146 lines
4.2 KiB
JavaScript
146 lines
4.2 KiB
JavaScript
const iconv = require('iconv-lite');
|
|
|
|
/**
|
|
* Decodes MIME encoded-words (e.g. =?UTF-8?B?...?= or =?ISO-8859-1?Q?...?=)
|
|
*/
|
|
function decodeMimeWords(str) {
|
|
if (!str || typeof str !== 'string') return str || '';
|
|
if (!str.includes('=?')) return str;
|
|
|
|
return str.replace(/=\?([^?]+)\?([BQbq])\?([^?]+)\?=/g, (fullMatch, charset, encoding, data) => {
|
|
try {
|
|
const cs = charset.toLowerCase();
|
|
const enc = encoding.toUpperCase();
|
|
|
|
let buf;
|
|
if (enc === 'B') {
|
|
buf = Buffer.from(data, 'base64');
|
|
} else if (enc === 'Q') {
|
|
const qDec = data.replace(/_/g, ' ').replace(/=([0-9A-Fa-f]{2})/g, (_, hex) => String.fromCharCode(parseInt(hex, 16)));
|
|
buf = Buffer.from(qDec, 'binary');
|
|
}
|
|
|
|
if (buf) {
|
|
if (iconv.encodingExists(cs)) {
|
|
return iconv.decode(buf, cs);
|
|
} else {
|
|
return buf.toString('utf-8');
|
|
}
|
|
}
|
|
} catch (_) {
|
|
return fullMatch;
|
|
}
|
|
return fullMatch;
|
|
});
|
|
}
|
|
|
|
/**
|
|
* Decodes raw Quoted-Printable strings (e.g. =E0=E8=E9=F9 or =C3=A8)
|
|
*/
|
|
function decodeQuotedPrintable(str, charset = 'utf-8') {
|
|
if (!str || typeof str !== 'string') return str || '';
|
|
if (!/=(?:[0-9A-Fa-f]{2}|\r?\n)/.test(str)) return str;
|
|
|
|
try {
|
|
const cleaned = str.replace(/=\r?\n/g, '');
|
|
const bytes = [];
|
|
for (let i = 0; i < cleaned.length; i++) {
|
|
if (cleaned[i] === '=' && i + 2 < cleaned.length && /[0-9A-Fa-f]{2}/.test(cleaned.substring(i + 1, i + 3))) {
|
|
bytes.push(parseInt(cleaned.substring(i + 1, i + 3), 16));
|
|
i += 2;
|
|
} else {
|
|
bytes.push(cleaned.charCodeAt(i));
|
|
}
|
|
}
|
|
const buf = Buffer.from(bytes);
|
|
if (iconv.encodingExists(charset)) {
|
|
return iconv.decode(buf, charset);
|
|
}
|
|
return buf.toString('utf-8');
|
|
} catch (_) {
|
|
return str;
|
|
}
|
|
}
|
|
|
|
/**
|
|
* Main text decoder & Mojibake repair utility
|
|
*/
|
|
function cleanAndDecodeText(rawInput, contentTypeHint = '') {
|
|
if (!rawInput && rawInput !== 0) return '';
|
|
|
|
let text = '';
|
|
|
|
// 1. Buffer / Base64 / String normalization
|
|
if (Buffer.isBuffer(rawInput)) {
|
|
let charset = 'utf-8';
|
|
const match = (contentTypeHint || '').match(/charset=["']?([^"';\s]+)/i);
|
|
if (match && match[1]) charset = match[1].toLowerCase();
|
|
|
|
try {
|
|
if (iconv.encodingExists(charset)) {
|
|
text = iconv.decode(rawInput, charset);
|
|
} else {
|
|
text = rawInput.toString('utf-8');
|
|
}
|
|
} catch (_) {
|
|
text = rawInput.toString('utf-8');
|
|
}
|
|
} else if (typeof rawInput === 'string') {
|
|
const cleaned = rawInput.replace(/\s+/g, '');
|
|
if (cleaned.length > 20 && cleaned.length % 4 === 0 && /^[A-Za-z0-9+/]*={0,2}$/.test(cleaned)) {
|
|
try {
|
|
const buf = Buffer.from(cleaned, 'base64');
|
|
const utfStr = buf.toString('utf-8');
|
|
// Only accept base64 decode if result doesn't look like corrupted binary noise
|
|
if (utfStr && !utfStr.includes('\uFFFD')) {
|
|
text = utfStr;
|
|
} else {
|
|
text = rawInput;
|
|
}
|
|
} catch (_) {
|
|
text = rawInput;
|
|
}
|
|
} else {
|
|
text = rawInput;
|
|
}
|
|
} else {
|
|
return String(rawInput);
|
|
}
|
|
|
|
// 2. Decode MIME words (e.g. =?UTF-8?B?...?= or =?ISO-8859-1?Q?...?=)
|
|
if (text.includes('=?')) {
|
|
text = decodeMimeWords(text);
|
|
}
|
|
|
|
// 3. Decode Quoted-Printable if present (and not inside HTML tags)
|
|
if (/=(?:[0-9A-Fa-f]{2})/.test(text) && !text.includes('<html') && !text.includes('<div') && !text.includes('<p>')) {
|
|
text = decodeQuotedPrintable(text);
|
|
}
|
|
|
|
// 4. Fix Mojibake / Double-UTF-8 / Win1252 / ISO-8859-1 mis-decoding
|
|
if (/(?:Ã[\s\xa0¡-ÿ]|â€[™œ]|–|Â[°©®])/.test(text)) {
|
|
try {
|
|
const reEncoded = iconv.encode(text, 'win1252');
|
|
const repaired = iconv.decode(reEncoded, 'utf-8');
|
|
if (repaired && !repaired.includes('\uFFFD')) {
|
|
text = repaired;
|
|
}
|
|
} catch (_) {}
|
|
}
|
|
|
|
// 6. Clean stray Mojibake/control characters like ý, ÿ, or U+FFFD () replacement symbols
|
|
if (/[ýÿ\uFFFD]/.test(text)) {
|
|
text = text.replace(/ý{1,}/g, '')
|
|
.replace(/ÿ{1,}/g, '')
|
|
.replace(/\uFFFD{1,}/g, '');
|
|
}
|
|
|
|
return text;
|
|
}
|
|
|
|
module.exports = {
|
|
cleanAndDecodeText,
|
|
decodeMimeWords,
|
|
decodeQuotedPrintable
|
|
};
|