This library leverages the most advanced Python features including the descriptor protocol and support for customizable metaclasses. The documentation below explains these features and how WorkToy leverages them to provide powerful and flexible tools for Python developers.
-
The Python metaclass -
worktoy.meta
-
The
worktoy.meta
Module - The
worktoy.keenum
module -
The
worktoy.ezdata
module -
The
worktoy.text
module
The stable version of WorkToy may be installed using the following command:
pip install worktoy
The development version, which is not for the faint of heart, may be
installed by passing the --pre
flag:
pip install worktoy --pre
The descriptor protocol in Python allows significant customisation of the
attribute access mechanism. To illustrate, let us implement a descriptor
class Integer
which wraps integer values. Such a class is intended
to be instantiated in the class bodies of other classes.
This discussion will now continue in the docstrings found in the implementation of this class.
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from typing import Any
class Integer:
"""Let us continue this discussion in the docstrings of the 'Integer'
class.
This descriptor class defines the following instance attributes:
- __fallback_value__: If instantiated without a default value, this
value is used.
- __default_value__: The default value of the descriptor instance.
- __field_name__: The name by which the descriptor instance appears in
the class body.
- __field_owner__: The class owning the descriptor instance.
- __pvt_name__: The name of the private variable used to store the
value of the attribute.
It also defines the following methods which each provides a docstring
explaining their purpose and function:
- __init__
- __set_name__
- __get__
- __set__
- __delete__ # DO NOT MISTAKE WITH __del__!
"""
__fallback_value__ = 0
__default_value__ = None
__field_name__ = None
__field_owner__ = None
def __init__(self, *args) -> None:
"""This constructor method accepts any number of positional arguments.
It then implements the under used and underappreciated 'for-loop'
ending with an 'else' clause. The code in the 'else' block runs
after the loop completes. The point is that the 'break' keyword also
applies to this block. So if 'break' is encountered, the code in the
'else' block does not run. To translate to natural language:
'Go through each positional argument and when you find an integer,
assign it to the default value. If you have not found any integer
after looking through each positional argument, use the fallback
value.'
It is the opinion of this author that the 'else' clause in a loop is
underused and underappreciated."""
for arg in args:
if isinstance(arg, int):
self.__default_value__ = arg
break
else:
self.__default_value__ = self.__fallback_value__
def __set_name__(self, owner: type, name: str) -> None:
"""This is the method that elevates the power of the descriptor
protocol beyond the mundane getter and setter pattern! This feature
was added in Python 3.6 released on December 23, 2016. Does this make it
a recent feature? Well, Minecraft Java 1.11 had been released on
November 14, 2016, meaning that this feature is about the same age as
totems of undying, shulker boxes and the observer block.
This method is invoked when the class owning the descriptor is created.
It informs the instance of the descriptor of its owner and the name
by which it appears in the class body of the owner. This means that the
descriptor instance is aware of its own name in the namespace of its
owner.
"""
self.__field_name__ = name
self.__field_owner__ = owner
self.__pvt_name__ = '__%s_value__' % (name,)
def __get__(self, instance: object, owner: type) -> Any:
"""This method is called when the descriptor instance is accessed. It
returns the value of the attribute. If the descriptor instance is
accessed through the owning class, the descriptor instance itself is
returned. For example:
class Owner:
num = Integer(69)
if __name__ == '__main__':
owner = Owner()
print(type(owner.num)) # <class 'int'>, the wrapped value instance
print(type(Owner.num)) # <class 'Integer'>, the descriptor instance
'owner.num' results in the following call:
'__get__(owner, Owner)'
'Owner.num' results in the following call:
'__get__(None, Owner)'
By making the above distinction, the descriptor instance object may
be accessed by going through the owning class. This is the most
common and highly recommended pattern. However, it also means that
ambiguity exists for the type-hint: When accessing through the
instance the hint should be 'int', but when accessing through the
class, the hint should be 'type'. For this reason, the 'Any' type is
used here.
"""
if instance is None:
return self
if getattr(instance, self.__pvt_name__, None) is None:
return self.__default_value__
return getattr(instance, self.__pvt_name__)
def __set__(self, instance: object, value: object) -> None:
"""This method is called when the attribute at the field name of the
descriptor instance is attempted to be set on the instance of the
owning class. Unlike the '__get__' method defined above, this method
is invoked only when the attribute is set on the instance. If set on
the owner, the descriptor instance itself is overwritten. This is
consistent with the pattern that access through the owning class
refers to the descriptor instance, whilst access through the owning
instance is managed by the descriptor class. For example:
class Owner:
num = Integer(420)
if __name__ == '__main__':
owner = Owner()
print(owner.num) # 420
owner.num = 69
print(owner.num) # 69
print(Owner.num) # <Integer object at 0x1EE7B00B5>
print(type(Owner.num)) # <class 'Integer'>
Owner.num = 69
print(Owner.num) # 69
print(type(Owner.num)) # <class 'int'>
owner.num = 69 # This results in the following call:
'__set__(owner, 69)'
Owner.num = 69 # This results in the following call:
'type(Owner).__setattr__(Owner, 'num', 69)'
"""
setattr(instance, self.__pvt_name__, value)
def __delete__(self, instance: object) -> None:
"""This method is called when the attribute is attempted to be
deleted. DO NOT MISTAKE WITH '__del__'! The '__del__' method is
called when the instance is destroyed. Both of these are outside the
scope of this discussion."""
delattr(instance, self.__pvt_name__)
In summary, the Integer
class defined above provides integer valued
attributes to other classes. The accessor functions implements only
trivial functionality here, but serves to illustrate the possibilities
for customization.
Python does provide a built-in class for creating descriptors: the
property
class. This class allows the use of a decorator to define
getter, setter and deleter functions for an attribute. Alternatively, the
property
may be instantiated in the class body with the getter, setter
and deleter functions as arguments. The following class has attributes
name
and number
both instances of property
implemented with
the decorator and the constructor respectively.
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
class OwningClass:
"""This class uses 'property' to implement the 'name' and 'number'
attributes. """
__fallback_number__ = 0
__fallback_name__ = 'Unnamed'
__inner_number__ = None
__inner_name__ = None
def __init__(self, *args, **kwargs) -> None:
self.__inner_number__ = kwargs.get('number', None)
self.__inner_name__ = kwargs.get('name', None)
for arg in args:
if isinstance(arg, int) and self.__inner_number__ is None:
self.__inner_number__ = arg
elif isinstance(arg, str) and self.__inner_name__ is None:
self.__inner_name__ = arg
@property
def name(self) -> str:
"""Name property"""
if self.__inner_name__ is None:
return self.__fallback_name__
return self.__inner_name__
@name.setter
def name(self, value: str) -> None:
"""Name setter"""
self.__inner_name__ = value
@name.deleter
def name(self) -> None:
"""Name deleter"""
del self.__inner_name__
def _getNumber(self, ) -> int:
"""Number getter"""
if self.__inner_number__ is None:
return self.__fallback_number__
return self.__inner_number__
def _setNumber(self, value: int) -> None:
"""Number setter"""
self.__inner_number__ = value
def _delNumber(self) -> None:
"""Number deleter"""
del self.__inner_number__
number = property(_getNumber, _setNumber, _delNumber, doc='Number')
Before proceeding further, let us briefly discuss the @decorator
syntax.
When using the keywords: def
, async def
and class
, we
begin a compound statement that creates a new object at the name
following the keyword. When such a compound statement is 'decorated' it
means that the created object is passed to the decorating function upon
creation and the object returned by the decorator is assigned to the name
instead.
A decorator could return an object wrapping the decorated object to augment its behaviour. Alternatively, it could record the decorated function for a particular purpose and return the object exactly as received. Or even a combination of the two.
Let us now examine a few decorators beginning with a simple notifier:
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from typing import Callable
def notify(callMeMaybe: Callable) -> Callable:
"""This decorator notifies the creation of the received object by
printing its name. """
print(callMeMaybe.__name__)
return callMeMaybe
@notify
class SomeClass:
"""This is just some class. """
@notify
def someMethod(self, ) -> None:
"""This is just some method. """
@notify
def nestedMethod() -> None:
"""It even has a nested method!"""
if __name__ == '__main__':
print('Entry point!')
someInstance = SomeClass()
The above code will output the following:
someMethod
SomeClass
Entry point!
The above may surprise some readers anticipating the output to begin with
the 'Entry point!' message. However, the class body is executed
immediately even before the if __name__ == '__main__':
block is
entered. During this execution, the someMethod
function is created
and passed to the notify
decorator. Upon completion of the class body
execution, the newly created class object is passed to the notify
decorator. Both happen prior to the normal entry point. But what about
the nested method? The someMethod
creates it when called, not when
created. In the example, the class body creates the someMethod
, but
nobody actually invokes it. Thus, the nested method remains uncreated.
Having introduced the Python descriptor protocol and the property
class, we shall now introduce the Field
class provided by the
worktoy.desc
module. This class aims at providing the same
functionality as the property
class. Class owning instances of
Field
can heavily customize the attribute access mechanism at a
particular name. In fact, the Field
descriptor provides no actual
functionality itself, but simply allows the owning class to define each
of the accessor functions. If multiple classes are to implement multiple
attributes with similar behaviour using the Field
class, it will
introduce significant boilerplate code. To implement general attribute
behaviour requires a different approach. Foreshadowing...
The following is a truncated version of the Field
class provided by
the worktoy.desc
module. This documentation will continue in the
docstrings below.
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from typing import Callable, Never
class Field:
"""Welcome back to the docstrings, where our discussion continues!
Please note that the 'Field' class found in 'worktoy.desc' is
significantly more complicated than the code below, which simply
illustrates the concept. """
# Accessor functions for the accessors of the attribute.
__getter_function__ = None
__setter_function__ = None
__deleter_function__ = None
def _getGetter(self, ) -> Callable:
"""This method returns the getter function. """
return self.__getter_function__
def _setGetter(self, value: Callable) -> None:
"""This method sets the getter function. """
self.__getter_function__ = value
def _getSetter(self, ) -> Callable:
"""This method returns the setter function. """
return self.__setter_function__
def _setSetter(self, value: Callable) -> None:
"""This method sets the setter function. """
self.__setter_function__ = value
def _getDeleter(self, ) -> Callable:
"""This method returns the deleter function. """
return self.__deleter_function__
def _setDeleter(self, value: Callable) -> None:
"""This method sets the deleter function. """
self.__deleter_function__ = value
# So much boilerplate code!
__field_name__ = None
__field_owner__ = None
def __set_name__(self, owner: type, name: str) -> None:
"""This is the same as in the previous 'Integer' class example."""
self.__field_name__ = name
self.__field_owner__ = owner
def _getPrivateName(self) -> str:
"""Parses a private name from the field name. """
return '__%s_value__' % (self.__field_name__,)
# We are finally ready to implement the descriptor protocol!
def __get__(self, instance: object, owner: type) -> object:
"""This method is called when the descriptor instance is accessed. """
if instance is None:
return self
pvtName = self._getPrivateName()
getFunc = self._getGetter() # Existence check omitted
return getFunc(instance)
def __set__(self, instance: object, value: object) -> None:
"""This method is called when the attribute at the field name of the
descriptor instance is attempted to be set on the instance of the
owning class. """
pvtName = self._getPrivateName()
setFunc = self._getSetter() # Existence check omitted
setFunc(instance, value)
def __delete__(self, instance: object) -> Never:
"""Outside the scope of this discussion."""
e = """This example does not implement attribute deletion!"""
raise TypeError(e)
# Public names for the accessor decorators:
def GET(self, func: Callable) -> Callable:
"""This decorator marks the function as the getter function. """
self._setGetter(func)
return func
def SET(self, func: Callable) -> Callable:
"""This decorator marks the function as the setter function. """
self._setSetter(func)
return func
def DEL(self, func: Callable) -> Callable:
"""This decorator marks the function as the deleter function. """
self._setDeleter(func)
return func
To illustrate the use of the Field
class, we will implement a class
encapsulation of a complex number having the real and imaginary parts as
separate attributes. The example will illustrate decorators, the
Field
class and the dangers of boilerplate code.
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.desc import Field
from worktoy.parse import maybe # None-aware filter
# The 'maybe' function provides None-awareness. It is similar to the
# null-coalescing operator in javascript.
# maybe(a, b) returns 'a' if 'a' is not None, otherwise it returns 'b'.
# Same as: a ?? b in javascript.
# Likely, the only redeeming feature of javascript.
class Complex:
"""More boilerplate??"""
__real_fallback__ = 0.
__imag_fallback__ = 0.
__real_part__ = None
__imag_part__ = None
RE = Field()
IM = Field()
@RE.GET
def _getReal(self) -> float:
"""Getter function for the real part."""
return maybe(self.__real_part__, self.__real_fallback__)
@RE.SET
def _setReal(self, value: float) -> None:
"""Setter function for the real part."""
self.__real_part__ = value
@IM.GET
def _getImag(self) -> float:
"""Getter function for the imaginary part."""
return maybe(self.__imag_part__, self.__imag_fallback__)
@IM.SET
def _setImag(self, value: float) -> None:
"""Setter function for the imaginary part."""
self.__imag_part__ = value
def __init__(self, *args) -> None:
x, y = None, None
for arg in args:
if isinstance(arg, int):
arg = float(arg)
if isinstance(arg, float) and x is None:
x = arg
elif isinstance(arg, float) and y is None:
y = arg
break
elif isinstance(arg, complex):
x, y = arg.real, arg.imag
break
else:
x, y = 69., 420.
self.RE, self.IM = x, y
def __add__(self, other: object) -> Complex:
"""This method is left as an exercise to the reader along with:
__sub__
__mul__
__truediv__
__pow__
__abs__
Any 'try-hard' reader may also implement the 'i' and 'r' versions."""
The Complex
class leverages the Field
implementation of the
descriptor protocol to provide attributes for the real and imaginary parts
of a complex number. Below is usage example:
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
if __name__ == '__main__':
z1 = Complex(69, 420)
z2 = Complex(1, 2)
z2.RE = 3
z2.IM = 4
z3 = z1 + z2
print(z3.RE, z3.IM)
z4 = 1337 * z1 # Hopefully, a try-hard reader implemented this!
print(z4.RE, z4.IM)
The above code will output the following:
70.0 422.0
93003.0 56180.0
In summary, both the Field
and property
classes provide a way to
customize the attribute access mechanism for each class. This comes at
the cost of significant boilerplate code for attributes that are frequently
behaving in the same way. In fact, most attributes might reasonably be
expected to behave as the Field
class does. The worktoy.desc
module provides a class for this exact purpose: the AttriBox
class.
Where Field
relies on the owning class itself to specify the accessor
functions, the AttriBox
class provides an attribute of a specified
class. This class is not instantiated until an instance of the owning
class calls the __get__
method. Only then will the inner object of
the specified class be created. The inner object is then placed on a
private variable belonging to the owning instance. When the __get__
is next called the inner object at the private variable is returned. When
instantiating the AttriBox
class, the following syntactic sugar
should be used: fieldName = AttriBox[FieldClass](*args, **kwargs)
.
The arguments placed in the parentheses after the brackets are those used
to instantiate the FieldClass
given in the brackets.
Below is an example of a class using the AttriBox
class to implement
a Circle
class. It uses the Point
class defined above to manage
the center of the circle. Notice how the Point
class itself is wrapped
in an AttriBox
instance. The area
attribute is defined using the
Field
class and illustrates the use of the Field
class to expose
a value as an attribute.
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.desc import AttriBox, Field
pi = 3.1415926535897932
class Point:
"""This class provides a point in 2D space. """
x = AttriBox[float](0)
y = AttriBox[float](0)
def __init__(self, *args) -> None:
for arg in args:
if isinstance(arg, int):
arg = float(arg)
if isinstance(arg, float) and self.x is None:
self.x = arg
elif isinstance(arg, float) and self.y is None:
self.y = arg
break
else:
self.x, self.y = 0, 0
class Circle:
"""This class uses the 'AttriBox' descriptor to manage the radius and
center, and it also illustrates a use case for the 'Field' class."""
radius = AttriBox[float](0)
center = AttriBox[Point](0, 0)
area = Field()
@area.GET
def _getArea(self) -> float:
return pi * self.radius ** 2
def __str__(self) -> str:
msg = """Circle centered at: (%.3f, %.3f), with radius: %.3f"""
return msg % (self.center.x, self.center.y, self.radius)
def __init__(self, *args) -> None:
"""This constructor is left as an exercise to the reader."""
if __name__ == '__main__':
circle = Circle(69, 420, 1337)
print(circle)
circle.radius = 80085
print(circle)
Running the code above will output the following:
Circle centered at: (69.000, 420.000), with radius: 4.000
Circle centered at: (69.000, 420.000), with radius: 1.000
So far the AttriBox
instantiation has used the following syntax:
"""Basic instantiation of the 'AttriBox' class."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.desc import AttriBox
class Owner:
"""Basic instantiation of the 'AttriBox' class."""
floatBox = AttriBox[float](69.)
intBox = AttriBox[int](420)
In the above example, the AttriBox
instantiates before the owning
class is even created. However, suppose the boxed class require the
owning instance to be passed to the constructor. This presents a
challenge as the AttriBox
instance exists before the owning class
event exists. Enter the THIS
object!
TL;DR
"""Advanced instantiation of the 'AttriBox' class."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.desc import AttriBox, THIS
class WhoDat:
"""Boxed class aware of its owning instance."""
__the_boss__ = None
def __init__(self, who: object) -> None:
self.__the_boss__ = who
def getBoss(self) -> object:
return self.__the_boss__
def __str__(self) -> str:
return 'Mah boss: %s' % (self.getBoss(),)
class Boss:
"""Owning class."""
whoDat = AttriBox[WhoDat](THIS)
name = AttriBox[str]()
def __init__(self, name: str) -> None:
self.name = name
def __str__(self) -> str:
return 'Mr. %s' % (self.name,)
if __name__ == '__main__':
boss = Boss('Guido')
print(boss.whoDat)
print(boss.whoDat.getBoss() is boss)
The above produces:
Mah boss: Mr. Guido
True
When the AttriBox.__get__
is called on the whoData
attribute, the
WhoDat
class instantiates, but the AttriBox
instance replaces
THIS
with the owning instance. This allows the WhoDat
instance to
be aware of its owning instance. Likewise, TYPE
would be replaced by
the owning class, BOX
with the AttriBox
instance and ATTR
with
the AttriBox
class (or subclass) itself.
As have been demonstrated and explained, the worktoy.desc
module
provides helpful, powerful and flexible implementations of the descriptor
protocol. The Field
allows classes to customize attribute access
behaviour in significant detail. The AttriBox
class provides a way to
set as attribute any class on another class in a single line. As
mentioned briefly, the class contained by the AttriBox
instance is
not instantiated until an instance of the owning class calls the
__get__
method.
The PySide6 library provides Python bindings for the Qt framework. Despite
involving bindings to a C++ library, the code itself remains Python and
not C++, thank the LORD. Nevertheless, certain errors do not have a
Pythonic representation. The AttriBox
clas was envisioned to provide
a convenient way to develop PySide6 applications, whilst remaining
oblivious to terms like "Segmentation Fault".
AttriBox
provides two features of particularly significance for
developing in PySide6: lazy instantiation and the THIS
object.
This refers to the fact that the AttriBox
is instantiated before its
inner class is. When an instance of the owning class calls the __get__
method, the inner class is instantiated. Not before. This seamlessly
satisfies the unintuitive-adjacent requirement that the first QObject
to be instantiated is the singular QCoreApplication
instance.
When instantiating any QObject
or subclass hereof, the constructor
may be passed another QObject
instance. This instance is then set as
the parent of the newly instantiated object. However, when placing an
instance of AttriBox
in the class body with a QObject
inside, the
parent class does not actually exist yet. (Unintuitive-adjacent).
Fortunately, THIS
provides a temporary placeholder for the owning
instance, such that when the class inside the AttriBox
is
instantiated, the THIS
object is replaced by the owning instance. For
example:
"""Using 'AttriBox' and 'THIS' in PySide6."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.desc import AttriBox, THIS
from PySide6.QtWidgets import QMainWindow, QWidget, QLabel, QVBoxLayout
from PySide6.QtCore import QObject
class MainWindow(QMainWindow):
"""This subclass of QMainWindow provides the main application window. """
baseWidget = AttriBox[QWidget](THIS)
baseLayout = AttriBox[QVBoxLayout]() # QLayout should NOT have parent
welcomeLabel = AttriBox[QLabel]('Welcome to AttriBox!', THIS, )
def __init__(self, *args, ) -> None:
for arg in args:
if isinstance(arg, QObject):
QMainWindow.__init__(self, arg)
else:
QMainWindow.__init__(self)
def initUi(self) -> None:
"""This method sets up the user interface"""
self.setWindowTitle("Welcome to WorkToy!")
self.welcomeLabel.setText("""Welcome to WorkToy!""")
self.baseLayout.addWidget(self.welcomeLabel)
self.baseWidget.setLayout(self.baseLayout)
# If passing 'THIS' to the layout box, it would set the window
# instance at the layout instead of the widget instance.
self.setCentralWidget(self.baseWidget)
def show(self) -> None:
"""This reimplementation calls 'initUi' before calling the parent
implementation"""
self.initUi()
QMainWindow.show(self)
The Python descriptor protocol provides powerful customization of
attribute behaviour, but at the cost of considerable amounts of
boilerplate code. The functionality may be seen as a feature of the class
owning the attribute or as a feature of the attribute class. The
worktoy.desc
implements a class for each. The Field
class allows
the owning class to define the attribute behaviour through the use of
decorators. The AttriBox
class provides an attribute of a specified
class on a single line. Both are subclasses of the AbstractDescriptor
class which provides the core functionality of the descriptor protocol.
The worktoy.desc
module exposes the powerful descriptor protocol. In
the following we shall see how worktoy.meta
exposes an even more
powerful Python feature: the metaclass. The remaining modules in the
worktoy
module combine these to achieve even greater power!
Python is the best programming language. Why? Because you feel happy when coding in Python. Your experience while programming depends on the syntax not on the underlying technology. Python made syntax king.
Python is freedom. This author might reject out of hand any contribution lacking type hints, but the Python interpreter will not. Thus, Python types are the best. Because they are voluntary.
Just like Python blesses voluntary efforts such as type-hints, it also permits things like:
# NO RIGHTS RESERVED
try:
crime()
except BaseException: # Don't show your parole officer
pass
While the above code is a sign of a severe personality disorder, Python permits it. It is said that Python is named after UK comedy efforts, but this author suspects a deeper meaning in the name being that of a snake: Granting free will. This freedom allows good code to be genuinely clear and intentional, reflecting a developer's honest effort to make the code understandable for others, rather than merely satisfying compiler demands.
There may be readers crying, having to wipe saliva of their screens having screamed about speed, GIL, memory usage, dynamic typing and so on. Objections relating to the permissive nature of Python miss the point: You don't have to. You can do better. You are free to choose. This leaves objections about performance, but again you are free to implement something faster, for example by using a just in time compiler such as provided by the Numba library as needed. As for memory uses, this author is presently using PyCharm, a Java based application, with an allowance of 8192MB of memory. Remaining objections are either outdated or soon to be outdated as is the case for the GIL, which is scheduled for removal in Python 3.14.
If you are still not convinced Python is the best, but are still reading, it signifies that you have an open mind. A personality trait indicating that you will love the subject of the following discussion.
It is likely that you have never heard of the Python metaclass. In fact, you may have quite negative associations with the word 'meta' on account of recent smooth-brained conduct of several multi-billion dollar companies.
Many concepts have implementations in most programming languages, but 'metaclass' is exclusive to Python. No other programming language has anything like it. Java reflections? No, no, no. Rust macros? Not even close! C++ templates? Get it out of here!
Understanding the Python metaclass does require some background. In the following sections, we will examine:
- The Python object
- Object Extensions (classes)
- The Python Function
- The
*
and**
operators -
The Python
lambda
Function (anonymous functions) - Class Instantiations
- The Custom Class
- The Custom Metaclass
- The Custom Namespace
Python operates on one fundamental idea: Everything is an object.
Everything. All numbers, all strings, all functions, all modules and
everything that you can reference. Even object
itself is an object.
This means that everything supports a core set of attributes and methods
defined on the core object
type.
With everything being an object, it is necessary to extend the
functionalities in the core object
type to create new types,
hereinafter classes. This allows objects to share the base object
,
while having additional functionalities depending on their class. Python
provides a number of special classes listed below:
-
object
- The base class for all classes. This class provides the most basic functionalities. -
int
- Extension for integers. The python interpreter uses heavily optimized C code to handle integers. This is the case for several classes on this list. -
float
- Extension for floating point numbers. This class provides a number of methods for manipulating floating point numbers. -
list
- Extension for lists of objects of dynamic size allowing members to be of any type. As the amount of data increases, the greater the performance penalty for the significant convenience. -
tuple
- Extension for tuples of objects of fixed size. This class is similar to the list class, but the size is fixed. This means that the tuple is immutable. While this is inflexible, it does allow instances to be used as keys in mappings. -
dict
- Extension for mappings. Objects of this class map keys to values. Keys be of a hashable type, meaning thatobject
itself is not sufficient. The hashables on this list are:int
,float
,str
andtuple
. -
set
- Extension for sets of objects. This class provides a number of methods for manipulating sets. The set class is optimized for membership testing. -
frozenset
- Provides an immutable version ofset
allowing it to be used as a key in mappings. -
str
- Extension for strings. This class provides a number of methods for manipulating strings. Theworktoy.text
module expands upon some of these.
To reiterate, everything is an object. Each object belongs to the
object
class but may additionally belong to a class that extends the
object
class. For example: 7
is an object. It is an instance of
object
by being an instance of int
which extends object
.
Classes are responsible for defining the instantiation of instances
belonging to them. Generally speaking, classes may be instantiated by
calling the class object treating it like a function. Classes may accept
or even require arguments when instantiated.
Before proceeding, we need to talk about functions. Python provides two
builtin extensions of object
that provide standalone objects that
implement functions: function
and lambda
. Both of these have
quite unique instantiation syntax and does not follow the conventions we
shall see later in this discussion.
Python allows the following syntax for creating a function. Please note
that all functions are still objects, and all functions created with the
syntax below belong to the same class function
. Unfortunately, this
class cannot be referred to directly. Which is super weird. Anyway, to
create a function, use the following syntax:
def multiplication(a: int, b: int) -> int:
"""This function returns the product of two integers."""
return a * b
The above function implements multiplication. It also provides the optional features: type hints and a docstring. The interpreter completely ignores these, but they are very helpful for humans. It is the opinion of this author that omitting type hints and docstrings is acceptable only when running a quick test. If anyone except you or God will ever read your code, it must have type hints and docstrings!
Below is the syntax that invokes the function:
result = multiplication(7, 8) # result is 56
In the function definition, the positional arguments were named a
and
b
. In the above invocation, the positional arguments were given
directly. Alternatively, they might have been given as keyword arguments:
result = multiplication(a=7, b=8) # result is 56
tluser = multiplication(b=8, a=7) # result is 56
When keyword arguments are used instead of positional arguments, the order is irrelevant, but names are required.
Suppose the function were to be invoked with the numbers from a
list: numbers = [7, 8]
, then we might invoke the multiplication
function as follows:
result = multiplication(numbers[0], numbers[1]) # result is 56
Imagine the function took more than two arguments. The above syntax would
still work, but would be cumbersome. Enter the star *
operator:
result = multiplication(*numbers) # result is 56
Wherever multiple positional arguments are expected, and we have a list or a tuple, the star operator unpacks it. This syntax will seem confusing, but it is very powerful and is used extensively in Python. It is also orders of magnitude more readable than the equivalent in C++ or Java.
This rant is left as an exercise to the reader
Besides function calls, the star operator conveniently concatenates lists
and tuples. Suppose we have two lists: a = [1, 2]
and b = [3, 4]
we may concatenate them in several ways:
a = [1, 2]
b = [3, 4]
ab = [a[0], a[1], b[0], b[1]] # Method 1: ab is [1, 2, 3, 4]
ab = a + b # Method 2: ab is [1, 2, 3, 4]
ab = [*a, *b] # Method 3: ab is [1, 2, 3, 4]
a.extend(b) # Method 4 modifies list 'a' in place.
a = [1, 2, 3, 4] # a is extended by b
Obviously, don't use the first method. The one relevant for the present discussion is the third, but the second and fourth have merit as well, but will not be used here. Finally, list comprehension is quite powerful as well but is the subject for a different discussion.
The single star is to lists and tuples as the double star is to
dictionaries. Suppose we have a dictionary: data = {'a': 1, 'b': 2}
then we may invoke the multiplication
function as follows:
data = {'a': 1, 'b': 2}
result = multiplication(**data) # result is 2
Like the star operator, the double star operator can be used to
concatenate two dictionaries. Suppose we have two dictionaries:
A = {'a': 1, 'b': 2}
and B = {'c': 3, 'd': 4}
. These may be
combined in several ways:
A = {'a': 1, 'b': 2}
B = {'c': 3, 'd': 4}
# Method 1
AB = {**A, **B} # AB is {'a': 1, 'b': 2, 'c': 3, 'd': 4}
# Method 2
AB = A | B
# Method 3 updates A in place
A |= B
A = {'a': 1, 'b': 2} # Resetting A
# Method 4 updates A in place
A.update(B)
As before, the first method is the most relevant for the present discussion. Unlike the example with lists, there is not really a method that is bad like the first method with lists.
In conclusion, the single and double star operators provide powerful unpacking of iterables and mappings respectively. Each have reasonable alternatives, but it is the opinion of this author that the star operators are preferred as they are unique to this use. The plus and pipe operators are used for addition and bitwise OR respectively. When the user first sees the plus or the pipe, they cannot immediately infer that the code is unpacking the operands. Not before having identified the types of the operands. In contrast, the star in front of an object without space immediately says unpacking.
Anyone having browsed through Python documentation or code may have
marvelled at the function signature: def someFunc(*args, **kwargs)
.
The signature means that the function accepts any number of positional
arguments as well as any number of keyword arguments. This allows one
function to accept multiple different argument signatures. While this may
be convenient, the ubiquitous use of this pattern is likely motivated by
the absense of function overloading in native Python. (Foreshadowing...)
Before getting back to class instantiation, we will round off this
discussion of functions with the lambda
function. The lambda
function is basically the anonymous function. The syntax of it is
lambda arguments: expression
. Whatever the expression on the right
hand side of the colon evaluates to is returned by the function. The
lambda
function allows inline function definition which is much more
condensed than the regular function definition as defined above. This
allows it to solve certain problems in one line, for example:
fb = lambda n: ('' if n % 3 else 'Fizz') + ('' if n % 5 else 'Buzz') or n
fb = lambda n: ('Fizz' * n % 3 < 1) + ('Buzz' * n % 5 < 1) or n
Besides flexing, the lambda
function is useful when working with
certain fields of mathematics, requiring implementation of many functions
that fit on one line. Below is an example of a series of functions
implementing Taylor series expansions. While type-hints should always be
used, the single line nature of the lambda
function makes it
impractical to include type-hints inside the function definition. This
author suggests instead the inclusion of type hints separately, for
example for the fizzBuzz
function above:
from typing import Callable
fb: Callable[[int], str]
The above signifies that fb
is a callable that takes an integer and
returns a string. Lambda functions will not fit type hints, so this seems
a reasonably helpful alternative.
"""Lambda function implementations of common mathematical functions."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from typing import Callable, TypeAlias
# int2int is a type alias for a mapping from int to int
int2int: TypeAlias = Callable[[int], int]
factorial: int2int
# The following functions take other functions as arguments
recursiveSum: Callable[[int2int, int], int]
taylorTerm: Callable[[float, int2int], float]
# The following maps term order to term value
expTerm: int2int
sinTerm: int2int
cosTerm: int2int
sinhTerm: int2int
coshTerm: int2int
# Combining the above allows implementation of the following functions.
# The float is the independent variable and the int is the number of
# terms to be used in the Taylor expansion.
exp: Callable[[float, int], float]
sin: Callable[[float, int], float]
cos: Callable[[float, int], float]
sinh: Callable[[float, int], float]
cosh: Callable[[float, int], float]
# Below are the actual implementations using type hints as indicated above.
factorial = lambda n: factorial(n - 1) * n if n else 1
recursiveSum = lambda F, n: F(n) + (recursiveSum(F, n - 1) if n else 0)
taylorTerm = lambda x, t: (lambda n: t(n) * x ** n / factorial(n))
expTerm = lambda n: 1
sinTerm = lambda n: (-1 if ((n - 1) % 4) else 1) if n % 2 else 0
cosTerm = lambda n: sinTerm(n + 1)
sinhTerm = lambda n: 1 if n % 2 else 0
coshTerm = lambda n: sinhTerm(n + 1)
exp = lambda x, n: recursiveSum(taylorTerm(x, expTerm), n)
sin = lambda x, n: recursiveSum(taylorTerm(x, sinTerm), n)
cos = lambda x, n: recursiveSum(taylorTerm(x, cosTerm), n)
sinh = lambda x, n: recursiveSum(taylorTerm(x, sinhTerm), n)
cosh = lambda x, n: recursiveSum(taylorTerm(x, coshTerm), n)
The above collection of functions implement recursive lambda functions to calculate function values of common mathematical functions including:
-
exp
: The exponential function. -
sin
: The sine function. -
cos
: The cosine function. -
sinh
: The hyperbolic sine function. -
cosh
: The hyperbolic cosine function.
The lambda functions implement Taylor-Maclaurin series expansions at a given number of terms and then begin by calculating the last term adding the previous term to it recursively, until the 0th term is reached. This implementation demonstrates the power of the recursive lambda function and is not at all flexing.
Since this discussion includes class instantiations, the previous section discussing functions will be quite relevant. We left the discussion of builtin Python classes having listed common ones. Generally speaking, Python classes have a general syntax for instantiation except for those listed. Below is the instantiation of the builtin classes.
-
object:
obj = object()
- This creates an object. Not particularly useful but does show the general syntax. -
int:
number = 69
- This creates an integer. -
float:
number = 420.0
- This creates a float. -
str:
message = 'Hello World!'
- This creates a string. -
list:
data = [1, 2, 3]
- This creates a list. -
tuple:
data = (1, 2, 3)
- This creates a tuple. -
?:
what = (1337)
- What does this create? Well, you might imagine that this creates a tuple, but it does not. The interpreter first removes the redundant parentheses and then the evaluation makes it an integer. To create a single element tuple, you must add the trailing comma:what = (1337,)
. This applies to one element tuples, as the comma separating the elements of a multi-element tuple sufficiently informs the interpreter that this is a tuple. The empty tuple requires no commas:empty = ()
. -
set:
data = {1, 2, 3}
- This creates a set. -
dict:
data = {'key': 'value'}
- This creates a dictionary. If the keys are strings, the general syntax may be of greater convenience:data = dict(key='value')
. Not requiring quotes around the keys. Although this syntax does not support non-string keys. -
?:
data = {}
- What does this create? Does it create an empty set or an empty dictionary. This author is not actually aware, and recommends insteadset()
ordict()
respectively when creating empty sets or dictionaries.
Except for list
and tuple
, the general class instantiation syntax
may be applied as seen below:
-
int:
number = int(69)
-
float:
number = float(420.0)
-
str:
message = str('Hello World!')
-
dict:
data = dict(key='value')
- This syntax is quite reasonable, but is limited to keys of string type.
Now let's have a look at what happens if we try to instantiate tuple
,
list
, set
or frozenset
using the general syntax:
-
list:
data = list(1, 2, 3)
- NOPE! This does not create the list predicted by common sense:data = [1, 2, 3]
. Instead, we are met by the following error message: "TypeError: list expected at most 1 argument, got 3". Instead, we must use the following syntax:data = list((1, 2, 3))
ordata = list([1, 2, 3])
. Now the attentive reader may begin to object, as one of the above require a list to already be defined and the other requires the tuple to be defined. Let's see how one might instantiate a tuple directly: -
tuple:
data = tuple(1, 2, 3)
- NOPE! This does not work either! We receive the exact same error message as before. Instead, we must use one of the following:data = tuple((1, 2, 3))
ordata = tuple([1, 2, 3])
. The logically sensitive readers now see a significant inconsistency in the syntax: One cannot in fact instantiate a tuple nor a list directly without having a list or tuple already created. This author suggests that the following syntax should be accepted:data = smartTuple(1, 2, 3)
and even:data = smartList(1, 2, 3)
. Perhaps this author is just being pedantic. The existing syntax is not a problem, and it's not like the suggested instantiation syntax is used anywhere else in Python. -
set:
data = set(1, 2, 3,)
This is correct syntax. So this works, but the suggestedsmartList
andsmartTuple
functions does not, OK sure, makes sense... -
frozenset:
data = frozenset([69, 420])
- This is correct syntax.
Let us have another look at the instantiations of dict
and of set
,
but not list
and tuple
.
def newDict(**kwargs) -> dict:
"""This function creates a new dictionary having the key value pairs
given by the keyword arguments. """
return dict(**kwargs) # Unpacking the keyword arguments creates the dict.
def newSet(*args) -> set:
"""This function creates a new set having the elements given by the
positional arguments. """
return set(args) # Unpacking the positional arguments creates the set.
def newList(*args) -> list:
"""As long as we don't use the word 'list', we can actually instantiate
a list in a reasonable way."""
return [*args, ] # Unpacking the positional arguments creates the list.
def newTuple(*args) -> tuple:
"""Same as for list, but remember the hanging comma!"""
return (*args,) # Unpacking the positional arguments creates the tuple.
In the previous section, we examined functions and builtin classes. To
reiterate, in the context of this discussion a class is an extension of
object
allowing objects to belong to different classes implementing
different extensions of object
. This raises a question: What
extension of object
contains object
extensions? If 7
is an
instance of the int
extension of object
, of what extension is
int
and instance. The answer is the type
. This extension of
object
provides all extensions of object
. This implies the
surprising that type
is an instance of itself.
The introduction of the type
class allows us to make the following
insightful statement:
7
is to int
as int
is to type
. This means that type
is responsible for instantiating new classes. A few readers may now begin
to see where this is going, but before we get there, let us examine how
type
creates a new class.
"""Sample class."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.desc import AttriBox
class PlanePoint:
"""Class representing a point in the plane """
x = AttriBox[float](0)
y = AttriBox[float](0)
def __init__(self, *args, **kwargs) -> None:
"""Constructor omitted..."""
def magnitude(self) -> float:
"""This method returns the magnitude of the point. """
return (self.x ** 2 + self.y ** 2) ** 0.5
if __name__ == '__main__':
P = PlanePoint(69, 420)
After the import statement, which is not the subject of the present
discussion, the first line of code encountered by the interpreter is the
class PlanePoint:
. The line omits some default values shown here:
class PlanePoint(object, metaclass=type)
. What the interpreter does
next is entirely up to the metaclass. Whatever object the metaclass
returns will be place at the name PlanePoint
. We will now look at
what the type
metaclass, which is the default, does when it creates a
class, but keep mind that the metaclass my do whatever it wants.
-
name:
PlanePoint
is recorded as the name of the class about to be created. -
bases: A tuple of the base classes is created. The
object
does not actually arrive in this tuple and thetype
provides implicitly.
Please note that it is possible to pass keyword arguments similarly to
the metaclass=type, but this is beyond the scope of the present
discussion. With the name and the bases, the metaclass now creates a
namespace object. The type
simply uses an empty dictionary. Then the
interpreter goes through the class body line by line look for assignments,
function definitions and even nested classes. Basically every statement
in the class body that assigns a value to a key and for each such pair
the __setitem__
method is called on the namespace object. The
implication of this is that where the value to be assigned is the return
value of a function, then that function is called during the class
creation process. This means that in the PlanePoint
class above, the
instances of AttriBox
are created before the class object is created.
When the interpreter finishes, it calls the __new__
method on the
metaclass and passes to it: the name, the bases, the namespace and any
keyword arguments initially passed to class creation. The interpreter
then waits for the metaclass to return the class object. When this
happens all the objects that implement __set_name__
has the method
called informing the descriptor instances that their owner has been
created. Finally, the interpreter applies the __init__
method of the
metaclass on the newly created class.
In summary:
- Setting up class creation The interpreter records the name of the class to be created, the base classes, the keyword arguments and which metaclass is responsible for creating the class.
-
Namespace creation The items collected are passed to the
__prepare__
method on the metaclass:namespace = type.__prepare__(name, bases, **kwargs)
-
Class Body Execution The interpreter goes through the class body
line by line and assigns the values to the namespace object:
namespace['x'] = AttriBox[float](0) # Creates the AttriBox object
-
Class Object Creation The namespace object is passed to the
__new__
method on the metaclass:cls = type.__new__(type, name, bases, namespace, **kwargs)
-
Descriptor Class Notification The objects implementing the descriptor
protocol are notified that the class object has been created:
AttriBox[float].__set_name__(PlanePoint, 'x')
-
type.__init__
The metaclass is called with the class object:type.__init__(cls, name, bases, namespace, **kwargs)
Although ontype
the__init__
method is a noop.
An impractical alternative to the above syntax is to create the new class
inline: PlanePoint = type('PlanePoint', (object,), {})
. Although,
this line has an empty dictionary where the namespace should have been.
This brings us to the actual subject of this discussion: The custom
metaclass. Because every step mentioned above may be customized by
subclassing type
. Doing so takes away every limitation. The line
discussed before:
"""The syntax can create anything you want!"""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
class AnyWayUWantIt(metaclass=MyMeta):
"""The syntax can create anything you want!"""
This line can create anything. A class for example, but anything. It can
create a string, it can return None
, it can create a new function,
any object possible may be created here.
This present discussion is about creating new classes, but readers are encouraged to experiment.
As mentioned, the type
object provides a very helpful class creation
process. What it does is defined in the heavily optimized C code of the
Python interpreter. This cannot be inspected as Python code. For the
purposes of this discussion, we will now create a custom metaclass that
does the same as the type
metaclass, but exposed as Python code.
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
class MetaType(type):
"""This custom metaclass illustrates the class creation process as it
is done by the 'type' metaclass. """
@classmethod
def __prepare__(mcls, name: str, bases: tuple, **kwargs) -> dict:
"""This method creates the namespace object, which for 'type' is
merely an empty dictionary. """
return dict()
def __new__(cls, name: str, bases: tuple, namespace: dict, **kw) -> type:
"""This method creates the class object. There is not much to see
here, as the 'type' metaclass does most of the work. This is normal
in custom metaclasses where this method, if implemented, performs
some tasks, creates the class object, possibly does some more tasks,
before returning the class object. """
cls = type.__new__(type, name, bases, namespace)
return cls
def __init__(cls, name: str, bases: tuple, namespace: dict, **kw) -> None:
"""A custom metaclass may implement this method. Doing so allows
further initialization after the '__set_name__' calls have finished. """
pass
def __call__(cls, *args, **kwargs) -> object:
"""This method is called when the class object is called. The
expected behaviour even from custom metaclasses, is for it to create
a new instance of the class object. Please note, that generally
speaking, custom classes are free to implement their own
instantiation in the form of the '__new__' and '__init__' methods. If
a custom metaclass does not intend to adhere to these, then when
encountering a class body that tries to implement them, the namespace
object should raise an error. Do not allow classes derived from the
custom metaclass to implement a function that you do not intend to
actually use. """
self = cls.__new__(cls, *args, **kwargs)
if isinstance(self, cls):
self.__init__(*args, **kwargs)
return self
def __instance_check__(cls, instance: object) -> bool:
"""Whenever the 'isinstance' function is called, this method on the
metaclass is responsible for determine if the instance should be
regarded an instance of the class object. """
otherCls = type(instance)
if cls is otherCls:
return True
for item in otherCls.__mro__:
if item is cls:
return True
return False
def __subclass_check__(cls, subclass: type) -> bool:
"""Similar to the above instance check method, this method is
responsible for deciding of the subclass provided should be regarded
as a subclass of the class object. """
if cls is subclass:
return True
for item in subclass.__mro__:
if item is cls:
return True
return False
Since the type
metaclass is heavily optimized in the C code of the
Python interpreter, the above implementation is for illustrative purposes
only. It shows what methods a custom metaclass may customize to achieve a
particular behaviour.
The custom namespace object must implement __getitem__
and
__setitem__
. Additionally, it must satisfy the key error preservation
and the type.__new__
method must receive a namespace of dict
-type.
This is elaborated below:
When a dictionary is accessed with a key that does not exist, a
KeyError
is raised. The interpreter relies on this behaviour to
handle lines in the class body that are not directly assignments
correctly. This is a particularly important requirement because failing
to raise the expected KeyError
will affect only classes that happen
to include a non-assignment line. Below is a list of known situations
that causes the issue:
-
Decorators: Unless the decorator is a function defined earlier in
the class body as an instance method able to receive a callable at the
self
argument, the decorator will cause the issue described. Please note that a static method would be able to receive a callable at the first position, but the static method decorator itself would cause the issue even sooner. - Function calls: If a function not defined previously in the class body is called during the class body without being assigned to a name, the error will occur.
The issue raises an error message that will not bring attention to the namespace object. Further, classes will frequently work fine, if they happen to not include any of the above non-assignments. In summary: failing to raise the expected error must be avoided at all costs, as it will cause undefined behaviour without any indication as to the to cause.
After the class body is executed the namespace object is passed to the
__new__
method on the metaclass. If the metaclass is intended to
create a new class object, the metaclass must eventually call the
__new__
method on the parent type
class. The type.__new__
method must receive a namespace object that is a subclass of dict
. It
is only at this stage the requirement is enforced. Thus, it is possible
to use a custom namespace object that is not a subclass of dict
, but
then it is necessary to implement functionality in the __new__
method
on the metaclass such that a dict
is passed to the type.__new__
call.
During class body execution the namespace object is passed the key value
pairs encountered. When using the empty dictionary as the namespace,
information is lost when a key receives multiple values as only the most
recently set value is retained. A custom namespace might collect all
values set at each name thus preserving all information. This application
is implemented in the worktoy.meta
module. Beyond the scope of this
module is the potential for the namespace object to dynamically change
during the class body execution. This potential is not explored here, but
readers are encouraged to experiment.
Preserving multiple values on the same key can only be provided for by a
custom namespace. An obvious use case would be function overloading. This
brings up an important distinction: A class implementing function
overloading is in some ways the exact same class as before. Only the
overloaded methods are different. Providing a custom namespace does not
actually result in classes that exhibit different behaviour. Achieving
this requires customization of the metaclass itself beyond the
__prepare__
method.
We have discussed class creation by use of type
, we have illustrated
what methods might be customized. In particular the custom namespace
returned by the __prepare__
method. This brings us to the
worktoy.meta
module. Our discussion will proceed with an examination
of the contents.
Below is a list of terms used in the worktoy.meta
module:
-
cls
- A newly created class object -
self
- A newly created object that is an instance of the newly created class. -
mcls
- The metaclass creating the new class. -
namespace
- This is where the class body is stored during class creation.
The worktoy.meta
module implements a pattern where the metaclass is
responsible for defining the functionality of the class, while the
namespace object is responsible for collecting items from the class body
execution. Rather than simply passing on the namespace object it receives,
the namespace object class is expected to implement a method called
compile
. The metaclass uses the dict
returned by the compile
when it calls the type.__new__
method.
This pattern is based on the separation of responsibilities: The namespace object class is responsible for processing the contents of the class body. The metaclass is responsible for defining the functionality of the class itself.
The worktoy.meta
module provides a decorator factory called
overload
used to mark an overloaded method with a type signature. The
Dispatcher
class contains a dictionary of functions keyed by their
type signatures. When calling an instance of this class, the types of the
arguments received determine what function to call. The BaseNamespace
class is a custom namespace object that collects overloaded functions and
replaces each such name with a relevant instance of the Dispatcher
. The
BaseMetaclass
class is a custom metaclass using the BaseNamespace
class as the namespace object. Finally, the BaseObject
class derives
from the BaseMetaclass
and implements function overloading.
Singleton classes are characterized by the fact that they are allowed
only one instance. The worktoy.meta
provides Singleton
class
derived from a custom metaclass. Subclasses of it are singletons. When
calling the class object of a subclass of Singleton
the single
instance of the class is returned. If the subclass implements
__init__
then it is called on the single instance. This allows
dynamic behaviour of singletons. If this is not desired, the singleton
subclass should provide functionality preventing the __init__
method
from running more than once.
The worktoy.meta
module provides base classes and a pattern for
custom metaclass creation and uses them to implement function overloading
in the BaseObject
class. Additionally, the module provides a
Singleton
class for creating singletons, which is based on a custom
metaclass derived from the module. Other parts of the worktoy
module
makes use of the worktoy.meta
in their implementation. This includes
the KeeNum
enumeration module and the ezdata
module.
The worktoy.keenum
module provides the KeeNum
enumeration class.
This class makes use of the worktoy.meta
module to create the
enumeration class. This discussion will demonstrate how to create
enumerations with this class. Every enumeration class must be indicated
in the class body using the worktoy.keenum.auto
function. Each such
instances may provide a public value by passing it to the auto
function. Please note however, that the public value is not used for any
purpose by the module. The KeeNum
implements a hidden value that it
uses internally.
"""Enumeration of weekdays using KeeNum."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.keenum import KeeNum, auto
class Weekday(KeeNum):
"""Enumeration of weekdays."""
MONDAY = auto()
TUESDAY = auto()
WEDNESDAY = auto()
THURSDAY = auto()
FRIDAY = auto()
SATURDAY = auto()
SUNDAY = auto()
In the documentation of the worktoy.desc
module, the PySide6
framework were mentioned as a use case for the AttriBox
class. Below
is a use case for the KeeNum
class in the PySide6 framework. In
fact, the Alignment
class shown below is a truncated version
of a enumeration class included in the ezside
module currently under
development.
"""Enumeration of alignment using KeeNum. """
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from PySide6.QtCore import Qt
from worktoy.keenum import KeeNum, auto
class Alignment(KeeNum):
"""Enumeration of alignment."""
CENTER = auto()
LEFT = auto()
RIGHT = auto()
TOP = auto()
BOTTOM = auto()
TOP_LEFT = auto()
TOP_RIGHT = auto()
BOTTOM_RIGHT = auto()
BOTTOM_LEFT = auto()
The KeeNum
class might also have been used to enumerate the different
accessor functions, which might have been useful in the worktoy.desc
.
"""Enumeration of accessor functions using KeeNum."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.keenum import KeeNum, auto
class Accessor(KeeNum):
"""Enumeration of accessor functions."""
GET = auto(getattr)
SET = auto(setattr)
DEL = auto(delattr)
In the above, the Accessor
class enumerates the accessor functions
getattr
, setattr
and delattr
. But the auto
function can
also be used to decorate enumerations, which makes their public values
functions.
"""Implementation of math functions using KeeNum"""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from typing import Callable, Any
from worktoy.keenum import KeeNum, auto
class Trig(KeeNum):
"""Enumeration of trigonometric functions."""
@classmethod
def factorial(cls, n: int) -> int:
"""This function returns the factorial of the argument."""
if n:
return n * cls.factorial(n - 1)
return 1
@classmethod
def recursiveSum(cls, callMeMaybe: Callable, n: int) -> float:
"""This function returns the sum of the function F from 0 to n."""
if n:
return callMeMaybe(n) + cls.recursiveSum(callMeMaybe, n - 1)
return callMeMaybe(n)
@classmethod
def taylorTerm(cls, x: float, callMeMaybe: Callable) -> Callable:
"""This function returns a function that calculates the nth term of a
Taylor series expansion."""
def polynomial(n: int) -> float:
return callMeMaybe(n) * x ** n / cls.factorial(n)
return polynomial
@auto
def SIN(self, x: float) -> float:
"""This method returns the sine of the argument."""
term = lambda n: [0, 1, 0, -1][n % 4]
return self.recursiveSum(self.taylorTerm(x, term), 17)
@auto
def COS(self, x: float) -> float:
"""This method returns the cosine of the argument."""
term = lambda n: [1, 0, -1, 0][n % 4]
return self.recursiveSum(self.taylorTerm(x, term), 17)
@auto
def SINH(self, x: float) -> float:
"""This method returns the hyperbolic sine of the argument."""
term = lambda n: n % 2
return self.recursiveSum(self.taylorTerm(x, term), 16)
@auto
def COSH(self, x: float) -> float:
"""This method returns the hyperbolic cosine of the argument."""
term = lambda n: (n + 1) % 2
return self.recursiveSum(self.taylorTerm(x, term), 16)
def __call__(self, *args, **kwargs) -> Any:
"""Calls are passed on to the public value"""
return self.value(self, *args, **kwargs)
The worktoy.ezdata
module provides the EZData
class, which
provides a dataclass based on the AttriBox
class. This is achieved by
leveraging the custom metaclass provided by the worktoy.meta
module.
The main convenience of the EZData
is the auto generated __init__
method that will populate fields with values given as positional
arguments or keyword arguments. The keys to the keyword arguments are the
field names.
Below is an example of the EZData
class in use:
"""Dataclass for a point in the plane using EZData."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.ezdata import EZData
from worktoy.desc import AttriBox
class PlanePoint(EZData):
"""Dataclass representing a point in the plane."""
x = AttriBox[float](0)
y = AttriBox[float](0)
def __str__(self, ) -> str:
"""String representation"""
return """(%.3f, %.3f)""" % (self.x, self.y)
if __name__ == '__main__':
P = PlanePoint(69, 420)
print(P)
P.x = 1337
print(P)
P.y = 80085 # Copilot suggested this for reals, lol
print(P)
The EZData
class supports fields with AttriBox
instances. As
explained in the documentation of the worktoy.desc
module, the
AttriBox
can use any class as the inner class. Thus, subclasses of
EZData
may use any number of fields of any class.
The worktoy.text
module provides a number of functions implementing
text formatting as listed below:
-
stringList
: This function allows creating a list of strings from a single string with separated values. The separator symbol may be provided at keyword argumentseparator
, but defaults to','
. Strings in the returned lists are stripped meaning that spaces are removed from the beginning and end of each string. -
monoSpace
: This function fixes the frustrating reality of managing longer strings in Python. Splitting a string over multiple lines provides only one good option for long strings and that is by using triple quotes. This option is great except for the fact that it preserves line breaks verbatim. ThemonoSpace
function receives a string and returns it with all continuous whitespace replaced by a single space. Additionally, strings may specify explicitly where line breaks and tabs should occur by include'<br>'
and'<tab>'
respectively. Once the initial space replacement is done, the function replaces the explicit line breaks and tabs with the appropriate symbol. -
wordWrap
: This function receives an int specifying the maximum line length and a string. The function returns the string with line breaks inserted at the appropriate places. The function does not break words in the middle, but instead moves the entire word to the next line. The function also removes any leading or trailing whitespace. -
typeMsg
: This function composes the message to be raised with aTypeError
exception when anobject
namedname
did not belong to the expected classcls
. -
joinWords
: This function receives a list of words which it concatenates into a single string, separated by commas except for the final two words which are separated by the word 'and'.
Below are examples of each of the above
"""Example of the 'stringList' function."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.text import stringList
if __name__ == '__main__':
baseString = """69, 420, 1337, 80085"""
baseList = stringList(baseString)
for item in baseList:
print(item)
"""Example of the 'monoSpace' function."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.text import monoSpace
if __name__ == '__main__':
baseString = """This is a string that is too long to fit on one line.
It is so long that it must be split over multiple lines. This is
frustrating because it is difficult to manage long strings in Python.
This is a problem that is solved by the 'monoSpace' function."""
print(baseString.count('\n'))
oneLine = monoSpace(baseString)
print(oneLine.count('\n'))
"""Example of the 'wordWrap' function."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.text import wordWrap
if __name__ == '__main__':
baseString = """This is a string that is too long to fit on one line.
It is so long that it must be split over multiple lines. This is
frustrating because it is difficult to manage long strings in Python.
This is a problem that is solved by the 'wordWrap' function."""
wrapped = wordWrap(40, baseString)
print(baseString.count('\n'))
print(len(wrapped))
print('\n'.join(wrapped))
"""Example of the 'typeMsg' function."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.text import typeMsg
if __name__ == '__main__':
susObject = 69 + 0j
susName = 'susObject'
expectedClass = float
e = typeMsg(susName, susObject, expectedClass)
print(e)
"""Example of the 'joinWords' function."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.text import joinWords
if __name__ == '__main__':
words = ['one', 'two', 'three', 'four', 'five']
print(joinWords(words))
This module provides two None
-aware functions:
-
maybe
: This functions returns the first positional argument it received that is different fromNone
. -
maybeType
: Same asmaybe
but ignoring arguments that are not of the expected type given as the first positional argument.
"""Example of the 'maybe' and 'maybeType' functions."""
# AGPL-3.0 license
# Copyright (c) 2024 Asger Jon Vistisen
from __future__ import annotations
from worktoy.parse import maybe, maybeType
someFalse = [0, '', dict(), set(), list(), 0j, .0, ]
if __name__ == '__main__':
for item in someFalse:
print(maybe(None, None, item, )) # item from 'someFalse'
print(maybeType(int, None, *someFalse)) # 0
print(maybeType(str, None, *someFalse)) # ''
print(maybeType(dict, None, *someFalse)) # {}
print(maybeType(set, None, *someFalse)) # set()
print(maybeType(list, None, *someFalse)) # []
print(maybeType(complex, None, *someFalse)) # 0j
print(maybeType(float, None, *someFalse)) # 0.0