Condition: Adding extra restrictions to types

Condition is pane's solution to field validation. In pane, features are built into the type system whenever possible, increasing composibility and flexibility. So to with value restrictions. These are implemented using t.Annotated.

We'll start with some examples:

>>> import typing as t
>>> from pane import convert, val_range, Condition, Positive

# built-in conditions
>>> convert(5.0, t.Annotated[float, Positive])
5.0

>>> convert(-1.0, t.Annotated[float, Positive])
Traceback (most recent call last):
...
pane.errors.ConvertError: Expected a positive float, instead got `-1.0` (failed condition 'positive')

>>> convert(6.0, t.Annotated[float, val_range(max=5.0)])
Traceback (most recent call last):
...
pane.errors.ConvertError: Expected a float satisfying v <= 5.0, instead got `6.0` (failed condition 'v <= 5.0')

# custom conditions
>>> convert([0, 1, 2, 4], t.List[t.Annotated[int, Condition(lambda v: v % 2 == 0, name='even')]])
Traceback (most recent call last):
...
pane.errors.ConvertError: Expected sequence of ints satisfying even
While parsing field '1':
  Expected an int satisfying even, instead got `1` (failed condition 'even')

These conditions can be applied at any nesting level, not limited to top-level fields. Conditions also support the & and | bitwise operators (interpreted as boolean operators).