unshorten 0.1.0

data/bin/unshorten

@@ -0,0 +1,10 @@
+#!/usr/bin/env ruby
+
+require 'unshorten'
+
+if ARGV.empty?
+    STDERR << "USAGE: unshorten URL\n"
+else
+    ARGV.each { |a| puts Unshorten[a] }
+end
+

data/lib/unshorten.rb

@@ -0,0 +1,86 @@
+require 'net/http'
+require 'uri'
+
+# Get original URLs from shortened ones.
+class Unshorten
+
+  # Cache entities limit
+  CACHE_SIZE_LIMIT = 1024
+
+  # Default options for unshorten
+  DEFAULT_OPTIONS = {
+    :max_level => 10,
+    :timeout => 2,
+    :use_cache => true,
+    :add_missing_http => true
+  }
+
+  @@cache = { }
+
+  class << self
+
+    # Unshorten a URL
+    #
+    # @param url [String] A shortened URL
+    # @param options [Hash] A set of options
+    # @option options [Integer] :max_level Max redirect times
+    # @option options [Integer] :timeout Timeout in seconds, for every request
+    # @option options [Boolean] :use_cache Use cached result if available
+    # @option options [Boolean] :add_missing_http add 'http://' if missing
+    # @see DEFAULT_OPTIONS
+    #
+    # @return Original url, a url that does not redirect
+    def unshorten(url, options = {})
+      DEFAULT_OPTIONS.each { |k, v| (options[k] = v) unless options.has_key? k }
+
+      follow(url, options)
+    end
+
+    alias :'[]' :unshorten
+
+    private
+
+    def expire_cache #:nodoc:
+      @@cache = { }
+    end
+
+    def mix_options(old, *new) #:nodoc:
+      options = old.dup
+      new.each { |n| n.each { |k, v| options[k] = v } }
+      options
+    end
+
+    def add_missing_http(url) #:nodoc:
+      if url =~ /^https?:/i
+        url
+      else
+        "http://#{url}"
+      end
+    end
+
+    def follow(url, options = DEFAULT_OPTIONS, level = 0) #:nodoc:
+      return @@cache[url] if options[:use_cache] and @@cache[url]
+
+      url = add_missing_http(url) if options[:add_missing_http]
+      return url if level >= options[:max_level]
+
+      uri = URI.parse(url)
+      http = Net::HTTP.new(uri.host, uri.port)
+      http.read_timeout = options[:timeout]
+
+      response = http.request_head(uri.path.empty? ? '/' : uri.path) rescue nil
+
+      if response.is_a? Net::HTTPRedirection and response['location'] then
+        expire_cache if @@cache.size > CACHE_SIZE_LIMIT
+        @@cache[url] = follow(response['location'], options, level + 1)
+      else
+        url
+      end
+    end
+
+  end
+
+end
+
+# vim:et:ts=2 sw=2
+

data/test/test_unshorten.rb

@@ -0,0 +1,16 @@
+require 'test/unit'
+require 'unshorten'
+
+class UnshortenTest < Test::Unit::TestCase
+    ORIGINAL_URL = 'http://dir.yahoo.com/Reference/Libraries/Library_and_Information_Science/Metadata/URIs___Universal_Resource_Identifiers/URLs___Uniform_Resource_Locators/URL_Shortening/'
+    SHORTENED_URL = 'http://tinyurl.com/j'
+
+    def test_unshorten_alias
+        assert_equal ORIGINAL_URL, Unshorten[SHORTENED_URL, :use_cache => false]
+    end
+
+    def test_option_max_level
+        assert_equal SHORTENED_URL, Unshorten.unshorten(SHORTENED_URL, :max_level => 0, :use_cache => false)
+    end
+
+end

metadata

@@ -0,0 +1,50 @@
+--- !ruby/object:Gem::Specification
+name: unshorten
+version: !ruby/object:Gem::Version
+  version: 0.1.0
+  prerelease: 
+platform: ruby
+authors:
+- Wu Jun
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2011-12-10 00:00:00.000000000 Z
+dependencies: []
+description: Get original URLs from shortened ones
+email: quark@lihdd.net
+executables:
+- unshorten
+extensions: []
+extra_rdoc_files: []
+files:
+- lib/unshorten.rb
+- test/test_unshorten.rb
+- bin/unshorten
+homepage: https://github.com/quark-zju/unshorten
+licenses: []
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  none: false
+  requirements:
+  - - ! '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 1.8.11
+signing_key: 
+specification_version: 3
+summary: Unshorten URLs
+test_files:
+- test/test_unshorten.rb
+has_rdoc: