Method Serialize

pkg/col/coldata/bytes.go:581–618

Serialize converts b into the "arrow-like" (which is arrow-compatible) format. We call this "arrow-like" because we're abusing the arrow format to get the best speed, possibly at the cost of increased allocations (when Bytes vector has been modified in-place many times via Sets at arbitrary positio

(n int, dataScratch []byte, offsetsScratch []int32)

//
// Note: it is assumed that n is not larger than MaxBatchSize.
func (b *Bytes) Serialize(n int, dataScratch []byte, offsetsScratch []int32) ([]byte, []int32) {
	if buildutil.CrdbTestBuild {
		if n > MaxBatchSize {
			colexecerror.InternalError(errors.AssertionFailedf(
				"too many bytes elements to serialize: %d vs MaxBatchSize of %d", n, MaxBatchSize,
			))
		}
	}
	data := dataScratch[:0]
	offsets := offsetsScratch[:0]

	// Handle the cases of 0 and 1 elements separately since then we cannot
	// store two offsets that we need.
	if n == 0 {
		offsets = append(offsets, 0)
		return data, offsets
	} else if n == 1 {
		data = append(data, b.Get(0)...)
		offsets = append(offsets, 0)
		offsets = append(offsets, int32(len(data)))
		return data, offsets
	}

	// Copy over b.elements treated as []byte as well as b.buffer into data.
	bytes := b.elementsAsBytes(n)
	if cap(data) < len(bytes)+len(b.buffer) {
		data = make([]byte, 0, len(bytes)+len(b.buffer))
	}
	data = append(data, bytes...)
	data = append(data, b.buffer...)

	// Now populate the offsets slice which conforms to the arrow format and has
	// the correct length.
	offsets = append(offsets, zeroInt32Slice[:n-1]...)
	offsets = append(offsets, int32(len(bytes)))
	offsets = append(offsets, int32(len(data)))
	return data, offsets
}

// Deserialize updates b according to the "arrow-like" format that was produced
// by Serialize.