Awesome

Nova Substrate SDK for Android (Kotlin/Java)

Nova Substrate SDK is a native Android library to help developers build native mobile apps for Substrate-based networks, e.g. Polkadot, Kusama & parachains

Nova Substrate SDK for Android (Kotlin/Java)
Table of contents

Bip39

Bip39 is the algorithm which provides an opportunity to use a list of words, called mnemonic, instead of a raw 32 byte seed. Library provides Bip39 class to work with mnemonics:

val bip39 = Bip39()

val newMnemonic = bip39.generateMnemonic(length = MnemonicLength.TWELVE) // twelve words
val entropy = bip39.generateEntropy(newMnemonic)
val theSameMnemonic = bip39.generateMnemonic(entropy)

To generate a seed, a passphrase is needed. Technically, it is a decoded derivation path (see Junction Decoder)

val seed = bip39.generateSeed(entropy, passphrase)

JSON import/export

Library provides support for decoding/encoding account information using JSON format, compatible with Polkadot.js

Import

Using JsonSeedDecoder you can perform decoding of the imported JSON:

val decoder = JsonSeedDecoder(..)

decoder.extractImportMetaData(myJson) // does not perform full decoding (skips secret decrypting). Faster
decoder.decode(myJson, password) // performs full decoding. Slower

Export

Using JsonSeedEncoder you can generate JSON out of account information:

val encoder = JsonSeedEncoder(..)

val json = encoder.generate(keypair, seed, password, name, encryptionType, genesis, addressByte)

Extensions

Library provides several extensions, that implement most common operations.

Hex

fun ByteArray.toHexString(withPrefix: Boolean = false): String
fun String.fromHex(): ByteArray
fun String.requirePrefix(prefix: String): String
fun String.requireHexPrefix(): String

Hashing

fun ByteArray.xxHash128(): ByteArray
fun ByteArray.xxHash64(): ByteArray

fun ByteArray.blake2b512(): ByteArray
fun ByteArray.blake2b256(): ByteArray
fun ByteArray.blake2b128(): ByteArray

fun XXHash64.hash(bytes: ByteArray, seed: Long = 0): ByteArray
fun BCMessageDigest.hashConcat(bytes: ByteArray): ByteArray
fun XXHash64.hashConcat(bytes: ByteArray): ByteArray

Icon generation

There's a support for default Polkadot.js icon generation using IconGenerator:

val generator = IconGenerator()
val drawable =  generator.getSvgImage(accountId, sizeInPixels)

Junction Decoder

JunctionDecoder provides support for derivation paths:

val derivationPath: String = ...
val decoder = JunctionDecoder()

val passphrase = decoder.getPassword(derivationPath) // retrieve passphrase to use in entropy -> seed generation
val decodedPath = decoder.decodeDerivationPath(derivationPath)

Runtime

You can create storage keys easily:

val accountId: ByteArray = ..

val bondedKey = Module.Staking.Bonded.storageKey(bytes)
val accountInfoKey = Module.System.Account.storageKey(bytes)

If you're missing some specific service/module, you can define it by your own:

 object Staking : Module("Staking") {

    object ActiveEra : Service<Unit>(Staking, "ActiveEra") {

        override fun storageKey(storageArgs: Unit): String {
            return StorageUtils.createStorageKey(
                service = this,
                identifier = null
            )
        }
    }
}

Scale

Library provides a convenient DSL to deal with scale encoding/decoding. Original codec reference: Link.

Define a schema

object AccountData : Schema<AccountData>() {
    val free by uint128()
    val reserved by uint128()
    val miscFrozen by uint128()
    val feeFrozen by uint128()
}

Create and encode structs

val struct = AccountData { data ->
    data[AccountData.free] = BigDecimal("1")
    data[AccountData.reserved] = BigInteger("0")
    data[AccountData.miscFrozen] = BigInteger("0")
    data[AccountData.feeFrozen] = BigInteger("0")
}

val inHex = struct.toHexString() // encode
val asBytes = struct.toByteArray() // or as byte array

Decode and use structs

val inHex = ...
val struct = AccountData.read(inHex)
val free = struct[AccountData.free]

Standart Data Types

Library provides the support for the following data types:

Numbers: uint8, uint16, uint32, uint64, uint128, uint(nBytes), compactInt, byte, long
Primitives: bool, string
Arrays:
- sizedByteArray(n) - only content is encoded/decoded, size is thus known in advance
- byteArray - size can vary, so the size is also encoded/decoded alongside with the content
Compound types:
- vector<D> - List of objects of some data type
- optional<D> - Nullable container for other data type
- pair<D1, D2>
- enum(D1, D2, D3...) - like union in C, stores only one value at once, but this value can have different data type
- enum<E : Enum> - for classical Kotlin enum

Custom Data Types

If the decoding/encoding cannot be done using standart data types, you can create your own by extending DataType<T>:

object Delimiter : DataType<Byte>() {
    override fun conformsType(value: Any?): Boolean {
        return value is Byte && value == 0
    }

    override fun read(reader: ScaleCodecReader): Byte {
        val read = reader.readByte()

        if (read != 0.toByte()) throw java.lang.IllegalArgumentException("Delimiter is not 0")

        return 0
    }

    override fun write(writer: ScaleCodecWriter, ignored: Byte) {
        writer.writeByte(0)
    }
}

And use it in your schema using custom() keyword:

object CustomTypeTest : Schema<CustomTypeTest>() {
    val delimiter by custom(Delimiter)
}

Default values

You can supply and default values for each field in the schema:

object DefaultValues : Schema<DefaultValues>() {
    val bytes by sizedByteArray(length = 10, default = ByteArray(10))
    val text by string(default = "Default")
    val bigInteger by uint128(default = BigInteger.TEN)
}

Nullable fields

By default, all fields are non-null. However, you can use optional() to change the default behavior:

object Person : Schema<Person>() {
    val friendName by string().optional() // friendName now is Field<String?>
}

SS58

SS58 is an address format using in substate ecosystem. You can encode/decode address using SS58Encoder:

val encoder = SS58Encoder()
val address = encoder.encode(publicKey, addressByte)
val accountId = encoder.decode(address)

WSRPC

Library provides an implementation of SocketService, which simplifies communication with the node: it provides a seamless error recovery, subscription mechanism.

Initialize socket

To create a socket service, you need to provide several parameters:

val reconnector = Reconnector(..) // to configure reconnect strategy and scheduling executor
val requestExecutor = RequestExecutor(..) // to configure sending executor
val socketService = SocketService(gson, logger, websocketFactory, reconnector, requestExecutor)

Use socket

socketService.start(url) // async connect
socketService.stop() // all subscriptions/pending requests are cancelled
socketService.switchUrl(newUrl) // stops current connection and start a new one

// execute single request
socketService.executeRequest(runtimeRequest, deliveryType, object : SocketService.ResponseListener<RpcResponse> {
            override fun onNext(response: RpcResponse) {
                // success
            }

            override fun onError(throwable: Throwable) {
                // unrecoverable error happened
            }
        })

// subscribe to changes
socketService.subscribe(runtimeRequest, object : SocketService.ResponseListener<SubscriptionChange> {
                override fun onNext(response: SubscriptionChange) {
                   // change arrived
                }

                override fun onError(throwable: Throwable) {
                   // unrecoverable error happened
                }
            })

Reconnect strategy

During setup of Reconnector, you can specify a ReconnectStrategy. There are several of them bundled with library:

ConstantReconnectStrategy
LinearReconnectStrategy
ExponentialReconnectStrategy. This is a default reconnect strategy.

You can create your own strategy by implementing ReconnectStrategy interface.

Delivery type

While sending a request, you can specify a DeliveryType. Currently, there are 3 of them:

AT_LEAST_ONCE - attempts to send request until succeeded. This is a default delivery type.
AT_MOST_ONCE - send request once, reports error if attempt failed.
ON_RECONNECT - similar to AT_LEAST_ONCE, but remembers request and sends it on each reconnect. Currently used for subscription initiation.

Using with coroutines

Library has an out-of-box support for coroutines:

scope.launch {
    val response = socketService.executeAsync(request, deliveryType) // suspend function
}

socketService.subscriptionFlow(request).onEach { change ->
        // do stuff here
}.launchIn(scope)

Mappers

The mappers for most common types are provided:

scale - For scale-encoded values
scaleCollection - For list of scale-encoded values
pojo - for json values
pojoList - for list of json values

All mappers return a nullable result by default. You can add nonNull() modifier to change this behavior. In case of null result, the RpcException will be thrown.

Usage

scale().nonNull().map(response, gson)

// or with coroutines adapter
socketService.executeAsync(request, deliveryType, mapper = scale().nonNull())

License

Nova Substrate SDK is available under the Apache 2.0 license. See the LICENSE file for more info. © Novasama Technologies GmbH 2023