chavinci-core

A modern Chavinci Core REST and RPC client to execute administrative tasks, multiwallet operations and queries about network and the blockchain.

Status

Installation

Install the package via yarn:

yarn add chavinci-core

or via npm:

Install the package via npm:

npm install chavinci-core --save

or via bun:

Install the package via bun:

bun install chavinci-core 

Usage

Client(...args)

Arguments

1. [agentOptions] (Object): Optional agent options to configure SSL/TLS.
2. [headers=false] (boolean): Whether to return the response headers.
3. [host=localhost] (string): The host to connect to.
4. [logger=debugnyan('chavinci-core')] (Function): Custom logger (by default, debugnyan).
5. [network=mainnet] (string): The network
6. [password] (string): The RPC server user password.
7. [port=[network]] (string): The RPC server port.
8. [ssl] (boolean|Object): Whether to use SSL/TLS with strict checking (boolean) or an expanded config (Object).
9. [ssl.enabled] (boolean): Whether to use SSL/TLS.
10. [ssl.strict] (boolean): Whether to do strict SSL/TLS checking (certificate must match host).
11. [timeout=30000] (number): How long until the request times out (ms).
12. [username] (number): The RPC server user name.
13. [version] (string): Which version to check methods for (read more).
14. [wallet] (string): Which wallet to manage (read more).

Examples

Using network mode

The network will automatically determine the port to connect to, just like the chavincid and chavinci-cli commands.

const Client = require('chavinci-core');
const client = new Client({ network: 'regtest' });

Setting a custom port

const client = new Client({ port: 28332 });

Connecting to an SSL/TLS server with strict checking enabled

By default, when ssl is enabled, strict checking is implicitly enabled.

const fs = require('fs');
const client = new Client({
  agentOptions: {
    ca: fs.readFileSync('/etc/ssl/chavincid/cert.pem')
  },
  ssl: true
});

Connecting to an SSL/TLS server without strict checking enabled

const client = new Client({
  ssl: {
    enabled: true,
    strict: false
  }
});

Using promises to process the response

client.getInfo().then((help) => console.log(help));

Using callbacks to process the response

Callback support was removed. Since every method returns a Promise, callbackify() (>[email protected]) can be used, or for older node versions you can use the npm package callbackify.

util.callbackify(() => client.getInfo())((error, help) => console.log(help));

Returning headers in the response

For compatibility with other Chavinci Core clients.

const client = new Client({ headers: true });

// Promise style with headers enabled:
client.getInfo().then(([body, headers]) => console.log(body, headers));

// Await style based on promises with headers enabled:
const [body, headers] = await client.getInfo();

Named parameters

Since version v0.14.0, it is possible to send commands via the JSON-RPC interface using named parameters instead of positional ones. This comes with the advantage of making the order of arguments irrelevant. It also helps improving the readability of certain function calls when leaving out arguments for their default value.

You must provide a version in the client arguments to enable named parameters.

const client = new Client({ version: '0.15.1' });

For instance, take the getBalance() call written using positional arguments:

const balance = await new Client().getBalance('*', 0);

It is functionally equivalent to using the named arguments account and minconf, leaving out include_watchonly (defaults to false):

const balance = await new Client({ version: '0.15.1' }).getBalance({
  account: '*',
  minconf: 0
});

This feature is available to all JSON-RPC methods that accept arguments.

Floating point number precision in JavaScript

Due to JavaScript's limited floating point precision, all big numbers (numbers with more than 15 significant digits) are returned as strings to prevent precision loss. This includes both the RPC and REST APIs.

Multiwallet

Since Chavinci Core v0.15.0, it's possible to manage multiple wallets using a single daemon. This enables use-cases such as managing a personal and a business wallet simultaneously in order to simplify accounting and accidental misuse of funds.

Historically, the accounts feature was supposed to offer similar functionality, but it has now been replaced by this more powerful feature.

To enable Multi Wallet support, start by specifying the number of added wallets you would like to have available and loaded on the server using the -wallet argument multiple times. For convenience, the chavinci-core docker image will be used, but it's not a requirement:

docker run --rm -it -p 18332:18332 chavinci-chain/chavinci-core:0.15-alpine \
  -printtoconsole \
  -server \
  -rpcauth='foo:e1fcea9fb59df8b0388f251984fe85$26431097d48c5b6047df8dee64f387f63835c01a2a463728ad75087d0133b8e6' \
  -regtest \
  -wallet=wallet1.dat \
  -wallet=wallet2.dat \
  -rpcallowip=172.17.0.0/16

Notice the rpcauth hash which has been previously generated for the password j1DuzF7QRUp-iSXjgewO9T_WT1Qgrtz_XWOHCMn_O-Y=. Do not copy and paste this hash ever beyond this exercise.

Instantiate a client for each wallet and execute commands targeted at each wallet:

const Client = require('chavinci-core');

const wallet1 = new Client({
  network: 'regtest',
  wallet: 'wallet1.dat',
  username: 'foo',
  password: 'j1DuzF7QRUp-iSXjgewO9T_WT1Qgrtz_XWOHCMn_O-Y='
});

const wallet2 = new Client({
  network: 'regtest',
  wallet: 'wallet2.dat',
  username: 'foo',
  password: 'j1DuzF7QRUp-iSXjgewO9T_WT1Qgrtz_XWOHCMn_O-Y='
});

(async function() {
  await wallet2.generate(100);

  console.log(await wallet1.getBalance());
  // => 0
  console.log(await wallet2.getBalance());
  // => 50
}());

Version Checking

By default, all methods are exposed on the client independently of the version it is connecting to. This is the most flexible option as defining methods for unavailable RPC calls does not cause any harm and the library is capable of handling a Method not found response error correctly.

const client = new Client();

client.command('foobar');
// => RpcError: -32601 Method not found

However, if you prefer to be on the safe side, you can enable strict version checking. This will validate all method calls before executing the actual RPC request:

const client = new Client({ version: '0.12.0' });

client.getHashesPerSec();
// => Method "gethashespersec" is not supported by version "0.12.0"

If you want to enable strict version checking for the bleeding edge version, you may set a very high version number to exclude recently deprecated calls:

const client = new Client({ version: `${Number.MAX_SAFE_INTEGER}.0.0` });

client.getWork();
// => Throws 'Method "getwork" is not supported by version "9007199254740991.0.0"'.

To avoid potential issues with prototype references, all methods are still enumerable on the library client prototype.

RPC

Start the chavincid with the RPC server enabled and optionally configure a username and password:

docker run --rm -it chavinci-chain/chavinci-core:0.12-alpine -printtoconsole -rpcuser=foo -rpcpassword=bar -server

These configuration values may also be set on the chavinci.conf file of your platform installation.

By default, port 8332 is used to listen for requests in mainnet mode, or 18332 in testnet and regtest modes (the regtest change will be changed to 18443 in 0.16). Use the network property to initialize the client on the desired mode and automatically set the respective default port. You can optionally set a custom port of your choice too.

The RPC services binds to the localhost loopback network interface, so use rpcbind to change where to bind to and rpcallowip to whitelist source IP access.

Methods

All RPC methods are exposed on the client interface as a camelcase'd version of those available on chavincid (see examples below).

For a more complete reference about which methods are available, check the RPC documentation on the Chavinci Core Developer Reference website.

Examples

client.createRawTransaction([{ txid: '1eb590cd06127f78bf38ab4140c4cdce56ad9eb8886999eb898ddf4d3b28a91d', vout: 0 }], { 'mgnucj8nYqdrPFh2JfZSB1NmUThUGnmsqe': 0.13 });
client.sendMany('test1', { mjSk1Ny9spzU2fouzYgLqGUD8U41iR35QN: 0.1, mgnucj8nYqdrPFh2JfZSB1NmUThUGnmsqe: 0.2 }, 6, 'Example Transaction');
client.sendToAddress('mmXgiR6KAhZCyQ8ndr2BCfEq1wNG2UnyG6', 0.1,  'sendtoaddress example', 'Nemo From Example.com');

Batch requests

Batch requests are support by passing an array to the command method with a method and optionally, parameters. The return value will be an array with all the responses.

const batch = [
  { method: 'getnewaddress', parameters: [] },
  { method: 'getnewaddress', parameters: [] }
]

new Client().command(batch).then((responses) => console.log(responses)));

// Or, using ES2015 destructuring.
new Client().command(batch).then(([firstAddress, secondAddress]) => console.log(firstAddress, secondAddress)));

Note that batched requests will only throw an error if the batch request itself cannot be processed. However, each individual response may contain an error akin to an individual request.

const batch = [
  { method: 'foobar', parameters: [] },
  { method: 'getnewaddress', parameters: [] }
]

new Client().command(batch).then(([address, error]) => console.log(address, error)));
// => `mkteeBFmGkraJaWN5WzqHCjmbQWVrPo5X3, { [RpcError: Method not found] message: 'Method not found', name: 'RpcError', code: -32601 }`.

REST

Support for the REST interface is still experimental and the API is still subject to change. These endpoints are also unauthenticated so there are certain risks which you should be aware, specifically of leaking sensitive data of the node if not correctly protected.

Error handling is still fragile so avoid passing user input.

Start the chavincid with the REST server enabled:

docker run --rm -it chavinci-chain/chavinci-core:0.12-alpine -printtoconsole -server -rest

These configuration values may also be set on the chavinci.conf file of your platform installation. Use txindex=1 if you'd like to enable full transaction query support (note: this will take a considerable amount of time on the first run).

Methods

Unlike RPC methods which are automatically exposed on the client, REST ones are handled individually as each method has its own specificity. The following methods are supported:

• getBlockByHash
• getBlockHeadersByHash
• getBlockchainInformation
• getMemoryPoolContent
• getMemoryPoolInformation
• getTransactionByHash
• getUnspentTransactionOutputs

getBlockByHash(hash, [options])

Given a block hash, returns a block, in binary, hex-encoded binary or JSON formats.

Arguments

1. hash (string): The block hash.
2. [options] (Object): The options object.
3. [options.extension=json] (string): Return in binary (bin), hex-encoded binary (hex) or JSON (json) format.

Example

client.getBlockByHash('0f9188f13cb7b2c71f2a335e3a4fc328bf5beb436012afca590b1a11466e2206', { extension: 'json' });

getBlockHeadersByHash(hash, count, [options])

Given a block hash, returns amount of block headers in upward direction.

Arguments

1. hash (string): The block hash.
2. count (number): The number of blocks to count in upward direction.
3. [options] (Object): The options object.
4. [options.extension=json] (string): Return in binary (bin), hex-encoded binary (hex) or JSON (json) format.

Example

client.getBlockHeadersByHash('0f9188f13cb7b2c71f2a335e3a4fc328bf5beb436012afca590b1a11466e2206', 1, { extension: 'json' });

getBlockchainInformation()

Returns various state info regarding block chain processing.

Example

client.getBlockchainInformation();

getMemoryPoolContent()

Returns transactions in the transaction memory pool.

Example

client.getMemoryPoolContent();

getMemoryPoolInformation()

Returns various information about the transaction memory pool. Only supports JSON as output format.

• size: the number of transactions in the transaction memory pool.
• bytes: size of the transaction memory pool in bytes.
• usage: total transaction memory pool memory usage.

Example

client.getMemoryPoolInformation();

getTransactionByHash(hash, [options])

Given a transaction hash, returns a transaction in binary, hex-encoded binary, or JSON formats.

Arguments

1. hash (string): The transaction hash.
2. [options] (Object): The options object.
3. [options.summary=false] (boolean): Whether to return just the transaction hash, thus saving memory.
4. [options.extension=json] (string): Return in binary (bin), hex-encoded binary (hex) or JSON (json) format.

Example

client.getTransactionByHash('b4dd08f32be15d96b7166fd77afd18aece7480f72af6c9c7f9c5cbeb01e686fe', { extension: 'json', summary: false });

getUnspentTransactionOutputs(outpoints, [options])

Query unspent transaction outputs (UTXO) for a given set of outpoints. See BIP64 for input and output serialisation.

Arguments

1. outpoints (array<Object>|Object): The outpoint to query in the format { id: '<txid>', index: '<index>' }.
2. [options] (Object): The options object.
3. [options.extension=json] (string): Return in binary (bin), hex-encoded binary (hex) or JSON (json) format.

Example

client.getUnspentTransactionOutputs([{
  id: '0f9188f13cb7b2c71f2a335e3a4fc328bf5beb436012afca590b1a11466e2206',
  index: 0
}, {
  id: '0f9188f13cb7b2c71f2a335e3a4fc328bf5beb436012afca590b1a11466e2206',
  index: 1
}], { extension: 'json' })

SSL

This client supports SSL out of the box. Simply pass the SSL public certificate to the client and optionally disable strict SSL checking which will bypass SSL validation (the connection is still encrypted but the server it is connecting to may not be trusted). This is, of course, discouraged unless for testing purposes when using something like self-signed certificates.

Generating a self-signed certificates for testing purposes

Please note that the following procedure should only be used for testing purposes.

Generate an self-signed certificate together with an unprotected private key:

openssl req -x509 -newkey rsa:4096 -keyout key.pem -out cert.pem -days 3650 -nodes

Connecting via SSL

On Chavinci Core <0.12, you can start the chavincid RPC server directly with SSL:

docker run --rm -it -v $(PWD)/ssl:/etc/ssl chavinci-chain/chavinci-core:0.11-alpine -printtoconsole -rpcuser=foo -rpcpassword=bar -rpcssl -rpcsslcertificatechainfile=/etc/ssl/chavincid/cert.pem -rpcsslprivatekeyfile=/etc/ssl/chavincid/key.pem -server

On Chavinci Core >0.12, use must use stunnel (brew install stunnel or sudo apt-get install stunnel4) or an HTTPS reverse proxy to configure SSL since the built-in support for SSL has been removed. The trade off with stunnel is performance and simplicity versus features, as it lacks more powerful capacities such as Basic Authentication and caching which are standard in reverse proxies.

You can use stunnel by configuring stunnel.conf with the following service requirements:

[chavinci]
accept = 28332
connect = 18332
cert = /etc/ssl/chavincid/cert.pem
key = /etc/ssl/chavincid/key.pem

The key option may be omitted if you concatenating your private and public certificates into a single stunnel.pem file.

On some versions of stunnel it is also possible to start a service using command line arguments. The equivalent would be:

stunnel -d 28332 -r 127.0.0.1:18332 -p stunnel.pem -P ''

Then pass the public certificate to the client:

const Client = require('chavinci-core');
const fs = require('fs');
const client = new Client({
  agentOptions: {
    ca: fs.readFileSync('/etc/ssl/chavincid/cert.pem')
  },
  port: 28332,
  ssl: true
});

Logging

By default, all requests made with chavinci-core are logged using uphold/debugnyan with chavinci-core as the logging namespace.

Please note that all sensitive data is obfuscated before calling the logger.

Example

Example output defining the environment variable DEBUG=chavinci-core:

const client = new Client();

client.getTransactionByHash('b4dd08f32be15d96b7166fd77afd18aece7480f72af6c9c7f9c5cbeb01e686fe');

// {
//   "name": "chavinci-core",
//   "hostname": "localhost",
//   "pid": 57908,
//   "level": 20,
//   "request": {
//     "headers": {
//       "host": "localhost:8332",
//       "accept": "application/json"
//     },
//     "id": "82cea4e5-2c85-4284-b9ec-e5876c84e67c",
//     "method": "GET",
//     "type": "request",
//     "uri": "http://localhost:8332/rest/tx/b4dd08f32be15d96b7166fd77afd18aece7480f72af6c9c7f9c5cbeb01e686fe.json"
//   },
//   "msg": "Making request 82cea4e5-2c85-4284-b9ec-e5876c84e67c to GET http://localhost:8332/rest/tx/b4dd08f32be15d96b7166fd77afd18aece7480f72af6c9c7f9c5cbeb01e686fe.json",
//   "time": "2017-02-07T14:40:35.020Z",
//   "v": 0
// }

Custom logger

A custom logger can be passed via the logger option and it should implement bunyan's log levels.

Tests

Currently the test suite is tailored for Docker (including docker-compose) due to the multitude of different chavincid configurations that are required in order to get the test suite passing.

To test using a local installation of node.js but with dependencies (e.g. chavincid) running inside Docker:

npm run dependencies
npm test

To test using Docker exclusively (similarly to what is done in Travis CI):

npm run testdocker

Release

npm version [<newversion> | major | minor | patch] -m "Release %s"

License

MIT