/// <summary>
/// Clone the <see cref="QuotedPrintableDecoder"/> with its current state.
/// </summary>
/// <remarks>
/// Creates a new <see cref="QuotedPrintableDecoder"/> with exactly the same state as the current decoder.
/// </remarks>
/// <returns>A new <see cref="QuotedPrintableDecoder"/> with identical state.</returns>
public IMimeDecoder Clone()
{
var decoder = new QuotedPrintableDecoder(rfc2047);
decoder.state = state;
decoder.saved = saved;
return(decoder);
}
/// <summary>
/// Clone the <see cref="QuotedPrintableDecoder"/> with its current state.
/// </summary>
/// <remarks>
/// Creates a new <see cref="QuotedPrintableDecoder"/> with exactly the same state as the current decoder.
/// </remarks>
/// <returns>A new <see cref="QuotedPrintableDecoder"/> with identical state.</returns>
public IMimeDecoder Clone ()
{
var decoder = new QuotedPrintableDecoder (rfc2047);
decoder.state = state;
decoder.saved = saved;
return decoder;
}
static unsafe string DecodeTokens (ParserOptions options, IList<Token> tokens, byte[] input, byte* inbuf, int length)
{
var decoded = new StringBuilder (length);
var qp = new QuotedPrintableDecoder (true);
var base64 = new Base64Decoder ();
var output = new byte[length];
Token token;
int len;
fixed (byte* outbuf = output) {
for (int i = 0; i < tokens.Count; i++) {
token = tokens[i];
if (token.Encoding != ContentEncoding.Default) {
// In order to work around broken mailers, we need to combine the raw
// decoded content of runs of identically encoded word tokens before
// converting to unicode strings.
ContentEncoding encoding = token.Encoding;
int codepage = token.CodePage;
IMimeDecoder decoder;
int outlen, n;
byte* outptr;
// find the end of this run (and measure the buffer length we'll need)
for (n = i + 1; n < tokens.Count; n++) {
if (tokens[n].Encoding != encoding || tokens[n].CodePage != codepage)
break;
}
// base64 / quoted-printable decode each of the tokens...
if (encoding == ContentEncoding.Base64)
decoder = base64;
else
decoder = qp;
outptr = outbuf;
outlen = 0;
do {
// Note: by not resetting the decoder state each loop, we effectively
// treat the payloads as one continuous block, thus allowing us to
// handle cases where a hex-encoded triplet of a quoted-printable
// encoded payload is split between 2 or more encoded-word tokens.
len = DecodeToken (tokens[i], decoder, inbuf, outptr);
outptr += len;
outlen += len;
i++;
} while (i < n);
decoder.Reset ();
i--;
var unicode = CharsetUtils.ConvertToUnicode (options, codepage, output, 0, outlen, out len);
decoded.Append (unicode, 0, len);
} else if (token.Is8bit) {
// *sigh* I hate broken mailers...
var unicode = CharsetUtils.ConvertToUnicode (options, input, token.StartIndex, token.Length, out len);
decoded.Append (unicode, 0, len);
} else {
// pure 7bit ascii, a breath of fresh air...
byte* inptr = inbuf + token.StartIndex;
byte* inend = inptr + token.Length;
while (inptr < inend)
decoded.Append ((char) *inptr++);
}
}
}
return decoded.ToString ();
}
public void TestQuotedPrintableDecode ()
{
const string input = "This is an ordinary text message in which my name (=ED=E5=EC=F9 =EF=E1 =E9=EC=E8=F4=F0)\nis in Hebrew (=FA=E9=F8=E1=F2).";
const string expected = "This is an ordinary text message in which my name (םולש ןב ילטפנ)\nis in Hebrew (תירבע).";
var encoding = Encoding.GetEncoding ("iso-8859-8");
var decoder = new QuotedPrintableDecoder ();
var output = new byte[4096];
string actual;
byte[] buf;
int n;
Assert.AreEqual (ContentEncoding.QuotedPrintable, decoder.Encoding);
buf = Encoding.ASCII.GetBytes (input);
n = decoder.Decode (buf, 0, buf.Length, output);
actual = encoding.GetString (output, 0, n);
Assert.AreEqual (expected, actual);
encoding = CharsetUtils.Latin1;
for (int i = 0; i < qpEncodedPatterns.Length; i++) {
decoder.Reset ();
buf = encoding.GetBytes (qpEncodedPatterns[i]);
n = decoder.Decode (buf, 0, buf.Length, output);
actual = encoding.GetString (output, 0, n);
Assert.AreEqual (qpDecodedPatterns[i], actual, "Failed to decode qpEncodedPatterns[{0}]", i);
}
}
public void TestQuotedPrintableDecode ()
{
const string input = "This is an ordinary text message in which my name (=ED=E5=EC=F9 =EF=E1 =E9=EC=E8=F4=F0)\nis in Hebrew (=FA=E9=F8=E1=F2).";
const string expected = "This is an ordinary text message in which my name (םולש ןב ילטפנ)\nis in Hebrew (תירבע).";
var encoding = Encoding.GetEncoding ("iso-8859-8");
var decoder = new QuotedPrintableDecoder ();
var output = new byte[4096];
var buf = Encoding.ASCII.GetBytes (input);
int n = decoder.Decode (buf, 0, buf.Length, output);
var actual = encoding.GetString (output, 0, n);
Assert.AreEqual (expected, actual);
}