Encoding-only constructs

Bit sequence

A bit sequence represents N bits, where N is a known number.

These N bits are stored in (N / 8) bytes when N is a multiple of 8; otherwise, a bit sequence uses (N / 8) + 1 bytes. The bits of the first byte correspond to positions 0 to 7 in the bit sequence, the bits of the second byte correspond to positions 8 to 15 etc., until position N - 1. The remaining bits, if any, must be unset.

Bit sequences are used to encode fields, sequences, and other lists with optional types in a space-optimized fashion.

Each position in the bit sequence encodes whether or not the element in the list has a value or no value. When the bit at position P is set in the bit sequence, the element at position P has a value; when the bit is unset, the element at position P has no value.

When encoding a struct, the leading bit sequence is not used for non-optional types. For example:

compact struct Contact {
    id: string
    age: uint8?
    name: string
    email: string?
    country: string
}

Position 0 in the bit sequence corresponds to the age field, and position 1 corresponds to the email field. All the other positions in the bit sequence associated with Contact are unused and must be unset.

Segment

A segment is a varuint62 size followed by size bytes, just like a Sequence<uint8>. The bytes that follow the size are called the body of the segment.