bt-fetch

bt-fetch

Interact with Bittorrent the same way you would websites via fetch() and bittorrent:// or bt:// URLs

How it works

• Uses WebTorrent to load torrents
• bittorrent:// URLs put the infohash of a torrent into the hostname portion
• bittorrent:// URLs can contain an individual file to prioritize downloading
• bittorrent:// URLs can point to a folder path to get a listing

API

import makeBTFetch from 'bt-fetch'

const fetch = makeBTFetch({
  // Use this if you want things to be downloaded somewhere specific
  folder: '~/.local/data/bt-fetch/'
  // You can also pass any arguments from `new WebTorrent` and `client.add`
})

// Get a file as text
const res = fetch('bittorrent://08ada5a7a6183aae1e09d831df6748d566095a10/example.html')

const content = await res.text()

// How big is the file in bytes (useful in HEAD requests)
const length = res.headers.get('Content-Length')
// How much of the file has been downloaded in bytes
const downloaded = res.headers.get('X-Downloaded')

// List files / subfolders as JSON
const res = await fetch('bittorrent://infohash/path/')

const files = await res.json()

// List the contents as a web page
const res = await fetch('bittorrent://infohash/path/', {headers: {
  Accept: 'text/html'
})

const page = await res.text()

// now also supports public keys bittorrent://64CharacterPublicKeyHere
const pubRes = fetch('bittorrent://1e267e045c1abcb9af26df782a048a1cfd2d26e6db23ff5026b213ce037301bf')

TODO

• Pass querystring params from magnet link info (trackers, etc)
• Support index.html resolution (with opt-out)
• Creating torrents using POST
• Mutable torrent record exchange (updates without DHT polling)
• Support some sort of DNS based on DNSLink
• Extension messages with EventSource a-la hypercore-fetch