The linked slidedeck holds many tips, of which I list the most interesting to me below

DATA and END

The __END__ keyword tells Ruby where a file ends – but you don't have to stop there. Any text you add after it is accessible via the DATA object.

Running the attached example file data.rb prints:

DATA is a File object to everything in the current file after "__END__"

No braces needed for 'special variables'

@instance, @@class, $global = [ 'instance', 'class', 'global' ]
puts "#@instance, #@@class, #$global"

...

The __END__ keyword tells Ruby where a file ends – but you don't have to stop there. Any text you add after it is accessible via the DATA object.

The attached example file data.rb looks like this:

puts DATA.read

__END__
DATA is a File object to everything in the current file after "__END__"

Running it with ruby data.rb prints:

DATA is a File object to everything in the current file after "__END__"

Ruby's mathn library changes Fixnum division to work with exact Rationals, so

2 / 3 => 0
2 / 3 * 3  => 0

require 'mathn'
2 / 3 => Rational(2,3)
2 / 3 * 3 => 2

While this might sometimes be quite neat, it's a nightmare if this gets required by some gem that suddenly redefines integer division across your whole project. Known culprits are the otherwise excellent distribution and [GetText](https://g...

If you get this warning on your local machine one of these steps might help:

Rebuilt the gem with the newer library

gem install --no-rdoc --no-ri nokogiri -- --with-xml2-include=/opt/local/include/libxml2 --with-xml2-lib=/opt/local/lib

If you still get the error, try to uninstall all nokogiri versions with

gem uninstall nokogiri

and install nokogiri again.

Fixing the issue on servers

However, on our servers this probably will not work. On the server, gems are stored in the ../shared/bundle/ruby/:version/gems dire...

If some of your JavaScripts fail on Internet Explorer, but only in staging or production environments, chances are that JavaScript compression is the culprit.

By default, Rails 3.2 compresses JavaScript with UglifyJS. I have seen a few cases where this actually breaks functioning JavaScript on IE (one example is the CKEditor).

I fixed this by switching to Yahoo's YUI Compressor.

To do this, do the following:

• replace the uglifier gem with the yui-compressor gem...

This describes how to migrate an existing cucumber test suite to Spreewald.

0. Add the gem

1. Include spreewald into your cucumber environment by putting
   require 'spreewald/web_steps'
   require 'spreewald/email_steps'
   # ...
   or just
   require 'spreewald/all_steps'
   into your support/env.rb.

2. Look through your step definitions for everything that might be included in Spreewald. Candidates are web_steps, shared_steps, table_steps, `em...

Cucumber_rails' old-school web-steps have been deprecated for a while, urging developers to write high-level step definitions that directly use Capybara or Webrat.

We think that's a bit drastic. More high-level steps are good, but ticking the odd check box with a general step is not always bad.

So we took the old web steps, improved them a bit, added some other favorites of ours (steps for emails, tables, [time travelling](/ma...

The ActionMailer in Rails 2 depends on a buggy version of TMail, which sometimes inserts a blank line into the mail header when sending a mail to multiple recipients. This makes the header end prematurely.

The reason why this is not exploding in your face all the time is that when you are relaying your e-mail through an MTA like Exim, it will fix this for you.

Fix for Rails if you don't have an awesome MTA

TMail is no longer maintained. The bug is fixed...

Simply use OpenStruct#to_h to receive an OpenStruct's hash representation.
In older Rubies you need OpenStruct#marshal_dump instead.

Ruby 2.0+

>> OpenStruct.new(foo: 23, bar: 42).to_h
=> { :foo => 23, :bar => 42 }

Ruby 1.9.3 and 1.8.7

>> OpenStruct.new(:foo => 23, :bar => 42).marshal_dump
=> { :foo => 23, :bar => 42 }

Both approaches will return the underlying hash table (a clone of it in Ruby 2.0+).

Ever wanted autocompletion for paths from paths.rb in Cucumber? This card lets you write your steps like this:

When I go to path *press tab now* # path is replaced with a list of all known Cucumber paths

This is how you do it

(key shortcuts apply for TextMate2)

1. Open the bundle editor (ctrl + alt +  + B)

2. Create a new Item ( + N), select "Command"

3. Paste this:
   ^
   #!/usr/bin/env ruby -wKU
   require File.join(ENV['TM_SUPPORT_PATH'], 'lib', 'ui.rb')

   cucumber_paths = File.join ENV['TM_PROJECT_DIRECTORY'], 'features'...

When you get an error like this:

Invalid gemspec in [/opt/www/foo-project.makandra.de/shared/bundle/ruby/1.8/specifications/carrierwave-0.6.2.gemspec]: Illformed requirement ["#<YAML::Syck::DefaultKey:0x7fda6f84d2e8> 1.1.4"]

... the machine's Rubygems needs to be updated.

If that happens on your local machine

• Manually remove the offending's gem files and specifications. The paths will be something like /usr/lib/ruby/gems/1.8/gems/your-broken-gem and `/usr/lib/ruby/gems/1.8/specificatio...

We frequently use the handy Paperclip Gem to manage file attachments.

If you need to move the files from local storage (i.e., your servers' harddisk) to Amazon S3, you can simply change settings for Paperclip to use the S3 storage adapter and use this script to migrate to S3. Put the snippet into a chore if you don't want to run that in the console.
YOUR_LOCAL_STORAGE_MODEL_DIRECTORY should be something like 'storage/your_model'.

Dir.glob(YOUR_LOCAL_STORAGE_MODEL_DIRECTORY**/*).each do |path|...

If you have an html_safe string, you won't be able to call gsub with a block and match reference variables like $1. They will be nil inside the block where you define replacements (as you already know).

This issue applies to both Rails 2 (with rails_xss) as well as Rails 3 applications.

Here is a fix to SafeBuffer#gsub. Note that it will only fix the $1 behavior, not give you a safe string in the end (see below).

Example

def test(input)...

In a nutshell: return statements inside blocks cause a method's return value to change. This is by design (and probably not even new to you, see below) -- but can be a problem, for example for the capture method of Rails.

Consider these methods:

def stuff
  puts 'yielding...'
  yield
  puts 'yielded.'
  true
end

We can call our stuff method with a block to yield. It works like t...

If you get errors from your development WEBrick that contain unicode salad, you are probably requesting the page via SSL. \
Since WEBrick does not speak SSL, so change the URL in your browser to use http instead of https.

The error looked like this for me:

[2012-09-06 11:19:07] ERROR bad URI `\000\026\000\020\000'.
[2012-09-06 11:19:07] ERROR bad URI `\000\026\000\020\000'.
[2012-09-06 11:19:07] ERROR bad Request-Line `\026\003\001\000�\001\000\000\177\003\001PHj�\031\006�L��'.
[2012-09-06 11:19:07] ERROR bad URI `\000\026\000\...

Tempfiles get deleted automatically

With the the ruby Tempfile class you can create temporary files. Those files only stick around as long as you have a reference to those. If no more variable points to them, the GC may finalize the object at some point and the file will be removed from the filesystem. If you would try to access your tempfile then using its path (which you stored previously), you would get an error because the file no longer exists.

Unlink your tempfiles when you're done with them
-...

RubyMine: Set specific Ruby version per project

I was experiencing the following problem:

 It seems your ruby installation is missing psych (for YAML output). 
 To eliminate this warning, please install libyaml and reinstall your ruby

So I tried to install the libyaml package via:

rvm pkg install libyaml

This is when I experienced the SSL certification problem mentioned above. This happens when your RVM certificates have expired. You can fix this by updating them via:

 rvm get stable

After doing that curl could verify the SSL certificate and I w...

For string columns, MySQL indexes the left side of a string. That means an index can speed a like query that has a wildcard on the right side:

SELECT * FROM foo WHERE field LIKE "bar%" # will be faster with an index

It can not speed up a query that has a variable left side:

SELECT * FROM foo WHERE field LIKE "%bar%" # will not be faster with an index

That also means if you use the ancestry gem you should index your ancestry column if you use scopes like descendants or `su...

In Selenium features the server and client are running in separate processes. Therefore, when mocking time with a tool like Timecop, the browser controlled by Selenium will still see the unmocked system time.

timemachine.js allows you to mock the client's time by monkey-patching into Javascript core classes. We use timemachine.js in combination with the Timecop gem to synchronize the local browser time to the ...

Testing with real live production data does come with at least one catch. All those real live users in your production environment have real live email addresses that receive real live emails.

The post includes monkey patch for ActionMailer that rewrites the domain of all recipients. It's a different take on the problem than our own mail_magnet gem.

So you got this error, even though your Gemfile bundles mysql2:

!!! Missing the mysql2 gem. Add it to your Gemfile: gem 'mysql2'

or

Please install the mysql adapter: `gem install activerecord-mysql-adapter` (mysql is not part of the bundle. Add it to Gemfile.)

The reason for this confusing error message is probably that your Gemfile says mysql2, but your database.yml still uses the mysql adapter. Change it to use the mysql2 adapter:

development:
  adapter: mysql2
  database: myproject_developm...

Our rspec_candy gem now gives you three matchers:

be_same_number_as

Tests if the given number is the "same" as the receiving number, regardless of whether you're comparing Fixnums (integers), Floats and BigDecimals:

100.should be_same_number_as(100.0)
50.4.should be_same_number_as(BigDecimal('50.4'))

Note that "same" means "same for your purposes". Internally the matcher compares normalized results of #to_s.

be_same_second_as

...