123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290 |
- "use strict";
- var Buffer = require("buffer").Buffer;
- // UTF-7 codec, according to https://tools.ietf.org/html/rfc2152
- // See also below a UTF-7-IMAP codec, according to http://tools.ietf.org/html/rfc3501#section-5.1.3
- exports.utf7 = Utf7Codec;
- exports.unicode11utf7 = 'utf7'; // Alias UNICODE-1-1-UTF-7
- function Utf7Codec(codecOptions, iconv) {
- this.iconv = iconv;
- };
- Utf7Codec.prototype.encoder = Utf7Encoder;
- Utf7Codec.prototype.decoder = Utf7Decoder;
- Utf7Codec.prototype.bomAware = true;
- // -- Encoding
- var nonDirectChars = /[^A-Za-z0-9'\(\),-\.\/:\? \n\r\t]+/g;
- function Utf7Encoder(options, codec) {
- this.iconv = codec.iconv;
- }
- Utf7Encoder.prototype.write = function(str) {
- // Naive implementation.
- // Non-direct chars are encoded as "+<base64>-"; single "+" char is encoded as "+-".
- return new Buffer(str.replace(nonDirectChars, function(chunk) {
- return "+" + (chunk === '+' ? '' :
- this.iconv.encode(chunk, 'utf16-be').toString('base64').replace(/=+$/, ''))
- + "-";
- }.bind(this)));
- }
- Utf7Encoder.prototype.end = function() {
- }
- // -- Decoding
- function Utf7Decoder(options, codec) {
- this.iconv = codec.iconv;
- this.inBase64 = false;
- this.base64Accum = '';
- }
- var base64Regex = /[A-Za-z0-9\/+]/;
- var base64Chars = [];
- for (var i = 0; i < 256; i++)
- base64Chars[i] = base64Regex.test(String.fromCharCode(i));
- var plusChar = '+'.charCodeAt(0),
- minusChar = '-'.charCodeAt(0),
- andChar = '&'.charCodeAt(0);
- Utf7Decoder.prototype.write = function(buf) {
- var res = "", lastI = 0,
- inBase64 = this.inBase64,
- base64Accum = this.base64Accum;
- // The decoder is more involved as we must handle chunks in stream.
- for (var i = 0; i < buf.length; i++) {
- if (!inBase64) { // We're in direct mode.
- // Write direct chars until '+'
- if (buf[i] == plusChar) {
- res += this.iconv.decode(buf.slice(lastI, i), "ascii"); // Write direct chars.
- lastI = i+1;
- inBase64 = true;
- }
- } else { // We decode base64.
- if (!base64Chars[buf[i]]) { // Base64 ended.
- if (i == lastI && buf[i] == minusChar) {// "+-" -> "+"
- res += "+";
- } else {
- var b64str = base64Accum + buf.slice(lastI, i).toString();
- res += this.iconv.decode(new Buffer(b64str, 'base64'), "utf16-be");
- }
- if (buf[i] != minusChar) // Minus is absorbed after base64.
- i--;
- lastI = i+1;
- inBase64 = false;
- base64Accum = '';
- }
- }
- }
- if (!inBase64) {
- res += this.iconv.decode(buf.slice(lastI), "ascii"); // Write direct chars.
- } else {
- var b64str = base64Accum + buf.slice(lastI).toString();
- var canBeDecoded = b64str.length - (b64str.length % 8); // Minimal chunk: 2 quads -> 2x3 bytes -> 3 chars.
- base64Accum = b64str.slice(canBeDecoded); // The rest will be decoded in future.
- b64str = b64str.slice(0, canBeDecoded);
- res += this.iconv.decode(new Buffer(b64str, 'base64'), "utf16-be");
- }
- this.inBase64 = inBase64;
- this.base64Accum = base64Accum;
- return res;
- }
- Utf7Decoder.prototype.end = function() {
- var res = "";
- if (this.inBase64 && this.base64Accum.length > 0)
- res = this.iconv.decode(new Buffer(this.base64Accum, 'base64'), "utf16-be");
- this.inBase64 = false;
- this.base64Accum = '';
- return res;
- }
- // UTF-7-IMAP codec.
- // RFC3501 Sec. 5.1.3 Modified UTF-7 (http://tools.ietf.org/html/rfc3501#section-5.1.3)
- // Differences:
- // * Base64 part is started by "&" instead of "+"
- // * Direct characters are 0x20-0x7E, except "&" (0x26)
- // * In Base64, "," is used instead of "/"
- // * Base64 must not be used to represent direct characters.
- // * No implicit shift back from Base64 (should always end with '-')
- // * String must end in non-shifted position.
- // * "-&" while in base64 is not allowed.
- exports.utf7imap = Utf7IMAPCodec;
- function Utf7IMAPCodec(codecOptions, iconv) {
- this.iconv = iconv;
- };
- Utf7IMAPCodec.prototype.encoder = Utf7IMAPEncoder;
- Utf7IMAPCodec.prototype.decoder = Utf7IMAPDecoder;
- Utf7IMAPCodec.prototype.bomAware = true;
- // -- Encoding
- function Utf7IMAPEncoder(options, codec) {
- this.iconv = codec.iconv;
- this.inBase64 = false;
- this.base64Accum = new Buffer(6);
- this.base64AccumIdx = 0;
- }
- Utf7IMAPEncoder.prototype.write = function(str) {
- var inBase64 = this.inBase64,
- base64Accum = this.base64Accum,
- base64AccumIdx = this.base64AccumIdx,
- buf = new Buffer(str.length*5 + 10), bufIdx = 0;
- for (var i = 0; i < str.length; i++) {
- var uChar = str.charCodeAt(i);
- if (0x20 <= uChar && uChar <= 0x7E) { // Direct character or '&'.
- if (inBase64) {
- if (base64AccumIdx > 0) {
- bufIdx += buf.write(base64Accum.slice(0, base64AccumIdx).toString('base64').replace(/\//g, ',').replace(/=+$/, ''), bufIdx);
- base64AccumIdx = 0;
- }
- buf[bufIdx++] = minusChar; // Write '-', then go to direct mode.
- inBase64 = false;
- }
- if (!inBase64) {
- buf[bufIdx++] = uChar; // Write direct character
- if (uChar === andChar) // Ampersand -> '&-'
- buf[bufIdx++] = minusChar;
- }
- } else { // Non-direct character
- if (!inBase64) {
- buf[bufIdx++] = andChar; // Write '&', then go to base64 mode.
- inBase64 = true;
- }
- if (inBase64) {
- base64Accum[base64AccumIdx++] = uChar >> 8;
- base64Accum[base64AccumIdx++] = uChar & 0xFF;
- if (base64AccumIdx == base64Accum.length) {
- bufIdx += buf.write(base64Accum.toString('base64').replace(/\//g, ','), bufIdx);
- base64AccumIdx = 0;
- }
- }
- }
- }
- this.inBase64 = inBase64;
- this.base64AccumIdx = base64AccumIdx;
- return buf.slice(0, bufIdx);
- }
- Utf7IMAPEncoder.prototype.end = function() {
- var buf = new Buffer(10), bufIdx = 0;
- if (this.inBase64) {
- if (this.base64AccumIdx > 0) {
- bufIdx += buf.write(this.base64Accum.slice(0, this.base64AccumIdx).toString('base64').replace(/\//g, ',').replace(/=+$/, ''), bufIdx);
- this.base64AccumIdx = 0;
- }
- buf[bufIdx++] = minusChar; // Write '-', then go to direct mode.
- this.inBase64 = false;
- }
- return buf.slice(0, bufIdx);
- }
- // -- Decoding
- function Utf7IMAPDecoder(options, codec) {
- this.iconv = codec.iconv;
- this.inBase64 = false;
- this.base64Accum = '';
- }
- var base64IMAPChars = base64Chars.slice();
- base64IMAPChars[','.charCodeAt(0)] = true;
- Utf7IMAPDecoder.prototype.write = function(buf) {
- var res = "", lastI = 0,
- inBase64 = this.inBase64,
- base64Accum = this.base64Accum;
- // The decoder is more involved as we must handle chunks in stream.
- // It is forgiving, closer to standard UTF-7 (for example, '-' is optional at the end).
- for (var i = 0; i < buf.length; i++) {
- if (!inBase64) { // We're in direct mode.
- // Write direct chars until '&'
- if (buf[i] == andChar) {
- res += this.iconv.decode(buf.slice(lastI, i), "ascii"); // Write direct chars.
- lastI = i+1;
- inBase64 = true;
- }
- } else { // We decode base64.
- if (!base64IMAPChars[buf[i]]) { // Base64 ended.
- if (i == lastI && buf[i] == minusChar) { // "&-" -> "&"
- res += "&";
- } else {
- var b64str = base64Accum + buf.slice(lastI, i).toString().replace(/,/g, '/');
- res += this.iconv.decode(new Buffer(b64str, 'base64'), "utf16-be");
- }
- if (buf[i] != minusChar) // Minus may be absorbed after base64.
- i--;
- lastI = i+1;
- inBase64 = false;
- base64Accum = '';
- }
- }
- }
- if (!inBase64) {
- res += this.iconv.decode(buf.slice(lastI), "ascii"); // Write direct chars.
- } else {
- var b64str = base64Accum + buf.slice(lastI).toString().replace(/,/g, '/');
- var canBeDecoded = b64str.length - (b64str.length % 8); // Minimal chunk: 2 quads -> 2x3 bytes -> 3 chars.
- base64Accum = b64str.slice(canBeDecoded); // The rest will be decoded in future.
- b64str = b64str.slice(0, canBeDecoded);
- res += this.iconv.decode(new Buffer(b64str, 'base64'), "utf16-be");
- }
- this.inBase64 = inBase64;
- this.base64Accum = base64Accum;
- return res;
- }
- Utf7IMAPDecoder.prototype.end = function() {
- var res = "";
- if (this.inBase64 && this.base64Accum.length > 0)
- res = this.iconv.decode(new Buffer(this.base64Accum, 'base64'), "utf16-be");
- this.inBase64 = false;
- this.base64Accum = '';
- return res;
- }
|