autolinker 0.1

data/test/helper.rb

@@ -0,0 +1,17 @@
+require 'rubygems'
+require 'bundler'
+
+begin
+  Bundler.setup(:default, :development)
+rescue Bundler::BundlerError => e
+  $stderr.puts e.message
+  $stderr.puts "Run `bundle install` to install missing gems"
+  exit e.status_code
+end
+require 'minitest/unit'
+require 'minitest/autorun'
+
+$LOAD_PATH.unshift(File.join(File.dirname(__FILE__), '..', 'lib'))
+$LOAD_PATH.unshift(File.dirname(__FILE__))
+require 'autolinker'
+

data/test/test_autolinker.rb

@@ -0,0 +1,270 @@
+require 'helper'
+
+class TestAutolinker < MiniTest::Unit::TestCase
+
+  def setup
+    @text_helper = Autolinker::TextHelper.new
+  end
+
+  def test_auto_link_within_tags
+    link_raw = 'http://www.rubyonrails.org/images/rails.png'
+    link_result = %Q(<img src="#{link_raw}" />)
+    assert_equal link_result, @text_helper.auto_link(link_result)
+  end
+
+  def test_auto_link_with_brackets
+    link1_raw = 'http://en.wikipedia.org/wiki/Sprite_(computer_graphics)'
+    link1_result = generate_result(link1_raw)
+    assert_equal link1_result, @text_helper.auto_link(link1_raw)
+    assert_equal "(link: #{link1_result})", @text_helper.auto_link("(link: #{link1_raw})")
+
+    link2_raw = 'http://en.wikipedia.org/wiki/Sprite_[computer_graphics]'
+    link2_result = generate_result(link2_raw)
+    assert_equal link2_result, @text_helper.auto_link(link2_raw)
+    assert_equal "[link: #{link2_result}]", @text_helper.auto_link("[link: #{link2_raw}]")
+
+    link3_raw = 'http://en.wikipedia.org/wiki/Sprite_{computer_graphics}'
+    link3_result = generate_result(link3_raw)
+    assert_equal link3_result, @text_helper.auto_link(link3_raw)
+    assert_equal "{link: #{link3_result}}", @text_helper.auto_link("{link: #{link3_raw}}")
+  end
+
+  def test_auto_link_with_options_hash
+    assert_equal 'Welcome to my new blog at <a class="menu" href="http://www.myblog.com/" target="_blank">http://www.myblog.com/</a>. Please e-mail me at <a class="menu" href="mailto:me@email.com" target="_blank">me@email.com</a>.',
+                 @text_helper.auto_link("Welcome to my new blog at http://www.myblog.com/. Please e-mail me at me@email.com.",
+                                        :link => :all, :html => { :class => "menu", :target => "_blank" })
+  end
+
+  def test_auto_link_with_multiple_trailing_punctuations
+    url = "http://youtube.com"
+    url_result = generate_result(url)
+    assert_equal url_result, @text_helper.auto_link(url)
+    assert_equal "(link: #{url_result}).", @text_helper.auto_link("(link: #{url}).")
+  end
+
+  def test_auto_link_with_block
+    url = "http://api.rubyonrails.com/Foo.html"
+    email = "fantabulous@shiznadel.ic"
+
+    assert_equal %(<p><a href="#{url}">#{url[0...7]}...</a><br /><a href="mailto:#{email}">#{email[0...7]}...</a><br /></p>), @text_helper.auto_link("<p>#{url}<br />#{email}<br /></p>") { |_url| truncate(_url, 10) }
+  end
+
+  def test_auto_link_with_block_with_html
+    pic = "http://example.com/pic.png"
+    url = "http://example.com/album?a&b=c"
+
+    assert_equal %(My pic: <a href="#{pic}"><img src="#{pic}" width="160px"></a> -- full album here #{generate_result(url)}), @text_helper.auto_link("My pic: #{pic} -- full album here #{url}") { |link|
+      if link =~ /\.(jpg|gif|png|bmp|tif)$/i
+        %(<img src="#{link}" width="160px">)
+      else
+        link
+      end
+    }
+  end
+
+  def test_auto_link_should_sanitize_input_when_sanitize_option_is_not_false
+    link_raw = %{http://www.rubyonrails.com?id=1&num=2}
+    malicious_script = '<script>alert("malicious!")</script>'
+    assert_equal %{<a href="http://www.rubyonrails.com?id=1&num=2">http://www.rubyonrails.com?id=1&num=2</a>}, @text_helper.auto_link("#{link_raw}#{malicious_script}")
+  end
+
+  def test_auto_link_should_sanitize_input_with_sanitize_options
+    link_raw = %{http://www.rubyonrails.com?id=1&num=2}
+    malicious_script = '<script>alert("malicious!")</script>'
+    text_with_attributes = %{<a href="http://ruby-lang-org" target="_blank" data-malicious="inject">Ruby</a>}
+
+    text_result = %{<a class="big" href="http://www.rubyonrails.com?id=1&num=2">http://www.rubyonrails.com?id=1&num=2</a><a href="http://ruby-lang-org" target="_blank">Ruby</a>}
+    assert_equal text_result, @text_helper.auto_link("#{link_raw}#{malicious_script}#{text_with_attributes}",
+                                                     :sanitize_options => { :attributes => ["target", "href"] },
+                                                     :html => { :class => 'big' })
+  end
+
+  def test_auto_link_should_not_sanitize_input_when_sanitize_option_is_false
+    link_raw = %{http://www.rubyonrails.com?id=1&num=2}
+    malicious_script = '<script>alert("malicious!")</script>'
+
+    assert_equal %{<a href="http://www.rubyonrails.com?id=1&num=2">http://www.rubyonrails.com?id=1&num=2</a><script>alert("malicious!")</script>}, @text_helper.auto_link("#{link_raw}#{malicious_script}", :sanitize => false)
+  end
+
+  def test_auto_link_other_protocols
+    ftp_raw = 'ftp://example.com/file.txt'
+    assert_equal %(Download #{generate_result(ftp_raw)}), @text_helper.auto_link("Download #{ftp_raw}")
+
+    file_scheme = 'file:///home/username/RomeoAndJuliet.pdf'
+    assert_equal generate_result(file_scheme), @text_helper.auto_link(file_scheme)
+  end
+
+  def test_auto_link_already_linked
+    linked1 = generate_result('Ruby On Rails', 'http://www.rubyonrails.com')
+    linked2 = %('<a href="http://www.example.com">www.example.com</a>')
+    linked3 = %('<a href="http://www.example.com" rel="nofollow">www.example.com</a>')
+    linked4 = %('<a href="http://www.example.com"><b>www.example.com</b></a>')
+    linked5 = %('<a href="#close">close</a> <a href="http://www.example.com"><b>www.example.com</b></a>')
+    linked6 = %('<a href="#close">close</a> <a href="http://www.example.com" target="_blank" data-ruby="ror"><b>www.example.com</b></a>')
+    assert_equal linked1, @text_helper.auto_link(linked1)
+    assert_equal linked2, @text_helper.auto_link(linked2)
+    assert_equal linked3, @text_helper.auto_link(linked3, :sanitize => false)
+    assert_equal linked4, @text_helper.auto_link(linked4)
+    assert_equal linked5, @text_helper.auto_link(linked5)
+    assert_equal linked6, @text_helper.auto_link(linked6, :sanitize_options => { :attributes => ["href", "target", "data-ruby"] })
+
+    linked_email = %Q(<a href="mailto:david@loudthinking.com">Mail me</a>)
+    assert_equal linked_email, @text_helper.auto_link(linked_email)
+  end
+
+
+  def test_auto_link_at_eol
+    url1 = "http://api.rubyonrails.com/Foo.html"
+    url2 = "http://www.ruby-doc.org/core/Bar.html"
+
+    assert_equal %(<p><a href="#{url1}">#{url1}</a><br /><a href="#{url2}">#{url2}</a><br /></p>), @text_helper.auto_link("<p>#{url1}<br />#{url2}<br /></p>")
+  end
+
+  def test_auto_link
+    email_raw = 'david@loudthinking.com'
+    email_result = %{<a href="mailto:#{email_raw}">#{email_raw}</a>}
+    link_raw = 'http://www.rubyonrails.com'
+    link_result = generate_result(link_raw)
+    link_result_with_options = %{<a href="#{link_raw}" target="_blank">#{link_raw}</a>}
+
+    assert_equal '', @text_helper.auto_link(nil)
+    assert_equal '', @text_helper.auto_link('')
+    assert_equal "#{link_result} #{link_result} #{link_result}", @text_helper.auto_link("#{link_raw} #{link_raw} #{link_raw}")
+
+    assert_equal %(hello #{email_result}), @text_helper.auto_link("hello #{email_raw}", :email_addresses)
+    assert_equal %(Go to #{link_result}), @text_helper.auto_link("Go to #{link_raw}", :urls)
+    assert_equal %(Go to #{link_raw}), @text_helper.auto_link("Go to #{link_raw}", :email_addresses)
+    assert_equal %(Go to #{link_result} and say hello to #{email_result}), @text_helper.auto_link("Go to #{link_raw} and say hello to #{email_raw}")
+    assert_equal %(<p>Link #{link_result}</p>), @text_helper.auto_link("<p>Link #{link_raw}</p>")
+    assert_equal %(<p>#{link_result} Link</p>), @text_helper.auto_link("<p>#{link_raw} Link</p>")
+    assert_equal %(<p>Link #{link_result_with_options}</p>), @text_helper.auto_link("<p>Link #{link_raw}</p>", :all, { :target => "_blank" })
+    assert_equal %(Go to #{link_result}.), @text_helper.auto_link(%(Go to #{link_raw}.))
+    assert_equal %(<p>Go to #{link_result}, then say hello to #{email_result}.</p>), @text_helper.auto_link(%(<p>Go to #{link_raw}, then say hello to #{email_raw}.</p>))
+    assert_equal %(#{link_result} #{link_result}), @text_helper.auto_link(%(#{link_result} #{link_raw}))
+
+    email2_raw = '+david@loudthinking.com'
+    email2_result = %{<a href="mailto:#{email2_raw}">#{email2_raw}</a>}
+    assert_equal email2_result, @text_helper.auto_link(email2_raw)
+    assert_equal email2_result, @text_helper.auto_link(email2_raw, :all)
+    assert_equal email2_result, @text_helper.auto_link(email2_raw, :email_addresses)
+
+    link2_raw = 'www.rubyonrails.com'
+    link2_result = generate_result(link2_raw, "http://#{link2_raw}")
+    assert_equal %(Go to #{link2_result}), @text_helper.auto_link("Go to #{link2_raw}", :urls)
+    assert_equal %(Go to #{link2_raw}), @text_helper.auto_link("Go to #{link2_raw}", :email_addresses)
+    assert_equal %(<p>Link #{link2_result}</p>), @text_helper.auto_link("<p>Link #{link2_raw}</p>")
+    assert_equal %(<p>#{link2_result} Link</p>), @text_helper.auto_link("<p>#{link2_raw} Link</p>")
+    assert_equal %(Go to #{link2_result}.), @text_helper.auto_link(%(Go to #{link2_raw}.))
+    assert_equal %(<p>Say hello to #{email_result}, then go to #{link2_result}.</p>), @text_helper.auto_link(%(<p>Say hello to #{email_raw}, then go to #{link2_raw}.</p>))
+
+    link3_raw = 'http://manuals.ruby-on-rails.com/read/chapter.need_a-period/103#page281'
+    link3_result = generate_result(link3_raw)
+    assert_equal %(Go to #{link3_result}), @text_helper.auto_link("Go to #{link3_raw}", :urls)
+    assert_equal %(Go to #{link3_raw}), @text_helper.auto_link("Go to #{link3_raw}", :email_addresses)
+    assert_equal %(<p>Link #{link3_result}</p>), @text_helper.auto_link("<p>Link #{link3_raw}</p>")
+    assert_equal %(<p>#{link3_result} Link</p>), @text_helper.auto_link("<p>#{link3_raw} Link</p>")
+    assert_equal %(Go to #{link3_result}.), @text_helper.auto_link(%(Go to #{link3_raw}.))
+    assert_equal %(<p>Go to #{link3_result}. Seriously, #{link3_result}? I think I'll say hello to #{email_result}. Instead.</p>),
+                 @text_helper.auto_link(%(<p>Go to #{link3_raw}. Seriously, #{link3_raw}? I think I'll say hello to #{email_raw}. Instead.</p>))
+
+    link4_raw = 'http://foo.example.com/controller/action?parm=value&p2=v2#anchor123'
+    link4_result = generate_result(link4_raw)
+    assert_equal %(<p>Link #{link4_result}</p>), @text_helper.auto_link("<p>Link #{link4_raw}</p>")
+    assert_equal %(<p>#{link4_result} Link</p>), @text_helper.auto_link("<p>#{link4_raw} Link</p>")
+
+    link5_raw = 'http://foo.example.com:3000/controller/action'
+    link5_result = generate_result(link5_raw)
+    assert_equal %(<p>#{link5_result} Link</p>), @text_helper.auto_link("<p>#{link5_raw} Link</p>")
+
+    link6_raw = 'http://foo.example.com:3000/controller/action+pack'
+    link6_result = generate_result(link6_raw)
+    assert_equal %(<p>#{link6_result} Link</p>), @text_helper.auto_link("<p>#{link6_raw} Link</p>")
+
+    link7_raw = 'http://foo.example.com/controller/action?parm=value&p2=v2#anchor-123'
+    link7_result = generate_result(link7_raw)
+    assert_equal %(<p>#{link7_result} Link</p>), @text_helper.auto_link("<p>#{link7_raw} Link</p>")
+
+    link8_raw = 'http://foo.example.com:3000/controller/action.html'
+    link8_result = generate_result(link8_raw)
+    assert_equal %(Go to #{link8_result}), @text_helper.auto_link("Go to #{link8_raw}", :urls)
+    assert_equal %(Go to #{link8_raw}), @text_helper.auto_link("Go to #{link8_raw}", :email_addresses)
+    assert_equal %(<p>Link #{link8_result}</p>), @text_helper.auto_link("<p>Link #{link8_raw}</p>")
+    assert_equal %(<p>#{link8_result} Link</p>), @text_helper.auto_link("<p>#{link8_raw} Link</p>")
+    assert_equal %(Go to #{link8_result}.), @text_helper.auto_link(%(Go to #{link8_raw}.))
+    assert_equal %(<p>Go to #{link8_result}. Seriously, #{link8_result}? I think I'll say hello to #{email_result}. Instead.</p>),
+                 @text_helper.auto_link(%(<p>Go to #{link8_raw}. Seriously, #{link8_raw}? I think I'll say hello to #{email_raw}. Instead.</p>))
+
+    link9_raw = 'http://business.timesonline.co.uk/article/0,,9065-2473189,00.html'
+    link9_result = generate_result(link9_raw)
+    assert_equal %(Go to #{link9_result}), @text_helper.auto_link("Go to #{link9_raw}", :urls)
+    assert_equal %(Go to #{link9_raw}), @text_helper.auto_link("Go to #{link9_raw}", :email_addresses)
+    assert_equal %(<p>Link #{link9_result}</p>), @text_helper.auto_link("<p>Link #{link9_raw}</p>")
+    assert_equal %(<p>#{link9_result} Link</p>), @text_helper.auto_link("<p>#{link9_raw} Link</p>")
+    assert_equal %(Go to #{link9_result}.), @text_helper.auto_link(%(Go to #{link9_raw}.))
+    assert_equal %(<p>Go to #{link9_result}. Seriously, #{link9_result}? I think I'll say hello to #{email_result}. Instead.</p>),
+                 @text_helper.auto_link(%(<p>Go to #{link9_raw}. Seriously, #{link9_raw}? I think I'll say hello to #{email_raw}. Instead.</p>))
+
+    link10_raw = 'http://www.mail-archive.com/ruby-talk@ruby-lang.org/'
+    link10_result = generate_result(link10_raw)
+    assert_equal %(<p>#{link10_result} Link</p>), @text_helper.auto_link("<p>#{link10_raw} Link</p>")
+
+    link11_raw = 'http://asakusa.rubyist.net/'
+    link11_result = generate_result(link11_raw)
+    assert_equal %(浅草.rbの公式サイトはこちら#{link11_result}), @text_helper.auto_link("浅草.rbの公式サイトはこちら#{link11_raw}")
+
+    link12_raw = 'http://tools.ietf.org/html/rfc3986'
+    link12_result = generate_result(link12_raw)
+    assert_equal %(<p>#{link12_result} text-after-nonbreaking-space</p>), @text_helper.auto_link("<p>#{link12_raw} text-after-nonbreaking-space</p>")
+  end
+
+  def test_auto_link_parsing
+    urls = %w(
+      http://www.rubyonrails.com
+      http://www.rubyonrails.com:80
+      http://www.rubyonrails.com/~minam
+      https://www.rubyonrails.com/~minam
+      http://www.rubyonrails.com/~minam/url%20with%20spaces
+      http://www.rubyonrails.com/foo.cgi?something=here
+      http://www.rubyonrails.com/foo.cgi?something=here&and=here
+      http://www.rubyonrails.com/contact;new
+      http://www.rubyonrails.com/contact;new%20with%20spaces
+      http://www.rubyonrails.com/contact;new?with=query&string=params
+      http://www.rubyonrails.com/~minam/contact;new?with=query&string=params
+      http://en.wikipedia.org/wiki/Wikipedia:Today%27s_featured_picture_%28animation%29/January_20%2C_2007
+      http://www.mail-archive.com/rails@lists.rubyonrails.org/
+      http://www.amazon.com/Testing-Equal-Sign-In-Path/ref=pd_bbs_sr_1?ie=UTF8&s=books&qid=1198861734&sr=8-1
+      http://en.wikipedia.org/wiki/Texas_hold'em
+      https://www.google.com/doku.php?id=gps:resource:scs:start
+      http://connect.oraclecorp.com/search?search[q]=green+france&search[type]=Group
+      http://of.openfoundry.org/projects/492/download#4th.Release.3
+      http://maps.google.co.uk/maps?f=q&q=the+london+eye&ie=UTF8&ll=51.503373,-0.11939&spn=0.007052,0.012767&z=16&iwloc=A
+      http://около.кола/колокола
+    )
+
+    urls.each do |url|
+      assert_equal generate_result(url), @text_helper.auto_link(url)
+    end
+  end
+
+  private
+  def generate_result(link_text, href = nil, escape = false)
+    href ||= link_text
+    if escape
+      %{<a href="#{CGI::escapeHTML href}">#{CGI::escapeHTML link_text}</a>}
+    else
+      %{<a href="#{href}">#{link_text}</a>}
+    end
+  end
+
+  def truncate(text, length, options = {})
+    options[:omission] ||= "..."
+
+    length_with_room_for_omission = length - options[:omission].length
+
+    stop = options[:separator] ?
+      (text.rindex(options[:separator], length_with_room_for_omission) || length_with_room_for_omission) : length_with_room_for_omission
+
+    (text.length > length ? text[0...stop] + options[:omission] : text).to_s
+  end
+
+end

metadata

@@ -0,0 +1,55 @@
+--- !ruby/object:Gem::Specification
+name: autolinker
+version: !ruby/object:Gem::Version
+  version: '0.1'
+platform: ruby
+authors:
+- Helmut Juskewycz
+autorequire: 
+bindir: bin
+cert_chain: []
+date: 2013-10-08 00:00:00.000000000 Z
+dependencies: []
+description: |2
+      An autolinking library that acts as a standalone replacement for Rails `auto_link`
+email: helmut@juskewycz
+executables: []
+extensions: []
+extra_rdoc_files: []
+files:
+- README.md
+- Rakefile
+- lib/autolinker.rb
+- lib/autolinker/text_helper.rb
+- lib/autolinker/html/node.rb
+- lib/autolinker/html/sanitizer.rb
+- lib/autolinker/html/tokenizer.rb
+- autolinker.gemspec
+- test/helper.rb
+- test/test_autolinker.rb
+homepage: http://github.com/hjuskewycz/autolinker
+licenses:
+- MIT
+metadata: {}
+post_install_message: 
+rdoc_options: []
+require_paths:
+- lib
+required_ruby_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+required_rubygems_version: !ruby/object:Gem::Requirement
+  requirements:
+  - - '>='
+    - !ruby/object:Gem::Version
+      version: '0'
+requirements: []
+rubyforge_project: 
+rubygems_version: 2.0.3
+signing_key: 
+specification_version: 4
+summary: Mostly autolinking
+test_files:
+- test/test_autolinker.rb