node package manager
Share your code. npm Orgs help your team discover, share, and reuse code. Create a free org ยป



I've been using this in my own production-level applications and it's worked quite wonderfully.

This takes care of revisions and annoyances in the background to help speed you up. But it doesn't abstract things too much. You're really just throwing around JS objects. Make sure you standardize your database code and your interfaces to make sure you aren't onion-skinning your data, nor losing records.

This works nicely with local CouchDB instances or a wonderful service called Cloudant that I love very much.

Getting Started

npm install couchdb-simple



For localhostin' your CouchDB server:

var couchdb = require("couchdb-simple");
var database = new couchdb();  

For Cloudant, and other hosted solutions:

var couchdb = require("couchdb-simple");

Reading from the database

You can read a list of entries:'/dir/', function( results, error ) {

    // Iterate on results.rows


or read a single entry:'/dir/entryid', function( results, error ) {

    // Results are accessible in results object


Writing to the database

Writing is easy, since you don't have to deal with revision information. At the same time, be careful not to wreck your existing data accidentally! Right now, I don't recommend writing without an entry ID, I think it's easier to handle those synchronously by declaring them before you write to the database.

var datastore = {};

database.write('/dir/entryid', datastore, function( results, error ) {


Removing from the database

Just like writing, revision data is handled behind the scenes so you can delete an entry in a single step. I should have named the method delete but I pushed this live too late :trollface:

database.remove('/dir/entryid', function( results, error ) {


Error Handling

There is an error object returned with every request that you can use to determine the success of the request:

if ( typeof error !== "undefined" && error ) {

    // Something's amiss about database request, handle it.

} else {

    // All's well!


This avoids try-catch patterns which really make cringe. I don't like things that supposed to run for a very long time to just break when something happens, so I find this method fairly graceful.

I use Cloudant for my Couchdb work, and it never really breaks so these are all over my code, but don't run that often. If interest is there, I can write in a HEAD method to check database availability.