Install

rvm 1.9.2

• install rvm

• follow pretty straight rvm documentation

• read notes

    rvm notes
  

  look at whether additional libraries are required

• install 1.9.2

    rvm install 1.9.2
    rvm use 1.9.2 --default
  

get source

	gem install bundler
	bundle install

Data

mongo shell

resources are on mongodb site

• shell?
• query?
• upsert?

recipe : admin role

launch mongo shell, with correct database

~/src/ruby/conference-on-rails (search)$ mongo
MongoDB shell version: 1.6.4
connecting to: test
> use conference_on_rails_development
switched to db conference_on_rails_development

find users whose email begins with randy, returning [admin, email] fields

> db.users.find({email : /^randy/}, {email : 1, admin : 1});
{ "_id" : ObjectId("4cb4e0cc1c94a27e7a000005"), "admin" : false, "email" : "[email protected]" }

turn such users to admin

> db.users.update({email : /^randy/}, {$set : {"admin" : true}});

rails console

it is there, for sure!

resource

• mongoid documentation

Revoke admin role to randy

ruby-1.9.2-p0 > randy = User.where(:email => /^randy/).first
 => #<User _id: 4cb4e0cc1c94a27e7a000005, email: "[email protected]", encrypted_password: "$2a$10$OVi7LsFCgOIf8QZ/9YSWRuNFaGiPbfEI4PEvTxf9eiH7f1sD5aM7.", password_salt: "$2a$10$OVi7LsFCgOIf8QZ/9YSWRu", remember_token: nil, remember_created_at: nil, reset_password_token: nil, sign_in_count: 4, current_sign_in_at: 2010-11-24 22:22:22 UTC, last_sign_in_at: 2010-10-18 20:52:29 UTC, current_sign_in_ip: "127.0.0.1", last_sign_in_ip: "127.0.0.1", first_name: "", last_name: "", bio: "yessssssssss", avatar: nil, admin: true>
ruby-1.9.2-p0 > randy.tap {|r|r.admin=false}.save!
 => true
ruby-1.9.2-p0 > randy.admin?
 => false

Tests

One line

cuke is red with selenium, and do not want to fix this (ff6 and selenium-webdriver-2.5.0) ?

time(rspec spec && ./script/cucumber --tags ~@javascript)

or cuke is green

time(rspec spec && ./script/cucumber)

Unit

see http://github.com/rspec/rspec

flow them!

	guard

or for one shot,

	rspec spec/

dunno why it fails :(

add a debugger statement, and type then

rspec -d spec/failing_spec.rb

!!! caution : I do not recommand debugging with a running spork drb server (personal taste)!!!

this practice is not related to rails (though pretty described in http://guides.rubyonrails.org/debugging_rails_applications.html)

Javascripts

rake jasmine

then goto http://localhost:8888 ... a graphical interface for test is quite annoying ...

And at this time, jasmine-1.0.1.1 webrick server resists to STOP, HUP... but not KILL :)

Integration kungfu

see

• http://github.com/aslakhellesoy/cucumber
• http://github.com/jnicklas/capybara

cucumber has dedicated environment

optional spork

There is no automation for command line

Have spork ran for cucumber on its own port

~/src/ruby/conference-on-rails (vote)$ spork cucumber --port 12345

Then, can use following options to have cuke connect to it

./script/cucumber --drb --port 12345 features/session_rate.feature

Gotcha

**@javascript tests using default js driver (selenium) are failing when spork is up**

** requires ff5 **

run them!

for one shot

./script/cucumber feature/this-one-is-red.feature

or

./script/cucumber --tags @this_one_is_red

see cucumber and tags, and capybara tags

dunno why it fails :(

Add a handy step before offending step

Then show me the page

then you go and see

Gotchas

Metrics

coverage, using simplecov

kill any spork drb server before, otherwise, coverage is not generated (at_exit spork)

COVERAGE=true rspec spec

no support for cucumber provided coverage, at this time

where the bloody site ?

seed database at least once

rake db:seed

look out for seeds under #{Rails.root}/db/seeds/development (there is an admin there)

then

rails s