url_expander 0.1.1

data/lib/url_expander/expanders/scrape/simurl.rb

@@ -0,0 +1,34 @@
+module UrlExpander
+  module Expanders
+    #
+    # Expand http://simurl.com/ URLS
+    # Usage:
+    # UrlExpander::Client.expand("http://simurl.com/fendaz")
+    #
+    class Simurl < UrlExpander::Expanders::Scrape
+      PATTERN = %r'(http://simurl\.com(/[\w/]+))'
+      attr_reader :parent_klass
+      
+      def initialize(short_url="", options={})
+        @parent_klass = self.class
+        super(short_url, options)
+      end
+      
+      # How to scarpe the url from a Qsr.li html document
+      def self.scrape_url(html)
+        doc = Hpricot(html)
+        str = doc.at("meta").attributes['content']
+        start_slice = str.index("=") +1
+        end_slice   = str.size - start_slice 
+        url = str.slice(start_slice, end_slice)
+        raise UrlExpander::Error.new('page not found',404) if url == "/index.php"
+        url
+      end
+      
+      class Request
+        include HTTParty
+        base_uri 'http://simurl.com'
+      end
+    end
+  end
+end

data/lib/url_expander/expanders/scrape.rb

@@ -0,0 +1,71 @@
+require 'Hpricot'
+
+module UrlExpander
+  module Expanders
+    
+    # Some websites don't follow the coding standards. They don't provide an
+    # api and they don't provide 301 redirect. The only way to get the shorten
+    # url is by parsing the returned html doc.
+    # 
+    # To use the Scrape class, define your class inside scrape folder. Your class must 
+    # provide the following:
+    #   def initialize(short_url="", options={})
+    #   def self.scrape_url(html)
+    #   class Request
+    #
+    # Example:
+    # class Qsrli < UrlExpander::Expanders::Scrape
+    #   PATTERN = %r'(http://qsr\.li(/[\w/]+))'
+    #   attr_reader :parent_klass, :xpath
+    # 
+    #   def initialize(short_url="", options={})
+    #     @parent_klass = self.class
+    #     super(short_url, options)
+    #   end
+    # 
+    #   def self.scrape_url(html)
+    #     doc = Hpricot(html)
+    #     doc.at('//*[@id="framecontent"]').attributes["src"]
+    #   end
+    # 
+    #   class Request
+    #     include HTTParty
+    #     base_uri 'http://qsr.li'
+    #   end
+    # end
+    class Scrape
+      attr_accessor :long_url
+      attr_reader   :parttern, :parent_klass
+      
+      def initialize(short_url="",options={})
+        if short_url.match(parent_klass::PATTERN)
+          path = $2
+        else
+          raise 'invalid pattern'
+        end
+        
+        @long_url = fetch_url(path)
+      end
+      
+      
+      protected
+      
+      def fetch_url(path)
+        url = nil
+        result = parent_klass::Request.get(path, :follow_redirects => false)
+        case result.response
+          when Net::HTTPMovedPermanently
+            url = result['Location']
+          when Net::HTTPFound
+            url = result['location']
+          when Net::HTTPOK
+            url = parent_klass.scrape_url(result.response.body)
+        end
+        
+        raise UrlExpander::Error.new('page not found',404) if url.nil?
+        url
+      end
+      
+    end
+  end
+end

data/lib/url_expander.rb

@@ -0,0 +1,62 @@
+require 'rubygems'
+require 'httparty'
+
+$:.unshift(File.dirname(__FILE__))
+
+require 'url_expander/expanders/expanders'
+
+module UrlExpander
+  class Client
+    # Expand a given url.
+    def self.expand(url="",options = {})
+      
+      # Setup the default options
+      options[:nested_shortening] = true unless options.has_key?(:nested_shortening)
+      if options[:nested_shortening]
+        options[:limit] = 10 unless options.has_key?(:limit)
+      end
+      options[:config_file] = '~/url_expander_credentials.yml' unless options[:config_file]
+      
+      # We Reached the maximum number of redirections, quit.
+      raise ArgumentError, 'HTTP redirect too deep' if options[:nested_shortening] && options[:limit] == 0
+      
+      # Make sure we have a url
+      raise ArgumentError.new('Expander requires a short url') if url.nil? || url.empty?
+      exclude_klasses = ['UrlExpander::Expanders::Basic', 'UrlExpander::Expanders::API', 'UrlExpander::Expanders::Scrape']
+      
+      # Get the names for all the expanders except the basic default one.
+      mod     = UrlExpander::Expanders
+      expanders = mod.constants.collect{|c| mod.const_get(c)}.select{|c| c.class == Class && !exclude_klasses.include?(c.to_s)}
+      expander_klass = nil
+      
+      # Find the correct expander
+      expanders.each do |exp|
+        if(exp::PATTERN.match(url))
+          expander_klass = exp
+        end
+      end
+      @expander = (!expander_klass.nil?) ? expander_klass.new(url,options) : nil
+      
+      if @expander.nil? && !options[:is_redirection]
+        raise ArgumentError.new('Unknow url') 
+      end
+      
+      if options[:nested_shortening] & !@expander.nil?
+        options[:limit] -= 1
+        options[:is_redirection] = true
+        UrlExpander::Client.expand(@expander.long_url, options)
+      else
+        (@expander.nil?) ? url : @expander.long_url
+      end
+    end
+  end
+  
+  class Error < StandardError
+    attr_reader :code
+    alias :msg :message
+    def initialize(msg, code)
+      @code = code
+      super("#{msg} - '#{code}'")
+    end
+  end
+end

data/test/helper.rb

@@ -0,0 +1,18 @@
+require 'rubygems'
+require 'bundler'
+begin
+  Bundler.setup(:default, :development)
+rescue Bundler::BundlerError => e
+  $stderr.puts e.message
+  $stderr.puts "Run `bundle install` to install missing gems"
+  exit e.status_code
+end
+require 'test/unit'
+require 'shoulda'
+
+$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+$LOAD_PATH.unshift(File.dirname(__FILE__))
+require 'url_expander'
+
+class Test::Unit::TestCase
+end

data/test/test_url_expander.rb

@@ -0,0 +1,7 @@
+require 'helper'
+
+class TestUrlExpander < Test::Unit::TestCase
+  should "probably rename this file and start testing for real" do
+    flunk "hey buddy, you should probably rename this file and start testing for real"
+  end
+end

metadata

@@ -0,0 +1,216 @@
+--- !ruby/object:Gem::Specification 
+name: url_expander
+version: !ruby/object:Gem::Version 
+  hash: 25
+  prerelease: 
+  segments: 
+  - 0
+  - 1
+  - 1
+  version: 0.1.1
+platform: ruby
+authors: 
+- Moski
+autorequire: 
+bindir: bin
+cert_chain: []
+
+date: 2011-07-16 00:00:00 +03:00
+default_executable: 
+dependencies: 
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :runtime
+  requirement: &id001 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 5
+        segments: 
+        - 0
+        - 6
+        - 1
+        version: 0.6.1
+  name: httparty
+  version_requirements: *id001
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :runtime
+  requirement: &id002 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  name: json
+  version_requirements: *id002
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :runtime
+  requirement: &id003 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  name: hpricot
+  version_requirements: *id003
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :development
+  requirement: &id004 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  name: shoulda
+  version_requirements: *id004
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :development
+  requirement: &id005 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 23
+        segments: 
+        - 1
+        - 0
+        - 0
+        version: 1.0.0
+  name: bundler
+  version_requirements: *id005
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :development
+  requirement: &id006 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ~>
+      - !ruby/object:Gem::Version 
+        hash: 7
+        segments: 
+        - 1
+        - 6
+        - 4
+        version: 1.6.4
+  name: jeweler
+  version_requirements: *id006
+- !ruby/object:Gem::Dependency 
+  prerelease: false
+  type: :development
+  requirement: &id007 !ruby/object:Gem::Requirement 
+    none: false
+    requirements: 
+    - - ">="
+      - !ruby/object:Gem::Version 
+        hash: 3
+        segments: 
+        - 0
+        version: "0"
+  name: rcov
+  version_requirements: *id007
+description: Expand short urls from shortning services shuch as bitly and tinyurl
+email: abushaikh@gmail.com
+executables: []
+
+extensions: []
+
+extra_rdoc_files: 
+- LICENSE.txt
+- README.rdoc
+files: 
+- .document
+- Gemfile
+- Gemfile.lock
+- LICENSE.txt
+- README.rdoc
+- Rakefile
+- VERSION
+- lib/url_expander.rb
+- lib/url_expander/expanders/api.rb
+- lib/url_expander/expanders/api/4sq.rb
+- lib/url_expander/expanders/api/bitly.rb
+- lib/url_expander/expanders/api/budurl.rb
+- lib/url_expander/expanders/api/cligs.rb
+- lib/url_expander/expanders/api/decenturl.rb
+- lib/url_expander/expanders/api/fxnws.rb
+- lib/url_expander/expanders/api/googl.rb
+- lib/url_expander/expanders/api/isgd.rb
+- lib/url_expander/expanders/api/nytims.rb
+- lib/url_expander/expanders/api/tcrnch.rb
+- lib/url_expander/expanders/api/xrlus.rb
+- lib/url_expander/expanders/basic.rb
+- lib/url_expander/expanders/basic/adjix.rb
+- lib/url_expander/expanders/basic/digbig.rb
+- lib/url_expander/expanders/basic/doiop.rb
+- lib/url_expander/expanders/basic/easyurljp.rb
+- lib/url_expander/expanders/basic/justas.rb
+- lib/url_expander/expanders/basic/moourl.rb
+- lib/url_expander/expanders/basic/notlong.rb
+- lib/url_expander/expanders/basic/nutshellurl.rb
+- lib/url_expander/expanders/basic/owly.rb
+- lib/url_expander/expanders/basic/shrtst.rb
+- lib/url_expander/expanders/basic/snipurl.rb
+- lib/url_expander/expanders/basic/tco.rb
+- lib/url_expander/expanders/basic/tighturl.rb
+- lib/url_expander/expanders/basic/tinycc.rb
+- lib/url_expander/expanders/basic/tinyurl.rb
+- lib/url_expander/expanders/basic/twurlnl.rb
+- lib/url_expander/expanders/basic/urlie.rb
+- lib/url_expander/expanders/basic/youtube.rb
+- lib/url_expander/expanders/expanders.rb
+- lib/url_expander/expanders/scrape.rb
+- lib/url_expander/expanders/scrape/qsrli.rb
+- lib/url_expander/expanders/scrape/shorl.rb
+- lib/url_expander/expanders/scrape/simurl.rb
+- test/helper.rb
+- test/test_url_expander.rb
+has_rdoc: true
+homepage: http://github.com/moski/url_expander
+licenses: 
+- MIT
+post_install_message: 
+rdoc_options: []
+
+require_paths: 
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+required_rubygems_version: !ruby/object:Gem::Requirement 
+  none: false
+  requirements: 
+  - - ">="
+    - !ruby/object:Gem::Version 
+      hash: 3
+      segments: 
+      - 0
+      version: "0"
+requirements: []
+
+rubyforge_project: 
+rubygems_version: 1.6.2
+signing_key: 
+specification_version: 3
+summary: Expand short url from different services
+test_files: []
+