Swarm.js
This library allows you to interact with the Swarm network from JavaScript. It:
-
Communicates with the network through the HTTP API;
-
Can be used either with a local node or a gateway;
-
Solves manifests recursively;
-
Enables you to upload/download raw data and directores;
-
Enables you to upload/download from disk or from pure JS;
-
Works on the browser and on Node.js;
-
Can automatically download the Swarm binaries safely and administer the local node for you.
Live demo!
Installing
npm install swarm-js
Basic usage
The simplest use case for Swarm is uploading/downloading raw data and directories. First, load the lib:
const swarm = require("swarm-js").at("http://swarm-gateways.net");
Upload raw data
To upload raw data, just call swarm.upload(buffer)
. It returns a promise with the uploaded hash.
const file = "test file";
swarm.upload(new Buffer(file)).then(hash => {
console.log("Uploaded file. Address:", hash);
})
Download raw data
To download raw data, just call swarm.download(hash)
. It returns a promise with the data buffer.
const fileHash = "a5c10851ef054c268a2438f10a21f6efe3dc3dcdcc2ea0e6a1a7a38bf8c91e23";
swarm.download(fileHash).then(buffer => {
console.log("Downloaded file:", buffer.toString());
});
Upload a directory
To upload a directory, just call swarm.upload(directory)
, where directory is an object mapping paths to entries, those containing a mime-type and the data (a buffer).
const dir = {
"/foo.txt": {type: "text/plain", data: new Buffer("sample file")},
"/bar.txt": {type: "text/plain", data: new Buffer("another file")}
};
swarm.upload(dir).then(hash => {
console.log("Uploaded directory. Address:", hash);
});
Download a directory
To dowwnload a directory, just call swarm.download(hash)
. Swarm.js will return a directory instead of a buffer by detecting the existence of a manifest on that hash.
const dirHash = "7e980476df218c05ecfcb0a2ca73597193a34c5a9d6da84d54e295ecd8e0c641";
swarm.download(dirHash).then(dir => {
console.log("Downloaded directory:");
for (let path in dir) {
console.log("-", path, ":", dir[path].data.toString());
}
});
Download a file/directory to disk (on Node.js)
swarm.download("DAPP_HASH", "/target/dir")
.then(path => console.log(`Downloaded DApp to ${path}.`))
.catch(console.log);
Upload raw data, a file or a directory from disk (on Node.js)
swarm.upload({
path: "/path/to/thing",
kind: "directory",
defaultFile: "/index.html"})
.then(console.log)
.catch(console.log);
Upload raw data, a file or a directory from disk (on Browser)
swarm.upload({pick: "file"})
Uploading a PDF
Uploading an Ethereum DApp
When it comes to decentralized applications (DApps), the Ethereum network is responsible for the back-end logic, while Swarm is responsible for hosting and serving the front-end code. Hosting a DApp on Swarm is as simple as creating a directory with some HTMLs and a default route (the "index.html"). This, too, can be done with Swarm.js either from disk, or with pure JavaScript. Here is a sneak peek:
const swarm = require("swarm-js").at("http://swarm-gateways.net");
const indexHtml =
`<html>
<body>
<h3><img src="ethereum_icon.png"/> Swarm.js example DApp</h3>
<p><a href="foo/test_text_1.txt">Test #1</a></p>
<p><a href="foo/test_text_2.txt">Test #2</a></p>
</body>
</html>`;
(...)
const exampleDApp = {
"" : {type: "text/html", data: indexHtml},
"/index.html" : {type: "text/html", data: indexHtml},
"/ethereum_icon.png" : {type: "image/png", data: ethereumIconPng},
"/foo/test_text_1.txt" : {type: "text/plain", data: testText1},
"/foo/test_text_2.txt" : {type: "text/plain", data: testText2}
}
swarm.upload(exampleDApp)
.then(console.log)
.catch(console.log);
When you run that script, it outputs a hash. You can then use that hash to access the uploaded DApp, by either using a Swarm-enabled browser such as Mist, or through a gateway. That demo DApp is live and can be accessed:
Running a local node
Rather than using a gateway, you might wish to run your own local node. For that, you can either download/install/run it yourself, and then use require("swarm-js").at("http://localhost:8500")
, or let Swarm.js take care of it. For that, you just need to write a small config JSON with some info:
const Swarm = require("swarm-js");
const config = {
account: "d849168d52ea5c40de1b0b973cfd96873c961963",
password: "sap",
dataDir: process.env.HOME + "/Library/Ethereum/swarm-js",
ensApi: process.env.HOME + "/Library/Ethereum/geth.ipc",
binPath: process.env.HOME + "/Library/Ethereum/swarm-js/bin/swarm"
};
Swarm.local(config, swarm => new Promise((resolve, reject) => {
swarm.upload(new Buffer("test")).then(hash => {
console.log("Uploaded data. Address:", hash);
resolve();
});
})).then(() => console.log("Done!"));
If you have any trouble with this, please open an issue to let me know. That function does everything required to start a local Swarm node, including downloading binaries (if not available yet) and manging the process. It then gives you a swarm
object pointing to the local node. If the Swarm process was started by Swarm.js
, it will be closed when you call resolve()
. While it is up, you're able to access it on your browser at http://localhost:8500
.
API
TODO: document the API.