A bit vector is an array data structure that compactly stores bits.

This library is based on 5 static different data structures:

• 8-bit vector: relies on an internal uint8
• 16-bit vector: relies on an internal uint16
• 32-bit vector: relies on an internal uint32
• 64-bit vector: relies on an internal uint64
• 128-bit vector: relies on two internal uint64 (for ASCII problems)

The rationale of using a static integer compared to a dynamic []byte is first of all to save memory. There is no structure and/or slice overhead. Hence, you might be interested in this library for memory-bound computation.

Also, the operations (get, set, etc.) are way more efficient. A simple benchmark shows that it's about 10 times more efficient than using a byte slice. Moreover, there is a guarantee that the internal bit vectors will not escape to the heap and remain only at the stack level.

Yet, the only drawback is to have a fixed-size bit vector (8, 16, 32, 64 or 128). If you require a dynamic bit vector, you should take a look at dropbox/godropbox for example.

go get github.com/teivah/bitvector

• 8-bit vector:

var bv bitvector.Len8

• 16-bit vector:

var bv bitvector.Len16

• 32-bit vector:

var bv bitvector.Len32

• 64-bit vector:

var bv bitvector.Len64

• 128-bit vector:

var bv bitvector.Ascii
// Or to reinitialize the bit vector
bv = bitvector.NewAscii()

• Set ith bit:

bv = bv.Set(i, true)
bv = bv.Set(i, false)

• Get ith bit:

b := bv.Get(i) // bool

• Toggle (flip) ith bit:

bv = bv.Toggle(i)

• Clear bits from index i (included) to index j (excluded):

bv = bv.Clear(i, j)

• Count the number of bits set to 1:

i := bv.Count() // uint8

• And operator:

bv := bv1.And(bv2)

• Or operator:

bv := bv1.Or(bv2)

• Xor operator:

bv := bv1.Xor(bv2)

• AndNot operator:

bv := bv1.AndNot(bv2)

• Push operator (left shift):

bv = bv.Push(2)

• Pop operator (right shift):

bv = bv.Pop(2)

• Convert the internal bit vector structure to a string:

s := bv.String() // string