Skip to content
Switch branches/tags

Name already in use

A tag already exists with the provided branch name. Many Git commands accept both tag and branch names, so creating this branch may cause unexpected behavior. Are you sure you want to create this branch?
Go to file
Cannot retrieve contributors at this time

How to Contribute to Bookshelf.js

  • Before sending a pull request for a feature or bug fix, be sure to have tests.
  • Use the same coding style as the rest of the codebase. There is a pre-commit hook that formats code automatically for you using prettier, so this should be taken care of already.
  • Fill in the issue or pull request templates when creating your entry. This will help clarify the scope of your proposal or the subject of your issue.
  • All pull requests should be made to the master branch.

Development Environment Setup

You'll need to have git installed obviously. Begin by forking the main repository and then getting the code from your forked copy:

git clone

Afterwards go to the bookshelf directory that was just created and install the dependencies:

npm install

At this point the only thing missing are the databases that will be used for running some of the tests of the automated test suite.

There are three options for setting up this part:

  • The first one is to use docker containers for the database servers. This is explained below.
  • Alternatively you can provide configuration options of the database servers and set them up manually. This is also explained further down.
  • Last option is to use a config file in case you already have your servers configured and don't want to change any of their configurations.

Using Docker Containers

You can install Docker and docker-compose easily on any Operating System. After picking the correct operating system installer (via package manager or download) just run the following command on the root of your cloned Bookshelf repository:

# This starts the test databases
docker-compose up -d

You can also teardown the databases with:

docker-compose down --remove-orphans

When using Docker you may run the tests many times against the same DB instances. The tests reset DB state on each run.

Note: admin privileges are needed, so you should act according to your chosen Operating System.

Manual Database Servers Setup

The two sections below deal with manually setting up the database servers needed for running tests. If you prefer an easier alternative you can use Docker containers instead, as explained in the previous section.


You can install MySQL easily on most linux distros by using their package manager. With Ubuntu this should do it:

sudo apt-get install mysql-server mysql-client

On OSX you can download a disk image directly from the MySQL Downloads page, or use one of the popular package managers like homebrew or MacPorts.

To run the test suite you will need to make sure it is possible to connect as the user root without the need for a password.

It is strongly recommended that you use the command line mysql client to access your MySQL instance since there can be problems connecting as the root user with some graphical clients like phpMyAdmin. To check if you can connect as root without needing a password use the following command:

mysql -u root

If you see an error like:

ERROR 1045 (28000): Access denied for user 'root'@'localhost' (using password: NO)

that means you can't login as root without a password. If you do know the root user's password just login with the known password like this:

mysql -u root -p

and enter the password when asked. Then just set an empty password for root like so:

USE mysql;
UPDATE user SET authentication_string = "" WHERE User = "root";

If you see an error like:

ERROR 1054 (42S22): Unknown column 'authentication_string' in 'field list'

It's because you are using an MySQL version older than 5.7. The authentication_string column has replaced the password column in this newer version. If you are using an older version just use password instead of authentication_string in the example above.

Note that you'll probably need to set the password to NULL instead of an empty string in MySQL versions 5.5 and older. The above example should work with versions 5.7 and newer.

If you have forgotten the root password you'll need to take some extra steps to reset it. Take a look at this Stack Overflow answer for further details.

If you still can't access the MySQL server without a root password after all this trouble then this answer on Ask Ubuntu has another alternative approach to the problem.


You can install PostgreSQL easily on most linux distros by using their package manager. With Ubuntu this should do it:

sudo apt-get install postgresql postgresql-client

On OSX the easiest way is probably by using PosgresApp. It should also be available to install via homebrew or MacPorts if you prefer.

In the case of PostgreSQL the requirement is to be able to connect as the postgres user on localhost also without the need for a password. This can be achieved by editing or adding the following line in the pg_hba.conf file:

host    all             all               trust

This file can be found in /etc/postgresql/9.4/main/ on most linux systems. The 9.4 part could be different depending on the version that is available in your distro. On OSX the location of this file will depend on the installation method chosen, but for the recommended PostgresApp install it will be in /Users/[yourusername]/Library/Application Support/Postgres/var-9.3/. Again, the var-9.3 part may be different depending on the version you installed.

The trust in the example above tells the locally running PostgreSQL server to ignore user passwords and always grant access on clients connecting locally. Do not use this setting in a production environment.

After editing the pg_hba.conf file you'll need to restart the PostgreSQL server for the changes to take effect.

Using a Configuration File

If you don't want to go to the trouble of performing the changes explained in the previous two sections you can instead use a config file that tells the test suite about your current database setup.

The tests will look for a BOOKSHELF_TEST environment variable that points to a config.js file with the connection details for each database server. This file must not be the same database config file you use for any other application, otherwise you risk data loss in that application.

Example config file:

module.exports = {
  mysql: {
    database: 'bookshelf_test',
    user: 'root',
    encoding: 'utf8'

  postgres: {
    user: 'myusername',
    database: 'bookshelf_test',
    password: 'secretpassword',
    host: 'localhost',
    port: 5432,
    charset: 'utf8',
    ssl: false

  sqlite3: {
    filename: ':memory:'

This file can be placed anywhere on your system and can have any name that you like, as long as the environment variable is pointing correctly to it. For convenience you can put it in your home directory and add the following line to your .bashrc or .zshrc:

export BOOKSHELF_TEST='/home/myusername/.bookshelf_config.js'

Database Creation

After having ensured the test suite can access both database servers just create a new database on each that will be used exclusively by Bookshelf.js:

CREATE DATABASE bookshelf_test;

Note that this step isn't necessary when using Docker.

Running the Tests

The easiest way to run the tests is by using docker & docker-compose as explained above.

The test suite requires that both MySQL and PostgreSQL servers have a database named bookshelf_test. See the sections above for further instructions.

Once you have your development environment properly setup, you can run the tests:

npm test

Always make sure all the tests are passing before sending a pull request.

Publishing a New Release

The release process is fairly well automated. You do need publishing rights to the bookshelf package and write access to the GitHub repository. Once that's set just make sure to follow the process explained below in the correct order.

  1. You should draft a new release on GitHub. This isn't strictly necessary, but is highly recommended. At this time you shouldn't publish it yet, but just save it as a draft instead.
  2. Make sure you're in the master branch with all the latest changes: git pull.
  3. Update the file and update the version number of package.json. For the changelog just follow the format of the previous update. In general you should link to PRs instead of issues when mentioning changes. If the PRs' descriptions are well written they should already include any associated issues. At this point there is no need to commit and/or push these changes since that is taken care of automatically by the release scripts.
  4. Just run npm publish and sit back.