Skip to content

Latest commit

 

History

History
1960 lines (1510 loc) · 77.5 KB

FAQ.md

File metadata and controls

1960 lines (1510 loc) · 77.5 KB

Frequently Asked Questions

Question Index

General

Messages

ImapClient

SmtpClient

General

Yes. MimeKit and MailKit are both completely free and open source. They are both covered under the MIT license.

In .NET Core, Microsoft decided to split out the non-Unicode text encodings into a separate NuGet package called System.Text.Encoding.CodePages.

MimeKit already pulls in a reference to this NuGet package, so you shouldn't need to add a reference to it in your project. That said, you will still need to register the encoding provider. It is recommended that you add the following line of code to your program initialization (e.g. the beginning of your program's Main() method):

System.Text.Encoding.RegisterProvider (System.Text.CodePagesEncodingProvider.Instance);

This only seems to happen in cases where the application is built for .NET Framework (v4.x) and seems to be most common for ASP.NET web applications that were built using Visual Studio 2019 (it is unclear whether this happens with Visual Studio 2022 as well).

The issue is that some (older?) versions of MSBuild do not correctly generate \*.dll.config, app.config and/or web.config files with proper assembly version binding redirects.

If this problem is happening to you, make sure to use MimeKit and MailKit >= v4.0 which include MimeKit.dll.config and MailKit.dll.config.

The next step is to manually edit your application's app.config (or web.config) to add a binding redirect for System.Runtime.CompilerServices.Unsafe:

<configuration>
  <runtime>
    <assemblyBinding xmlns="urn:schemas-microsoft-com:asm.v1">
      <dependentAssembly>
        <assemblyIdentity name="System.Runtime.CompilerServices.Unsafe" publicKeyToken="b03f5f7f11d50a3a" culture="neutral" />
        <bindingRedirect oldVersion="0.0.0.0-6.0.0.0" newVersion="6.0.0.0" />
      </dependentAssembly>
    </assemblyBinding>
  </runtime>
</configuration>

When you get an exception with that error message, it usually means that you are encountering one of the following scenarios:

1. The mail server does not support SSL on the specified port.

There are 2 different ways to use SSL/TLS encryption with mail servers.

The first way is to enable SSL/TLS encryption immediately upon connecting to the SMTP, POP3 or IMAP server. This method requires an "SSL port" because the standard port defined for the protocol is meant for plain-text communication.

The second way is via a STARTTLS command (aka STLS for POP3) that is optionally supported by the server.

Below is a table of the protocols supported by MailKit and the standard plain-text ports (which either do not support any SSL/TLS encryption at all or only via the STARTTLS command extension) and the SSL ports which require SSL/TLS encryption immediately upon a successful connection to the remote host.

Protocol Standard Port SSL Port
SMTP 25 or 587 465
POP3 110 995
IMAP 143 993

It is important to use the correct SecureSocketOptions for the port that you are connecting to.

If you are connecting to one of the standard ports above, you will need to use SecureSocketOptions.None, SecureSocketOptions.StartTls or SecureSocketOptions.StartTlsWhenAvailable.

If you are connecting to one of the SSL ports, you will need to use SecureSocketOptions.SslOnConnect.

You could also try using SecureSocketOptions.Auto which works by choosing the appropriate option to use by comparing the specified port to the ports in the above table.

2. The mail server that you are connecting to is using an expired (or otherwise untrusted) SSL certificate.

Often times, mail servers will use self-signed certificates instead of using a certificate that has been signed by a trusted Certificate Authority. Another potential pitfall is when locally installed anti-virus software replaces the certificate in order to scan web traffic for viruses.

When your system is unable to validate the mail server's certificate because it is not signed by a known and trusted Certificate Authority, the above error will occur.

If you are on a Linux system or are running a web service in a Linux container, it might be possible to use the following command to install the standard set of Certificate Authority root certificates using the following command:

apt update && apt install -y ca-certificates

Another option is to work around this problem by supplying a custom RemoteCertificateValidationCallback and setting it on the client's ServerCertificateValidationCallback property.

In the simplest example, you could do something like this (although I would strongly recommend against it in production use):

using (var client = new SmtpClient ()) {
    client.ServerCertificateValidationCallback = (s,c,h,e) => true;

    client.Connect (hostName, port, SecureSocketOptions.Auto);

    // ...
}

A better solution might be to compare the certificate's common name, issuer, serial number, and fingerprint to known values to make sure that the certificate can be trusted. Take the following code snippet as an example of how to do this:

bool MyServerCertificateValidationCallback (object sender, X509Certificate certificate, X509Chain chain, SslPolicyErrors sslPolicyErrors)
{
    if (sslPolicyErrors == SslPolicyErrors.None)
        return true;

    // Note: The following code casts to an X509Certificate2 because it's easier to get the
    // values for comparison, but it's possible to get them from an X509Certificate as well.
    if (certificate is X509Certificate2 certificate2) {
        var cn = certificate2.GetNameInfo (X509NameType.SimpleName, false);
        var fingerprint = certificate2.Thumbprint;
        var serial = certificate2.SerialNumber;
        var issuer = certificate2.Issuer;

        return cn == "imap.gmail.com" && issuer == "CN=GTS CA 1O1, O=Google Trust Services, C=US" &&
            serial == "00A15434C2695FB1880300000000CBF786" &&
            fingerprint == "F351BCB631771F19AF41DFF22EB0A0839092DA51";
    }

    return false;
}

The downside of the above example is that it requires hard-coding known values for "trusted" mail server certificates which can quickly become unwieldy to deal with if your program is meant to be used with a wide range of mail servers.

The best approach would be to prompt the user with a dialog explaining that the certificate is not trusted for the reasons enumerated by the SslPolicyErrors argument as well as potentially the errors provided in the X509Chain. If the user wishes to accept the risks of trusting the certificate, your program could then return true.

For more details on writing a custom SSL certificate validation callback, it may be worth checking out the SslCertificateValidation.cs example.

3. A Certificate Authority CRL server for one or more of the certificates in the chain is temporarily unavailable.

Most Certificate Authorities are probably pretty good at keeping their CRL and/or OCSP servers up 24/7, but occasionally they do go down or are otherwise unreachable due to other network problems between you and the server. When this happens, it becomes impossible to check the revocation status of one or more of the certificates in the chain.

To ignore revocation checks, you can set the CheckCertificateRevocation property of the IMAP, POP3 or SMTP client to false before you connect:

using (var client = new SmtpClient ()) {
    client.CheckCertificateRevocation = false;

    client.Connect (hostName, port, SecureSocketOptions.Auto);

    // ...
}

4. The server does not support the same set of SSL/TLS protocols that the client is configured to use.

MailKit attempts to keep up with the latest security recommendations and so is continuously removing older SSL and TLS protocols that are no longer considered secure from the default configuration. This often means that MailKit's SMTP, POP3 and IMAP clients will fail to connect to servers that are still using older SSL and TLS protocols. Currently, the SSL and TLS protocols that are not supported by default are: SSL v2.0, SSL v3.0, TLS v1.0 and TLS v1.1.

You can override MailKit's default set of supported SSL and TLS protocols by setting the value of the SslProtocols property on your SMTP, POP3 or IMAP client.

For example:

using (var client = new SmtpClient ()) {
    // Allow SSLv3.0 and all versions of TLS
    client.SslProtocols = SslProtocols.Ssl3 | SslProtocols.Tls | SslProtocols.Tls11 | SslProtocols.Tls12 | SslProtocols.Tls13;

    client.Connect ("smtp.gmail.com", 465, true);

    // ...
}

All of MailKit's client implementations have a constructor that takes a nifty IProtocolLogger interface for logging client/server communications. Out of the box, you can use the handy ProtocolLogger class. Here are some examples of how to use it:

// log to a file called 'imap.log'
var client = new ImapClient (new ProtocolLogger ("imap.log"));
// log to standard output (i.e. the console)
var client = new ImapClient (new ProtocolLogger (Console.OpenStandardOutput ()));

Note: When submitting a protocol log as part of a bug report, make sure to scrub any sensitive information including your authentication credentials. This information will generally be the base64 encoded blob immediately following an AUTHENTICATE or AUTH command (depending on the type of server). The only exception to this case is if you are authenticating with NTLM in which case I may need this information, but only if the bug/error is in the authentication step.

By default, GMail's POP3 and IMAP server does not behave like standard POP3 or IMAP servers and hides messages from clients using those protocols (as well as having other non-standard behavior).

If you want to configure your GMail POP3 or IMAP settings to behave the way POP3 and IMAP are intended to behave according to their protocol specifications, you'll need to log in to your GMail account via your web browser and navigate to the Forwarding and POP/IMAP tab of your GMail Settings page and set your options to look like this:

GMail POP3 and IMAP Settings

As of September 30th, 2024, authentication using only a username and password is no longer supported by Google.

There are now only 2 options to choose from:

  1. Use OAuth 2.0 authentication
  2. Use an "App password"

To use an App password, you will first need to turn on 2-Step Verification. Once 2-Step Verification is turned on, you can generate an App password.

Then, assuming that your GMail account is [email protected], you would use the following code snippet to connect to GMail via IMAP:

using (var client = new ImapClient ()) {
    client.Connect ("imap.gmail.com", 993, SecureSocketOptions.SslOnConnect);
    client.Authenticate ("[email protected]", "password");

    // do stuff...

    client.Disconnect (true);
}

Connecting via POP3 or SMTP is identical except for the host names and ports (and, of course, you'd use a Pop3Client or SmtpClient as appropriate).

The first thing you need to do is follow Google's instructions for obtaining OAuth 2.0 credentials for your application.

(Or, as an alternative set of step-by-step instructions, you can follow the directions that I have written in GMailOAuth2.md.)

Once you've done that, the easiest way to obtain an access token is to use Google's Google.Apis.Auth library:

const string GMailAccount = "[email protected]";

var clientSecrets = new ClientSecrets {
    ClientId = "XXX.apps.googleusercontent.com",
    ClientSecret = "XXX"
};

var codeFlow = new GoogleAuthorizationCodeFlow (new GoogleAuthorizationCodeFlow.Initializer {
    // Cache tokens in ~/.local/share/google-filedatastore/CredentialCacheFolder on Linux/Mac
    DataStore = new FileDataStore ("CredentialCacheFolder", false),
    Scopes = new [] { "https://mail.google.com/" },
    ClientSecrets = clientSecrets,
    LoginHint = GMailAccount
});

// Note: For a web app, you'll want to use AuthorizationCodeWebApp instead.
var codeReceiver = new LocalServerCodeReceiver ();
var authCode = new AuthorizationCodeInstalledApp (codeFlow, codeReceiver);

var credential = await authCode.AuthorizeAsync (GMailAccount, CancellationToken.None);

if (credential.Token.IsStale)
    await credential.RefreshTokenAsync (CancellationToken.None);

var oauth2 = new SaslMechanismOAuthBearer (credential.UserId, credential.Token.AccessToken);

using (var client = new ImapClient ()) {
    await client.ConnectAsync ("imap.gmail.com", 993, SecureSocketOptions.SslOnConnect);
    await client.AuthenticateAsync (oauth2);
    await client.DisconnectAsync (true);
}

Messages

To construct a message with attachments, the first thing you'll need to do is create a multipart/mixed container which you'll then want to add the message body to first. Once you've added the body, you can then add MIME parts to it that contain the content of the files you'd like to attach, being sure to set the Content-Disposition header value to attachment. You'll probably also want to set the filename parameter on the Content-Disposition header as well as the name parameter on the Content-Type header. The most convenient way to do this is to use the MimePart.FileName property which will set both parameters for you as well as setting the Content-Disposition header value to attachment if it has not already been set to something else.

var message = new MimeMessage ();
message.From.Add (new MailboxAddress ("Joey", "[email protected]"));
message.To.Add (new MailboxAddress ("Alice", "[email protected]"));
message.Subject = "How you doin?";

// create our message text, just like before (except don't set it as the message.Body)
var body = new TextPart ("plain") {
    Text = @"Hey Alice,

What are you up to this weekend? Monica is throwing one of her parties on
Saturday and I was hoping you could make it.

Will you be my +1?

-- Joey
"
};

// create an image attachment for the file located at path
var attachment = new MimePart ("image", "gif") {
    Content = new MimeContent (File.OpenRead (path), ContentEncoding.Default),
    ContentDisposition = new ContentDisposition (ContentDisposition.Attachment),
    ContentTransferEncoding = ContentEncoding.Base64,
    FileName = Path.GetFileName (path)
};

// now create the multipart/mixed container to hold the message text and the
// image attachment
var multipart = new Multipart ("mixed");
multipart.Add (body);
multipart.Add (attachment);

// now set the multipart/mixed as the message body
message.Body = multipart;

A simpler way to construct messages with attachments is to take advantage of the BodyBuilder class.

var message = new MimeMessage ();
message.From.Add (new MailboxAddress ("Joey", "[email protected]"));
message.To.Add (new MailboxAddress ("Alice", "[email protected]"));
message.Subject = "How you doin?";

var builder = new BodyBuilder ();

// Set the plain-text version of the message text
builder.TextBody = @"Hey Alice,

What are you up to this weekend? Monica is throwing one of her parties on
Saturday and I was hoping you could make it.

Will you be my +1?

-- Joey
";

// We may also want to attach a calendar event for Monica's party...
builder.Attachments.Add (@"C:\Users\Joey\Documents\party.ics");

// Now we just need to set the message body and we're done
message.Body = builder.ToMessageBody ();

For more information, see Creating Messages.

(Note: for the TL;DR version, skip to the end)

MIME is a tree structure of parts. There are multiparts which contain other parts (even other multiparts). There are message parts which contain messages. And finally, there are leaf-node parts which contain content.

There are a few common message structures:

  1. The message contains only a text/plain or text/html part (easy, just use that).

  2. The message contains a multipart/alternative which will typically look a bit like this:

    multipart/alternative
       text/plain
       text/html
    
  3. Same as above, but the html part is inside a multipart/related so that it can embed images:

    multipart/alternative
       text/plain
       multipart/related
          text/html
          image/jpeg
          image/png
    
  4. The message contains a textual body part as well as some attachments:

    multipart/mixed
       text/plain or text/html
       application/octet-stream
       application/zip
    
  5. the same as above, but with the first part replaced with either #2 or #3. To illustrate:

    multipart/mixed
       multipart/alternative
          text/plain
          text/html
       application/octet-stream
       application/zip
    

    or...

    multipart/mixed
       multipart/alternative
          text/plain
          multipart/related
             text/html
             image/jpeg
             image/png
       application/octet-stream
       application/zip
    

Now, if you don't care about any of that and just want to get the text of the first text/plain or text/html part you can find, that's easy.

MimeMessage has two convenience properties for this: TextBody and HtmlBody.

MimeMessage.HtmlBody, as the name implies, will traverse the MIME structure for you and find the most appropriate body part with a Content-Type of text/html that can be interpreted as the message body. Likewise, the TextBody property can be used to get the text/plain version of the message body.

For more information, see Working with Messages.

In most cases, a message with a body that has a MIME-type of multipart/mixed containing more than a single part probably has attachments. As illustrated above, the first part of a multipart/mixed is typically the textual body of the message, but it is not always quite that simple.

In general, MIME attachments will have a Content-Disposition header with a value of attachment. To get the list of body parts matching this criteria, you can use the MimeMessage.Attachments property.

Unfortunately, not all mail clients follow this convention and so you may need to write your own custom logic. For example, you may wish to treat all body parts having a name or filename parameter set on them:

var attachments = message.BodyParts.OfType<MimePart> ().Where (part => !string.IsNullOrEmpty (part.FileName));

A more sophisticated approach is to treat body parts not referenced by the main textual body part of the message as attachments. In other words, treat any body part not used for rendering the message as an attachment. For an example on how to do this, consider the following code snippets:

/// <summary>
/// Visits a MimeMessage and generates HTML suitable to be rendered by a browser control.
/// </summary>
class HtmlPreviewVisitor : MimeVisitor
{
    List<MultipartRelated> stack = new List<MultipartRelated> ();
    List<MimeEntity> attachments = new List<MimeEntity> ();
    readonly string tempDir;
    string body;

    /// <summary>
    /// Creates a new HtmlPreviewVisitor.
    /// </summary>
    /// <param name="tempDirectory">A temporary directory used for storing image files.</param>
    public HtmlPreviewVisitor (string tempDirectory)
    {
        tempDir = tempDirectory;
    }

    /// <summary>
    /// The list of attachments that were in the MimeMessage.
    /// </summary>
    public IList<MimeEntity> Attachments {
        get { return attachments; }
    }

    /// <summary>
    /// The HTML string that can be set on the BrowserControl.
    /// </summary>
    public string HtmlBody {
        get { return body ?? string.Empty; }
    }

    protected override void VisitMultipartAlternative (MultipartAlternative alternative)
    {
        // walk the multipart/alternative children backwards from greatest level of faithfulness to the least faithful
        for (int i = alternative.Count - 1; i >= 0 && body == null; i--)
            alternative[i].Accept (this);
    }

    protected override void VisitMultipartRelated (MultipartRelated related)
    {
        var root = related.Root;

        // push this multipart/related onto our stack
        stack.Add (related);

        // visit the root document
        root.Accept (this);

        // pop this multipart/related off our stack
        stack.RemoveAt (stack.Count - 1);
    }

    // look up the image based on the img src url within our multipart/related stack
    bool TryGetImage (string url, out MimePart image)
    {
        UriKind kind;
        int index;
        Uri uri;

        if (Uri.IsWellFormedUriString (url, UriKind.Absolute))
            kind = UriKind.Absolute;
        else if (Uri.IsWellFormedUriString (url, UriKind.Relative))
            kind = UriKind.Relative;
        else
            kind = UriKind.RelativeOrAbsolute;

        try {
            uri = new Uri (url, kind);
        } catch {
            image = null;
            return false;
        }

        for (int i = stack.Count - 1; i >= 0; i--) {
            if ((index = stack[i].IndexOf (uri)) == -1)
                continue;

            image = stack[i][index] as MimePart;
            return image != null;
        }

        image = null;

        return false;
    }

    /// <summary>
    /// Get a file:// URI for the image attachment.
    /// </summary>
    /// <remarks>
    /// Saves the image attachment to a temp file and returns a file:// URI for the
    /// temp file.
    /// </remarks>
    /// <returns>The file:// URI.</returns>
    /// <param name="image">The image attachment.</param>
    /// <param name="url">The original HTML image URL.</param>
    string GetFileUri (MimePart image, string url)
    {
        string fileName = url.Replace (':', '_').Replace ('\\', '_').Replace ('/', '_');

        string path = Path.Combine (tempDir, fileName);

        if (!File.Exists (path)) {
            using (var output = File.Create (path))
                image.Content.DecodeTo (output);
        }

        return "file://" + path.Replace ('\\', '/');
    }

    /// <summary>
    /// Get a data: URI for the image attachment.
    /// </summary>
    /// <remarks>
    /// Encodes the image attachment into a string suitable for setting as a src= attribute value in
    /// an img tag.
    /// </remarks>
    /// <returns>The data: URI.</returns>
    /// <param name="image">The image attachment.</param>
    string GetDataUri (MimePart image)
    {
        using (var memory = new MemoryStream ()) {
            image.Content.DecodeTo (memory);
            var buffer = memory.GetBuffer ();
            var length = (int) memory.Length;
            var base64 = Convert.ToBase64String (buffer, 0, length);

            return string.Format ("data:{0};base64,{1}", image.ContentType.MimeType, base64);
        }
    }

    // Replaces <img src=...> urls that refer to images embedded within the message with
    // "file://" urls that the browser control will actually be able to load.
    void HtmlTagCallback (HtmlTagContext ctx, HtmlWriter htmlWriter)
    {
        if (ctx.TagId == HtmlTagId.Meta && !ctx.IsEndTag) {
            bool isContentType = false;

            ctx.WriteTag (htmlWriter, false);

            // replace charsets with "utf-8" since our output will be in utf-8 (and not whatever the original charset was)
            foreach (var attribute in ctx.Attributes) {
                if (attribute.Id == HtmlAttributeId.Charset) {
                    htmlWriter.WriteAttributeName (attribute.Name);
                    htmlWriter.WriteAttributeValue ("utf-8");
                } else if (isContentType && attribute.Id == HtmlAttributeId.Content) {
                    htmlWriter.WriteAttributeName (attribute.Name);
                    htmlWriter.WriteAttributeValue ("text/html; charset=utf-8");
                } else {
                    if (attribute.Id == HtmlAttributeId.HttpEquiv && attribute.Value != null
                        && attribute.Value.Equals ("Content-Type", StringComparison.OrdinalIgnoreCase))
                        isContentType = true;

                    htmlWriter.WriteAttribute (attribute);
                }
            }
        } else if (ctx.TagId == HtmlTagId.Image && !ctx.IsEndTag && stack.Count > 0) {
            ctx.WriteTag (htmlWriter, false);

            // replace the src attribute with a file:// URL
            foreach (var attribute in ctx.Attributes) {
                if (attribute.Id == HtmlAttributeId.Src) {
                    if (!TryGetImage (attribute.Value, out var image)) {
                        htmlWriter.WriteAttribute (attribute);
                        continue;
                    }

                    // Note: you can either use a "file://" URI or you can use a
                    // "data:" URI, the choice is yours.
                    var uri = GetFileUri (image, attribute.Value);
                    //var uri = GetDataUri (image);

                    htmlWriter.WriteAttributeName (attribute.Name);
                    htmlWriter.WriteAttributeValue (uri);
                } else {
                    htmlWriter.WriteAttribute (attribute);
                }
            }
        } else if (ctx.TagId == HtmlTagId.Body && !ctx.IsEndTag) {
            ctx.WriteTag (htmlWriter, false);

            // add and/or replace oncontextmenu="return false;"
            foreach (var attribute in ctx.Attributes) {
                if (attribute.Name.Equals ("oncontextmenu", StringComparison.OrdinalIgnoreCase))
                   continue;

                htmlWriter.WriteAttribute (attribute);
            }

            htmlWriter.WriteAttribute ("oncontextmenu", "return false;");
        } else {
            // pass the tag through to the output
            ctx.WriteTag (htmlWriter, true);
        }
    }

    protected override void VisitTextPart (TextPart entity)
    {
        TextConverter converter;

        if (body != null) {
            // since we've already found the body, treat this as an attachment
            attachments.Add (entity);
            return;
        }

        if (entity.IsHtml) {
            converter = new HtmlToHtml {
                HtmlTagCallback = HtmlTagCallback
            };
        } else if (entity.IsFlowed) {
            var flowed = new FlowedToHtml ();
            string delsp;

            if (entity.ContentType.Parameters.TryGetValue ("delsp", out delsp))
                flowed.DeleteSpace = delsp.Equals ("yes", StringComparison.OrdinalIgnoreCase);

            converter = flowed;
        } else {
            converter = new TextToHtml ();
        }

        body = converter.Convert (entity.Text);
    }

    protected override void VisitTnefPart (TnefPart entity)
    {
        // extract any attachments in the MS-TNEF part
        attachments.AddRange (entity.ExtractAttachments ());
    }

    protected override void VisitMessagePart (MessagePart entity)
    {
        // treat message/rfc822 parts as attachments
        attachments.Add (entity);
    }

    protected override void VisitMimePart (MimePart entity)
    {
        // realistically, if we've gotten this far, then we can treat this as an attachment
        // even if the IsAttachment property is false.
        attachments.Add (entity);
    }
}

And the way you'd use this visitor might look something like this:

void Render (MimeMessage message)
{
    var tmpDir = Path.Combine (Path.GetTempPath (), message.MessageId);
    var visitor = new HtmlPreviewVisitor (tmpDir);

    Directory.CreateDirectory (tmpDir);

    message.Accept (visitor);

    DisplayHtml (visitor.HtmlBody);
    DisplayAttachments (visitor.Attachments);
}

Once you've rendered the message using the above technique, you'll have a list of attachments that were not used, even if they did not match the simplistic criteria used by the MimeMessage.Attachments property.

The MimeKit API was designed to use the existing MIME format for serialization. In light of this, the ability to use the .NET serialization API and format did not make much sense to support.

You can easily serialize a MimeMessage to a stream using the WriteTo methods.

For more information on this topic, see the following other two topics:

One of the more common operations that MimeKit is meant for is parsing email messages from arbitrary streams. There are two ways of accomplishing this task.

The first way is to use one of the Load methods on MimeMessage:

// Load a MimeMessage from a stream
var message = MimeMessage.Load (stream);

Or you can load a message from a file path:

// Load a MimeMessage from a file path
var message = MimeMessage.Load ("message.eml");

The second way is to use the MimeParser class. For the most part, using the MimeParser directly is not necessary unless you wish to parse a Unix mbox file stream. However, this is how you would do it:

// Load a MimeMessage from a stream
var parser = new MimeParser (stream, MimeFormat.Entity);
var message = parser.ParseMessage ();

For Unix mbox file streams, you would use the parser like this:

// Load every message from a Unix mbox
var parser = new MimeParser (stream, MimeFormat.Mbox);
while (!parser.IsEndOfStream) {
    var message = parser.ParseMessage ();

    // do something with the message
}

One you've got a MimeMessage, you can save it to a file using the WriteTo method:

message.WriteTo ("message.eml");

The WriteTo method also has overloads that allow you to write the message to a Stream instead.

By default, the WriteTo method will save the message using DOS line-endings on Windows and Unix line-endings on Unix-based systems such as macOS and Linux. You can override this behavior by passing a FormatOptions argument to the method:

// clone the default formatting options
var format = FormatOptions.Default.Clone ();

// override the line-endings to be DOS no matter what platform we are on
format.NewLineFormat = NewLineFormat.Dos;

message.WriteTo (format, "message.eml");

Note: While it may seem like you can safely use the ToString method to serialize a message, DON'T DO IT! This is not safe! MIME messages cannot be accurately represented as strings due to the fact that each MIME part of the message may be encoded in a different character set, thus making it impossible to convert the message into a unicode string using a single charset to do the conversion (which is exactly what ToString does).

If you've already got a MimePart that represents the attachment that you'd like to save, here's how you might save it:

using (var stream = File.Create (fileName))
    attachment.Content.DecodeTo (stream);

Pretty simple, right?

But what if your attachment is actually a MessagePart?

To save the content of a message/rfc822 part, you'd use the following code snippet:

using (var stream = File.Create (fileName))
    attachment.Message.WriteTo (stream);

If you are iterating over all of the attachments in a message, you might do something like this:

foreach (var attachment in message.Attachments) {
    var fileName = attachment.ContentDisposition?.FileName ?? attachment.ContentType.Name;

    using (var stream = File.Create (fileName)) {
        if (attachment is MessagePart) {
            var rfc822 = (MessagePart) attachment;

            rfc822.Message.WriteTo (stream);
        } else {
            var part = (MimePart) attachment;

            part.Content.DecodeTo (stream);
        }
    }
}

The From, To, and Cc properties of a MimeMessage are all of type InternetAddressList. An InternetAddressList is a list of InternetAddress items. This is where most people start to get lost because an InternetAddress is an abstract class that only really has a Name property.

As you've probably already discovered, the Name property contains the name of the person (if available), but what you want is his or her email address, not their name.

To get the email address, you'll need to figure out what subclass of address each InternetAddress really is. There are 2 subclasses of InternetAddress: GroupAddress and MailboxAddress.

A GroupAddress is a named group of more InternetAddress items that are contained within the Members property. To get an idea of what a group address represents, consider the following examples:

To: My Friends: Joey <[email protected]>, Monica <[email protected]>, "Mrs. Chanandler Bong"
    <[email protected]>, Ross <[email protected]>, Rachel <[email protected]>;

In the above example, the To header's InternetAddressList will contain only 1 item which will be a GroupAddress with a Name value of My Friends. The Members property of the GroupAddress will contain 5 more InternetAddress items (which will all be instances of MailboxAddress).

The above example, however, is not very likely to ever be seen in messages you deal with. A far more common example would be the one below:

To: undisclosed-recipients:;

Most of the time, the From, To, and Cc headers will only contain mailbox addresses. As you will notice, a MailboxAddress has an Address property which will contain the email address of the mailbox. In the following example, the Address property will contain the value [email protected]:

To: John Smith <[email protected]>

If you only care about getting a flattened list of the mailbox addresses in a From, To, or Cc header, you can do something like this:

foreach (var mailbox in message.To.Mailboxes)
    Console.WriteLine ("{0}'s email address is {1}", mailbox.Name, mailbox.Address);

An attachment filename is stored as a MIME parameter on the Content-Disposition header. Unfortunately, the original MIME specifications did not specify a method for encoding non-ASCII filenames. In 1997, rfc2184 (later updated by rfc2231) was published which specified an encoding mechanism to use for encoding them. Since there was a window in time where the MIME specifications did not define a way to encode them, some mail client developers decided to use the mechanism described by rfc2047 which was meant for encoding non-ASCII text in headers. While this may at first seem logical, the problem with this approach was that rfc2047 encoded-word tokens are not allowed to be in quotes (as well as some other issues) and so another, more appropriate, encoding mechanism was needed.

Outlook is one of those mail clients which decided to encode filenames using the mechanism described in rfc2047 and until Outlook 2007, did not support filenames encoded using the mechanism defined in rfc2231.

As of MimeKit v1.2.18, it is possible to configure MimeKit to use the rfc2047 encoding mechanism for filenames (and other Content-Disposition and Content-Type parameter values) by setting the encoding method on each individual Parameter:

Parameter param;

if (attachment.ContentDisposition.Parameters.TryGetValue ("filename", out param))
    param.EncodingMethod = ParameterEncodingMethod.Rfc2047;

Or:

foreach (var param in attachment.ContentDisposition.Parameters) {
    param.EncodingMethod = ParameterEncodingMethod.Rfc2047;
}

Some PGP-enabled mail clients, such as Thunderbird, embed encrypted PGP blurbs within the text/plain body of the message rather than using the PGP/MIME format that MimeKit prefers.

These messages often look something like this:

Return-Path: <[email protected]>
Received: from [127.0.0.1] (hostname.example.com. [201.95.8.17])
    by mx.google.com with ESMTPSA id l67sm26628445yha.8.2014.04.27.13.49.44
    for <[email protected]>
    (version=TLSv1 cipher=ECDHE-RSA-RC4-SHA bits=128/128);
    Sun, 27 Apr 2014 13:49:44 -0700 (PDT)
Message-ID: <[email protected]>
Date: Sun, 27 Apr 2014 17:49:43 -0300
From: Die-Hard PGP Fan <[email protected]>
User-Agent: Mozilla/5.0 (Windows NT 6.3; WOW64; rv:24.0) Gecko/20100101 Thunderbird/24.4.0
MIME-Version: 1.0
To: undisclosed-recipients:;
Subject: Test of inline encrypted PGP blocks
X-Enigmail-Version: 1.6
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 8bit
X-Antivirus: avast! (VPS 140427-1, 27/04/2014), Outbound message
X-Antivirus-Status: Clean

-----BEGIN PGP MESSAGE-----
Charset: ISO-8859-1
Version: GnuPG v2.0.22 (MingW32)
Comment: Using GnuPG with Thunderbird - http://www.enigmail.net/

SGFoISBJIGZvb2xlZCB5b3UsIHRoaXMgdGV4dCBpc24ndCBhY3R1YWxseSBlbmNy
eXB0ZWQgd2l0aCBQR1AsCml0J3MgYWN0dWFsbHkgb25seSBiYXNlNjQgZW5jb2Rl
ZCEKCkknbSBqdXN0IHVzaW5nIHRoaXMgYXMgYW4gZXhhbXBsZSwgdGhvdWdoLCBz
byBpdCBkb2Vzbid0IHJlYWxseSBtYXR0ZXIuCgpGb3IgdGhlIHNha2Ugb2YgYXJn
dW1lbnQsIHdlJ2xsIHByZXRlbmQgdGhhdCB0aGlzIGlzIGFjdHVhbGx5IGFuIGVu
Y3J5cHRlZApibHVyYi4gTW1ta2F5PyBUaGFua3MuCg==
-----END PGP MESSAGE-----

To deal with these kinds of messages, I've added a method to OpenPgpContext called GetDecryptedStream which can be used to get the raw decrypted stream.

There are actually 2 variants of this method:

public Stream GetDecryptedStream (Stream encryptedData, out DigitalSignatureCollection signatures)

and

public Stream GetDecryptedStream (Stream encryptedData)

The first variant is useful in cases where the encrypted PGP blurb is also digitally signed, allowing you to get your hands on the list of digital signatures in order for you to verify each of them.

To decrypt the content of the message, you'll want to locate the TextPart (in this case, it'll just be message.Body) and then do this:

static Stream DecryptEmbeddedPgp (TextPart text)
{
    using (var memory = new MemoryStream ()) {
        text.Content.DecodeTo (memory);
        memory.Position = 0;

        using (var ctx = new MyGnuPGContext ()) {
            return ctx.GetDecryptedStream (memory);
        }
    }
}

What you do with that decrypted stream is up to you. It's up to you to figure out what the decrypted content is (is it text? a jpeg image? a video?) and how to display it to the user.

Replying to a message is fairly simple. For the most part, you'd just create the reply message the same way you'd create any other message. There are only a few slight differences:

  1. In the reply message, you'll want to prefix the Subject header with "Re: " if the prefix doesn't already exist in the message you are replying to (in other words, if you are replying to a message with a Subject of "Re: party tomorrow night!", you would not prefix it with another "Re: ").
  2. You will want to set the reply message's In-Reply-To header to the value of the Message-Id header in the original message.
  3. You will want to copy the original message's References header into the reply message's References header and then append the original message's Message-Id header.
  4. You will probably want to "quote" the original message's text in the reply.
  5. If you are generating an automatic reply, you should also follow RFC3834 and set the Auto-Submitted value to auto-replied.

If this logic were to be expressed in code, it might look something like this:

public static MimeMessage Reply (MimeMessage message, MailboxAddress from, bool replyToAll)
{
    var reply = new MimeMessage ();

    reply.From.Add (from);

    // reply to the sender of the message
    if (message.ReplyTo.Count > 0) {
        reply.To.AddRange (message.ReplyTo);
    } else if (message.From.Count > 0) {
        reply.To.AddRange (message.From);
    } else if (message.Sender != null) {
        reply.To.Add (message.Sender);
    }

    if (replyToAll) {
        // include all of the other original recipients - TODO: remove ourselves from these lists
        reply.To.AddRange (message.To);
        reply.Cc.AddRange (message.Cc);
    }

    // set the reply subject
    if (!message.Subject?.StartsWith ("Re:", StringComparison.OrdinalIgnoreCase))
        reply.Subject = "Re: " + (message.Subject ?? string.Empty);
    else
        reply.Subject = message.Subject;

    // construct the In-Reply-To and References headers
    if (!string.IsNullOrEmpty (message.MessageId)) {
        reply.InReplyTo = message.MessageId;
        foreach (var id in message.References)
            reply.References.Add (id);
        reply.References.Add (message.MessageId);
    }

    // if this is an automatic reply, be sure to specify this using the Auto-Submitted header in order to avoid (infinite) mail loops
    reply.Headers.Add (HeaderId.AutoSubmitted, "auto-replied");

    // quote the original message text
    using (var quoted = new StringWriter ()) {
        var sender = message.Sender ?? message.From.Mailboxes.FirstOrDefault ();

        quoted.WriteLine ("On {0}, {1} wrote:", message.Date.ToString ("f"), !string.IsNullOrEmpty (sender.Name) ? sender.Name : sender.Address);
        using (var reader = new StringReader (message.TextBody)) {
            string line;

            while ((line = reader.ReadLine ()) != null) {
                quoted.Write ("> ");
                quoted.WriteLine (line);
            }
        }

        reply.Body = new TextPart ("plain") {
            Text = quoted.ToString ()
        };
    }

    return reply;
}

But what if you wanted to reply to a message and quote the HTML formatting of the original message body (assuming it has an HTML body) while still including the embedded images?

This gets a bit more complicated, but it's still doable...

The first thing we'd need to do is implement our own MimeVisitor to handle this:

public class ReplyVisitor : MimeVisitor
{
    readonly Stack<Multipart> stack = new Stack<Multipart> ();
    MimeMessage original, reply;
    MailboxAddress from;
    bool replyToAll;
    int isRelated;

    /// <summary>
    /// Creates a new ReplyVisitor.
    /// </summary>
    public ReplyVisitor (MailboxAddress from, bool replyToAll)
    {
        this.replyToAll = replyToAll;
        this.from = from;
    }

    /// <summary>
    /// Gets the reply.
    /// </summary>
    /// <value>The reply.</value>
    public MimeMessage Reply {
        get { return reply; }
    }

    void Push (MimeEntity entity)
    {
        var multipart = entity as Multipart;

        if (reply.Body == null) {
            reply.Body = entity;
        } else {
            var parent = stack.Peek ();
            parent.Add (entity);
        }

        if (multipart != null)
            stack.Push (multipart);
    }

    void Pop ()
    {
        stack.Pop ();
    }

    static string GetOnDateSenderWrote (MimeMessage message)
    {
        var sender = message.Sender != null ? message.Sender : message.From.Mailboxes.FirstOrDefault ();
        var name = sender != null ? (!string.IsNullOrEmpty (sender.Name) ? sender.Name : sender.Address) : "an unknown sender";

        return string.Format ("On {0}, {1} wrote:", message.Date.ToString ("f"), name);
    }

    /// <summary>
    /// Visit the specified message.
    /// </summary>
    /// <param name="message">The message.</param>
    public override void Visit (MimeMessage message)
    {
        reply = new MimeMessage ();
        original = message;

        stack.Clear ();

        reply.From.Add (from.Clone ());

        // reply to the sender of the message
        if (message.ReplyTo.Count > 0) {
            reply.To.AddRange (message.ReplyTo);
        } else if (message.From.Count > 0) {
            reply.To.AddRange (message.From);
        } else if (message.Sender != null) {
            reply.To.Add (message.Sender);
        }

        if (replyToAll) {
            // include all of the other original recipients - TODO: remove ourselves from these lists
            reply.To.AddRange (message.To);
            reply.Cc.AddRange (message.Cc);
        }

        // set the reply subject
        if (!message.Subject?.StartsWith ("Re:", StringComparison.OrdinalIgnoreCase))
            reply.Subject = "Re: " + (message.Subject ?? string.Empty);
        else
            reply.Subject = message.Subject;

        // construct the In-Reply-To and References headers
        if (!string.IsNullOrEmpty (message.MessageId)) {
            reply.InReplyTo = message.MessageId;
            foreach (var id in message.References)
                reply.References.Add (id);
            reply.References.Add (message.MessageId);
        }

        base.Visit (message);
    }

    /// <summary>
    /// Visit the specified entity.
    /// </summary>
    /// <param name="entity">The MIME entity.</param>
    /// <exception cref="System.NotSupportedException">
    /// Only Visit(MimeMessage) is supported.
    /// </exception>
    public override void Visit (MimeEntity entity)
    {
        throw new NotSupportedException ();
    }

    protected override void VisitMultipartAlternative (MultipartAlternative alternative)
    {
        var multipart = new MultipartAlternative ();

        Push (multipart);

        for (int i = 0; i < alternative.Count; i++)
            alternative[i].Accept (this);

        Pop ();
    }

    protected override void VisitMultipartRelated (MultipartRelated related)
    {
        var multipart = new MultipartRelated ();
        var root = related.Root;

        Push (multipart);

        root.Accept (this);

        isRelated++;
        for (int i = 0; i < related.Count; i++) {
            if (related[i] != root)
                related[i].Accept (this);
        }
        isRelated--;

        Pop ();
    }

    protected override void VisitMultipart (Multipart multipart)
    {
        foreach (var part in multipart) {
            if (part is MultipartAlternative)
                part.Accept (this);
            else if (part is MultipartRelated)
                part.Accept (this);
            else if (part is TextPart)
                part.Accept (this);
        }
    }

    void HtmlTagCallback (HtmlTagContext ctx, HtmlWriter htmlWriter)
    {
        if (ctx.TagId == HtmlTagId.Body && !ctx.IsEmptyElementTag) {
            if (ctx.IsEndTag) {
                // end our opening <blockquote>
                htmlWriter.WriteEndTag (HtmlTagId.BlockQuote);

                // pass the </body> tag through to the output
                ctx.WriteTag (htmlWriter, true);
            } else {
                // pass the <body> tag through to the output
                ctx.WriteTag (htmlWriter, true);

                // prepend the HTML reply with "On {DATE}, {SENDER} wrote:"
                htmlWriter.WriteStartTag (HtmlTagId.P);
                htmlWriter.WriteText (GetOnDateSenderWrote (original));
                htmlWriter.WriteEndTag (HtmlTagId.P);

                // Wrap the original content in a <blockquote>
                htmlWriter.WriteStartTag (HtmlTagId.BlockQuote);
                htmlWriter.WriteAttribute (HtmlAttributeId.Style, "border-left: 1px #ccc solid; margin: 0 0 0 .8ex; padding-left: 1ex;");

                ctx.InvokeCallbackForEndTag = true;
            }
        } else {
            // pass the tag through to the output
            ctx.WriteTag (htmlWriter, true);
        }
    }

    string QuoteText (string text)
    {
        using (var quoted = new StringWriter ()) {
            quoted.WriteLine (GetOnDateSenderWrote (original));

            using (var reader = new StringReader (text)) {
                string line;

                while ((line = reader.ReadLine ()) != null) {
                    quoted.Write ("> ");
                    quoted.WriteLine (line);
                }
            }

            return quoted.ToString ();
        }
    }

    protected override void VisitTextPart (TextPart entity)
    {
        string text;

        if (entity.IsHtml) {
            var converter = new HtmlToHtml {
                HtmlTagCallback = HtmlTagCallback
            };

            text = converter.Convert (entity.Text);
        } else if (entity.IsFlowed) {
            var converter = new FlowedToText ();

            text = converter.Convert (entity.Text);
            text = QuoteText (text);
        } else {
            // quote the original message text
            text = QuoteText (entity.Text);
        }

        var part = new TextPart (entity.ContentType.MediaSubtype.ToLowerInvariant ()) {
            Text = text
        };

        Push (part);
    }

    protected override void VisitMessagePart (MessagePart entity)
    {
        // don't descend into message/rfc822 parts
    }

    protected override void VisitMimePart (MimePart entity)
    {
        if (isRelated > 0 || !entity.IsAttachment) {
            var parent = stack.Peek ();
            parent.Add (entity);
        }
    }
}
public static MimeMessage Reply (MimeMessage message, MailboxAddress from, bool replyToAll)
{
    var visitor = new ReplyVisitor (from, replyToAll);

    visitor.Visit (message);

    return visitor.Reply;
}

There are 2 common ways of forwarding a message: attaching the original message as an attachment and inlining the message body much like replying typically does. Which method you choose is up to you.

To forward a message by attaching it as an attachment, you would do something like this:

public static MimeMessage Forward (MimeMessage original, MailboxAddress from, IEnumerable<InternetAddress> to)
{
    var message = new MimeMessage ();
    message.From.Add (from);
    message.To.AddRange (to);

    // set the forwarded subject
    if (!original.Subject?.StartsWith ("FW:", StringComparison.OrdinalIgnoreCase))
        message.Subject = "FW: " + (original.Subject ?? string.Empty);
    else
        message.Subject = original.Subject;

    // create the main textual body of the message
    var text = new TextPart ("plain") { Text = "Here's the forwarded message:" };

    // create the message/rfc822 attachment for the original message
    var rfc822 = new MessagePart { Message = original };
    
    // create a multipart/mixed container for the text body and the forwarded message
    var multipart = new Multipart ("mixed");
    multipart.Add (text);
    multipart.Add (rfc822);

    // set the multipart as the body of the message
    message.Body = multipart;

    return message;
}

To forward a message by inlining the original message's text content, you can do something like this:

public static MimeMessage Forward (MimeMessage original, MailboxAddress from, IEnumerable<InternetAddress> to)
{
    var message = new MimeMessage ();
    message.From.Add (from);
    message.To.AddRange (to);

    // set the forwarded subject
    if (!original.Subject?.StartsWith ("FW:", StringComparison.OrdinalIgnoreCase))
        message.Subject = "FW: " + (original.Subject ?? string.Empty);
    else
        message.Subject = original.Subject;

    // quote the original message text
    using (var text = new StringWriter ()) {
        text.WriteLine ();
        text.WriteLine ("-------- Original Message --------");
        text.WriteLine ("Subject: {0}", original.Subject ?? string.Empty);
        text.WriteLine ("Date: {0}", DateUtils.FormatDate (original.Date));
        text.WriteLine ("From: {0}", original.From);
        text.WriteLine ("To: {0}", original.To);
        text.WriteLine ();

        text.Write (original.TextBody);

        message.Body = new TextPart ("plain") {
            Text = text.ToString ()
        };
    }

    return message;
}

Keep in mind that not all messages will have a TextBody available, so you'll have to find a way to handle those cases.

.NET Core (and ASP.NET Core by extension) and .NET 5 only provide the Unicode encodings, ASCII and ISO-8859-1 by default. Other text encodings are not available to your application unless your application registers the encoding provider that provides all of the additional encodings.

First, add a package reference for the System.Text.Encoding.CodePages nuget package to your project and then register the additional text encodings using the following code snippet:

System.Text.Encoding.RegisterProvider (System.Text.CodePagesEncodingProvider.Instance);

Note: The above code snippet should be safe to call in .NET Framework versions >= 4.6 as well.

ImapClient

If the folder is open (via Open), then the ImapFolder.Unread property will be kept up to date (at least as-of the latest command issued to the server).

If the folder isn't open, then you will need to query the unread state of the folder using the Status method with the appropriate StatusItems flag(s).

For example, to get the total and unread counts, you can do this:

folder.Status (StatusItems.Count | StatusItems.Unread);

int total = folder.Count;
int unread = folder.Unread;

The obvious solution is:

var query = SearchQuery.DeliveredAfter (dateRange.BeginDate)
    .And (SearchQuery.DeliveredBefore (dateRange.EndDate));
var results = folder.Search (query);

However, it has been reported to me that this doesn't work reliably depending on the IMAP server implementation.

If you find that this query doesn't get the expected results for your IMAP server, here's another solution that should always work:

var query = SearchQuery.Not (SearchQuery.DeliveredBefore (dateRange.BeginDate)
    .Or (SearchQuery.DeliveredAfter (dateRange.EndDate)));
var results = folder.Search (query);

If you get an InvalidOperationException with the message, "The ImapClient is currently busy processing a command.", it means that you are trying to use the ImapClient and/or one of its ImapFolders from multiple threads.

To avoid this situation, you'll need to lock the SyncRoot property of the ImapClient and ImapFolder objects when performing operations on them.

For example:

lock (client.SyncRoot) {
    client.NoOp ();
}

Note: Locking the SyncRoot is only necessary when using the synchronous API's. All Async() method variants already do this locking for you.

If you get this exception, it's probably because you thought you had to open the destination folder that you passed as an argument to one of the CopyTo or MoveTo methods. When you opened that destination folder, you also inadvertently closed the source folder which is why you are getting this exception.

The IMAP server can only have a single folder open at a time. Whenever you open a folder, you automatically close the previously opened folder.

When copying or moving messages from one folder to another, you only need to have the source folder open.

If you look at the source code for the ImapFolder.MoveTo() method, what you'll notice is that there are several code paths depending on the features that the IMAP server supports.

If the IMAP server supports the MOVE extension, then MailKit's MoveTo() method will use the MOVE command. I suspect that your server does not support the MOVE command or you probably wouldn't be seeing what you are seeing.

When the IMAP server does not support the MOVE command, MailKit has to use the COPY command to copy the message(s) to the destination folder. Once the COPY command has completed, it will then mark the messages that you asked it to move for deletion by setting the \Deleted flag on those messages.

If the server supports the UIDPLUS extension, then MailKit will attempt to EXPUNGE the subset of messages that it just marked for deletion, however, if the UIDPLUS extension is not supported by the IMAP server, then it cannot safely expunge just that subset of messages and so it stops there.

My guess is that your server supports neither MOVE nor UIDPLUS and that is why clients like Outlook continue to see the messages in your folder. I believe, however, that Outlook has a setting to show deleted messages with a strikeout (which you probably have disabled).

So to answer your question more succinctly: After calling folder.MoveTo (...);, if you are confident that the messages marked for deletion should be expunged, call folder.Expunge ();

The way to mark messages as read using the IMAP protocol is to set the \Seen flag on the message(s).

To do this using MailKit, you will first need to know either the index(es) or the UID(s) of the messages that you would like to set the \Seen flag on. Once you have that information, you will want to call one of the AddFlags methods on the ImapFolder. For example:

folder.AddFlags (uids, MessageFlags.Seen, true);

To mark messages as unread, you would remove the \Seen flag, like so:

folder.RemoveFlags (uids, MessageFlags.Seen, true);

Assuming your IMAP server does not support the QRESYNC extension (which simplifies this procedure a ton), here is some simple code to illustrate how to go about re-synchronizing your cache with the remote IMAP server.

/// <summary>
/// Just a simple class to represent the cached information about a message.
/// </summary>
class CachedMessageInfo
{
    public UniqueId UniqueId;
    public MessageFlags Flags;
    public HashSet<string> Keywords;
    public Envelope Envelope;
    public BodyPart Body;
}

/// <summary>
/// Resynchronize the cache with the remote IMAP folder.
/// </summary>
/// <param name="folder">The IMAP folder.</param>
/// <param name="cache">The local cache of message metadata.</param>
/// <param name="cachedUidValidity">The cached UIDVALIDITY value of the IMAP folder from a previous session.</param>
static void ResyncFolder (ImapFolder folder, List<CachedMessageInfo> cache, ref uint cachedUidValidity)
{
    IList<IMessageSummary> summaries;

    // Step 1: Open the folder.

    // Note: we only need read-only access to update our cache, but depending on
    // what you plan to do with the folder after resynchronizing, you may want
    // top open the folder in read-write mode instead.
    folder.Open (FolderAccess.ReadOnly);

    if (cache.Count > 0) {
        if (folder.UidValidity == cachedUidValidity) {
            // Step 2: Remove messages from our cache that no longer exist on the server.

            // get the full list of UIDs on the server...
            var all = folder.Search (SearchQuery.All);

            // remove any messages from our cache that no longer exist...
            for (int i = 0; i < cache.Count; i++) {
                if (!all.Contains (cache[i].UniqueId)) {
                    cache.RemoveAt (i);
                    i--;
                }
            }

            // Step 3: Sync any flag changes for our cached messages.

            // get a list of known uids... astute observers will note that an easy
            // optimization to make here would be to merge this loop with the above
            // loop.
            var known = new UniqueIdSet (SortOrder.Ascending);
            for (int i = 0; i < cache.Count; i++)
                known.Add (cache[i].UniqueId);

            // fetch the flags for our known messages...
            summaries = folder.Fetch (known, MessageSummaryItems.Flags);
            for (int i = 0; i < summaries.Count; i++) {
                // Note: the indexes should match up with our cache, but it wouldn't
                // hurt to add error checking to make sure. I'm not bothering to here
                // for simplicity reasons.
                cache[i].Flags = summaries[i].Flags.Value;
                cache[i].Keywords = summaries[i].Keywords;
            }
        } else {
            // The UIDVALIDITY of the folder has changed. This means that our entire
            // cache is obsolete. We need to clear our cache and start from scratch.
            cachedUidValidity = folder.UidValidity;
            cache.Clear ();
        }
    } else {
        // We have nothing cached, so just start from scratch.
        cachedUidValidity = folder.UidValidity;
    }

    // Step 4: Fetch the messages we don't already know about and add them to our cache.

    summaries = folder.Fetch (cache.Count, -1, MessageSummaryItems.UniqueId | MessageSummaryItems.Flags | MessageSummaryItems.Envelope | MessageSummaryItems.BodyStructure);
    for (int i = 0; i < summaries.Count; i++) {
        cache.Add (new CachedMessageInfo {
            UniqueId = summaries[i].UniqueId,
            Flags = summaries[i].Flags.Value,
            Keywords = summaries[i].Keywords,
            Envelope = summaries[i].Envelope,
            Body = summaries[i].Body
        });
    }

    // Tada! Now we are resynchronized with the server!
}
var result = await GetPublicClientOAuth2CredentialsAsync ("IMAP", "[email protected]");

// Note: We always use result.Account.Username instead of `Username` because the user may have selected an alternative account.
var oauth2 = new SaslMechanismOAuth2 (result.Account.Username, result.AccessToken);

using (var client = new ImapClient ()) {
    await client.ConnectAsync ("outlook.office365.com", 993, SecureSocketOptions.SslOnConnect);
    await client.AuthenticateAsync (oauth2);

    // ...

    await client.DisconnectAsync (true);
}

Notes:

  1. The GetPublicClientOAuth2CredentialsAsync() method used in this example code snippet can be found in the ExchangeOAuth2.md documentation.
  2. Some users have reported that they need to use "[email protected]\\sharedMailboxName" as their username instead of "[email protected]".

SmtpClient

It seems to be a common misunderstanding that messages sent via SMTP will magically show up in the account's "Sent Mail" folder.

In order for the message to show up in the "Sent Mail" folder, you will need to append the message to the "Sent Mail" folder yourself because the SMTP protocol does not support doing this automatically.

If the "Sent Mail" folder is a local mbox folder, you'll need to append it like this:

using (var mbox = File.Open ("C:\\path\\to\\Sent Mail.mbox", FileMode.Append, FileAccess.Write)) {
    var marker = string.Format ("From MAILER-DAEMON {0}{1}", DateTime.Now.ToString (CultureInfo.InvariantCulture, "ddd MMM d HH:mm:ss yyyy"), Environment.NewLine);
    var bytes = Encoding.ASCII.GetBytes (marker);
    
    // Write the mbox marker bytes.
    mbox.Write (bytes, 0, bytes.Length);
    
    // Write the message, making sure to escape any line that looks like an mbox From-marker.
    using (var filtered = new FilteredStream (stream)) {
        filtered.Add (new MboxFromMarker ());
        message.WriteTo (filtered);
        filtered.Flush ();
    }
    
    mbox.Flush ();
}

If the "Sent Mail" folder exists on an IMAP server, you would need to do something more like this:

using (var client = new ImapClient ()) {
    client.Connect ("imap.server.com", 993, SecureSocketOptions.SslOnConnect);
    client.Authenticate ("username", "password");
    
    IMailFolder sentMail;
    
    if (client.Capabilities.HasFlag (ImapCapabilities.SpecialUse)) {
        sentMail = client.GetFolder (SpecialFolder.Sent);
    } else {
        var personal = client.GetFolder (client.PersonalNamespaces[0]);
        
        // Note: This assumes that the "Sent Mail" folder lives at the root of the folder hierarchy
        // and is named "Sent Mail" as opposed to "Sent" or "Sent Items" or any other variation.
        sentMail = personal.GetSubfolder ("Sent Mail");
    }
    
    sentMail.Append (message, MessageFlags.Seen);
    
    client.Disconnect (true);
}

Based on Microsoft's referencesource, when SmtpDeliveryMethod.SpecifiedPickupDirectory is used, the SmtpClient saves the message to the specified pickup directory location using a randomly generated filename based on Guid.NewGuid ().ToString () + ".eml", so to achieve the same results with MailKit, you could do something like this:

public static void SaveToPickupDirectory (MimeMessage message, string pickupDirectory)
{
    do {
        // Generate a random file name to save the message to.
        var path = Path.Combine (pickupDirectory, Guid.NewGuid ().ToString () + ".eml");
        Stream stream;

        try {
            // Attempt to create the new file.
            stream = File.Open (path, FileMode.CreateNew);
        } catch (IOException) {
            // If the file already exists, try again with a new Guid.
            if (File.Exists (path))
                continue;

            // Otherwise, fail immediately since it probably means that there is
            // no graceful way to recover from this error.
            throw;
        }

        try {
            using (stream) {
                // IIS pickup directories expect the message to be "byte-stuffed"
                // which means that lines beginning with "." need to be escaped
                // by adding an extra "." to the beginning of the line.
                //
                // Use an SmtpDataFilter "byte-stuff" the message as it is written
                // to the file stream. This is the same process that an SmtpClient
                // would use when sending the message in a `DATA` command.
                using (var filtered = new FilteredStream (stream)) {
                    filtered.Add (new SmtpDataFilter ());

                    // Make sure to write the message in DOS (<CR><LF>) format.
                    var options = FormatOptions.Default.Clone ();
                    options.NewLineFormat = NewLineFormat.Dos;

                    message.WriteTo (options, filtered);
                    filtered.Flush ();
                    return;
                }
            }
        } catch {
            // An exception here probably means that the disk is full.
            //
            // Delete the file that was created above so that incomplete files are not
            // left behind for IIS to send accidentally.
            File.Delete (path);
            throw;
        }
    } while (true);
}

The first thing I need to make clear is that requesting a notification does not guarantee that you'll actually get one. In order for you to receive a notification that the message was read by its recipient, the recipient's mail client needs to know how to send such a notification and that the user has enabled it to do so.

That said, here's how you can request a notification when the recipient reads the message that has been sent:

// Add the following header to tell the recipient's client that you want to receive a
// notification when the message has been read by the user.
message.Headers[HeaderId.DispositionNotificationTo] = new MailboxAddress ("My Name", "[email protected]").ToString (true);

For more information on this topic, read rfc3798.

A read receipt notification comes in the form of a MIME message with a top-level MIME part with a MIME-type of multipart/report that has a report-type parameter with a value of disposition-notification.

You could check for this in code like this:

var report = message.Body as MultipartReport;
if (report != null && report.ReportType.Equals ("disposition-notification", StringComparison.OrdinalIgnoreCase)) {
    // This is a read receipt notification.
}

The first part of the multipart/report will be a human-readable explanation of the notification.

The second part will have a MIME-type of message/disposition-notification and be represented by a MessageDispositionNotification.

This notification part will contain a list of header-like fields containing information about the message that this notification is for such as the Original-Message-Id, Original-Recipient, etc.

var notification = report[1] as MessageDispositionNotification;
if (notification != null) {
    // Get the Message-Id of the message this notification is for...
    var messageId = notification.Fields["Original-Message-Id"];
}

For more information on this topic, read rfc3798.