michael@0: // NOTE: Requires testharness.js
michael@0: // http://www.w3.org/2008/webapps/wiki/Harness
michael@0: 
michael@0: test(
michael@0:   function() {
michael@0:     var badStrings = [
michael@0:       { input: '\ud800', expected: '\ufffd' }, // Surrogate half
michael@0:       { input: '\udc00', expected: '\ufffd' }, // Surrogate half
michael@0:       { input: 'abc\ud800def', expected: 'abc\ufffddef' }, // Surrogate half
michael@0:       { input: 'abc\udc00def', expected: 'abc\ufffddef' }, // Surrogate half
michael@0:       { input: '\udc00\ud800', expected: '\ufffd\ufffd' } // Wrong order
michael@0:     ];
michael@0: 
michael@0:     badStrings.forEach(
michael@0:       function(t) {
michael@0:         var encoded = new TextEncoder('utf-8').encode(t.input);
michael@0:         var decoded = new TextDecoder('utf-8').decode(encoded);
michael@0:         assert_equals(t.expected, decoded);
michael@0:       });
michael@0:   },
michael@0:   "bad data"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function() {
michael@0:     var bad = [
michael@0:       { encoding: 'utf-8', input: [0xC0] }, // ends early
michael@0:       { encoding: 'utf-8', input: [0xC0, 0x00] }, // invalid trail
michael@0:       { encoding: 'utf-8', input: [0xC0, 0xC0] }, // invalid trail
michael@0:       { encoding: 'utf-8', input: [0xE0] }, // ends early
michael@0:       { encoding: 'utf-8', input: [0xE0, 0x00] }, // invalid trail
michael@0:       { encoding: 'utf-8', input: [0xE0, 0xC0] }, // invalid trail
michael@0:       { encoding: 'utf-8', input: [0xE0, 0x80, 0x00] }, // invalid trail
michael@0:       { encoding: 'utf-8', input: [0xE0, 0x80, 0xC0] }, // invalid trail
michael@0:       { encoding: 'utf-8', input: [0xFC, 0x80, 0x80, 0x80, 0x80, 0x80] }, // > 0x10FFFF
michael@0:       { encoding: 'utf-16le', input: [0x00] }, // truncated code unit
michael@0:       { encoding: 'utf-16le', input: [0x00, 0xd8] }, // surrogate half
michael@0:       { encoding: 'utf-16le', input: [0x00, 0xd8, 0x00, 0x00] }, // surrogate half
michael@0:       { encoding: 'utf-16le', input: [0x00, 0xdc, 0x00, 0x00] }, // trail surrogate
michael@0:       { encoding: 'utf-16le', input: [0x00, 0xdc, 0x00, 0xd8] }  // swapped surrogates
michael@0:       // TODO: Single byte encoding cases
michael@0:     ];
michael@0: 
michael@0:     bad.forEach(
michael@0:       function(t) {
michael@0:         assert_throws({name: 'EncodingError'}, function () {
michael@0:           new TextDecoder(t.encoding, {fatal: true}).decode(new Uint8Array(t.input));
michael@0:         });
michael@0:       });
michael@0:   },
michael@0:   "fatal flag"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function() {
michael@0:     var encodings = [
michael@0:       { label: 'utf-8', encoding: 'utf-8' },
michael@0:       { label: 'utf-16', encoding: 'utf-16le' },
michael@0:       { label: 'utf-16le', encoding: 'utf-16le' },
michael@0:       { label: 'utf-16be', encoding: 'utf-16be' },
michael@0:       { label: 'ascii', encoding: 'windows-1252' },
michael@0:       { label: 'iso-8859-1', encoding: 'windows-1252' }
michael@0:     ];
michael@0: 
michael@0:     encodings.forEach(
michael@0:       function(test) {
michael@0:         assert_equals(new TextDecoder(test.label.toLowerCase()).encoding, test.encoding);
michael@0:         assert_equals(new TextDecoder(test.label.toUpperCase()).encoding, test.encoding);
michael@0:       });
michael@0:   },
michael@0:   "Encoding names are case insensitive"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function() {
michael@0:     var utf8_bom = [0xEF, 0xBB, 0xBF];
michael@0:     var utf8 = [0x7A, 0xC2, 0xA2, 0xE6, 0xB0, 0xB4, 0xF0, 0x9D, 0x84, 0x9E, 0xF4, 0x8F, 0xBF, 0xBD];
michael@0: 
michael@0:     var utf16le_bom = [0xff, 0xfe];
michael@0:     var utf16le = [0x7A, 0x00, 0xA2, 0x00, 0x34, 0x6C, 0x34, 0xD8, 0x1E, 0xDD, 0xFF, 0xDB, 0xFD, 0xDF];
michael@0: 
michael@0:     var utf16be_bom = [0xfe, 0xff];
michael@0:     var utf16be = [0x00, 0x7A, 0x00, 0xA2, 0x6C, 0x34, 0xD8, 0x34, 0xDD, 0x1E, 0xDB, 0xFF, 0xDF, 0xFD];
michael@0: 
michael@0:     var string = "z\xA2\u6C34\uD834\uDD1E\uDBFF\uDFFD"; // z, cent, CJK water, G-Clef, Private-use character
michael@0: 
michael@0:     // missing BOMs
michael@0:     assert_equals(new TextDecoder('utf-8').decode(new Uint8Array(utf8)), string);
michael@0:     assert_equals(new TextDecoder('utf-16le').decode(new Uint8Array(utf16le)), string);
michael@0:     assert_equals(new TextDecoder('utf-16be').decode(new Uint8Array(utf16be)), string);
michael@0: 
michael@0:     // matching BOMs
michael@0:     assert_equals(new TextDecoder('utf-8').decode(new Uint8Array(utf8_bom.concat(utf8))), string);
michael@0:     assert_equals(new TextDecoder('utf-16le').decode(new Uint8Array(utf16le_bom.concat(utf16le))), string)
michael@0:     assert_equals(new TextDecoder('utf-16be').decode(new Uint8Array(utf16be_bom.concat(utf16be))), string);
michael@0: 
michael@0:     // matching BOMs split
michael@0:     var decoder8 = new TextDecoder('utf-8');
michael@0:     assert_equals(decoder8.decode(new Uint8Array(utf8_bom.slice(0, 1)), {stream: true}), '');
michael@0:     assert_equals(decoder8.decode(new Uint8Array(utf8_bom.slice(1).concat(utf8))), string);
michael@0:     assert_equals(decoder8.decode(new Uint8Array(utf8_bom.slice(0, 2)), {stream: true}), '');
michael@0:     assert_equals(decoder8.decode(new Uint8Array(utf8_bom.slice(2).concat(utf8))), string);
michael@0:     var decoder16le = new TextDecoder('utf-16le');
michael@0:     assert_equals(decoder16le.decode(new Uint8Array(utf16le_bom.slice(0, 1)), {stream: true}), '');
michael@0:     assert_equals(decoder16le.decode(new Uint8Array(utf16le_bom.slice(1).concat(utf16le))), string);
michael@0:     var decoder16be = new TextDecoder('utf-16be');
michael@0:     assert_equals(decoder16be.decode(new Uint8Array(utf16be_bom.slice(0, 1)), {stream: true}), '');
michael@0:     assert_equals(decoder16be.decode(new Uint8Array(utf16be_bom.slice(1).concat(utf16be))), string);
michael@0: 
michael@0:     // mismatching BOMs
michael@0:     assert_not_equals(new TextDecoder('utf-8').decode(new Uint8Array(utf16le_bom.concat(utf8))), string);
michael@0:     assert_not_equals(new TextDecoder('utf-8').decode(new Uint8Array(utf16be_bom.concat(utf8))), string);
michael@0:     assert_not_equals(new TextDecoder('utf-16le').decode(new Uint8Array(utf8_bom.concat(utf16le))), string);
michael@0:     assert_not_equals(new TextDecoder('utf-16le').decode(new Uint8Array(utf16be_bom.concat(utf16le))), string);
michael@0:     assert_not_equals(new TextDecoder('utf-16be').decode(new Uint8Array(utf8_bom.concat(utf16be))), string);
michael@0:     assert_not_equals(new TextDecoder('utf-16be').decode(new Uint8Array(utf16le_bom.concat(utf16be))), string);
michael@0:   },
michael@0:   "Byte-order marks"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function () {
michael@0:     assert_equals(new TextDecoder("utf-8").encoding, "utf-8"); // canonical case
michael@0:     assert_equals(new TextDecoder("UTF-16").encoding, "utf-16le"); // canonical case and name
michael@0:     assert_equals(new TextDecoder("UTF-16BE").encoding, "utf-16be"); // canonical case and name
michael@0:     assert_equals(new TextDecoder("iso8859-1").encoding, "windows-1252"); // canonical case and name
michael@0:     assert_equals(new TextDecoder("iso-8859-1").encoding, "windows-1252"); // canonical case and name
michael@0:   },
michael@0:   "Encoding names"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function () {
michael@0:     ["utf-8", "utf-16le", "utf-16be"].forEach(function (encoding) {
michael@0:       var string = "\x00123ABCabc\x80\xFF\u0100\u1000\uFFFD\uD800\uDC00\uDBFF\uDFFF";
michael@0:       var encoded = new TextEncoder(encoding).encode(string);
michael@0: 
michael@0:       for (var len = 1; len <= 5; ++len) {
michael@0:         var out = "", decoder = new TextDecoder(encoding);
michael@0:         for (var i = 0; i < encoded.length; i += len) {
michael@0:           var sub = [];
michael@0:           for (var j = i; j < encoded.length && j < i + len; ++j) {
michael@0:             sub.push(encoded[j]);
michael@0:           }
michael@0:           out += decoder.decode(new Uint8Array(sub), {stream: true});
michael@0:         }
michael@0:         out += decoder.decode();
michael@0:         assert_equals(out, string, "streaming decode " + encoding);
michael@0:       }
michael@0:     });
michael@0:   },
michael@0:   "Streaming Decode"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function () {
michael@0:     var jis = [0x82, 0xC9, 0x82, 0xD9, 0x82, 0xF1];
michael@0:     var expected = "\u306B\u307B\u3093"; // Nihon
michael@0:     assert_equals(new TextDecoder("shift_jis").decode(new Uint8Array(jis)), expected);
michael@0:   },
michael@0:   "Shift_JIS Decode"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function () {
michael@0:     var encodings = ["utf-8", "ibm866", "iso-8859-2", "iso-8859-3", "iso-8859-4", "iso-8859-5", "iso-8859-6", "iso-8859-7", "iso-8859-8", "iso-8859-8-i", "iso-8859-10", "iso-8859-13", "iso-8859-14", "iso-8859-15", "iso-8859-16", "koi8-r", "koi8-u", "macintosh", "windows-874", "windows-1250", "windows-1251", "windows-1252", "windows-1253", "windows-1254", "windows-1255", "windows-1256", "windows-1257", "windows-1258", "x-mac-cyrillic", "gbk", "gb18030", "hz-gb-2312", "big5", "euc-jp", "iso-2022-jp", "shift_jis", "euc-kr", "x-user-defined"];
michael@0: 
michael@0:     encodings.forEach(function (encoding) {
michael@0:       var string = '', bytes = [];
michael@0:       for (var i = 0; i < 128; ++i) {
michael@0: 
michael@0:         // Encodings that have escape codes in 0x00-0x7F
michael@0:         if (encoding === "hz-gb-2312" && i === 0x7E)
michael@0:           continue;
michael@0:         if (encoding === "iso-2022-jp" && i === 0x1B)
michael@0:           continue;
michael@0: 
michael@0:         string += String.fromCharCode(i);
michael@0:         bytes.push(i);
michael@0:       }
michael@0:       var ascii_encoded = new TextEncoder('utf-8').encode(string);
michael@0:       assert_equals(new TextDecoder(encoding).decode(ascii_encoded), string, encoding);
michael@0:       //assert_array_equals(new TextEncoder(encoding).encode(string), bytes, encoding);
michael@0:     });
michael@0:   },
michael@0:   "Supersets of ASCII decode ASCII correctly"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function () {
michael@0:     assert_throws({name: 'EncodingError'}, function() { new TextDecoder("utf-8", {fatal: true}).decode(new Uint8Array([0xff])); });
michael@0:     // This should not hang:
michael@0:     new TextDecoder("utf-8").decode(new Uint8Array([0xff]));
michael@0: 
michael@0:     assert_throws({name: 'EncodingError'}, function() { new TextDecoder("utf-16", {fatal: true}).decode(new Uint8Array([0x00])); });
michael@0:     // This should not hang:
michael@0:     new TextDecoder("utf-16").decode(new Uint8Array([0x00]));
michael@0: 
michael@0:     assert_throws({name: 'EncodingError'}, function() { new TextDecoder("utf-16be", {fatal: true}).decode(new Uint8Array([0x00])); });
michael@0:     // This should not hang:
michael@0:     new TextDecoder("utf-16be").decode(new Uint8Array([0x00]));
michael@0:   },
michael@0:   "Non-fatal errors at EOF"
michael@0: );
michael@0: 
michael@0: test(
michael@0:   function () {
michael@0: 
michael@0:     var utf_encodings = ["utf-8", "utf-16le", "utf-16be"];
michael@0: 
michael@0:     var legacy_encodings = ["ibm866", "iso-8859-2", "iso-8859-3", "iso-8859-4", "iso-8859-5", "iso-8859-6", "iso-8859-7", "iso-8859-8", "iso-8859-8-i", "iso-8859-10", "iso-8859-13", "iso-8859-14", "iso-8859-15", "iso-8859-16", "koi8-r", "koi8-u", "macintosh", "windows-874", "windows-1250", "windows-1251", "windows-1252", "windows-1253", "windows-1254", "windows-1255", "windows-1256", "windows-1257", "windows-1258", "x-mac-cyrillic", "gbk", "gb18030", "hz-gb-2312", "big5", "euc-jp", "iso-2022-jp", "shift_jis", "euc-kr", "x-user-defined"];
michael@0: 
michael@0:     utf_encodings.forEach(function(encoding) {
michael@0:       assert_equals(new TextDecoder(encoding).encoding, encoding);
michael@0:       assert_equals(new TextEncoder(encoding).encoding, encoding);
michael@0:     });
michael@0: 
michael@0:     legacy_encodings.forEach(function(encoding) {
michael@0:       assert_equals(new TextDecoder(encoding).encoding, encoding);
michael@0:       assert_throws({name: 'TypeError'}, function() { new TextEncoder(encoding); });
michael@0:     });
michael@0:   },
michael@0:   "Non-UTF encodings supported only for decode, not encode"
michael@0: );