twitter-text 1.4.16 → 1.4.17

data/lib/autolink.rb

@@ -1,3 +1,5 @@
+# encoding: UTF-8
+
 require 'set'
 
 module Twitter
@@ -148,14 +150,11 @@ module Twitter
       url_entities = {}
       if options[:url_entities]
         options[:url_entities].each do |entity|
-          entity = entity.with_indifferent_access
-          url_entities[entity[:url]] = entity
+          url_entities[entity["url"]] = entity
         end
         options.delete(:url_entities)
       end
 
-      html_attrs = html_attrs_for_options(options)
-
       Twitter::Rewriter.rewrite_urls(text) do |url|
         # In the case of t.co URLs, don't allow additional path characters
         after = ""
@@ -171,11 +170,67 @@ module Twitter
         end
 
         display_url = url
-        if url_entities[url] && url_entities[url][:display_url]
-          display_url = url_entities[url][:display_url]
+        link_text = html_escape(display_url)
+        if url_entities[url] && url_entities[url]["display_url"]
+          display_url = url_entities[url]["display_url"]
+          expanded_url = url_entities[url]["expanded_url"]
+          if !options[:title]
+            options[:title] = expanded_url
+          end
+
+          # Goal: If a user copies and pastes a tweet containing t.co'ed link, the resulting paste
+          # should contain the full original URL (expanded_url), not the display URL.
+          #
+          # Method: Whenever possible, we actually emit HTML that contains expanded_url, and use
+          # font-size:0 to hide those parts that should not be displayed (because they are not part of display_url).
+          # Elements with font-size:0 get copied even though they are not visible.
+          # Note that display:none doesn't work here. Elements with display:none don't get copied.
+          #
+          # Additionally, we want to *display* ellipses, but we don't want them copied.  To make this happen we
+          # wrap the ellipses in a tco-ellipsis class and provide an onCopy handler that sets display:none on
+          # everything with the tco-ellipsis class.
+          # 
+          # Exception: pic.twitter.com images, for which expandedUrl = "https://twitter.com/#!/username/status/1234/photo/1
+          # For those URLs, display_url is not a substring of expanded_url, so we don't do anything special to render the elided parts.
+          # For a pic.twitter.com URL, the only elided part will be the "https://", so this is fine.
+          display_url_sans_ellipses = display_url.sub("…", "")
+          if expanded_url.include?(display_url_sans_ellipses)
+            display_url_index = expanded_url.index(display_url_sans_ellipses)
+            before_display_url = expanded_url.slice(0, display_url_index)
+            # Portion of expanded_url that comes after display_url
+            after_display_url = expanded_url.slice(display_url_index + display_url_sans_ellipses.length, 999999)
+            preceding_ellipsis = display_url.match(/^…/) ? "…" : ""
+            following_ellipsis = display_url.match(/…$/) ? "…" : ""
+            # As an example: The user tweets "hi http://longdomainname.com/foo"
+            # This gets shortened to "hi http://t.co/xyzabc", with display_url = "…nname.com/foo"
+            # This will get rendered as:
+            # <span class='tco-ellipsis'> <!-- This stuff should get displayed but not copied -->
+            #   …
+            #   <!-- There's a chance the onCopy event handler might not fire. In case that happens,
+            #        we include an &nbsp; here so that the … doesn't bump up against the URL and ruin it.
+            #        The &nbsp; is inside the tco-ellipsis span so that when the onCopy handler *does*
+            #        fire, it doesn't get copied.  Otherwise the copied text would have two spaces in a row,
+            #        e.g. "hi  http://longdomainname.com/foo".
+            #   <span style='font-size:0'>&nbsp;</span>
+            # </span>
+            # <span style='font-size:0'>  <!-- This stuff should get copied but not displayed -->
+            #   http://longdomai
+            # </span>
+            # <span class='js-display-url'> <!-- This stuff should get displayed *and* copied -->
+            #   nname.com/foo
+            # </span>
+            # <span class='tco-ellipsis'> <!-- This stuff should get displayed but not copied -->
+            #   <span style='font-size:0'>&nbsp;</span>
+            #   …
+            # </span>
+            invisible = "style='font-size:0; line-height:0'"
+            link_text = "<span class='tco-ellipsis'>#{preceding_ellipsis}<span #{invisible}>&nbsp;</span></span><span #{invisible}>#{html_escape before_display_url}</span><span class='js-display-url'>#{html_escape display_url_sans_ellipses}</span><span #{invisible}>#{after_display_url}</span><span class='tco-ellipsis'><span #{invisible}>&nbsp;</span>#{following_ellipsis}</span>"
+          end
         end
 
-        %(<a href="#{href}"#{html_attrs}>#{html_escape(display_url)}</a>#{after})
+        html_attrs = html_attrs_for_options(options)
+
+        %(<a href="#{href}"#{html_attrs}>#{link_text}</a>#{after})
       end
     end
 

data/lib/regex.rb

@@ -65,7 +65,19 @@ module Twitter
           regex_range(0xc0, 0xd6),
           regex_range(0xd8, 0xf6),
           regex_range(0xf8, 0xff),
-          regex_range(0x015f)
+          regex_range(0x0100, 0x024f),
+          regex_range(0x0253, 0x0254),
+          regex_range(0x0256, 0x0257),
+          regex_range(0x0259),
+          regex_range(0x025b),
+          regex_range(0x0263),
+          regex_range(0x0268),
+          regex_range(0x026f),
+          regex_range(0x0272),
+          regex_range(0x0289),
+          regex_range(0x028b),
+          regex_range(0x02bb),
+          regex_range(0x1e00, 0x1eff)
     ].join('').freeze
 
     NON_LATIN_HASHTAG_CHARS = [
@@ -74,13 +86,48 @@ module Twitter
       regex_range(0x0500, 0x0527), # Cyrillic Supplement
       regex_range(0x2de0, 0x2dff), # Cyrillic Extended A
       regex_range(0xa640, 0xa69f), # Cyrillic Extended B
-      # Hangul (Korean)
+      regex_range(0x0591, 0x05bd), # Hebrew
+      regex_range(0x05bf),
+      regex_range(0x05c1, 0x05c2),
+      regex_range(0x05c4, 0x05c5),
+      regex_range(0x05c7),
+      regex_range(0x05d0, 0x05ea),
+      regex_range(0x05f0, 0x05f2),
+      regex_range(0xfb12, 0xfb28), # Hebrew Presentation Forms
+      regex_range(0xfb2a, 0xfb36),
+      regex_range(0xfb38, 0xfb3c),
+      regex_range(0xfb3e),
+      regex_range(0xfb40, 0xfb41),
+      regex_range(0xfb43, 0xfb44),
+      regex_range(0xfb46, 0xfb4f),
+      regex_range(0x0610, 0x061a), # Arabic
+      regex_range(0x0620, 0x065f),
+      regex_range(0x066e, 0x06d3),
+      regex_range(0x06d5, 0x06dc),
+      regex_range(0x06de, 0x06e8),
+      regex_range(0x06ea, 0x06ef),
+      regex_range(0x06fa, 0x06fc),
+      regex_range(0x06ff),
+      regex_range(0x0750, 0x077f), # Arabic Supplement
+      regex_range(0x08a0),         # Arabic Extended A
+      regex_range(0x08a2, 0x08ac),
+      regex_range(0x08e4, 0x08fe),
+      regex_range(0xfb50, 0xfbb1), # Arabic Pres. Forms A
+      regex_range(0xfbd3, 0xfd3d),
+      regex_range(0xfd50, 0xfd8f),
+      regex_range(0xfd92, 0xfdc7),
+      regex_range(0xfdf0, 0xfdfb),
+      regex_range(0xfe70, 0xfe74), # Arabic Pres. Forms B
+      regex_range(0xfe76, 0xfefc),
+      regex_range(0x200c, 0x200c), # Zero-Width Non-Joiner
+      regex_range(0x0e01, 0x0e3a), # Thai
+      regex_range(0x0e40, 0x0e4e), # Hangul (Korean)
       regex_range(0x1100, 0x11ff), # Hangul Jamo
       regex_range(0x3130, 0x3185), # Hangul Compatibility Jamo
       regex_range(0xA960, 0xA97F), # Hangul Jamo Extended-A
       regex_range(0xAC00, 0xD7AF), # Hangul Syllables
       regex_range(0xD7B0, 0xD7FF), # Hangul Jamo Extended-B
-      regex_range(0xFFA1, 0xFFDC) # Half-width Hangul
+      regex_range(0xFFA1, 0xFFDC)  # Half-width Hangul
     ].join('').freeze
     REGEXEN[:latin_accents] = /[#{LATIN_ACCENTS}]+/o
 

data/spec/autolinking_spec.rb

@@ -546,6 +546,23 @@ describe Twitter::Autolink do
   end
 
   describe "autolinking options" do
+    it "should show display_url when :url_entities provided" do
+      linked = TestAutolink.new.auto_link("http://t.co/0JG5Mcq", :url_entities => [{
+        "url" => "http://t.co/0JG5Mcq",
+        "display_url" => "blog.twitter.com/2011/05/twitte…",
+        "expanded_url" => "http://blog.twitter.com/2011/05/twitter-for-mac-update.html",
+        "indices" => [
+          84,
+          103
+        ]
+      }])
+      html = Nokogiri::HTML(linked)
+      html.search('a').should_not be_empty
+      html.search('a[@href="http://t.co/0JG5Mcq"]').should_not be_empty
+      html.search('span[@class=js-display-url]').inner_text.should == "blog.twitter.com/2011/05/twitte"
+      html.inner_text.should == " http://blog.twitter.com/2011/05/twitter-for-mac-update.html …"
+    end
+
     it "should apply :url_class as a CSS class" do
       linked = TestAutolink.new.auto_link("http://example.com/", :url_class => 'myclass')
       linked.should have_autolinked_url('http://example.com/')

data/twitter-text.gemspec

@@ -2,7 +2,7 @@
 
 Gem::Specification.new do |s|
   s.name = "twitter-text"
-  s.version = "1.4.16"
+  s.version = "1.4.17"
   s.authors = ["Matt Sanford", "Patrick Ewing", "Ben Cherry", "Britt Selvitelle",
                "Raffi Krikorian", "J.P. Cummins", "Yoshimasa Niwa", "Keita Fujii"]
   s.email = ["matt@twitter.com", "patrick.henry.ewing@gmail.com", "bcherry@gmail.com", "bs@brittspace.com",

metadata

@@ -1,13 +1,13 @@
 --- !ruby/object:Gem::Specification 
 name: twitter-text
 version: !ruby/object:Gem::Version 
-  hash: 39
+  hash: 37
   prerelease: 
   segments: 
   - 1
   - 4
-  - 16
-  version: 1.4.16
+  - 17
+  version: 1.4.17
 platform: ruby
 authors: 
 - Matt Sanford
@@ -22,7 +22,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2012-02-07 00:00:00 -08:00
+date: 2012-02-23 00:00:00 -08:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency