@arso-project/sonar-client

sonar-client

A JavaScript client for Sonar.

The client talks to sonar-server over HTTP. Authentication is not yet implemented, so currently the expectation is that a server is running on localhost.

Usage

npm install @arso-project/sonar-client

const Client = require('@arso-project/sonar-client')
const client = new Client()

API

const client = new SonarClient(opts)

Create a new SonarClient. opts is an object with optional keys:

• endpoint: The API endpoint to talk to. Default is http://localhost:9191/api
• island: The island to talk to. Default is default.
• cache: Cache records locally in memory. Default is true

Islands

await client.info()

Get a list of all islands available on this server.

await client.createIsland(name, [opts])

Creates a island with name name on the Sonar server. The name may not contain whitespaces. opts is an optional object with:

• key: Hex string of an existing island. Will then sync this island instead of creating a new, empty island.
• alias: When setting key, alias is required and is your nick name within this island.

await client.updateIsland(config, [key])

Updates the config of the current island or the island with the optional parameter key.

Currently only the config option share (boolean) is supported that controls if an island is shared via p2p.

Schemas

await client.getSchemas()

Get all schemas in the current island.

await client.getSchema(schemaName)

Get schema schemaName in this island. Throws if the schema does not exist.

await client.putSchema(schemaName, schema)

Add an new schema to the island. schemaName is a string, schema a JSON schema object.

Sources

client.putSource(key, info)

Adds a source new source with key key to the island. key should be the key of the local writer of a clone of this island.

Database

await client.get({ schema, id }, [opts])

Get records by schema and id. This returns an array of matching records.

await client.put(record)

Put a new record into the database. record is an object that looks like this:

{
  schema: "string",
  id: "string",
  value: object
}

If id is empty, a new id will be generated. schema and value are required.

await client.query(name, args, [opts])

Query the database. Returns an array of matching records. Records may have a meta property that includes query-specific metadata (e.g. the score for search queries).

• name: The name of a supported query. Options at the moment are search, records, history and indexes
• args: The arguments for the query. Depends on the query being used.
  • For records: { schema, name, id }
  • For history: { from: timestamp, to: timestamp }
  • For search: Either a "string" for a simple full-text search, or an tantivy query object (to be documented)
  • For indexes: { schema, prop, value, from, to, reverse, limit } (to be documented)

await client.sync([views])

Wait until all running operations are finished. This returns once all currently running indexing batches are finished. If you did any put or delete operations before this function will return once those are commited to all indexes. In case of pending incoming messages from remote feeds, the current indexing run is awaited.

• views: Optional. String of a view name or array of view names to wait for being finished. View names are the same as the query names above.

Files

await client.getDrives()

Get a list of Hyperdrives that are part of this island.

await client.writeFile(path, file, [opts])

Write a file.

• path: The path to the file. The first path segment must be the key or alias of a hyperdrive registered in this island.
• file: Either a readable stream or a buffer.

await client.readFile(path, [opts])

Read a file.

• path: The path to the file. The first path segment must be the key or alias of a hyperdrive registered in this island.

Returns a readable stream to the file. If opts.stream is false returns a Buffer.

await client.statFile(path)

Get the stat info for a path.

await client.readDir(path)

Get the directory listing for a path.

Resources

Resources are files with attached metadata.

await client.createResource(value, [opts])

Create a new resource file.

value is an object with a required filename and optional prefix. This will create a new resource for a file in the default writable hyperdrive at path prefix/filename.

Options are:

• force: Create resource even if a file for the path already exists (default false).
• update: Update resource if a resource for the file already exists.

Returns the resource record object.

await client.writeResourceFile(record, file, [opts])

Write the file attached to a resource.

• record is a resource record as returned from createResourceFile or as returned from a query.

• file is a readable stream or a buffer.

await client.readResourceFile(record, [opts])

Read the file attached to a resource.

• record is a resource record as returned from createResourceFile or as returned from a query.

• opts are the same as readFile

Subscriptions

await client.pullSubscription(name, [opts])

Pull the subscription name. Returns an object with a batch of messages:

{
  messages: [], // Array of records
  finished: bool, // true if no more records are available
  cursor: number // the lseq of the last record returned
}

await client.ackSubscription(name, lseq)

Set the subscription cursor to lseq. Note that you cannot acknowledge individual messages, but set the cursor (position in local total log of an island). The lseq is available on each record. Usually, you want to process messages linearily as returned from pullSubscription and set the cursor after each processed message.

Commands

TODO.