Recent Posts - Page 26

Apache, Typo, and Feedburner

September 08, 2006

A few weeks ago, I started using FeedBurner and posted a blog entry about how I configured Lighttpd to handle that so it didn’t disrupt everyone that subscribes to my RSS feed. This was working great… until the other day.. when I upgraded to Typo trunk. When I did this, I decided to start using mongrel::cluster and pound. Two days later… I’m noticing that my subscriber count has dropped over thousand people in a day… was it something that I said?

A ha! I was handling the redirect with Lighttpd and had replaced it with Pound.

So, I am now delegating this to Apache.

# Redirect typo feeds to FeedBurner
RewriteCond %{HTTP_USER_AGENT} !^FeedBurner.*$
RewriteRule /xml/(atom|rss|rss20)/feed.xml$ http://feeds.feedburner.com/RobbyOnRails [R=temp,L]

…and all was well again.

UPDATE

The rewrite condition should be !^FeedBurner.*$ not !^FeedBurner$.

Read More

Injecting Rails into your Legacy World

September 06, 2006

Many of you know that I have an unhealthy obsession with databases… in particular legacy databases. I’m not the only one it seems. :-)

I’m hoping to compile a few stories of how Rails enthusiasts have been able to take advantage of the Ruby on Rails framework within a Legacy environment, whether that be to talk with legacy databases, web services, or a piece of cauliflower. If you have a story (good or bad) of how you used Rails with a legacy system, please email me and let’s talk. :-)

Also, I would like to announce that I have started a new google group that focuses on the topic of using Rails with legacy environments. Check out the new Rails and Legacy group.

If you’re down in California, you can see my talk, “Rails meets the Legacy World” at the Ruby on Rails Seminar, which is being hosted at the AJAXWorld Conference & Expo.

Read More

Typo upgraded to trunk

September 05, 2006

I upgraded my blog to Typo trunk this evening. I hadn’t updated since late last year… so it was long overdue. I only had one issue with a migration from a while back.

old revision: 761

current revision: 1258.

I’m impressed that the upgrade worked so nicely… last year the migrations were a bit rough. :-)

UPDATE

I’m also now running this on mongrel with pound! I followed the configuration here.

Read More

PL/Ruby loves RubyGems and DRb

August 22, 2005

I admit it. I have had a torrid love affair with procedural languages ever since I started playing with PostgreSQL. The ability to share logic amongst all the applications touching the same database server.. was…well… a breath of fresh air.

What is a procedural language in Postgresql?

PostgreSQL docs describe them as, “…allows user-defined functions to be written in other languages besides SQL and C. “

Well, PostgreSQL has PLs for Perl, Python, Java, C, PHP… and even RUBY!

CREATE FUNCTION ruby_max(int4, int4) RETURNS int4 AS
' if args[0].to_i > args[1].to_i return args[0] else return
args[1] end ' LANGUAGE 'plruby';

PL/PGSQL is nice and all, but it’s not as fun as playing with Ruby. PL/Perl… well is perl, and PL/Python… is python. Both PL/Perl and PL/Python have untrusted variants. You see, they don’t want your PostgreSQL server to do anything harmful to the machine by being able to do stuff like system(‘cat /dev/null > /etc/passwd). But for some people, (like me) they want the flexibility of their language anyways. :-)

Note: Never do this if your system user that runs PostgreSQL has privileges to do anything harmful on your system.

The PL/Ruby documentation is minimal at the moment, but covers enough to get you started. I don’t know if many people are using it out there… but hopefully that is about to change! I’ve played with it a bit, but always wanted to be able to do stuff like require ‘rubygems’, but this is a feature of an untrusted language. I even found myself digging around in C code to see if I could figure out how to hack the plruby language to skip over those checks… but I am not a C programmer and I got lost in some header files.

Then it hit me. “Why haven’t you emailed the author?”

So I emailed the author of PL/Ruby, Guy Decoux, who responded pretty quickly with the answer to my dreams! Okay, I do have bigger dreams than this… but you get the idea.

First of all, some of you might be thinking, “Why on Earth would you want to do this?”

Well, here is a simple example of how it could be used with RedCloth Let’s say that I want to be able to perform the following query from within SQL.

SELECT redcloth('**strong text** and *emphasized text*');

Why not do this in the application? Well, I do actually have a case where I have an older PHP application that I will be porting to Ruby in the future, but would like to give the application some access to some of the features of Ruby that I will be using, such as RedCloth.

Okay, so show me an example of one of these scary PostgreSQL functions.

CREATE FUNCTION redcloth(text) RETURNS text AS '

  require "rubygems"
  require "redcloth"

  content = args[0]

  rc = RedCloth.new(content)

  return rc.to_html

' LANGUAGE 'plruby';

“Wait! You said this would be scary!?”

Well, PL/Ruby allows you to write… plain ole Ruby within your functions. (do you see where I am getting here?)

PL/Ruby meets RedCloth

rb=# SELECT redcloth('*strong text* and _emphasized text_');
                         redcloth
------------------------------------------------------------------
 <p><strong>strong text</strong> and <em>emphasized text</em></p>
(1 row)

PL/Ruby meets ShortURL

CREATE FUNCTION rubyurlize(text) RETURNS text AS '

  require "rubygems"
  require "shorturl"

  return ShortURL.shorten(args[0])

' LANGUAGE 'plruby';

…which allows for

rb=# SELECT
rb-#   rubyurlize('http://www.robbyonrails.com/') as link1,
rb-#   rubyurlize('http://moulon.inra.fr/ruby/plruby.html') as link2;

      link1           |         link2
--------------------------+------------------------
 http://rubyurl.com/lyoKm | http://rubyurl.com/dTo
(1 row)

PostgreSQL meets DRb

Okay, this is one of the reasons why I wanted to play with PL/Ruby a bit more. Distributed Ruby Objects… from PostreSQL?

What is DRb?

If you don’t know already… per the description in RDOC, “dRuby is a distributed object system for Ruby. It allows an object in one Ruby process to invoke methods on an object in another Ruby process on the same or a different machine.”

It basically allows you to share an object to other machines… at the same time!

mmm…distributed objects…

DRb Object

Here is a simple ruby script that you would run from the shell. It creates a DRb object which accepts connections at localhost:9000.

#!/usr/bin/ruby

require 'drb'

class MyRemoteObject
  def say(str)
    return "You say #{str}. I say #{str.reverse.upcase}!"
  end
end

server = MyRemoteObject.new

DRb.start_service('druby://localhost:9000', server)
DRb.thread.join

Start me up!

$ ruby mydrb.rb

Now that we have DRb running and listening for connections…we need a client to connect to it.

DRb function in PL/Ruby

Here is a very simple DRb client script and I just drop that into a PostgreSQL function.

CREATE FUNCTION drb_test(text) RETURNS text AS '

  require "drb"

  DRb.start_service

  ro = DRbObject.new(nil, "druby://localhost:9000")

  return ro.say(args[0])

' LANGUAGE 'plruby';

The result?

rb=# SELECT drb_test('Potato');
       drb_test
-------------------------------
 You say Potato. I say OTATOP!
(1 row)

Are we having fun yet?

Okay, so how do I manage to get this to work? Well… for that, you will have to read my blog post, Installing untrusted PL/Ruby for PostgreSQL

Let’s all go get some coffee (or tea) and start playing with PL/Ruby today!

Read More