From 7332ffb1f08c9414119aa0a59ec8334c7599bfd8 Mon Sep 17 00:00:00 2001 From: Wojciech Mandrysz Date: Mon, 3 Oct 2016 18:59:34 +0200 Subject: JS: Replaced fromCodePoint/codePointAt with fromCharCode/charCodeAt because of functions limited availability, fixed typo in tests. --- js/binary/decoder.js | 21 +++++++++------------ js/binary/decoder_test.js | 6 +++--- js/binary/encoder.js | 15 ++++++++++++--- 3 files changed, 24 insertions(+), 18 deletions(-) (limited to 'js') diff --git a/js/binary/decoder.js b/js/binary/decoder.js index 62f7b8b6..e4fb9148 100644 --- a/js/binary/decoder.js +++ b/js/binary/decoder.js @@ -905,12 +905,12 @@ jspb.BinaryDecoder.prototype.readString = function(length) { var bytes = this.bytes_; var cursor = this.cursor_; var end = cursor + length; - var codepoints = []; + var codeUnits = []; while (cursor < end) { var c = bytes[cursor++]; if (c < 128) { // Regular 7-bit ASCII. - codepoints.push(c); + codeUnits.push(c); } else if (c < 192) { // UTF-8 continuation mark. We are out of sync. This // might happen if we attempted to read a character @@ -918,11 +918,11 @@ jspb.BinaryDecoder.prototype.readString = function(length) { continue; } else if (c < 224) { // UTF-8 with two bytes. var c2 = bytes[cursor++]; - codepoints.push(((c & 31) << 6) | (c2 & 63)); + codeUnits.push(((c & 31) << 6) | (c2 & 63)); } else if (c < 240) { // UTF-8 with three bytes. var c2 = bytes[cursor++]; var c3 = bytes[cursor++]; - codepoints.push(((c & 15) << 12) | ((c2 & 63) << 6) | (c3 & 63)); + codeUnits.push(((c & 15) << 12) | ((c2 & 63) << 6) | (c3 & 63)); } else if (c < 248) { // UTF-8 with 4 bytes. var c2 = bytes[cursor++]; var c3 = bytes[cursor++]; @@ -932,20 +932,17 @@ jspb.BinaryDecoder.prototype.readString = function(length) { var codepoint = ((c & 7) << 18) | ((c2 & 63) << 12) | ((c3 & 63) << 6) | (c4 & 63); // Surrogates formula from wikipedia. // 1. Subtract 0x10000 from codepoint - codepoint -= 65536; + codepoint -= 0x10000; // 2. Split this into the high 10-bit value and the low 10-bit value - var low = codepoint & 1023; - var high = (codepoint >> 10) & 1023; // 3. Add 0xD800 to the high value to form the high surrogate - high += 55296; // 4. Add 0xDC00 to the low value to form the low surrogate: - low += 56320; - codepoints.push(high); - codepoints.push(low); + var low = (codepoint & 1023) + 0xDC00; + var high = ((codepoint >> 10) & 1023) + 0xD800; + codeUnits.push(high, low) } } - var result = String.fromCodePoint.apply(null, codepoints); + var result = String.fromCharCode.apply(null, codeUnits); this.cursor_ = cursor; return result; }; diff --git a/js/binary/decoder_test.js b/js/binary/decoder_test.js index 12da72a7..9f947b99 100644 --- a/js/binary/decoder_test.js +++ b/js/binary/decoder_test.js @@ -218,19 +218,19 @@ describe('binaryDecoderTest', function() { var ascii = "ASCII should work in 3, 2, 1..." var utf8_two_bytes = "©"; - var utf8_tree_bytes = "❄"; + var utf8_three_bytes = "❄"; var utf8_four_bytes = "😁"; encoder.writeString(ascii); encoder.writeString(utf8_two_bytes); - encoder.writeString(utf8_tree_bytes); + encoder.writeString(utf8_three_bytes); encoder.writeString(utf8_four_bytes); var decoder = jspb.BinaryDecoder.alloc(encoder.end()); assertEquals(ascii, decoder.readString(ascii.length)); assertEquals(utf8_two_bytes, decoder.readString(utf8_two_bytes.length)); - assertEquals(utf8_tree_bytes, decoder.readString(utf8_tree_bytes.length)); + assertEquals(utf8_three_bytes, decoder.readString(utf8_three_bytes.length)); assertEquals(utf8_four_bytes, decoder.readString(utf8_four_bytes.length)); }); diff --git a/js/binary/encoder.js b/js/binary/encoder.js index 59c4ccb9..fe5e34e9 100644 --- a/js/binary/encoder.js +++ b/js/binary/encoder.js @@ -409,10 +409,19 @@ jspb.BinaryEncoder.prototype.writeFixedHash64 = function(hash) { */ jspb.BinaryEncoder.prototype.writeString = function(value) { var oldLength = this.buffer_.length; - - // UTF16 to UTF8 conversion loop swiped from goog.crypt.stringToUtf8ByteArray. + for (var i = 0; i < value.length; i++) { - var c = value.codePointAt(i); + + var c = value.charCodeAt(i); + // Look for surrogates + if (c >= 0xD800 && c <= 0xDBFF && i + 1 < value.length) { + var second = value.charCodeAt(i + 1); + if (second >= 0xDC00 && second <= 0xDFFF) { // low surrogate + // http://mathiasbynens.be/notes/javascript-encoding#surrogate-formulae + c = (c - 0xD800) * 0x400 + second - 0xDC00 + 0x10000; + } + } + if (c < 128) { this.buffer_.push(c); } else if (c < 2048) { -- cgit v1.2.3