Source

A Source is SoapJS's low-level data-access boundary — one per database collection/table. Adapters and repositories build on it. MongoDB's implementation is MongoSource.

In current code the interface is Source<DocumentType> and the MongoDB class is MongoSource. (Earlier versions called this component a "collection".)

Creating a Source

import { SoapMongo, MongoSource } from '@soapjs/soap-node-mongo';

const soapMongo = await SoapMongo.create(mongoConfig);
const source = new MongoSource(soapMongo, 'characters');

You rarely use a Source directly — wrap it in a DatabaseContext with a mapper and hand it to a repository:

import { DatabaseContext } from '@soapjs/soap';

const context = new DatabaseContext(source, new CharacterMapper(), soapMongo.sessions);
const repo = new CharacterRepositoryMongo(context);

The Source contract

interface Source<DocumentType> {
  find(query?): Promise<DocumentType[]>;
  count(query?): Promise<number>;
  aggregate<T>(query): Promise<T[]>;
  insert(query): Promise<DocumentType[]>;
  update(query): Promise<UpdateStats>;
  remove(query): Promise<RemoveStats>;
  native<ResultType>(query): Promise<ResultType>;   // since @soapjs/soap 0.10 / soap-node-mongo 0.7
}

• find / count accept FindParams / CountParams (built with Where) or a RepositoryQuery.

• aggregate runs a portable AggregationParams.

• native() is the escape hatch for queries the abstraction can't express — it runs a client-specific query verbatim (see Using Native Queries).

Read/write split

Give each side its own connection — read Source on a replica, write Source on the primary:

const writeSource = new MongoSource(soapMongoWrite, 'characters'); // primary
const readSource  = new MongoSource(soapMongoRead, 'characters');  // secondaryPreferred

See the demo for a full read/write split.