Chris@1295: # Redmine - project management software
Chris@1295: # Copyright (C) 2006-2012  Jean-Philippe Lang
Chris@1295: #
Chris@1295: # This program is free software; you can redistribute it and/or
Chris@1295: # modify it under the terms of the GNU General Public License
Chris@1295: # as published by the Free Software Foundation; either version 2
Chris@1295: # of the License, or (at your option) any later version.
Chris@1295: #
Chris@1295: # This program is distributed in the hope that it will be useful,
Chris@1295: # but WITHOUT ANY WARRANTY; without even the implied warranty of
Chris@1295: # MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
Chris@1295: # GNU General Public License for more details.
Chris@1295: #
Chris@1295: # You should have received a copy of the GNU General Public License
Chris@1295: # along with this program; if not, write to the Free Software
Chris@1295: # Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.
Chris@1295: 
Chris@1295: require File.expand_path('../../../../test_helper', __FILE__)
Chris@1295: 
Chris@1295: class Redmine::CodesetUtilTest < ActiveSupport::TestCase
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_from_latin1
Chris@1295:     with_settings :repositories_encodings => 'UTF-8,ISO-8859-1' do
Chris@1295:       s1 = "Texte encod\xc3\xa9"
Chris@1295:       s2 = "Texte encod\xe9"
Chris@1295:       s3 = s2.dup
Chris@1295:       if s1.respond_to?(:force_encoding)
Chris@1295:         s1.force_encoding("UTF-8")
Chris@1295:         s2.force_encoding("ASCII-8BIT")
Chris@1295:         s3.force_encoding("UTF-8")
Chris@1295:       end
Chris@1295:       assert_equal s1, Redmine::CodesetUtil.to_utf8_by_setting(s2)
Chris@1295:       assert_equal s1, Redmine::CodesetUtil.to_utf8_by_setting(s3)
Chris@1295:     end
Chris@1295:   end
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_from_euc_jp
Chris@1295:     with_settings :repositories_encodings => 'UTF-8,EUC-JP' do
Chris@1295:       s1 = "\xe3\x83\xac\xe3\x83\x83\xe3\x83\x89\xe3\x83\x9e\xe3\x82\xa4\xe3\x83\xb3"
Chris@1295:       s2 = "\xa5\xec\xa5\xc3\xa5\xc9\xa5\xde\xa5\xa4\xa5\xf3"
Chris@1295:       s3 = s2.dup
Chris@1295:       if s1.respond_to?(:force_encoding)
Chris@1295:         s1.force_encoding("UTF-8")
Chris@1295:         s2.force_encoding("ASCII-8BIT")
Chris@1295:         s3.force_encoding("UTF-8")
Chris@1295:       end
Chris@1295:       assert_equal s1, Redmine::CodesetUtil.to_utf8_by_setting(s2)
Chris@1295:       assert_equal s1, Redmine::CodesetUtil.to_utf8_by_setting(s3)
Chris@1295:     end
Chris@1295:   end
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_should_be_converted_all_latin1
Chris@1295:     with_settings :repositories_encodings => 'ISO-8859-1' do
Chris@1295:       s1 = "\xc3\x82\xc2\x80"
Chris@1295:       s2 = "\xC2\x80"
Chris@1295:       s3 = s2.dup
Chris@1295:       if s1.respond_to?(:force_encoding)
Chris@1295:         s1.force_encoding("UTF-8")
Chris@1295:         s2.force_encoding("ASCII-8BIT")
Chris@1295:         s3.force_encoding("UTF-8")
Chris@1295:       end
Chris@1295:       assert_equal s1, Redmine::CodesetUtil.to_utf8_by_setting(s2)
Chris@1295:       assert_equal s1, Redmine::CodesetUtil.to_utf8_by_setting(s3)
Chris@1295:     end
Chris@1295:   end
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_blank_string
Chris@1295:     assert_equal "",  Redmine::CodesetUtil.to_utf8_by_setting("")
Chris@1295:     assert_equal nil, Redmine::CodesetUtil.to_utf8_by_setting(nil)
Chris@1295:   end
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_returns_ascii_as_utf8
Chris@1295:     s1 = "ASCII"
Chris@1295:     s2 = s1.dup
Chris@1295:     if s1.respond_to?(:force_encoding)
Chris@1295:       s1.force_encoding("UTF-8")
Chris@1295:       s2.force_encoding("ISO-8859-1")
Chris@1295:     end
Chris@1295:     str1 = Redmine::CodesetUtil.to_utf8_by_setting(s1)
Chris@1295:     str2 = Redmine::CodesetUtil.to_utf8_by_setting(s2)
Chris@1295:     assert_equal s1, str1
Chris@1295:     assert_equal s1, str2
Chris@1295:     if s1.respond_to?(:force_encoding)
Chris@1295:       assert_equal "UTF-8", str1.encoding.to_s
Chris@1295:       assert_equal "UTF-8", str2.encoding.to_s
Chris@1295:     end
Chris@1295:   end
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_invalid_utf8_sequences_should_be_stripped
Chris@1295:     with_settings :repositories_encodings => '' do
Chris@1295:       # s1 = File.read("#{RAILS_ROOT}/test/fixtures/encoding/iso-8859-1.txt")
Chris@1295:       s1 = "Texte encod\xe9 en ISO-8859-1."
Chris@1295:       s1.force_encoding("ASCII-8BIT") if s1.respond_to?(:force_encoding)
Chris@1295:       str = Redmine::CodesetUtil.to_utf8_by_setting(s1)
Chris@1295:       if str.respond_to?(:force_encoding)
Chris@1295:         assert str.valid_encoding?
Chris@1295:         assert_equal "UTF-8", str.encoding.to_s
Chris@1295:       end
Chris@1295:       assert_equal "Texte encod? en ISO-8859-1.", str
Chris@1295:     end
Chris@1295:   end
Chris@1295: 
Chris@1295:   def test_to_utf8_by_setting_invalid_utf8_sequences_should_be_stripped_ja_jis
Chris@1295:     with_settings :repositories_encodings => 'ISO-2022-JP' do
Chris@1295:       s1 = "test\xb5\xfetest\xb5\xfe"
Chris@1295:       s1.force_encoding("ASCII-8BIT") if s1.respond_to?(:force_encoding)
Chris@1295:       str = Redmine::CodesetUtil.to_utf8_by_setting(s1)
Chris@1295:       if str.respond_to?(:force_encoding)
Chris@1295:         assert str.valid_encoding?
Chris@1295:         assert_equal "UTF-8", str.encoding.to_s
Chris@1295:       end
Chris@1295:       assert_equal "test??test??", str
Chris@1295:     end
Chris@1295:   end
Chris@1295: end