Rails quick tips #7: Project-specific .irbrc

In a previous installment of this series, I wrote about adding a project-specific .pryrc to Ruby/Rails projects. However, over the past couple of years, IRB gained some nice features and I don’t use Pry much anymore because it’s one less dependency I need to worry about. That doesn’t mean I want to give up on per-project configuration files though, which turned out to be slightly more complicated with IRB.

Configuration file lookup

The way IRB and Pry load their configuration is significantly different. Pry looks for $XDG_CONFIG_HOME/pry/pryrc or ~/.pryrc first and then loads the .pryrc file in the current directory if it exists (source). IRB will look for other configuration files only if ~/.irbrc doesn’t exist and it will do so in the following order (source):

IRB reads from ~/.irbrc when it’s invoked.

If ~/.irbrc doesn’t exist, irb will try to read in the following order:

• .irbrc
• irb.rc
• _irbrc
• $irbrc

This is a bit of a problem since I often work on more than one Ruby/Rails project and want to keep my general configuration in ~/.irbrc while being able to override/augment it with an additional local file.

The solution

This is what I came up with to solve this problem.

1. Add the IRBRC environment variable to the project’s .envrc. I pretty much always use direnv, so this works well for me.

   export IRBRC="./.irbrc.local"
   

2. Load the global .irbrc from the local file and add whatever else I need:

   load File.expand_path("~/.irbrc")
   
   # just an example
   def user
     @user ||= User.first
   end
   

3. Add .irbrc.local and .irbrc.local_history to either .gitignore or ~/.gitignore (I did the latter).

While I do prefer Pry’s approach overall, this is easy enough to set up and configure.