Strange decoding of body text about s22.imap HOT 10 CLOSED

I have found that replacing the DecodeQP method with this...
public static string DecodeQuotedPrintables(string input)
{
var occurences = new Regex(@"=[0-9A-Z]{2}", RegexOptions.Multiline);
var matches = occurences.Matches(input);
foreach (Match match in matches)
{
char hexChar = (char)Convert.ToInt32(match.Groups[0].Value.Substring(1), 16);
input = input.Replace(match.Groups[0].Value, hexChar.ToString());
}
return input.Replace("=\r\n", "");
}

resolves the issue. I don't know enough about QP to know what is correct and what isn't.

from s22.imap.

I am having a similar issue. When saving some messages, there is no body when it gets saved. Some work, some don't. The attachments are coming. In said message, there doesn't seem to be quoted printable. The messages that work do have this.

from s22.imap.

Previous post removed, incorrect fix. indexer (i) + 2 was correct.

from s22.imap.

On 12/5/2012 5:34 AM, trickbooter wrote:

Previous post removed, incorrect fix. indexer (i) + 2 was correct.

—
Reply to this email directly or view it on GitHub
#21 (comment).

Is there something i can do on my end to fix this issue? By your
comment, I have to override QPDecode.

Thanks,

Jason

from s22.imap.

Thanks for reporting, I'll look into it. Can you post an example mail message (the "raw" message with mail headers and body) or forward one to me so that I can re-produce the problem?

from s22.imap.

I did read that QP encodes to a certain length (80 chars?) and that a special sequence is used to denote this, =\r\n. This needs to be dealt with. For me, this was the crux of the issue (see code below).

internal static string QPDecode(string value, Encoding encoding) {
            try {
                using (MemoryStream m = new MemoryStream()) {
                    for (int i = 0; i < value.Length; i++) {
                        if(value[i] == '=' && value[i+1] == '\r' && value[i+2]=='\n') {
                            i = i + 2;
                        } else if (value[i] == '=') {
                            string hex = value.Substring(i + 1, 2);
                            m.WriteByte(Convert.ToByte(hex, 16));
                            i = i + 2;
                        } else {
                            m.WriteByte(Convert.ToByte(value[i]));
                        }
                    }
                    return encoding.GetString(m.ToArray());
                }
            } catch {
                throw new FormatException("value is not a valid quoted-printable " +
                    "encoded string");
            }
        }

Using the existing method, the code decodes my message as (snipped to preserve confidentiality and shown as the c# raw string hence the \ escape char on the "...)
\r\n\r\n<rm xmlns=3D"http://www

after replacing the QPDecode' method with the one shown above, the same message decodes as follows...
\r\n\r\n<rm xmlns="http://www

I can forward you an email if you like, but I'd rather not post an email to this forum as it will reveal some private info. Have you got an email I can send to in confidence?

Many Thanks

Paul

from s22.imap.

Hi! That's a Soft Line Breaks. Non-significant line break.
Check out Rule #5 (Soft Line Breaks) in RFC 1521 (http://www.ietf.org/rfc/rfc1521.txt).
You should just skip \r\n (or \n) in QPDecode:

    string hex = value.Substring(i + 1, 2);
    if (hex != Environment.NewLine)
        m.WriteByte(Convert.ToByte(hex, 16));
    i = i + 2;

Or could be New Line in value string represented as \n, while Environment.NewLine is \r\n? If so, then code should be:

    string hex = value.Substring(i + 1, 2);
    if (hex=="\r\n")
    {
        i += 2;
    }
    else if (hex.StartsWith("\n"))
    {
        i += 1;
    }
    else
    {
        m.WriteByte(Convert.ToByte(hex, 16));
        i += 2;
    }

from s22.imap.

Agreed, I understand what they are. However, the standard QP Decode method
works for my incoming emails only after I made the changes I presented in
my last email.

On 13 January 2013 11:50, rocknroll-mind [email protected] wrote:

Hi! That's a Soft Line Breaks. Non-significant line break.
Check out Rule #5 (Soft Line Breaks) in RFC 1521 (
http://www.ietf.org/rfc/rfc1521.txt).
You should just skip \r\n (or \n) in QPDecode:

string hex = value.Substring(i + 1, 2);
if (hex != Environment.NewLine)
    m.WriteByte(Convert.ToByte(hex, 16));
i = i + 2;

Or could be New Line in value string represented as \n, while
Environment.NewLine is \r\n? If so, then code should be:

string hex = value.Substring(i + 1, 2);
if (hex=="\r\n")
{
    i += 2;
}
else if (hex.StartsWith("\n"))
{
    i += 1;
}
else
{
    m.WriteByte(Convert.ToByte(hex, 16));
    i += 2;
}

—
Reply to this email directly or view it on GitHubhttps://github.com//issues/21#issuecomment-12192661.

from s22.imap.

You are right, I just submitted a patch to address it.

from s22.imap.

Thanks. It is an excellent library. Well maintain, functionally coded.

On Saturday, 26 January 2013 at 15:31, smiley22 wrote:

You are right, I just submitted a patch to address it.

—
Reply to this email directly or view it on GitHub (#21 (comment)).

from s22.imap.