twitst4tz

utf16.js (5011B)

---

 "use strict";
 var Buffer = require("safer-buffer").Buffer;
 
 // Note: UTF16-LE (or UCS2) codec is Node.js native. See encodings/internal.js
 
 // == UTF16-BE codec. ==========================================================
 
 exports.utf16be = Utf16BECodec;
 function Utf16BECodec() {
 }
 
 Utf16BECodec.prototype.encoder = Utf16BEEncoder;
 Utf16BECodec.prototype.decoder = Utf16BEDecoder;
 Utf16BECodec.prototype.bomAware = true;
 
 
 // -- Encoding
 
 function Utf16BEEncoder() {
 }
 
 Utf16BEEncoder.prototype.write = function(str) {
     var buf = Buffer.from(str, 'ucs2');
     for (var i = 0; i < buf.length; i += 2) {
         var tmp = buf[i]; buf[i] = buf[i+1]; buf[i+1] = tmp;
     }
     return buf;
 }
 
 Utf16BEEncoder.prototype.end = function() {
 }
 
 
 // -- Decoding
 
 function Utf16BEDecoder() {
     this.overflowByte = -1;
 }
 
 Utf16BEDecoder.prototype.write = function(buf) {
     if (buf.length == 0)
         return '';
 
     var buf2 = Buffer.alloc(buf.length + 1),
         i = 0, j = 0;
 
     if (this.overflowByte !== -1) {
         buf2[0] = buf[0];
         buf2[1] = this.overflowByte;
         i = 1; j = 2;
     }
 
     for (; i < buf.length-1; i += 2, j+= 2) {
         buf2[j] = buf[i+1];
         buf2[j+1] = buf[i];
     }
 
     this.overflowByte = (i == buf.length-1) ? buf[buf.length-1] : -1;
 
     return buf2.slice(0, j).toString('ucs2');
 }
 
 Utf16BEDecoder.prototype.end = function() {
 }
 
 
 // == UTF-16 codec =============================================================
 // Decoder chooses automatically from UTF-16LE and UTF-16BE using BOM and space-based heuristic.
 // Defaults to UTF-16LE, as it's prevalent and default in Node.
 // http://en.wikipedia.org/wiki/UTF-16 and http://encoding.spec.whatwg.org/#utf-16le
 // Decoder default can be changed: iconv.decode(buf, 'utf16', {defaultEncoding: 'utf-16be'});
 
 // Encoder uses UTF-16LE and prepends BOM (which can be overridden with addBOM: false).
 
 exports.utf16 = Utf16Codec;
 function Utf16Codec(codecOptions, iconv) {
     this.iconv = iconv;
 }
 
 Utf16Codec.prototype.encoder = Utf16Encoder;
 Utf16Codec.prototype.decoder = Utf16Decoder;
 
 
 // -- Encoding (pass-through)
 
 function Utf16Encoder(options, codec) {
     options = options || {};
     if (options.addBOM === undefined)
         options.addBOM = true;
     this.encoder = codec.iconv.getEncoder('utf-16le', options);
 }
 
 Utf16Encoder.prototype.write = function(str) {
     return this.encoder.write(str);
 }
 
 Utf16Encoder.prototype.end = function() {
     return this.encoder.end();
 }
 
 
 // -- Decoding
 
 function Utf16Decoder(options, codec) {
     this.decoder = null;
     this.initialBytes = [];
     this.initialBytesLen = 0;
 
     this.options = options || {};
     this.iconv = codec.iconv;
 }
 
 Utf16Decoder.prototype.write = function(buf) {
     if (!this.decoder) {
         // Codec is not chosen yet. Accumulate initial bytes.
         this.initialBytes.push(buf);
         this.initialBytesLen += buf.length;
         
         if (this.initialBytesLen < 16) // We need more bytes to use space heuristic (see below)
             return '';
 
         // We have enough bytes -> detect endianness.
         var buf = Buffer.concat(this.initialBytes),
             encoding = detectEncoding(buf, this.options.defaultEncoding);
         this.decoder = this.iconv.getDecoder(encoding, this.options);
         this.initialBytes.length = this.initialBytesLen = 0;
     }
 
     return this.decoder.write(buf);
 }
 
 Utf16Decoder.prototype.end = function() {
     if (!this.decoder) {
         var buf = Buffer.concat(this.initialBytes),
             encoding = detectEncoding(buf, this.options.defaultEncoding);
         this.decoder = this.iconv.getDecoder(encoding, this.options);
 
         var res = this.decoder.write(buf),
             trail = this.decoder.end();
 
         return trail ? (res + trail) : res;
     }
     return this.decoder.end();
 }
 
 function detectEncoding(buf, defaultEncoding) {
     var enc = defaultEncoding || 'utf-16le';
 
     if (buf.length >= 2) {
         // Check BOM.
         if (buf[0] == 0xFE && buf[1] == 0xFF) // UTF-16BE BOM
             enc = 'utf-16be';
         else if (buf[0] == 0xFF && buf[1] == 0xFE) // UTF-16LE BOM
             enc = 'utf-16le';
         else {
             // No BOM found. Try to deduce encoding from initial content.
             // Most of the time, the content has ASCII chars (U+00**), but the opposite (U+**00) is uncommon.
             // So, we count ASCII as if it was LE or BE, and decide from that.
             var asciiCharsLE = 0, asciiCharsBE = 0, // Counts of chars in both positions
                 _len = Math.min(buf.length - (buf.length % 2), 64); // Len is always even.
 
             for (var i = 0; i < _len; i += 2) {
                 if (buf[i] === 0 && buf[i+1] !== 0) asciiCharsBE++;
                 if (buf[i] !== 0 && buf[i+1] === 0) asciiCharsLE++;
             }
 
             if (asciiCharsBE > asciiCharsLE)
                 enc = 'utf-16be';
             else if (asciiCharsBE < asciiCharsLE)
                 enc = 'utf-16le';
         }
     }
 
     return enc;
 }