Added code block normalization routing for import

Gemfile

@@ -199,6 +199,8 @@ gem 'fast_blank' #, github: "SamSaffron/fast_blank"
 # this provides a very efficient lru cache
 gem 'lru_redux'
 
+gem 'htmlentities', require: false
+
 # IMPORTANT: mini profiler monkey patches, so it better be required last
 #  If you want to amend mini profiler to do the monkey patches in the railstie
 #  we are open to it. by deferring require to the initializer we can configure discourse installs without it

Gemfile.lock

@@ -130,6 +130,7 @@ GEM
     highline (1.6.20)
     hike (1.2.3)
     hiredis (0.5.1)
+    htmlentities (4.3.2)
     i18n (0.6.9)
     image_optim (0.9.1)
       exifr (~> 1.1.3)
@@ -420,6 +421,7 @@ DEPENDENCIES
   handlebars-source (= 1.3.0)
   highline
   hiredis
+  htmlentities
   image_optim (= 0.9.1)
   librarian (>= 0.0.25)
   listen (= 0.7.3)

lib/import/normalize.rb

@@ -0,0 +1,12 @@
+# markdown normalizer to be used by importers
+#
+#
+require 'htmlentities'
+module Import::Normalize
+  def self.normalize_code_blocks(code, lang=nil)
+    coder = HTMLEntities.new
+    code.gsub(/<pre>\s*<code>\n?(.*?)\n?<\/code>\s*<\/pre>/m) {
+        "\n```#{lang}\n#{coder.decode($1)}\n```\n"
+    }
+  end
+end

spec/components/import/normalize_spec.rb

@@ -0,0 +1,21 @@
+require "spec_helper"
+require_dependency "import/normalize"
+
+describe Import::Normalize do
+  describe "#normalize_code_blocks" do
+    it "normalizes 2 code blocks correctly" do
+      markdown = <<MD
+      &nbsp;
+      <pre>
+        <code>
+        I am a te&nbsp;&quot;
+        </code></pre>
+        test &nbsp;
+        <pre><code>this is a &quot;&quot;</code></pre>
+MD
+      expected = "      &nbsp;\n      \n```\n        I am a te \"\n        \n```\n\n        test &nbsp;\n        \n```\nthis is a \"\"\n```\n\n"
+      Import::Normalize.normalize_code_blocks(markdown).should == expected
+    end
+  end
+end
+