Mike's burogu

Nature Taking a Byte Out of Technology

Michael — Mon, 07 Nov 2011 09:43:30 +0000

This is the second time in 2011, when nature has reminded us of its divine power. The Fukushima Daiichi nuclear incident in Japan earlier this year, affected much of the NAND RAM supplies in the region impacting the likes of Sony and even Apple.

Right now, we learn about the flooding in Thailand and its potential effects on HDD supplies. Prices are being pushed upwards as a result.

I wonder, although sincerely hope not, if the next global crisis will be a result of multi-regional devastation in Asia. Oh, I hope not…

Rails 3.1: Starting the Week on Sunday, as in the Middle East

Michael — Fri, 21 Oct 2011 21:14:18 +0000

Was rather interesting to find out that ‘Monday’ is hardcoded into Rails as the start of the week, so working around this involved creating a custom initialiser that reopened two classes as follows:

…and here’s the proof

1
2
3
4
5
6
7
8

pry(main)> Time.now.traditional_beginning_of_week
=> 2011-10-16 00:00:00 +0300
pry(main)> Time.now.traditional_end_of_week
=> 2011-10-22 23:59:59 +0300
pry(main)> Time.now.traditional_beginning_of_week.strftime('%e %A, %B %Y')
=> "16 Sunday, October 2011"
pry(main)> Time.now.traditional_end_of_week.strftime('%e %A, %B %Y')
=> "22 Saturday, October 2011"

URI Obfuscation in Rails

Michael — Thu, 20 Oct 2011 20:06:43 +0000

Was toying with my local yarddoc server and was perusing Paperclip’s rdoc and happend across an interesting note on URI obfuscation. John has indeed taken a very interesting approach and I’m looking forward to applying this to my custom S3 to API tool.

Rails 3.1: Plupload POST to Controller Causes Devise Session Logout (CSRF Protection Bypass)

Michael — Wed, 19 Oct 2011 18:16:41 +0000

Another interesting morning thanks to CSRF protecting kicking in when an upload with Plupload would POST to a controller’s action and subsequent reloads or visits to any parts caused Devise to log me out.

This StackOverflow discussion and a blast from the past hinted in the right direction as for the issue.

Thanks to Rene (rvanlieshout) though the solution was far simpler. Rather than passing the csrf_token via the HTTP header, I passed it in the POST params hash by leveraging Plupload’s multipart_params option

1	multipart_params: {authenticity_token: false}

and passed via the BeforeUpload event where ‘up’ is the Uploader instance

1	up.settings.multipart_params.authenticity_token = '<%=j @token.to_s %>';

Rails 3.1: Bypassing Same-Origin Policy by Proxying JSON Request via Apache2 (mod_proxy)

Michael — Wed, 19 Oct 2011 17:52:05 +0000

Few days ago I started looking at tackling the challenge of finding a viable solution to large online uploads and integrating them into an existing Rails powered site. Since a 5 GB limit on each video was acceptable, I decided to leverage Amazon S3 as the storage solution and the route I ended up taking has resulted in creating a Rails app that essentially handles the upload to S3 and provides an API to those uploaded assets.

I liked the idea of using an HTML5 based upload mechanism, especially due to the clean approach taken in this article by AAAA.

However, once I had completed all the cool parts of getting the API to work, I dropped Devise in authenticate admin login — as well as leverage its :token_authenticatable module in the future for authenticated API access. For the sake of this article, the host app was hosted at www.bar.org, and the uploader app at foo.bar.org. The Same-Origin Policy for Javascript dictates failure for such access. Trying to connect raised

XMLHttpRequest cannot load http://foo.bar.org/api/v1/videos.json. Origin null is not allowed by Access-Control-Allow-Origin.

You can read more about the specifics on StackOverflow — my options were therefore limited to JSONP or proxying my JSON request through Apache itself! Rather than having to deal with a lot more JS, I decided to take the far more straightforward approach by proxying the request via /videos and all this required was and update to the virtual host config by adding

1
2
3
4
5
6
7
8

ProxyRequests Off

<Proxy *>
Order Allow,Deny
Allow from all
Proxy>

ProxyPass /videos http://foo.bar.org/api/v1/videos.json

Huge Thanks to David Workman (workmad3) and Josh Rendek (namidark) for being orsum and their superb advice!

Pushing to Multiple Git Repositories Simultaneously

Michael — Sun, 16 Oct 2011 21:20:10 +0000

Setup your .git/config file with the following entry, and add as many repeating url entries for all the repos that will be updated simultaneously:

1
2
3
4

[remote "all"]
# http://jeetworks.org/node/22
url = git@github.com:core_app
url = git@bitbucket.org:bsodmike/core_app.git

1	git push all --dry-run -v

You want to be careful if attempting git push all –all –dry-run -v as this will create new branches for all local branches not present in your remotes. Thankfully I ran a dry-run so although it seemed like I had gone and spawned a tone of branches, it was just a dry-run.

1
2
3
4
5
6
7
8
9
10

$ git push all --all --dry-run -v
Pushing to git@github.com:core_app
To git@github.com:core_app
* [new branch] datetime_massage_exp -> datetime_massage_exp
* [new branch] jquery_attempt -> jquery_attempt
Pushing to git@bitbucket.org:bsodmike/core_app.git
To git@bitbucket.org:bsodmike/core_app.git
* [new branch] datetime_massage_exp -> datetime_massage_exp
* [new branch] development -> development
* [new branch] jquery_attempt -> jquery_attempt

Rails: Create a User Profile on Registration with Devise

Michael — Fri, 30 Sep 2011 12:10:01 +0000

The trick to the above was just discussed on StackOverflow. You will have to of course override the default registration controller and flesh your own and setup the new action as follows

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

= simple_form_for(resource, :as => resource_name, :url => registration_path(resource_name)) do |f|

= devise_error_messages!

= f.input :username, :label => 'Username'
= f.input :email, :label => 'Email'
= f.input :password, :label => 'Password'
= f.input :password_confirmation, :label => 'Password confirm'

// make a new profile in memory
= resource.build_profile

// start fields for profile
= f.fields_for :profile do |f|
= f.label :bod_day
= f.text_field :bod_day
// end fields for profile

= f.button :submit, t(:submit, :scope => :register)

As with accepts_nested_attributes_for :profile in the User model, one would also include attr_accessible …, :profile_attributes.

Rails to_json or as_json?

Michael — Thu, 29 Sep 2011 18:51:07 +0000

Thanks to Jonathan Julian’s fantastic post on the subject, here’s some code that I used today in a project. I set about defining an instance method as_json in the model I was interested in, in this case Employee

1
2
3

def as_json(options={})
{ :id => self.id, :name => self.name, :title => self.title, :email => self.email, :phone => self.phone, :image => self.image }
end

and in the controller

1
2
3
4
5
6
7
8

class EmployeesController < ApplicationController
def index
@employees = Employee.all
respond_to do |format|
format.json { render :json => @employees }
end
end
end

Visiting ‘/employees.json’ yields the expected JSON

1	[{"id":1,"name":"John Doe","title":"Prof.","email":"john@foo.com","phone":"21343245252343","image":"john.jpg"}]

Update: You can also pass details from your controller to as_json as follows:

1	format.json { render :json => @employees.as_json(:fax => "23423545") }

You can then access this in the as_json method via options[:fax] and update the JSON object tree response as you wish.

Guard causing REPLs to lag… and stop working?

Michael — Thu, 29 Sep 2011 13:17:18 +0000

Becky was kind enough to send me a link to the patch on Github, the solution simply requiring an update to version 0.8.0 of Guard by adding the following your Gemfile

1	gem 'guard', :git => 'git://github.com/guard/guard.git'

This will ensure the rails console and pry to continue working with Guard…

HOWTO: Rails 3.1 Engines with RSpec

Michael — Wed, 21 Sep 2011 21:37:44 +0000

You’ll need RVM and at least Ruby 1.9.2-p290 installed to proceed. Run the following to generate the engine

1	rails plugin new epic_fail --skip-test-unit --full --mountable --dummy-path=spec/dummy

Add the following .rvmrc file to your engine’s root folder

1	rvm use 1.9.2@rails31 --create --rvmrc

This will automatically generate the files and folder structure of your engine. At the time of writing my gemspec looked like

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30

$:.push File.expand_path("../lib", __FILE__)

# Maintain your gem's version:
require "epic_fail/version"

# Describe your gem and declare its dependencies:
Gem::Specification.new do |s|
s.name = "epic_fail"
s.version = EpicFail::VERSION
s.authors = ["TODO: Your name"]
s.email = ["TODO: Your email"]
s.homepage = "TODO"
s.summary = "TODO: Summary of EpicFail."
s.description = "TODO: Description of EpicFail."

s.files = Dir["{app,config,db,lib}/**/*"] + ["MIT-LICENSE", "Rakefile", "README.rdoc"]

s.add_dependency "rails", "~> 3.1.0"
s.add_dependency "jquery-rails"

s.add_development_dependency "sqlite3"
s.add_development_dependency "thin"
s.add_development_dependency "rspec-rails"
s.add_development_dependency "capybara"
s.add_development_dependency "factory_girl_rails"
s.add_development_dependency "guard-rspec"
s.add_development_dependency "rb-fsevent"
s.add_development_dependency "yard"
s.add_development_dependency "pry"
end

The development dependencies comprises of my preferred testing stack for TDD with RSpec and BDD with RSpec and Capybara. Guard to keep a watch on the various files that may be updated (rb-fsevent is for Mac only), Factory Girl for factories, and my REPL replacement — pry.

At this stage you will want to perform

1 2	rails g rspec:install guard init rspec

Set about updating your spec_helper.rb

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

# This file is copied to spec/ when you run 'rails generate rspec:install'
ENV["RAILS_ENV"] ||= 'test'
require File.expand_path("../dummy/config/environment", __FILE__)
require 'rspec/rails'
require 'capybara/rspec'
require 'pry'

ENGINE_RAILS_ROOT = File.join(File.dirname(__FILE__), '../')

# Requires supporting ruby files with custom matchers and macros, etc,
# in spec/support/ and its subdirectories.
Dir[File.join(ENGINE_RAILS_ROOT, "spec/support/**/*.rb")].each {|f| require f }

RSpec.configure do |config|
# == Mock Framework
#
# If you prefer to use mocha, flexmock or RR, uncomment the appropriate line:
#
# config.mock_with :mocha
# config.mock_with :flexmock
# config.mock_with :rr
config.mock_with :rspec

# Remove this line if you're not using ActiveRecord or ActiveRecord fixtures
#config.fixture_path = "#{::Rails.root}/spec/fixtures"

# If you're not using ActiveRecord, or you'd prefer not to run each of your
# examples within a transaction, remove the following line or assign false
# instead of true.
config.use_transactional_fixtures = true

config.include Capybara::DSL, :example_group => { :file_path => /\bspec\/requests\// }
config.include EpicFail::Engine.routes.url_helpers
end

and added to the engine’s Rakefile

1
2
3
4
5
6
7
8
9
10

require 'rake'
require 'rspec/core'
require 'rspec/core/rake_task'

RSpec::Core::RakeTask.new(:spec) do |spec|
spec.pattern = FileList['spec/**/*_spec.rb']
end

desc 'Default: Run all specs.'
task :default => :spec

I’ve made some very specific tweaks to the above with regards to testing – these have been gleaned from the included references. I also wanted to decouple the gemspec of the dummy/app from the engine itself and to do this made some changes in boot.rb

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

require 'rubygems'
# gemfile = File.expand_path('../../../../Gemfile', __FILE__)
#
# if File.exist?(gemfile)
# ENV['BUNDLE_GEMFILE'] = gemfile
# require 'bundler'
# Bundler.setup
# end

# Set up gems listed in the Gemfile.
ENV['BUNDLE_GEMFILE'] ||= File.expand_path('../../Gemfile', __FILE__)

require 'bundler/setup' if File.exists?(ENV['BUNDLE_GEMFILE'])

$:.unshift File.expand_path('../../../../lib', __FILE__)

and the added Gemfile in spec/dummy

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17

source "http://rubygems.org"

gem "rails", "~> 3.1.0"
gem "epic_fail", :path => '../../../epic_fail'
gem "jquery-rails"

group :development, :test do
gem "sqlite3"
gem "thin"
gem "rspec-rails"
gem "capybara"
gem "factory_girl_rails"
gem "guard-rspec"
gem "rb-fsevent"
gem "yard"
gem "pry"
end

Notice how the engine it referenced via relative paths. There were several tweaks made to the Railtie itself and how routes were dynamically generated in the dummy/app.

Github Source: https://github.com/bsodmike/EpicFail

Review the commits made to my first engine, EpicFail — fork it if you like as well!

As per the latest commit, I’ve deployed some middleware to handle exception notification – as a simple test. It also includes modification to print Rails bootstrap initializer points whenever the environment is loaded.

One good thing that came out of this little adventure was the discovery of a little bug in the plugin_new generator – my approach was to incorporate the fix within the engine, however, Pastorino has applied the fix to the dummy/app’s Gemfile itself.

It’s certainly been a thrilling 48 hours and I’m completely enjoying my pursuits with Rails Engines thus far.

Mike's burogu

Nature Taking a Byte Out of Technology

Rails 3.1: Starting the Week on Sunday, as in the Middle East

URI Obfuscation in Rails

Rails 3.1: Plupload POST to Controller Causes Devise Session Logout (CSRF Protection Bypass)

Rails 3.1: Bypassing Same-Origin Policy by Proxying JSON Request via Apache2 (mod_proxy)

Pushing to Multiple Git Repositories Simultaneously

Rails: Create a User Profile on Registration with Devise

Rails to_json or as_json?

Guard causing REPLs to lag… and stop working?

HOWTO: Rails 3.1 Engines with RSpec

Github Source: https://github.com/bsodmike/EpicFail

References