Wednesday, April 9, 2008

Getting postgresql and git to play nice on Mac OS X

If you installed postgresql83 using macports, you may run into some trouble building the apr dependency.

In file included from misc/unix/rand.c:39:
/opt/local/include/uuid.h:94: error: conflicting types for 'uuid_t'
/usr/include/unistd.h:133: error: previous declaration of 'uuid_t' was here
uuid_t is defined in the port of apr(git dependency) and ossp-uuid(postgresql83 dependency). Simply uninstalling postgresql83 and ossp-uuid does not solve the problem. Apparently macports is happy to let one port overwrite files from another without a huge red flag. The commands below should do everything you need.

port uninstall postgresql83-server postgresql83 ossp-uuid
port clean apr
port install git-core +svn
port install postgresql83-server

Tuesday, December 18, 2007

Saturday, September 15, 2007

Easily display svn revision

In one of my projects we'll occasionally have production servers that don't get restarted on a deploy and this is a nightmare when it comes to tracking down bugs. Following the easy svn info parsing with YAML, we create a constant that holds our revision that gets set only when the server is starting.

Put this in your environment.rb:

Revision = YAML::load(`svn info $RAILS_ROOT`)["Revision"]

We can then easily display it along with other debug information to certain user accounts.

Note: since this is set in environment.rb it will only be set once at application startup. This is great for performance but will show stale revision information in development mode. This should be a non-issue for everyone since this exists solely for debugging production server failed restarts.

Sunday, August 19, 2007

Less is more with redgreen

When running tests, it's normal to put the most basic functionality tests at the begining, but when you run a test with a lot of errors you'll end up with more complex end tests staring you in the face. I used to try to solve what was in front of me, so I would start at the end. This is obviously harder and basically a waste of time, so my next move was to run the test case and pipe it through less.
ruby test.rb | less
That worked perfectly fine until I installed redgreen to highlight my successes and failures. Now less ends up escaping all of the color codes and I'm left with a messy screen of numbers and reversed colors.
ruby test.rb | less -R
Shouldn't less default to displaying color codes? Alias that!

Thursday, June 21, 2007

"duck typing made easy".gsub(/y$/, 'ier')

I was reading through the ever amazing Beast Forum code and came across this bit in distance_of_time_in_words method of application helper.

from_time = from_time.to_time if from_time.respond_to?(:to_time)

Duck typing is great, and you should use it when possible, but look at how ugly it makes things. You have to type to_time twice, once as a method and once as a symbol. The object has to be typed twice as well, maybe if I had some fancy Textmate snippet or something to help me. Or some slightly clever Ruby code.

def method_missing(method, *args)
  return super unless method.to_s =~ /_if_respond_to$/
  method_name = method.to_s.gsub(/_if_respond_to$/, '')
  return self unless self.respond_to? method_name
  return self.send(method_name)

Now we can easily do:

# => Thu Jun 20 17:25:13 -0400 2007

# => "123"

# What if it doesn't respond_to? your method?
# => 123

Friday, June 15, 2007

Proper cache expiry with after_commit

I'm using cache_fu to handle all of my ActiveRecord memcaching these days. It is an amazingly simple and powerful addition to AR to easily use memcached, and I highly recommend it. I've really only had this one problem with it, the suggested cache expiry is:

after_save :expire_cache
after_destroy :expire_cache

I noticed a troubling thing about using after_save to expire caches, after_save is still within the transaction that save is automatically wrapped in. I assume this is to protect against an error during the after_save call, so you can roll back the database to it's previous state. When you're expiring caches though, you need to make sure the underlying data is actually changed before you expire the cache, otherwise you risk caching the old data before the new data is committed. This could go unnoticed, but if you're using optimistic locking, it raises exceptions when you try to save stale data.

To be a little more clear, a normal looks something like this:

SQL (0.000363)   BEGIN
Post Update (0.000572)   UPDATE posts SET "created_at" = '2007-06-13 21:15:40.212720',
"last_edited_at" = NULL, "user_id" = 8, "body" = 'hello world', "updated_at" = '2007-06-16
18:45:24.412415', "topic_id" = 7 WHERE "id" = 11
SQL (0.000893)   COMMIT

Notice how everything is wrapped nicely with a BEGIN and COMMIT.

Now after adding an after_save method that simply logs "Expire Cache!" we can see the order of events:

SQL (0.000363)   BEGIN
Post Update (0.000591)   UPDATE posts SET "created_at" = '2007-06-13 21:15:40.212720',
"last_edited_at" = NULL, "user_id" = 8, "body" = 'hello world', "updated_at" = '2007-06-16
18:49:49.869814', "topic_id" = 7 WHERE "id" = 11
Expire Cache!
SQL (0.000874)   COMMIT

If we imagine that Expire Cache! took 10 seconds to run, we can see that there is a measurable amount of time between Expire Cache! and the COMMIT. Database servers don't want to hand out incomplete or just wrong data, so they will serve the "old" data during this time, switching to the UPDATEd data after the COMMIT. If this post gets requested again during that 10 second window, it will be cached as the "old" data. Now we have a database with one value, and a cache with another, but they both think they have the correct data. We need to move the expire cache operation outside of the commit to remove this problem area.

Enter after_commit

We add a callback right after any save or destroy operation, which does the operation and then calls after_commit. Theoretically this works with update_attribute and any other AcitveRecord write operation, but I haven't fully tested those cases.

module ActiveRecord
  class Base

    class << self

      # Class methods

      def after_commit(*callbacks, &block)
        callbacks << block if block_given?
        write_inheritable_array(:after_commit, callbacks)


    # Instance Methods

    def save_with_after_commit_callback(*args)
      value = save_without_after_commit_callback(args)
      return value
    alias_method_chain :save, :after_commit_callback

    def save_with_after_commit_callback!(*args)
      value = save_without_after_commit_callback!
      return value
    alias_method_chain :save!, :after_commit_callback

    def destroy_with_after_commit_callback
      value = destroy_without_after_commit_callback
      return value
    alias_method_chain :destroy, :after_commit_callback


Since we wrapped save and destroy to call after_commit, we need only add one callback to expire caches now:

after_commit :expire_cache

And here is the log:

SQL (0.000365)   BEGIN
Post Update (0.000772)   UPDATE posts SET "created_at" = '2007-06-13 21:15:40.212720',
"last_edited_at" = NULL, "user_id" = 8, "body" = 'hello world', "updated_at" = '2007-06-16
19:05:55.108429', "topic_id" = 7 WHERE "id" = 11
SQL (0.000929)   COMMIT
Expire Cache!

Friday, May 4, 2007


Update: I have a patch awaiting approval on test-spec to add should.differ, change, increase and decrease. These read a lot better than the assert_difference/change methods.
# Checks multiple methods on object before and after the block asserting that there is
# a different value returned for each method. This is especially good for non-numeric fields.
def assert_changes(object, methods = nil)
  methods = [methods] unless methods.is_a? Array

  initial_values = {}
  for method in methods
    initial_values[method] = object.send(method)

  object.reload if object.respond_to? :reload

  for method in methods
    assert_not_equal initial_values[method], object.send(method), "#{object}##{method}"

alias assert_change assert_changes