From f4208ae83fd43e0cdd663d82a73fabfb65f327bb Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?R=C3=A9gis=20Hanol?= <regis@hanol.fr>
Date: Tue, 12 Aug 2014 00:01:58 +0200
Subject: [PATCH] FEATURE: normalize whitespaces in topic title/post content

---
 lib/post_creator.rb                  | 2 ++
 lib/post_revisor.rb                  | 5 ++++-
 lib/text_cleaner.rb                  | 8 ++++++++
 spec/components/text_cleaner_spec.rb | 8 ++++++++
 4 files changed, 22 insertions(+), 1 deletion(-)

diff --git a/lib/post_creator.rb b/lib/post_creator.rb
index 4d6433a89e3..a0145bbc778 100644
--- a/lib/post_creator.rb
+++ b/lib/post_creator.rb
@@ -203,6 +203,8 @@ class PostCreator
   end
 
   def setup_post
+    @opts[:raw] = TextCleaner.normalize_whitespaces(@opts[:raw]).strip
+
     post = @topic.posts.new(raw: @opts[:raw],
                             user: @user,
                             reply_to_post_number: @opts[:reply_to_post_number])
diff --git a/lib/post_revisor.rb b/lib/post_revisor.rb
index d33f2b36768..dea3b55a4e8 100644
--- a/lib/post_revisor.rb
+++ b/lib/post_revisor.rb
@@ -17,7 +17,10 @@ class PostRevisor
   #  :skip_validation ask ActiveRecord to skip validations
   #
   def revise!(editor, new_raw, opts = {})
-    @editor, @new_raw, @opts = editor, new_raw, opts
+    @editor = editor
+    @opts = opts
+    @new_raw = TextCleaner.normalize_whitespaces(new_raw).strip
+
     return false unless should_revise?
     @post.acting_user = @editor
     revise_post
diff --git a/lib/text_cleaner.rb b/lib/text_cleaner.rb
index a74ee4ead69..da12624bc95 100644
--- a/lib/text_cleaner.rb
+++ b/lib/text_cleaner.rb
@@ -36,10 +36,18 @@ class TextCleaner
     text.sub!(/\s+([!?]\s*)\z/, '\1') if opts[:remove_extraneous_space]
     # Fixes interior spaces
     text.gsub!(/ +/, ' ') if opts[:fixes_interior_spaces]
+    # Normalize whitespaces
+    text = normalize_whitespaces(text)
     # Strip whitespaces
     text.strip! if opts[:strip_whitespaces]
 
     text
   end
 
+  @@whitespaces_regexp = Regexp.new("(\u00A0|\u1680|\u180E|[\u2000-\u200B]|\u2028|\u2029|\u202F|\u205F|\u3000|\uFEFF)", "u").freeze
+
+  def self.normalize_whitespaces(text)
+    text.gsub(@@whitespaces_regexp, ' ')
+  end
+
 end
diff --git a/spec/components/text_cleaner_spec.rb b/spec/components/text_cleaner_spec.rb
index 07d51136541..0603ba3d09c 100644
--- a/spec/components/text_cleaner_spec.rb
+++ b/spec/components/text_cleaner_spec.rb
@@ -191,4 +191,12 @@ describe TextCleaner do
 
   end
 
+  describe "#normalize_whitespaces" do
+    it "normalize whitespaces" do
+      whitespaces = "\u0020\u00A0\u1680\u180E\u2000\u2001\u2002\u2003\u2004\u2005\u2006\u2007\u2008\u2009\u200A\u200B\u2028\u2029\u202F\u205F\u3000\uFEFF"
+      whitespaces.strip.should_not == ""
+      TextCleaner.normalize_whitespaces(whitespaces).strip.should == ""
+    end
+  end
+
 end