123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188 |
- "use strict";
- var Buffer = require("buffer").Buffer;
- // Export Node.js internal encodings.
- module.exports = {
- // Encodings
- utf8: { type: "_internal", bomAware: true},
- cesu8: { type: "_internal", bomAware: true},
- unicode11utf8: "utf8",
- ucs2: { type: "_internal", bomAware: true},
- utf16le: "ucs2",
- binary: { type: "_internal" },
- base64: { type: "_internal" },
- hex: { type: "_internal" },
- // Codec.
- _internal: InternalCodec,
- };
- //------------------------------------------------------------------------------
- function InternalCodec(codecOptions, iconv) {
- this.enc = codecOptions.encodingName;
- this.bomAware = codecOptions.bomAware;
- if (this.enc === "base64")
- this.encoder = InternalEncoderBase64;
- else if (this.enc === "cesu8") {
- this.enc = "utf8"; // Use utf8 for decoding.
- this.encoder = InternalEncoderCesu8;
- // Add decoder for versions of Node not supporting CESU-8
- if (new Buffer('eda0bdedb2a9', 'hex').toString() !== '💩') {
- this.decoder = InternalDecoderCesu8;
- this.defaultCharUnicode = iconv.defaultCharUnicode;
- }
- }
- }
- InternalCodec.prototype.encoder = InternalEncoder;
- InternalCodec.prototype.decoder = InternalDecoder;
- //------------------------------------------------------------------------------
- // We use node.js internal decoder. Its signature is the same as ours.
- var StringDecoder = require('string_decoder').StringDecoder;
- if (!StringDecoder.prototype.end) // Node v0.8 doesn't have this method.
- StringDecoder.prototype.end = function() {};
- function InternalDecoder(options, codec) {
- StringDecoder.call(this, codec.enc);
- }
- InternalDecoder.prototype = StringDecoder.prototype;
- //------------------------------------------------------------------------------
- // Encoder is mostly trivial
- function InternalEncoder(options, codec) {
- this.enc = codec.enc;
- }
- InternalEncoder.prototype.write = function(str) {
- return new Buffer(str, this.enc);
- }
- InternalEncoder.prototype.end = function() {
- }
- //------------------------------------------------------------------------------
- // Except base64 encoder, which must keep its state.
- function InternalEncoderBase64(options, codec) {
- this.prevStr = '';
- }
- InternalEncoderBase64.prototype.write = function(str) {
- str = this.prevStr + str;
- var completeQuads = str.length - (str.length % 4);
- this.prevStr = str.slice(completeQuads);
- str = str.slice(0, completeQuads);
- return new Buffer(str, "base64");
- }
- InternalEncoderBase64.prototype.end = function() {
- return new Buffer(this.prevStr, "base64");
- }
- //------------------------------------------------------------------------------
- // CESU-8 encoder is also special.
- function InternalEncoderCesu8(options, codec) {
- }
- InternalEncoderCesu8.prototype.write = function(str) {
- var buf = new Buffer(str.length * 3), bufIdx = 0;
- for (var i = 0; i < str.length; i++) {
- var charCode = str.charCodeAt(i);
- // Naive implementation, but it works because CESU-8 is especially easy
- // to convert from UTF-16 (which all JS strings are encoded in).
- if (charCode < 0x80)
- buf[bufIdx++] = charCode;
- else if (charCode < 0x800) {
- buf[bufIdx++] = 0xC0 + (charCode >>> 6);
- buf[bufIdx++] = 0x80 + (charCode & 0x3f);
- }
- else { // charCode will always be < 0x10000 in javascript.
- buf[bufIdx++] = 0xE0 + (charCode >>> 12);
- buf[bufIdx++] = 0x80 + ((charCode >>> 6) & 0x3f);
- buf[bufIdx++] = 0x80 + (charCode & 0x3f);
- }
- }
- return buf.slice(0, bufIdx);
- }
- InternalEncoderCesu8.prototype.end = function() {
- }
- //------------------------------------------------------------------------------
- // CESU-8 decoder is not implemented in Node v4.0+
- function InternalDecoderCesu8(options, codec) {
- this.acc = 0;
- this.contBytes = 0;
- this.accBytes = 0;
- this.defaultCharUnicode = codec.defaultCharUnicode;
- }
- InternalDecoderCesu8.prototype.write = function(buf) {
- var acc = this.acc, contBytes = this.contBytes, accBytes = this.accBytes,
- res = '';
- for (var i = 0; i < buf.length; i++) {
- var curByte = buf[i];
- if ((curByte & 0xC0) !== 0x80) { // Leading byte
- if (contBytes > 0) { // Previous code is invalid
- res += this.defaultCharUnicode;
- contBytes = 0;
- }
- if (curByte < 0x80) { // Single-byte code
- res += String.fromCharCode(curByte);
- } else if (curByte < 0xE0) { // Two-byte code
- acc = curByte & 0x1F;
- contBytes = 1; accBytes = 1;
- } else if (curByte < 0xF0) { // Three-byte code
- acc = curByte & 0x0F;
- contBytes = 2; accBytes = 1;
- } else { // Four or more are not supported for CESU-8.
- res += this.defaultCharUnicode;
- }
- } else { // Continuation byte
- if (contBytes > 0) { // We're waiting for it.
- acc = (acc << 6) | (curByte & 0x3f);
- contBytes--; accBytes++;
- if (contBytes === 0) {
- // Check for overlong encoding, but support Modified UTF-8 (encoding NULL as C0 80)
- if (accBytes === 2 && acc < 0x80 && acc > 0)
- res += this.defaultCharUnicode;
- else if (accBytes === 3 && acc < 0x800)
- res += this.defaultCharUnicode;
- else
- // Actually add character.
- res += String.fromCharCode(acc);
- }
- } else { // Unexpected continuation byte
- res += this.defaultCharUnicode;
- }
- }
- }
- this.acc = acc; this.contBytes = contBytes; this.accBytes = accBytes;
- return res;
- }
- InternalDecoderCesu8.prototype.end = function() {
- var res = 0;
- if (this.contBytes > 0)
- res += this.defaultCharUnicode;
- return res;
- }
|