truncate_html 0.5.1 → 0.9.3

checksums.yaml

@@ -0,0 +1,7 @@
+---
+SHA1:
+  metadata.gz: 8b62f471b431ebc1482936c69165f634440e9f97
+  data.tar.gz: 55123d5396db552d9a0ab64b834de98afbe17688
+SHA512:
+  metadata.gz: 9d0fcc597a3e4e8f39a08bf5e0cfda14cf692dc6753727a2812ffa9338e4567ecdbd51a7751f2a8649aca6a8fa8335c8c3609313673e9156225d7df9d5472590
+  data.tar.gz: 78b8a0c36a8bf3c12390d7606ee2a4561006787dd527c166fcdd9c8104c1cd7f23680e6cb337e6b0b9d47bfa680932666c87a39400c07374371381d2d1729e79

data/.gitignore

@@ -0,0 +1,7 @@
+pkg
+coverage
+profiling
+tmp
+spec/rails_root/log/*
+log/*.log
+.bundle

data/.travis.yml

@@ -0,0 +1,6 @@
+language: ruby
+rvm:
+  - 1.9.2
+  - 1.9.3
+  - 2.0.0
+  - jruby-19mode # JRuby in 1.9 mode

data/Gemfile

@@ -1,7 +1,4 @@
-# Edit this Gemfile to bundle your application's dependencies.
-source 'http://gemcutter.org'
-group :development, :test do
-  gem 'jeweler'
-  gem "rails", "= 3.0.3"
-  gem 'rspec-rails', '= 2.3'
-end
+source "http://rubygems.org"
+
+# Specify your gem's dependencies in truncate_html.gemspec
+gemspec

data/Gemfile.lock

@@ -1,92 +1,105 @@
+PATH
+  remote: .
+  specs:
+    truncate_html (0.9.3)
+
 GEM
-  remote: http://gemcutter.org/
+  remote: http://rubygems.org/
   specs:
-    abstract (1.0.0)
-    actionmailer (3.0.3)
-      actionpack (= 3.0.3)
-      mail (~> 2.2.9)
-    actionpack (3.0.3)
-      activemodel (= 3.0.3)
-      activesupport (= 3.0.3)
-      builder (~> 2.1.2)
-      erubis (~> 2.6.6)
-      i18n (~> 0.4)
-      rack (~> 1.2.1)
-      rack-mount (~> 0.6.13)
-      rack-test (~> 0.5.6)
-      tzinfo (~> 0.3.23)
-    activemodel (3.0.3)
-      activesupport (= 3.0.3)
-      builder (~> 2.1.2)
-      i18n (~> 0.4)
-    activerecord (3.0.3)
-      activemodel (= 3.0.3)
-      activesupport (= 3.0.3)
-      arel (~> 2.0.2)
-      tzinfo (~> 0.3.23)
-    activeresource (3.0.3)
-      activemodel (= 3.0.3)
-      activesupport (= 3.0.3)
-    activesupport (3.0.3)
-    arel (2.0.7)
-    builder (2.1.2)
-    diff-lcs (1.1.2)
-    erubis (2.6.6)
-      abstract (>= 1.0.0)
-    git (1.2.5)
-    i18n (0.5.0)
-    jeweler (1.5.2)
-      bundler (~> 1.0.0)
-      git (>= 1.2.5)
-      rake
-    mail (2.2.15)
-      activesupport (>= 2.3.6)
+    actionmailer (3.2.13)
+      actionpack (= 3.2.13)
+      mail (~> 2.5.3)
+    actionpack (3.2.13)
+      activemodel (= 3.2.13)
+      activesupport (= 3.2.13)
+      builder (~> 3.0.0)
+      erubis (~> 2.7.0)
+      journey (~> 1.0.4)
+      rack (~> 1.4.5)
+      rack-cache (~> 1.2)
+      rack-test (~> 0.6.1)
+      sprockets (~> 2.2.1)
+    activemodel (3.2.13)
+      activesupport (= 3.2.13)
+      builder (~> 3.0.0)
+    activerecord (3.2.13)
+      activemodel (= 3.2.13)
+      activesupport (= 3.2.13)
+      arel (~> 3.0.2)
+      tzinfo (~> 0.3.29)
+    activeresource (3.2.13)
+      activemodel (= 3.2.13)
+      activesupport (= 3.2.13)
+    activesupport (3.2.13)
+      i18n (= 0.6.1)
+      multi_json (~> 1.0)
+    arel (3.0.2)
+    builder (3.0.4)
+    diff-lcs (1.2.2)
+    erubis (2.7.0)
+    hike (1.2.1)
+    i18n (0.6.1)
+    journey (1.0.4)
+    json (1.7.7)
+    mail (2.5.3)
       i18n (>= 0.4.0)
       mime-types (~> 1.16)
       treetop (~> 1.4.8)
-    mime-types (1.16)
-    polyglot (0.3.1)
-    rack (1.2.1)
-    rack-mount (0.6.13)
-      rack (>= 1.0.0)
-    rack-test (0.5.7)
+    mime-types (1.22)
+    multi_json (1.7.2)
+    polyglot (0.3.3)
+    rack (1.4.5)
+    rack-cache (1.2)
+      rack (>= 0.4)
+    rack-ssl (1.3.3)
+      rack
+    rack-test (0.6.2)
       rack (>= 1.0)
-    rails (3.0.3)
-      actionmailer (= 3.0.3)
-      actionpack (= 3.0.3)
-      activerecord (= 3.0.3)
-      activeresource (= 3.0.3)
-      activesupport (= 3.0.3)
+    rails (3.2.13)
+      actionmailer (= 3.2.13)
+      actionpack (= 3.2.13)
+      activerecord (= 3.2.13)
+      activeresource (= 3.2.13)
+      activesupport (= 3.2.13)
       bundler (~> 1.0)
-      railties (= 3.0.3)
-    railties (3.0.3)
-      actionpack (= 3.0.3)
-      activesupport (= 3.0.3)
+      railties (= 3.2.13)
+    railties (3.2.13)
+      actionpack (= 3.2.13)
+      activesupport (= 3.2.13)
+      rack-ssl (~> 1.3.2)
       rake (>= 0.8.7)
-      thor (~> 0.14.4)
-    rake (0.8.7)
-    rspec (2.3.0)
-      rspec-core (~> 2.3.0)
-      rspec-expectations (~> 2.3.0)
-      rspec-mocks (~> 2.3.0)
-    rspec-core (2.3.1)
-    rspec-expectations (2.3.0)
-      diff-lcs (~> 1.1.2)
-    rspec-mocks (2.3.0)
-    rspec-rails (2.3.0)
-      actionpack (~> 3.0)
-      activesupport (~> 3.0)
-      railties (~> 3.0)
-      rspec (~> 2.3.0)
-    thor (0.14.6)
-    treetop (1.4.9)
+      rdoc (~> 3.4)
+      thor (>= 0.14.6, < 2.0)
+    rake (10.0.4)
+    rdoc (3.12.2)
+      json (~> 1.4)
+    rspec-core (2.13.1)
+    rspec-expectations (2.13.0)
+      diff-lcs (>= 1.1.3, < 2.0)
+    rspec-mocks (2.13.0)
+    rspec-rails (2.13.0)
+      actionpack (>= 3.0)
+      activesupport (>= 3.0)
+      railties (>= 3.0)
+      rspec-core (~> 2.13.0)
+      rspec-expectations (~> 2.13.0)
+      rspec-mocks (~> 2.13.0)
+    sprockets (2.2.2)
+      hike (~> 1.2)
+      multi_json (~> 1.0)
+      rack (~> 1.0)
+      tilt (~> 1.1, != 1.3.0)
+    thor (0.18.1)
+    tilt (1.3.6)
+    treetop (1.4.12)
+      polyglot
       polyglot (>= 0.3.1)
-    tzinfo (0.3.24)
+    tzinfo (0.3.37)
 
 PLATFORMS
   ruby
 
 DEPENDENCIES
-  jeweler
-  rails (= 3.0.3)
-  rspec-rails (= 2.3)
+  rails (~> 3.2.13)
+  rspec-rails (~> 2.13)
+  truncate_html!

data/README.md

@@ -0,0 +1,97 @@
+TruncateHtml
+============
+
+[![Build Status](https://secure.travis-ci.org/hgmnz/truncate_html.png?branch=master)](http://travis-ci.org/hgmnz/truncate_html)
+[![Code Climate](https://codeclimate.com/github/hgmnz/truncate_html.png)](https://codeclimate.com/github/hgmnz/truncate_html)
+
+truncate_html cuts off a string of HTML and takes care of closing any lingering open tags. There are many ways to solve this. This library does not have any dependencies, and [parses HTML using regular expressions](http://stackoverflow.com/questions/1732348/regex-match-open-tags-except-xhtml-self-contained-tags/1732454#1732454).
+
+It can be used with or without Rails.
+
+Example
+-------
+
+```ruby
+some_html = '<ul><li><a href="http://whatever">This is a link</a></li></ul>'
+truncate_html(some_html, length: 15, omission: '...(continued)')
+  => <ul><li><a href="http://whatever">This...(continued)</a></li></ul>
+```
+
+A few notes:
+
+* By default, it will truncate on word boundary.
+  To truncate the HTML string strictly at the specified length, pass in the `word_boundary: false` option.
+* If the input HTML is nil, it will return an empty string.
+* The omission text's length does count toward the resulting string's length.
+* `<script>` tags will pass right through - they will not count toward the resulting string's length, or be truncated.
+
+* The default options are:
+```
+length: 100
+omission: '...'
+word_boundary: /\S/
+```
+
+You may also set global configuration options.
+For example, place the following on application boot,
+something like `config/initializers/truncate_html.rb`
+
+```ruby
+TruncateHtml.configure do |config|
+  config.length        = 50
+  config.omission      = '...(continued)'
+end
+```
+
+If you really want, you can even set a custom word boundary regexp.
+For example, to truncate at the end of the nearest sentence:
+
+```ruby
+TruncateHtml.configure do |config|
+  config.word_boundary = /\S[\.\?\!]/
+end
+```
+
+You can also truncate the HTML at a specific point not based on length but content.
+To do that, place the `:break_token` in your source. This allows the truncation to be
+data driven, breaking after a leading paragraph or sentence. If the
+`:break_token` is in your content before the specified `:length`, `:length` will be
+ignored and the content truncated at `:break_token`.
+If the `:break_token` is in your content after the specified `:length`,
+`:break_token` will be ignored and the content truncated at `:length`.
+
+```ruby
+TruncateHtml.configure do |config|
+  config.break_token = '<!-- truncate -->'
+end
+```
+Installation
+------------
+
+The latest gem version for the Rails 2.x series is 0.3.2.
+To use truncate_html on a Rails 2 app, please install the 0.3.2 version:
+
+    gem install truncate_html -v 0.3.2
+
+For Rails 3, use the latest truncate_html:
+
+    gem install truncate_html
+
+Issues or Suggestions
+---------------------
+
+Found an issue or have a suggestion? Please report it on [Github's issue tracker](http://github.com/hgmnz/truncate_html/issues).
+
+Testing
+-------
+
+```shell
+bundle
+rake
+```
+
+All green? Go hack.
+
+Copyright (c) 2009 - 2010 Harold A. Giménez, released under the MIT license
+
+Thanks to all the [contributors](https://github.com/hgmnz/truncate_html/contributors)!

data/Rakefile

@@ -34,36 +34,17 @@ RSpec::Core::RakeTask.new(:spec) do |t|
   t.pattern   = "./spec/**/*_spec.rb"
 end
 
-begin
-  require 'metric_fu'
-  MetricFu::Configuration.run do |config|
-    config.metrics  = [:saikuro, :flog, :flay, :reek, :roodi, :rcov]
-    config.graphs   = [:flog, :flay, :reek, :roodi, :rcov]
-    config.rcov     = { :environment => 'test',
-                        :test_files => ['spec/**/*_spec.rb'],
-                        :rcov_opts => ["--sort coverage",
-                                       "--text-coverage",
-                                       "--profile",
-                                       "--exclude /gems/,/Library/"]}
-    config.graph_engine = :bluff
-  end
-rescue LoadError
-  puts "Install metric_fu to run code metrics"
-end
-
-begin
-  require 'jeweler'
-  Jeweler::Tasks.new do |gem|
-    gem.name = "truncate_html"
-    gem.summary = %Q{Uses an API similar to Rails' truncate helper to truncate HTML and close any lingering open tags.}
-    gem.description = %Q{Truncates html so you don't have to}
-    gem.email = "harold.gimenez@gmail.com"
-    gem.homepage = "http://github.com/hgimenez/truncate_html"
-    gem.authors = ["Harold A. Gimenez"]
-    # gem is a Gem::Specification... see http://www.rubygems.org/read/chapter/20 for additional settings
+namespace :gem do
+  desc "Release to rubygems"
+  task :release do
+    require File.expand_path('lib/truncate_html/version', File.dirname(__FILE__))
+    version = TruncateHtml::VERSION
+    message = "Bump to version #{version}"
+    system "git tag -a -m '#{message}' v#{version}"
+    system "git push origin master"
+    system "git push origin $(git tag | tail -1)"
+    system "gem build truncate_html.gemspec"
+    system "gem push truncate_html-#{version}.gem"
+    system "rm truncate_html-#{version}.gem"
   end
-  Jeweler::GemcutterTasks.new
-
-rescue LoadError
-  puts "Jeweler (or a dependency) not available. Install it with: sudo gem install jeweler"
 end

data/VERSION

@@ -1 +1 @@
-0.5.1
+0.9.2

data/lib/app/helpers/truncate_html_helper.rb

@@ -3,7 +3,7 @@ module TruncateHtmlHelper
   def truncate_html(html, options={})
     return '' if html.nil?
     html_string = TruncateHtml::HtmlString.new(html)
-    TruncateHtml::HtmlTruncator.new(html_string).truncate(options).html_safe
+    TruncateHtml::HtmlTruncator.new(html_string, options).truncate.html_safe
   end
 
 end

data/lib/truncate_html/configuration.rb

@@ -1,6 +1,6 @@
 module TruncateHtml
   class Configuration
-    attr_accessor :length, :omission, :word_boundary
+    attr_accessor :length, :omission, :word_boundary, :break_token
   end
 
   class << self
@@ -11,5 +11,4 @@ module TruncateHtml
     self.configuration ||= Configuration.new
     yield configuration
   end
-
 end

data/lib/truncate_html/html_string.rb

@@ -1,39 +1,40 @@
+# encoding: utf-8
 module TruncateHtml
   class HtmlString < String
 
-    UNPAIRED_TAGS = %w(br hr img)
+    UNPAIRED_TAGS = %w(br hr img).freeze
+    REGEX = /(?:<script.*>.*<\/script>)+|<\/?[^>]+>|[[[:alpha:]][0-9]\|`~!@#\$%^&*\(\)\-_\+=\[\]{}:;'²³§",\.\/?]+|\s+|[[:punct:]]/.freeze
 
     def initialize(original_html)
       super(original_html)
     end
 
     def html_tokens
-      scan(regex).map do
-        |token| token.gsub(
-          #remove newline characters
-            /\n/,''
-        ).gsub(
-          #clean out extra consecutive whitespace
-            /\s+/, ' '
+      scan(REGEX).map do |token|
+        HtmlString.new(
+          token.gsub(
+            /\n/,' ' #replace newline characters with a whitespace
+          ).gsub(
+            /\s+/, ' ' #clean out extra consecutive whitespace
+          )
         )
-      end.map { |token| HtmlString.new(token) }
+      end
     end
 
     def html_tag?
-      self =~ /<\/?[^>]+>/ ? true : false
+      /<\/?[^>]+>/ === self && !html_comment?
     end
 
     def open_tag?
-      self =~ /<(?!(?:#{UNPAIRED_TAGS.join('|')}|script|\/))[^>]+>/i ? true : false
+      /<(?!(?:#{UNPAIRED_TAGS.join('|')}|script|\/))[^>]+>/i === self
     end
 
-    def matching_close_tag
-      gsub(/<(\w+)\s?.*>/, '</\1>').strip
+    def html_comment?
+      /<\s?!--.*-->/ === self
     end
 
-    private
-    def regex
-      /(?:<script.*>.*<\/script>)+|<\/?[^>]+>|[#{"[[:alpha:]]" if RUBY_VERSION >= '1.9'}\w\|`~!@#\$%^&*\(\)\-_\+=\[\]{}:;'",\.\/?]+|\s+/
+    def matching_close_tag
+      gsub(/<(\w+)\s?.*>/, '</\1>').strip
     end
 
   end

data/lib/truncate_html/html_truncator.rb

@@ -1,67 +1,97 @@
 module TruncateHtml
   class HtmlTruncator
 
-    def initialize(original_html)
-      @original_html  = original_html
-    end
-
-    def truncate(options = {})
-      length        = options[:length]       || TruncateHtml.configuration.length
-      @omission     = options[:omission]     || TruncateHtml.configuration.omission
-      @word_boundary = (options.has_key?(:word_boundary) ? options[:word_boundary] : TruncateHtml.configuration.word_boundary)
+    def initialize(original_html, options = {})
+      @original_html   = original_html
+      length           = options[:length]       || TruncateHtml.configuration.length
+      @omission        = options[:omission]     || TruncateHtml.configuration.omission
+      @word_boundary   = (options.has_key?(:word_boundary) ? options[:word_boundary] : TruncateHtml.configuration.word_boundary)
+      @break_token     = options[:break_token] || TruncateHtml.configuration.break_token || nil
       @chars_remaining = length - @omission.length
-      @open_tags, @truncated_html = [], ['']
+      @open_tags, @closing_tags, @truncated_html = [], [], ['']
+    end
 
+    def truncate
+      return @omission if @chars_remaining < 0
       @original_html.html_tokens.each do |token|
-        #if truncate_more?(token)
-        if @chars_remaining > 0
-          process_token(token)
-        else
+        if @chars_remaining <= 0 || truncate_token?(token)
           close_open_tags
           break
+        else
+          process_token(token)
         end
       end
-      @truncated_html.join
+
+      build_output
     end
 
     private
 
-      def process_token(token)
-        append_to_result(token)
-        if token.html_tag?
-          if token.open_tag?
-            @open_tags << token
-          else
-            remove_latest_open_tag(token)
-          end
-        else
-          @chars_remaining -= (@word_boundary ? token.length : token[0, @chars_remaining].length)
+    def word_boundary
+      if @word_boundary == true
+        TruncateHtml.configuration.word_boundary
+      else
+        @word_boundary
+      end
+    end
+
+    def build_output
+      out = @truncated_html.join
+      if word_boundary
+        term_regexp = Regexp.new("^.*#{word_boundary.source}")
+        match = out.match(term_regexp)
+        if match && match[0] != out
+          out = match[0] + @closing_tags.join
         end
       end
 
-      def append_to_result(token)
-        if @word_boundary
-          @truncated_html << token if token.html_tag? || @chars_remaining - token.length >= 0  
+      out
+    end
+
+    def process_token(token)
+      append_to_result(token)
+      if token.html_tag?
+        if token.open_tag?
+          @open_tags << token
         else
-          @truncated_html << token[0, @chars_remaining]
+          remove_latest_open_tag(token)
+        end
+      elsif !token.html_comment?
+        @chars_remaining -= (@word_boundary ? token.length : token[0, @chars_remaining].length)
+        if @chars_remaining <= 0
+          @truncated_html[-1] = @truncated_html[-1].rstrip + @omission
         end
       end
+    end
 
-      def close_open_tags
-        @truncated_html[-1] = @truncated_html[-1].rstrip + @omission
-        @open_tags.reverse_each do |open_tag|
-          @truncated_html << open_tag.matching_close_tag
-        end
+    def append_to_result(token)
+      if token.html_tag? || token.html_comment?
+        @truncated_html << token
+      elsif @word_boundary
+        @truncated_html << token if (@chars_remaining - token.length) >= 0
+      else
+        @truncated_html << token[0, @chars_remaining]
+      end
+    end
+
+    def close_open_tags
+      @open_tags.reverse_each do |open_tag|
+        @closing_tags << open_tag.matching_close_tag
+        @truncated_html << open_tag.matching_close_tag
       end
+    end
 
-      def remove_latest_open_tag(close_tag)
-        (0...@open_tags.length).to_a.reverse.each do |index|
-          if @open_tags[index].matching_close_tag == close_tag
-            @open_tags.delete_at(index)
-            break
-          end
+    def remove_latest_open_tag(close_tag)
+      (0...@open_tags.length).to_a.reverse.each do |index|
+        if @open_tags[index].matching_close_tag == close_tag
+          @open_tags.delete_at(index)
+          break
         end
       end
+    end
 
+    def truncate_token?(token)
+      @break_token and token == @break_token
+    end
   end
 end

data/lib/truncate_html/version.rb

@@ -0,0 +1,3 @@
+module TruncateHtml
+  VERSION = "0.9.3"
+end

data/lib/truncate_html.rb

@@ -1,12 +1,14 @@
+require File.join(File.dirname(__FILE__), 'truncate_html', 'version')
 require File.join(File.dirname(__FILE__), 'truncate_html', 'html_truncator')
 require File.join(File.dirname(__FILE__), 'truncate_html', 'html_string')
 require File.join(File.dirname(__FILE__), 'truncate_html', 'configuration')
 require File.join(File.dirname(__FILE__), 'app', 'helpers', 'truncate_html_helper')
 
 TruncateHtml.configure do |config|
-  config.length       = 100
-  config.omission     = '...'
-  config.word_boundary = true
+  config.length        = 100
+  config.omission      = '...'
+  config.word_boundary = /\S/
 end
 
-ActionController::Base.helper(TruncateHtmlHelper)
+
+ActionController::Base.helper(TruncateHtmlHelper) if defined?(ActionController)

data/spec/helpers/truncate_html_helper_spec.rb

@@ -17,20 +17,20 @@ describe TruncateHtmlHelper do
   end
 
   before(:each) do
-    @html_truncator_mock = mock(TruncateHtml::HtmlTruncator)
     @original_html = '<p>foo</p>'
     @original_html.stub!(:html_safe).and_return(@original_html)
   end
 
-  it 'creates an instance of HtmlTruncator and calls truncate() on it' do
-    @html_truncator_mock.stub!(:truncate).and_return(@original_html)
-    TruncateHtml::HtmlTruncator.should_receive(:new).and_return(@html_truncator_mock)
+  it 'creates an instance of HtmlTruncator and calls truncate on it' do
+    truncator = double(truncate: @original_html)
+    TruncateHtml::HtmlTruncator.should_receive(:new).and_return(truncator)
     truncator.truncate_html(@original_html)
   end
 
-  it 'calls truncate() on the HtmlTruncator object' do
-    TruncateHtml::HtmlTruncator.stub!(:new).and_return(@html_truncator_mock)
-    @html_truncator_mock.should_receive(:truncate).with({}).once.and_return(@original_html)
+  it 'calls truncate on the HtmlTruncator object' do
+    truncator = double(truncate: @original_html)
+    TruncateHtml::HtmlTruncator.stub!(:new).and_return(truncator)
+    truncator.should_receive(:truncate).and_return(@original_html)
     truncator.truncate_html('foo')
   end
 

data/spec/rails_root/Gemfile

@@ -1,6 +1,6 @@
 # Edit this Gemfile to bundle your application's dependencies.
 source 'http://gemcutter.org'
 group :development, :test do
-  gem "rails", "= 3.0.3"
-  gem 'rspec-rails', '= 2.3'
+  gem "rails", "~> 3.2.13"
+  gem 'rspec-rails', '~> 2.13'
 end