html_truncator 0.1.0 → 0.1.1

data/README.md

@@ -63,10 +63,10 @@ Rails has a `truncate` helper, but as the doc says:
 
 I know there are some Ruby code to truncate HTML, like:
 
-* https://github.com/hgimenez/truncate_html
-* https://gist.github.com/101410
-* http://henrik.nyh.se/2008/01/rails-truncate-html-helper
-* http://blog.madebydna.com/all/code/2010/06/04/ruby-helper-to-cleanly-truncate-html.html
+* [https://github.com/hgimenez/truncate_html](https://github.com/hgimenez/truncate_html)
+* [https://gist.github.com/101410](https://gist.github.com/101410)
+* [http://henrik.nyh.se/2008/01/rails-truncate-html-helper](http://henrik.nyh.se/2008/01/rails-truncate-html-helper)
+* [http://blog.madebydna.com/all/code/2010/06/04/ruby-helper-to-cleanly-truncate-html.html](http://blog.madebydna.com/all/code/2010/06/04/ruby-helper-to-cleanly-truncate-html.html)
 
 But I'm not pleased with these solutions: they are either based on regexp for
 parsing the content (too fragile), they don't put the ellipsis where expected,

data/lib/html_truncator.rb

@@ -4,38 +4,46 @@ require "nokogiri"
 class HTML_Truncator
   def self.truncate(text, max_words, ellipsis="...")
     doc = Nokogiri::HTML::DocumentFragment.parse(text)
-    doc.truncate(max_words, ellipsis)
+    doc.truncate(max_words, ellipsis).first
   end
 end
 
 
 class Nokogiri::HTML::DocumentFragment
-  def truncate(max_words, ellipsis)
-    inner_truncate(max_words, ellipsis).first
+  def ellipsable?
+    true
   end
 end
 
 class Nokogiri::XML::Node
   def truncate(max_words, ellipsis)
-    inner, remaining = inner_truncate(max_words, ellipsis)
+    return ["", 1, ellipsis] if max_words == 0 && !ellipsable?
+    inner, remaining, ellipsis = inner_truncate(max_words, ellipsis)
     children.remove
     add_child Nokogiri::HTML::DocumentFragment.parse(inner)
-    [to_xml, max_words - remaining]
+    [to_xml(:indent => 0), max_words - remaining, ellipsis]
   end
 
   def inner_truncate(max_words, ellipsis)
     inner, remaining = "", max_words
     self.children.each do |node|
-      txt, nb = node.truncate(remaining, ellipsis)
+      txt, nb, ellipsis = node.truncate(remaining, ellipsis)
       remaining -= nb
+      if remaining < 0
+        inner += txt
+        if ellipsable?
+          inner += ellipsis
+          ellipsis = ""
+        end
+        break
+      end
       inner += txt
-      break if remaining < 0
     end
-    [inner, remaining]
+    [inner, remaining, ellipsis]
   end
 
-  def nb_words
-    inner_text.split.length
+  def ellipsable?
+    %w(p ol ul li div header article nav section footer aside dd dt dl).include? name
   end
 end
 
@@ -43,8 +51,7 @@ class Nokogiri::XML::Text
   def truncate(max_words, ellipsis)
     words    = content.split
     nb_words = words.length
-    return [to_xhtml, nb_words] if nb_words <= max_words
-    return [ellipsis, 1] if max_words == 0
-    [words.slice(0, max_words).join(' ') + ellipsis, nb_words]
+    return [to_xhtml, nb_words, ellipsis] if nb_words <= max_words && max_words > 0
+    [words.slice(0, max_words).join(' '), nb_words, ellipsis]
   end
 end

metadata

@@ -5,8 +5,8 @@ version: !ruby/object:Gem::Version
   segments: 
   - 0
   - 1
-  - 0
-  version: 0.1.0
+  - 1
+  version: 0.1.1
 platform: ruby
 authors: 
 - Bruno Michel
@@ -14,7 +14,7 @@ autorequire:
 bindir: bin
 cert_chain: []
 
-date: 2011-01-09 00:00:00 +01:00
+date: 2011-01-12 00:00:00 +01:00
 default_executable: 
 dependencies: 
 - !ruby/object:Gem::Dependency