Interact with Cabal using PouchDB.
npm install @garbados/pouchdb-cabal
Interact with Cabal using PouchDB. Comparable to cabal-core.
Cabal is a P2P chat application that uses kappa-db to store and index information. This project demonstrates a minimum viable alternate client, where a distinct database is used for storage and indexing. It's deadly to lock yourself into a single database, and this project shows that in the Cabal ecosystem, you don't have to.
Plus I like using PouchDB.
As PouchDB supports plugins, you can easily extend pouchdb-cabal to include methods pertaining to more types. For example:
``js`
PouchDB.plugin(require('@garbados/pouchdb-cabal'))
PouchDB.plugin({
publishChannelJoin: async function (channel) {
return this.publish({
type: 'channel/join',
content: { channel }
})
},
publishChannelLeave: async function (channel) {
return this.publish({
type: 'channel/leave',
content: { channel }
})
}
})
Alternatively, you can read the setupCabal and swarm methods in ./index.js, and the fromHypercore method in pouchdb-hypercore to learn how to connect to a cabal yourself, so you can plug it into your favorite database!
Use NPM:
`
$ npm i -S @garbados/pouchdb-cabal
The plugin adds and modifies the following methods to a PouchDB instance:
Update a document without providing a _rev if it differs from the version on disk. This is used on startup to ensure that the database has the latest versions of its design documents.
Given the parameters to create a multifeed, attach and ready the objects needed to interact with a cabal. Must be run before interacting with cabal.
Returns the latest nickname for the given hypercore.
Returns the current flags for a given hypercore as an object whose keys are hypercore keys and whose values are lists of flags. For example:
`js``
const result = await db.getFlags('...')
console.log(result)
>>> { '{key}': ['mod'] }
Retrieve messages in a given channel name from earliest to most recent. Results look like this:
`
{
"_id": "{key}@{seq}",
"_rev": ".-...",
".key": "{key-whomst-posted}",
".seq": {log-entry-number},
"type": "chat/text",
"content": {
"channel": "rituals",
"text": "we are gathered here today to pray cthulhu eats us first"
},
"timestamp": 123456789
}
opts is passed to db.query. You can use limit to control how many results are returned at once, and you can use startkey and endkey to paginate by timestamp.
As getChannel but sorts from most recent to the earliest.
Publish a message to the multifeed. This can be a message of any type. A timestamp is added before the message is appended to the log. The message will subsequently be saved to PouchDB. The promise resolves to the sequence number of the published message.
Publish a about message to the multifeed identifying yourself with a nickname nick. The promise resolves to the sequence number of the published message.
Publish a chat/text message to the multifeed, posting the string text to the channel channel. The promise resolves to the sequence number of the published message.
Publish a chat/topic message to the multifeed, posting the string text as the topic of channel channel. The promise resolves to the sequence number of the published message.
Join the swarm for the cabal and begin downloading blocks. opts is passed directly to hyperswarm.
Halt peering the multifeed and close open file handles.
Wraps db.destroy to also halt peering any hypercores and close any open file handles.
If you know your way around CouchDB design documents, you can check out the design documents used to index users, messages, flags, etc., in order to use db.query() to make advanced queries against this data. You can also use db.update() to add your own design documents in your own plugin.
npm test`