60 lines
1.8 KiB
JavaScript
60 lines
1.8 KiB
JavaScript
"use strict";
|
|
const iconvLite = require("iconv-lite");
|
|
const supportedNames = require("./supported-names.json");
|
|
const labelsToNames = require("./labels-to-names.json");
|
|
|
|
const supportedNamesSet = new Set(supportedNames);
|
|
|
|
// https://encoding.spec.whatwg.org/#concept-encoding-get
|
|
exports.labelToName = label => {
|
|
label = String(label).trim().toLowerCase();
|
|
|
|
return labelsToNames[label] || null;
|
|
};
|
|
|
|
// https://encoding.spec.whatwg.org/#decode
|
|
exports.decode = (uint8Array, fallbackEncodingName) => {
|
|
let encoding = fallbackEncodingName;
|
|
if (!exports.isSupported(encoding)) {
|
|
throw new RangeError(`"${encoding}" is not a supported encoding name`);
|
|
}
|
|
|
|
const bomEncoding = exports.getBOMEncoding(uint8Array);
|
|
if (bomEncoding !== null) {
|
|
encoding = bomEncoding;
|
|
// iconv-lite will strip BOMs for us, so no need to do the extra byte removal that the spec does.
|
|
// Note that we won't end up in the x-user-defined case when there's a bomEncoding.
|
|
}
|
|
|
|
if (encoding === "x-user-defined") {
|
|
// https://encoding.spec.whatwg.org/#x-user-defined-decoder
|
|
let result = "";
|
|
for (const byte of uint8Array) {
|
|
if (byte <= 0x7F) {
|
|
result += String.fromCodePoint(byte);
|
|
} else {
|
|
result += String.fromCodePoint(0xF780 + byte - 0x80);
|
|
}
|
|
}
|
|
return result;
|
|
}
|
|
|
|
return iconvLite.decode(uint8Array, encoding);
|
|
};
|
|
|
|
// https://github.com/whatwg/html/issues/1910#issuecomment-254017369
|
|
exports.getBOMEncoding = uint8Array => {
|
|
if (uint8Array[0] === 0xFE && uint8Array[1] === 0xFF) {
|
|
return "UTF-16BE";
|
|
} else if (uint8Array[0] === 0xFF && uint8Array[1] === 0xFE) {
|
|
return "UTF-16LE";
|
|
} else if (uint8Array[0] === 0xEF && uint8Array[1] === 0xBB && uint8Array[2] === 0xBF) {
|
|
return "UTF-8";
|
|
}
|
|
|
|
return null;
|
|
};
|
|
|
|
exports.isSupported = name => {
|
|
return supportedNamesSet.has(String(name));
|
|
};
|