Skip to content

Emurgo/yoroi-graphql-migration-backend

Repository files navigation

yoroi-graphql-migration-backend

Background

Adrestia is the new codename for all Cardano tooling and includes tooling that will have long-term support.

Adrestia recommends powering light clients with GraphQL. This is done by using cardano-db-sync to dump the database to a database and then using cardano-graphql to serve that data.

Purpose of this project

This provides an API useful to light wallets for Cardano.

For some endpoints, we use SQL queries directly (either because they aren't possible in GraphQL or because we need the performance of raw SQL queries)

However, we initially wanted every query to be in GraphQL so that any project using yoroi-graphql-migration-backend can use our REST API as a intermediate step to running GraphQL queries directly.

Requirements

To run this, you will need to run the following

  1. cardano-node
  2. cardano-db-sync

Building

Development build (with hot reloading):

# install the right version of Node
nvm use
nvm install

# install dependencies
npm install

# run the server
npm run dev

The server will then run at http://localhost:8082. You can query using curl (ex: curl http://localhost:8082/bestblock)

This is no easy way to configure runtime settings. However, you can edit lines 23-26 of src/index.ts to change port settings, graphql uri, et cetera.

Tests

There are test which run by querying your local cardano-db-sync and cardano-graphql. You can run it by doing the following

# run the server on a terminal window
npm run dev

# run the tests on a different terminal
npm run test

API

For addresses, refer to CIP5 for how they should be encoded. Notably, we support

  • addr & addr_test
  • stake & stake_test
  • addr_vkh

We recommend querying using payment key hashes (addr_vkh) when possible (otherwise you may miss addresses for a wallet such as mangled based addresses or other address types like pointer addresses)

txs/utxoForAddresses Input

Up to 50 addresses in the request

{
  // byron addresses, bech32 address, bech32 stake addresses or addr_vkh
  addresses: Array<string>
}

Output

Array<{
  utxo_id: string, // concat tx_hash and tx_index
  tx_hash: string,
  tx_index: number,
  block_num: number, // NOTE: not slot_no
  receiver: string,
  amount: string,
  dataHash: string,
  assets: Asset[],
}>
v2/txs/utxoAtPoint This endpoint is based on the current `/txs/utxoForAddresses`. It adds capabilities for passing a reference block and pagination information.

This endpoint basically takes a "snapshot" of how was the UTxO information for the given addresses up to referenceBlock, meaning UTxOs created in transactions from blocks after referenceBlockHash won't be included, but also that UTxOs spent only in blocks after referenceBlockHash will actually be included in the response.

OBS: we don't actually take snapshots, such information can be inferred from on-chain data.

Input

Up to 50 addresses in the request

{
  // byron addresses, bech32 address, bech32 stake addresses or addr_vkh
  addresses: Array<string>,
  page: number,
  pageSize: number,
  referenceBlockHash?: string // the hash of the block
}

Output

Array<{
  utxo_id: string, // concat tx_hash and tx_index
  tx_hash: string,
  tx_index: number,
  block_num: number, // NOTE: not slot_no
  receiver: string,
  amount: string,
  assets: Asset[],
}>
v2/txs/utxoDiffSincePoint Returns a diff of inputs and outputs between two points. See the comments next to the request parameters for more information. This endpoint is better used in combination with `v2/txs/utxoAtPoint`. After making a request to `v2/txs/utxoAtPoint`, the clients can keep their local copy of the UTxO state by calling `v2/txs/utxoDiffSincePoint` with `afterPoint` or `afterBestBlocks` being built using the block from the previous request as a reference point and making any changes established by the diff, so essentially discarding outputs which have been returned in the diff as inputs and including the new outputs returned by the diff.

Input

{
  // byron addresses, bech32 address, bech32 stake addresses or addr_vkh
  addresses: Array<string>,
  untilBlockHash: string, // only transactions up to this block (including it) will be considered for generating the diff
  // when starting to call this endpoint, the client should set the `blockHash` field. Then the client should pass the returned `lastDiffPointSelected` as `afterPoint` for the next page.
  afterPoint: {
    blockHash: string, // only transactions AFTER this clock will be considered for generating the diff
    ...
  },
  afterBestBlocks?: Array<string> // only transactions after the latest block from this array will be included. The inclusion of `afterBestBlocks` in the request will also add 2 new fields to the response (see the response bellow for more details)
  diffLimit: number // number of diff items to return
}

Output

{
  diffItems: Array<{
    type: "output" | "input",
    id: string, // in the format {TX hash}:{output index}
    amount: string,
    assets?: Asset[], // only included for outputs
    block_num?: number // only included for outputs
  }>,
  // An opaque object to mark the end of the returned page (whose size is limited by `diffLimit` of the input). The client should pass this object as the `afterPoint` input for the next page.
  lastDiffPointSelected: Object,
  lastFoundBestBlock?: string, // only included if `afterBestBlocks` was supplied. This will be the latest found block from `afterBestBlocks`
  lastFoundSafeBlock?: string // only included if `afterBestBlocks` was supplied. This will be the latest safe block from `afterBestBlocks`, safe block being the block with the highest depth up to a maximum, determined at runtime by configuration
}
account/registrationHistory Input
{
  // bech32 stake address
  addresses: Array<string>
}

Output

{
  [addresses: string]: Array<{|
    slot: number,
    txIndex: number,
    certIndex: number,
    certType: "StakeRegistration"|"StakeDeregistration",
  |}>
}
account/state Input
{
  // bech32 stake addresses
  addresses: Array<string>
}

Output

{
  [addresses: string]: null | {|
    poolOperator: null, // not implemented yet
    remainingAmount: string, // current remaining awards
    rewards: string, //all the rewards every added (not implemented yet)
    withdrawals: string // all the withdrawals that have ever happened (not implemented yet)
  |}
}
account/rewardHistory Input
{
  // bech32 stake address
  addresses: Array<string>
}

Output

{
  [addresses: string]: Array<{
    epoch: number,
    reward: string,
    poolHash: string,
  }>
}
pool/info Input
{
  poolIds: Array<string> // operator key (pool id)
}

Output

{
  [poolId: string]: null | {|
    info: {
      name?: string,
      description?: string,
      ticker?: string,
      ... // other stuff from SMASH.
    },
    history: Array<{|
      epoch: number,
      slot: number,
      tx_ordinal: number
      cert_ordinal: number
      payload: Certificate // see `v2/txs/history`
    |}>
  |}
}
pool/delegationHistory Input
{
  poolRanges: Dictionary<string, Dictionary<string, {fromEpoch: number, toEpoch?: number}>> // operator key (pool id), fromEpoch and toEpoch are inclusive
}

Output

[
  {|
      epoch: number;    
      poolHash: string;
      slot: number;
      tx_ordinal: number
      cert_ordinal: number;
      payload: Certificate | null;
      info: {
          name?: string;
          description?: string;
          ticket?: string;
          homepage?: string;
      }
  |}
]
txs/utxoSumForAddresses Input

Up to 50 addresses in the request

{
  addresses: Array<string>
}

Output

{
  sum: ?string,
  tokensBalance: [
    amount: string,
    assetId: string
  ]
}
v2/addresses/filterUsed Input

Up to 50 addresses in the request

{
  // byron addresses, bech32 address or addr_vkh
  addresses: Array<string>
}

Output

Array<string>
v2/txs/history Since short rollbacks are common (by design) in Cardano Shelley, your app needs to be ready for this. The pagination mechanism should help make this easy for you.

To handle pagination, we use an after and untilBlock field that refers to positions inside the chain. Usually, pagination works as follows:

  1. Query the bestblock endpoint to get the current tip of the chain (and call this untilBlock)
  2. Look up the last transaction your application has saved locally (and call this after)
  3. Query everything between untilBlock and after. If untilBlock no long exists, requery. If after no long exists, mark the transaction as failed and re-query with an earlier transaction
  4. If more results were returned than the maximum responses you can receive for one query, find the most recent transction included in the response and set this as the new after and then query again (with the same value for untilBlock)

Note: this endpoint will throw an error if either the untilBlock or after fields no longer exist inside the blockchain (allowing your app to handle rollbacks). Notably, the error codes are

  • 'REFERENCE_BLOCK_MISMATCH'
  • 'REFERENCE_TX_NOT_FOUND'
  • 'REFERENCE_BEST_BLOCK_MISMATCH'

Input

Up to 50 addresses in the request

{
  // byron addresses, bech32 address, bech32 stake addresses or addr_vkh
  addresses: Array<string>,
  // omitting "after" means you query starting from the genesis block
  after?: {
    block: string, // block hash
    tx: string, // tx hash
  },
  untilBlock: string, // block hash - inclusive
}

Output

Up to 50 transactions are returned. Use pagination with the after field to get more.

Array<{
  // information that is only present if block is included in the blockchain
  block_num: null | number,
  block_hash: null | string,
  tx_ordinal: null | number,
  time: null | string, // timestamp with timezone
  epoch: null | number,
  slot: null | number,

  // information that is always present
  type: 'byron' | 'shelley',
  hash: string,
  last_update: string, // timestamp with timezone
  tx_state: 'Successful' | 'Failed' | 'Pending',
  inputs: Array<{ // these will be ordered by the input transaction id asc
    address: string,
    amount: string,
    id: string, // concatenation of txHash || index
    index: number,
    txHash: string, 
    assets: Asset[]
  }>,
  collateral_inputs: Array<{
    address: string,
    amount: string,
    id: string, // concatenation of txHash || index
    index: number,
    txHash: string,
    assets: Asset[]
  }>,
  outputs: Array<{ //these will be ordered by transaction index asc.
    address: string,
    amount: string,
    dataHash: string,
    assets: Asset[]
  }>,
  withdrawals: Array<{| address: string, // hex
    amount: string
  |}>,
  certificates: Array<{|
    kind: 'StakeRegistration',
    rewardAddress:string, //hex
  |} | {|
    kind: 'StakeDeregistration',
    rewardAddress:string, // hex
  |} | {|
    kind: 'StakeDelegation',
    rewardAddress:string, // hex
    poolKeyHash: string, // hex
  |} | {|
    kind: 'PoolRegistration',
    poolParams: {|
      operator: string, // hex
      vrfKeyHash: string, // hex
      pledge: string,
      cost: string,
      margin: number,
      rewardAccount: string, // hex
      poolOwners: Array<string>,  // hex
      relays: Array<{| ipv4: string|null,
        ipv6: string|null,
        dnsName: string|null,
        dnsSrvName: string|null,
        port: string|null |}>,
      poolMetadata: null | {|
        url: string,
        metadataHash: string, //hex
      |},
    |},
  |} | {|
    type: 'PoolRetirement',
    poolKeyHash: string, // hex
    epoch: number,
  |} {|
    type: 'MoveInstantaneousRewardsCert',
    rewards: { [addresses: string]: string } // dictionary of stake addresses to their reward amounts in lovelace
    pot: 0 | 1 // 0 = Reserves, 1 = Treasury
  |}>,
  valid_contract: boolean, // False if the contract is invalid. True if the contract is valid or there is no contract.
  script_size: number, // The sum of the script sizes (in bytes) of scripts in the transaction.
}>
v2/txs/get This endpoint returns the transactions' information given their hashes (or ids).

Since short rollbacks are common (by design) in Cardano Shelley, your app needs to be ready for this.

Input

Up to 100 tx hashes in the request

{
 txHashes: string[],
}

Output

Up to 100 transactions are returned. Transactions which are not yet on-chain will be ignored and won't be included in the response. The txHashes sent in the request are transformed into keys under the txs object, and the value corresponding to this key is the transaction information

txs: {
  "<txHash>": {
    // information that is only present if block is included in the blockchain
    block_num: null | number,
    block_hash: null | string,
    tx_ordinal: null | number,
    time: null | string, // timestamp with timezone
    epoch: null | number,
    slot: null | number,

    // information that is always present
    type: 'byron' | 'shelley',
    hash: string,
    last_update: string, // timestamp with timezone
    tx_state: 'Successful' | 'Failed' | 'Pending',
    inputs: Array<{ // these will be ordered by the input transaction id asc
      address: string,
      amount: string,
      id: string, // concatenation of txHash || index
      index: number,
      txHash: string, 
      assets: Asset[]
    }>,
    collateral_inputs: Array<{
      address: string,
      amount: string,
      id: string, // concatenation of txHash || index
      index: number,
      txHash: string,
      assets: Asset[]
    }>,
    outputs: Array<{ //these will be ordered by transaction index asc.
      address: string,
      amount: string,
      dataHash: string,
      assets: Asset[]
    }>,
    withdrawals: Array<{| address: string, // hex
      amount: string
    |}>,
    certificates: Array<{|
      kind: 'StakeRegistration',
      rewardAddress:string, //hex
    |} | {|
      kind: 'StakeDeregistration',
      rewardAddress:string, // hex
    |} | {|
      kind: 'StakeDelegation',
      rewardAddress:string, // hex
      poolKeyHash: string, // hex
    |} | {|
      kind: 'PoolRegistration',
      poolParams: {|
        operator: string, // hex
        vrfKeyHash: string, // hex
        pledge: string,
        cost: string,
        margin: number,
        rewardAccount: string, // hex
        poolOwners: Array<string>,  // hex
        relays: Array<{| ipv4: string|null,
          ipv6: string|null,
          dnsName: string|null,
          dnsSrvName: string|null,
          port: string|null |}>,
        poolMetadata: null | {|
          url: string,
          metadataHash: string, //hex
        |},
      |},
    |} | {|
      type: 'PoolRetirement',
      poolKeyHash: string, // hex
      epoch: number,
    |} {|
      type: 'MoveInstantaneousRewardsCert',
      rewards: { [addresses: string]: string } // dictionary of stake addresses to their reward amounts in lovelace
      pot: 0 | 1 // 0 = Reserves, 1 = Treasury
    |}>,
    valid_contract: boolean, // False if the contract is invalid. True if the contract is valid or there is no contract.
    script_size: number, // The sum of the script sizes (in bytes) of scripts in the transaction.
  }
}
v2/bestblock Input

None (GET request)

Output

{
  // 0 if no blocks in db
  height: number,
  // null when no blocks in db
  epoch: null | number,
  slot: null | number,
  globalSlot: null | number,
  hash: null | string,
}
GET v2/tipStatus Input

None (GET request)

Output

{
  safeBlock: {
    height: number,
    epoch: null | number,
    slot: null | number,
    globalSlot: null | number,
    hash: null | string,
  },
  bestBlock:  {
    height: number,
    epoch: null | number,
    slot: null | number,
    globalSlot: null | number,
    hash: null | string,
  }
}
POST v2/tipStatus Input
{
  reference: {
    bestBlocks: string[]
  }
}

Output

{
  safeBlock: string,
  bestBlock: string,
  reference: {
    lastFoundSafeBlock: string,
    lastFoundBestBlock: string
  }
}
txs/signed Input
{
  // base64 encoding of the transaction
  signedTx: string,
}

Output

{
  // this is calculated based on the submitted `signedTx`, and will be an exact match of the transaction ID on the blockchain once the transaction is confirmed
  txId: string
}
tx/status This endpoint is used to return the current on-chain status of up to 100 transactions, given their ids. Currently, we return only the depth, meaning the number of blocks on top of the transactions

Input

{
  "txHashes": string[]
}

Output: the txHashes sent in the request are transformed into keys under the depth field, and the value corresponding to this key will be the number of blocks on top of the transaction

{
  depth: {
    "<txHash>": number
  },
  submissionStatus?: {
    "<txHash>": "WAITING" | "FAILED" | "MAX_RETRY_REACHED" | "SUCCEESS"
  }
}
status

This endpoint is used to test whether or not the server can still be reached and get any manually flagged errors.

Input

None (GET request)

Output

{
  isServerOk: boolean, // heartbeat endpoint for server. IF you want the node status, use v2/importerhealthcheck instead
  isMaintenance: boolean, // manually set and indicates you should disable ADA integration in your app until it returns false. Use to avoid weird app-side behavior during server upgrades.
  serverTime: number, // in millisecond unix time
  isQueueOnline: boolean, // indicates if the backend is using the TX queue to submit the signed transactions. Essentially, it returns true if the USE_SIGNED_TX_QUEUE env var is set to "true".
}
v2/importerhealthcheck This endpoint is used to check whether or not the underlying node is properly syncing

Input

None (GET request)

Output

200 status if things look good. Error if node is not syncing

multiAsset/supply This endpoint is used to get current supplies of given multi assets

Input

{
  // list of multi assets to get supplies of
  assets: Array<{
    policy: string,
    name: string
  }>
}

Output

{
  // current supplies of given assets.
  // entry for an asset is null if it is not found.
  supplies: {
    "${asset.policy}.${asset.name}": number | null
  }
}
txs/io/:tx_hash This endpoint is used to get inputs and outputs of a transaction with the given hash

Input

None (GET request)

Output

{
  inputs: Array<{ // these will be ordered by the input transaction id asc
    address: string,
    amount: string,
    id: string, // concatenation of txHash || index
    index: number,
    txHash: string, 
    assets: Asset[]
  }>,
  collateralInputs: Array<{
    address: string,
    amount: string,
    id: string, // concatenation of txHash || index
    index: number,
    txHash: string,
    assets: Asset[]
  }>,
  outputs: Array<{ //these will be ordered by transaction index asc.
    address: string,
    amount: string,
    dataHash: string,
    assets: Asset[]
  }>,
}
txs/io/:tx_hash/o/:index This endpoint is used to get a single output with the given index of a transaction with the given hash

Input

None (GET request)

Output

{
  output: {
    address: string,
    amount: string,
    dataHash: string,
    assets: Asset[]
  },
}
/oracles/getDatapoints This endpoint is used to return specific data (data point) of a specified oracle.

There are two usages of this endpoint - with and without source. This is because when calling with source, ticker becomes mandatory.

  1. without specifying source:

Input

{
  addresses: Array<string>, // mandatory, bech32 addresses of trusted oracles
  ticker?: string,  // optional. If not set, fetch all available tickers. If set, fetch only that ticker
  blockNum?: number, // optional. If not set, fetch latest `count` data and order desc.
  // If set, find `count` nearest (absolute) values around that block
  // e.g. with block_no = 100, count = 3 and data at blocks 85,90,100,115,135, returned data will be for blocks:
  // 100 (absolute distance 0), 90 (absolute distance 10) and 115 (absolute distance 15 - same as block 85, but more recent)
  // i.e. nearest blocks to the block specified
  // in case of a draw we display the more recent block
  count?: number, // optional, default 1, max 10
}

Output

[addresses: string]: undefined | {
  blockDistance: number | null,
  blockNumber: number,
  txHash: string,
  txIndex: number,
  payload: any, // if a ticker is specified, then array of JSON with data,
               // if not specified then all tickers are returned in form of [ticker: string]: Array<any>
}
  1. with source (and thus also ticker) specified:

Input

{
  addresses: Array<string>, // mandatory, bech32 addresses of trusted oracles
  ticker: string,  // mandatory. When filtering with source, tickers are mandatory
  blockNum?: number, // optional. If not set, fetch latest `count` data and order desc.
  // If set, find `count` nearest (absolute) values around that block
  // e.g. with block_no = 100, count = 3 and data at blocks 85,90,100,115,135, returned data will be for blocks:
  // 100 (absolute distance 0), 90 (absolute distance 10) and 115 (absolute distance 15 - same as block 85, but more recent)
  // i.e. nearest blocks to the block specified
  // in case of a draw we display the more recent block
  source: string, // mandatory. Source of the data
  count?: number, // optional, default 1, max 10
}

Output

[addresses: string]: undefined | {
  blockDistance: number | null,
  blockNumber: number,
  txHash: string,
  txIndex: number,
  payload: any, // JSON with data
}
/oracles/getTickers This endpoint is used to return all available tickers of a specified array of oracles.

Input

{
  addresses: Array<string>, // bech32 addresses of trusted oracles
}

Output

[addresses: string]: undefined | {
  Array<{
    ticker: string,
    latestBlock: number,
  }>
}
/multiAsset/policyIdExists This endpoint is used to check if given policyIds and (optionally) fingerprints already exist on chain.

Number of policyIds need to be in [0, 100]

Number of fingerprints need to be in [0, 100]

Input

{
  policyIds: Array<string>, // hex encoded policyIds that will be checked,
  fingerprints?: Array<string>, // fingerprints that will be checked,
}

Output

{
  policyIdResults: Array<{
    [policyId: string]: boolean
  }>,
  fingerprintResults?: Array<{
    [fingerprint: string]: boolean
  }>
}
/multiAsset/metadata Retrieves on-chain metadata for assets.

Input

{
  assets: Array<{
    nameHex: string,
    policy: string,
  }>
}

or

{
  assets: Array<{
    name: string,
    policy: string,
  }>
}

Prefer the first one, where the asset names are passes as hex strings via the nameHex field.

Output

{
  // the key in this case is a combination of policy and name, separated by a dot: "policy"."name"
  [asset: string]: {
    key: string, // the metadata label key. e.g. 721 in case of NFTs
    metadata: any
  }
}
/asset/:fingerprint/mintTxs Retrieves all minting transactions for the given asset fingerprint together with its metadata, if any.

Output

{
  policy: string, // hex-encoded policy
  name: string, // hex-encoded asset name
  txs: Array<{
    hash: string,
    block: {
      slot: number,
      epoch: number
    },
    metadata?: {
      key: number,
      json: any
    }
  }>
}
multiAsset/validateNFT/:fingerprint Retrieves image from an NFT and validates it. In case this was already done for the given NFT, simply return the validation results

Input

None (GET request)

Outputs:

200 OK (when the validation already happened)

{
  status: string,
  contents: Array<string>,
  originalStatus: string,
  thirdPartyReport: any
}

204 No Content (when NFT is sent for validation)

{
}
POST v2.1/txs/summaries Return transaction summaries for addresses.

Input

{
  // byron addresses, bech32 address, bech32 stake addresses or addr_vkh
  addresses: Array<string>,
  before: {
    blockHash: string, // return txs before this block
    txHash?: string, // if specified, also return txs in the block but before this tx
  },
}

Output

{
  [address: string]: [ // for each address in the input address list that has txs
    {
      txHash: string, // hash of the tx
      blockHash: string, // hash of the block containing the tx
      txBlockIndex: number, // the index of the tx in the block
      epoch: number, // epoch number of the block containing the tx
      slot: number, // absolute block number of the block containing the tx
    }
  ]
}