simon-p-r/level-mongo

mongo interface to leveldb

level-mongo

A basic mongo query interface for leveldb backend, future version will try to allow more advanced querying and partial & full updates

Install

$ npm install level-mongo --save

Usage

Valid options for constructor object exposed by module are shown below

options:

• collections - object with keys being name of collection (sublevel-level) to create
  • key - unique identifier for each json record within collection
• location - string required to define base directory where leveldb is persisted on disk
• config - object with the following keys
  • keyEncoding - string, valid options are one of 'hex', 'utf8', 'ascii', 'binary', 'base64', 'ucs2', 'utf16le'
  • createIfMissing - boolean, default true. Create db if missing
  • errorIfExists - boolean, default false. Error if db already exists
  • compression - boolean, default true. Set to true to use snappy compression end for leveldb files
  • cacheSize - number, default 8 x 1024 x 1024. LRU cache for leveldb

'use strict';

const LevelMongo = require('level-mongo');

const db = new LevelMongo({
   config: {
       keyEncoding: 'ascii',
       compression: false,
       cacheSize: 4 * 1024 * 1024
   },
   collections: {
       users: {
           key: '_id'
       }
   },
   location: './db'
});

db.open((err) => {

   db.collections.users.findOne({ _id: 'abcd'}, (err, doc) => {

       console.log(err, doc);

       db.close((err) => {

           console.log(err);

       });
   });

});

Mongo methods partially implemented on each collection object

• findOne - find record by key

• find - return all records within collection

• insertOne - insert record, unique key must be embedded within object being inserted

• insertMany - batch insert operation

• deleteOne - find and delete one record by key

• deleteMany - deletes all records within collection

• updateOne - updates one record by merging modifier object into original document

• count - return number of records in collection

• getKeys - returns all keys stored within collection

TODO

• improve indexing - in memory only? B-Tree?
• add schema parameter and validation option to define data shape
• findOne
• findMany
• insertOne
• insertMany
• updateOne
• updatedMany
• count
• allow querying by document properties as opposed to just be id
• abstract shared logic better
• add update operators
  • $set - partially implemented
  • $unset - partially implemented
  • $inc
  • $mul
  • $rename
  • $min
  • $max
  • $currentDate
• add array operators
• add find projection
  • Return fields
  • Exclude fields