diff options
Diffstat (limited to 'node_modules/iconv-lite/encodings/internal.js')
| -rw-r--r-- | node_modules/iconv-lite/encodings/internal.js | 218 |
1 files changed, 218 insertions, 0 deletions
diff --git a/node_modules/iconv-lite/encodings/internal.js b/node_modules/iconv-lite/encodings/internal.js new file mode 100644 index 0000000..4e5c3ff --- /dev/null +++ b/node_modules/iconv-lite/encodings/internal.js @@ -0,0 +1,218 @@ +"use strict" +var Buffer = require("safer-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 === "utf8") { this.encoder = InternalEncoderUtf8 } 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 (Buffer.from("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 + +function InternalDecoder (options, codec) { + this.decoder = new StringDecoder(codec.enc) +} + +InternalDecoder.prototype.write = function (buf) { + if (!Buffer.isBuffer(buf)) { + buf = Buffer.from(buf) + } + + return this.decoder.write(buf) +} + +InternalDecoder.prototype.end = function () { + return this.decoder.end() +} + +// ------------------------------------------------------------------------------ +// Encoder is mostly trivial + +function InternalEncoder (options, codec) { + this.enc = codec.enc +} + +InternalEncoder.prototype.write = function (str) { + return Buffer.from(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 Buffer.from(str, "base64") +} + +InternalEncoderBase64.prototype.end = function () { + return Buffer.from(this.prevStr, "base64") +} + +// ------------------------------------------------------------------------------ +// CESU-8 encoder is also special. + +function InternalEncoderCesu8 (options, codec) { +} + +InternalEncoderCesu8.prototype.write = function (str) { + var buf = Buffer.alloc(str.length * 3); var 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; var contBytes = this.contBytes; var accBytes = this.accBytes + var 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 +} + +// ------------------------------------------------------------------------------ +// check the chunk boundaries for surrogate pair + +function InternalEncoderUtf8 (options, codec) { + this.highSurrogate = "" +} + +InternalEncoderUtf8.prototype.write = function (str) { + if (this.highSurrogate) { + str = this.highSurrogate + str + this.highSurrogate = "" + } + + if (str.length > 0) { + var charCode = str.charCodeAt(str.length - 1) + if (charCode >= 0xd800 && charCode < 0xdc00) { + this.highSurrogate = str[str.length - 1] + str = str.slice(0, str.length - 1) + } + } + + return Buffer.from(str, this.enc) +} + +InternalEncoderUtf8.prototype.end = function () { + if (this.highSurrogate) { + var str = this.highSurrogate + this.highSurrogate = "" + return Buffer.from(str, this.enc) + } +} |