utf32.js 9.75 KB
'use strict';

var Buffer = require('safer-buffer').Buffer;

// == UTF32-LE/BE codec. ==========================================================

exports._utf32 = Utf32Codec;

function Utf32Codec(codecOptions, iconv) {
    this.iconv = iconv;
    this.bomAware = true;
    this.isLE = codecOptions.isLE;
}

exports.utf32le = { type: '_utf32', isLE: true };
exports.utf32be = { type: '_utf32', isLE: false };

// Aliases
exports.ucs4le = 'utf32le';
exports.ucs4be = 'utf32be';

Utf32Codec.prototype.encoder = Utf32Encoder;
Utf32Codec.prototype.decoder = Utf32Decoder;

// -- Encoding

function Utf32Encoder(options, codec) {
    this.isLE = codec.isLE;
    this.highSurrogate = 0;
}

Utf32Encoder.prototype.write = function(str) {
    var src = Buffer.from(str, 'ucs2');
    var dst = Buffer.alloc(src.length * 2);
    var write32 = this.isLE ? dst.writeUInt32LE : dst.writeUInt32BE;
    var offset = 0;

    for (var i = 0; i < src.length; i += 2) {
        var code = src.readUInt16LE(i);
        var isHighSurrogate = (0xD800 <= code && code < 0xDC00);
        var isLowSurrogate = (0xDC00 <= code && code < 0xE000);

        if (this.highSurrogate) {
            if (isHighSurrogate || !isLowSurrogate) {
                // There shouldn't be two high surrogates in a row, nor a high surrogate which isn't followed by a low
                // surrogate. If this happens, keep the pending high surrogate as a stand-alone semi-invalid character
                // (technically wrong, but expected by some applications, like Windows file names).
                write32.call(dst, this.highSurrogate, offset);
                offset += 4;
            }
            else {
                // Create 32-bit value from high and low surrogates;
                var codepoint = (((this.highSurrogate - 0xD800) << 10) | (code - 0xDC00)) + 0x10000;

                write32.call(dst, codepoint, offset);
                offset += 4;
                this.highSurrogate = 0;

                continue;
            }
        }

        if (isHighSurrogate)
            this.highSurrogate = code;
        else {
            // Even if the current character is a low surrogate, with no previous high surrogate, we'll
            // encode it as a semi-invalid stand-alone character for the same reasons expressed above for
            // unpaired high surrogates.
            write32.call(dst, code, offset);
            offset += 4;
            this.highSurrogate = 0;
        }
    }

    if (offset < dst.length)
        dst = dst.slice(0, offset);

    return dst;
};

Utf32Encoder.prototype.end = function() {
    // Treat any leftover high surrogate as a semi-valid independent character.
    if (!this.highSurrogate)
        return;

    var buf = Buffer.alloc(4);

    if (this.isLE)
        buf.writeUInt32LE(this.highSurrogate, 0);
    else
        buf.writeUInt32BE(this.highSurrogate, 0);

    this.highSurrogate = 0;

    return buf;
};

// -- Decoding

function Utf32Decoder(options, codec) {
    this.isLE = codec.isLE;
    this.badChar = codec.iconv.defaultCharUnicode.charCodeAt(0);
    this.overflow = [];
}

Utf32Decoder.prototype.write = function(src) {
    if (src.length === 0)
        return '';

    var i = 0;
    var codepoint = 0;
    var dst = Buffer.alloc(src.length + 4);
    var offset = 0;
    var isLE = this.isLE;
    var overflow = this.overflow;
    var badChar = this.badChar;

    if (overflow.length > 0) {
        for (; i < src.length && overflow.length < 4; i++)
            overflow.push(src[i]);
        
        if (overflow.length === 4) {
            // NOTE: codepoint is a signed int32 and can be negative.
            // NOTE: We copied this block from below to help V8 optimize it (it works with array, not buffer).
            if (isLE) {
                codepoint = overflow[i] | (overflow[i+1] << 8) | (overflow[i+2] << 16) | (overflow[i+3] << 24);
            } else {
                codepoint = overflow[i+3] | (overflow[i+2] << 8) | (overflow[i+1] << 16) | (overflow[i] << 24);
            }
            overflow.length = 0;

            offset = _writeCodepoint(dst, offset, codepoint, badChar);
        }
    }

    // Main loop. Should be as optimized as possible.
    for (; i < src.length - 3; i += 4) {
        // NOTE: codepoint is a signed int32 and can be negative.
        if (isLE) {
            codepoint = src[i] | (src[i+1] << 8) | (src[i+2] << 16) | (src[i+3] << 24);
        } else {
            codepoint = src[i+3] | (src[i+2] << 8) | (src[i+1] << 16) | (src[i] << 24);
        }
        offset = _writeCodepoint(dst, offset, codepoint, badChar);
    }

    // Keep overflowing bytes.
    for (; i < src.length; i++) {
        overflow.push(src[i]);
    }

    return dst.slice(0, offset).toString('ucs2');
};

function _writeCodepoint(dst, offset, codepoint, badChar) {
    // NOTE: codepoint is signed int32 and can be negative. We keep it that way to help V8 with optimizations.
    if (codepoint < 0 || codepoint > 0x10FFFF) {
        // Not a valid Unicode codepoint
        codepoint = badChar;
    } 

    // Ephemeral Planes: Write high surrogate.
    if (codepoint >= 0x10000) {
        codepoint -= 0x10000;

        var high = 0xD800 | (codepoint >> 10);
        dst[offset++] = high & 0xff;
        dst[offset++] = high >> 8;

        // Low surrogate is written below.
        var codepoint = 0xDC00 | (codepoint & 0x3FF);
    }

    // Write BMP char or low surrogate.
    dst[offset++] = codepoint & 0xff;
    dst[offset++] = codepoint >> 8;

    return offset;
};

Utf32Decoder.prototype.end = function() {
    this.overflow.length = 0;
};

// == UTF-32 Auto codec =============================================================
// Decoder chooses automatically from UTF-32LE and UTF-32BE using BOM and space-based heuristic.
// Defaults to UTF-32LE. http://en.wikipedia.org/wiki/UTF-32
// Encoder/decoder default can be changed: iconv.decode(buf, 'utf32', {defaultEncoding: 'utf-32be'});

// Encoder prepends BOM (which can be overridden with (addBOM: false}).

exports.utf32 = Utf32AutoCodec;
exports.ucs4 = 'utf32';

function Utf32AutoCodec(options, iconv) {
    this.iconv = iconv;
}

Utf32AutoCodec.prototype.encoder = Utf32AutoEncoder;
Utf32AutoCodec.prototype.decoder = Utf32AutoDecoder;

// -- Encoding

function Utf32AutoEncoder(options, codec) {
    options = options || {};

    if (options.addBOM === undefined)
        options.addBOM = true;

    this.encoder = codec.iconv.getEncoder(options.defaultEncoding || 'utf-32le', options);
}

Utf32AutoEncoder.prototype.write = function(str) {
    return this.encoder.write(str);
};

Utf32AutoEncoder.prototype.end = function() {
    return this.encoder.end();
};

// -- Decoding

function Utf32AutoDecoder(options, codec) {
    this.decoder = null;
    this.initialBufs = [];
    this.initialBufsLen = 0;
    this.options = options || {};
    this.iconv = codec.iconv;
}

Utf32AutoDecoder.prototype.write = function(buf) {
    if (!this.decoder) { 
        // Codec is not chosen yet. Accumulate initial bytes.
        this.initialBufs.push(buf);
        this.initialBufsLen += buf.length;

        if (this.initialBufsLen < 32) // We need more bytes to use space heuristic (see below)
            return '';

        // We have enough bytes -> detect endianness.
        var encoding = detectEncoding(this.initialBufs, this.options.defaultEncoding);
        this.decoder = this.iconv.getDecoder(encoding, this.options);

        var resStr = '';
        for (var i = 0; i < this.initialBufs.length; i++)
            resStr += this.decoder.write(this.initialBufs[i]);

        this.initialBufs.length = this.initialBufsLen = 0;
        return resStr;
    }

    return this.decoder.write(buf);
};

Utf32AutoDecoder.prototype.end = function() {
    if (!this.decoder) {
        var encoding = detectEncoding(this.initialBufs, this.options.defaultEncoding);
        this.decoder = this.iconv.getDecoder(encoding, this.options);

        var resStr = '';
        for (var i = 0; i < this.initialBufs.length; i++)
            resStr += this.decoder.write(this.initialBufs[i]);

        var trail = this.decoder.end();
        if (trail)
            resStr += trail;

        this.initialBufs.length = this.initialBufsLen = 0;
        return resStr;
    }

    return this.decoder.end();
};

function detectEncoding(bufs, defaultEncoding) {
    var b = [];
    var charsProcessed = 0;
    var invalidLE = 0, invalidBE = 0;   // Number of invalid chars when decoded as LE or BE.
    var bmpCharsLE = 0, bmpCharsBE = 0; // Number of BMP chars when decoded as LE or BE.

    outer_loop:
    for (var i = 0; i < bufs.length; i++) {
        var buf = bufs[i];
        for (var j = 0; j < buf.length; j++) {
            b.push(buf[j]);
            if (b.length === 4) {
                if (charsProcessed === 0) {
                    // Check BOM first.
                    if (b[0] === 0xFF && b[1] === 0xFE && b[2] === 0 && b[3] === 0) {
                        return 'utf-32le';
                    }
                    if (b[0] === 0 && b[1] === 0 && b[2] === 0xFE && b[3] === 0xFF) {
                        return 'utf-32be';
                    }
                }

                if (b[0] !== 0 || b[1] > 0x10) invalidBE++;
                if (b[3] !== 0 || b[2] > 0x10) invalidLE++;

                if (b[0] === 0 && b[1] === 0 && (b[2] !== 0 || b[3] !== 0)) bmpCharsBE++;
                if ((b[0] !== 0 || b[1] !== 0) && b[2] === 0 && b[3] === 0) bmpCharsLE++;

                b.length = 0;
                charsProcessed++;

                if (charsProcessed >= 100) {
                    break outer_loop;
                }
            }
        }
    }

    // Make decisions.
    if (bmpCharsBE - invalidBE > bmpCharsLE - invalidLE)  return 'utf-32be';
    if (bmpCharsBE - invalidBE < bmpCharsLE - invalidLE)  return 'utf-32le';

    // Couldn't decide (likely all zeros or not enough data).
    return defaultEncoding || 'utf-32le';
}