absence¶

🕳️ A Python library package which provides a sentinel for absent values - a falsey, immutable singleton that represents the absence of a value in contexts where None or False may be valid values.

Key Features ⭐¶

1️⃣ Absence Sentinel: A falsey singleton which represents absence.
🏭 Absence Factory: Create custom absence sentinels for package-specific or arbitrary needs.
𝒇 Predicate Functions: Determine if a value is absent.
🔍 Type Support: Type alias for optional values which may be absent. (Similar to typing.Optional and its relation to None.)
🌟 Builtins Integration: Can install singleton and predicate function into Python builtins.

Installation 📦¶

pip install absence

Examples 💡¶

Use the absent sentinel to represent missing values:

>>> from dataclasses import dataclass
>>> from absence import absent, is_absent, Absential
>>> @dataclass
... class User:
...     name: str | None
...     email: str | None
>>> def apply_partial_update(
...     user: User,
...     name: Absential[ str | None ] = absent,
...     email: Absential[ str | None ] = absent,
... ) -> User:
...     ''' Updates user fields if values provided.
...
...         Absent value means "don't change".
...         None value means "clear field".
...     '''
...     if not is_absent( name ): user.name = name
...     if not is_absent( email ): user.email = email
...     return user
>>> user = User( name = 'Alice', email = 'alice@example.com' )
>>> # Clear name but leave email unchanged
>>> updated = apply_partial_update( user, name = None )
>>> updated.name  # Cleared to None
>>> updated.email  # Unchanged
'alice@example.com'
>>> # Update both fields
>>> updated = apply_partial_update( user, name = 'Bob', email = 'bob@example.com' )
>>> updated.name
'Bob'
>>> updated.email
'bob@example.com'

Create package-specific absence sentinels:

>>> from absence import AbsenceFactory
>>> MISSING = AbsenceFactory( )
>>> bool( MISSING )
False

Use Cases 🎯¶

🔄 Optional Arguments: When None is a valid argument value but you need to detect absence.
🎯 Sentinel Values: When you need a unique, falsey object to represent missing or invalid states.
🧩 Type Safety: When you want explicit typing for values that may be absent.

Comparison with Alternatives 🤔¶

Alternative	Truthy?	Unique?	Picklable?	Scope
`object()`	Yes	Yes	No	Arbitrary
PEP 661 Sentinels	Optional	Yes	Yes	Per-Module
`dataclasses.MISSING`	Yes	Yes	No	Global
`typing.NoDefault`	Yes	Yes	Yes	Global
`absence.absent`	No	Yes	No	Global

The absent sentinel combines falsey behavior with global uniqueness, making it particularly suitable for representing missing values in contexts where None might be a valid value. The companion AbsenceFactory allows creation of arbitrary absence sentinels, when needed, such as for specific packages.

See PEP 661 (“Sentinel Values”), typing.NoDefault, and dataclasses.MISSING for more details on alternatives.