Delicatable treats for the rubies.

Been a long time. shouldn’t of left ya without a dope plugin to step to. meh. But for reals, I’ve been working way too much. Here’s a tasty sampler of some new and revived github projects:

• ActiveRecord Dumper
• Asset Tag Extensions
• Ruby Prof Request Filters
• Google Base4r extensions

ActiveRecord Dumper

script/plugin install http://github.com/blythedunham/ar_dumper

Dumper. I barely know her! I wrote this ages ago back in the Rails 1 dot 0h(mg) daze, but still works like a charm.

Easily export records to csv, yaml, or xml using the :find options, or (new) specify an array of activerecords with the :records option. AR Dumper dumps to strings or files with a paginated option, and it supports full customization of displayed headers and content with options such as  nly, :export, :procs, :methods

Book.dumper :yml, :find => {:conditions => ['author_name like ?', 'snowgiraffe'] }

Book.dumper :csv,   nly => [:author_name, :title], :records => @my_books

Asset Tag Extensions

script/plugin install git://github.com/blythedunham/javascript_tag_extensions

Your high performance web site client loading slowly? In addition to combining asset files, one of the best ways to improve client load speed is to move the javascript files and inline script tags to the bottom of the page. Even better, execute inline javascript after the document loads.  To load files at the bottom of the page use the

include_javascript_tag 'myjsfile', :defer => true

Similarly, replacing calls to javascript_tag with Asset Tag Extension’s inline_javascript renders the script at the bottom of the page. Use the n_load => true functionality use JQuery OnLoad functionality.

  <%  inline_javascript   n_load => true do %>
    alert("I love grapes");
  <% end %>

Ruby Prof Request Filters

script/plugin install http://github.com/blythedunham/ruby_prof_request_filters

Need to instantly profile a page? Output ruby prof results to your browser by adding the following params to the url: ruby_prof=true Flat graphs, html graph, whatever you need, this little tool comes in handy when your solving big problems.

Google Base: Base4r

I did some work on Dan Dukeson’s Google Base API For Ruby. Add, modify and delete items from Google Base (formerly Froogle) with Ruby.

Install

script/plugin install git://github.com/blythedunham/static_record_cache.git

Documentation

class SomeStaticClass < ActiveRecord::Base
  acts_as_static_record
end

#cache hits
SomeStaticClass.find_by_name('blah')
SomeStaticClass.find_by_id(5)
SomeStaticClass.find :all
assoc_class.some_static_class

One can specify the cache key for faster lookups and also customize the finder SQL for the class. More documentation can be found in the rdocs.

Another, simpler method of caching can be used by folks using the active record context plugin by technoweenie. Extending StaticRecordCache will permanently cache data for the class regardless of context. This however, only caches by id.

class TelephoneCarriers < ActiveRecord::Base
  extend StaticActiveRecordContext
end

Be sure to install active_record_cache plugin:

script/plugin install http://svn.techno-weenie.net/projects/plugins/active_record_context/

script/plugin install http://arperftoolkit.rubyforge.org/svn/trunk/conditional_page_cache

At the top of the controller, indicate which actions you wish to cache with conditionally_cache.

class MyController < Application

  conditionally_cache :action_one, :action_two

  def cache_page?(current_action_name)
    @user.demo?
  end

  def cache_action?(action_name)
    !@user.admin?
  end

end

To use page caching, add a method cache_page?(current_action_name) on the controller. The page will be cached if this method returns true.

Similarly, return true from method cache_action?(current_action_name) when you want to action cache the page. This method is the same as the one used by the action caching plugin which works great along with this one.

playing pinball in Vegas reminded me of mongrel multi caching funtime

We really like cache with our rails at Spongecell and have been really successful with Tom Fake’s Action Cache plugin when we are unable to page cache highly demanded content.

One little concurrency issue that I encountered running with the Action cache plugin on multiple mongrels was one mongrel would delete (expire) a meta data file while another was simultaneously writing the meta or data file. Anyhoo, the result was an angry corrupt file that was served to other web clients. This made some of customers angry. Then I was angry. To avoid all the hatred, I patched the method cached_entry in action_cache.rb with this little change to remove the cache on exceptions. action_cache.diff

        rescue
          remove_cache_item(controller)
          return nil
        end

With that said, here are my favorite features which you will love too!

• The timer expiry. So… you can be super lazy and just expire your cache say every 10 minutes without having to write complex sweeping code that may or may not slow down some other actions.
  @response.time_to_live = 10.minutes

• Easy integration with  Nginx X-Accel-Redirect (and Apache’s X-SendFile) to tell your web server to send the cached file downstream. This allows you to still use before and after filters, but  offloads the work the Nginx and free up your mongrels a bit to bark at someone else.

For more information on rails caching, check out Gregg from Rail Envy’s sweet post.