Looking for a lightweight alternative to heavy Signals/Slots libraries in an asynchronous and multi-threaded environment?
PynneX is a pure-Python (asyncio-based) library that streamlines event-driven concurrency without the overhead of larger GUI frameworks or external dependencies.
Modern Python applications often blend async I/O and multithreading. Typical Signals/Slots solutions from GUI toolkits or external libraries can impose extra dependencies, especially when you only need concurrency handling rather than full UI features.
PynneX offers a focused approach:
- Decorator-based signals and slots for clean, event-driven code
- Built-in thread-safety, so you don’t manually deal with locks or queues
- Easy background task handling via
@nx_with_worker - Seamless integration with asyncio (async or sync slots)
- No external dependencies beyond Python 3.10+ (for improved asyncio support)
As a result, events flow safely across threads and coroutines without “callback spaghetti,” giving you a cleaner concurrency model in pure Python.
- Pure Python: No external dependencies needed
- Async/Await Friendly: Slots can be synchronous or asynchronous, integrating naturally with
asyncio - Thread-Safe: Automatically manages signal emissions and slot executions across thread boundaries
- Flexible Connection Types: Direct or queued connections, chosen based on whether caller/callee share the same thread
- Worker Thread Pattern: Decorator
@nx_with_workerprovides a dedicated thread & event loop, simplifying background tasks - Familiar Decorators:
@nx_signal,@nx_slot,@nx_with_worker; also available withoutnx_prefix - Thread-Safe Properties: Use
@nx_propertyto emit signals on value changes, with automatic thread dispatch - Weak Reference: If you connect a slot with
weak=True, the connection is removed automatically once the receiver is garbage-collected
PynneX depends on Python’s asyncio. You must have a running event loop (e.g., asyncio.run(...)) for certain features like async slots or cross-thread calls.
If no event loop is running, PynneX raises a RuntimeError instead of creating one behind the scenes—this ensures predictable concurrency behavior.
pip install pynnexPynneX requires Python 3.10+, leveraging newer asyncio improvements. Alternatively, clone from GitHub and install locally:
git clone https://github.com/nexconnectio/pynnex.git
cd pynnex
pip install -e .For development (includes tests and linting tools):
pip install -e ".[dev]
Here’s the simplest “Hello, Signals/Slots” example. Once installed, run the snippet below:
# hello_pynnex.py
from pynnex import with_signals, signal, slot
@with_signals
class Greeter:
@signal
def greet(self):
"""Signal emitted when greeting happens."""
pass
def say_hello(self):
self.greet.emit("Hello from PynneX!")
@with_signals
class Printer:
@slot
def on_greet(self, message):
print(message)
greeter = Greeter()
printer = Printer()
# Connect the signal to the slot
greeter.greet.connect(printer, printer.on_greet)
# Fire the signal
greeter.say_hello()Output:
Hello from PynneX!
By simply defining signal and slot, you can set up intuitive event handling that also works smoothly in multithreaded contexts.
Below are some brief examples. For more, see the docs/ directory.
from pynnex import with_signals, signal, slot
@with_signals
class Counter:
def __init__(self):
self.count = 0
@signal
def count_changed(self):
pass
def increment(self):
self.count += 1
self.count_changed.emit(self.count)
@with_signals
class Display:
@slot
async def on_count_changed(self, value):
print(f"Count is now: {value}")
# Connect and use
counter = Counter()
display = Display()
counter.count_changed.connect(display, display.on_count_changed)
counter.increment() # Will print: "Count is now: 1"@with_signals
class AsyncDisplay:
@slot
async def on_count_changed(self, value):
await asyncio.sleep(1) # Simulate async operation
print(f"Count updated to: {value}")
# Usage in async context
async def main():
counter = Counter()
display = AsyncDisplay()
counter.count_changed.connect(display, display.on_count_changed)
counter.increment()
# Wait for async processing
await asyncio.sleep(1.1)
asyncio.run(main())- Signals: Declared with
@nx_signal. Signals are attributes of a class that can be emitted to notify interested parties. - Slots: Declared with
@nx_slot. Slots are methods that respond to signals. Slots can be synchronous or async functions. - Connections: Use
signal.connect(receiver, slot)to link signals to slots. Connections can also be made directly to functions or lambdas.
PynneX automatically detects whether the signal emission and slot execution occur in the same thread or different threads:
- Auto Connection: When connection_type is AUTO_CONNECTION (default), PynneX checks whether the slot is a coroutine function or whether the caller and callee share the same thread affinity. If they are the same thread and slot is synchronous, it uses direct connection. Otherwise, it uses queued connection.
- Direct Connection: If signal and slot share the same thread affinity, the slot is invoked directly.
- Queued Connection: If they differ, the call is queued to the slot’s thread/event loop, ensuring thread safety.
This mechanism frees you from manually dispatching calls across threads.
The @nx_property decorator provides thread-safe property access with automatic signal emission:
@with_signals
class Example:
def __init__(self):
super().__init__()
self._data = None
@signal
def updated(self):
"""Signal emitted when data changes."""
pass
@nx_property(notify=updated)
def data(self):
"""Thread-safe property with change notification."""
return self._data
@data.setter
def data(self, value):
self._data = value
e = Example()
e.data = 42 # Thread-safe property set; emits 'updated' signal on changeFor background work, PynneX provides a @nx_with_worker decorator that:
- Spawns a dedicated event loop in a worker thread.
- Allows you to queue async tasks to this worker.
- Enables easy start/stop lifecycle management.
- Integrates with signals and slots for thread-safe updates to the main
Worker Example
from pynnex import nx_with_worker, signal
@with_worker
class DataProcessor:
@signal
def processing_done(self):
"""Emitted when processing completes"""
async def run(self, *args, **kwargs):
# The main entry point for the worker thread’s event loop
# Wait for tasks or stopping signal
await self.wait_for_stop()
async def process_data(self, data):
# Perform heavy computation in the worker thread
result = await heavy_computation(data)
self.processing_done.emit(result)
processor = DataProcessor()
processor.start()
# Queue a task to run in the worker thread:
processor.queue_task(processor.process_data(some_data))
# Stop the worker gracefully
processor.stop()- Usage Guide: Learn how to define signals/slots, manage threads, and structure your event-driven code.
- API Reference: Detailed documentation of classes, decorators, and functions.
- Examples: Practical use cases, including UI integration, async operations, and worker pattern usage.
- Logging Guidelines: Configure logging levels and handlers for debugging.
- Testing Guide: earn how to run tests and contribute safely.
Configure logging to diagnose issues:
import logging
logging.getLogger('pynnex').setLevel(logging.DEBUG)For more details, see the Logging Guidelines.
Pynnex uses pytest for testing:
# Run all tests
pytest
# Run with verbose output
pytest -v
# Run specific test file
pytest tests/unit/test_signal.pySee the Testing Guide for more details.
We welcome contributions! Please read our Contributing Guidelines before submitting PRs.
If PynneX has helped simplify your async/multithreaded workflows, please consider sponsoring us. All funds go toward infrastructure, documentation, and future development.
Please note that financial contributions support only the project's maintenance and do not grant financial rewards to individual contributors.
PynneX is licensed under the MIT License. See LICENSE for details.
