test/ruby/test_transcode.rb

   1 # -*- encoding: ASCII-8BIT -*-   # make sure this runs in binary mode
   2 # some of the comments are in UTF-8
   3
   4 require 'test/unit'
   5 class TestTranscode < Test::Unit::TestCase
   6   def setup_really_needed? # trick to create all the necessary encodings
   7     all_encodings = [ 'ISO-8859-1', 'ISO-8859-2',
   8                       'ISO-8859-3', 'ISO-8859-4',
   9                       'ISO-8859-5', 'ISO-8859-6',
  10                       'ISO-8859-7', 'ISO-8859-8',
  11                       'ISO-8859-9', 'ISO-8859-10',
  12                       'ISO-8859-11', 'ISO-8859-13',
  13                       'ISO-8859-14', 'ISO-8859-15',
  14                       'UTF-16BE'
  15                     ]
  16     all_encodings.each do |enc|
  17       'abc'.encode(enc, 'UTF-8')
  18     end
  19   end
  20
  21   def test_errors
  22     assert_raise(ArgumentError) { 'abc'.encode }
  23     assert_raise(ArgumentError) { 'abc'.encode! }
  24     assert_raise(ArgumentError) { 'abc'.encode('foo', 'bar') }
  25     assert_raise(ArgumentError) { 'abc'.encode!('foo', 'bar') }
  26     assert_raise(ArgumentError) { 'abc'.force_encoding('utf-8').encode('foo') }
  27     assert_raise(ArgumentError) { 'abc'.force_encoding('utf-8').encode!('foo') }
  28     assert_raise(Encoding::ConversionUndefined) { "\x80".encode('utf-8','ASCII-8BIT') }
  29     assert_raise(Encoding::InvalidByteSequence) { "\x80".encode('utf-8','US-ASCII') }
  30     assert_raise(Encoding::ConversionUndefined) { "\xA5".encode('utf-8','iso-8859-3') }
  31   end
  32
  33   def test_arguments
  34     assert_equal('abc', 'abc'.force_encoding('utf-8').encode('iso-8859-1'))
  35     # check that encoding is kept when no conversion is done
  36     assert_equal('abc'.force_encoding('Shift_JIS'), 'abc'.force_encoding('Shift_JIS').encode('Shift_JIS'))
  37     assert_equal('abc'.force_encoding('Shift_JIS'), 'abc'.force_encoding('Shift_JIS').encode!('Shift_JIS'))
  38     # assert that encoding is correctly set
  39     assert_equal("D\u00FCrst".encoding, "D\xFCrst".force_encoding('iso-8859-1').encode('utf-8').encoding)
  40     # check that Encoding can be used as parameter
  41     assert_equal("D\u00FCrst", "D\xFCrst".encode('utf-8', Encoding.find('ISO-8859-1')))
  42     assert_equal("D\u00FCrst", "D\xFCrst".encode(Encoding.find('utf-8'), 'ISO-8859-1'))
  43     assert_equal("D\u00FCrst", "D\xFCrst".encode(Encoding.find('utf-8'), Encoding.find('ISO-8859-1')))
  44   end
  45
  46   def test_length
  47     assert_equal("\u20AC"*20, ("\xA4"*20).encode('utf-8', 'iso-8859-15'))
  48     assert_equal("\u20AC"*20, ("\xA4"*20).encode!('utf-8', 'iso-8859-15'))
  49     assert_equal("\u20AC"*2000, ("\xA4"*2000).encode('utf-8', 'iso-8859-15'))
  50     assert_equal("\u20AC"*2000, ("\xA4"*2000).encode!('utf-8', 'iso-8859-15'))
  51     assert_equal("\u20AC"*200000, ("\xA4"*200000).encode('utf-8', 'iso-8859-15'))
  52     assert_equal("\u20AC"*200000, ("\xA4"*200000).encode!('utf-8', 'iso-8859-15'))
  53   end
  54
  55   def check_both_ways(utf8, raw, encoding)
  56     assert_equal(utf8.force_encoding('utf-8'), raw.encode('utf-8', encoding))
  57     assert_equal(raw.force_encoding(encoding), utf8.encode(encoding, 'utf-8'))
  58   end
  59
  60   def test_encodings
  61     check_both_ways("\u307E\u3064\u3082\u3068 \u3086\u304D\u3072\u308D",
  62         "\x82\xdc\x82\xc2\x82\xe0\x82\xc6 \x82\xe4\x82\xab\x82\xd0\x82\xeb", 'shift_jis') # まつもと ゆきひろ
  63     check_both_ways("\u307E\u3064\u3082\u3068 \u3086\u304D\u3072\u308D",
  64         "\xa4\xde\xa4\xc4\xa4\xe2\xa4\xc8 \xa4\xe6\xa4\xad\xa4\xd2\xa4\xed", 'euc-jp')
  65     check_both_ways("\u677E\u672C\u884C\u5F18", "\x8f\xbc\x96\x7b\x8d\x73\x8d\x4f", 'shift_jis') # 松本行弘
  66     check_both_ways("\u677E\u672C\u884C\u5F18", "\xbe\xbe\xcb\xdc\xb9\xd4\xb9\xb0", 'euc-jp')
  67     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-1') # Dürst
  68     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-2')
  69     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-3')
  70     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-4')
  71     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-9')
  72     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-10')
  73     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-13')
  74     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-14')
  75     check_both_ways("D\u00FCrst", "D\xFCrst", 'iso-8859-15')
  76     check_both_ways("r\u00E9sum\u00E9", "r\xE9sum\xE9", 'iso-8859-1') # résumé
  77     check_both_ways("\u0065\u006C\u0151\u00ED\u0072\u00E1\u0073", "el\xF5\xEDr\xE1s", 'iso-8859-2') # előírás
  78     check_both_ways("\u043F\u0435\u0440\u0435\u0432\u043E\u0434",
  79          "\xDF\xD5\xE0\xD5\xD2\xDE\xD4", 'iso-8859-5') # перевод
  80     check_both_ways("\u0643\u062A\u0628", "\xE3\xCA\xC8", 'iso-8859-6') # كتب
  81     check_both_ways("\u65E5\u8A18", "\x93\xFA\x8BL", 'shift_jis') # 日記
  82     check_both_ways("\u65E5\u8A18", "\xC6\xFC\xB5\xAD", 'euc-jp')
  83     check_both_ways("\uC560\uC778\uAD6C\uD568\u0020\u6734\uC9C0\uC778",
  84          "\xBE\xD6\xC0\xCE\xB1\xB8\xC7\xD4\x20\xDA\xD3\xC1\xF6\xC0\xCE", 'euc-kr') # 애인구함 朴지인
  85     check_both_ways("\uC544\uD58F\uD58F\u0020\uB620\uBC29\uD6BD\uB2D8\u0020\uC0AC\uB791\uD716",
  86          "\xBE\xC6\xC1\x64\xC1\x64\x20\x8C\x63\xB9\xE6\xC4\x4F\xB4\xD4\x20\xBB\xE7\xB6\xFB\xC5\x42", 'cp949') # 아햏햏 똠방횽님 사랑휖
  87   end
  88
  89   def test_twostep
  90     assert_equal("D\xFCrst".force_encoding('iso-8859-2'), "D\xFCrst".encode('iso-8859-2', 'iso-8859-1'))
  91   end
  92
  93   def test_ascii_range
  94     encodings = [
  95       'US-ASCII', 'ASCII-8BIT',
  96       'ISO-8859-1', 'ISO-8859-2', 'ISO-8859-3',
  97       'ISO-8859-4', 'ISO-8859-5', 'ISO-8859-6',
  98       'ISO-8859-7', 'ISO-8859-8', 'ISO-8859-9',
  99       'ISO-8859-10', 'ISO-8859-11', 'ISO-8859-13',
 100       'ISO-8859-14', 'ISO-8859-15',
 101       'EUC-JP', 'SHIFT_JIS', 'EUC-KR'
 102     ]
 103     all_ascii = (0..127).to_a.pack 'C*'
 104     encodings.each do |enc|
 105       test_start = all_ascii
 106       assert_equal(test_start, test_start.encode('UTF-8',enc).encode(enc).force_encoding('ASCII-8BIT'))
 107     end
 108   end
 109
 110   def test_all_bytes
 111     encodings_8859 = [
 112       'ISO-8859-1', 'ISO-8859-2',
 113       #'ISO-8859-3', # not all bytes used
 114       'ISO-8859-4', 'ISO-8859-5',
 115       #'ISO-8859-6', # not all bytes used
 116       #'ISO-8859-7', # not all bytes used
 117       #'ISO-8859-8', # not all bytes used
 118       'ISO-8859-9', 'ISO-8859-10',
 119       #'ISO-8859-11', # not all bytes used
 120       #'ISO-8859-12', # not available
 121       'ISO-8859-13','ISO-8859-14','ISO-8859-15',
 122       #'ISO-8859-16', # not available
 123     ]
 124     all_bytes = (0..255).to_a.pack 'C*'
 125     encodings_8859.each do |enc|
 126       test_start = all_bytes
 127       assert_equal(test_start, test_start.encode('UTF-8',enc).encode(enc).force_encoding('ASCII-8BIT'))
 128     end
 129   end
 130
 131   def check_utf_16_both_ways(utf8, raw)
 132     copy = raw.dup
 133     0.step(copy.length-1, 2) { |i| copy[i+1], copy[i] = copy[i], copy[i+1] }
 134     check_both_ways(utf8, raw, 'utf-16be')
 135     check_both_ways(utf8, copy, 'utf-16le')
 136   end
 137
 138   def test_utf_16
 139     check_utf_16_both_ways("abc", "\x00a\x00b\x00c")
 140     check_utf_16_both_ways("\u00E9", "\x00\xE9");
 141     check_utf_16_both_ways("\u00E9\u0070\u00E9\u0065", "\x00\xE9\x00\x70\x00\xE9\x00\x65") # épée
 142     check_utf_16_both_ways("\u677E\u672C\u884C\u5F18", "\x67\x7E\x67\x2C\x88\x4C\x5F\x18") # 松本行弘
 143     check_utf_16_both_ways("\u9752\u5C71\u5B66\u9662\u5927\u5B66", "\x97\x52\x5C\x71\x5B\x66\x96\x62\x59\x27\x5B\x66") # 青山学院大学
 144     check_utf_16_both_ways("Martin D\u00FCrst", "\x00M\x00a\x00r\x00t\x00i\x00n\x00 \x00D\x00\xFC\x00r\x00s\x00t") # Martin Dürst
 145     # BMP
 146     check_utf_16_both_ways("\u0000", "\x00\x00")
 147     check_utf_16_both_ways("\u007F", "\x00\x7F")
 148     check_utf_16_both_ways("\u0080", "\x00\x80")
 149     check_utf_16_both_ways("\u0555", "\x05\x55")
 150     check_utf_16_both_ways("\u04AA", "\x04\xAA")
 151     check_utf_16_both_ways("\u0333", "\x03\x33")
 152     check_utf_16_both_ways("\u04CC", "\x04\xCC")
 153     check_utf_16_both_ways("\u00F0", "\x00\xF0")
 154     check_utf_16_both_ways("\u070F", "\x07\x0F")
 155     check_utf_16_both_ways("\u07FF", "\x07\xFF")
 156     check_utf_16_both_ways("\u0800", "\x08\x00")
 157     check_utf_16_both_ways("\uD7FF", "\xD7\xFF")
 158     check_utf_16_both_ways("\uE000", "\xE0\x00")
 159     check_utf_16_both_ways("\uFFFF", "\xFF\xFF")
 160     check_utf_16_both_ways("\u5555", "\x55\x55")
 161     check_utf_16_both_ways("\uAAAA", "\xAA\xAA")
 162     check_utf_16_both_ways("\u3333", "\x33\x33")
 163     check_utf_16_both_ways("\uCCCC", "\xCC\xCC")
 164     check_utf_16_both_ways("\uF0F0", "\xF0\xF0")
 165     check_utf_16_both_ways("\u0F0F", "\x0F\x0F")
 166     check_utf_16_both_ways("\uFF00", "\xFF\x00")
 167     check_utf_16_both_ways("\u00FF", "\x00\xFF")
 168     # outer planes
 169     check_utf_16_both_ways("\u{10000}", "\xD8\x00\xDC\x00")
 170     check_utf_16_both_ways("\u{FFFFF}", "\xDB\xBF\xDF\xFF")
 171     check_utf_16_both_ways("\u{100000}", "\xDB\xC0\xDC\x00")
 172     check_utf_16_both_ways("\u{10FFFF}", "\xDB\xFF\xDF\xFF")
 173     check_utf_16_both_ways("\u{105555}", "\xDB\xD5\xDD\x55")
 174     check_utf_16_both_ways("\u{55555}", "\xD9\x15\xDD\x55")
 175     check_utf_16_both_ways("\u{AAAAA}", "\xDA\x6A\xDE\xAA")
 176     check_utf_16_both_ways("\u{33333}", "\xD8\x8C\xDF\x33")
 177     check_utf_16_both_ways("\u{CCCCC}", "\xDA\xF3\xDC\xCC")
 178     check_utf_16_both_ways("\u{8F0F0}", "\xD9\xFC\xDC\xF0")
 179     check_utf_16_both_ways("\u{F0F0F}", "\xDB\x83\xDF\x0F")
 180     check_utf_16_both_ways("\u{8FF00}", "\xD9\xFF\xDF\x00")
 181     check_utf_16_both_ways("\u{F00FF}", "\xDB\x80\xDC\xFF")
 182   end
 183
 184   def check_utf_32_both_ways(utf8, raw)
 185     copy = raw.dup
 186     0.step(copy.length-1, 4) do |i|
 187       copy[i+3], copy[i+2], copy[i+1], copy[i] = copy[i], copy[i+1], copy[i+2], copy[i+3]
 188     end
 189     check_both_ways(utf8, raw, 'utf-32be')
 190     #check_both_ways(utf8, copy, 'utf-32le')
 191   end
 192
 193   def test_utf_32
 194     check_utf_32_both_ways("abc", "\x00\x00\x00a\x00\x00\x00b\x00\x00\x00c")
 195     check_utf_32_both_ways("\u00E9", "\x00\x00\x00\xE9");
 196     check_utf_32_both_ways("\u00E9\u0070\u00E9\u0065",
 197       "\x00\x00\x00\xE9\x00\x00\x00\x70\x00\x00\x00\xE9\x00\x00\x00\x65") # épée
 198     check_utf_32_both_ways("\u677E\u672C\u884C\u5F18",
 199       "\x00\x00\x67\x7E\x00\x00\x67\x2C\x00\x00\x88\x4C\x00\x00\x5F\x18") # 松本行弘
 200     check_utf_32_both_ways("\u9752\u5C71\u5B66\u9662\u5927\u5B66",
 201       "\x00\x00\x97\x52\x00\x00\x5C\x71\x00\x00\x5B\x66\x00\x00\x96\x62\x00\x00\x59\x27\x00\x00\x5B\x66") # 青山学院大学
 202     check_utf_32_both_ways("Martin D\u00FCrst",
 203       "\x00\x00\x00M\x00\x00\x00a\x00\x00\x00r\x00\x00\x00t\x00\x00\x00i\x00\x00\x00n\x00\x00\x00 \x00\x00\x00D\x00\x00\x00\xFC\x00\x00\x00r\x00\x00\x00s\x00\x00\x00t") # Martin Dürst
 204     # BMP
 205     check_utf_32_both_ways("\u0000", "\x00\x00\x00\x00")
 206     check_utf_32_both_ways("\u007F", "\x00\x00\x00\x7F")
 207     check_utf_32_both_ways("\u0080", "\x00\x00\x00\x80")
 208     check_utf_32_both_ways("\u0555", "\x00\x00\x05\x55")
 209     check_utf_32_both_ways("\u04AA", "\x00\x00\x04\xAA")
 210     check_utf_32_both_ways("\u0333", "\x00\x00\x03\x33")
 211     check_utf_32_both_ways("\u04CC", "\x00\x00\x04\xCC")
 212     check_utf_32_both_ways("\u00F0", "\x00\x00\x00\xF0")
 213     check_utf_32_both_ways("\u070F", "\x00\x00\x07\x0F")
 214     check_utf_32_both_ways("\u07FF", "\x00\x00\x07\xFF")
 215     check_utf_32_both_ways("\u0800", "\x00\x00\x08\x00")
 216     check_utf_32_both_ways("\uD7FF", "\x00\x00\xD7\xFF")
 217     check_utf_32_both_ways("\uE000", "\x00\x00\xE0\x00")
 218     check_utf_32_both_ways("\uFFFF", "\x00\x00\xFF\xFF")
 219     check_utf_32_both_ways("\u5555", "\x00\x00\x55\x55")
 220     check_utf_32_both_ways("\uAAAA", "\x00\x00\xAA\xAA")
 221     check_utf_32_both_ways("\u3333", "\x00\x00\x33\x33")
 222     check_utf_32_both_ways("\uCCCC", "\x00\x00\xCC\xCC")
 223     check_utf_32_both_ways("\uF0F0", "\x00\x00\xF0\xF0")
 224     check_utf_32_both_ways("\u0F0F", "\x00\x00\x0F\x0F")
 225     check_utf_32_both_ways("\uFF00", "\x00\x00\xFF\x00")
 226     check_utf_32_both_ways("\u00FF", "\x00\x00\x00\xFF")
 227     # outer planes
 228     check_utf_32_both_ways("\u{10000}", "\x00\x01\x00\x00")
 229     check_utf_32_both_ways("\u{FFFFF}", "\x00\x0F\xFF\xFF")
 230     check_utf_32_both_ways("\u{100000}","\x00\x10\x00\x00")
 231     check_utf_32_both_ways("\u{10FFFF}","\x00\x10\xFF\xFF")
 232     check_utf_32_both_ways("\u{105555}","\x00\x10\x55\x55")
 233     check_utf_32_both_ways("\u{55555}", "\x00\x05\x55\x55")
 234     check_utf_32_both_ways("\u{AAAAA}", "\x00\x0A\xAA\xAA")
 235     check_utf_32_both_ways("\u{33333}", "\x00\x03\x33\x33")
 236     check_utf_32_both_ways("\u{CCCCC}", "\x00\x0C\xCC\xCC")
 237     check_utf_32_both_ways("\u{8F0F0}", "\x00\x08\xF0\xF0")
 238     check_utf_32_both_ways("\u{F0F0F}", "\x00\x0F\x0F\x0F")
 239     check_utf_32_both_ways("\u{8FF00}", "\x00\x08\xFF\x00")
 240     check_utf_32_both_ways("\u{F00FF}", "\x00\x0F\x00\xFF")
 241   end
 242
 243   def test_invalid_ignore
 244     # arguments only
 245     assert_nothing_raised { 'abc'.encode('utf-8', invalid: :ignore) }
 246     # check handling of UTF-8 ill-formed subsequences
 247     assert_equal("\x00\x41\x00\x3E\x00\x42".force_encoding('UTF-16BE'),
 248       "\x41\xC2\x3E\x42".encode('UTF-16BE', 'UTF-8', invalid: :ignore))
 249     assert_equal("\x00\x41\x00\xF1\x00\x42".force_encoding('UTF-16BE'),
 250       "\x41\xC2\xC3\xB1\x42".encode('UTF-16BE', 'UTF-8', invalid: :ignore))
 251     assert_equal("\x00\x42".force_encoding('UTF-16BE'),
 252       "\xF0\x80\x80\x42".encode('UTF-16BE', 'UTF-8', invalid: :ignore))
 253     assert_equal(''.force_encoding('UTF-16BE'),
 254       "\x82\xAB".encode('UTF-16BE', 'UTF-8', invalid: :ignore))
 255
 256     assert_equal("\e$B!!\e(B".force_encoding("ISO-2022-JP"),
 257       "\xA1\xA1\xFF".encode("ISO-2022-JP", "EUC-JP", invalid: :ignore))
 258     assert_equal("\e$B\x24\x22\x24\x24\e(B".force_encoding("ISO-2022-JP"),
 259       "\xA4\xA2\xFF\xA4\xA4".encode("ISO-2022-JP", "EUC-JP", invalid: :ignore))
 260     assert_equal("\e$B\x24\x22\x24\x24\e(B".force_encoding("ISO-2022-JP"),
 261       "\xA4\xA2\xFF\xFF\xA4\xA4".encode("ISO-2022-JP", "EUC-JP", invalid: :ignore))
 262   end
 263
 264   def test_invalid_replace
 265     # arguments only
 266     assert_nothing_raised { 'abc'.encode('UTF-8', invalid: :replace) }
 267     assert_equal("\xEF\xBF\xBD".force_encoding("UTF-8"),
 268       "\x80".encode("UTF-8", "UTF-16BE", invalid: :replace))
 269     assert_equal("\xFF\xFD".force_encoding("UTF-16BE"),
 270       "\x80".encode("UTF-16BE", "UTF-8", invalid: :replace))
 271     assert_equal("\xFD\xFF".force_encoding("UTF-16LE"),
 272       "\x80".encode("UTF-16LE", "UTF-8", invalid: :replace))
 273     assert_equal("\x00\x00\xFF\xFD".force_encoding("UTF-32BE"),
 274       "\x80".encode("UTF-32BE", "UTF-8", invalid: :replace))
 275     assert_equal("\xFD\xFF\x00\x00".force_encoding("UTF-32LE"),
 276       "\x80".encode("UTF-32LE", "UTF-8", invalid: :replace))
 277
 278     assert_equal("\uFFFD!",
 279       "\xdc\x00\x00!".encode("utf-8", "utf-16be", :invalid=>:replace))
 280     assert_equal("\uFFFD!",
 281       "\xd8\x00\x00!".encode("utf-8", "utf-16be", :invalid=>:replace))
 282
 283     assert_equal("\uFFFD!",
 284       "\x00\xdc!\x00".encode("utf-8", "utf-16le", :invalid=>:replace))
 285     assert_equal("\uFFFD!",
 286       "\x00\xd8!\x00".encode("utf-8", "utf-16le", :invalid=>:replace))
 287
 288     assert_equal("\uFFFD!",
 289       "\x01\x00\x00\x00\x00\x00\x00!".encode("utf-8", "utf-32be", :invalid=>:replace), "[ruby-dev:35726]")
 290     assert_equal("\uFFFD!",
 291       "\x00\xff\x00\x00\x00\x00\x00!".encode("utf-8", "utf-32be", :invalid=>:replace))
 292     assert_equal("\uFFFD!",
 293       "\x00\x00\xd8\x00\x00\x00\x00!".encode("utf-8", "utf-32be", :invalid=>:replace))
 294
 295     assert_equal("\uFFFD!",
 296       "\x00\x00\x00\xff!\x00\x00\x00".encode("utf-8", "utf-32le", :invalid=>:replace))
 297     assert_equal("\uFFFD!",
 298       "\x00\x00\xff\x00!\x00\x00\x00".encode("utf-8", "utf-32le", :invalid=>:replace))
 299     assert_equal("\uFFFD!",
 300       "\x00\xd8\x00\x00!\x00\x00\x00".encode("utf-8", "utf-32le", :invalid=>:replace))
 301
 302     assert_equal("\uFFFD!",
 303       "\xff!".encode("utf-8", "euc-jp", :invalid=>:replace))
 304     assert_equal("\uFFFD!",
 305       "\xa1!".encode("utf-8", "euc-jp", :invalid=>:replace))
 306     assert_equal("\uFFFD!",
 307       "\x8f\xa1!".encode("utf-8", "euc-jp", :invalid=>:replace))
 308
 309     assert_equal("?",
 310       "\xdc\x00".encode("EUC-JP", "UTF-16BE", :invalid=>:replace), "[ruby-dev:35776]")
 311     assert_equal("ab?cd?ef",
 312       "\0a\0b\xdc\x00\0c\0d\xdf\x00\0e\0f".encode("EUC-JP", "UTF-16BE", :invalid=>:replace))
 313
 314     assert_equal("\e$B!!\e(B?".force_encoding("ISO-2022-JP"),
 315       "\xA1\xA1\xFF".encode("ISO-2022-JP", "EUC-JP", invalid: :replace))
 316     assert_equal("\e$B\x24\x22\e(B?\e$B\x24\x24\e(B".force_encoding("ISO-2022-JP"),
 317       "\xA4\xA2\xFF\xA4\xA4".encode("ISO-2022-JP", "EUC-JP", invalid: :replace))
 318     assert_equal("\e$B\x24\x22\e(B??\e$B\x24\x24\e(B".force_encoding("ISO-2022-JP"),
 319       "\xA4\xA2\xFF\xFF\xA4\xA4".encode("ISO-2022-JP", "EUC-JP", invalid: :replace))
 320   end
 321
 322   def test_undef_replace
 323     assert_equal("?", "\u20AC".encode("EUC-JP", :undef=>:replace), "[ruby-dev:35709]")
 324   end
 325
 326   def test_shift_jis
 327     check_both_ways("\u3000", "\x81\x40", 'shift_jis') # full-width space
 328     check_both_ways("\u00D7", "\x81\x7E", 'shift_jis') # ~
 329     check_both_ways("\u00F7", "\x81\x80", 'shift_jis') #
 330     check_both_ways("\u25C7", "\x81\x9E", 'shift_jis') #
 331     check_both_ways("\u25C6", "\x81\x9F", 'shift_jis') #
 332     check_both_ways("\u25EF", "\x81\xFC", 'shift_jis') #
 333     check_both_ways("\u6A97", "\x9F\x40", 'shift_jis') # @
 334     check_both_ways("\u6BEF", "\x9F\x7E", 'shift_jis') # ~
 335     check_both_ways("\u9EBE", "\x9F\x80", 'shift_jis') #
 336     check_both_ways("\u6CBE", "\x9F\x9E", 'shift_jis') #
 337     check_both_ways("\u6CBA", "\x9F\x9F", 'shift_jis') #
 338     check_both_ways("\u6ECC", "\x9F\xFC", 'shift_jis') #
 339     check_both_ways("\u6F3E", "\xE0\x40", 'shift_jis') # @
 340     check_both_ways("\u70DD", "\xE0\x7E", 'shift_jis') # ~
 341     check_both_ways("\u70D9", "\xE0\x80", 'shift_jis') #
 342     check_both_ways("\u71FC", "\xE0\x9E", 'shift_jis') #
 343     check_both_ways("\u71F9", "\xE0\x9F", 'shift_jis') #
 344     check_both_ways("\u73F1", "\xE0\xFC", 'shift_jis') #
 345     assert_raise(Encoding::ConversionUndefined) { "\xEF\x40".encode("utf-8", 'shift_jis') }
 346     assert_raise(Encoding::ConversionUndefined) { "\xEF\x7E".encode("utf-8", 'shift_jis') }
 347     assert_raise(Encoding::ConversionUndefined) { "\xEF\x80".encode("utf-8", 'shift_jis') }
 348     assert_raise(Encoding::ConversionUndefined) { "\xEF\x9E".encode("utf-8", 'shift_jis') }
 349     assert_raise(Encoding::ConversionUndefined) { "\xEF\x9F".encode("utf-8", 'shift_jis') }
 350     assert_raise(Encoding::ConversionUndefined) { "\xEF\xFC".encode("utf-8", 'shift_jis') }
 351     assert_raise(Encoding::ConversionUndefined) { "\xF0\x40".encode("utf-8", 'shift_jis') }
 352     assert_raise(Encoding::ConversionUndefined) { "\xF0\x7E".encode("utf-8", 'shift_jis') }
 353     assert_raise(Encoding::ConversionUndefined) { "\xF0\x80".encode("utf-8", 'shift_jis') }
 354     assert_raise(Encoding::ConversionUndefined) { "\xF0\x9E".encode("utf-8", 'shift_jis') }
 355     assert_raise(Encoding::ConversionUndefined) { "\xF0\x9F".encode("utf-8", 'shift_jis') }
 356     assert_raise(Encoding::ConversionUndefined) { "\xF0\xFC".encode("utf-8", 'shift_jis') }
 357     check_both_ways("\u9ADC", "\xFC\x40", 'shift_jis') # @
 358     assert_raise(Encoding::ConversionUndefined) { "\xFC\x7E".encode("utf-8", 'shift_jis') }
 359     assert_raise(Encoding::ConversionUndefined) { "\xFC\x80".encode("utf-8", 'shift_jis') }
 360     assert_raise(Encoding::ConversionUndefined) { "\xFC\x9E".encode("utf-8", 'shift_jis') }
 361     assert_raise(Encoding::ConversionUndefined) { "\xFC\x9F".encode("utf-8", 'shift_jis') }
 362     assert_raise(Encoding::ConversionUndefined) { "\xFC\xFC".encode("utf-8", 'shift_jis') }
 363     check_both_ways("\u677E\u672C\u884C\u5F18", "\x8f\xbc\x96\x7b\x8d\x73\x8d\x4f", 'shift_jis') # {sO
 364     check_both_ways("\u9752\u5C71\u5B66\u9662\u5927\u5B66", "\x90\xC2\x8E\x52\x8A\x77\x89\x40\x91\xE5\x8A\x77", 'shift_jis') # Rw@w
 365     check_both_ways("\u795E\u6797\u7FA9\u535A", "\x90\x5F\x97\xD1\x8B\x60\x94\x8E", 'shift_jis') # _ы`
 366   end
 367
 368   def test_iso_2022_jp
 369     assert_raise(Encoding::InvalidByteSequence) { "\x1b(A".encode("utf-8", "iso-2022-jp") }
 370     assert_raise(Encoding::InvalidByteSequence) { "\x1b$(A".encode("utf-8", "iso-2022-jp") }
 371     assert_raise(Encoding::InvalidByteSequence) { "\x1b$C".encode("utf-8", "iso-2022-jp") }
 372     assert_raise(Encoding::InvalidByteSequence) { "\x0e".encode("utf-8", "iso-2022-jp") }
 373     assert_raise(Encoding::InvalidByteSequence) { "\x80".encode("utf-8", "iso-2022-jp") }
 374     assert_raise(Encoding::InvalidByteSequence) { "\x1b$(Dd!\x1b(B".encode("utf-8", "iso-2022-jp") }
 375     assert_raise(Encoding::ConversionUndefined) { "\u9299".encode("iso-2022-jp") }
 376     assert_raise(Encoding::ConversionUndefined) { "\uff71\uff72\uff73\uff74\uff75".encode("iso-2022-jp") }
 377     assert_raise(Encoding::InvalidByteSequence) { "\x1b(I12345\x1b(B".encode("utf-8", "iso-2022-jp") }
 378     assert_equal("\xA1\xA1".force_encoding("euc-jp"),
 379                  "\e$B!!\e(B".encode("EUC-JP", "ISO-2022-JP"))
 380     assert_equal("\e$B!!\e(B".force_encoding("ISO-2022-JP"),
 381                  "\xA1\xA1".encode("ISO-2022-JP", "EUC-JP"))
 382   end
 383
 384   def test_iso_2022_jp_1
 385     # check_both_ways("\u9299", "\x1b$(Dd!\x1b(B", "iso-2022-jp-1") # JIS X 0212 区68 点01 銙
 386   end
 387
 388   def test_unicode_public_review_issue_121 # see http://www.unicode.org/review/pr-121.html
 389     # assert_equal("\x00\x61\xFF\xFD\x00\x62".force_encoding('UTF-16BE'),
 390     #   "\x61\xF1\x80\x80\xE1\x80\xC2\x62".encode('UTF-16BE', 'UTF-8', invalid: :replace)) # option 1
 391     assert_equal("\x00\x61\xFF\xFD\xFF\xFD\xFF\xFD\x00\x62".force_encoding('UTF-16BE'),
 392       "\x61\xF1\x80\x80\xE1\x80\xC2\x62".encode('UTF-16BE', 'UTF-8', invalid: :replace)) # option 2
 393     assert_equal("\x61\x00\xFD\xFF\xFD\xFF\xFD\xFF\x62\x00".force_encoding('UTF-16LE'),
 394       "\x61\xF1\x80\x80\xE1\x80\xC2\x62".encode('UTF-16LE', 'UTF-8', invalid: :replace)) # option 2
 395     # assert_equal("\x00\x61\xFF\xFD\xFF\xFD\xFF\xFD\xFF\xFD\xFF\xFD\xFF\xFD\x00\x62".force_encoding('UTF-16BE'),
 396     # "\x61\xF1\x80\x80\xE1\x80\xC2\x62".encode('UTF-16BE', 'UTF-8', invalid: :replace)) # option 3
 397   end
 398 end