Click here to Skip to main content
Click here to Skip to main content

MIME Message Composer/Analyser

, 18 Apr 2004
Rate this:
Please Sign up or sign in to vote.
A C++ implementation of MIME


This implementation is designed to provide an easy-to-use, well-encapsulated component for the development of MIME-conformant applications, with less MIME knowledge. However, because MIME is open and extensible, to provide an easy way to support any customized extension is also important. Here are some features of this implementation:

  • Meet the requirements of MIME-conformance.
  • Written in ANSI C++, easy to migrate to UNIX.
  • Full control on all (standard and extended) header fields and body parts. Provide wrappers on raw data to interpret the standard fields.
  • Provide standard encoding mechanism (Quoted-Printable and Base64) and support all the standard media types.
  • Seamless interface to support extended media types and encoding mechanisms.
  • Provide standard header encoding/decoding for non-ASCII text, and folding/unfolding for long line text.

Using the code

Although there're more than 10 classes in the source files, the only 3 or 4 classes you need to work with, include:

CMimeEnvironment The global environment to manage encoding mechanisms and options.
CMimeField Represents a field of a MIME body part header. Provides methods to manage the name, value, parameters and charset of the field.
CMimeBody Represents a body part of a MIME message. It's derived from CMimeHeader class, provides methods to manage the content and header fields of the body part.
CMimeMessage Represents a MIME message. It's derived from CMimeBody class, provides extra functionality to access RFC822 message header fields (e.g. From, To, Date, Subject).

Compose a message

This example demonstrates how to create a complex multipart message with a text part, a file attachment, an attached message, and an embedded multipart entity.

    CMimeMessage mail;

    // Initialize message header
    mail.SetDate(); // set 'Date' field to the current time
      Nick Name <>, 
      \"Nick Name\" <>");
    mail.SetSubject("Test message");
    mail.SetFieldValue("X-Priority", "3 (Normal)"); // extended field
    mail.SetFieldValue("X-My-Field", "My value");   // user-defined field

    // Initialize body part header
    // generate a boundary delimeter automatically
    // if the parameter is NULL

    // Add a text body part
    // Content-Type is not specified, so the default
    // value "text/plain" is implicitly used
    // Content-Transfer-Encoding is not specified
    // so the default value "7bit" is implicitly used
    CMimeBody* pBp = mail.CreatePart();
    pBp->SetText("Hi, there");  // set the content of the body part

    // Add a file attachment body part
    pBp = mail.CreatePart();
    pBp->SetContentDescription("enclosed photo");
    // if Content-Type is not specified, it'll be
    // set to "image/jpeg" by ReadFromFile()

    // Generate a simple message
    CMimeMessage mail2;
    mail2.SetSubject("This is an attached message");
    mail2.SetText("Content of attached message.\r\n");

    // Attach the message
    pBp = mail.CreatePart();
    pBp->SetContentDescription("enclosed message");
    // if Content-Type is not specified, it'll be
    // set to "message/rfc822" by SetMessage()

    // Add an embeded multipart
    pBp = mail.CreatePart();
    CMimeBody *pBpChild = pBp->CreatePart();
    pBpChild->SetText("Content of Part 1\r\n");
    pBpChild = pBp->CreatePart();
    pBpChild->SetText("Content of Part 2\r\n");

    // Store the message to buffer    
    // fold the long lines in the headers
    int nSize = mail.GetLength();
    char* pBuff = new char[nSize];
    nSize = mail.Store(pBuff, nSize);
    delete pBuff;

Analyze a message

To analyze a message is the same and simple: call Load() to load the message object from buffer, then call FindFirstPart()/FindNextPart() functions to iterate its child body parts. But the fact that a multipart entity could have embedded multipart entities brings some complexities. We have to call FindFirstPart()/FindNextPart() on each child body part recursively to iterate all the descendant parts. So CMimeBody class provides an alternative way, to retrieve a list of all the descendant body parts, by calling GetBodyPartList().

    CMimeMessage mail;
    int nLoadedSize = mail.Load(pBuff, nDataSize);

    // Analyze the message header
    const char* pszField;
    pszField = mail.GetSubject();
    if (pszField != NULL)
        printf("Subject: %s\r\n", pszField);
    pszField = mail.GetFrom();
    if (pszField != NULL)
        printf("From: %s\r\n", pszField);
    pszField = mail.GetFieldValue("X-Priority");
    if (pszField != NULL)
        printf("X-Priority: %s\r\n", pszField);

    // Iterate all the descendant body parts
    CMimeBody::CBodyList bodies;
    int nCount = mail.GetBodyPartList(bodies);
    CMimeBody::CBodyList::const_iterator it;
    for (it=bodies.begin(); it!=bodies.end(); it++)
        CMimeBody* pBP = *it;

        // Iterate all the header fields of this body part:
        CMimeHeader::CFieldList& fds = pBP->Fields();
        CMimeHeader::CFieldList::const_iterator itfd;
        for (itfd=fds.begin(); itfd!=fds.end(); itfd++)
            const CMimeField& fd = *itfd;
            printf("%s: %s\r\n", fd.GetName(), fd.GetValue());

        if (pBP->IsText())
            string strText;
            printf("Content: %s\r\n", strText.c_str());
        else if (pBP->IsAttachment())
            string strName = pBP->GetName();
            printf("File name: %s\r\n", strName.c_str());
            printf("File size: %d\r\n", pBP->GetContentLength());
            strName = "d:\\download\\" + strName;

Header encoding and folding

The header encoding is to encode any non-ASCII text data in message headers to 'encoded-word'. For example, a field like "Subject: Bonne Année" will be encoded to "Subject: =?iso-8859-1?Q?Bonne=20Ann=E9e?=" and "From: René <>" will be encoded to "From: =?iso-8859-1?Q?Ren=E9?= <>".

During the period of storing (invoking CMimeBody::Store() function), the header encoding is performed automatically if a field contains non-ASCII text, and the 'charset' of this field has been specified. The 'charset' of a field can be specified by calling CMimeBody::SetFieldCharset(), or by passing a charset string to functions like CMimeBody::SetFieldValue(), CMimeMessage::SetSubject(), CMimeMessage::SetFrom(), etc.

Instead of specifying charset for each field repeatedly, you can specify the 'global charset' by calling CMimeEnvironment::SetGlobalCharset(). When the header encoding is being performed, the global charset is used if the field's charset is empty. If the global charset is empty too, the header encoding doesn't occur. So you can leave both charsets empty if you don't need the header encoding.

    mail.SetSubject("Bonne Année", "iso-8859-1");
    mail.SetFieldValue("Content-Description", "carte bonne année");

Furthermore, header folding is performed automatically too. In general, any header lines longer than 76 bytes will be folded by inserting "\r\n\t" at the position of a space (SPACE or TAB). For address fields (such as From, To, CC, etc.), the folding occurs between addresses, after the separating comma. However, the header folding can be turned on/off by calling CMimeEnvironment::SetAutoFolding() function with bAutoFolding parameter set to true/false.

Support of custom encoding

To support any Content-Transfer-Encoding mechanisms other than Quoted-Printable and Base64, derive a class from CMimeCodeBase, then register this class by invoking REGISTER_MIMECODER().

class CMyCoder : public CMimeCodeBase
    virtual int GetEncodeLength() const;
    virtual int GetDecodeLength() const;
    virtual int Encode(unsigned char* pbOutput, int nMaxSize) const;
    virtual int Decode(unsigned char* pbOutput, int nMaxSize);

    CMimeMessage mail;
    REGISTER_MIMECODER("my_coding_name", CMyCoder);
    mail.Store(pbBuff, nSize);
    mail.Load(pbBuff, nSize);

After REGISTER_MIMECODER() being called, for any body parts with Content-Transfer-Encoding specified to my_coding_name, CMyCoder class will be used to encode/decode the content by Store()/Load() function.

Support of extended media type

The use of extended media types is rare but still supported. Similar to support for custom encoding, derive a class from CMimeBody, then register the class by invoking REGISTER_MEDIATYPE().

class CMyBodyPart : public CMimeBody
    virtual void Clear();
    virtual int GetLength() const;
    virtual int Store(char* pszData, int nMaxSize) const;
    virtual int Load(const char* pszData, int nDataSize);

    CMimeMessage mail;
    REGISTER_MEDIATYPE("my_media_type", CMyBodyPart);
    CMyBodyPart* pBP = (CMyBodyPart*) mail.CreatePart("my_media_type");
    mail.Store(pbBuff, nSize);
    mail.Load(pbBuff, nSize);

After REGISTER_MEDIATYPE() being called, for any body parts with Content-Type specified to my_media_type, CMyBodyPart object will be created to represent this body part.


  • 17 Feb 2004 - updated download
  • 16 Apr 2004 - updated download


This article has no explicit license attached to it but may contain usage terms in the article text or the download files themselves. If in doubt please contact the author via the discussion board below.

A list of licenses authors might use can be found here


About the Author

Web Developer
Canada Canada
Nick Adams is one of my favorite figures in Hemingway's stories. I use it because Jeff Lee has been occupied on Codeproject.

Comments and Discussions

BugThanks, but, i got some problems PinmemberXie Jinghui2-Aug-12 23:57 
QuestionSubject decode Pinmemberpiccoola30-Jul-12 2:12 
GeneralMy vote of 5 Pinmemberytfrdfiw16-Jul-12 16:29 
GeneralMy vote of 5 PinmemberXie Jinghui23-Aug-11 0:15 
GeneralMIME Message Composer/Analyser Pinmemberaaa945260415-Feb-11 6:42 
Generaljust a dumm question Pinmemberkampi4-Jan-11 4:08 
Questionhow to use utf-8 Pinmemberdie_for_rock_vn4-Dec-10 9:33 
How to use utf-8 for subject and content?
i tried:
but it doesn't work.
And how to analyze a unicode(utf-8) message?
AnswerRe: how to use utf-8 Pinmemberdie_for_rock_vn7-Dec-10 22:19 
Generalit's good PinmemberMember 41124855-Aug-10 23:12 
GeneralLATEST CODE Pinmemberrichardwhitehead23-Sep-09 1:12 
GeneralRe: LATEST CODE Pinmemberrichardwhitehead8-Mar-10 22:36 
GeneralRe: LATEST CODE PinmemberJose Schmidt12-Oct-10 6:56 
GeneralRe: LATEST CODE Pinmemberrichardwhitehead12-Oct-10 10:44 
GeneralRe: LATEST CODE PinmemberJose Schmidt12-Oct-10 12:04 
GeneralVery nice work! Pinmembermedium21-Sep-09 11:31 
GeneralLicense PinmemberMember 384073731-Aug-08 5:18 
GeneralUsing string function strstr() may be lead to bug Pinmemberzhtsh22-Jul-08 21:32 
QuestionHow to SetFileName PinmemberPelle191110-Jul-08 10:38 
GeneralEncoded or not encoded, that is the question... PinmemberPelle19111-Jul-08 1:32 
GeneralBig thanks PinmemberJerry Evans13-Mar-08 12:20 
GeneralRe: Big thanks Pinmemberawilbourn22-Mar-08 9:57 
GeneralRe: Big thanks PinmemberJerry Evans22-Mar-08 10:38 
GeneralRe: Big thanks Pinmemberawilbourn22-Mar-08 14:11 
GeneralRe: Big thanks PinmemberJerry Evans22-Mar-08 15:19 
GeneralRe: Big thanks PinmemberOmar.Pessoa24-Sep-08 9:44 
GeneralDouble dots in begin of line for plain/text attachments [modified] PinmemberCoderCZ17-Oct-07 3:39 
QuestionHow to accomodate emails written in chinese language? PinmemberAnindyaM29-Aug-07 18:41 
AnswerRe: How to accomodate emails written in chinese language? Pinmemberkanitamildasan25-Sep-07 19:29 
QuestionHow to get the "Received" Field Pinmemberannewang66315-Aug-07 22:24 
Generaliso-8859-2 in subject Pinmemberbenicky8-Jun-07 8:20 
GeneralRe: iso-8859-2 in subject [modified] PinmemberCoderCZ9-Jul-07 0:37 
QuestionHow do I handle message/partial content types Pinmemberpeterdelchiappo13-Oct-06 0:03 
QuestionLicense terms Pinmemberxquiky1-Oct-06 13:59 
GeneralGood work Pinmemberwaldermort31-Aug-06 21:05 
GeneralLoad function is too slow PinmemberKFC12330-Jul-06 20:26 
GeneralCMimeBody::ReadFromFile losing last block on large file Pinmemberrinzai7-Jun-06 6:03 
GeneralRe: CMimeBody::ReadFromFile losing last block on large file Pinmemberrinzai7-Jun-06 6:10 
GeneralRe: CMimeBody::ReadFromFile losing last block on large file Pinmemberrinzai7-Jun-06 6:54 
GeneralRe: CMimeBody::ReadFromFile losing last block on large file Pinmembergu7-Sep-06 3:58 
GeneralIt need 2 times of program memory for CMimeBody::Load PinmemberEaway11-May-06 0:27 
QuestionCMimeBody::IsAttachment()? PinmemberEaway19-Apr-06 19:18 
QuestionVedio? Pinmemberthenickname10-Jan-06 10:22 
QuestionReadFromFile problem with large files PinmemberRobin Minto6-Jan-06 13:44 
Generalvery useful Pinmemberpglidden31-Dec-05 8:25 
QuestionHow to decode japanese characters Pinmemberjuncki15-Dec-05 10:24 
AnswerRe: How to decode japanese characters Pinmemberjuncki15-Dec-05 14:47 
QuestionRe: How to decode japanese characters PinmemberAbhinav agrawal14-May-06 5:00 
QuestionHow to use the MIME message, i.e how to send it out? Pinmemberkezhu8-Dec-05 15:53 
Questionnecessary sample for parsing message ! Pinmemberjsli@ukr.net28-Nov-05 23:44 
Question8bit and CP_UTF8 PinmemberdennisV10-Nov-05 12:20 

General General    News News    Suggestion Suggestion    Question Question    Bug Bug    Answer Answer    Joke Joke    Rant Rant    Admin Admin   

Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages.

| Advertise | Privacy | Terms of Use | Mobile
Web04 | 2.8.150326.1 | Last Updated 19 Apr 2004
Article Copyright 2004 by nickadams
Everything else Copyright © CodeProject, 1999-2015
Layout: fixed | fluid