Metadata-Version: 2.4
Name: bvwx
Version: 0.13.4
Summary: Bit Vectors With Xes
Project-URL: Repository, https://github.com/cjdrake/bvwx.git
Requires-Python: >=3.12
Description-Content-Type: text/markdown

# Bit Vectors With Xes

BVWX (pronounced bih-vuh-wax, like "bivouacs") is a Python library that
implements a family of hardware-oriented, bit vector data types and operators.

[Read the docs!](https://bvwx.rtfd.org) (WIP)

[![Documentation Status](https://readthedocs.org/projects/bvwx/badge/?version=latest)](https://bvwx.readthedocs.io/en/latest/?badge=latest)


## Features

### Data Types

The fundamental BVWX data type is an ordered collection of "bits".
Unlike standard Boolean algebra where bit values are restricted to {`0`, `1`},
BVWX extends this to include hardware-oriented values `W` (don't care),
and `X` (exception).

This lifted, four-state logic consists of:

* `0` - False
* `1` - True
* `W` - Weak Unknown: either True or False, propagates *optimistically*
* `X` - Strong Unknown: neither True nor False, propagates *pessimistically*

`W` may also be denoted by `-` in string literals.
The `-` character is a convention from the [Espresso][1] PLA file format.
Wherever it hampers readability, we will instead use `W`.

Collections of bits can be organized into arbitrary shapes using the
multi-dimensional `Array` and one-dimensional `Vec` class factories.

`Enum`, `Struct`, and `Union` class factories can be extended to create
user-defined abstract date types.

### Operators

BVWX implements several operators necessary for implementing Boolean algorithms.
This includes standard NOT, OR, AND, XOR, ITE (if-then-else), and MUX functions,
but also efficient implementations of arithmetic, comparison, shift, rotate,
extend, concatenate, pack, encode/decode, and bit count.

See [Operators](https://bvwx.readthedocs.io/en/latest/reference.html#operators)
section of the reference documentation for a full list.


## Example

```python
>>> from bvwx import *

>>> x0 = bits(["4b----", "4b1111", "4b0000", "4bXXXX"])
>>> x1 = bits(["4b-10X", "4b-10X", "4b-10X", "4b-10X"])

>>> # Bitwise Operators
>>> ~x0
bits(["4b----", "4b0000", "4b1111", "4bXXXX"])
>>> ~x1
bits(["4b-01X", "4b-01X", "4b-01X", "4b-01X"])
>>> x0 | x1
bits(["4b-1-X", "4b111X", "4b-10X", "4bXXXX"])
>>> x0 & x1
bits(["4b--0X", "4b-10X", "4b000X", "4bXXXX"])
>>> x0 ^ x1
bits(["4b---X", "4b-01X", "4b-10X", "4bXXXX"])

>>> # Enums
>>> class Color(Enum):
...     RED   = "2b00"
...     GREEN = "2b01"
...     BLUE  = "2b10"
...
>>> Color.GREEN & Color.BLUE
Color.RED

>>> # Structs
>>> class Pixel(Struct):
...     r: Vec[8]
...     g: Vec[8]
...     b: Vec[8]
...
>>> maize = Pixel(r="8hFF", g="8hCB", b="8h05")
>>> blue = Pixel(r="8h00", g="8h27", b="8h4C")
>>> maize & blue
Pixel(
    r=bits("8b0000_0000"),
    g=bits("8b0000_0011"),
    b=bits("8b0000_0100"),
)

>>> # And much more ...
```


## Installing

BVWX is available on [PyPI](https://pypi.org):

    $ pip install bvwx

It requires Python 3.12+


## Developing

BVWX's repository is on [GitHub](https://github.com):

    $ git clone https://github.com/cjdrake/bvwx.git

It is 100% Python, and has no runtime dependencies.
Development dependencies are listed in `requirements-dev.txt`.


[//]: Links:

[1]: https://ptolemy.berkeley.edu/projects/embedded/pubs/downloads/espresso
