Nion utility classes.
A utility library of useful Python objects.
- Bindings, Converters, Models
- Process, Threads
- Publish and Subscribe
- Reference Counting
These objects are primarily used within the Nion Python libraries, but may be useful in general usage too.
This project is funded by Nion Co. as part of the Nion Swift imaging and analysis platform. The code is available under the MIT License.
Requires Python 3.5 or later.
Getting Help and Contributing
If you find a bug, please file an issue on GitHub. You can also contact us directly at firstname.lastname@example.org.
This is primarily a library focused on providing support to higher level Nion Python libraries. If you are using this in your own project, we will accept bug fixes and minor feature improvements via pull requests. For larger features or additions, please contact us to discuss.
This library includes some direct tests, but is also tested via other Nion Python projects. Any contribution will need to pass the entire suite of tests. New contributions should be submitted with new tests.
Summary of Features
Events can be used by objects to notify other objects when something of interest occurs. The source object “fires” the event, optionally passing parameters, and the “listener” receives a function call. The source object determines when to fire an event. The event can have multiple listeners. The listeners are called synchronously in the order in which they are added, and the source can fire unconditionally, or until a listener returns True or False.
from nion.utils import Event class Telescope: def __init__(self): self.new_planet_detected_event = Event.Event() def on_external_event(self, coordinates): self.new_planet_detected_event.fire(coordinates) def handle_new_planet(coordinates): print("New planet coordinates at " + str(coordinates)) telescope = Telescope() listener = telescope.new_planet_detected_event.listen(handle_new_planet) listener.close() # when finished
The Observable based class defines five basic events for watching for direct changes to an object such as a property changing, an object being set or cleared, or an item being inserted or removed from a list. The observable is used along with events to implement bindings.
from nion.utils import Observable class MyClass(Observable.Observable): def __init__(self): self.__weight = 1.0 @property def weight(self): return self.__weight @weight.setter def weight(self, new_weight): self.__weight = new_weight self.notify_set_property("weight", new_weight) myc = MyClass() def property_changed(key, value): if key == "weight": print("The weight changed " + str(value)) listener = myc.property_changed_event.listen(property_changed) listener.close() # when finished
Bindings, Converters, Models
Bindings connect a value in a source object to a value in a target object. Bindings can be one way bindings from source to target, or two way bindings from source to target and from target to source. Bindings can bind property values, lists, or an item in a tuple between the source and target. Bindings work using the Observable events and subsequently are implemented via Events.
Bindings can optionally make value conversions between the source and target. For instance, a binding between a float property and a user interface text field will need to convert from float to string and back. Converters between value and strings are included for integer, float, percentage, check state, and UUID to strings.
Classes for integer and float based points, sizes, and rectangles are included. Additional geometry routines are also included, for instance: line midpoint.
The PersistentObject based class defines a basic structure for storing objects and their relationship to each other into a persistent storage context. PersistentObjects can store basic properties, single objects (to-one relationship) and lists of objects (to-many relationship). Subclasses must strictly notify the PersistentObject of changes to their persistent data and follow certain guidelines. Doing so allows the object to be stored persistently and restored from persistent storage.
Properties in the PersistentObject can have validators, converters, change notifications, and more. Items and relationships have change notifications and more.
The PersistentStorageContext defines an interfaces which manages a collection of PersistentObjects. It must be able to store a simple dict structure for properties, items, and lists.
Process defines classes to facilitate a threaded queue, which executes its items serially, and thread set which executes the most recent item in the set.
ThreadPool defines a threaded dispatcher with the ability to limit dispatch frequency and a thread pool with the ability to execute explicitly without threads for testing.
Publish and Subscribe
Publish and subscribe implements a basic publish and subscribe mechanism. It is should be considered experimental and is not recommended for use.
The ReferenceCounted base class provides an explicitly reference counted object that is unique from regular Python reference counting in that it provides precise control of when the reference is acquired and released. The about_to_delete method is called when reference count reaches zero.
The Stream classes provide a async-based stream of values that can be controlled using standard reactive operators such as sample, debounce, and combine. The stream source is an Event named value_stream and the source object must provide both the value_stream and a value property.