => {
+export const TaskComponent = React.memo((props: TaskProps) => {
-  const {
+  const { fullAccessPath, docString, status, addNotification, displayName, id } = props;
-    fullAccessPath,
-    docString,
-    value: runningTask,
-    addNotification,
-    displayName,
-    id,
-  } = props;
-
-  // Conditional rendering based on the 'render' prop.
-  if (!props.render) {
-    return null;
-  }
 
   const renderCount = useRenderCount();
   const formRef = useRef(null);
   const [spinning, setSpinning] = useState(false);
-  const name = fullAccessPath.split(".").at(-1)!;
-  const parentPath = fullAccessPath.slice(0, -(name.length + 1));
 
   useEffect(() => {
     let message: string;
 
-    if (runningTask === null) {
+    if (status === "RUNNING") {
-      message = `${fullAccessPath} task was stopped.`;
-    } else {
       message = `${fullAccessPath} was started.`;
+    } else {
+      message = `${fullAccessPath} was stopped.`;
     }
+
     addNotification(message);
     setSpinning(false);
-  }, [props.value]);
+  }, [status]);
 
   const execute = async (event: React.FormEvent) => {
     event.preventDefault();
-    let method_name: string;
 
-    if (runningTask !== undefined && runningTask !== null) {
+    const method_name = status == "RUNNING" ? "stop" : "start";
-      method_name = `stop_${name}`;
-    } else {
-      method_name = `start_${name}`;
-    }
 
-    const accessPath = [parentPath, method_name].filter((element) => element).join(".");
+    const accessPath = [fullAccessPath, method_name]
+      .filter((element) => element)
+      .join(".");
     setSpinning(true);
     runMethod(accessPath);
   };
 
   return (
-    <div className="component asyncMethodComponent" id={id}>
+    <div className="component taskComponent" id={id}>
       {process.env.NODE_ENV === "development" && <div>Render count: {renderCount}</div>}
       <Form onSubmit={execute} ref={formRef}>
         <InputGroup>
@@ -76,7 +60,7 @@ export const AsyncMethodComponent = React.memo((props: AsyncMethodProps) => {
           <Button id={`button-${id}`} type="submit">
             {spinning ? (
               <Spinner size="sm" role="status" aria-hidden="true" />
-            ) : runningTask === "RUNNING" ? (
+            ) : status === "RUNNING" ? (
               "Stop "
             ) : (
               "Start "
@@ -88,4 +72,4 @@ export const AsyncMethodComponent = React.memo((props: AsyncMethodProps) => {
   );
 });
 
-AsyncMethodComponent.displayName = "AsyncMethodComponent";
+TaskComponent.displayName = "TaskComponent";

frontend/src/types/SerializedObject.ts

@@ -77,7 +77,12 @@ type SerializedException = SerializedObjectBase & {
   type: "Exception";
 };
 
-type DataServiceTypes = "DataService" | "Image" | "NumberSlider" | "DeviceConnection";
+type DataServiceTypes =
+  | "DataService"
+  | "Image"
+  | "NumberSlider"
+  | "DeviceConnection"
+  | "Task";
 
 type SerializedDataService = SerializedObjectBase & {
   name: string;

mkdocs.yml

@@ -6,7 +6,11 @@ nav:
   - Getting Started: getting-started.md
   - User Guide:
     - Components Guide: user-guide/Components.md
-    - Interacting with pydase Services: user-guide/interaction/main.md
+    - Interacting with pydase Services: user-guide/interaction/README.md
+    - Achieving Service Persistence: user-guide/Service_Persistence.md
+    - Understanding Tasks: user-guide/Tasks.md
+    - Understanding Units: user-guide/Understanding-Units.md
+    - Validating Property Setters: user-guide/Validating-Property-Setters.md
   - Developer Guide: 
     - Developer Guide: dev-guide/README.md
     - API Reference: dev-guide/api.md
@@ -40,10 +44,35 @@ markdown_extensions:
 
 
 plugins:
-  - include-markdown
+- include-markdown
-  - search
+- search
-  - mkdocstrings
+- mkdocstrings:
-  - swagger-ui-tag
+    handlers:
+      python:
+        paths: [src]  # search packages in the src folder
+        import:
+          - https://docs.python.org/3/objects.inv
+          - https://docs.pydantic.dev/latest/objects.inv
+          - https://confz.readthedocs.io/en/latest/objects.inv
+        options:
+          show_source: true
+          inherited_members: true
+          merge_init_into_class: true
+          show_signature_annotations: true
+          signature_crossrefs: true
+          separate_signature: true
+          docstring_options:
+            ignore_init_summary: true
+          # docstring_section_style: list
+          heading_level: 2
+          parameter_headings: true
+          show_root_heading: true
+          show_root_full_path: true
+          show_symbol_type_heading: true
+          show_symbol_type_toc: true
+          # summary: true
+          unwrap_annotated: true
+- swagger-ui-tag
 
 watch:
   - src/pydase

poetry.lock

@@ -769,6 +769,20 @@ python-dateutil = ">=2.8.1"
 [package.extras]
 dev = ["flake8", "markdown", "twine", "wheel"]
 
+[[package]]
+name = "griffe"
+version = "1.1.0"
+description = "Signatures for entire Python programs. Extract the structure, the frame, the skeleton of your project, to generate API documentation or find breaking changes in your API."
+optional = false
+python-versions = ">=3.8"
+files = [
+    {file = "griffe-1.1.0-py3-none-any.whl", hash = "sha256:38ccc5721571c95ae427123074cf0dc0d36bce7c9701ab2ada9fe0566ff50c10"},
+    {file = "griffe-1.1.0.tar.gz", hash = "sha256:c6328cbdec0d449549c1cc332f59227cd5603f903479d73e4425d828b782ffc3"},
+]
+
+[package.dependencies]
+colorama = ">=0.4"
+
 [[package]]
 name = "h11"
 version = "0.14.0"
@@ -1212,21 +1226,24 @@ beautifulsoup4 = ">=4.11.1"
 
 [[package]]
 name = "mkdocstrings"
-version = "0.22.0"
+version = "0.25.2"
 description = "Automatic documentation from sources, for MkDocs."
 optional = false
-python-versions = ">=3.7"
+python-versions = ">=3.8"
 files = [
-    {file = "mkdocstrings-0.22.0-py3-none-any.whl", hash = "sha256:2d4095d461554ff6a778fdabdca3c00c468c2f1459d469f7a7f622a2b23212ba"},
+    {file = "mkdocstrings-0.25.2-py3-none-any.whl", hash = "sha256:9e2cda5e2e12db8bb98d21e3410f3f27f8faab685a24b03b06ba7daa5b92abfc"},
-    {file = "mkdocstrings-0.22.0.tar.gz", hash = "sha256:82a33b94150ebb3d4b5c73bab4598c3e21468c79ec072eff6931c8f3bfc38256"},
+    {file = "mkdocstrings-0.25.2.tar.gz", hash = "sha256:5cf57ad7f61e8be3111a2458b4e49c2029c9cb35525393b179f9c916ca8042dc"},
 ]
 
 [package.dependencies]
+click = ">=7.0"
 Jinja2 = ">=2.11.1"
 Markdown = ">=3.3"
 MarkupSafe = ">=1.1"
-mkdocs = ">=1.2"
+mkdocs = ">=1.4"
 mkdocs-autorefs = ">=0.3.1"
+mkdocstrings-python = {version = ">=0.5.2", optional = true, markers = "extra == \"python\""}
+platformdirs = ">=2.2.0"
 pymdown-extensions = ">=6.3"
 
 [package.extras]
@@ -1234,6 +1251,21 @@ crystal = ["mkdocstrings-crystal (>=0.3.4)"]
 python = ["mkdocstrings-python (>=0.5.2)"]
 python-legacy = ["mkdocstrings-python-legacy (>=0.2.1)"]
 
+[[package]]
+name = "mkdocstrings-python"
+version = "1.10.8"
+description = "A Python handler for mkdocstrings."
+optional = false
+python-versions = ">=3.8"
+files = [
+    {file = "mkdocstrings_python-1.10.8-py3-none-any.whl", hash = "sha256:bb12e76c8b071686617f824029cb1dfe0e9afe89f27fb3ad9a27f95f054dcd89"},
+    {file = "mkdocstrings_python-1.10.8.tar.gz", hash = "sha256:5856a59cbebbb8deb133224a540de1ff60bded25e54d8beacc375bb133d39016"},
+]
+
+[package.dependencies]
+griffe = ">=0.49"
+mkdocstrings = ">=0.25"
+
 [[package]]
 name = "multidict"
 version = "6.0.5"
@@ -2464,4 +2496,4 @@ multidict = ">=4.0"
 [metadata]
 lock-version = "2.0"
 python-versions = "^3.10"
-content-hash = "54e25a68577a912301aa9125e3f3de545e03a199a79b2153c106285b92febbba"
+content-hash = "7131eddc2065147a18c145bb6da09492f03eb7fe050e968109cecb6044d17ed6"

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "pydase"
-version = "0.9.0"
+version = "0.10.5"
 description = "A flexible and robust Python library for creating, managing, and interacting with data services, with built-in support for web and RPC servers, and customizable features for diverse use cases."
 authors = ["Mose Mueller <mosmuell@ethz.ch>"]
 readme = "README.md"
@@ -38,7 +38,7 @@ optional = true
 [tool.poetry.group.docs.dependencies]
 mkdocs-material = "^9.5.30"
 mkdocs-include-markdown-plugin = "^3.9.1"
-mkdocstrings = "^0.22.0"
+mkdocstrings = {extras = ["python"], version = "^0.25.2"}
 pymdown-extensions = "^10.1"
 mkdocs-swagger-ui-tag = "^0.6.10"
 

src/pydase/client/client.py

@@ -31,10 +31,7 @@ class NotifyDict(TypedDict):
 
 def asyncio_loop_thread(loop: asyncio.AbstractEventLoop) -> None:
     asyncio.set_event_loop(loop)
-    try:
+    loop.run_forever()
-        loop.run_forever()
-    except RuntimeError:
-        logger.debug("Tried starting even loop, but it is running already")
 
 
 class ProxyClass(ProxyClassMixin, pydase.components.DeviceConnection):
@@ -43,10 +40,10 @@ class ProxyClass(ProxyClassMixin, pydase.components.DeviceConnection):
     via a socket.io client in an asyncio environment.
 
     Args:
-        sio_client (socketio.AsyncClient):
+        sio_client:
             The socket.io client instance used for asynchronous communication with the
             pydase service server.
-        loop (asyncio.AbstractEventLoop):
+        loop:
             The event loop in which the client operations are managed and executed.
 
     This class is used to create a proxy object that behaves like a local representation
@@ -54,26 +51,27 @@ class ProxyClass(ProxyClassMixin, pydase.components.DeviceConnection):
     while actually communicating over network protocols.
     It can also be used as an attribute of a pydase service itself, e.g.
 
-        ```python
+    ```python
-        import pydase
+    import pydase
 
 
-        class MyService(pydase.DataService):
+    class MyService(pydase.DataService):
-            proxy = pydase.Client(
+        proxy = pydase.Client(
-                hostname="...", port=8001, block_until_connected=False
+            hostname="...", port=8001, block_until_connected=False
-            ).proxy
+        ).proxy
 
 
-        if __name__ == "__main__":
+    if __name__ == "__main__":
-            service = MyService()
+        service = MyService()
-            server = pydase.Server(service, web_port=8002).run()
+        server = pydase.Server(service, web_port=8002).run()
-        ```
+    ```
     """
 
     def __init__(
         self, sio_client: socketio.AsyncClient, loop: asyncio.AbstractEventLoop
     ) -> None:
         super().__init__()
+        pydase.components.DeviceConnection.__init__(self)
         self._initialise(sio_client=sio_client, loop=loop)
 
 
@@ -84,19 +82,16 @@ class Client:
     connection, disconnection, and updates, and ensures that the proxy object is
     up-to-date with the server state.
 
-    Attributes:
-        proxy (ProxyClass):
-            A proxy object representing the remote service, facilitating interaction as
-            if it were local.
-
     Args:
-        url (str):
+        url:
             The URL of the pydase Socket.IO server. This should always contain the
             protocol and the hostname.
+
             Examples:
-                - wss://my-service.example.com  # for secure connections, use wss
+
-                - ws://localhost:8001
+            - `wss://my-service.example.com`  # for secure connections, use wss
-        block_until_connected (bool):
+            - `ws://localhost:8001`
+        block_until_connected:
             If set to True, the constructor will block until the connection to the
             service has been established. This is useful for ensuring the client is
             ready to use immediately after instantiation. Default is True.
@@ -112,6 +107,8 @@ class Client:
         self._sio = socketio.AsyncClient()
         self._loop = asyncio.new_event_loop()
         self.proxy = ProxyClass(sio_client=self._sio, loop=self._loop)
+        """A proxy object representing the remote service, facilitating interaction as
+        if it were local."""
         self._thread = threading.Thread(
             target=asyncio_loop_thread, args=(self._loop,), daemon=True
         )

src/pydase/client/proxy_loader.py

@@ -351,7 +351,7 @@ class ProxyLoader:
     ) -> Any:
         # Custom types like Components or DataService classes
         component_class = cast(
-            type, Deserializer.get_component_class(serialized_object["type"])
+            type, Deserializer.get_service_base_class(serialized_object["type"])
         )
         class_bases = (
             ProxyClassMixin,

src/pydase/components/coloured_enum.py

@@ -7,58 +7,59 @@ class ColouredEnum(Enum):
 
     This class extends the standard Enum but requires its values to be valid CSS
     colour codes. Supported colour formats include:
-        - Hexadecimal colours
+
-        - Hexadecimal colours with transparency
+    - Hexadecimal colours
-        - RGB colours
+    - Hexadecimal colours with transparency
-        - RGBA colours
+    - RGB colours
-        - HSL colours
+    - RGBA colours
-        - HSLA colours
+    - HSL colours
-        - Predefined/Cross-browser colour names
+    - HSLA colours
+    - Predefined/Cross-browser colour names
+
     Refer to the this website for more details on colour formats:
     (https://www.w3schools.com/cssref/css_colours_legal.php)
 
     The behavior of this component in the UI depends on how it's defined in the data
     service:
-        - As property with a setter or as attribute: Renders as a dropdown menu,
+
-            allowing users to select and change its value from the frontend.
+    - As property with a setter or as attribute: Renders as a dropdown menu, allowing
-        - As property without a setter: Displays as a coloured box with the key of the
+    users to select and change its value from the frontend.
-            `ColouredEnum` as text inside, serving as a visual indicator without user
+    - As property without a setter: Displays as a coloured box with the key of the
-            interaction.
+    `ColouredEnum` as text inside, serving as a visual indicator without user
+    interaction.
 
     Example:
-    --------
+        ```python
-    ```python
+        import pydase.components as pyc
-    import pydase.components as pyc
+        import pydase
-    import pydase
 
-    class MyStatus(pyc.ColouredEnum):
+        class MyStatus(pyc.ColouredEnum):
-        PENDING = "#FFA500"  # Orange
+            PENDING = "#FFA500"  # Orange
-        RUNNING = "#0000FF80"  # Transparent Blue
+            RUNNING = "#0000FF80"  # Transparent Blue
-        PAUSED = "rgb(169, 169, 169)"  # Dark Gray
+            PAUSED = "rgb(169, 169, 169)"  # Dark Gray
-        RETRYING = "rgba(255, 255, 0, 0.3)"  # Transparent Yellow
+            RETRYING = "rgba(255, 255, 0, 0.3)"  # Transparent Yellow
-        COMPLETED = "hsl(120, 100%, 50%)"  # Green
+            COMPLETED = "hsl(120, 100%, 50%)"  # Green
-        FAILED = "hsla(0, 100%, 50%, 0.7)"  # Transparent Red
+            FAILED = "hsla(0, 100%, 50%, 0.7)"  # Transparent Red
-        CANCELLED = "SlateGray"  # Slate Gray
+            CANCELLED = "SlateGray"  # Slate Gray
 
-    class StatusExample(pydase.DataService):
+        class StatusExample(pydase.DataService):
-        _status = MyStatus.RUNNING
+            _status = MyStatus.RUNNING
 
-        @property
+            @property
-        def status(self) -> MyStatus:
+            def status(self) -> MyStatus:
-            return self._status
+                return self._status
 
-        @status.setter
+            @status.setter
-        def status(self, value: MyStatus) -> None:
+            def status(self, value: MyStatus) -> None:
-            # Custom logic here...
+                # Custom logic here...
-            self._status = value
+                self._status = value
 
-    # Example usage:
+        # Example usage:
-    my_service = StatusExample()
+        my_service = StatusExample()
-    my_service.status = MyStatus.FAILED
+        my_service.status = MyStatus.FAILED
-    ```
+        ```
 
-    Note
+    Note:
-    ----
+        Each enumeration name and value must be unique. This means that you should use
-    Each enumeration name and value must be unique. This means that you should use
+        different colour formats when you want to use a colour multiple times.
-    different colour formats when you want to use a colour multiple times.
     """

src/pydase/components/device_connection.py

@@ -1,6 +1,7 @@
 import asyncio
 
 import pydase.data_service
+import pydase.task.decorator
 
 
 class DeviceConnection(pydase.data_service.DataService):
@@ -19,22 +20,26 @@ class DeviceConnection(pydase.data_service.DataService):
     to the device. This method should update the `self._connected` attribute to reflect
     the connection status:
 
-    >>> class MyDeviceConnection(DeviceConnection):
+    ```python
-    ...     def connect(self) -> None:
+    class MyDeviceConnection(DeviceConnection):
-    ...         # Implementation to connect to the device
+        def connect(self) -> None:
-    ...         # Update self._connected to `True` if connection is successful,
+            # Implementation to connect to the device
-    ...         # `False` otherwise
+            # Update self._connected to `True` if connection is successful,
-    ...         ...
+            # `False` otherwise
+            ...
+    ```
 
     Optionally, if additional logic is needed to determine the connection status,
     the `connected` property can also be overridden:
 
-    >>> class MyDeviceConnection(DeviceConnection):
+    ```python
-    ...     @property
+    class MyDeviceConnection(DeviceConnection):
-    ...     def connected(self) -> bool:
+        @property
-    ...         # Custom logic to determine connection status
+        def connected(self) -> bool:
-    ...         return some_custom_condition
+            # Custom logic to determine connection status
-    ...
+            return some_custom_condition
+
+    ```
 
     Frontend Representation
     -----------------------
@@ -48,7 +53,6 @@ class DeviceConnection(pydase.data_service.DataService):
     def __init__(self) -> None:
         super().__init__()
         self._connected = False
-        self._autostart_tasks["_handle_connection"] = ()  # type: ignore
         self._reconnection_wait_time = 10.0
 
     def connect(self) -> None:
@@ -66,6 +70,7 @@ class DeviceConnection(pydase.data_service.DataService):
         """
         return self._connected
 
+    @pydase.task.decorator.task(autostart=True)
     async def _handle_connection(self) -> None:
         """Automatically tries reconnecting to the device if it is not connected.
         This method leverages the `connect` method and the `connected` property to

src/pydase/components/number_slider.py

@@ -11,76 +11,74 @@ class NumberSlider(DataService):
     This class models a UI slider for a data service, allowing for adjustments of a
     parameter within a specified range and increments.
 
-    Parameters:
+    Args:
-    -----------
+        value:
-    value (float, optional):
+            The initial value of the slider. Defaults to 0.
-        The initial value of the slider. Defaults to 0.
+        min_:
-    min (float, optional):
+            The minimum value of the slider. Defaults to 0.
-        The minimum value of the slider. Defaults to 0.
+        max_:
-    max (float, optional):
+            The maximum value of the slider. Defaults to 100.
-        The maximum value of the slider. Defaults to 100.
+        step_size:
-    step_size (float, optional):
+            The increment/decrement step size of the slider. Defaults to 1.0.
-        The increment/decrement step size of the slider. Defaults to 1.0.
 
     Example:
-    --------
+        ```python
-    ```python
+        class MySlider(pydase.components.NumberSlider):
-    class MySlider(pydase.components.NumberSlider):
+            def __init__(
-        def __init__(
+                self,
-            self,
+                value: float = 0.0,
-            value: float = 0.0,
+                min_: float = 0.0,
-            min_: float = 0.0,
+                max_: float = 100.0,
-            max_: float = 100.0,
+                step_size: float = 1.0,
-            step_size: float = 1.0,
+            ) -> None:
-        ) -> None:
+                super().__init__(value, min_, max_, step_size)
-            super().__init__(value, min_, max_, step_size)
 
-        @property
+            @property
-        def min(self) -> float:
+            def min(self) -> float:
-            return self._min
+                return self._min
 
-        @min.setter
+            @min.setter
-        def min(self, value: float) -> None:
+            def min(self, value: float) -> None:
-            self._min = value
+                self._min = value
 
-        @property
+            @property
-        def max(self) -> float:
+            def max(self) -> float:
-            return self._max
+                return self._max
 
-        @max.setter
+            @max.setter
-        def max(self, value: float) -> None:
+            def max(self, value: float) -> None:
-            self._max = value
+                self._max = value
 
-        @property
+            @property
-        def step_size(self) -> float:
+            def step_size(self) -> float:
-            return self._step_size
+                return self._step_size
 
-        @step_size.setter
+            @step_size.setter
-        def step_size(self, value: float) -> None:
+            def step_size(self, value: float) -> None:
-            self._step_size = value
+                self._step_size = value
 
-        @property
+            @property
-        def value(self) -> float:
+            def value(self) -> float:
-            return self._value
+                return self._value
 
-        @value.setter
+            @value.setter
-        def value(self, value: float) -> None:
+            def value(self, value: float) -> None:
-            if value < self._min or value > self._max:
+                if value < self._min or value > self._max:
-                raise ValueError(
+                    raise ValueError(
-                    "Value is either below allowed min or above max value."
+                        "Value is either below allowed min or above max value."
-                )
+                    )
 
-            self._value = value
+                self._value = value
 
-    class MyService(pydase.DataService):
+        class MyService(pydase.DataService):
-        def __init__(self) -> None:
+            def __init__(self) -> None:
-            self.voltage = MyService()
+                self.voltage = MyService()
 
-    # Modifying or accessing the voltage value:
+        # Modifying or accessing the voltage value:
-    my_service = MyService()
+        my_service = MyService()
-    my_service.voltage.value = 5
+        my_service.voltage.value = 5
-    print(my_service.voltage.value)  # Output: 5
+        print(my_service.voltage.value)  # Output: 5
-    ```
+        ```
     """
 
     def __init__(

src/pydase/config.py

@@ -5,19 +5,31 @@ from confz import BaseConfig, EnvSource
 
 
 class OperationMode(BaseConfig):  # type: ignore[misc]
-    environment: Literal["development", "production"] = "development"
+    environment: Literal["testing", "development", "production"] = "development"
+    """The service's operation mode."""
 
     CONFIG_SOURCES = EnvSource(allow=["ENVIRONMENT"])
 
 
 class ServiceConfig(BaseConfig):  # type: ignore[misc]
+    """Service configuration.
+
+    Variables can be set through environment variables prefixed with `SERVICE_` or an
+    `.env` file containing those variables.
+    """
+
     config_dir: Path = Path("config")
+    """Configuration directory"""
     web_port: int = 8001
+    """Web server port"""
 
     CONFIG_SOURCES = EnvSource(allow_all=True, prefix="SERVICE_", file=".env")
 
 
 class WebServerConfig(BaseConfig):  # type: ignore[misc]
+    """The service's web server configuration."""
+
     generate_web_settings: bool = False
+    """Should generate web_settings.json file"""
 
     CONFIG_SOURCES = EnvSource(allow=["GENERATE_WEB_SETTINGS"])

src/pydase/data_service/abstract_data_service.py

@@ -1,15 +1,7 @@
 from __future__ import annotations
 
-from typing import TYPE_CHECKING, Any
-
 from pydase.observer_pattern.observable.observable import Observable
 
-if TYPE_CHECKING:
-    from pydase.data_service.data_service import DataService
-    from pydase.data_service.task_manager import TaskManager
-
 
 class AbstractDataService(Observable):
-    __root__: DataService
+    pass
-    _task_manager: TaskManager
-    _autostart_tasks: dict[str, tuple[Any]]

src/pydase/data_service/data_service.py

@@ -5,12 +5,12 @@ from typing import Any
 
 import pydase.units as u
 from pydase.data_service.abstract_data_service import AbstractDataService
-from pydase.data_service.task_manager import TaskManager
 from pydase.observer_pattern.observable.observable import (
     Observable,
 )
 from pydase.utils.helpers import (
     get_class_and_instance_attributes,
+    is_descriptor,
     is_property_attribute,
 )
 from pydase.utils.serialization.serializer import (
@@ -24,11 +24,6 @@ logger = logging.getLogger(__name__)
 class DataService(AbstractDataService):
     def __init__(self) -> None:
         super().__init__()
-        self._task_manager = TaskManager(self)
-
-        if not hasattr(self, "_autostart_tasks"):
-            self._autostart_tasks = {}
-
         self.__check_instance_classes()
 
     def __setattr__(self, __name: str, __value: Any) -> None:
@@ -74,7 +69,7 @@ class DataService(AbstractDataService):
         if not issubclass(
             value_class,
             (int | float | bool | str | list | dict | Enum | u.Quantity | Observable),
-        ):
+        ) and not is_descriptor(__value):
             logger.warning(
                 "Class '%s' does not inherit from DataService. This may lead to"
                 " unexpected behaviour!",

src/pydase/data_service/data_service_observer.py

@@ -8,7 +8,10 @@ from pydase.observer_pattern.observable.observable_object import ObservableObjec
 from pydase.observer_pattern.observer.property_observer import (
     PropertyObserver,
 )
-from pydase.utils.helpers import get_object_attr_from_path
+from pydase.utils.helpers import (
+    get_object_attr_from_path,
+    normalize_full_access_path_string,
+)
 from pydase.utils.serialization.serializer import (
     SerializationPathError,
     SerializedObject,
@@ -53,7 +56,7 @@ class DataServiceObserver(PropertyObserver):
         cached_value = cached_value_dict.get("value")
         if (
             all(part[0] != "_" for part in full_access_path.split("."))
-            and cached_value != value
+            and cached_value != dump(value)["value"]
         ):
             logger.debug("'%s' changed to '%s'", full_access_path, value)
 
@@ -99,7 +102,8 @@ class DataServiceObserver(PropertyObserver):
         )
 
     def _notify_dependent_property_changes(self, changed_attr_path: str) -> None:
-        changed_props = self.property_deps_dict.get(changed_attr_path, [])
+        normalized_attr_path = normalize_full_access_path_string(changed_attr_path)
+        changed_props = self.property_deps_dict.get(normalized_attr_path, [])
         for prop in changed_props:
             # only notify about changing attribute if it is not currently being
             # "changed" e.g. when calling the getter of a property within another
@@ -124,8 +128,10 @@ class DataServiceObserver(PropertyObserver):
         object.
 
         Args:
-            callback (Callable[[str, Any, dict[str, Any]]): The callback function to be
+            callback:
-            registered. The function should have the following signature:
+                The callback function to be registered. The function should have the
+                following signature:
+
                 - full_access_path (str): The full dot-notation access path of the
                   changed attribute. This path indicates the location of the changed
                   attribute within the observable object's structure.

src/pydase/data_service/state_manager.py

@@ -33,17 +33,19 @@ def load_state(func: Callable[..., Any]) -> Callable[..., Any]:
     the value should be loaded from the JSON file.
 
     Example:
-    >>>     class Service(pydase.DataService):
+        ```python
-    ...         _name = "Service"
+        class Service(pydase.DataService):
-    ...
+            _name = "Service"
-    ...         @property
+
-    ...         def name(self) -> str:
+            @property
-    ...             return self._name
+            def name(self) -> str:
-    ...
+                return self._name
-    ...         @name.setter
+
-    ...         @load_state
+            @name.setter
-    ...         def name(self, value: str) -> None:
+            @load_state
-    ...             self._name = value
+            def name(self, value: str) -> None:
+                self._name = value
+        ```
     """
 
     func._load_state = True  # type: ignore[attr-defined]
@@ -85,13 +87,11 @@ class StateManager:
     StateManager provides a snapshot of the DataService's state that is sufficiently
     accurate for initial rendering and interaction.
 
-    Attributes:
+    Args:
-        cache (dict[str, Any]):
+        service:
-            A dictionary cache of the DataService's state.
-        filename (str):
-            The file name used for storing the DataService's state.
-        service (DataService):
             The DataService instance whose state is being managed.
+        filename:
+            The file name used for storing the DataService's state.
 
     Note:
         The StateManager's cache updates are triggered by notifications and do not
@@ -200,9 +200,11 @@ class StateManager:
         It also handles type-specific conversions for the new value before setting it.
 
         Args:
-            path: A dot-separated string indicating the hierarchical path to the
+            path:
+                A dot-separated string indicating the hierarchical path to the
                 attribute.
-            value: The new value to set for the attribute.
+            serialized_value:
+                The serialized representation of the new value to set for the attribute.
         """
 
         try:

src/pydase/data_service/task_manager.py

@@ -1,225 +0,0 @@
-from __future__ import annotations
-
-import asyncio
-import inspect
-import logging
-from enum import Enum
-from typing import TYPE_CHECKING, Any
-
-from pydase.data_service.abstract_data_service import AbstractDataService
-from pydase.utils.helpers import (
-    function_has_arguments,
-    get_class_and_instance_attributes,
-    is_property_attribute,
-)
-
-if TYPE_CHECKING:
-    from collections.abc import Callable
-
-    from .data_service import DataService
-
-logger = logging.getLogger(__name__)
-
-
-class TaskStatus(Enum):
-    RUNNING = "running"
-
-
-class TaskManager:
-    """
-    The TaskManager class is a utility designed to manage asynchronous tasks. It
-    provides functionality for starting, stopping, and tracking these tasks. The class
-    is primarily used by the DataService class to manage its tasks.
-
-    A task in TaskManager is any asynchronous function. To add a task, you simply need
-    to define an async function within your class that extends TaskManager. For example:
-
-    ```python
-    class MyService(DataService):
-        async def my_task(self):
-            # Your task implementation here
-            pass
-    ```
-
-    With the above definition, TaskManager automatically creates `start_my_task` and
-    `stop_my_task` methods that can be used to control the task.
-
-    TaskManager also supports auto-starting tasks. If there are tasks that should start
-    running as soon as an instance of your class is created, you can define them in
-    `self._autostart_tasks` in your class constructor (__init__ method). Here's how:
-
-    ```python
-    class MyService(DataService):
-        def __init__(self):
-            self._autostart_tasks = {
-                "my_task": (*args)  # Replace with actual arguments
-            }
-            self.wait_time = 1
-            super().__init__()
-
-        async def my_task(self, *args):
-            while True:
-                # Your task implementation here
-                await asyncio.sleep(self.wait_time)
-    ```
-
-    In the above example, `my_task` will start running as soon as
-    `_start_autostart_tasks` is called which is done when the DataService instance is
-    passed to the `pydase.Server` class.
-
-    The responsibilities of the TaskManager class are:
-
-    - Track all running tasks: Keeps track of all the tasks that are currently running.
-    This allows for monitoring of task statuses and for making sure tasks do not
-    overlap.
-    - Provide the ability to start and stop tasks: Automatically creates methods to
-    start and stop each task.
-    - Emit notifications when the status of a task changes: Has a built-in mechanism for
-    emitting notifications when a task starts or stops. This is used to update the user
-    interfaces, but can also be used to write logs, etc.
-    """
-
-    def __init__(self, service: DataService) -> None:
-        self.service = service
-
-        self.tasks: dict[str, asyncio.Task[None]] = {}
-        """A dictionary to keep track of running tasks. The keys are the names of the
-        tasks and the values are TaskDict instances which include the task itself and
-        its kwargs.
-        """
-
-        self._set_start_and_stop_for_async_methods()
-
-    @property
-    def _loop(self) -> asyncio.AbstractEventLoop:
-        return asyncio.get_running_loop()
-
-    def _set_start_and_stop_for_async_methods(self) -> None:
-        for name in dir(self.service):
-            # circumvents calling properties
-            if is_property_attribute(self.service, name):
-                continue
-
-            method = getattr(self.service, name)
-            if inspect.iscoroutinefunction(method):
-                if function_has_arguments(method):
-                    logger.info(
-                        "Async function %a is defined with at least one argument. If "
-                        "you want to use it as a task, remove the argument(s) from the "
-                        "function definition.",
-                        method.__name__,
-                    )
-                    continue
-
-                # create start and stop methods for each coroutine
-                setattr(
-                    self.service, f"start_{name}", self._make_start_task(name, method)
-                )
-                setattr(self.service, f"stop_{name}", self._make_stop_task(name))
-
-    def _initiate_task_startup(self) -> None:
-        if self.service._autostart_tasks is not None:
-            for service_name, args in self.service._autostart_tasks.items():
-                start_method = getattr(self.service, f"start_{service_name}", None)
-                if start_method is not None and callable(start_method):
-                    start_method(*args)
-                else:
-                    logger.warning(
-                        "No start method found for service '%s'", service_name
-                    )
-
-    def start_autostart_tasks(self) -> None:
-        self._initiate_task_startup()
-        attrs = get_class_and_instance_attributes(self.service)
-
-        for attr_value in attrs.values():
-            if isinstance(attr_value, AbstractDataService):
-                attr_value._task_manager.start_autostart_tasks()
-            elif isinstance(attr_value, list):
-                for item in attr_value:
-                    if isinstance(item, AbstractDataService):
-                        item._task_manager.start_autostart_tasks()
-
-    def _make_stop_task(self, name: str) -> Callable[..., Any]:
-        """
-        Factory function to create a 'stop_task' function for a running task.
-
-        The generated function cancels the associated asyncio task using 'name' for
-        identification, ensuring proper cleanup. Avoids closure and late binding issues.
-
-        Args:
-            name (str): The name of the coroutine task, used for its identification.
-        """
-
-        def stop_task() -> None:
-            # cancel the task
-            task = self.tasks.get(name, None)
-            if task is not None:
-                self._loop.call_soon_threadsafe(task.cancel)
-
-        return stop_task
-
-    def _make_start_task(
-        self, name: str, method: Callable[..., Any]
-    ) -> Callable[..., Any]:
-        """
-        Factory function to create a 'start_task' function for a coroutine.
-
-        The generated function starts the coroutine as an asyncio task, handling
-        registration and monitoring.
-        It uses 'name' and 'method' to avoid the closure and late binding issue.
-
-        Args:
-            name (str): The name of the coroutine, used for task management.
-            method (callable): The coroutine to be turned into an asyncio task.
-        """
-
-        def start_task() -> None:
-            def task_done_callback(task: asyncio.Task[None], name: str) -> None:
-                """Handles tasks that have finished.
-
-                Removes a task from the tasks dictionary, calls the defined
-                callbacks, and logs and re-raises exceptions."""
-
-                # removing the finished task from the tasks i
-                self.tasks.pop(name, None)
-
-                # emit the notification that the task was stopped
-                self.service._notify_changed(name, None)
-
-                exception = task.exception()
-                if exception is not None:
-                    # Handle the exception, or you can re-raise it.
-                    logger.error(
-                        "Task '%s' encountered an exception: %s: %s",
-                        name,
-                        type(exception).__name__,
-                        exception,
-                    )
-                    raise exception
-
-            async def task() -> None:
-                try:
-                    await method()
-                except asyncio.CancelledError:
-                    logger.info("Task '%s' was cancelled", name)
-
-            if not self.tasks.get(name):
-                # creating the task and adding the task_done_callback which checks
-                # if an exception has occured during the task execution
-                task_object = self._loop.create_task(task())
-                task_object.add_done_callback(
-                    lambda task: task_done_callback(task, name)
-                )
-
-                # Store the task and its arguments in the '__tasks' dictionary. The
-                # key is the name of the method, and the value is a dictionary
-                # containing the task object and the updated keyword arguments.
-                self.tasks[name] = task_object
-
-                # emit the notification that the task was started
-                self.service._notify_changed(name, TaskStatus.RUNNING)
-            else:
-                logger.error("Task '%s' is already running!", name)
-
-        return start_task

src/pydase/frontend/index.html

@@ -6,7 +6,7 @@
     <meta name="viewport" content="width=device-width, initial-scale=1.0" />
     <meta name="theme-color" content="#000000" />
     <meta name="description" content="Web site displaying a pydase UI." />
-  <script type="module" crossorigin src="/assets/index-D7tStNHJ.js"></script>
+  <script type="module" crossorigin src="/assets/index-BjsjosWf.js"></script>
   <link rel="stylesheet" crossorigin href="/assets/index-D2aktF3W.css">
 </head>
 

src/pydase/observer_pattern/observable/decorators.py

@@ -17,10 +17,10 @@ def validate_set(
     getter and check against the desired value.
 
     Args:
-        timeout (float):
+        timeout:
             The maximum time (in seconds) to wait for the value to be within the
             precision boundary.
-        precision (float | None):
+        precision:
             The acceptable deviation from the desired value. If None, the value must be
             exact.
     """
@@ -44,13 +44,11 @@ def has_validate_set_decorator(prop: property) -> bool:
     Checks if a property setter has been decorated with the `validate_set` decorator.
 
     Args:
-        prop (property):
+        prop:
             The property to check.
 
     Returns:
-        bool:
+        True if the property setter has the `validate_set` decorator, False otherwise.
-            True if the property setter has the `validate_set` decorator, False
-            otherwise.
     """
 
     property_setter = prop.fset
@@ -68,11 +66,11 @@ def _validate_value_was_correctly_set(
     specified `precision` and time `timeout`.
 
     Args:
-        obj (Observable):
+        obj:
             The instance of the class containing the property.
-        name (str):
+        name:
             The name of the property to validate.
-        value (Any):
+        value:
             The desired value to check against.
 
     Raises:

src/pydase/observer_pattern/observable/observable.py

@@ -6,7 +6,7 @@ from pydase.observer_pattern.observable.decorators import (
     has_validate_set_decorator,
 )
 from pydase.observer_pattern.observable.observable_object import ObservableObject
-from pydase.utils.helpers import is_property_attribute
+from pydase.utils.helpers import is_descriptor, is_property_attribute
 
 logger = logging.getLogger(__name__)
 
@@ -22,7 +22,9 @@ class Observable(ObservableObject):
             - {"__annotations__"}
         }
         for name, value in class_attrs.items():
-            if isinstance(value, property) or callable(value):
+            if isinstance(value, property) or callable(value) or is_descriptor(value):
+                # Properties, methods and descriptors have to be stored as class
+                # attributes to work properly. So don't make it an instance attribute.
                 continue
             self.__dict__[name] = self._initialise_new_objects(name, value)
 

src/pydase/observer_pattern/observer/observer.py

@@ -24,8 +24,7 @@ class Observer(ABC):
         self.on_change_start(changing_attribute)
 
     @abstractmethod
-    def on_change(self, full_access_path: str, value: Any) -> None:
+    def on_change(self, full_access_path: str, value: Any) -> None: ...
-        ...
 
     def on_change_start(self, full_access_path: str) -> None:
         return

src/pydase/observer_pattern/observer/property_observer.py

@@ -5,6 +5,7 @@ from typing import Any
 
 from pydase.observer_pattern.observable.observable import Observable
 from pydase.observer_pattern.observer.observer import Observer
+from pydase.utils.helpers import is_descriptor
 
 logger = logging.getLogger(__name__)
 
@@ -60,18 +61,28 @@ class PropertyObserver(Observer):
     def _process_nested_observables_properties(
         self, obj: Observable, deps: dict[str, Any], prefix: str
     ) -> None:
-        for k, value in vars(obj).items():
+        for k, value in {**vars(type(obj)), **vars(obj)}.items():
+            actual_value = value
             prefix = (
                 f"{prefix}." if prefix != "" and not prefix.endswith(".") else prefix
             )
             parent_path = f"{prefix}{k}"
-            if isinstance(value, Observable):
+
+            # Get value from descriptor
+            if not isinstance(value, property) and is_descriptor(value):
+                actual_value = getattr(obj, k)
+
+            if isinstance(actual_value, Observable):
                 new_prefix = f"{parent_path}."
                 deps.update(
-                    self._get_properties_and_their_dependencies(value, new_prefix)
+                    self._get_properties_and_their_dependencies(
+                        actual_value, new_prefix
+                    )
                 )
             elif isinstance(value, list | dict):
-                self._process_collection_item_properties(value, deps, parent_path)
+                self._process_collection_item_properties(
+                    actual_value, deps, parent_path
+                )
 
     def _process_collection_item_properties(
         self,

src/pydase/server/server.py

@@ -13,6 +13,8 @@ from pydase.config import ServiceConfig
 from pydase.data_service.data_service_observer import DataServiceObserver
 from pydase.data_service.state_manager import StateManager
 from pydase.server.web_server import WebServer
+from pydase.task.autostart import autostart_service_tasks
+from pydase.utils.helpers import current_event_loop_exists
 
 HANDLED_SIGNALS = (
     signal.SIGINT,  # Unix signal 2. Sent by Ctrl+C.
@@ -35,18 +37,18 @@ class AdditionalServerProtocol(Protocol):
 
     Args:
         data_service_observer:
-          Observer for the DataService, handling state updates and communication to
+            Observer for the DataService, handling state updates and communication to
-          connected clients through injected callbacks. Can be utilized to access the
+            connected clients through injected callbacks. Can be utilized to access the
-          service and state manager, and to add custom state-update callbacks.
+            service and state manager, and to add custom state-update callbacks.
         host:
-          Hostname or IP address where the server is accessible. Commonly '0.0.0.0' to
+            Hostname or IP address where the server is accessible. Commonly '0.0.0.0' to
-          bind to all network interfaces.
+            bind to all network interfaces.
         port:
-          Port number on which the server listens. Typically in the range 1024-65535
+            Port number on which the server listens. Typically in the range 1024-65535
-          (non-standard ports).
+            (non-standard ports).
         **kwargs:
-          Any additional parameters required for initializing the server. These
+            Any additional parameters required for initializing the server. These
-          parameters are specific to the server's implementation.
+            parameters are specific to the server's implementation.
     """
 
     def __init__(
@@ -64,18 +66,17 @@ class AdditionalServerProtocol(Protocol):
 
 
 class AdditionalServer(TypedDict):
-    """
+    """A TypedDict that represents the configuration for an additional server to be run
-    A TypedDict that represents the configuration for an additional server to be run
     alongside the main server.
-
-    This class is used to specify the server type, the port on which the server should
-    run, and any additional keyword arguments that should be passed to the server when
-    it's instantiated.
     """
 
     server: type[AdditionalServerProtocol]
+    """Server adhering to the
+    [`AdditionalServerProtocol`][pydase.server.server.AdditionalServerProtocol]."""
     port: int
+    """Port on which the server should run."""
     kwargs: dict[str, Any]
+    """Additional keyword arguments that will be passed to the server's constructor """
 
 
 class Server:
@@ -83,29 +84,20 @@ class Server:
     The `Server` class provides a flexible server implementation for the `DataService`.
 
     Args:
-        service: DataService
+        service:
             The DataService instance that this server will manage.
-        host: str
+        host:
-            The host address for the server. Default is '0.0.0.0', which means all
+            The host address for the server. Defaults to `'0.0.0.0'`, which means all
             available network interfaces.
-        web_port: int
+        web_port:
-            The port number for the web server. Default is
+            The port number for the web server. Defaults to
-            `pydase.config.ServiceConfig().web_port`.
+            [`ServiceConfig().web_port`][pydase.config.ServiceConfig.web_port].
-        enable_web: bool
+        enable_web:
-            Whether to enable the web server. Default is True.
+            Whether to enable the web server.
-        filename: str | Path | None
+        filename:
             Filename of the file managing the service state persistence.
-            Defaults to None.
+        additional_servers:
-        additional_servers : list[AdditionalServer]
+            A list of additional servers to run alongside the main server.
-            A list of additional servers to run alongside the main server. Each entry in
-            the list should be a dictionary with the following structure:
-                - server: A class that adheres to the AdditionalServerProtocol. This
-                    class should have an `__init__` method that accepts the DataService
-                    instance, port, host, and optional keyword arguments, and a `serve`
-                    method that is a coroutine responsible for starting the server.
-                - port: The port on which the additional server will be running.
-                - kwargs: A dictionary containing additional keyword arguments that will
-                    be passed to the server's `__init__` method.
 
             Here's an example of how you might define an additional server:
 
@@ -145,8 +137,8 @@ class Server:
             )
             server.run()
             ```
-        **kwargs: Any
+        **kwargs:
-          Additional keyword arguments.
+            Additional keyword arguments.
     """
 
     def __init__(  # noqa: PLR0913
@@ -166,13 +158,18 @@ class Server:
         self._web_port = web_port
         self._enable_web = enable_web
         self._kwargs = kwargs
-        self._loop: asyncio.AbstractEventLoop
         self._additional_servers = additional_servers
         self.should_exit = False
         self.servers: dict[str, asyncio.Future[Any]] = {}
         self._state_manager = StateManager(self._service, filename)
         self._observer = DataServiceObserver(self._state_manager)
         self._state_manager.load_state()
+        autostart_service_tasks(self._service)
+        if not current_event_loop_exists():
+            self._loop = asyncio.new_event_loop()
+            asyncio.set_event_loop(self._loop)
+        else:
+            self._loop = asyncio.get_event_loop()
 
     def run(self) -> None:
         """
@@ -180,7 +177,7 @@ class Server:
 
         This method should be called to start the server after it's been instantiated.
         """
-        asyncio.run(self.serve())
+        self._loop.run_until_complete(self.serve())
 
     async def serve(self) -> None:
         process_id = os.getpid()
@@ -196,10 +193,8 @@ class Server:
         logger.info("Finished server process [%s]", process_id)
 
     async def startup(self) -> None:
-        self._loop = asyncio.get_running_loop()
         self._loop.set_exception_handler(self.custom_exception_handler)
         self.install_signal_handlers()
-        self._service._task_manager.start_autostart_tasks()
 
         for server in self._additional_servers:
             addin_server = server["server"](
@@ -214,7 +209,7 @@ class Server:
             )
 
             server_task = self._loop.create_task(addin_server.serve())
-            server_task.add_done_callback(self.handle_server_shutdown)
+            server_task.add_done_callback(self._handle_server_shutdown)
             self.servers[server_name] = server_task
         if self._enable_web:
             self._web_server = WebServer(
@@ -225,10 +220,10 @@ class Server:
             )
             server_task = self._loop.create_task(self._web_server.serve())
 
-            server_task.add_done_callback(self.handle_server_shutdown)
+            server_task.add_done_callback(self._handle_server_shutdown)
             self.servers["web"] = server_task
 
-    def handle_server_shutdown(self, task: asyncio.Task[Any]) -> None:
+    def _handle_server_shutdown(self, task: asyncio.Task[Any]) -> None:
         """Handle server shutdown. If the service should exit, do nothing. Else, make
         the service exit."""
 

src/pydase/server/web_server/sio_setup.py

@@ -23,6 +23,7 @@ logger = logging.getLogger(__name__)
 
 # These functions can be monkey-patched by other libraries at runtime
 dump = pydase.utils.serialization.serializer.dump
+sio_client_manager = None
 
 
 class UpdateDict(TypedDict):
@@ -54,12 +55,15 @@ class RunMethodDict(TypedDict):
     exposed DataService.
 
     Attributes:
-        name (str): The name of the method to be run.
+        name:
-        parent_path (str): The access path for the parent object of the method to be
+            The name of the method to be run.
-            run. This is used to construct the full access path for the method. For
+        parent_path:
-            example, for an method with access path 'attr1.list_attr[0].method_name',
+            The access path for the parent object of the method to be run. This is used
-            'attr1.list_attr[0]' would be the parent_path.
+            to construct the full access path for the method. For example, for an method
-        kwargs (dict[str, Any]): The arguments passed to the method.
+            with access path 'attr1.list_attr[0].method_name', 'attr1.list_attr[0]'
+            would be the parent_path.
+        kwargs:
+            The arguments passed to the method.
     """
 
     name: str
@@ -76,23 +80,30 @@ def setup_sio_server(
     Sets up and configures a Socket.IO asynchronous server.
 
     Args:
-        observer (DataServiceObserver):
+        observer:
-          The observer managing state updates and communication.
+            The observer managing state updates and communication.
-        enable_cors (bool):
+        enable_cors:
-          Flag indicating whether CORS should be enabled for the server.
+            Flag indicating whether CORS should be enabled for the server.
-        loop (asyncio.AbstractEventLoop):
+        loop:
-          The event loop in which the server will run.
+            The event loop in which the server will run.
 
     Returns:
-        socketio.AsyncServer: The configured Socket.IO asynchronous server.
+        The configured Socket.IO asynchronous server.
     """
 
     state_manager = observer.state_manager
 
     if enable_cors:
-        sio = socketio.AsyncServer(async_mode="aiohttp", cors_allowed_origins="*")
+        sio = socketio.AsyncServer(
+            async_mode="aiohttp",
+            cors_allowed_origins="*",
+            client_manager=sio_client_manager,
+        )
     else:
-        sio = socketio.AsyncServer(async_mode="aiohttp")
+        sio = socketio.AsyncServer(
+            async_mode="aiohttp",
+            client_manager=sio_client_manager,
+        )
 
     setup_sio_events(sio, state_manager)
     setup_logging_handler(sio)
@@ -127,15 +138,15 @@ def setup_sio_server(
 def setup_sio_events(sio: socketio.AsyncServer, state_manager: StateManager) -> None:  # noqa: C901
     @sio.event  # type: ignore
     async def connect(sid: str, environ: Any) -> None:
-        logging.debug("Client [%s] connected", click.style(str(sid), fg="cyan"))
+        logger.debug("Client [%s] connected", click.style(str(sid), fg="cyan"))
 
     @sio.event  # type: ignore
     async def disconnect(sid: str) -> None:
-        logging.debug("Client [%s] disconnected", click.style(str(sid), fg="cyan"))
+        logger.debug("Client [%s] disconnected", click.style(str(sid), fg="cyan"))
 
     @sio.event  # type: ignore
     async def service_serialization(sid: str) -> SerializedObject:
-        logging.debug(
+        logger.debug(
             "Client [%s] requested service serialization",
             click.style(str(sid), fg="cyan"),
         )

src/pydase/server/web_server/web_server.py

@@ -25,41 +25,51 @@ API_VERSION = "v1"
 
 class WebServer:
     """
-    Represents a web server that adheres to the AdditionalServerProtocol, designed to
+    Represents a web server that adheres to the
-    work with a DataService instance. This server facilitates client-server
+    [`AdditionalServerProtocol`][pydase.server.server.AdditionalServerProtocol],
-    communication and state management through web protocols and socket connections.
+    designed to work with a [`DataService`][pydase.DataService] instance. This server
+    facilitates client-server communication and state management through web protocols
+    and socket connections.
 
-    The WebServer class initializes and manages a web server environment using FastAPI
+    The WebServer class initializes and manages a web server environment aiohttp and
-    and Socket.IO, allowing for HTTP and WebSocket communications. It incorporates CORS
+    Socket.IO, allowing for HTTP and Socket.IO communications. It incorporates CORS
-    (Cross-Origin Resource Sharing) support, custom CSS, and serves a frontend static
+    (Cross-Origin Resource Sharing) support, custom CSS, and serves a static files
-    files directory. It also initializes web server settings based on configuration
+    directory. It also initializes web server settings based on configuration files or
-    files or generates default settings if necessary.
+    generates default settings if necessary.
 
     Configuration for the web server (like service configuration directory and whether
     to generate new web settings) is determined in the following order of precedence:
+
     1. Values provided directly to the constructor.
     2. Environment variable settings (via configuration classes like
-      `pydase.config.ServiceConfig` and `pydase.config.WebServerConfig`).
+      [`ServiceConfig`][pydase.config.ServiceConfig] and
+      [`WebServerConfig`][pydase.config.WebServerConfig]).
     3. Default values defined in the configuration classes.
 
     Args:
-        data_service_observer (DataServiceObserver): Observer for the DataService,
+        data_service_observer:
-          handling state updates and communication to connected clients.
+            Observer for the [`DataService`][pydase.DataService], handling state updates
-        host (str): Hostname or IP address where the server is accessible. Commonly
+            and communication to connected clients.
-          '0.0.0.0' to bind to all network interfaces.
+        host:
-        port (int): Port number on which the server listens. Typically in the range
+            Hostname or IP address where the server is accessible. Commonly '0.0.0.0'
-          1024-65535 (non-standard ports).
+            to bind to all network interfaces.
-        css (str | Path | None, optional): Path to a custom CSS file for styling the
+        port:
-          frontend. If None, no custom styles are applied. Defaults to None.
+            Port number on which the server listens. Typically in the range 1024-65535
-        enable_cors (bool, optional): Flag to enable or disable CORS policy. When True,
+            (non-standard ports).
-          CORS is enabled, allowing cross-origin requests. Defaults to True.
+        css:
-        config_dir (Path | None, optional): Path to the configuration
+            Path to a custom CSS file for styling the frontend. If None, no custom
-          directory where the web settings will be stored. Defaults to
+            styles are applied. Defaults to None.
-          `pydase.config.ServiceConfig().config_dir`.
+        enable_cors:
-        generate_new_web_settings (bool | None, optional): Flag to enable or disable
+            Flag to enable or disable CORS policy. When True, CORS is enabled, allowing
-          generation of new web settings if the configuration file is missing. Defaults
+            cross-origin requests. Defaults to True.
-          to `pydase.config.WebServerConfig().generate_new_web_settings`.
+        config_dir:
-        **kwargs (Any): Additional unused keyword arguments.
+            Path to the configuration directory where the web settings will be stored.
+            Defaults to
+            [`ServiceConfig().config_dir`][pydase.config.ServiceConfig.config_dir].
+        generate_web_settings:
+            Flag to enable or disable generation of new web settings if the
+            configuration file is missing. Defaults to
+            [`WebServerConfig().generate_web_settings`][pydase.config.WebServerConfig.generate_web_settings].
     """
 
     def __init__(  # noqa: PLR0913

src/pydase/task/autostart.py

@@ -0,0 +1,46 @@
+from typing import Any
+
+import pydase.data_service.data_service
+import pydase.task.task
+from pydase.task.task_status import TaskStatus
+from pydase.utils.helpers import is_property_attribute
+
+
+def autostart_service_tasks(
+    service: pydase.data_service.data_service.DataService,
+) -> None:
+    """Starts the service tasks defined with the `autostart` keyword argument.
+
+    This method goes through the attributes of the passed service and its nested
+    [`DataService`][pydase.DataService] instances and calls the start method on
+    autostart-tasks.
+    """
+
+    for attr in dir(service):
+        if is_property_attribute(service, attr) or attr in {
+            "_observers",
+            "__dict__",
+        }:  # prevent eval of property attrs and recursion
+            continue
+
+        val = getattr(service, attr)
+        if isinstance(val, pydase.task.task.Task):
+            if val.autostart and val.status == TaskStatus.NOT_RUNNING:
+                val.start()
+            else:
+                continue
+        else:
+            autostart_nested_service_tasks(val)
+
+
+def autostart_nested_service_tasks(
+    service: pydase.data_service.data_service.DataService | list[Any] | dict[Any, Any],
+) -> None:
+    if isinstance(service, pydase.DataService):
+        autostart_service_tasks(service)
+    elif isinstance(service, list):
+        for entry in service:
+            autostart_nested_service_tasks(entry)
+    elif isinstance(service, dict):
+        for entry in service.values():
+            autostart_nested_service_tasks(entry)

src/pydase/task/decorator.py

@@ -0,0 +1,145 @@
+import logging
+from collections.abc import Callable, Coroutine
+from typing import Any, Generic, TypeVar, overload
+
+from pydase.data_service.data_service import DataService
+from pydase.task.task import Task
+
+logger = logging.getLogger(__name__)
+
+R = TypeVar("R")
+
+
+class PerInstanceTaskDescriptor(Generic[R]):
+    """
+    A descriptor class that provides a unique [`Task`][pydase.task.task.Task] object
+    for each instance of a [`DataService`][pydase.data_service.data_service.DataService]
+    class.
+
+    The `PerInstanceTaskDescriptor` is used to transform an asynchronous function into a
+    task that is managed independently for each instance of a `DataService` subclass.
+    This allows tasks to be initialized, started, and stopped on a per-instance basis,
+    providing better control over task execution within the service.
+
+    The `PerInstanceTaskDescriptor` is not intended to be used directly. Instead, it is
+    used internally by the `@task` decorator to manage task objects for each instance of
+    the service class.
+    """
+
+    def __init__(
+        self,
+        func: Callable[[Any], Coroutine[None, None, R]]
+        | Callable[[], Coroutine[None, None, R]],
+        autostart: bool = False,
+    ) -> None:
+        self.__func = func
+        self.__autostart = autostart
+        self.__task_instances: dict[object, Task[R]] = {}
+
+    def __set_name__(self, owner: type[DataService], name: str) -> None:
+        """Stores the name of the task within the owning class. This method is called
+        automatically when the descriptor is assigned to a class attribute.
+        """
+
+        self.__task_name = name
+
+    @overload
+    def __get__(
+        self, instance: None, owner: type[DataService]
+    ) -> "PerInstanceTaskDescriptor[R]":
+        """Returns the descriptor itself when accessed through the class."""
+
+    @overload
+    def __get__(self, instance: DataService, owner: type[DataService]) -> Task[R]:
+        """Returns the `Task` object associated with the specific `DataService`
+        instance.
+        If no task exists for the instance, a new `Task` object is created and stored
+        in the `__task_instances` dictionary.
+        """
+
+    def __get__(
+        self, instance: DataService | None, owner: type[DataService]
+    ) -> "Task[R] | PerInstanceTaskDescriptor[R]":
+        if instance is None:
+            return self
+
+        # Create a new Task object for this instance, using the function's name.
+        if instance not in self.__task_instances:
+            self.__task_instances[instance] = instance._initialise_new_objects(
+                self.__task_name,
+                Task(self.__func.__get__(instance, owner), autostart=self.__autostart),
+            )
+
+        return self.__task_instances[instance]
+
+
+def task(
+    *, autostart: bool = False
+) -> Callable[
+    [
+        Callable[[Any], Coroutine[None, None, R]]
+        | Callable[[], Coroutine[None, None, R]]
+    ],
+    PerInstanceTaskDescriptor[R],
+]:
+    """
+    A decorator to define an asynchronous function as a per-instance task within a
+    [`DataService`][pydase.DataService] class.
+
+    This decorator transforms an asynchronous function into a
+    [`Task`][pydase.task.task.Task] object that is unique to each instance of the
+    `DataService` class. The resulting `Task` object provides methods like `start()`
+    and `stop()` to control the execution of the task, and manages the task's lifecycle
+    independently for each instance of the service.
+
+    The decorator is particularly useful for defining tasks that need to run
+    periodically or perform asynchronous operations, such as polling data sources,
+    updating databases, or any recurring job that should be managed within the context
+    of a `DataService`.
+    time.
+
+    Args:
+        autostart:
+            If set to True, the task will automatically start when the service is
+            initialized. Defaults to False.
+
+    Returns:
+        A decorator that wraps an asynchronous function in a
+        [`PerInstanceTaskDescriptor`][pydase.task.decorator.PerInstanceTaskDescriptor]
+        object, which, when accessed, provides an instance-specific
+        [`Task`][pydase.task.task.Task] object.
+
+    Example:
+        ```python
+        import asyncio
+
+        import pydase
+        from pydase.task.decorator import task
+
+
+        class MyService(pydase.DataService):
+            @task(autostart=True)
+            async def my_task(self) -> None:
+                while True:
+                    # Perform some periodic work
+                    await asyncio.sleep(1)
+
+
+        if __name__ == "__main__":
+            service = MyService()
+            pydase.Server(service=service).run()
+        ```
+
+        In this example, `my_task` is defined as a task using the `@task` decorator, and
+        it will start automatically when the service is initialized because
+        `autostart=True` is set. You can manually start or stop the task using
+        `service.my_task.start()` and `service.my_task.stop()`, respectively.
+    """
+
+    def decorator(
+        func: Callable[[Any], Coroutine[None, None, R]]
+        | Callable[[], Coroutine[None, None, R]],
+    ) -> PerInstanceTaskDescriptor[R]:
+        return PerInstanceTaskDescriptor(func, autostart=autostart)
+
+    return decorator

src/pydase/task/task.py

@@ -0,0 +1,148 @@
+import asyncio
+import inspect
+import logging
+from collections.abc import Callable, Coroutine
+from typing import (
+    Generic,
+    TypeVar,
+)
+
+import pydase.data_service.data_service
+from pydase.task.task_status import TaskStatus
+from pydase.utils.helpers import current_event_loop_exists
+
+logger = logging.getLogger(__name__)
+
+R = TypeVar("R")
+
+
+class Task(pydase.data_service.data_service.DataService, Generic[R]):
+    """A class representing a task within the `pydase` framework.
+
+    The `Task` class wraps an asynchronous function and provides methods to manage its
+    lifecycle, such as `start()` and `stop()`. It is typically used to perform periodic
+    or recurring jobs in a [`DataService`][pydase.DataService], like reading
+    sensor data, updating databases, or executing other background tasks.
+
+    When a function is decorated with the [`@task`][pydase.task.decorator.task]
+    decorator, it is replaced by a `Task` instance that controls the execution of the
+    original function.
+
+    Args:
+        func:
+            The asynchronous function that this task wraps. It must be a coroutine
+            without arguments.
+        autostart:
+            If set to True, the task will automatically start when the service is
+            initialized. Defaults to False.
+
+    Example:
+        ```python
+        import asyncio
+
+        import pydase
+        from pydase.task.decorator import task
+
+
+        class MyService(pydase.DataService):
+            @task(autostart=True)
+            async def my_task(self) -> None:
+                while True:
+                    # Perform some periodic work
+                    await asyncio.sleep(1)
+
+
+        if __name__ == "__main__":
+            service = MyService()
+            pydase.Server(service=service).run()
+        ```
+
+        In this example, `my_task` is defined as a task using the `@task` decorator, and
+        it will start automatically when the service is initialized because
+        `autostart=True` is set. You can manually start or stop the task using
+        `service.my_task.start()` and `service.my_task.stop()`, respectively.
+    """
+
+    def __init__(
+        self,
+        func: Callable[[], Coroutine[None, None, R | None]],
+        *,
+        autostart: bool = False,
+    ) -> None:
+        super().__init__()
+        self._autostart = autostart
+        self._func_name = func.__name__
+        self._func = func
+        self._task: asyncio.Task[R | None] | None = None
+        self._status = TaskStatus.NOT_RUNNING
+        self._result: R | None = None
+
+        if not current_event_loop_exists():
+            self._loop = asyncio.new_event_loop()
+            asyncio.set_event_loop(self._loop)
+        else:
+            self._loop = asyncio.get_event_loop()
+
+    @property
+    def autostart(self) -> bool:
+        """Defines if the task should be started automatically when the
+        [`Server`][pydase.Server] starts."""
+        return self._autostart
+
+    @property
+    def status(self) -> TaskStatus:
+        """Returns the current status of the task."""
+        return self._status
+
+    def start(self) -> None:
+        """Starts the asynchronous task if it is not already running."""
+        if self._task:
+            return
+
+        def task_done_callback(task: asyncio.Task[R | None]) -> None:
+            """Handles tasks that have finished.
+
+            Updates the task status, calls the defined callbacks, and logs and re-raises
+            exceptions.
+            """
+
+            self._task = None
+            self._status = TaskStatus.NOT_RUNNING
+
+            exception = task.exception()
+            if exception is not None:
+                # Handle the exception, or you can re-raise it.
+                logger.error(
+                    "Task '%s' encountered an exception: %s: %s",
+                    self._func_name,
+                    type(exception).__name__,
+                    exception,
+                )
+                raise exception
+
+            self._result = task.result()
+
+        async def run_task() -> R | None:
+            if inspect.iscoroutinefunction(self._func):
+                logger.info("Starting task %r", self._func_name)
+                self._status = TaskStatus.RUNNING
+                res: Coroutine[None, None, R | None] = self._func()
+                try:
+                    return await res
+                except asyncio.CancelledError:
+                    logger.info("Task '%s' was cancelled", self._func_name)
+                    return None
+            logger.warning(
+                "Cannot start task %r. Function has not been bound yet", self._func_name
+            )
+            return None
+
+        logger.info("Creating task %r", self._func_name)
+        self._task = self._loop.create_task(run_task())
+        self._task.add_done_callback(task_done_callback)
+
+    def stop(self) -> None:
+        """Stops the running asynchronous task by cancelling it."""
+
+        if self._task:
+            self._task.cancel()

src/pydase/task/task_status.py

@@ -0,0 +1,8 @@
+import enum
+
+
+class TaskStatus(enum.Enum):
+    """Possible statuses of a [`Task`][pydase.task.task.Task]."""
+
+    RUNNING = "running"
+    NOT_RUNNING = "not_running"

src/pydase/units.py

@@ -21,18 +21,20 @@ def convert_to_quantity(
     Convert a given value into a pint.Quantity object with the specified unit.
 
     Args:
-        value (QuantityDict | float | int | Quantity):
+        value:
             The value to be converted into a Quantity object.
+
             - If value is a float or int, it will be directly converted to the specified
               unit.
             - If value is a dict, it must have keys 'magnitude' and 'unit' to represent
               the value and unit.
             - If value is a Quantity object, it will remain unchanged.\n
-        unit (str, optional): The target unit for conversion. If empty and value is not
+        unit:
-            a Quantity object, it will assume a unitless quantity.
+            The target unit for conversion. If empty and value is not a Quantity object,
+            it will assume a unitless quantity.
 
     Returns:
-        Quantity: The converted value as a pint.Quantity object with the specified unit.
+        The converted value as a pint.Quantity object with the specified unit.
 
     Examples:
         >>> convert_to_quantity(5, 'm')
@@ -42,9 +44,9 @@ def convert_to_quantity(
         >>> convert_to_quantity(10.0 * u.units.V)
         <Quantity(10.0, 'volt')>
 
-    Notes:
+    Note:
-        - If unit is not provided and value is a float or int, the resulting Quantity
+        If unit is not provided and value is a float or int, the resulting Quantity will
-          will be unitless.
+        be unitless.
     """
 
     if isinstance(value, int | float):

src/pydase/utils/decorators.py

@@ -10,9 +10,9 @@ class FunctionDefinitionError(Exception):
 
 
 def frontend(func: Callable[..., Any]) -> Callable[..., Any]:
-    """
+    """Decorator to mark a [`DataService`][pydase.DataService] method for frontend
-    Decorator to mark a DataService method for frontend rendering. Ensures that the
+    rendering. Ensures that the method does not contain arguments, as they are not
-    method does not contain arguments, as they are not supported for frontend rendering.
+    supported for frontend rendering.
     """
 
     if function_has_arguments(func):

src/pydase/utils/helpers.py

@@ -114,8 +114,6 @@ def get_class_and_instance_attributes(obj: object) -> dict[str, Any]:
 
     If an attribute exists at both the instance and class level,the value from the
     instance attribute takes precedence.
-    The __root__ object is removed as this will lead to endless recursion in the for
-    loops.
     """
 
     return dict(chain(type(obj).__dict__.items(), obj.__dict__.items()))
@@ -162,6 +160,12 @@ def get_object_attr_from_path(target_obj: Any, path: str) -> Any:
     return get_object_by_path_parts(target_obj, path_parts)
 
 
+def get_task_class() -> type:
+    from pydase.task.task import Task
+
+    return Task
+
+
 def get_component_classes() -> list[type]:
     """
     Returns references to the component classes in a list.
@@ -196,3 +200,48 @@ def function_has_arguments(func: Callable[..., Any]) -> bool:
 
     # Check if there are any parameters left which would indicate additional arguments.
     return len(parameters) > 0
+
+
+def is_descriptor(obj: object) -> bool:
+    """Check if an object is a descriptor."""
+
+    # Exclude functions, methods, builtins and properties
+    if (
+        inspect.isfunction(obj)
+        or inspect.ismethod(obj)
+        or inspect.isbuiltin(obj)
+        or isinstance(obj, property)
+    ):
+        return False
+
+    # Check if it has any descriptor methods
+    return any(hasattr(obj, method) for method in ("__get__", "__set__", "__delete__"))
+
+
+def current_event_loop_exists() -> bool:
+    """Check if an event loop has been set."""
+    import asyncio
+
+    return asyncio.get_event_loop_policy()._local._loop is not None  # type: ignore
+
+
+def normalize_full_access_path_string(s: str) -> str:
+    """Normalizes a string representing a full access path by converting double quotes
+    to single quotes.
+
+    This function is useful for ensuring consistency in strings that represent access
+    paths containing dictionary keys, by replacing all double quotes (`"`) with single
+    quotes (`'`).
+
+    Args:
+        s (str): The input string to be normalized.
+
+    Returns:
+        A new string with all double quotes replaced by single quotes.
+
+    Example:
+        >>> normalize_full_access_path_string('dictionary["first"].my_task')
+        "dictionary['first'].my_task"
+    """
+
+    return s.replace('"', "'")

src/pydase/utils/serialization/deserializer.py

@@ -6,7 +6,9 @@ from typing import TYPE_CHECKING, Any, NoReturn, cast
 import pydase
 import pydase.components
 import pydase.units as u
-from pydase.utils.helpers import get_component_classes
+from pydase.utils.helpers import (
+    get_component_classes,
+)
 from pydase.utils.serialization.types import (
     SerializedDatetime,
     SerializedException,
@@ -23,6 +25,7 @@ logger = logging.getLogger(__name__)
 class Deserializer:
     @classmethod
     def deserialize(cls, serialized_object: SerializedObject) -> Any:
+        """Deserialize `serialized_object` (a `dict`) to a Python object."""
         type_handler: dict[str | None, None | Callable[..., Any]] = {
             None: None,
             "int": cls.deserialize_primitive,
@@ -48,9 +51,9 @@ class Deserializer:
             return handler(serialized_object)
 
         # Custom types like Components or DataService classes
-        component_class = cls.get_component_class(serialized_object["type"])
+        service_base_class = cls.get_service_base_class(serialized_object["type"])
-        if component_class:
+        if service_base_class:
-            return cls.deserialize_component_type(serialized_object, component_class)
+            return cls.deserialize_data_service(serialized_object, service_base_class)
 
         return None
 
@@ -109,11 +112,11 @@ class Deserializer:
         raise exception(serialized_object["value"])
 
     @staticmethod
-    def get_component_class(type_name: str | None) -> type | None:
+    def get_service_base_class(type_name: str | None) -> type | None:
         for component_class in get_component_classes():
             if type_name == component_class.__name__:
                 return component_class
-        if type_name == "DataService":
+        if type_name in ("DataService", "Task"):
             import pydase
 
             return pydase.DataService
@@ -136,7 +139,7 @@ class Deserializer:
         return property(get, set)
 
     @classmethod
-    def deserialize_component_type(
+    def deserialize_data_service(
         cls, serialized_object: SerializedObject, base_class: type
     ) -> Any:
         def create_proxy_class(serialized_object: SerializedObject) -> type:
@@ -159,4 +162,5 @@ class Deserializer:
 
 
 def loads(serialized_object: SerializedObject) -> Any:
+    """Deserialize `serialized_object` (a `dict`) to a Python object."""
     return Deserializer.deserialize(serialized_object)

src/pydase/utils/serialization/serializer.py

@@ -9,12 +9,14 @@ from typing import TYPE_CHECKING, Any, Literal, cast
 
 import pydase.units as u
 from pydase.data_service.abstract_data_service import AbstractDataService
-from pydase.data_service.task_manager import TaskStatus
+from pydase.task.task_status import TaskStatus
 from pydase.utils.decorators import render_in_frontend
 from pydase.utils.helpers import (
     get_attribute_doc,
     get_component_classes,
     get_data_service_class_reference,
+    get_task_class,
+    is_property_attribute,
     parse_full_access_path,
     parse_serialized_key,
 )
@@ -52,8 +54,27 @@ class SerializationPathError(Exception):
 
 
 class Serializer:
+    """Serializes objects into
+    [`SerializedObject`][pydase.utils.serialization.types.SerializedObject]
+    representations.
+    """
+
     @classmethod
     def serialize_object(cls, obj: Any, access_path: str = "") -> SerializedObject:  # noqa: C901
+        """Serialize `obj` to a
+        [`SerializedObject`][pydase.utils.serialization.types.SerializedObject].
+
+        Args:
+            obj:
+                Object to be serialized.
+            access_path:
+                String corresponding to the full access path of the object. This will be
+                prepended to the full_access_path in the SerializedObject entries.
+
+        Returns:
+            Dictionary representation of `obj`.
+        """
+
         result: SerializedObject
 
         if isinstance(obj, Exception):
@@ -261,6 +282,10 @@ class Serializer:
         if component_base_cls:
             obj_type = component_base_cls.__name__  # type: ignore
 
+        elif isinstance(obj, get_task_class()):
+            # Check if obj is a pydase task
+            obj_type = "Task"
+
         # Get the set of DataService class attributes
         data_service_attr_set = set(dir(get_data_service_class_reference()))
         # Get the set of the object attributes
@@ -275,29 +300,15 @@ class Serializer:
             if key.startswith("_"):
                 continue  # Skip attributes that start with underscore
 
-            # Skip keys that start with "start_" or "stop_" and end with an async
-            # method name
-            if key.startswith(("start_", "stop_")) and key.split("_", 1)[1] in {
-                name
-                for name, _ in inspect.getmembers(
-                    obj, predicate=inspect.iscoroutinefunction
-                )
-            }:
-                continue
-
             val = getattr(obj, key)
 
             path = f"{access_path}.{key}" if access_path else key
             serialized_object = cls.serialize_object(val, access_path=path)
 
-            # If there's a running task for this method
-            if serialized_object["type"] == "method" and key in obj._task_manager.tasks:
-                serialized_object["value"] = TaskStatus.RUNNING.name
-
             value[key] = serialized_object
 
             # If the DataService attribute is a property
-            if isinstance(getattr(obj.__class__, key, None), property):
+            if is_property_attribute(obj, key):
                 prop: property = getattr(obj.__class__, key)
                 value[key]["readonly"] = prop.fset is None
                 value[key]["doc"] = get_attribute_doc(prop)  # overwrite the doc
@@ -313,6 +324,19 @@ class Serializer:
 
 
 def dump(obj: Any) -> SerializedObject:
+    """Serialize `obj` to a
+    [`SerializedObject`][pydase.utils.serialization.types.SerializedObject].
+
+    The [`Serializer`][pydase.utils.serialization.serializer.Serializer] is used for
+    encoding.
+
+    Args:
+        obj:
+            Object to be serialized.
+
+    Returns:
+        Dictionary representation of `obj`.
+    """
     return Serializer.serialize_object(obj)
 
 
@@ -321,12 +345,13 @@ def set_nested_value_by_path(
 ) -> None:
     """
     Set a value in a nested dictionary structure, which conforms to the serialization
-    format used by `pydase.utils.serializer.Serializer`, using a dot-notation path.
+    format used by [`Serializer`][pydase.utils.serialization.serializer.Serializer],
+    using a dot-notation path.
 
     Args:
         serialization_dict:
             The base dictionary representing data serialized with
-            `pydase.utils.serializer.Serializer`.
+            [`Serializer`][pydase.utils.serialization.serializer.Serializer].
         path:
             The dot-notation path (e.g., 'attr1.attr2[0].attr3') indicating where to
             set the value.
@@ -334,8 +359,8 @@ def set_nested_value_by_path(
             The new value to set at the specified path.
 
     Note:
-        - If the index equals the length of the list, the function will append the
+        If the index equals the length of the list, the function will append the
-          serialized representation of the 'value' to the list.
+        serialized representation of the 'value' to the list.
     """
 
     path_parts = parse_full_access_path(path)
@@ -438,26 +463,24 @@ def get_container_item_by_key(
 ) -> SerializedObject:
     """
     Retrieve an item from a container specified by the passed key. Add an item to the
-    container if allow_append is set to True.
+    container if `allow_append` is set to `True`.
 
     If specified keys or indexes do not exist, the function can append new elements to
     dictionaries and to lists if `allow_append` is True and the missing element is
     exactly the next sequential index (for lists).
 
     Args:
-        container: dict[str, SerializedObject] | list[SerializedObject]
+        container:
             The container representing serialized data.
-        key: str
+        key:
             The key name representing the attribute in the dictionary, which may include
             direct keys or indexes (e.g., 'attr_name', '["key"]' or '[0]').
-        allow_append: bool
+        allow_append:
             Flag to allow appending a new entry if the specified index is out of range
             by exactly one position.
 
     Returns:
-        SerializedObject
+        The dictionary or list item corresponding to the specified attribute and index.
-            The dictionary or list item corresponding to the specified attribute and
-            index.
 
     Raises:
         SerializationPathError:
@@ -485,13 +508,12 @@ def get_data_paths_from_serialized_object(  # noqa: C901
     Recursively extracts full access paths from a serialized object.
 
     Args:
-        serialized_obj (SerializedObject):
+        serialized_obj:
             The dictionary representing the serialization of an object. Produced by
             `pydase.utils.serializer.Serializer`.
 
     Returns:
-        list[str]:
+        A list of strings, each representing a full access path in the serialized
-            A list of strings, each representing a full access path in the serialized
             object.
     """
 
@@ -532,12 +554,11 @@ def generate_serialized_data_paths(
     Recursively extracts full access paths from a serialized DataService class instance.
 
     Args:
-        data (dict[str, SerializedObject]):
+        data:
             The value of the "value" key of a serialized DataService class instance.
 
     Returns:
-        list[str]:
+        A list of strings, each representing a full access path in the serialized
-            A list of strings, each representing a full access path in the serialized
             object.
     """
 
@@ -556,3 +577,6 @@ def serialized_dict_is_nested_object(serialized_dict: SerializedObject) -> bool:
     # We are excluding Quantity here as the value corresponding to the "value" key is
     # a dictionary of the form {"magnitude": ..., "unit": ...}
     return serialized_dict["type"] != "Quantity" and (isinstance(value, dict | list))
+
+
+__all__ = ["Serializer", "dump"]

src/pydase/utils/serialization/types.py

@@ -98,7 +98,9 @@ class SerializedException(SerializedObjectBase):
     type: Literal["Exception"]
 
 
-DataServiceTypes = Literal["DataService", "Image", "NumberSlider", "DeviceConnection"]
+DataServiceTypes = Literal[
+    "DataService", "Image", "NumberSlider", "DeviceConnection", "Task"
+]
 
 
 class SerializedDataService(SerializedObjectBase):
@@ -123,3 +125,21 @@ SerializedObject = (
     | SerializedQuantity
     | SerializedNoValue
 )
+"""
+This type can be any of the following:
+
+- SerializedBool
+- SerializedFloat
+- SerializedInteger
+- SerializedString
+- SerializedDatetime
+- SerializedList
+- SerializedDict
+- SerializedNoneType
+- SerializedMethod
+- SerializedException
+- SerializedDataService
+- SerializedEnum
+- SerializedQuantity
+- SerializedNoValue
+"""

tests/components/test_device_connection.py

@@ -3,6 +3,7 @@ import asyncio
 import pydase
 import pydase.components.device_connection
 import pytest
+from pydase.task.autostart import autostart_service_tasks
 from pytest import LogCaptureFixture
 
 
@@ -19,10 +20,9 @@ async def test_reconnection(caplog: LogCaptureFixture) -> None:
             self._connected = True
 
     service_instance = MyService()
+    autostart_service_tasks(service_instance)
 
     assert service_instance._connected is False
 
-    service_instance._task_manager.start_autostart_tasks()
-
     await asyncio.sleep(0.01)
     assert service_instance._connected is True

tests/data_service/test_data_service.py

@@ -36,8 +36,7 @@ def test_unexpected_type_change_warning(caplog: LogCaptureFixture) -> None:
 
 
 def test_basic_inheritance_warning(caplog: LogCaptureFixture) -> None:
-    class SubService(DataService):
+    class SubService(DataService): ...
-        ...
 
     class SomeEnum(Enum):
         HI = 0
@@ -57,11 +56,9 @@ def test_basic_inheritance_warning(caplog: LogCaptureFixture) -> None:
         def name(self) -> str:
             return self._name
 
-        def some_method(self) -> None:
+        def some_method(self) -> None: ...
-            ...
 
-        async def some_task(self) -> None:
+        async def some_task(self) -> None: ...
-            ...
 
     ServiceClass()
 
@@ -129,17 +126,12 @@ def test_exposing_methods(caplog: LogCaptureFixture) -> None:
                 return "some method"
 
     class ClassWithTask(pydase.DataService):
-        async def some_task(self, sleep_time: int) -> None:
+        @frontend
-            pass
+        def some_method(self) -> str:
+            return "some method"
 
     ClassWithTask()
 
-    assert (
-        "Async function 'some_task' is defined with at least one argument. If you want "
-        "to use it as a task, remove the argument(s) from the function definition."
-        in caplog.text
-    )
-
 
 def test_dynamically_added_attribute(caplog: LogCaptureFixture) -> None:
     class MyService(DataService):

tests/data_service/test_data_service_cache.py

@@ -1,7 +1,6 @@
 import logging
 
 import pydase
-import pytest
 from pydase.data_service.data_service_observer import DataServiceObserver
 from pydase.data_service.state_manager import StateManager
 
@@ -33,35 +32,3 @@ def test_nested_attributes_cache_callback() -> None:
         ]
         == "Ciao"
     )
-
-
-@pytest.mark.asyncio(scope="function")
-async def test_task_status_update() -> None:
-    class ServiceClass(pydase.DataService):
-        name = "World"
-
-        async def my_method(self) -> None:
-            pass
-
-    service_instance = ServiceClass()
-    state_manager = StateManager(service_instance)
-    DataServiceObserver(state_manager)
-
-    assert (
-        state_manager.cache_manager.get_value_dict_from_cache("my_method")["type"]
-        == "method"
-    )
-    assert (
-        state_manager.cache_manager.get_value_dict_from_cache("my_method")["value"]
-        is None
-    )
-
-    service_instance.start_my_method()  # type: ignore
-    assert (
-        state_manager.cache_manager.get_value_dict_from_cache("my_method")["type"]
-        == "method"
-    )
-    assert (
-        state_manager.cache_manager.get_value_dict_from_cache("my_method")["value"]
-        == "RUNNING"
-    )

tests/data_service/test_data_service_observer.py

@@ -5,7 +5,7 @@ import pydase
 import pytest
 from pydase.data_service.data_service_observer import DataServiceObserver
 from pydase.data_service.state_manager import StateManager
-from pydase.utils.serialization.serializer import SerializationError
+from pydase.utils.serialization.serializer import SerializationError, dump
 
 logger = logging.getLogger()
 
@@ -146,3 +146,41 @@ def test_private_attribute_does_not_have_to_be_serializable() -> None:
         service_instance.change_publ_attr()
 
     service_instance.change_priv_attr()
+
+
+def test_normalized_attr_path_in_dependent_property_changes(
+    caplog: pytest.LogCaptureFixture,
+) -> None:
+    class SubService(pydase.DataService):
+        _prop = 10.0
+
+        @property
+        def prop(self) -> float:
+            return self._prop
+
+    class MyService(pydase.DataService):
+        def __init__(self) -> None:
+            super().__init__()
+            self.service_dict = {"one": SubService()}
+
+    service_instance = MyService()
+    state_manager = StateManager(service=service_instance)
+    observer = DataServiceObserver(state_manager=state_manager)
+
+    assert observer.property_deps_dict["service_dict['one']._prop"] == [
+        "service_dict['one'].prop"
+    ]
+
+    # We can use dict key path encoded with double quotes
+    state_manager.set_service_attribute_value_by_path(
+        'service_dict["one"]._prop', dump(11.0)
+    )
+    assert service_instance.service_dict["one"].prop == 11.0
+    assert "'service_dict[\"one\"].prop' changed to '11.0'" in caplog.text
+
+    # We can use dict key path encoded with single quotes
+    state_manager.set_service_attribute_value_by_path(
+        "service_dict['one']._prop", dump(12.0)
+    )
+    assert service_instance.service_dict["one"].prop == 12.0
+    assert "'service_dict[\"one\"].prop' changed to '12.0'" in caplog.text

tests/data_service/test_task_manager.py

@@ -1,135 +0,0 @@
-import asyncio
-import logging
-
-import pydase
-import pytest
-from pydase.data_service.data_service_observer import DataServiceObserver
-from pydase.data_service.state_manager import StateManager
-from pytest import LogCaptureFixture
-
-logger = logging.getLogger("pydase")
-
-
-@pytest.mark.asyncio(scope="function")
-async def test_autostart_task_callback(caplog: LogCaptureFixture) -> None:
-    class MyService(pydase.DataService):
-        def __init__(self) -> None:
-            super().__init__()
-            self._autostart_tasks = {  # type: ignore
-                "my_task": (),  # type: ignore
-                "my_other_task": (),  # type: ignore
-            }
-
-        async def my_task(self) -> None:
-            logger.info("Triggered task.")
-
-        async def my_other_task(self) -> None:
-            logger.info("Triggered other task.")
-
-    # Your test code here
-    service_instance = MyService()
-    state_manager = StateManager(service_instance)
-    DataServiceObserver(state_manager)
-    service_instance._task_manager.start_autostart_tasks()
-
-    assert "'my_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-    assert "'my_other_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-
-
-@pytest.mark.asyncio(scope="function")
-async def test_DataService_subclass_autostart_task_callback(
-    caplog: LogCaptureFixture,
-) -> None:
-    class MySubService(pydase.DataService):
-        def __init__(self) -> None:
-            super().__init__()
-            self._autostart_tasks = {  # type: ignore
-                "my_task": (),
-                "my_other_task": (),
-            }
-
-        async def my_task(self) -> None:
-            logger.info("Triggered task.")
-
-        async def my_other_task(self) -> None:
-            logger.info("Triggered other task.")
-
-    class MyService(pydase.DataService):
-        sub_service = MySubService()
-
-    service_instance = MyService()
-    state_manager = StateManager(service_instance)
-    DataServiceObserver(state_manager)
-    service_instance._task_manager.start_autostart_tasks()
-
-    assert "'sub_service.my_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-    assert "'sub_service.my_other_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-
-
-@pytest.mark.asyncio(scope="function")
-async def test_DataService_subclass_list_autostart_task_callback(
-    caplog: LogCaptureFixture,
-) -> None:
-    class MySubService(pydase.DataService):
-        def __init__(self) -> None:
-            super().__init__()
-            self._autostart_tasks = {  # type: ignore
-                "my_task": (),
-                "my_other_task": (),
-            }
-
-        async def my_task(self) -> None:
-            logger.info("Triggered task.")
-
-        async def my_other_task(self) -> None:
-            logger.info("Triggered other task.")
-
-    class MyService(pydase.DataService):
-        sub_services_list = [MySubService() for i in range(2)]
-
-    service_instance = MyService()
-    state_manager = StateManager(service_instance)
-    DataServiceObserver(state_manager)
-    service_instance._task_manager.start_autostart_tasks()
-
-    assert (
-        "'sub_services_list[0].my_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-    )
-    assert (
-        "'sub_services_list[0].my_other_task' changed to 'TaskStatus.RUNNING'"
-        in caplog.text
-    )
-    assert (
-        "'sub_services_list[1].my_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-    )
-    assert (
-        "'sub_services_list[1].my_other_task' changed to 'TaskStatus.RUNNING'"
-        in caplog.text
-    )
-
-
-@pytest.mark.asyncio(scope="function")
-async def test_start_and_stop_task_methods(caplog: LogCaptureFixture) -> None:
-    class MyService(pydase.DataService):
-        def __init__(self) -> None:
-            super().__init__()
-
-        async def my_task(self) -> None:
-            while True:
-                logger.debug("Logging message")
-                await asyncio.sleep(0.1)
-
-    # Your test code here
-    service_instance = MyService()
-    state_manager = StateManager(service_instance)
-    DataServiceObserver(state_manager)
-    service_instance.start_my_task()
-    await asyncio.sleep(0.01)
-
-    assert "'my_task' changed to 'TaskStatus.RUNNING'" in caplog.text
-    assert "Logging message" in caplog.text
-    caplog.clear()
-
-    service_instance.stop_my_task()
-    await asyncio.sleep(0.01)
-    assert "Task 'my_task' was cancelled" in caplog.text

tests/task/test_task.py

@@ -0,0 +1,291 @@
+import asyncio
+import logging
+
+import pydase
+import pytest
+from pydase.data_service.data_service_observer import DataServiceObserver
+from pydase.data_service.state_manager import StateManager
+from pydase.task.autostart import autostart_service_tasks
+from pydase.task.decorator import task
+from pydase.task.task_status import TaskStatus
+from pytest import LogCaptureFixture
+
+logger = logging.getLogger("pydase")
+
+
+@pytest.mark.asyncio(scope="function")
+async def test_start_and_stop_task(caplog: LogCaptureFixture) -> None:
+    class MyService(pydase.DataService):
+        @task()
+        async def my_task(self) -> None:
+            logger.info("Triggered task.")
+            while True:
+                await asyncio.sleep(1)
+
+    # Your test code here
+    service_instance = MyService()
+    state_manager = StateManager(service_instance)
+    DataServiceObserver(state_manager)
+
+    autostart_service_tasks(service_instance)
+    await asyncio.sleep(0.1)
+    assert service_instance.my_task.status == TaskStatus.NOT_RUNNING
+
+    service_instance.my_task.start()
+    await asyncio.sleep(0.1)
+    assert service_instance.my_task.status == TaskStatus.RUNNING
+
+    assert "'my_task.status' changed to 'TaskStatus.RUNNING'" in caplog.text
+    assert "Triggered task." in caplog.text
+    caplog.clear()
+
+    service_instance.my_task.stop()
+    await asyncio.sleep(0.1)
+    assert service_instance.my_task.status == TaskStatus.NOT_RUNNING
+    assert "Task 'my_task' was cancelled" in caplog.text
+
+
+@pytest.mark.asyncio(scope="function")
+async def test_autostart_task(caplog: LogCaptureFixture) -> None:
+    class MyService(pydase.DataService):
+        @task(autostart=True)
+        async def my_task(self) -> None:
+            logger.info("Triggered task.")
+            while True:
+                await asyncio.sleep(1)
+
+    # Your test code here
+    service_instance = MyService()
+    state_manager = StateManager(service_instance)
+    DataServiceObserver(state_manager)
+
+    autostart_service_tasks(service_instance)
+
+    await asyncio.sleep(0.1)
+    assert service_instance.my_task.status == TaskStatus.RUNNING
+
+    assert "'my_task.status' changed to 'TaskStatus.RUNNING'" in caplog.text
+
+
+@pytest.mark.asyncio(scope="function")
+async def test_nested_list_autostart_task(
+    caplog: LogCaptureFixture,
+) -> None:
+    class MySubService(pydase.DataService):
+        @task(autostart=True)
+        async def my_task(self) -> None:
+            logger.info("Triggered task.")
+            while True:
+                await asyncio.sleep(1)
+
+    class MyService(pydase.DataService):
+        sub_services_list = [MySubService() for i in range(2)]
+
+    service_instance = MyService()
+    state_manager = StateManager(service_instance)
+    DataServiceObserver(state_manager)
+    autostart_service_tasks(service_instance)
+
+    await asyncio.sleep(0.1)
+    assert service_instance.sub_services_list[0].my_task.status == TaskStatus.RUNNING
+    assert service_instance.sub_services_list[1].my_task.status == TaskStatus.RUNNING
+
+    assert (
+        "'sub_services_list[0].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+    assert (
+        "'sub_services_list[1].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+
+
+@pytest.mark.asyncio(scope="function")
+async def test_nested_dict_autostart_task(
+    caplog: LogCaptureFixture,
+) -> None:
+    class MySubService(pydase.DataService):
+        @task(autostart=True)
+        async def my_task(self) -> None:
+            logger.info("Triggered task.")
+            while True:
+                await asyncio.sleep(1)
+
+    class MyService(pydase.DataService):
+        sub_services_dict = {"first": MySubService(), "second": MySubService()}
+
+    service_instance = MyService()
+    state_manager = StateManager(service_instance)
+    DataServiceObserver(state_manager)
+
+    autostart_service_tasks(service_instance)
+
+    await asyncio.sleep(0.1)
+
+    assert (
+        service_instance.sub_services_dict["first"].my_task.status == TaskStatus.RUNNING
+    )
+    assert (
+        service_instance.sub_services_dict["second"].my_task.status
+        == TaskStatus.RUNNING
+    )
+
+    assert (
+        "'sub_services_dict[\"first\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"second\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+
+
+@pytest.mark.asyncio(scope="function")
+async def test_manual_start_with_multiple_service_instances(
+    caplog: LogCaptureFixture,
+) -> None:
+    class MySubService(pydase.DataService):
+        @task()
+        async def my_task(self) -> None:
+            logger.info("Triggered task.")
+            while True:
+                await asyncio.sleep(1)
+
+    class MyService(pydase.DataService):
+        sub_services_list = [MySubService() for i in range(2)]
+        sub_services_dict = {"first": MySubService(), "second": MySubService()}
+
+    service_instance = MyService()
+    state_manager = StateManager(service_instance)
+    DataServiceObserver(state_manager)
+
+    autostart_service_tasks(service_instance)
+
+    await asyncio.sleep(0.1)
+
+    assert (
+        service_instance.sub_services_list[0].my_task.status == TaskStatus.NOT_RUNNING
+    )
+    assert (
+        service_instance.sub_services_list[1].my_task.status == TaskStatus.NOT_RUNNING
+    )
+    assert (
+        service_instance.sub_services_dict["first"].my_task.status
+        == TaskStatus.NOT_RUNNING
+    )
+    assert (
+        service_instance.sub_services_dict["second"].my_task.status
+        == TaskStatus.NOT_RUNNING
+    )
+
+    service_instance.sub_services_list[0].my_task.start()
+    await asyncio.sleep(0.01)
+
+    assert service_instance.sub_services_list[0].my_task.status == TaskStatus.RUNNING
+    assert (
+        "'sub_services_list[0].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+    assert (
+        "'sub_services_list[1].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"first\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"second\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+
+    service_instance.sub_services_list[0].my_task.stop()
+    await asyncio.sleep(0.01)
+
+    assert "Task 'my_task' was cancelled" in caplog.text
+    caplog.clear()
+
+    service_instance.sub_services_list[1].my_task.start()
+    await asyncio.sleep(0.01)
+
+    assert service_instance.sub_services_list[1].my_task.status == TaskStatus.RUNNING
+    assert (
+        "'sub_services_list[0].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_list[1].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"first\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"second\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+
+    service_instance.sub_services_list[1].my_task.stop()
+    await asyncio.sleep(0.01)
+
+    assert "Task 'my_task' was cancelled" in caplog.text
+    caplog.clear()
+
+    service_instance.sub_services_dict["first"].my_task.start()
+    await asyncio.sleep(0.01)
+
+    assert (
+        service_instance.sub_services_dict["first"].my_task.status == TaskStatus.RUNNING
+    )
+    assert (
+        "'sub_services_list[0].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_list[1].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"first\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"second\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+
+    service_instance.sub_services_dict["first"].my_task.stop()
+    await asyncio.sleep(0.01)
+
+    assert "Task 'my_task' was cancelled" in caplog.text
+    caplog.clear()
+
+    service_instance.sub_services_dict["second"].my_task.start()
+    await asyncio.sleep(0.01)
+
+    assert (
+        service_instance.sub_services_dict["second"].my_task.status
+        == TaskStatus.RUNNING
+    )
+    assert (
+        "'sub_services_list[0].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_list[1].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"first\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        not in caplog.text
+    )
+    assert (
+        "'sub_services_dict[\"second\"].my_task.status' changed to 'TaskStatus.RUNNING'"
+        in caplog.text
+    )
+
+    service_instance.sub_services_dict["second"].my_task.stop()
+    await asyncio.sleep(0.01)
+
+    assert "Task 'my_task' was cancelled" in caplog.text

tests/utils/serialization/test_serializer.py

@@ -1,4 +1,3 @@
-import asyncio
 import enum
 from datetime import datetime
 from enum import Enum
@@ -8,7 +7,7 @@ import pydase
 import pydase.units as u
 import pytest
 from pydase.components.coloured_enum import ColouredEnum
-from pydase.data_service.task_manager import TaskStatus
+from pydase.task.task_status import TaskStatus
 from pydase.utils.decorators import frontend
 from pydase.utils.serialization.serializer import (
     SerializationPathError,
@@ -214,11 +213,9 @@ async def test_method_serialization() -> None:
             return "some method"
 
         async def some_task(self) -> None:
-            while True:
+            pass
-                await asyncio.sleep(10)
 
     instance = ClassWithMethod()
-    instance.start_some_task()  # type: ignore
 
     assert dump(instance)["value"] == {
         "some_method": {
@@ -234,7 +231,7 @@ async def test_method_serialization() -> None:
         "some_task": {
             "full_access_path": "some_task",
             "type": "method",
-            "value": TaskStatus.RUNNING.name,
+            "value": None,
             "readonly": True,
             "doc": None,
             "async": True,