htmlbeautifier 0.0.10 → 0.0.11

checksums.yaml

@@ -1,7 +1,7 @@
 ---
 SHA1:
-  metadata.gz: ef0b7742f4aeee06c65142a6c5cfdd735b462f94
-  data.tar.gz: 59a6d1b39bdc3ccf32ce4f10aa673e448802ec79
+  metadata.gz: 6f0dba839e3aa0da9272b5789530d578eb210cb4
+  data.tar.gz: 156a6912f4c8980d3ca06c5e43865c535af0238e
 SHA512:
-  metadata.gz: 2133a2a77b572ca42386e4a3c3bc7d08c292962d4fe4869de237ca0a120b92fc6e2614bf927fde5ea3533864c68af756f60dcb5d3cd76067677e8593b8ad1702
-  data.tar.gz: cb1d72451e92bf6bd2d0a7bb74e34956c35d5b83e9f9d2e4f39e1d1d9bb75ee483f97956d2d8b2c0c39c8f41e776e1d11fd3f22f6d4c5bad6f6fc090d269c28d
+  metadata.gz: 2a322f76309db244d0eb0d5b4860b3a6681efd0f8c6219b039baeb1e756c78df756995d6975574a4db7b9a559ff92fa6f87c5dcc6fef59c1dfbdcc5cbc4a1321
+  data.tar.gz: 8fe4f9811c1789f532a27a73862e0ad6b9495a96172e14ddc87cab6dcb47de7c722bd2a0252e49ae2ef540233e4c0281cf052340927831cd0c1480289af74be2

data/bin/htmlbeautifier

@@ -1,23 +1,39 @@
 #!/usr/bin/env ruby
 require 'htmlbeautifier'
+require 'optparse'
 require 'fileutils'
 
-def beautify(name, input, output)
-  HtmlBeautifier::Beautifier.new(output).scan(input)
+def beautify(name, input, output, options)
+  beautifier = HtmlBeautifier::Beautifier.new(output)
+  beautifier.tab_stops = options[:tab_stops]
+  beautifier.scan(input)
   output << "\n"
 rescue => e
   raise "Error parsing #{name}: #{e}"
 end
 
+options = {:tab_stops => 2}
+parser = OptionParser.new do |opts|
+  opts.banner = "Usage: #{__FILE__} [options] [file ...]"
+  opts.on("-t", "--tab-stops NUMBER", Integer, "Set number of tab stops") do |num|
+    options[:tab_stops] = num
+  end
+  opts.on("-h", "--help", "Display this help message and exit") do
+    puts opts
+    exit
+  end
+end
+parser.parse!
+
 if ARGV.any?
   ARGV.each do |path|
     input = File.read(path)
     temppath = path + ".tmp"
     File.open(temppath, "w") do |output|
-      beautify path, input, output
+      beautify path, input, output, options
     end
     FileUtils.mv temppath, path
   end
 else
-  beautify "standard input", $stdin.read, $stdout
+  beautify "standard input", $stdin.read, $stdout, options
 end

data/lib/htmlbeautifier/builder.rb

@@ -14,7 +14,7 @@ module HtmlBeautifier
 
     def initialize(output, tab_stops)
       @level = 0
-      @new_line = true
+      @new_line = false
       @tab = ' ' * tab_stops
       @output = output
       @ie_cc_levels = []
@@ -30,15 +30,14 @@ module HtmlBeautifier
     end
 
     def emit(s)
-      if (@new_line)
-        @output << (@tab * @level)
+      if @new_line && !@output.empty?
+        @output << ("\n" + @tab * @level)
       end
       @output << s
       @new_line = false
     end
 
-    def whitespace(*x)
-      emit "\n"
+    def new_line(*_args)
       @new_line = true
     end
 
@@ -59,10 +58,10 @@ module HtmlBeautifier
         lines.pop while lines.last.strip.empty?
         indentation = lines.first[/^ +/]
 
-        whitespace
+        new_line
         lines.each do |line|
           emit line.rstrip.sub(/^#{indentation}/, '')
-          whitespace
+          new_line
         end
 
         outdent
@@ -70,10 +69,28 @@ module HtmlBeautifier
       emit closing
     end
 
+    def preformatted_block(opening, content, closing)
+      emit opening
+      emit content
+      emit closing
+    end
+
     def standalone_element(e)
       emit e
     end
 
+    def close_block_element(e)
+      outdent
+      emit e
+      new_line
+    end
+
+    def open_block_element(e)
+      new_line
+      emit e
+      indent
+    end
+
     def close_element(e)
       outdent
       emit e
@@ -97,8 +114,8 @@ module HtmlBeautifier
     end
 
     def text(t)
-      emit(t.strip)
-      whitespace if t =~ /\s$/
+      emit t.chomp
+      new_line if t.end_with? $/
     end
   end
 end

data/lib/htmlbeautifier/html_parser.rb

@@ -7,6 +7,12 @@ module HtmlBeautifier
       area | base | br | col | command | embed | hr | img | input | keygen |
       link | meta | param | source | track | wbr
     )}mix
+    HTML_BLOCK_ELEMENTS = %r{(?:
+      address | blockquote | center | dd | dir | div | dl | dt | fieldset |
+      form | h1 | h2 | h3 | h4 | h5 | h6 | hr | isindex | li | menu |
+      noframes | noscript | ol | p | pre | table | tbody | td | tfoot | th |
+      thead | tr | ul
+    )}mix
 
     def initialize
       super do |p|
@@ -20,20 +26,26 @@ module HtmlBeautifier
           :standalone_element
         p.map %r{<!.*?>}m,
           :standalone_element
-        p.map %r{(<script#{ELEMENT_CONTENT}>)(.*?)(</script>)}m,
+        p.map %r{(<script#{ELEMENT_CONTENT}>)(.*?)(</script>)}mi,
           :foreign_block
-        p.map %r{(<style#{ELEMENT_CONTENT}>)(.*?)(</style>)}m,
+        p.map %r{(<style#{ELEMENT_CONTENT}>)(.*?)(</style>)}mi,
           :foreign_block
+        p.map %r{(<pre#{ELEMENT_CONTENT}>)(.*?)(</pre>)}mi,
+          :preformatted_block
         p.map %r{<#{ELEMENT_CONTENT}/>}m,
           :standalone_element
         p.map %r{<#{HTML_VOID_ELEMENTS}(?: #{ELEMENT_CONTENT})?>}m,
           :standalone_element
+        p.map %r{</#{HTML_BLOCK_ELEMENTS}>}m,
+          :close_block_element
+        p.map %r{<#{HTML_BLOCK_ELEMENTS}(?: #{ELEMENT_CONTENT})?>}m,
+          :open_block_element
         p.map %r{</#{ELEMENT_CONTENT}>}m,
           :close_element
         p.map %r{<#{ELEMENT_CONTENT}>}m,
           :open_element
-        p.map %r{\s+},
-          :whitespace
+        p.map %r{\s*\r?\n\s*}m,
+          :new_line
         p.map %r{[^<]+},
           :text
       end

data/lib/htmlbeautifier/version.rb

@@ -2,7 +2,7 @@ module HtmlBeautifier #:nodoc:
   module VERSION #:nodoc:
     MAJOR = 0
     MINOR = 0
-    TINY  = 10
+    TINY  = 11
 
     STRING = [MAJOR, MINOR, TINY].join('.')
   end

data/test/test_html_beautifier_integration.rb

@@ -26,8 +26,7 @@ class TestHtmlBeautifierIntegration < Test::Unit::TestCase
       Heading 1
       </h1>
       </div>
-      <div id="somethingElse">
-      <p>Lorem Ipsum</p>
+      <div id="somethingElse"><p>Lorem Ipsum</p>
       <% if @x %>
       <% @ys.each do |y| %>
       <p>
@@ -44,12 +43,8 @@ class TestHtmlBeautifierIntegration < Test::Unit::TestCase
       <col style="width: 50%;">
       </colgroup>
       <tbody>
-      <tr>
-      <td>First column</td>
-      </tr>
-      <tr>
-      <td>Second column</td>
-      </tr>
+      <tr><td>First column</td></tr><tr>
+      <td>Second column</td></tr>
       </tbody>
       </table>
       </body>

data/test/test_html_beautifier_regression.rb

@@ -83,7 +83,6 @@ class HtmlBeautifierRegressionTest < Test::Unit::TestCase
     assert_beautifies expected, source
   end
 
-
   def test_should_indent_styles
     source = code(%q(
       <style>
@@ -287,4 +286,39 @@ class HtmlBeautifierRegressionTest < Test::Unit::TestCase
     ))
     assert_beautifies source, source
   end
+
+  def test_should_not_modify_pre_content
+    source = code(%q(
+      <div>
+        <pre>   Preformatted   text
+
+                should  <em>not  be </em>
+                      modified,
+                ever!
+
+        </pre>
+      </div>
+    ))
+    assert_beautifies source, source
+  end
+
+  def test_should_add_newline_after_block_elements
+    source = code(%q(
+      <section><h1>Title</h1><p>Lorem <em>ipsum</em></p>
+      <ol>
+        <li>First</li><li>Second</li></ol>
+      </section>
+    ))
+    expected = code(%(
+      <section>
+        <h1>Title</h1>
+        <p>Lorem <em>ipsum</em></p>
+        <ol>
+          <li>First</li>
+          <li>Second</li>
+        </ol>
+      </section>
+    ))
+    assert_beautifies expected, source
+  end
 end

metadata

@@ -1,27 +1,27 @@
 --- !ruby/object:Gem::Specification
 name: htmlbeautifier
 version: !ruby/object:Gem::Version
-  version: 0.0.10
+  version: 0.0.11
 platform: ruby
 authors:
 - Paul Battley
 autorequire: 
 bindir: bin
 cert_chain: []
-date: 2014-09-28 00:00:00.000000000 Z
+date: 2014-12-29 00:00:00.000000000 Z
 dependencies:
 - !ruby/object:Gem::Dependency
   name: rake
   requirement: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '0'
   type: :development
   prerelease: false
   version_requirements: !ruby/object:Gem::Requirement
     requirements:
-    - - ">="
+    - - "~>"
       - !ruby/object:Gem::Version
         version: '0'
 description: A normaliser/beautifier for HTML that also understands embedded Ruby.