loader

ao Wasm Loader

This module takes an ao Wasm ArrayBuffer and returns a handle function, that given an ao-process message, will produce a result.

• Usage
  • Using a File
  • Using fetch
  • Result Object
• Contributing
  • Publish a new Version of the package

Usage

The @permaweb/ao-loader MUST receive an ArrayBuffer that contains the Wasm to be invoked:

import AoLoader from "@permaweb/ao-loader";

/* ao READ-ONLY Env Variables */
const env = {
  Process: {
    Id: "2",
    Tags: [
      { name: "Authority", value: "XXXXXX" },
    ],
  },
};

// Create the handle function that executes the Wasm
const handle = await AoLoader(wasmBinary, {
  format: "wasm32-unknown-emscripten2",
  inputEncoding: "JSON-1",
  outputEncoding: "JSON-1", 
  memoryLimit: "524288000", // in bytes
  computeLimit: 9e12.toString(),
  extensions: []
});

// To spawn a process, pass null as the buffer
const result = await handle(null, {
  Owner: "OWNER_ADDRESS",
  Target: "XXXXX",
  From: "YYYYYY",
  Tags: [
    { name: "Action", value: "Ping" },
  ],
  Data: "ping",
}, env);

// To evaluate a message on an existing process

const options = {
  format: "wasm32-unknown-emscripten2",
  inputEncoding: "JSON-1",
  outputEncoding: "JSON-1", 
  memoryLimit: "524288000", // in bytes
  computeLimit: 9e12.toString(),
  extensions: []
}
const handle = await AoLoader(wasmBinary, options);
const buffer = await LoadFromCache();

const result = await handle(buffer, {
  Owner: "OWNER_ADDRESS",
  Tags: [
    { name: "Action", value: "Balance" },
    { name: "Target", value: "vh-NTHVvlKZqRxc8LyyTNok65yQ55a_PJ1zWLb9G2JI" },
  ],
}, env);

saveToCache(result.Memory);
console.log(result.Output);

Using a File

You can use fs to a load a Wasm file from the local filesystem as a Node Buffer. Since a Node Buffer implements ArrayBuffer, it can be passed directly to the AoLoader directly!

To get back a Node Buffer, make sure to NOT pass an encoding parameter to readFile

import AoLoader from "@permaweb/ao-loader";
import fs from "fs";

async function main() {
  const wasmBinary = fs.readFileSync("process.wasm");
  const options = {
    format: "wasm32-unknown-emscripten2",
    inputEncoding: "JSON-1",
    outputEncoding: "JSON-1", 
    memoryLimit: "524288000", // in bytes
    computeLimit: 9e12.toString(),
    extensions: []
  }
  const handle = AoLoader(wasmBinary, options);
  const result = await handle(...);
}

Using fetch

You can also use native fetch to retrieve the Wasm as an ArrayBuffer. This is great if you're fetching a Wasm contract published on Arweave:

import AoLoader from "@permaweb/ao-loader";

async function main() {
  const tx_id = '...'
  const wasmBinary = await fetch(`https://arweave.net/${tx_id}`)
    .then(res => res.arrayBuffer())
  const options = {
    format: "wasm32-unknown-emscripten2",
    inputEncoding: "JSON-1",
    outputEncoding: "JSON-1", 
    memoryLimit: "524288000", // in bytes
    computeLimit: 9e12.toString(),
    extensions: []
  }

  const handle = AoLoader(wasmBinary, options);
  const result = await handle(...)
}

Result Object

The Result Object returns a Successful Result:

{
  Output,
  Messages,
  Spawns,
  Assignments,
  GasUsed
}

Or an unSuccessful Result:

{
  Error
}

Contributing

Publish a new Version of the package

We use a Github workflow to build and publish new version of the Loader to NPM. To publish a new version, go to the ao Loader workflow and click the Run Workflow button. Provide the semver compatible version you would like to bump to, and then click Run Workflow. This will trigger a Workflow Dispatch that will bump the version is the manifest files, build the module, and finally publish it to NPM