dom/encoding/test/test_BOMEncoding.js

changeset 0
6474c204b198
     1.1 --- /dev/null	Thu Jan 01 00:00:00 1970 +0000
     1.2 +++ b/dom/encoding/test/test_BOMEncoding.js	Wed Dec 31 06:09:35 2014 +0100
     1.3 @@ -0,0 +1,132 @@
     1.4 +/*
     1.5 + * test_TextDecoderBOMEncoding.js
     1.6 + * bug 764234 tests
     1.7 +*/
     1.8 +
     1.9 +function runTextDecoderBOMEnoding()
    1.10 +{
    1.11 +  test(testDecodeValidBOMUTF16, "testDecodeValidBOMUTF16");
    1.12 +  test(testBOMEncodingUTF8, "testBOMEncodingUTF8");
    1.13 +  test(testMoreBOMEncoding, "testMoreBOMEncoding");
    1.14 +}
    1.15 +
    1.16 +function testDecodeValidBOMUTF16() {
    1.17 +  var expectedString = "\"\u0412\u0441\u0435 \u0441\u0447\u0430\u0441\u0442\u043B\u0438\u0432\u044B\u0435 \u0441\u0435\u043C\u044C\u0438 \u043F\u043E\u0445\u043E\u0436\u0438 \u0434\u0440\u0443\u0433 \u043D\u0430 \u0434\u0440\u0443\u0433\u0430, \u043A\u0430\u0436\u0434\u0430\u044F \u043D\u0435\u0441\u0447\u0430\u0441\u0442\u043B\u0438\u0432\u0430\u044F \u0441\u0435\u043C\u044C\u044F \u043D\u0435\u0441\u0447\u0430\u0441\u0442\u043B\u0438\u0432\u0430 \u043F\u043E-\u0441\u0432\u043E\u0435\u043C\u0443.\"";
    1.18 +
    1.19 +  // Testing UTF-16BE
    1.20 +  var data = [0xFE, 0xFF, 0x00, 0x22, 0x04, 0x12, 0x04, 0x41, 0x04, 0x35, 0x00, 0x20, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x4B, 0x04, 0x35, 0x00, 0x20, 0x04, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x38, 0x00, 0x20, 0x04, 0x3F, 0x04, 0x3E, 0x04, 0x45, 0x04, 0x3E, 0x04, 0x36, 0x04, 0x38, 0x00, 0x20, 0x04, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x00, 0x20, 0x04, 0x3D, 0x04, 0x30, 0x00, 0x20, 0x04, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x04, 0x30, 0x00, 0x2C, 0x00, 0x20, 0x04, 0x3A, 0x04, 0x30, 0x04, 0x36, 0x04, 0x34, 0x04, 0x30, 0x04, 0x4F, 0x00, 0x20, 0x04, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x04, 0x4F, 0x00, 0x20, 0x04, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x4F, 0x00, 0x20, 0x04, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x00, 0x20, 0x04, 0x3F, 0x04, 0x3E, 0x00, 0x2D, 0x04, 0x41, 0x04, 0x32, 0x04, 0x3E, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x43, 0x00, 0x2E, 0x00, 0x22];
    1.21 +  testBOMCharset({encoding: "utf-16be", data: data, expected: expectedString,
    1.22 +                  msg: "decoder valid UTF-16BE test."});
    1.23 +}
    1.24 +
    1.25 +function testBOMEncodingUTF8() {
    1.26 +
    1.27 +  // basic utf-8 test with valid encoding and byte stream. no byte om provided.
    1.28 +  var data = [0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27];
    1.29 +  var expectedString = " !\"#$%&'";
    1.30 +  testBOMCharset({encoding: "utf-8", data: data, expected: expectedString,
    1.31 +                  msg: "utf-8 encoding."});
    1.32 +
    1.33 +  // test valid encoding provided with valid byte OM also provided.
    1.34 +  data = [0xEF, 0xBB, 0xBF, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27];
    1.35 +  expectedString = " !\"#$%&'";
    1.36 +  testBOMCharset({encoding: "utf-8", data: data, expected: expectedString,
    1.37 +                  msg: "valid utf-8 encoding provided with VALID utf-8 BOM test."});
    1.38 +
    1.39 +  // test valid encoding provided with invalid byte OM also provided.
    1.40 +  data = [0xFF, 0xFE, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27];
    1.41 +  testBOMCharset({encoding: "utf-8", fatal: true, data: data, error: "EncodingError",
    1.42 +                  msg: "valid utf-8 encoding provided with invalid utf-8 fatal BOM test."});
    1.43 +
    1.44 +  // test valid encoding provided with invalid byte OM also provided.
    1.45 +  data = [0xFF, 0xFE, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27];
    1.46 +  expectedString = "\ufffd\ufffd !\"#$%&'";
    1.47 +  testBOMCharset({encoding: "utf-8", data: data, expected: expectedString,
    1.48 +                  msg: "valid utf-8 encoding provided with invalid utf-8 BOM test."});
    1.49 +
    1.50 +  // test empty encoding provided with invalid byte OM also provided.
    1.51 +  data = [0xFF, 0xFE, 0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27];
    1.52 +  testBOMCharset({encoding: "", data: data, error: "TypeError",
    1.53 +                  msg: "empty encoding provided with invalid utf-8 BOM test."});
    1.54 +}
    1.55 +
    1.56 +function testMoreBOMEncoding() {
    1.57 +  var expectedString = "\"\u0412\u0441\u0435 \u0441\u0447\u0430\u0441\u0442\u043B\u0438\u0432\u044B\u0435 \u0441\u0435\u043C\u044C\u0438 \u043F\u043E\u0445\u043E\u0436\u0438 \u0434\u0440\u0443\u0433 \u043D\u0430 \u0434\u0440\u0443\u0433\u0430, \u043A\u0430\u0436\u0434\u0430\u044F \u043D\u0435\u0441\u0447\u0430\u0441\u0442\u043B\u0438\u0432\u0430\u044F \u0441\u0435\u043C\u044C\u044F \u043D\u0435\u0441\u0447\u0430\u0441\u0442\u043B\u0438\u0432\u0430 \u043F\u043E-\u0441\u0432\u043E\u0435\u043C\u0443.\"";
    1.58 +
    1.59 + // Testing user provided encoding is UTF-16BE & bom encoding is utf-16le
    1.60 +  var data = [0xFF, 0xFE, 0x00, 0x22, 0x04, 0x12, 0x04, 0x41, 0x04, 0x35, 0x00, 0x20, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x4B, 0x04, 0x35, 0x00, 0x20, 0x04, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x38, 0x00, 0x20, 0x04, 0x3F, 0x04, 0x3E, 0x04, 0x45, 0x04, 0x3E, 0x04, 0x36, 0x04, 0x38, 0x00, 0x20, 0x04, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x00, 0x20, 0x04, 0x3D, 0x04, 0x30, 0x00, 0x20, 0x04, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x04, 0x30, 0x00, 0x2C, 0x00, 0x20, 0x04, 0x3A, 0x04, 0x30, 0x04, 0x36, 0x04, 0x34, 0x04, 0x30, 0x04, 0x4F, 0x00, 0x20, 0x04, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x04, 0x4F, 0x00, 0x20, 0x04, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x4F, 0x00, 0x20, 0x04, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x00, 0x20, 0x04, 0x3F, 0x04, 0x3E, 0x00, 0x2D, 0x04, 0x41, 0x04, 0x32, 0x04, 0x3E, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x43, 0x00, 0x2E, 0x00, 0x22];
    1.61 +
    1.62 +  testBOMCharset({encoding: "utf-16be", fatal: true, data: data, expected: "\ufffe" + expectedString,
    1.63 +                  msg: "test decoder invalid BOM encoding for utf-16be fatal."});
    1.64 +
    1.65 +  testBOMCharset({encoding: "utf-16be", data: data, expected: "\ufffe" + expectedString,
    1.66 +                  msg: "test decoder invalid BOM encoding for utf-16be."});
    1.67 +
    1.68 +  // Testing user provided encoding is UTF-16LE & bom encoding is utf-16be
    1.69 +  var dataUTF16 = [0xFE, 0xFF, 0x22, 0x00, 0x12, 0x04, 0x41, 0x04, 0x35, 0x04, 0x20, 0x00, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x4B, 0x04, 0x35, 0x04, 0x20, 0x00, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x38, 0x04, 0x20, 0x00, 0x3F, 0x04, 0x3E, 0x04, 0x45, 0x04, 0x3E, 0x04, 0x36, 0x04, 0x38, 0x04, 0x20, 0x00, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x04, 0x20, 0x00, 0x3D, 0x04, 0x30, 0x04, 0x20, 0x00, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x04, 0x30, 0x04, 0x2C, 0x00, 0x20, 0x00, 0x3A, 0x04, 0x30, 0x04, 0x36, 0x04, 0x34, 0x04, 0x30, 0x04, 0x4F, 0x04, 0x20, 0x00, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x04, 0x4F, 0x04, 0x20, 0x00, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x4F, 0x04, 0x20, 0x00, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x04, 0x20, 0x00, 0x3F, 0x04, 0x3E, 0x04, 0x2D, 0x00, 0x41, 0x04, 0x32, 0x04, 0x3E, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x43, 0x04, 0x2E, 0x00, 0x22, 0x00];
    1.70 +  testBOMCharset({encoding: "utf-16le", fatal: true, data: dataUTF16, expected: "\ufffe" + expectedString,
    1.71 +                  msg: "test decoder invalid BOM encoding for utf-16le fatal."});
    1.72 +
    1.73 +  testBOMCharset({encoding: "utf-16le", data: dataUTF16, expected: "\ufffe" + expectedString,
    1.74 +                  msg: "test decoder invalid BOM encoding for utf-16le."});
    1.75 +
    1.76 +  // Testing user provided encoding is UTF-16 & bom encoding is utf-16be
    1.77 +  testBOMCharset({encoding: "utf-16", fatal: true, data: dataUTF16, expected: "\ufffe" + expectedString,
    1.78 +                  msg: "test decoder invalid BOM encoding for utf-16 fatal."});
    1.79 +
    1.80 +  testBOMCharset({encoding: "utf-16", data: dataUTF16, expected: "\ufffe" + expectedString,
    1.81 +                  msg: "test decoder invalid BOM encoding for utf-16."});
    1.82 +
    1.83 +  // Testing user provided encoding is UTF-16 & bom encoding is utf-16le
    1.84 +  dataUTF16 = [0xFF, 0xFE, 0x22, 0x00, 0x12, 0x04, 0x41, 0x04, 0x35, 0x04, 0x20, 0x00, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x4B, 0x04, 0x35, 0x04, 0x20, 0x00, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x38, 0x04, 0x20, 0x00, 0x3F, 0x04, 0x3E, 0x04, 0x45, 0x04, 0x3E, 0x04, 0x36, 0x04, 0x38, 0x04, 0x20, 0x00, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x04, 0x20, 0x00, 0x3D, 0x04, 0x30, 0x04, 0x20, 0x00, 0x34, 0x04, 0x40, 0x04, 0x43, 0x04, 0x33, 0x04, 0x30, 0x04, 0x2C, 0x00, 0x20, 0x00, 0x3A, 0x04, 0x30, 0x04, 0x36, 0x04, 0x34, 0x04, 0x30, 0x04, 0x4F, 0x04, 0x20, 0x00, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x04, 0x4F, 0x04, 0x20, 0x00, 0x41, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x4C, 0x04, 0x4F, 0x04, 0x20, 0x00, 0x3D, 0x04, 0x35, 0x04, 0x41, 0x04, 0x47, 0x04, 0x30, 0x04, 0x41, 0x04, 0x42, 0x04, 0x3B, 0x04, 0x38, 0x04, 0x32, 0x04, 0x30, 0x04, 0x20, 0x00, 0x3F, 0x04, 0x3E, 0x04, 0x2D, 0x00, 0x41, 0x04, 0x32, 0x04, 0x3E, 0x04, 0x35, 0x04, 0x3C, 0x04, 0x43, 0x04, 0x2E, 0x00, 0x22, 0x00];
    1.85 +  testBOMCharset({encoding: "utf-16", fatal: true, data: dataUTF16, expected: expectedString,
    1.86 +                  msg: "test decoder BOM encoding for utf-16 fatal."});
    1.87 +
    1.88 +  testBOMCharset({encoding: "utf-16", data: dataUTF16, expected: expectedString,
    1.89 +                  msg: "test decoder BOM encoding for utf-16."});
    1.90 +
    1.91 +  // Testing user provided encoding is UTF-8 & bom encoding is utf-16be
    1.92 +  data = [0xFE, 0xFF, 0x22, 0xd0, 0x92, 0xd1, 0x81, 0xd0, 0xb5, 0x20, 0xd1, 0x81, 0xd1, 0x87, 0xd0, 0xb0, 0xd1, 0x81, 0xd1, 0x82, 0xd0, 0xbb, 0xd0, 0xb8, 0xd0, 0xb2, 0xd1, 0x8b, 0xd0, 0xb5, 0x20, 0xd1, 0x81, 0xd0, 0xb5, 0xd0, 0xbc, 0xd1, 0x8c, 0xd0, 0xb8, 0x20, 0xd0, 0xbf, 0xd0, 0xbe, 0xd1, 0x85, 0xd0, 0xbe, 0xd0, 0xb6, 0xd0, 0xb8, 0x20, 0xd0, 0xb4, 0xd1, 0x80, 0xd1, 0x83, 0xd0, 0xb3, 0x20, 0xd0, 0xbd, 0xd0, 0xb0, 0x20, 0xd0, 0xb4, 0xd1, 0x80, 0xd1, 0x83, 0xd0, 0xb3, 0xd0, 0xb0, 0x2c, 0x20, 0xd0, 0xba, 0xd0, 0xb0, 0xd0, 0xb6, 0xd0, 0xb4, 0xd0, 0xb0, 0xd1, 0x8f, 0x20, 0xd0, 0xbd, 0xd0, 0xb5, 0xd1, 0x81, 0xd1, 0x87, 0xd0, 0xb0, 0xd1, 0x81, 0xd1, 0x82, 0xd0, 0xbb, 0xd0, 0xb8, 0xd0, 0xb2, 0xd0, 0xb0, 0xd1, 0x8f, 0x20, 0xd1, 0x81, 0xd0, 0xb5, 0xd0, 0xbc, 0xd1, 0x8c, 0xd1, 0x8f, 0x20, 0xd0, 0xbd, 0xd0, 0xb5, 0xd1, 0x81, 0xd1, 0x87, 0xd0, 0xb0, 0xd1, 0x81, 0xd1, 0x82, 0xd0, 0xbb, 0xd0, 0xb8, 0xd0, 0xb2, 0xd0, 0xb0, 0x20, 0xd0, 0xbf, 0xd0, 0xbe, 0x2d, 0xd1, 0x81, 0xd0, 0xb2, 0xd0, 0xbe, 0xd0, 0xb5, 0xd0, 0xbc, 0xd1, 0x83, 0x2e, 0x22];
    1.93 +
    1.94 +  testBOMCharset({encoding: "utf-8", fatal: true, data: data, error: "EncodingError",
    1.95 +                  msg: "test decoder invalid BOM encoding for valid utf-8 fatal provided label."});
    1.96 +
    1.97 +  testBOMCharset({encoding: "utf-8", data: data, expected: "\ufffd\ufffd" + expectedString,
    1.98 +                  msg: "test decoder invalid BOM encoding for valid utf-8 provided label."});
    1.99 +
   1.100 +  // Testing user provided encoding is non-UTF & bom encoding is utf-16be
   1.101 +  data = [0xFE, 0xFF, 0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, 0xa8, 0xa9, 0xaa, 0xab, 0xac, 0xad, 0xaf, 0xb0, 0xb1, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6, 0xb7, 0xb8, 0xb9, 0xba, 0xbb, 0xbc, 0xbd, 0xbe, 0xbf, 0xc0, 0xc1, 0xc2, 0xc3, 0xc4, 0xc5, 0xc6, 0xc7, 0xc8, 0xc9, 0xca, 0xcb, 0xcc, 0xcd, 0xce, 0xcf, 0xd0, 0xd1, 0xd3, 0xd4, 0xd5, 0xd6, 0xd7, 0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0xdd, 0xde, 0xdf, 0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef, 0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe];
   1.102 +
   1.103 +  expectedString = "\u03CE\uFFFD\u2019\xA3\u20AC\u20AF\xA6\xA7\xA8\xA9\u037A\xAB\xAC\xAD\u2015"
   1.104 +                 + "\xB0\xB1\xB2\xB3\u0384\u0385\u0386\xB7\u0388\u0389\u038A\xBB\u038C\xBD\u038E\u038F"
   1.105 +                 + "\u0390\u0391\u0392\u0393\u0394\u0395\u0396\u0397\u0398\u0399\u039A\u039B\u039C\u039D\u039E\u039F"
   1.106 +                 + "\u03A0\u03A1\u03A3\u03A4\u03A5\u03A6\u03A7\u03A8\u03A9\u03AA\u03AB\u03AC\u03AD\u03AE\u03AF"
   1.107 +                 + "\u03B0\u03B1\u03B2\u03B3\u03B4\u03B5\u03B6\u03B7\u03B8\u03B9\u03BA\u03BB\u03BC\u03BD\u03BE\u03BF"
   1.108 +                 + "\u03C0\u03C1\u03C2\u03C3\u03C4\u03C5\u03C6\u03C7\u03C8\u03C9\u03CA\u03CB\u03CC\u03CD\u03CE";
   1.109 +
   1.110 +  testBOMCharset({encoding: "greek", fatal: true, data: data, error: "EncodingError",
   1.111 +                  msg: "test decoder encoding provided with invalid BOM encoding for greek."});
   1.112 +
   1.113 +  testBOMCharset({encoding: "greek", data: data, expected: expectedString,
   1.114 +                  msg: "test decoder encoding provided with invalid BOM encoding for greek."});
   1.115 +}
   1.116 +
   1.117 +function testBOMCharset(test)
   1.118 +{
   1.119 +  var outText;
   1.120 +  try {
   1.121 +    var decoder = 'fatal' in test ?
   1.122 +      new TextDecoder(test.encoding, {fatal: test.fatal}) :
   1.123 +      new TextDecoder(test.encoding);
   1.124 +    outText = decoder.decode(new Uint8Array(test.data));
   1.125 +  } catch (e) {
   1.126 +    assert_equals(e.name, test.error, test.msg);
   1.127 +    return;
   1.128 +  }
   1.129 +  assert_true(!test.error, test.msg);
   1.130 +
   1.131 +  if (outText !== test.expected) {
   1.132 +    assert_equals(escape(outText), escape(test.expected), test.msg + " Code points do not match expected code points.");
   1.133 +  }
   1.134 +}
   1.135 +

mercurial