Javascript DID Resolver
This library is intended as a simple common interface for javascript applications to resolve DID documents from Decentralized Identifiers (DIDs).
This is intended to support the proposed Decentralized Identifiers spec from the W3C Credentials Community Group.
The library does not implement any specific DID method, but allows DID method implementors to release npm packages that applications can add.
Breaking changes (0.1.0)
This removes the global resolver. We've learnt that this ended up causing many problems. Instead you should now instantiate a Resolver object and use that to resolve did's within your consuming app/library.
Changes required for developers of DID resolvers
You now no longer register your DID resolvers using the global registerMethod()
function. Instead export the resolver function and show how to configure it in a local resolver object.
Configure Resolver
object
You are now required to preconfigure a resolver during instantiation. The Resolver
constructor expects a registry of methods mapped to a resolver function. For example:
{
ethr: resolve,
web: resolve
}
Each method resolver should expose a function called getResolver
which will return an object containing one of these key/value pairs. Then you can flatten them into one object to pass into the Resolver
constructor.
import { Resolver } from 'did-resolver'
import ethr from 'ethr-did-resolver'
import web from 'web-did-resolver'
import sov from 'sov-did-resolver'
ethrResolver = ethr.getResolver()
webResolver = web.getResolver()
const resolver = new Resolver({
...ethrResolver,
...webResolver,
})
const resolver = new Resolver(ethrResolver)
Resolving a DID document
The resolver presents a simple resolve()
function that returns a ES6 Promise returning the DID document.
resolver.resolve('did:uport:2nQtiQG6Cgm1GYTBaaKAgr76uY7iSexUkqX/some/path#fragment=123').then(doc => console.log)
const doc = await resolver.resolve('did:uport:2nQtiQG6Cgm1GYTBaaKAgr76uY7iSexUkqX/some/path#fragment=123')
Caching
Resolving DID Documents can be expensive. It is in most cases best to cache DID documents. Caching has to be specifically enabled using the cache
parameter
The built in cache uses a Map, but does not have an automatic TTL, so entries don't expire. This is fine in most web, mobile and serverless contexts. If you run a long running process you may want to use an existing configurable caching system.
The built in Cache can be enabled by passing in a true
value to the constructor:
const resolver = new DIDResolver({
ethr,
web
}, true)
Here is an example using js-cache
which has not been tested.
var cache = require('js-cache')
const customCache : DIDCache = (parsed, resolve) => {
if (parsed.params && parsed.params['no-cache'] === 'true') return await resolve()
const cached = cache.get(parsed.didUrl)
if (cached !== undefined) return cached
const doc = await resolve()
cache.set(parsed, doc, 60000)
return doc
}
const resolver = new DIDResolver({
ethr,
web
}, customCache)
Implementing a DID method
Each DID method will have its own methods for looking up an identifier on its respective blockchain or other decentralized storage mechanism.
To avoid misconfiguration, method implementers should export a getResolver()
function which returns an object mapping the method name to a resolve(did: string, parsed: ParsedDID, didResolver: DIDResolver)
function. e.g. { ethr: resolve }
.
The resolve function should accept a did string, and an object of type ParsedDID
export function getResolver() {
async function resolve(
did: string,
parsed: ParsedDID,
didResolver: Resolver
): Promise<DIDDocument> {
console.log(parsed)
const didDoc = ...
const parentDID = await didResolver.resolve(...)
return didDoc
}
return { myMethod: resolve }
}
The MyMethod getResolver()
result could then be passed into the DIDResolver constructor. Note that it should be flattened if used with other methods as well.
import { DIDResolver } from 'did-resolver'
import MyMethod from 'mymethod-did-resolver'
const myResolver = MyMethod.getResolver()
const resolver = new DIDResolver(myResolver)