nokogiri_truncate_html 0.0.1 → 0.0.2

data/lib/nokogiri_truncate_html/truncate_document.rb

@@ -0,0 +1,66 @@
+require 'nokogiri'
+require 'cgi'
+
+module NokogiriTruncateHtml
+  class TruncateFinished < Exception
+  end
+
+  class TruncateDocument < Nokogiri::XML::SAX::Document
+
+    def initialize
+      # We use the xhtml "flavour" so that we can decode
+      # apostrophes. That is the only difference between
+      # xhtml1 and html4.
+      @encoder = HTMLEntities.new('xhtml1')
+      @discard_first_element = false
+    end
+
+    def length=(length)
+      @length = length
+    end
+
+    def omission=(omission)
+      @omission = omission
+    end
+
+    def output
+      while @tags.size > 0
+        @output << "</#{@tags.pop}>"
+      end
+      @output
+    end
+
+    def start_document
+      @output, @chars_remaining, @tags = '', @length, []
+      @discard_first_element = false
+    end
+
+    def characters(string)
+      text = @encoder.decode(string)
+      @output << CGI.escapeHTML(text[0, @chars_remaining])
+      @chars_remaining -= text.length
+      if @chars_remaining < 0
+        @output << @omission
+        raise TruncateFinished 
+      end
+    end
+
+    def start_element(name, attrs = [])
+      unless @discard_first_element
+        return if %w(html body).include? name
+        return @discard_first_element = true
+      end
+
+      if %w(br embed hr img input param).include? name
+        @output << "<#{name}#{' ' if attrs.size > 0 }#{attrs.map { |attr,val| "#{attr}=\"#{val}\"" }.join(' ')} />"
+      else
+        @output << "<#{name}#{' ' if attrs.size > 0 }#{attrs.map { |attr,val| "#{attr}=\"#{val}\"" }.join(' ')}>"
+        @tags.push name
+      end
+    end
+
+    def end_element(name)
+      @output << "</#{@tags.pop}>" if @tags.last == name
+    end
+  end
+end

data/lib/nokogiri_truncate_html/truncate_html_helper.rb

@@ -0,0 +1,37 @@
+require 'active_support/core_ext/module/attribute_accessors.rb'
+require 'active_support/core_ext/string/output_safety'
+require 'nokogiri_truncate_html/truncate_document'
+
+module NokogiriTruncateHtml
+  module TruncateHtmlHelper
+    mattr_accessor :parser
+    mattr_accessor :document
+
+    # Truncates html respecting tags and html entities.
+    #
+    # The API is the same as ActionView::Helpers::TextHelper#truncate. It uses Nokogiri and HtmlEntities for entity awareness.
+    #
+    # Examples:
+    #  truncate_html '<p>Hello <strong>World</strong></p>', :length => 7 # => '<p>Hello <strong>W&hellip;</strong></p>'
+    #  truncate_html '<p>Hello &amp; Goodbye</p>', :length => 7          # => '<p>Hello &amp;&hellip;</p>'
+    def truncate_html(input, *args)
+      self.document ||= TruncateDocument.new#(TruncateHtmlHelper.flavor) #, length, omission)
+      self.parser ||= Nokogiri::HTML::SAX::Parser.new(document)
+
+      # support both 2.2 & earlier APIs
+      options = args.extract_options!
+      length = options[:length] || args[0] || 30
+      omission = options[:omission] || args[1] || '&hellip;'
+
+      # Adding div around the input is a hack. It gets removed in TruncateDocument.
+      input = "<div>#{input}</div>"
+      self.document.length = length
+      self.document.omission = omission
+      begin
+        self.parser.parse_memory(input)
+      rescue TruncateFinished
+      end
+      self.document.output.html_safe
+    end
+  end
+end

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: nokogiri_truncate_html
 version: !ruby/object:Gem::Version
-  version: 0.0.1
+  version: 0.0.2
   prerelease: 
 platform: ruby
 authors:
@@ -83,6 +83,8 @@ executables: []
 extensions: []
 extra_rdoc_files: []
 files:
+- lib/nokogiri_truncate_html/truncate_html_helper.rb
+- lib/nokogiri_truncate_html/truncate_document.rb
 - lib/nokogiri_truncate_html.rb
 homepage: https://github.com/Springest/truncate_html
 licenses: []