This data structure efficiently stores sparse data (slices) in memory. It supports both set and get operations for specific segments, treating the data as a contiguous slice. The underlying structure dynamically allocates and stores only the bits that are set, optimizing memory usage.
It merges small contiguous chunks (16384 entries by default, configurable with WithMinContiguous) into a larger slice for more efficient storage and retrieval speed.
// Create a byte store that will merge chunks of up to 1kiB.
s := store.NewStore[byte](store.WithMinContiguous[byte](1 << 10))
// Create an input buffer of 1MiB.
in := bytes.Repeat([]byte{1, 2, 3, 4}, 1<<20/4)
// Populate the values at an offset of 1GiB minus 1MiB.
s.Set(1<<30-1<<20, in)
// Get the occupancy (how many values are populated).
occupancy := s.Occupancy() // -> 1MiB
// Get the length (how much data the structure represents).
length := s.Length() // -> 1GiB
fmt.Printf("Occupancy: %v, length: %v\n", occupancy, length)
// -> Occupancy: 1048576, length: 1073741824
// Create an output buffer of 1kiB.
out := make([]byte, 1<<10)
// Check if the data would be complete at an offset of 1GiB minus 1MiB
s.Has(1<<30-1<<20, len(out)) // -> true
// Retrieve data at an offset of 1GiB.
has := s.Get(1<<30-1<<20, out) // -> true
fmt.Printf("For 1KiB at 1GiB-1MiB (complete = %v): %v\n", has, out[:4])
// -> For 1kiB at 1GiB-1MiB (complete = true): [1 2 3 4]
// Create a new output buffer of 1kiB.
out = make([]byte, 1<<10)
// Check if the data would be complete at an offset of 1GiB.
s.Has(1<<30, len(out)) // -> false
// Retrieve data at an offset of 1GiB.
has = s.Get(1<<30, out) // -> false
fmt.Printf("For 1KiB at 1GiB (complete = %v): %v\n", has, out[:4])
// -> For 1kiB at 1GiB (complete = false): [0 0 0 0]