From 5a8f37dc6f800c985f4a0571eb674ce7b56d70f4 Mon Sep 17 00:00:00 2001
From: Martin Edenhofer <me@edenhofer.de>
Date: Wed, 5 Dec 2012 02:27:56 +0100
Subject: [PATCH] Moved to own encode.rb for utf8 input checks/validation.

---
 app/models/channel/email_parser.rb | 46 ++++--------------------------
 lib/encode.rb                      | 31 ++++++++++++++++++++
 lib/rss.rb                         |  4 +--
 3 files changed, 39 insertions(+), 42 deletions(-)
 create mode 100644 lib/encode.rb

diff --git a/app/models/channel/email_parser.rb b/app/models/channel/email_parser.rb
index 08a3b02b4..812767dd9 100644
--- a/app/models/channel/email_parser.rb
+++ b/app/models/channel/email_parser.rb
@@ -1,42 +1,8 @@
 # encoding: utf-8
 
 require 'mail'
-#require 'iconv'
+
 class Channel::EmailParser
-  def conv (charset, string)
-
-    # if no charset is given, use LATIN1 as default
-    if !charset || charset == 'US-ASCII' || charset == 'ASCII-8BIT'
-      charset = 'LATIN1'
-    end
-
-    # return if string is false
-    return string if !string
-
-    # validate already existing utf8 strings
-    if charset.downcase == 'utf8' || charset.downcase == 'utf-8'
-      begin
-
-        # return if encoding is valid
-        utf8 = string.force_encoding('UTF-8')
-        return utf8 if utf8.valid_encoding?
-
-        # try to encode from Windows-1252 to utf8
-        string.encode!( 'UTF-8', 'Windows-1252' )
-
-      rescue EncodingError => e
-        puts "Bad encoding: #{new_value.inspect}"
-        string.encode!( 'UTF-8', invalid: :replace, undef: :replace, replace: '?' )
-      end
-      return string
-    end
-#    puts '-------' + charset
-#    puts string
-
-    # convert string
-    string.encode!( 'UTF-8', charset.upcase )
-#    Iconv.conv( 'UTF8', charset, string )
-  end
 
 =begin
 
@@ -97,7 +63,7 @@ class Channel::EmailParser
 
     # set all headers
     mail.header.fields.each { |field|
-      data[field.name.downcase.to_sym] = conv( 'utf8', field.to_s )
+      data[field.name.downcase.to_sym] = Encode.conv( 'utf8', field.to_s )
     }
 
     # set extra headers
@@ -124,7 +90,7 @@ class Channel::EmailParser
       # text attachment/body exists
       if mail.text_part
         data[:body] = mail.text_part.body.decoded
-        data[:body] = conv( mail.text_part.charset, data[:body] )
+        data[:body] = Encode.conv( mail.text_part.charset, data[:body] )
 
       # html attachment/body may exists and will be converted to text
       else
@@ -132,7 +98,7 @@ class Channel::EmailParser
         if mail.html_part.body
           filename = 'html-email'
           data[:body] = mail.html_part.body.to_s
-          data[:body] = conv( mail.html_part.charset.to_s, data[:body] )
+          data[:body] = Encode.conv( mail.html_part.charset.to_s, data[:body] )
           data[:body] = html2ascii( data[:body] )
 
         # any other attachments
@@ -219,7 +185,7 @@ class Channel::EmailParser
       # text part
       if !mail.mime_type || mail.mime_type.to_s ==  '' || mail.mime_type.to_s.downcase == 'text/plain'
         data[:body] = mail.body.decoded
-        data[:body] = conv( mail.charset, data[:body] )
+        data[:body] = Encode.conv( mail.charset, data[:body] )
 
       # html part
       else
@@ -227,7 +193,7 @@ class Channel::EmailParser
         if mail.mime_type.to_s.downcase == 'text/html'
           filename = 'html-email'
           data[:body] = mail.body.decoded
-          data[:body] = conv( mail.charset, data[:body] )
+          data[:body] = Encode.conv( mail.charset, data[:body] )
           data[:body] = html2ascii( data[:body] )
 
         # any other attachments
diff --git a/lib/encode.rb b/lib/encode.rb
new file mode 100644
index 000000000..9daf07bd1
--- /dev/null
+++ b/lib/encode.rb
@@ -0,0 +1,31 @@
+#require 'iconv'
+class Encode
+  def self.conv (charset, string)
+
+    # if no charset is given, use LATIN1 as default
+    if !charset || charset == 'US-ASCII' || charset == 'ASCII-8BIT'
+      charset = 'LATIN1'
+    end
+
+    # return if string is false
+    return string if !string
+
+    # validate already existing utf8 strings
+    if charset.downcase == 'utf8' || charset.downcase == 'utf-8'
+      begin
+
+        # return if encoding is valid
+        utf8 = string.force_encoding('UTF-8')
+        return utf8 if utf8.valid_encoding?
+
+        # try to encode from Windows-1252 to utf8
+        string.encode!( 'UTF-8', 'Windows-1252' )
+
+      rescue EncodingError => e
+        puts "Bad encoding: #{new_value.inspect}"
+        string.encode!( 'UTF-8', invalid: :replace, undef: :replace, replace: '?' )
+      end
+      return string
+    end
+  end
+end
\ No newline at end of file
diff --git a/lib/rss.rb b/lib/rss.rb
index 84f5704b1..838cbb6cb 100644
--- a/lib/rss.rb
+++ b/lib/rss.rb
@@ -16,8 +16,8 @@ module RSS
       rss.items.each { |item|
         record = {
           :id        => item.id,
-          :title     => item.title,
-          :summary   => item.summary,
+          :title     => Encode.conv( 'utf8', item.title ),
+          :summary   => Encode.conv( 'utf8', item.summary ),
           :link      => item.link,
           :published => item.published
         }