native-proton

Makes things smaller

install

npm i @vandeurenglenn/proto-array

usage

import { encode, decode } from '@vandeurenglenn/proto-array'

const proto = {
  index: 0,
  hash: '0xf1a1'
}

const data = {
  index: 5,
  hash: '0xa1f1'
}

const encoded = encode(proto, data)
const decoded = decode(proto, encoded)

proto key syntax

Keys in the proto object control field behavior using a ?-separated syntax:

| Syntax | Description | |---|---| | key | Required field | | key? | Optional field — omitted from output if empty/undefined | | key??min:N | Optional field with minimum length of N |

supported types

The proto value determines the field type. Use any of the following:

| Value | Encoded type | |---|---| | '' or any string | string | | 0 or any number | number | | true / false | boolean | | BigInt(0) | bigint | | {} or new Object() | object (JSON) | | [] or new Array() | array (JSON) | | new Uint8Array() | uint8Array (raw bytes) |

compression

Pass true as the third argument to encode/decode to enable pako deflate compression:

const compressed = encode(proto, data, true)
const decoded = decode(proto, compressed, true)

example

import { encode, decode } from '@vandeurenglenn/proto-array'

const proto = {
  'hash?': '',          // optional string
  input: { test: 1 },  // required object
  count: 0,            // required number
  id: BigInt(1),       // required bigint
  bytes: new Uint8Array() // required raw bytes
}

const encoded = encode(proto, {
  input: { test: 1 },
  count: 42,
  id: BigInt(9),
  bytes: new Uint8Array(8)
})

const decoded = decode(proto, encoded)

explanation

When encoding, property names are excluded from the output — only values are stored in order using the provided proto as a schema. When decoding, property names are restored from the proto. This produces significantly smaller output than JSON, especially for repeated messages with the same shape.