truncato 0.7.4 → 0.7.5

data/README.md

@@ -9,7 +9,7 @@ In your `Gemfile`
 ```ruby
 gem 'truncato'
 ```
- 
+
 ## Usage
 
 ```ruby
@@ -23,12 +23,15 @@ The configuration options are:
 * `tail`: The string to append when the truncation occurs ('...' by default)
 * `count_tags`: Boolean value indicating whether tags size should be considered when truncating (`true` by default)
 * `filtered_attributes`: Array of attribute names that will be removed from the output. This allows you to make the truncated string shorter by excluding the content of attributes you can discard in some given context, e.g HTML `style` attribute.
+* `tail_before_final_tag`: Boolean value indicating whether to apply a tail before the final closing tag (`false` by default)
+* `comments`: Boolean value indicating whether to include comments in parsed results (`false` by default)
+* `count_tail`: Boolean value indicating whether to include the tail within the bounds of the provided max length (`false` by default)
 
 ## Performance
 
 Truncato was designed with performance in mind. Its main motivation was that existing libs couldn't truncate a multiple-MB document into a few-KB one in a reasonable time. It uses the [Nokogiri](http://nokogiri.org/) SAX parser.
 
-There is a benchmark included that generates a synthetic XML of 4MB and truncates it to 400 KB. You can run the benchmark using 
+There is a benchmark included that generates a synthetic XML of 4MB and truncates it to 400 KB. You can run the benchmark using
 
 ```ruby
 rake truncato:benchmark

data/VERSION

@@ -1 +1 @@
-0.7.4
+0.7.5

data/lib/truncato/truncated_sax_document.rb

@@ -6,13 +6,13 @@ class TruncatedSaxDocument < Nokogiri::XML::SAX::Document
 
   def initialize(options)
     @html_coder = HTMLEntities.new
-    capture_options(options)
+    capture_options options
     init_parsing_state
   end
 
   def start_element name, attributes
     return if @max_length_reached || artificial_root_name?(name)
-    @closing_tags.push name
+    @closing_tags.push name unless single_tag_element? name
     append_to_truncated_string opening_tag(name, attributes), overriden_tag_length
   end
 
@@ -23,10 +23,25 @@ class TruncatedSaxDocument < Nokogiri::XML::SAX::Document
     append_to_truncated_string @html_coder.encode(string_to_append), string_to_append.length
   end
 
+  def comment string
+    if @comments
+      return if @max_length_reached
+      remaining_length = max_length - @estimated_length - 1
+      string_to_append = comment_tag(string).length > remaining_length ? truncate_comment(comment_tag(string), remaining_length) : comment_tag(string)
+      append_to_truncated_string string_to_append
+    end
+  end
+
+  def comment_tag comment
+    "<!--#{comment}-->"
+  end
+
   def end_element name
     return if @max_length_reached || artificial_root_name?(name)
-    @closing_tags.pop
-    append_to_truncated_string closing_tag(name), overriden_tag_length
+    unless single_tag_element? name
+      @closing_tags.pop
+      append_to_truncated_string closing_tag(name), overriden_tag_length
+    end
   end
 
   def end_document
@@ -36,40 +51,55 @@ class TruncatedSaxDocument < Nokogiri::XML::SAX::Document
   private
 
   def capture_options(options)
-    @max_length = options[:max_length]
-    @count_tags = options [:count_tags]
-    @tail = options[:tail]
-    @filtered_attributes = options[:filtered_attributes] || []
+    @max_length            = options[:max_length]
+    @count_tags            = options [:count_tags]
+    @count_tail            = options.fetch(:count_tail, false)
+    @tail                  = options[:tail]
+    @filtered_attributes   = options[:filtered_attributes] || []
+    @tail_before_final_tag = options.fetch(:tail_before_final_tag, false)
+    @comments              = options.fetch(:comments, false)
   end
 
   def init_parsing_state
-    @truncated_string = ""
-    @closing_tags = []
-    @estimated_length = 0
+    @truncated_string   = ""
+    @closing_tags       = []
+    @estimated_length   = @count_tail ? tail_length : 0
     @max_length_reached = false
   end
 
+  def tail_length
+    tail.match(/^&\w+;$/).nil? ? tail.length : 1
+  end
+
+  def single_tag_element? name
+    ["br", "img"].include? name
+  end
+
   def append_to_truncated_string string, overriden_length=nil
     @truncated_string << string
     increase_estimated_length(overriden_length || string.length)
   end
 
   def opening_tag name, attributes
-    attributes_string = attributes_to_string(attributes)
-    "<#{name}#{attributes_string}>"
+    attributes_string = attributes_to_string attributes
+    if single_tag_element? name
+      "<#{name}#{attributes_string} />"
+    else
+      "<#{name}#{attributes_string}>"
+    end
   end
 
-  def attributes_to_string(attributes)
+  def attributes_to_string attributes
     return "" if attributes.empty?
     attributes_string = concatenate_attributes_declaration attributes
     attributes_string.rstrip
   end
 
-  def concatenate_attributes_declaration(attributes)
+  def concatenate_attributes_declaration attributes
     attributes.inject(' ') do |string, attribute|
       key, value = attribute
-      next string if @filtered_attributes.include?(key)
-      string << "#{key}='#{@html_coder.encode(value)}' "
+      next string if @filtered_attributes.include? key
+      string << "#{key}='#{@html_coder.encode value}' "
     end
   end
 
@@ -87,24 +117,42 @@ class TruncatedSaxDocument < Nokogiri::XML::SAX::Document
   end
 
   def truncate_string string, remaining_length
-    @tail_appended = true
-    "#{string[0..remaining_length]}#{tail}"
+    if @tail_before_final_tag
+      string[0..remaining_length]
+    else
+      @tail_appended = true
+      "#{string[0..remaining_length]}#{tail}"
+    end
+  end
+
+  def truncate_comment string, remaining_length
+    if @tail_before_final_tag
+      string[0..remaining_length]
+    else
+      @tail_appended = true
+      "#{string[0..remaining_length]}#{tail}-->"
+    end
   end
 
   def close_truncated_document
+    append_tail_between_closing_tags if @tail_before_final_tag
     append_to_truncated_string tail unless @tail_appended
     append_closing_tags
   end
 
   def append_closing_tags
-    @closing_tags.reverse.each { |name| append_to_truncated_string closing_tag(name) }
+    @closing_tags.reverse.each { |name| append_to_truncated_string closing_tag name }
   end
 
   def overriden_tag_length
     @count_tags ? nil : 0
   end
 
-  def artificial_root_name?(name)
+  def artificial_root_name? name
     name == Truncato::ARTIFICIAL_ROOT_NAME
   end
+
+  def append_tail_between_closing_tags
+    append_to_truncated_string closing_tag(@closing_tags.delete_at (@closing_tags.length - 1)) if @closing_tags.length > 1
+  end
 end

data/lib/truncato/truncato.rb

@@ -26,7 +26,7 @@ module Truncato
   private
 
   def self.truncate_html source, options
-    self.do_truncate_html(source, options) ? self.do_truncate_html(with_articial_root(source), options) : nil
+    self.do_truncate_html(source, options) ? self.do_truncate_html(with_artificial_root(source), options) : nil
   end
 
   def self.do_truncate_html source, options
@@ -37,7 +37,7 @@ module Truncato
     truncated_string.empty? ? nil : truncated_string
   end
 
-  def self.with_articial_root(source)
+  def self.with_artificial_root(source)
     "<#{ARTIFICIAL_ROOT_NAME}>#{source}</#{ARTIFICIAL_ROOT_NAME}>"
   end
 

data/spec/truncato/truncato_spec.rb

@@ -33,6 +33,67 @@ describe "Truncato" do
                        expected: "<div>some text 0</div><div><p>some text 1</p><p>som...</p></div>"
   end
 
+  describe "include tail as part of max_length" do
+    it_should_truncate "html text with a tag (counting tail)", with: {max_length: 4, count_tail: true, count_tags: false},
+                        source: "<p>some text</p>",
+                        expected: "<p>s...</p>"
+
+    it_should_truncate "html text with a tag (counting tail)", with: {max_length: 6, count_tail: true, count_tags: false}, source: "<p>some text</p>", expected: "<p>som...</p>"
+
+    it_should_truncate "html text with a tag (counting tail)", with: {max_length: 16, count_tail: true, count_tags: false},
+                        source: "<p>some text</p><div><span>some other text</span></div>",
+                        expected: "<p>some text</p><div><span>some...</span></div>"
+
+    it_should_truncate "html text with a tag (counting tail and including tail before final tag)", with: {max_length: 16, count_tail: true, count_tags: false, tail_before_final_tag: true},
+                        source: "<p>some text</p><div><span>some other text</span></div>",
+                        expected: "<p>some text</p><div><span>some</span>...</div>"
+
+    it_should_truncate "html text, counting special html characters as one character",
+                        with: {max_length: 16, count_tail: true, count_tags: false, tail_before_final_tag: true, tail: '&hellip;'},
+                        source: "<p>some text</p><div><span>some other text</span></div>",
+                        expected: "<p>some text</p><div><span>some o</span>&hellip;</div>"
+  end
+
+  describe "insert tail between two or more final tags" do
+    it_should_truncate "html text as normal when tail_before_final_tag option is not set",
+                        with: {max_length: 4, count_tags: false},
+                        source: "<p><span>some text</span>some more text</p>",
+                        expected: "<p><span>some...</span></p>"
+
+    it_should_truncate "html text when tail_before_final_tag: true by inserting tail before the final tag, and after any other closing tags",
+                        with: {max_length: 4, count_tags: false, tail_before_final_tag: true},
+                        source: "<p><span>some text</span>some more text</p>",
+                        expected: "<p><span>some</span>...</p>"
+  end
+
+  describe "single html tag elements" do
+    it_should_truncate "html text with <br /> element without adding a closing tag", with: {max_length: 9},
+                       source: "<div><p><br />some text 1</p><p>some text 2</p></div>",
+                       expected: "<div><p><br />...</p></div>"
+
+    it_should_truncate "html text with <img /> element without adding a closing tag", with: {max_length: 9},
+                       source: "<div><p><img src='some_path' />some text 1</p><p>some text 2</p></div>",
+                       expected: "<div><p><img src='some_path' />...</p></div>"
+  end
+
+  describe "comment html element" do
+    it_should_truncate "html text and ignore <!-- a comment --> element by default", with: {max_length: 20},
+                       source: "<!-- a comment --><p>some text 1</p>",
+                       expected: "<p>some text 1</p>"
+
+    it_should_truncate "html text with <!-- a comment --> element", with: {max_length: 30, comments: true},
+                       source: "<!-- a comment --><p>some text 1</p>",
+                       expected: "<!-- a comment --><p>some text...</p>"
+
+    it_should_truncate "html text with <!-- a comment --> element that exceeds the max_length", with: {max_length: 5, comments: true},
+                       source: "<!-- a comment --><p>some text 1</p>",
+                       expected: "<!-- ...-->"
+
+    it_should_truncate "html text with <!-- a comment --> element with other elements that exceeds max_length", with: {max_length: 20, comments: true},
+                       source: "<!-- a comment --><p>some text 1</p>",
+                       expected: "<!-- a comment --><p>...</p>"
+  end
+
   describe "html attributes" do
     it_should_truncate "html text with 1 attributes", with: {max_length: 3, count_tags: false},
                        source: "<p attr1='1'>some text</p>",
@@ -54,15 +115,13 @@ describe "Truncato" do
                        source: "<p attr1='&gt;some'>text</p>",
                        expected: "<p attr1='&gt;some'>tex...</p>"
 
-    it_should_truncate "html text with 2 attributes filtering one of them", with: {max_length: 30, count_tags: false, filtered_attributes: ['attr2']},
-                       source: "<p attr1='1'>some text</p>",
-                       expected: "<p attr1='1'>some text</p>"
+    it_should_truncate "html text with 2 attributes filtering one of them", with: {max_length: 90, count_tags: false, filtered_attributes: ['attr2']},
+                       source: "<p attr1='1'>some text</p><p attr2='2'>filtered text</p>",
+                       expected: "<p attr1='1'>some text</p><p>filtered text</p>"
 
     it_should_truncate "html text with 2 attributes filtering all of them", with: {max_length: 3, count_tags: false, filtered_attributes: ['attr1', 'attr2']},
                        source: "<p attr1='1' attr2='2'>some text</p>",
                        expected: "<p>som...</p>"
   end
 
-
 end
-

data/truncato.gemspec

@@ -5,11 +5,11 @@
 
 Gem::Specification.new do |s|
   s.name = "truncato"
-  s.version = "0.7.4"
+  s.version = "0.7.5"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["Jorge Manrubia"]
-  s.date = "2013-04-14"
+  s.date = "2013-04-27"
   s.description = "Ruby tool for truncating HTML strings keeping a valid HTML markup"
   s.email = "jorge.manrubia@gmail.com"
   s.extra_rdoc_files = [
@@ -52,6 +52,8 @@ Gem::Specification.new do |s|
       s.add_development_dependency(%q<bundler>, ["~> 1.3"])
       s.add_development_dependency(%q<bundler>, ["~> 1.3"])
       s.add_development_dependency(%q<jeweler>, ["~> 1.8.4"])
+      s.add_development_dependency(%q<bundler>, ["~> 1.3"])
+      s.add_development_dependency(%q<jeweler>, ["~> 1.8.4"])
     else
       s.add_dependency(%q<truncato>, [">= 0"])
       s.add_dependency(%q<nokogiri>, ["~> 1.5.5"])
@@ -59,6 +61,8 @@ Gem::Specification.new do |s|
       s.add_dependency(%q<bundler>, ["~> 1.3"])
       s.add_dependency(%q<bundler>, ["~> 1.3"])
       s.add_dependency(%q<jeweler>, ["~> 1.8.4"])
+      s.add_dependency(%q<bundler>, ["~> 1.3"])
+      s.add_dependency(%q<jeweler>, ["~> 1.8.4"])
     end
   else
     s.add_dependency(%q<truncato>, [">= 0"])
@@ -67,6 +71,8 @@ Gem::Specification.new do |s|
     s.add_dependency(%q<bundler>, ["~> 1.3"])
     s.add_dependency(%q<bundler>, ["~> 1.3"])
     s.add_dependency(%q<jeweler>, ["~> 1.8.4"])
+    s.add_dependency(%q<bundler>, ["~> 1.3"])
+    s.add_dependency(%q<jeweler>, ["~> 1.8.4"])
   end
 end
 

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification
 name: truncato
 version: !ruby/object:Gem::Version
-  version: 0.7.4
+  version: 0.7.5
   prerelease: 
 platform: ruby
 authors:
@@ -9,7 +9,7 @@ authors:
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2013-04-14 00:00:00.000000000 Z
+date: 2013-04-27 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: truncato
@@ -107,6 +107,38 @@ dependencies:
     - - ~>
       - !ruby/object:Gem::Version
         version: 1.8.4
+- !ruby/object:Gem::Dependency
+  name: bundler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.3'
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: '1.3'
+- !ruby/object:Gem::Dependency
+  name: jeweler
+  requirement: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.8.4
+  type: :development
+  prerelease: false
+  version_requirements: !ruby/object:Gem::Requirement
+    none: false
+    requirements:
+    - - ~>
+      - !ruby/object:Gem::Version
+        version: 1.8.4
 description: Ruby tool for truncating HTML strings keeping a valid HTML markup
 email: jorge.manrubia@gmail.com
 executables: []
@@ -148,7 +180,7 @@ required_ruby_version: !ruby/object:Gem::Requirement
       version: '0'
       segments:
       - 0
-      hash: 1642571317414252409
+      hash: 3957383429942337949
 required_rubygems_version: !ruby/object:Gem::Requirement
   none: false
   requirements: