LiquidVRAM

Paid feature

This is a paid feature. Please check the pricing page for our different subscription tiers.

Hydra supports reading and writing LiquidApps' vRAM tables. Unlike standard eosio::multi_index tables, vRAM tables are stored and encoded across multiple tables. Hydra decodes this table data on the fly when using getTableRowsScoped, so they can be used like any other EOSIO multi-index table. At the same time, it's possible to use Hydra's exisiting fixture loading feature to load initial vRAM table data.

Consider a token contract storing the user account balances in a vRAM table:

coldtoken#

TABLE account {
    asset    balance;
    uint64_t primary_key()const { return balance.symbol.code().raw(); }
};

typedef dapp::multi_index<"accounts"_n, account> accounts_t;
typedef eosio::multi_index<".accounts"_n, account> accounts_t_v_abi;
TABLE shardbucket {
    std::vector<char> shard_uri;
    uint64_t shard;
    uint64_t primary_key() const { return shard; }
};
typedef eosio::multi_index<"accounts"_n, shardbucket> accounts_t_abi;

Reading table data#

The transfer function can be tested by simply reading the accounts table that the dapp::multi_index was defined on:

test("it can transfer the vram balance", async () => {
  expect.assertions(1);

  await vram.contract.transfer(
    {
      from: vram.accountName,
      to: user.accountName,
      quantity: `1.0000 TOKENS`,
      memo: `transfer balance`,
    },
    [{ actor: vram.accountName, permission: `active` }]
  );

  // table is actually a shard table but hydra decodes it on the fly
  expect(vram.getTableRowsScoped(`accounts`)).toEqual({
    user: [
      {
        balance: "1.0000 TOKENS",
      },
    ],
    vram: [
      {
        balance: "41.0000 TOKENS",
      },
    ],
  });
});

Under the hood, Hydra transforms the vRAM table data to what one would expect.

Loading table data#

Using the hydraload action, one can inject data into any vRAM table. Enable it by including the header file and annotating each dapp::multi_index type:

// #include it at the top
#include "hydra.hpp"

CONTRACT_START()
 public:
  // ...
  // HYDRA_FIXUTRE_ACTION takes tuples of dapp multi-indexes as arguments
  typedef dapp::multi_index<"accounts"_n, account> accounts_t;

  HYDRA_FIXTURE_ACTION(
      ((accounts)(account)(accounts_t))
  )

  // ...

// don't forget to add hydraload to the list of actions here too (!)
CONTRACT_END((create)(issue)(transfer)(hydraload))

Afterwards, specify the data to load in your tests:

test("it loads vram fixtures", async () => {
  expect.assertions(1);

  await vram.resetTables();

  const tx = await vram.loadFixtures(`accounts`, {
    user: [
      {
        balance: "1.0000 TOKENS",
      },
    ],
    vram: [
      {
        balance: "41.0000 TOKENS",
      },
    ],
  });

  // table is actually a shard table but hydra decodes it on the fly
  expect(vram.getTableRowsScoped(`accounts`)).toEqual({
    user: [
      {
        balance: "1.0000 TOKENS",
      },
    ],
    vram: [
      {
        balance: "41.0000 TOKENS",
      },
    ],
  });
});