crystal binary encoder protocol

Made for performance and smallest possible binary size

Its a very simple implementation, it just follows the order of the object and concatenate binary representations of the original information. This should lead in very fast serialization and smallest possible binary format

downsides

• if mistakes are made in encoding, data will be lost, its very minimal

example

import encoder

mut e := encoder.new()
a := AStruct{
    items: ['a', 'b']
    nr: 10
    privkey: []u8{len: 5, init: u8(0xf8)}
}
e.add_list_string(a.items)
e.add_int(a.nr)
_, privkey := ed25519.generate_key()!
e.add_bytes(privkey)

println(e.data)
mut d := encoder.decoder_new(e.data)
mut aa := AStruct{}
aa.items = d.get_list_string()
aa.nr = d.get_int()
aa.privkey = d.get_bytes()

assert a == aa

fn decode[T](data []u8) !T

fn decoder_new(data []u8) Decoder

fn encode[T](obj T) ![]u8

example see https://github.com/vlang/v/blob/master/examples/compiletime/reflection.v

fn new() Encoder

enum DataType{ string int bytes u8 u16 u32 u64 time list_string list_int list_u8 list_u16 list_u32 list_u64 map_string map_bytes }

struct Decoder {
pub mut:
	version u8 = 1 // is important
	data    []u8
}

fn (mut d Decoder) get_string() string

fn (mut d Decoder) get_int() int

fn (mut d Decoder) get_bytes() []u8

fn (mut d Decoder) get_u8() u8

adds u16 length of string in bytes + the bytes

fn (mut d Decoder) get_u16() u16

fn (mut d Decoder) get_u32() u32

fn (mut d Decoder) get_u64() u64

fn (mut d Decoder) get_i64() i64

fn (mut d Decoder) get_time() time.Time

fn (mut d Decoder) get_ourtime() ourtime.OurTime

fn (mut d Decoder) get_list_string() []string

fn (mut d Decoder) get_list_int() []int

fn (mut d Decoder) get_list_u8() []u8

fn (mut d Decoder) get_list_u16() []u16

fn (mut d Decoder) get_list_u32() []u32

fn (mut d Decoder) get_list_u64() []u64

fn (mut d Decoder) get_map_string() map[string]string

fn (mut d Decoder) get_map_bytes() map[string][]u8

struct Encoder {
pub mut:
	data []u8
	// datatypes []DataType
}

fn (mut b Encoder) add_string(data string)

adds u16 length of string in bytes + the bytes

fn (mut b Encoder) add_int(data int)

Please note that unlike C and Go, int is always a 32 bit integer. We borrow the add_u32() function to handle the encoding of a 32 bit type

fn (mut b Encoder) add_bytes(data []u8)

add bytes or bytestring

fn (mut b Encoder) add_u8(data u8)

fn (mut b Encoder) add_u16(data u16)

fn (mut b Encoder) add_u32(data u32)

fn (mut b Encoder) add_u64(data u64)

fn (mut b Encoder) add_i64(data i64)

fn (mut b Encoder) add_time(data time.Time)

fn (mut b Encoder) add_ourtime(data ourtime.OurTime)

fn (mut b Encoder) add_list_string(data []string)

fn (mut b Encoder) add_list_int(data []int)

fn (mut b Encoder) add_list_u8(data []u8)

fn (mut b Encoder) add_list_u16(data []u16)

fn (mut b Encoder) add_list_u32(data []u32)

fn (mut b Encoder) add_list_u64(data []u64)

fn (mut b Encoder) add_map_string(data map[string]string)

when complicated hash e.g. map of other object need to serialize each sub object

fn (mut b Encoder) add_map_bytes(data map[string][]u8)

when complicated hash e.g. map of other object need to serialize each sub object