Skip to content

neume-network/core

Repository files navigation

@neume-network/core

dry run prettier unit tests

installation

We're still building neume network so things are drastically changing all the time. It may be that the main branch isn't working. But we're tagging individual git commits that end up producing good results. Check changelog.md for all available tags.

prerequsites

neume-network-core is dependent on an Ethereum full node JSON-RPC interface. Consider running your own node or choose an existing service from ethereumnodes.com.

You'll receive the best results by running Erigon and colocating the neume network crawler on the same machine as communication through a local socket is vastly more performant compared to extraction over the network.

# Clone the repository
git clone [email protected]:neume-network/core.git

# Copy the example .env file
# ⚠️ Be sure to update the variables in `.env` with the appropriate values!
cp .env-copy .env

# Install the dependencies
npm i

how to run?

The easiet way to run is using is via npm scripts.

However, there may be one issue to get started with, which is syncing an Ethereum full node and downloading all event logs manually. To turbo start you, we can offer to download a prefiltered flat file of event logs from all music NFT platforms we currently crawl called the call-block-logs-transformation file. We recommend everyone to host their own nodes of course, do it anon!

# The following line assumes your is DATA_DIR=data
curl https://raw.githubusercontent.com/neume-network/data/main/results/call-block-logs-transformation > \
  data/call-block-logs-transformation
npm run dev

It'll default to using the ./crawl_path.mjs and config.mjs file.

components

Each component of neume is published as a npm package. Below are the main components of neume.

@neume-network/core

The core is the entrypoint for neume.

@neume-network/strategies

Strategies are like React components which can extract, transform and load data. For example, we can define a strategy that will extract tokenURI from given a list NFT addresses and then transform the tokenURI to change ipfs:// to https://ipfs.io/ or ar:// to https://arweave.net/. For more information visit neume-network/strategies.

component contract

@neume-network/core guarantees the existence of variables and folders to other @neume-network packages like @neume-network/strategies. Below, we outline what core is currently guaranteeing:

existence and definition of environment variables

@neume-network/core must guarantee the existence and definition of the following environment variables:

RPC_HTTP_HOST=https://
DATA_DIR=data
EXTRACTION_WORKER_CONCURRENCY=12
IPFS_HTTPS_GATEWAY=https://
ARWEAVE_HTTPS_GATEWAY=https://
  • If RPC_HTTP_HOST requires Bearer-token authorization, users must define RPC_API_KEY to be used in an HTTP Authorization: Bearer ${RPC_API_KEY} header.
  • If IPFS_HTTPS_GATEWAY requires Bearer-token authorization, users must define IPFS_HTTPS_GATEWAY_KEY to be used in an HTTP Authorization: Bearer ${IPFS_HTTPS_GATEWAY_KEY} header.

managing DATA_DIR and the file system directory

A directory containing the outputs of all @neume-network/strategies is called the DATA_DIR. Apart from guaranteeing the environment variable's existence, @neume-network/core must guarantee the directory's existence on the user's file system.

JavaScript Usage

neume can be imported as a JavaScript utility to run strategies.

import { boot } from "@neume-network/core";

const crawlPath = [[{ name: "get-xkcd", extractor: {} }]];
const config = {
  queue: {
    options: {
      concurrent: 1,
    },
  },
};
(async () => {
  await boot(crawlPath, config);
})();

CLI

neume has a CLI which can be used to run strategies.

CLI options

Usage: neume.mjs <options>

Options:
  --help     Show help                                                 [boolean]
  --version  Show version number                                       [boolean]
  --path     Sequence of strategies that the crawler will follow.     [required]
  --config   Configuration for neume CLI
example
./neume.mjs --path crawl_path.mjs --config config.mjs
  • crawl_path.mjs: A crawl path that runs all strategies supported by neume to crawl music NFTs.
  • config.mjs: A configuration which should be modified according to your environment. Configuration should follow the schema given in @neume-network/schema.

Crawl Path

It is the path taken by the neume crawler. It is defined as a .mjs file and should default export the desired path. The path must follow the schema given in @neume-network/schema.

Example:

An understanding of neume strategies is required to understand the example

The below crawl path will first run web3subgraph strategy and then soundxyz-call-tokenuri and zora-call-tokenuri. soundxyz-call-tokenuri and zora-call-tokenuri will run after web3subgraph has completed and they both will run in parallel.

extractor.args are inputs for extractor of the corresponding strategy. For example, 0xabefbc9fd2f806065b4f3c237d4b59d9a97bcac7/9956 will be an input to the extractor of web3subgraph.

Similary, transformer.args.slice(1) are inputs for transformer of the corresponding strategy and transformer.args[0] is the file to be transformed.

Both extractor.args and transformer.args are optional.

export default [
  [
    {
      name: "web3subgraph",
      extractor: { args: ["0xabefbc9fd2f806065b4f3c237d4b59d9a97bcac7/9956"] },
      transformer: {
        args: ["path/to/file", "some-argument"],
      },
    },
  ],
  [
    {
      name: "soundxyz-call-tokenuri",
      extractor: {
        args: ["path/to/file"],
      },
      transformer: {},
    },
    {
      name: "zora-call-tokenuri",
      extractor: {
        args: ["path/to/file"],
      },
      transformer: {},
    },
  ],
];

contributing

See contributing.md

license

Licensed as SPDX-License-Identifier: GPL-3.0-only