cupakromer

TL;DR

Both methods are nearly identical in perf. Differences are due more to underlying forwarded messages' implementation than how the message forwarding is defined.

Summary

Error rates between each type, for the most part, negate any real ranking. A real stat analysis would be necessary to make a real determination.

Grouped by Command style:

cupakromer

# With a clean project setup
$ bundle --binstubs --standalone

$ bundle binstubs rspec-core

$ cat bin/rspec
#!/usr/bin/env ruby
#
# This file was generated by Bundler.
#

cupakromer

Some suggested updates to the slides in https://github.com/jessitron/generatron. These suggestions are based off of the array of built-in matchers that RSpec 3 provides (https://www.relishapp.com/rspec/rspec-expectations/v/3-0/docs/built-in-matchers). As well as the composability / compounding features now available: http://myronmars.to/n/dev-blog/2014/01/new-in-rspec-3-composable-matchers

Slide 24

expect(influence.relevance).to (be <= 100).and (be >= 0)

You can also just make this a range matcher:

cupakromer

def capture_backtrace(with_error: StandardError, &action)
  action.call
  :no_error_raised
rescue with_error => e
  e.backtrace
end

# This fails if SocketError is not raised,
# or if the message doesn't match
it "raises a socket error reporting the address info" do

cupakromer

require 'rails_helper'

RSpec.describe User, type: :model do

  def build_valid_user(opts = {})
    valid_opts = {
      name: "Example User",
      email: "user@example.com",
      password: "foobar",
      password_confirmation: "foobar",

cupakromer

require 'domain/exportable_widget'
require 'support/preconditions'

RSpec.describe ExportableWidget do

  # We don't know yet if we will have multiple exportable objects, but this
  # is simply the documentation of our intended shared protocol if we do later.
  shared_examples_for "an exportable" do
    required_attribute :exportable

cupakromer

# As far as I was aware, symbol-to-proc was supposed to be roughly equivalent to:
#
#     Proc.new { |v| v.send(self) }
#
# However, the following produces very odd results. Why?!?
[[3, 4], [5, 6], [3, 4], [5, 6]].each_with_index.map(&:first)
# => [[], [5], [3, 4], [5, 6]]

[[3, 4], [5, 6], [3, 4], [5, 6]].each_with_index.map{ |x| x.first }
# => [3, 5, 3, 5]

cupakromer

class Calculator
  def reduce(operator)
    fail "You shouldn't be calling this directly!"
  end
end

def uses_a_duck_type(calculation)
  calculation.reduce(:+)
end

cupakromer

Command Line Tricks

• !! Repeat Previous Command
• !-1 Repeat Previous Command
• !-2 Repeat 2nd Previous Command
• !ps Execute Previous Command STARTING WITH ps
• !^ First Argument From Previous Command
• !:2 2nd Argument From Previous Command
• !$ Last Argument From Previous Command
• !!:$ Last Argument From Previous Command

cupakromer

gem 'jekyll', '~> 2.0.0.alpha'
gem 'nokogiri'
gem 'redcarpet'
gem 'rouge'