snatch 1.0.7 → 1.0.8

data/VERSION

@@ -1 +1 @@
-1.0.7
+1.0.8

data/lib/snatch/clean.rb

@@ -22,5 +22,13 @@ class Snatch
 
       File.open(@file_name, 'w') { |f| f.write @doc.to_xhtml }
     end
+
+	private
+
+	  # Sadly we can't use this as our XSLT won't output valid XHTML
+		# def clean_xhtml
+		# 	xsl_io = File.open(File.expand_path('../../../xsl/pretty_print.xsl', __FILE__))
+		# 	Nokogiri::XSLT(xsl_io).apply_to(@doc)
+		# end
   end
 end

data/lib/snatch/clean/html.rb

@@ -63,17 +63,17 @@ class Snatch
       end
 
       def update
-        @doc.css('base, meta[generator]').each { |node| node.remove }
+        @doc.css('base, meta[name=generator]').each { |node| node.remove }
 
         @doc.search('//comment()').remove
 
+        klass = Class.new { include HrefFixMethods }.new
         HrefFixMethods.instance_methods.each do |m|
-          klass = Class.new { include HrefFixMethods }.new
           @doc.css('a[href]').each { |a| klass.send m, a }
         end
 
+        klass = Class.new { include SrcFixMethods }.new
         SrcFixMethods.instance_methods.each do |m|
-          klass = Class.new { include SrcFixMethods }.new
           @doc.css('[src]').each { |a| klass.send m, a }
         end
       end

data/snatch.gemspec

@@ -5,7 +5,7 @@
 
 Gem::Specification.new do |s|
   s.name = %q{snatch}
-  s.version = "1.0.7"
+  s.version = "1.0.8"
 
   s.required_rubygems_version = Gem::Requirement.new(">= 0") if s.respond_to? :required_rubygems_version=
   s.authors = ["James Conroy-Finn"]
@@ -36,7 +36,8 @@ Gem::Specification.new do |s|
      "spec/snatch_spec.rb",
      "spec/spec.opts",
      "spec/spec_helper.rb",
-     "spec/support/matchers/nokogiri.rb"
+     "spec/support/matchers/nokogiri.rb",
+     "xsl/pretty_print.xsl"
   ]
   s.homepage = %q{http://github.com/jcf/snatch}
   s.rdoc_options = ["--charset=UTF-8"]

data/spec/snatch/clean/html_spec.rb

@@ -14,7 +14,7 @@ describe Snatch::Clean::HTML do
 
     it 'should remove any base tags' do
       mock_nodes = [(mock_node = mock('nokogiri_node'))]
-      @html.doc.should_receive(:css).with('base, meta[generator]').and_return(mock_nodes)
+      @html.doc.should_receive(:css).with('base, meta[name=generator]').and_return(mock_nodes)
       @html.doc.should_receive(:search).and_return(mock_node)
       mock_node.should_receive(:remove).twice.and_return([])
       mock_nodes.should_receive(:each).and_yield(mock_node)
@@ -23,7 +23,7 @@ describe Snatch::Clean::HTML do
 
     it 'should remove comments and the generator meta tag' do
       mock_nodes = [(mock_node = mock('nokogiri_node'))]
-      @html.doc.should_receive(:css).with('base, meta[generator]').and_return([])
+      @html.doc.should_receive(:css).with('base, meta[name=generator]').and_return([])
       @html.doc.should_receive(:search).with('//comment()').and_return(mock_nodes)
       mock_nodes.should_receive(:remove).and_return([mock_node])
       @html.send(:update) 

data/xsl/pretty_print.xsl

@@ -0,0 +1,47 @@
+<xsl:stylesheet version="1.0" xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
+  <xsl:output method="xml" encoding="ISO-8859-1"/>
+  <xsl:param name="indent-increment" select="'   '"/>
+
+  <xsl:template name="newline">
+    <xsl:text disable-output-escaping="yes">
+</xsl:text>
+  </xsl:template>
+
+  <xsl:template match="comment() | processing-instruction()">
+    <xsl:param name="indent" select="''"/>
+    <xsl:call-template name="newline"/>
+    <xsl:value-of select="$indent"/>
+    <xsl:copy />
+  </xsl:template>
+
+  <xsl:template match="text()">
+    <xsl:param name="indent" select="''"/>
+    <xsl:call-template name="newline"/>
+    <xsl:value-of select="$indent"/>
+    <xsl:value-of select="normalize-space(.)"/>
+  </xsl:template>
+
+  <xsl:template match="text()[normalize-space(.)='']"/>
+
+  <xsl:template match="*">
+    <xsl:param name="indent" select="''"/>
+    <xsl:call-template name="newline"/>
+    <xsl:value-of select="$indent"/>
+      <xsl:choose>
+       <xsl:when test="count(child::*) > 0">
+        <xsl:copy>
+         <xsl:copy-of select="@*"/>
+         <xsl:apply-templates select="*|text()">
+           <xsl:with-param name="indent" select="concat ($indent, $indent-increment)"/>
+         </xsl:apply-templates>
+         <xsl:call-template name="newline"/>
+         <xsl:value-of select="$indent"/>
+        </xsl:copy>
+       </xsl:when>
+       <xsl:otherwise>
+        <xsl:copy-of select="."/>
+       </xsl:otherwise>
+     </xsl:choose>
+  </xsl:template>
+</xsl:stylesheet>
+

metadata

@@ -1,7 +1,7 @@
 --- !ruby/object:Gem::Specification 
 name: snatch
 version: !ruby/object:Gem::Version 
-  version: 1.0.7
+  version: 1.0.8
 platform: ruby
 authors: 
 - James Conroy-Finn
@@ -62,6 +62,7 @@ files:
 - spec/spec.opts
 - spec/spec_helper.rb
 - spec/support/matchers/nokogiri.rb
+- xsl/pretty_print.xsl
 has_rdoc: true
 homepage: http://github.com/jcf/snatch
 licenses: []