more-abc 2.2.6

An extension of the abc module with many similar features added.

more_abc

This module is an extension of the abc module, with many similar features added.

Installation

pip install more-abc

Usage

ABCMixin

ABCMixin enforces implementation of initialize, validate, and to_dict on subclasses, and provides is_valid() and get_info() for free.

from more_abc import ABCMixin

class User(ABCMixin):
    def __init__(self, name: str, age: int):
        self.name = name
        self.age = age
        self.initialize()

    def initialize(self):
        self.active = True

    def validate(self) -> bool:
        return isinstance(self.name, str) and self.age >= 0

    def to_dict(self) -> dict:
        return {"name": self.name, "age": self.age, "active": self.active}

user = User("Alice", 30)
print(user.is_valid())   # True
print(user.get_info())   # {'data': {'name': 'Alice', 'age': 30, 'active': True}, 'is_valid': True, 'class_name': 'User'}
print(repr(user))        # User({'name': 'Alice', 'age': 30, 'active': True})

ABCException

ABCException is an abstract base for custom exceptions. Subclasses must implement get_message().

from more_abc import ABCException

class NotFoundError(ABCException):
    def get_message(self) -> str:
        return f"Class {self.cls!r} was not found."

raise NotFoundError(cls="MyClass")
# NotFoundError: Class 'MyClass' was not found.

ABCWarning

ABCWarning works the same way as ABCException but for warnings.

import warnings
from more_abc import ABCWarning

class DeprecatedWarning(ABCWarning):
    def get_message(self) -> str:
        return f"{self.cls!r} is deprecated and will be removed in a future version."

warnings.warn(DeprecatedWarning(cls="OldClass"))
# DeprecatedWarning: 'OldClass' is deprecated and will be removed in a future version.

ABCEnum

ABCEnum is an Enum base class that supports abstract methods. Use it when you want to define an enum interface that concrete subclasses must implement.

from abc import abstractmethod
from more_abc import ABCEnum

class Direction(ABCEnum):
    NORTH = "N"
    SOUTH = "S"
    EAST  = "E"
    WEST  = "W"

    @abstractmethod
    def opposite(self) -> "Direction": ...

# Direction.NORTH  →  TypeError: Can't instantiate abstract class Direction …

class CardinalDirection(Direction):
    NORTH = "N"
    SOUTH = "S"
    EAST  = "E"
    WEST  = "W"

    def opposite(self) -> "CardinalDirection":
        _opp = {"N": "S", "S": "N", "E": "W", "W": "E"}
        return CardinalDirection(_opp[self.value])

print(CardinalDirection.NORTH.opposite())  # CardinalDirection.SOUTH

ABCEnumMeta is the underlying combined metaclass (ABCMeta + EnumMeta) and is available for advanced use cases.

ABCIntEnum

ABCIntEnum is an IntEnum base class that supports abstract methods. Members compare equal to their integer values.

from abc import abstractmethod
from more_abc import ABCIntEnum

class Permission(ABCIntEnum):
    READ    = 1
    WRITE   = 2
    EXECUTE = 4

    @abstractmethod
    def label(self) -> str: ...

class FilePermission(Permission):
    READ    = 1
    WRITE   = 2
    EXECUTE = 4

    def label(self) -> str:
        return self.name.lower()

print(FilePermission.READ.label())   # "read"
print(FilePermission.READ > 0)       # True  (int comparison)

ABCFlag

ABCFlag is a Flag base class that supports abstract methods. Members can be combined with bitwise operators.

from abc import abstractmethod
from more_abc import ABCFlag

class Access(ABCFlag):
    READ    = 1
    WRITE   = 2
    EXECUTE = 4

    @abstractmethod
    def describe(self) -> str: ...

class FileAccess(Access):
    READ    = 1
    WRITE   = 2
    EXECUTE = 4

    def describe(self) -> str:
        return f"FileAccess({self.name})"

rw = FileAccess.READ | FileAccess.WRITE
print(rw)  # FileAccess.READ|WRITE

ABCIntFlag

ABCIntFlag is an IntFlag base class that supports abstract methods. Combines integer semantics with bitwise flag operations.

from abc import abstractmethod
from more_abc import ABCIntFlag

class Mode(ABCIntFlag):
    READ    = 0o4
    WRITE   = 0o2
    EXECUTE = 0o1

    @abstractmethod
    def to_octal(self) -> str: ...

class UnixMode(Mode):
    READ    = 0o4
    WRITE   = 0o2
    EXECUTE = 0o1

    def to_octal(self) -> str:
        return oct(self.value)

print(UnixMode.READ.to_octal())              # '0o4'
print((UnixMode.READ | UnixMode.WRITE) == 6) # True

abstract_class

abstract_class is a decorator that converts any regular class into an ABC and marks the specified method names as abstract, without requiring manual ABCMeta or ABC inheritance.

from more_abc import abstract_class

@abstract_class('run', 'stop')
class Worker:
    def run(self): ...
    def stop(self): ...

class MyWorker(Worker):
    def run(self):
        print("running")

    def stop(self):
        print("stopped")

w = MyWorker()
w.run()   # running
w.stop()  # stopped

Attempting to instantiate without implementing all abstract methods raises TypeError:

class BadWorker(Worker):
    def run(self):
        print("running")
# missing stop()

BadWorker()  # TypeError: Can't instantiate abstract class BadWorker without an implementation for abstract method 'stop'

Methods listed in abstract_class that don't exist on the decorated class are automatically added as abstract stubs:

@abstract_class('process', 'cleanup')
class Pipeline:
    pass  # neither method defined — both become abstract stubs

class MyPipeline(Pipeline):
    def process(self): ...
    def cleanup(self): ...

abstractdataclass

abstractdataclass is a drop-in replacement for @dataclass that automatically gives the class ABCMeta as its metaclass, so you can use @abstractmethod without manually inheriting from ABC.

from abc import abstractmethod
from more_abc import abstractdataclass

@abstractdataclass
class Shape:
    color: str

    @abstractmethod
    def area(self) -> float: ...

@abstractdataclass(frozen=True)
class Circle(Shape):
    radius: float

    def area(self) -> float:
        return 3.14159 * self.radius ** 2

c = Circle(color="red", radius=5.0)
print(c.area())   # 78.53975
print(c)          # Circle(color='red', radius=5.0)

Attempting to instantiate an abstract class raises TypeError as expected:

Shape(color="blue")  # TypeError: Can't instantiate abstract class Shape ...

Type aliases

ABCclassType and ABCMetaclassType mirror the pattern from the types module.

from more_abc import ABCclassType, ABCMetaclassType
from abc import ABC, ABCMeta

assert ABCclassType is type(ABC)       # True
assert ABCMetaclassType is type(ABCMeta)  # True

def accepts_abc_class(cls: ABCclassType):
    print(f"Got an ABC class: {cls}")

accepts_abc_class(ABC)

AbstractLogHandler

AbstractLogHandler is an abstract base for logging.Handler. Subclasses must implement configure() and emit(). The concrete close() method handles thread-safe resource cleanup automatically.

import logging
from more_abc import AbstractLogHandler

class PrintHandler(AbstractLogHandler):
    def configure(self, config: dict) -> None:
        self._handler_config.update(config)

    def emit(self, record: logging.LogRecord) -> None:
        print(self.format(record))

handler = PrintHandler(level=logging.DEBUG)
handler.configure({"prefix": "[LOG]"})

logger = logging.getLogger("demo")
logger.addHandler(handler)
logger.warning("something happened")

AbstractLogFormatter

AbstractLogFormatter is an abstract base for logging.Formatter. Subclasses must implement format().

import logging
from more_abc import AbstractLogFormatter

class UpperFormatter(AbstractLogFormatter):
    def format(self, record: logging.LogRecord) -> str:
        return f"[{record.levelname}] {record.getMessage().upper()}"

handler = logging.StreamHandler()
handler.setFormatter(UpperFormatter())

AbstractLogFilter

AbstractLogFilter is an abstract base for logging.Filter. Subclasses must implement filter(), returning True to allow a record through or False to discard it.

import logging
from more_abc import AbstractLogFilter

class ErrorOnlyFilter(AbstractLogFilter):
    def filter(self, record: logging.LogRecord) -> bool:
        return record.levelno >= logging.ERROR

handler = logging.StreamHandler()
handler.addFilter(ErrorOnlyFilter())

abc re-exports

more_abc re-exports all public symbols from the standard abc module, so you can use it as a single import for everything ABC-related.

# Instead of:
from abc import ABC, ABCMeta, abstractmethod
from more_abc import ABCMixin, abstractdataclass

# You can do:
from more_abc import ABC, ABCMeta, abstractmethod, ABCMixin, abstractdataclass

Available re-exports: ABC, ABCMeta, abstractmethod, abstractproperty, get_cache_token.

Sortable / Filterable / Transformable

more_abc.collections_abc provides three ABC families for custom collection types, each following the same Base* / *Mixin / * pattern.

Sortable — in-place sorting via __sort__:

from more_abc import Sortable

class NumberList(Sortable):
    def __init__(self, data: list):
        self._data = list(data)

    def __sort__(self, reverse=False):
        self._data.sort(reverse=reverse)

    def __copy__(self):
        return NumberList(self._data)

nl = NumberList([3, 1, 2])
nl.sort()                    # in-place: [1, 2, 3]
asc = nl.sorted(reverse=True)  # new copy: [3, 2, 1]

Filterable — predicate filtering via __filter__:

from more_abc import Filterable

class NumberList(Filterable):
    def __init__(self, data: list):
        self._data = list(data)

    def __filter__(self, predicate):
        return NumberList([x for x in self._data if predicate(x)])

nl = NumberList([1, 2, 3, 4, 5])
evens = nl.filter(lambda x: x % 2 == 0)   # [2, 4]
odds  = nl.reject(lambda x: x % 2 == 0)   # [1, 3, 5]

Transformable — element-wise mapping via __transform__:

from more_abc import Transformable

class NumberList(Transformable):
    def __init__(self, data: list):
        self._data = list(data)

    def __transform__(self, func):
        return NumberList([func(x) for x in self._data])

nl = NumberList([1, 2, 3])
doubled = nl.map(lambda x: x * 2)  # [2, 4, 6]

BaseSortable, SortableMixin, BaseFilterable, FilterableMixin, BaseTransformable, and TransformableMixin are also exported for advanced composition.

AbstractRawIO

AbstractRawIO is an abstract base for io.RawIOBase. Subclasses must implement read(), readinto(), and write().

from more_abc import AbstractRawIO

class MemoryRawIO(AbstractRawIO):
    def __init__(self, data: bytes):
        self._buf = bytearray(data)
        self._pos = 0

    def read(self, size=-1) -> bytes:
        if size == -1:
            chunk = bytes(self._buf[self._pos:])
            self._pos = len(self._buf)
        else:
            chunk = bytes(self._buf[self._pos:self._pos + size])
            self._pos += len(chunk)
        return chunk

    def readinto(self, b) -> int:
        data = self.read(len(b))
        n = len(data)
        b[:n] = data
        return n

    def write(self, b) -> int:
        self._buf[self._pos:self._pos + len(b)] = b
        self._pos += len(b)
        return len(b)

AbstractBufferedIO

AbstractBufferedIO is an abstract base for io.BufferedIOBase. Subclasses must implement read(), read1(), and write().

from more_abc import AbstractBufferedIO

class SimpleBufferedIO(AbstractBufferedIO):
    def __init__(self, raw: bytes):
        self._buf = bytearray(raw)
        self._pos = 0

    def read(self, size=None) -> bytes:
        if size is None:
            chunk = bytes(self._buf[self._pos:])
            self._pos = len(self._buf)
        else:
            chunk = bytes(self._buf[self._pos:self._pos + size])
            self._pos += len(chunk)
        return chunk

    def read1(self, size=-1) -> bytes:
        return self.read(size)

    def write(self, b) -> int:
        self._buf += b
        return len(b)

AbstractTextIO

AbstractTextIO is an abstract base for io.TextIOBase. Subclasses must implement read(), readline(), and write().

from more_abc import AbstractTextIO

class StringIO(AbstractTextIO):
    def __init__(self, text: str = ""):
        self._lines = text.splitlines(keepends=True)
        self._pos = 0

    def read(self, size=None) -> str:
        text = "".join(self._lines[self._pos:])
        self._pos = len(self._lines)
        return text if size is None else text[:size]

    def readline(self, size=-1) -> str:
        if self._pos >= len(self._lines):
            return ""
        line = self._lines[self._pos]
        self._pos += 1
        return line if size == -1 else line[:size]

    def write(self, s: str) -> int:
        self._lines.extend(s.splitlines(keepends=True))
        return len(s)

AbstractJSONEncoder

AbstractJSONEncoder is an abstract base for json.JSONEncoder. Subclasses must implement default(), encode(), and iterencode().

from more_abc import AbstractJSONEncoder
from datetime import datetime

class DateTimeEncoder(AbstractJSONEncoder):
    def default(self, o):
        if isinstance(o, datetime):
            return o.isoformat()
        return super().default(o)

    def encode(self, o):
        return super().encode(o)

    def iterencode(self, o, _one_shot=False):
        return super().iterencode(o, _one_shot)

import json
data = {"timestamp": datetime(2024, 1, 15, 10, 30)}
result = json.dumps(data, cls=DateTimeEncoder)
print(result)  # {"timestamp": "2024-01-15T10:30:00"}

AbstractJSONDecoder

AbstractJSONDecoder is an abstract base for json.JSONDecoder. Subclasses must implement decode() and raw_decode().

from more_abc import AbstractJSONDecoder
import json

class UpperCaseDecoder(AbstractJSONDecoder):
    def decode(self, s):
        obj = super().decode(s)
        if isinstance(obj, dict):
            return {k.upper(): v for k, v in obj.items()}
        return obj

    def raw_decode(self, s, idx=0):
        return super().raw_decode(s, idx)

data = '{"name": "alice", "age": 30}'
result = json.loads(data, cls=UpperCaseDecoder)
print(result)  # {'NAME': 'alice', 'AGE': 30}