anonDB

A generic interface for data storage. Supported databases:

• SQLite
• PostgreSQL
• IndexedDB

Usage

const schema = [
  {
    name: 'User',
    primaryKey: 'id',
    rows: [
      {
        name: 'id',
        unique: true,
        type: 'String',
        default: () => nanoid(),
      },
      {
        name: 'createdAt',
        type: 'Int',
        default: () => +new Date(),
      },
      ['username', 'String', { unique: true }]
    ]
  },
]

NodeJS

const { DB, SQLiteConnector } = require('anondb/node')
const schema = require('./schema')

async function () {
  const db = await SQLiteConnector.create(schema)
  const doc = await db.create('User', {
    username: 'Chance',
  })
  const thatDoc = await db.findOne('User', {
    where: {
      username: 'Chance',
    }
  })
}

Browser

import { IndexedDBConnector } from 'anondb/web'

async function () {
  const db = await IndexedDBConnector.create(schema)
  const doc = await db.create('User', {
    username: 'Chance',
  })
  const thatDoc = await db.findOne('User', {
    where: {
      username: 'Chance',
    }
  })
}

Interface

This is the interface for a DB object. This is the primary way to interact with data storage. A DB object might be backed by an SQL database, indexedDB, or a simple in memory structure.

Common Types

Where

The where clause is used in many different queries. It allows matches based on simple property comparison or more complex operators including lt, lte, gt, gte, ne (not equal), and nin (not in array).

A where clause may contain the top level keys AND or OR. Each may contain an array of where clauses to be combined using and/or logic respectively.

A where clause using all possible comparison operators is shown below.

{
  textField: 'simple value',
  // An array compares using the IN operator
  integerField: [1, 2, 3, 4, 5],
  // The IN operator may be used on any field type
  textField2: ['this', 'that', 'the other'],
  integer2Field: {
    gt: 5, // requires that this field be greater than 5
    lt: 10, // multiple operators may be used, they are combined using AND logic
  },
  textField2: { nin: [ 'not', 'these', 'words' ] },
  // a top level key named OR allows you to specify conditions combined by OR logic
  OR: [
    {
      // This returns documents where 5 < integer2Field < 10 or integer2Field is 12
      integer2Field: 12,
    },
    {
      textField: 'another simple value'
    }
  ],
  // a top level key named AND allows you to specify conditions combined by AND logic
  AND: [
    {
      integer2Field: { gt: 10 },
    },
    {
      integer2Field: { lt: 20 },
    }
  ]
}

OrderBy

The orderBy field may be used to order the returned documents. The field is a simple object with keys corresponding to either asc or desc. Muliple keys may be specified.

orderBy: {
  integerField: 'asc',
  stringField: 'desc',
}

Include

The include field may be used to retrieve related documents. The include field is an object with field names mapped to either true or an object specifying nested documents to load.

include: {
  // Load the car model document
  model: true,
  // Load the make document, and the company document that exists on the make document
  make: {
    company: true,
  }
}

Create

Creates one or many documents.

create(collection: string, doc: any): Promise<any>

collection: The collection of documents to insert the new document(s) into. This can be thought of as a table. Collections must be specific in the schema. doc: The document (or array of documents) to be inserted into the collection.

Returns: The document (or documents) inserted.

Find One

Returns a single document matching the provided arguments.

findOne(collection: string, options: FindOneOptions): Promise<any>

• collection: The collection to search.
• options: An object with the following properties.

FindOneOptions:

• where: WhereClause
• orderBy: OrderByClause
• include: IncludeClause

Find Many

Returns many documents matching the provided arguments.

findMany(collection: string, options: FindManyOptions): Promise<any[]>

• collection: The collection to search.
• options: An object with the following properties.

FindManyOptions:

• where: WhereClause
• orderBy: OrderByClause
• include: IncludeClause
• limit: Integer representing the maximum number of documents to return.

Count

Returns the number of documents matching a where clause.

count(collection: string, where: WhereClause): Promise<number>

• collection: The collection to count.
• where: A WhereClause filtering the documents.

Update

Updates one or more documents returning the number of documents affected.

update(collection: string, options: UpdateOptions): Promise<number>

UpdateOptions:

• where: WhereClause
• update: An object with new keys and values.

Upsert

Upserts a document by creating it if it doesn't exist, updating if it does exist.

upsert(collection: string, options: UpsertOptions): Promise<number>

UpsertOptions:

• where: WhereClause to determine if the document exists.
• create: The document to be created if necessary.
• update: An object with new values to be updated if necessary.
• constraintKey (optional): The key to use when determining if a document exists. Use this if the where clause includes multiple unique fields. Underlying implementations may not be able to determine which key to use when comparing.

Delete

Delete documents matching a clause.

delete(collection: string, options: DeleteOptions): Promise<number>

DeleteOptions:

• where: WhereClause

Transaction

Batch database operations to be executed at once. If any operation in the transaction fails all operations will be rolled back.

transaction(operation: (db: TransactionDB) => void): Promise<void>

operation: A function performing database operations.

TransactionDB: A pseudo object with the following functions: create, update, upsert, delete, onCommit, onError, onComplete. Each function behaves the same as the normal version BUT the functions are not asynchronous, they return immediately but data is not persisted until the transaction is committed. onCommit allows a function to be registered as a callback when the transaction is successfully committed. onError registers a callback that is executed if the transaction errors. onComplete is executed regardless of success or error.

Once the operation function finishes executing the transaction will be applied.

Close

A function for closing and tearing down database instances. Call this when disposing of a DB reference.

close(): Promise<void>

Schema

The schema is used to define how data should be stored. Each schema is an array of collections. Each collection specifies information about the documents contained within. A collection may be thought of as a table.

Collection

A collection may specify the following keys:

• name - required - The name of the collection.
• primaryKey - optional - The primary key for the collection. This may be an array of keys. Each key must be specified in the row section.
• rows - required - An array of rows that are a part of each document in the collection.

Rows

A row may specify the following keys:

• name - required - The name of the row.
• optional - optional - A boolean indicating whether a value for this field is required.
• unique - optional - A boolean indicating whether duplicate values for this field are allowed in the collection.
• type - required - One of String, Int, Bool, Object. Objects are automatically serialized and deserialized using JSON.
• default - optional - A function or value to be used as the default value.
• relation - optional - An object specifying information for a virtual row.

A relation must include the following keys:

• localField: The name of the local field to match against the remote document.
• foreignField: The name of the remote field to match against the local document.
• foreignTable: The remote collection.