mongoose-harmony-gridfs

mongoose-harmony-gridfs

Mongoose plugin for GridFS based on ES6 Harmony

Purpose

Store large documents in a MongoDB database using Mongoose with a painless syntax. This project should only be used with generator heavy projects, such as Koa projects.

Installation

You can download this plug-in through NPM.

npm install mongoose-harmony-gridfs 

Usage

var mongoose = require('mongoose'),
    plugin = require('mongoose-harmony-gridfs')

var schema = mongoose.Schema({
    name: String,
    joinedOn: Date
});

var settings = {
    keys: ['photo', 'thumbnail']
};

ProfileSchema.plugin(plugin, settings); // Attach GridFS
var Profile = module.exports = mongoose.model('Profile', schema); // Ready to use GridFS

Initial setup

Schema setup

You must attach this plug-in to your schema in order to begin using GridFS. This is necessary because the plug-in needs to know what keys you will be using as document storage. Once the plug-in has been attached, all new instances will contain GridFS functionality.

var user = new Profile() // Has GridFS functionality

Please be wary of key usage. Your original schema should not contain the same keys as the ones included in the plug-in.

Settings

• key: Array of strings that determine what keys will be used by the plug-in. Without this option, the plug-in will do nothing.
• bucket: String that determines what GridFS repository should be used. By default it is 'fs'.
• linkPropertyName: String that determines what the property that is responsible for linking files will be named. By default it is '_gfs'.

CRUD operations

Saving

Once you've created a new instance, you can start saving buffers right away.

var user = new Profile()
var data = "This is a random string!"
user.photo = new Buffer(data) // This sends document information immediately to the database

Information is sent to the database immediately. Any existing information that was previously stored in the database is removed and replaced with the new buffer.

Loading

If you can yield values because you are calling from a generator, you can yield for a database value. This will return a buffer with all contents in the file.

var photo = yield user.photo;

If you need to provide error handling, you can wrap this statement in a try/catch.

var photo;
try {
    photo = yield user.photo;
} catch (err){
    // Do something with the error
}

If you are operating within a function, you can use Q's done function to retrieve the information in a callback. Similarly, you can use the fail function to catch errors.

var promise = user.photo;
var photo;
promise.done(function(data){
    photo = data;
}).fail(function(err){
    throw err;
});

Future development

There are plans to extend this plug-in to include the ability to read streams of files and to be able to gather and store metadata. If this is something that interests you, feel free to fork and open a pull request. I will also need tests and those should be coming soon.