Skip to content

Common Lisp library for handling bit vectors, bit vector arithmetic, and type conversions.

License

Notifications You must be signed in to change notification settings

thephoeron/bit-smasher

Repository files navigation

BIT-SMASHER

Quicklisp MIT License

Common Lisp library for handling bit vectors, bit vector arithmetic, and type conversions.

Available in Quicklisp as of July 2014 release.

Documentation available at: https://thephoeron.common-lisp.dev/bit-smasher/

Supported Platforms

The current release of BIT-SMASHER compiles without warning and passes all tests on 64-bit versions of the following Lisp implementations:

  • SBCL 2.2.5
  • Clozure CL 1.12.1
  • CLISP 2.49.92
  • ABCL 1.9.0
  • Allegro CL 10.1

It compiles with warnings on:

  • LispWorks 8.0.1

It compiles with style-warnings on:

  • ECL 21.2.1

It does not build on:

  • CLASP CL 1.0.0 (clasp-boehmprecise-1.0.0-316-gf3b9992a5)
  • CMUCL 21d

Usage Notes and Limitations

This library was designed to complement the set of functions included in the Common Lisp specification for handling bit-vectors, by adding relevant lookup, conversion, arithmetic, measurement, and predicate functions. For documentation and tutorials on the bit-vector functions included in the Common Lisp standard, please refer to:

BIT-SMASHER only handles the set of non-negative integers. As such, arithmetic on bit-vectors may not always produce the results you expect—return values of all arithmetic functions are given as the bit-vector of the absolute ceiling value. Manual conversion of negative integers, floats, fractions, or complex numbers will trigger an error.

Examples

The conversion functions allow you to convert universally between bit-vectors, octet-vectors, hexadecimal strings, and non-negative integers.

; universal type-casting style functions
(bits<- "F0") => #*11110000
(bits<- 240) => #*11110000
(int<- #*11110000) => 240

; manual conversions without type-checking
(hex->bits "F0") => #*11110000
(int->bits 10) => #*00001010
(octets->bits (int->octets 244)) => #*11110100
; etc., etc...

Bit-vectors are returned zero-padded to the next full byte.

(bits<- 255) => #*11111111
(bits<- 256) => #*0000000100000000

Arithmetic on bit-vectors can be achieved through the functions bit-sum, bit-difference, bit-product, bit-quotient, bit-floor, bit-ceiling, lshift, and rshift. There are also the shorthand macros, bit+, bit-, bit*, bit/, <<, and >>. As stated above, the bit-vector arithmetic functions return the absolute ceiling value of the operation. So,

(bit- #*0000 #*0010) => #*00000010 ; +2, not -2

The measurement functions wide-bit-length and min-bit-length tell you the maximum and minimum number of bits needed to store a value, respectively. They operate on bit-vectors, octet-vectors, hexadecimal strings, and non-negative integers.

(wide-bit-length 256) => 16
(min-bit-length 256) => 9

There is also the measurement function byte-length that returns the total number of bytes required to store an integer, bit-vector, or hexadecimal value; or the actual length of byte vector or simple byte array.

(byte-length "A0FF") => 2
(byte-length 65536) => 3

In addition to the built-in CL predicate function, bit-vector-p, BIT-SMASHER adds the predicate function twos-complement-p, when you need to test the minimum bit length for the two's complement rule. This is required where padding bit-vectors, octet-vectors, or hex-strings with leading zeros up to a set word-length is expected.

(twos-complement-p 256) => NIL
(twos-complement-p 255) => T

License

Copyright © 2014–2022, "the Phoeron" Colin J.E. Lupton and the Contributors. This project is released under the MIT License; please see bit-smasher/LICENSE for more information.

Releases

No releases published

Packages

No packages published

Contributors 4

  •  
  •  
  •  
  •