From: ghorner <gabriel.horner@...>
Date: 2009-07-24T03:50:06+09:00
Subject: thoughts on rewriting irb?

Hi,
  I wrote up a post on a mini irb which does better irb completion,
persistent history and error handling in 7 lines:
http://tagaholic.me/2009/07/23/mini-irb-and-mini-script-console.html

If you're curious the code looks like this:
%w{readline rubygems bond bond/completion}.each {|e| require e }
history_file = File.join(ENV["HOME"], '.mini_irb_history')
IO.readlines(history_file).each {|e| Readline::HISTORY << e.chomp } if
File.exists?(history_file)
while (input = Readline.readline('>> ', true)) != 'exit'
  begin puts "=> #{eval(input).inspect}"; rescue Exception; puts
"Error: #{$!}" end
 end
File.open(history_file, 'w') {|f| f.write Readline::HISTORY.to_a.join
("\n") }

Since that probably came out horribly, you could also just see this
snippet: http://gist.github.com/68608

At first, it may seem like a joke to compare it to irb but imho it
does the brunt of what a repl should do. All of irb's features:
workspaces, jobs, config options, command framework are features which
the majority of ruby programmers don't seem to use. Personally the
only thing I'd port to mini-irb is a ruby lexer to detect multiple
lines of ruby. So what are other people's two cents? If you could
rewrite irb, what features would you give it?