A couple of days ago, Tim Lucas wrote a cool article called “instance_eval brings sexy back” where he demonstrated how to use instance_eval to improve the usability of the match method. The downside, however, was that Tim’s technique required manually defining accessor methods each time match was used.

Myles Byrne rapidly responded with a cuter solution:

class MatchData
def matchnames(*names)
names.each_with_index do |name, index|
self.instance_eval “def #{name}; self[#{index+1}] end”
end
self
end
end

time_components = /(\\d+):(\\d+):(\\d+)/.match(“17:00:34″).matchnames(:hours, :mins, :secs)
time_components.hours

I quickly realized, however, that I never use match in this way. I prefer to use String#match instead (Regexp#match doesn’t click to me, it seems back to front). Read More

There’s more about Amazon S3 and Marcel Molina’s hot new library coming as the first day of the Ruby Advent Calendar (this Friday!), so I don’t want to say too much about it yet. For anyone already enjoying this library, however, I put together a scrappy program that lets me copy files up to S3 from the command line easily:

#!/usr/bin/env ruby

require ‘rubygems’
require ‘aws/s3′

local_file = ARGV[0]
bucket = ARGV[1]
mime_type = ARGV[2] || "application/octet-stream"

AWS::S3::Base.establish_connection!(
:access_key_id => ‘REPLACE_ME’,
:secret_access_key => ‘REPLACE_ME’
)

base_name = File.basename(local_file)

puts "Uploading #{local_file} as ‘#{base_name}’ to ‘#{bucket}’"

AWS::S3::S3Object.store(
base_name,
File.open(local_file),
bucket,
:content_type => mime_type
)

puts "Uploaded!"

Just name it s3cp (or similar), chmod it, and then you can do stuff like:

s3cp ~/somefile.whatever bucket Read More