Remove invalid utf8 sequences from commit comments and author name (#4773).

git-svn-id: svn+ssh://rubyforge.org/var/svn/redmine/trunk@3466 e93f8b46-1217-0410-a6f0-8f06a7374b81

Remove invalid utf8 sequences from commit comments and author name (#4773).
git-svn-id: svn+ssh://rubyforge.org/var/svn/redmine/trunk@3466 e93f8b46-1217-0410-a6f0-8f06a7374b81
46aa8555 · Jean-Philippe Lang · 6e689215 · 46aa8555 · 46aa8555 · 46aa8555
Commit 46aa8555 authored Feb 20, 2010 by Jean-Philippe Lang
Hide whitespace changes
Inline Side-by-side

Showing with 27 additions and 5 deletions

changeset.rb app/models/changeset.rb +8 -3

iso-8859-1.txt test/fixtures/encoding/iso-8859-1.txt +1 -0

changeset_test.rb test/unit/changeset_test.rb +18 -2

No files found.
--- a/app/models/changeset.rb
+++ b/app/models/changeset.rb
 # Redmine - project management software
-# Copyright (C) 2006-2008  Jean-Philippe Lang
+# Copyright (C) 2006-2010  Jean-Philippe Lang
 #
 # This program is free software; you can redistribute it and/or
 # modify it under the terms of the GNU General Public License
@@ -57,6 +57,10 @@ class Changeset < ActiveRecord::Base
    super
  end
  
+  def committer=(arg)
+    write_attribute(:committer, self.class.to_utf8(arg.to_s))
+  end
+
  def project
    repository.project
  end
@@ -180,11 +184,12 @@ class Changeset < ActiveRecord::Base
    encoding = Setting.commit_logs_encoding.to_s.strip
    unless encoding.blank? || encoding == 'UTF-8'
      begin
-        return Iconv.conv('UTF-8', encoding, str)
+        str = Iconv.conv('UTF-8', encoding, str)
      rescue Iconv::Failure
        # do nothing here
      end
    end
-    str
+    # removes invalid UTF8 sequences
+    Iconv.conv('UTF-8//IGNORE', 'UTF-8', str + '  ')[0..-3]
  end
 end
--- a/test/fixtures/encoding/iso-8859-1.txt
+++ b/test/fixtures/encoding/iso-8859-1.txt
+Texte encodé en ISO-8859-1.
\ No newline at end of file
--- a/test/unit/changeset_test.rb
+++ b/test/unit/changeset_test.rb
-# redMine - project management software
-# Copyright (C) 2006-2007  Jean-Philippe Lang
+# encoding: utf-8
+#
+# Redmine - project management software
+# Copyright (C) 2006-2010  Jean-Philippe Lang
 #
 # This program is free software; you can redistribute it and/or
 # modify it under the terms of the GNU General Public License
@@ -117,4 +119,18 @@ class ChangesetTest < ActiveSupport::TestCase
    changeset = Changeset.find_by_revision('10')
    assert_nil changeset.next
  end
+  
+  def test_comments_should_be_converted_to_utf8
+    with_settings :commit_logs_encoding => 'ISO-8859-1' do
+      c = Changeset.new
+      c.comments = File.read("#{RAILS_ROOT}/test/fixtures/encoding/iso-8859-1.txt")
+      assert_equal "Texte encodé en ISO-8859-1.", c.comments
+    end
+  end
+  
+  def test_invalid_utf8_sequences_in_comments_should_be_stripped
+    c = Changeset.new
+    c.comments = File.read("#{RAILS_ROOT}/test/fixtures/encoding/iso-8859-1.txt")
+    assert_equal "Texte encod en ISO-8859-1.", c.comments
+  end
 end