'Protocol'에 해당되는 글 2건

  1. 2009.08.31 RFC 3501 (imap4 protocol)
  2. 2009.08.10 SMTP protocol (RFC 821) - (ESMTP는 RFC 1821)


RFC3501 - INTERNET MESSAGE ACCESS PROTOCOL - VERSION 4rev1

Network Working Group                                         M. Crispin
Request for Comments: 3501                      University of Washington
Obsoletes: 2060                                               March 2003
Category: Standards Track

            INTERNET MESSAGE ACCESS PROTOCOL - VERSION 4rev1

Status of this Memo

   This document specifies an Internet standards track protocol for the
   Internet community, and requests discussion and suggestions for
   improvements.  Please refer to the current edition of the "Internet
   Official Protocol Standards" (STD 1) for the standardization state
   and status of this protocol.  Distribution of this memo is unlimited.

Copyright Notice

   Copyright (C) The Internet Society (2003).  All Rights Reserved.

Abstract

   The Internet Message Access Protocol, Version 4rev1 (IMAP4rev1)
   allows a client to access and manipulate electronic mail messages on
   a server.  IMAP4rev1 permits manipulation of mailboxes (remote
   message folders) in a way that is functionally equivalent to local
   folders.  IMAP4rev1 also provides the capability for an offline
   client to resynchronize with the server.

   IMAP4rev1 includes operations for creating, deleting, and renaming
   mailboxes, checking for new messages, permanently removing messages,
   setting and clearing flags, RFC 2822 and RFC 2045 parsing, searching,
   and selective fetching of message attributes, texts, and portions
   thereof.  Messages in IMAP4rev1 are accessed by the use of numbers.
   These numbers are either message sequence numbers or unique
   identifiers.

   IMAP4rev1 supports a single server.  A mechanism for accessing
   configuration information to support multiple IMAP4rev1 servers is
   discussed in RFC 2244.

   IMAP4rev1 does not specify a means of posting mail; this function is
   handled by a mail transfer protocol such as RFC 2821.

Table of Contents

   IMAP4rev1 Protocol Specification ................................  4
   1.      How to Read This Document ...............................  4
   1.1.    Organization of This Document ...........................  4
   1.2.    Conventions Used in This Document .......................  4
   1.3.    Special Notes to Implementors ...........................  5
   2.      Protocol Overview .......................................  6
   2.1.    Link Level ..............................................  6
   2.2.    Commands and Responses ..................................  6
   2.2.1.  Client Protocol Sender and Server Protocol Receiver .....  6
   2.2.2.  Server Protocol Sender and Client Protocol Receiver .....  7
   2.3.    Message Attributes ......................................  8
   2.3.1.  Message Numbers .........................................  8
   2.3.1.1.        Unique Identifier (UID) Message Attribute .......  8
   2.3.1.2.        Message Sequence Number Message Attribute ....... 10
   2.3.2.  Flags Message Attribute ................................. 11
   2.3.3.  Internal Date Message Attribute ......................... 12
   2.3.4.  [RFC-2822] Size Message Attribute ....................... 12
   2.3.5.  Envelope Structure Message Attribute .................... 12
   2.3.6.  Body Structure Message Attribute ........................ 12
   2.4.    Message Texts ........................................... 13
   3.      State and Flow Diagram .................................. 13
   3.1.    Not Authenticated State ................................. 13
   3.2.    Authenticated State ..................................... 13
   3.3.    Selected State .......................................... 13
   3.4.    Logout State ............................................ 14
   4.      Data Formats ............................................ 16
   4.1.    Atom .................................................... 16
   4.2.    Number .................................................. 16
   4.3.    String .................................................. 16
   4.3.1.  8-bit and Binary Strings ................................ 17
   4.4.    Parenthesized List ...................................... 17
   4.5.    NIL ..................................................... 17
   5.      Operational Considerations .............................. 18
   5.1.    Mailbox Naming .......................................... 18
   5.1.1.  Mailbox Hierarchy Naming ................................ 19
   5.1.2.  Mailbox Namespace Naming Convention ..................... 19
   5.1.3.  Mailbox International Naming Convention ................. 19
   5.2.    Mailbox Size and Message Status Updates ................. 21
   5.3.    Response when no Command in Progress .................... 21
   5.4.    Autologout Timer ........................................ 22
   5.5.    Multiple Commands in Progress ........................... 22
   6.      Client Commands ........................................  23
   6.1.    Client Commands - Any State ............................  24
   6.1.1.  CAPABILITY Command .....................................  24
   6.1.2.  NOOP Command ...........................................  25
   6.1.3.  LOGOUT Command .........................................  26

   6.2.    Client Commands - Not Authenticated State ..............  26
   6.2.1.  STARTTLS Command .......................................  27
   6.2.2.  AUTHENTICATE Command ...................................  28
   6.2.3.  LOGIN Command ..........................................  30
   6.3.    Client Commands - Authenticated State ..................  31
   6.3.1.  SELECT Command .........................................  32
   6.3.2.  EXAMINE Command ........................................  34
   6.3.3.  CREATE Command .........................................  34
   6.3.4.  DELETE Command .........................................  35
   6.3.5.  RENAME Command .........................................  37
   6.3.6.  SUBSCRIBE Command ......................................  39
   6.3.7.  UNSUBSCRIBE Command ....................................  39
   6.3.8.  LIST Command ...........................................  40
   6.3.9.  LSUB Command ...........................................  43
   6.3.10. STATUS Command .........................................  44
   6.3.11. APPEND Command .........................................  46
   6.4.    Client Commands - Selected State .......................  47
   6.4.1.  CHECK Command ..........................................  47
   6.4.2.  CLOSE Command ..........................................  48
   6.4.3.  EXPUNGE Command ........................................  49
   6.4.4.  SEARCH Command .........................................  49
   6.4.5.  FETCH Command ..........................................  54
   6.4.6.  STORE Command ..........................................  58
   6.4.7.  COPY Command ...........................................  59
   6.4.8.  UID Command ............................................  60
   6.5.    Client Commands - Experimental/Expansion ...............  62
   6.5.1.  X<atom> Command ........................................  62
   7.      Server Responses .......................................  62
   7.1.    Server Responses - Status Responses ....................  63
   7.1.1.  OK Response ............................................  65
   7.1.2.  NO Response ............................................  66
   7.1.3.  BAD Response ...........................................  66
   7.1.4.  PREAUTH Response .......................................  67
   7.1.5.  BYE Response ...........................................  67
   7.2.    Server Responses - Server and Mailbox Status ...........  68
   7.2.1.  CAPABILITY Response ....................................  68
   7.2.2.  LIST Response ..........................................  69
   7.2.3.  LSUB Response ..........................................  70
   7.2.4   STATUS Response ........................................  70
   7.2.5.  SEARCH Response ........................................  71
   7.2.6.  FLAGS Response .........................................  71
   7.3.    Server Responses - Mailbox Size ........................  71
   7.3.1.  EXISTS Response ........................................  71
   7.3.2.  RECENT Response ........................................  72
   7.4.    Server Responses - Message Status ......................  72
   7.4.1.  EXPUNGE Response .......................................  72
   7.4.2.  FETCH Response .........................................  73
   7.5.    Server Responses - Command Continuation Request ........  79

   8.      Sample IMAP4rev1 connection ............................  80
   9.      Formal Syntax ..........................................  81
   10.     Author's Note ..........................................  92
   11.     Security Considerations ................................  92
   11.1.   STARTTLS Security Considerations .......................  92
   11.2.   Other Security Considerations ..........................  93
   12.     IANA Considerations ....................................  94
   Appendices .....................................................  95
   A.      References .............................................  95
   B.      Changes from RFC 2060 ..................................  97
   C.      Key Word Index ......................................... 103
   Author's Address ............................................... 107
   Full Copyright Statement ....................................... 108

IMAP4rev1 Protocol Specification

1.      How to Read This Document

1.1.    Organization of This Document

   This document is written from the point of view of the implementor of
   an IMAP4rev1 client or server.  Beyond the protocol overview in
   section 2, it is not optimized for someone trying to understand the
   operation of the protocol.  The material in sections 3 through 5
   provides the general context and definitions with which IMAP4rev1
   operates.

   Sections 6, 7, and 9 describe the IMAP commands, responses, and
   syntax, respectively.  The relationships among these are such that it
   is almost impossible to understand any of them separately.  In
   particular, do not attempt to deduce command syntax from the command
   section alone; instead refer to the Formal Syntax section.

1.2.    Conventions Used in This Document

   "Conventions" are basic principles or procedures.  Document
   conventions are noted in this section.

   In examples, "C:" and "S:" indicate lines sent by the client and
   server respectively.

   The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT",
   "SHOULD", "SHOULD NOT", "MAY", and "OPTIONAL" in this document are to
   be interpreted as described in [KEYWORDS].

   The word "can" (not "may") is used to refer to a possible
   circumstance or situation, as opposed to an optional facility of the
   protocol.

   "User" is used to refer to a human user, whereas "client" refers to
   the software being run by the user.

   "Connection" refers to the entire sequence of client/server
   interaction from the initial establishment of the network connection
   until its termination.

   "Session" refers to the sequence of client/server interaction from
   the time that a mailbox is selected (SELECT or EXAMINE command) until
   the time that selection ends (SELECT or EXAMINE of another mailbox,
   CLOSE command, or connection termination).

   Characters are 7-bit US-ASCII unless otherwise specified.  Other
   character sets are indicated using a "CHARSET", as described in
   [MIME-IMT] and defined in [CHARSET].  CHARSETs have important
   additional semantics in addition to defining character set; refer to
   these documents for more detail.

   There are several protocol conventions in IMAP.  These refer to
   aspects of the specification which are not strictly part of the IMAP
   protocol, but reflect generally-accepted practice.  Implementations
   need to be aware of these conventions, and avoid conflicts whether or
   not they implement the convention.  For example, "&" may not be used
   as a hierarchy delimiter since it conflicts with the Mailbox
   International Naming Convention, and other uses of "&" in mailbox
   names are impacted as well.

1.3.    Special Notes to Implementors

   Implementors of the IMAP protocol are strongly encouraged to read the
   IMAP implementation recommendations document [IMAP-IMPLEMENTATION] in
   conjunction with this document, to help understand the intricacies of
   this protocol and how best to build an interoperable product.

   IMAP4rev1 is designed to be upwards compatible from the [IMAP2] and
   unpublished IMAP2bis protocols.  IMAP4rev1 is largely compatible with
   the IMAP4 protocol described in RFC 1730; the exception being in
   certain facilities added in RFC 1730 that proved problematic and were
   subsequently removed.  In the course of the evolution of IMAP4rev1,
   some aspects in the earlier protocols have become obsolete.  Obsolete
   commands, responses, and data formats which an IMAP4rev1
   implementation can encounter when used with an earlier implementation
   are described in [IMAP-OBSOLETE].

   Other compatibility issues with IMAP2bis, the most common variant of
   the earlier protocol, are discussed in [IMAP-COMPAT].  A full
   discussion of compatibility issues with rare (and presumed extinct)

   variants of [IMAP2] is in [IMAP-HISTORICAL]; this document is
   primarily of historical interest.

   IMAP was originally developed for the older [RFC-822] standard, and
   as a consequence several fetch items in IMAP incorporate "RFC822" in
   their name.  With the exception of RFC822.SIZE, there are more modern
   replacements; for example, the modern version of RFC822.HEADER is
   BODY.PEEK[HEADER].  In all cases, "RFC822" should be interpreted as a
   reference to the updated [RFC-2822] standard.

2.      Protocol Overview

2.1.    Link Level

   The IMAP4rev1 protocol assumes a reliable data stream such as that
   provided by TCP.  When TCP is used, an IMAP4rev1 server listens on
   port 143.

2.2.    Commands and Responses

   An IMAP4rev1 connection consists of the establishment of a
   client/server network connection, an initial greeting from the
   server, and client/server interactions.  These client/server
   interactions consist of a client command, server data, and a server
   completion result response.

   All interactions transmitted by client and server are in the form of
   lines, that is, strings that end with a CRLF.  The protocol receiver
   of an IMAP4rev1 client or server is either reading a line, or is
   reading a sequence of octets with a known count followed by a line.

2.2.1.  Client Protocol Sender and Server Protocol Receiver

   The client command begins an operation.  Each client command is
   prefixed with an identifier (typically a short alphanumeric string,
   e.g., A0001, A0002, etc.) called a "tag".  A different tag is
   generated by the client for each command.

   Clients MUST follow the syntax outlined in this specification
   strictly.  It is a syntax error to send a command with missing or
   extraneous spaces or arguments.

   There are two cases in which a line from the client does not
   represent a complete command.  In one case, a command argument is
   quoted with an octet count (see the description of literal in String
   under Data Formats); in the other case, the command arguments require
   server feedback (see the AUTHENTICATE command).  In either case, the

   server sends a command continuation request response if it is ready
   for the octets (if appropriate) and the remainder of the command.
   This response is prefixed with the token "+".

        Note: If instead, the server detected an error in the
        command, it sends a BAD completion response with a tag
        matching the command (as described below) to reject the
        command and prevent the client from sending any more of the
        command.

        It is also possible for the server to send a completion
        response for some other command (if multiple commands are
        in progress), or untagged data.  In either case, the
        command continuation request is still pending; the client
        takes the appropriate action for the response, and reads
        another response from the server.  In all cases, the client
        MUST send a complete command (including receiving all
        command continuation request responses and command
        continuations for the command) before initiating a new
        command.

   The protocol receiver of an IMAP4rev1 server reads a command line
   from the client, parses the command and its arguments, and transmits
   server data and a server command completion result response.

2.2.2.  Server Protocol Sender and Client Protocol Receiver

   Data transmitted by the server to the client and status responses
   that do not indicate command completion are prefixed with the token
   "*", and are called untagged responses.

   Server data MAY be sent as a result of a client command, or MAY be
   sent unilaterally by the server.  There is no syntactic difference
   between server data that resulted from a specific command and server
   data that were sent unilaterally.

   The server completion result response indicates the success or
   failure of the operation.  It is tagged with the same tag as the
   client command which began the operation.  Thus, if more than one
   command is in progress, the tag in a server completion response
   identifies the command to which the response applies.  There are
   three possible server completion responses: OK (indicating success),
   NO (indicating failure), or BAD (indicating a protocol error such as
   unrecognized command or command syntax error).

   Servers SHOULD enforce the syntax outlined in this specification
   strictly.  Any client command with a protocol syntax error, including
   (but not limited to) missing or extraneous spaces or arguments,

   SHOULD be rejected, and the client given a BAD server completion
   response.

   The protocol receiver of an IMAP4rev1 client reads a response line
   from the server.  It then takes action on the response based upon the
   first token of the response, which can be a tag, a "*", or a "+".

   A client MUST be prepared to accept any server response at all times.
   This includes server data that was not requested.  Server data SHOULD
   be recorded, so that the client can reference its recorded copy
   rather than sending a command to the server to request the data.  In
   the case of certain server data, the data MUST be recorded.

   This topic is discussed in greater detail in the Server Responses
   section.

2.3.    Message Attributes

   In addition to message text, each message has several attributes
   associated with it.  These attributes can be retrieved individually
   or in conjunction with other attributes or message texts.

2.3.1.  Message Numbers

   Messages in IMAP4rev1 are accessed by one of two numbers; the unique
   identifier or the message sequence number.

2.3.1.1.        Unique Identifier (UID) Message Attribute

   A 32-bit value assigned to each message, which when used with the
   unique identifier validity value (see below) forms a 64-bit value
   that MUST NOT refer to any other message in the mailbox or any
   subsequent mailbox with the same name forever.  Unique identifiers
   are assigned in a strictly ascending fashion in the mailbox; as each
   message is added to the mailbox it is assigned a higher UID than the
   message(s) which were added previously.  Unlike message sequence
   numbers, unique identifiers are not necessarily contiguous.

   The unique identifier of a message MUST NOT change during the
   session, and SHOULD NOT change between sessions.  Any change of
   unique identifiers between sessions MUST be detectable using the
   UIDVALIDITY mechanism discussed below.  Persistent unique identifiers
   are required for a client to resynchronize its state from a previous
   session with the server (e.g., disconnected or offline access
   clients); this is discussed further in [IMAP-DISC].

   Associated with every mailbox are two values which aid in unique
   identifier handling: the next unique identifier value and the unique
   identifier validity value.

   The next unique identifier value is the predicted value that will be
   assigned to a new message in the mailbox.  Unless the unique
   identifier validity also changes (see below), the next unique
   identifier value MUST have the following two characteristics.  First,
   the next unique identifier value MUST NOT change unless new messages
   are added to the mailbox; and second, the next unique identifier
   value MUST change whenever new messages are added to the mailbox,
   even if those new messages are subsequently expunged.

        Note: The next unique identifier value is intended to
        provide a means for a client to determine whether any
        messages have been delivered to the mailbox since the
        previous time it checked this value.  It is not intended to
        provide any guarantee that any message will have this
        unique identifier.  A client can only assume, at the time
        that it obtains the next unique identifier value, that
        messages arriving after that time will have a UID greater
        than or equal to that value.

   The unique identifier validity value is sent in a UIDVALIDITY
   response code in an OK untagged response at mailbox selection time.
   If unique identifiers from an earlier session fail to persist in this
   session, the unique identifier validity value MUST be greater than
   the one used in the earlier session.

        Note: Ideally, unique identifiers SHOULD persist at all
        times.  Although this specification recognizes that failure
        to persist can be unavoidable in certain server
        environments, it STRONGLY ENCOURAGES message store
        implementation techniques that avoid this problem.  For
        example:

         1) Unique identifiers MUST be strictly ascending in the
            mailbox at all times.  If the physical message store is
            re-ordered by a non-IMAP agent, this requires that the
            unique identifiers in the mailbox be regenerated, since
            the former unique identifiers are no longer strictly
            ascending as a result of the re-ordering.

         2) If the message store has no mechanism to store unique
            identifiers, it must regenerate unique identifiers at
            each session, and each session must have a unique
            UIDVALIDITY value.

         3) If the mailbox is deleted and a new mailbox with the
            same name is created at a later date, the server must
            either keep track of unique identifiers from the
            previous instance of the mailbox, or it must assign a
            new UIDVALIDITY value to the new instance of the
            mailbox.  A good UIDVALIDITY value to use in this case
            is a 32-bit representation of the creation date/time of
            the mailbox.  It is alright to use a constant such as
            1, but only if it guaranteed that unique identifiers
            will never be reused, even in the case of a mailbox
            being deleted (or renamed) and a new mailbox by the
            same name created at some future time.

         4) The combination of mailbox name, UIDVALIDITY, and UID
            must refer to a single immutable message on that server
            forever.  In particular, the internal date, [RFC-2822]
            size, envelope, body structure, and message texts
            (RFC822, RFC822.HEADER, RFC822.TEXT, and all BODY[...]
            fetch data items) must never change.  This does not
            include message numbers, nor does it include attributes
            that can be set by a STORE command (e.g., FLAGS).

2.3.1.2.        Message Sequence Number Message Attribute

   A relative position from 1 to the number of messages in the mailbox.
   This position MUST be ordered by ascending unique identifier.  As
   each new message is added, it is assigned a message sequence number
   that is 1 higher than the number of messages in the mailbox before
   that new message was added.

   Message sequence numbers can be reassigned during the session.  For
   example, when a message is permanently removed (expunged) from the
   mailbox, the message sequence number for all subsequent messages is
   decremented.  The number of messages in the mailbox is also
   decremented.  Similarly, a new message can be assigned a message
   sequence number that was once held by some other message prior to an
   expunge.

   In addition to accessing messages by relative position in the
   mailbox, message sequence numbers can be used in mathematical
   calculations.  For example, if an untagged "11 EXISTS" is received,
   and previously an untagged "8 EXISTS" was received, three new
   messages have arrived with message sequence numbers of 9, 10, and 11.
   Another example, if message 287 in a 523 message mailbox has UID
   12345, there are exactly 286 messages which have lesser UIDs and 236
   messages which have greater UIDs.

2.3.2.  Flags Message Attribute

   A list of zero or more named tokens associated with the message.  A
   flag is set by its addition to this list, and is cleared by its
   removal.  There are two types of flags in IMAP4rev1.  A flag of
   either type can be permanent or session-only.

   A system flag is a flag name that is pre-defined in this
   specification.  All system flags begin with "\".  Certain system
   flags (\Deleted and \Seen) have special semantics described
   elsewhere.  The currently-defined system flags are:

        \Seen
           Message has been read

        \Answered
           Message has been answered

        \Flagged
           Message is "flagged" for urgent/special attention

        \Deleted
           Message is "deleted" for removal by later EXPUNGE

        \Draft
           Message has not completed composition (marked as a draft).

        \Recent
           Message is "recently" arrived in this mailbox.  This session
           is the first session to have been notified about this
           message; if the session is read-write, subsequent sessions
           will not see \Recent set for this message.  This flag can not
           be altered by the client.

           If it is not possible to determine whether or not this
           session is the first session to be notified about a message,
           then that message SHOULD be considered recent.

           If multiple connections have the same mailbox selected
           simultaneously, it is undefined which of these connections
           will see newly-arrived messages with \Recent set and which
           will see it without \Recent set.

   A keyword is defined by the server implementation.  Keywords do not
   begin with "\".  Servers MAY permit the client to define new keywords
   in the mailbox (see the description of the PERMANENTFLAGS response
   code for more information).

   A flag can be permanent or session-only on a per-flag basis.
   Permanent flags are those which the client can add or remove from the
   message flags permanently; that is, concurrent and subsequent
   sessions will see any change in permanent flags.  Changes to session
   flags are valid only in that session.

        Note: The \Recent system flag is a special case of a
        session flag.  \Recent can not be used as an argument in a
        STORE or APPEND command, and thus can not be changed at
        all.

2.3.3.  Internal Date Message Attribute

   The internal date and time of the message on the server.  This
   is not the date and time in the [RFC-2822] header, but rather a
   date and time which reflects when the message was received.  In
   the case of messages delivered via [SMTP], this SHOULD be the
   date and time of final delivery of the message as defined by
   [SMTP].  In the case of messages delivered by the IMAP4rev1 COPY
   command, this SHOULD be the internal date and time of the source
   message.  In the case of messages delivered by the IMAP4rev1
   APPEND command, this SHOULD be the date and time as specified in
   the APPEND command description.  All other cases are
   implementation defined.

2.3.4.  [RFC-2822] Size Message Attribute

   The number of octets in the message, as expressed in [RFC-2822]
   format.

2.3.5.  Envelope Structure Message Attribute

   A parsed representation of the [RFC-2822] header of the message.
   Note that the IMAP Envelope structure is not the same as an
   [SMTP] envelope.

2.3.6.  Body Structure Message Attribute

   A parsed representation of the [MIME-IMB] body structure
   information of the message.

2.4.    Message Texts

   In addition to being able to fetch the full [RFC-2822] text of a
   message, IMAP4rev1 permits the fetching of portions of the full
   message text.  Specifically, it is possible to fetch the
   [RFC-2822] message header, [RFC-2822] message body, a [MIME-IMB]
   body part, or a [MIME-IMB] header.

3.      State and Flow Diagram

   Once the connection between client and server is established, an
   IMAP4rev1 connection is in one of four states.  The initial
   state is identified in the server greeting.  Most commands are
   only valid in certain states.  It is a protocol error for the
   client to attempt a command while the connection is in an
   inappropriate state, and the server will respond with a BAD or
   NO (depending upon server implementation) command completion
   result.

3.1.    Not Authenticated State

   In the not authenticated state, the client MUST supply
   authentication credentials before most commands will be
   permitted.  This state is entered when a connection starts
   unless the connection has been pre-authenticated.

3.2.    Authenticated State

   In the authenticated state, the client is authenticated and MUST
   select a mailbox to access before commands that affect messages
   will be permitted.  This state is entered when a
   pre-authenticated connection starts, when acceptable
   authentication credentials have been provided, after an error in
   selecting a mailbox, or after a successful CLOSE command.

3.3.    Selected State

   In a selected state, a mailbox has been selected to access.
   This state is entered when a mailbox has been successfully
   selected.

3.4.    Logout State

   In the logout state, the connection is being terminated.  This
   state can be entered as a result of a client request (via the
   LOGOUT command) or by unilateral action on the part of either
   the client or server.

   If the client requests the logout state, the server MUST send an
   untagged BYE response and a tagged OK response to the LOGOUT
   command before the server closes the connection; and the client
   MUST read the tagged OK response to the LOGOUT command before
   the client closes the connection.

   A server MUST NOT unilaterally close the connection without
   sending an untagged BYE response that contains the reason for
   having done so.  A client SHOULD NOT unilaterally close the
   connection, and instead SHOULD issue a LOGOUT command.  If the
   server detects that the client has unilaterally closed the
   connection, the server MAY omit the untagged BYE response and
   simply close its connection.

                   +----------------------+
                   |connection established|
                   +----------------------+
                              ||
                              \/
            +--------------------------------------+
            |          server greeting             |
            +--------------------------------------+
                      || (1)       || (2)        || (3)
                      \/           ||            ||
            +-----------------+    ||            ||
            |Not Authenticated|    ||            ||
            +-----------------+    ||            ||
             || (7)   || (4)       ||            ||
             ||       \/           \/            ||
             ||     +----------------+           ||
             ||     | Authenticated  |<=++       ||
             ||     +----------------+  ||       ||
             ||       || (7)   || (5)   || (6)   ||
             ||       ||       \/       ||       ||
             ||       ||    +--------+  ||       ||
             ||       ||    |Selected|==++       ||
             ||       ||    +--------+           ||
             ||       ||       || (7)            ||
             \/       \/       \/                \/
            +--------------------------------------+
            |               Logout                 |
            +--------------------------------------+
                              ||
                              \/
                +-------------------------------+
                |both sides close the connection|
                +-------------------------------+

         (1) connection without pre-authentication (OK greeting)
         (2) pre-authenticated connection (PREAUTH greeting)
         (3) rejected connection (BYE greeting)
         (4) successful LOGIN or AUTHENTICATE command
         (5) successful SELECT or EXAMINE command
         (6) CLOSE command, or failed SELECT or EXAMINE command
         (7) LOGOUT command, server shutdown, or connection closed

4.      Data Formats

   IMAP4rev1 uses textual commands and responses.  Data in
   IMAP4rev1 can be in one of several forms: atom, number, string,
   parenthesized list, or NIL.  Note that a particular data item
   may take more than one form; for example, a data item defined as
   using "astring" syntax may be either an atom or a string.

4.1.    Atom

   An atom consists of one or more non-special characters.

4.2.    Number

   A number consists of one or more digit characters, and
   represents a numeric value.

4.3.    String

   A string is in one of two forms: either literal or quoted
   string.  The literal form is the general form of string.  The
   quoted string form is an alternative that avoids the overhead of
   processing a literal at the cost of limitations of characters
   which may be used.

   A literal is a sequence of zero or more octets (including CR and
   LF), prefix-quoted with an octet count in the form of an open
   brace ("{"), the number of octets, close brace ("}"), and CRLF.
   In the case of literals transmitted from server to client, the
   CRLF is immediately followed by the octet data.  In the case of
   literals transmitted from client to server, the client MUST wait
   to receive a command continuation request (described later in
   this document) before sending the octet data (and the remainder
   of the command).

   A quoted string is a sequence of zero or more 7-bit characters,
   excluding CR and LF, with double quote (<">) characters at each
   end.

   The empty string is represented as either "" (a quoted string
   with zero characters between double quotes) or as {0} followed
   by CRLF (a literal with an octet count of 0).

     Note: Even if the octet count is 0, a client transmitting a
     literal MUST wait to receive a command continuation request.

4.3.1.  8-bit and Binary Strings

   8-bit textual and binary mail is supported through the use of a
   [MIME-IMB] content transfer encoding.  IMAP4rev1 implementations MAY
   transmit 8-bit or multi-octet characters in literals, but SHOULD do
   so only when the [CHARSET] is identified.

   Although a BINARY body encoding is defined, unencoded binary strings
   are not permitted.  A "binary string" is any string with NUL
   characters.  Implementations MUST encode binary data into a textual
   form, such as BASE64, before transmitting the data.  A string with an
   excessive amount of CTL characters MAY also be considered to be
   binary.

4.4.    Parenthesized List

   Data structures are represented as a "parenthesized list"; a sequence
   of data items, delimited by space, and bounded at each end by
   parentheses.  A parenthesized list can contain other parenthesized
   lists, using multiple levels of parentheses to indicate nesting.

   The empty list is represented as () -- a parenthesized list with no
   members.

4.5.    NIL

   The special form "NIL" represents the non-existence of a particular
   data item that is represented as a string or parenthesized list, as
   distinct from the empty string "" or the empty parenthesized list ().

        Note: NIL is never used for any data item which takes the
        form of an atom.  For example, a mailbox name of "NIL" is a
        mailbox named NIL as opposed to a non-existent mailbox
        name.  This is because mailbox uses "astring" syntax which
        is an atom or a string.  Conversely, an addr-name of NIL is
        a non-existent personal name, because addr-name uses
        "nstring" syntax which is NIL or a string, but never an
        atom.

5.      Operational Considerations

   The following rules are listed here to ensure that all IMAP4rev1
   implementations interoperate properly.

5.1.    Mailbox Naming

   Mailbox names are 7-bit.  Client implementations MUST NOT attempt to
   create 8-bit mailbox names, and SHOULD interpret any 8-bit mailbox
   names returned by LIST or LSUB as UTF-8.  Server implementations
   SHOULD prohibit the creation of 8-bit mailbox names, and SHOULD NOT
   return 8-bit mailbox names in LIST or LSUB.  See section 5.1.3 for
   more information on how to represent non-ASCII mailbox names.

        Note: 8-bit mailbox names were undefined in earlier
        versions of this protocol.  Some sites used a local 8-bit
        character set to represent non-ASCII mailbox names.  Such
        usage is not interoperable, and is now formally deprecated.

   The case-insensitive mailbox name INBOX is a special name reserved to
   mean "the primary mailbox for this user on this server".  The
   interpretation of all other names is implementation-dependent.

   In particular, this specification takes no position on case
   sensitivity in non-INBOX mailbox names.  Some server implementations
   are fully case-sensitive; others preserve case of a newly-created
   name but otherwise are case-insensitive; and yet others coerce names
   to a particular case.  Client implementations MUST interact with any
   of these.  If a server implementation interprets non-INBOX mailbox
   names as case-insensitive, it MUST treat names using the
   international naming convention specially as described in section
   5.1.3.

   There are certain client considerations when creating a new mailbox
   name:

   1)    Any character which is one of the atom-specials (see the Formal
         Syntax) will require that the mailbox name be represented as a
         quoted string or literal.

   2)    CTL and other non-graphic characters are difficult to represent
         in a user interface and are best avoided.

   3)    Although the list-wildcard characters ("%" and "*") are valid
         in a mailbox name, it is difficult to use such mailbox names
         with the LIST and LSUB commands due to the conflict with
         wildcard interpretation.

   4)    Usually, a character (determined by the server implementation)
         is reserved to delimit levels of hierarchy.

   5)    Two characters, "#" and "&", have meanings by convention, and
         should be avoided except when used in that convention.

5.1.1.  Mailbox Hierarchy Naming

   If it is desired to export hierarchical mailbox names, mailbox names
   MUST be left-to-right hierarchical using a single character to
   separate levels of hierarchy.  The same hierarchy separator character
   is used for all levels of hierarchy within a single name.

5.1.2.  Mailbox Namespace Naming Convention

   By convention, the first hierarchical element of any mailbox name
   which begins with "#" identifies the "namespace" of the remainder of
   the name.  This makes it possible to disambiguate between different
   types of mailbox stores, each of which have their own namespaces.

        For example, implementations which offer access to USENET
        newsgroups MAY use the "#news" namespace to partition the
        USENET newsgroup namespace from that of other mailboxes.
        Thus, the comp.mail.misc newsgroup would have a mailbox
        name of "#news.comp.mail.misc", and the name
        "comp.mail.misc" can refer to a different object (e.g., a
        user's private mailbox).

5.1.3.  Mailbox International Naming Convention

   By convention, international mailbox names in IMAP4rev1 are specified
   using a modified version of the UTF-7 encoding described in [UTF-7].
   Modified UTF-7 may also be usable in servers that implement an
   earlier version of this protocol.

   In modified UTF-7, printable US-ASCII characters, except for "&",
   represent themselves; that is, characters with octet values 0x20-0x25
   and 0x27-0x7e.  The character "&" (0x26) is represented by the
   two-octet sequence "&-".

   All other characters (octet values 0x00-0x1f and 0x7f-0xff) are
   represented in modified BASE64, with a further modification from
   [UTF-7] that "," is used instead of "/".  Modified BASE64 MUST NOT be
   used to represent any printing US-ASCII character which can represent
   itself.

   "&" is used to shift to modified BASE64 and "-" to shift back to
   US-ASCII.  There is no implicit shift from BASE64 to US-ASCII, and
   null shifts ("-&" while in BASE64; note that "&-" while in US-ASCII
   means "&") are not permitted.  However, all names start in US-ASCII,
   and MUST end in US-ASCII; that is, a name that ends with a non-ASCII
   ISO-10646 character MUST end with a "-").

   The purpose of these modifications is to correct the following
   problems with UTF-7:

      1) UTF-7 uses the "+" character for shifting; this conflicts with
         the common use of "+" in mailbox names, in particular USENET
         newsgroup names.

      2) UTF-7's encoding is BASE64 which uses the "/" character; this
         conflicts with the use of "/" as a popular hierarchy delimiter.

      3) UTF-7 prohibits the unencoded usage of "\"; this conflicts with
         the use of "\" as a popular hierarchy delimiter.

      4) UTF-7 prohibits the unencoded usage of "~"; this conflicts with
         the use of "~" in some servers as a home directory indicator.

      5) UTF-7 permits multiple alternate forms to represent the same
         string; in particular, printable US-ASCII characters can be
         represented in encoded form.

      Although modified UTF-7 is a convention, it establishes certain
      requirements on server handling of any mailbox name with an
      embedded "&" character.  In particular, server implementations
      MUST preserve the exact form of the modified BASE64 portion of a
      modified UTF-7 name and treat that text as case-sensitive, even if
      names are otherwise case-insensitive or case-folded.

      Server implementations SHOULD verify that any mailbox name with an
      embedded "&" character, used as an argument to CREATE, is: in the
      correctly modified UTF-7 syntax, has no superfluous shifts, and
      has no encoding in modified BASE64 of any printing US-ASCII
      character which can represent itself.  However, client
      implementations MUST NOT depend upon the server doing this, and
      SHOULD NOT attempt to create a mailbox name with an embedded "&"
      character unless it complies with the modified UTF-7 syntax.

      Server implementations which export a mail store that does not
      follow the modified UTF-7 convention MUST convert to modified
      UTF-7 any mailbox name that contains either non-ASCII characters
      or the "&" character.

           For example, here is a mailbox name which mixes English,
           Chinese, and Japanese text:
           ~peter/mail/&U,BTFw-/&ZeVnLIqe-

           For example, the string "&Jjo!" is not a valid mailbox
           name because it does not contain a shift to US-ASCII
           before the "!".  The correct form is "&Jjo-!".  The
           string "&U,BTFw-&ZeVnLIqe-" is not permitted because it
           contains a superfluous shift.  The correct form is
           "&U,BTF2XlZyyKng-".

5.2.    Mailbox Size and Message Status Updates

   At any time, a server can send data that the client did not request.
   Sometimes, such behavior is REQUIRED.  For example, agents other than
   the server MAY add messages to the mailbox (e.g., new message
   delivery), change the flags of the messages in the mailbox (e.g.,
   simultaneous access to the same mailbox by multiple agents), or even
   remove messages from the mailbox.  A server MUST send mailbox size
   updates automatically if a mailbox size change is observed during the
   processing of a command.  A server SHOULD send message flag updates
   automatically, without requiring the client to request such updates
   explicitly.

   Special rules exist for server notification of a client about the
   removal of messages to prevent synchronization errors; see the
   description of the EXPUNGE response for more detail.  In particular,
   it is NOT permitted to send an EXISTS response that would reduce the
   number of messages in the mailbox; only the EXPUNGE response can do
   this.

   Regardless of what implementation decisions a client makes on
   remembering data from the server, a client implementation MUST record
   mailbox size updates.  It MUST NOT assume that any command after the
   initial mailbox selection will return the size of the mailbox.

5.3.    Response when no Command in Progress

   Server implementations are permitted to send an untagged response
   (except for EXPUNGE) while there is no command in progress.  Server
   implementations that send such responses MUST deal with flow control
   considerations.  Specifically, they MUST either (1) verify that the
   size of the data does not exceed the underlying transport's available
   window size, or (2) use non-blocking writes.

5.4.    Autologout Timer

   If a server has an inactivity autologout timer, the duration of that
   timer MUST be at least 30 minutes.  The receipt of ANY command from
   the client during that interval SHOULD suffice to reset the
   autologout timer.

5.5.    Multiple Commands in Progress

   The client MAY send another command without waiting for the
   completion result response of a command, subject to ambiguity rules
   (see below) and flow control constraints on the underlying data
   stream.  Similarly, a server MAY begin processing another command
   before processing the current command to completion, subject to
   ambiguity rules.  However, any command continuation request responses
   and command continuations MUST be negotiated before any subsequent
   command is initiated.

   The exception is if an ambiguity would result because of a command
   that would affect the results of other commands.  Clients MUST NOT
   send multiple commands without waiting if an ambiguity would result.
   If the server detects a possible ambiguity, it MUST execute commands
   to completion in the order given by the client.

   The most obvious example of ambiguity is when a command would affect
   the results of another command, e.g., a FETCH of a message's flags
   and a STORE of that same message's flags.

   A non-obvious ambiguity occurs with commands that permit an untagged
   EXPUNGE response (commands other than FETCH, STORE, and SEARCH),
   since an untagged EXPUNGE response can invalidate sequence numbers in
   a subsequent command.  This is not a problem for FETCH, STORE, or
   SEARCH commands because servers are prohibited from sending EXPUNGE
   responses while any of those commands are in progress.  Therefore, if
   the client sends any command other than FETCH, STORE, or SEARCH, it
   MUST wait for the completion result response before sending a command
   with message sequence numbers.

        Note: UID FETCH, UID STORE, and UID SEARCH are different
        commands from FETCH, STORE, and SEARCH.  If the client
        sends a UID command, it must wait for a completion result
        response before sending a command with message sequence
        numbers.

   For example, the following non-waiting command sequences are invalid:

      FETCH + NOOP + STORE
      STORE + COPY + FETCH
      COPY + COPY
      CHECK + FETCH

   The following are examples of valid non-waiting command sequences:

      FETCH + STORE + SEARCH + CHECK
      STORE + COPY + EXPUNGE

      UID SEARCH + UID SEARCH may be valid or invalid as a non-waiting
      command sequence, depending upon whether or not the second UID
      SEARCH contains message sequence numbers.

6.      Client Commands

   IMAP4rev1 commands are described in this section.  Commands are
   organized by the state in which the command is permitted.  Commands
   which are permitted in multiple states are listed in the minimum
   permitted state (for example, commands valid in authenticated and
   selected state are listed in the authenticated state commands).

   Command arguments, identified by "Arguments:" in the command
   descriptions below, are described by function, not by syntax.  The
   precise syntax of command arguments is described in the Formal Syntax
   section.

   Some commands cause specific server responses to be returned; these
   are identified by "Responses:" in the command descriptions below.
   See the response descriptions in the Responses section for
   information on these responses, and the Formal Syntax section for the
   precise syntax of these responses.  It is possible for server data to
   be transmitted as a result of any command.  Thus, commands that do
   not specifically require server data specify "no specific responses
   for this command" instead of "none".

   The "Result:" in the command description refers to the possible
   tagged status responses to a command, and any special interpretation
   of these status responses.

   The state of a connection is only changed by successful commands
   which are documented as changing state.  A rejected command (BAD
   response) never changes the state of the connection or of the
   selected mailbox.  A failed command (NO response) generally does not
   change the state of the connection or of the selected mailbox; the
   exception being the SELECT and EXAMINE commands.

6.1.    Client Commands - Any State

   The following commands are valid in any state: CAPABILITY, NOOP, and
   LOGOUT.

6.1.1.  CAPABILITY Command

   Arguments:  none

   Responses:  REQUIRED untagged response: CAPABILITY

   Result:     OK - capability completed
               BAD - command unknown or arguments invalid

      The CAPABILITY command requests a listing of capabilities that the
      server supports.  The server MUST send a single untagged
      CAPABILITY response with "IMAP4rev1" as one of the listed
      capabilities before the (tagged) OK response.

      A capability name which begins with "AUTH=" indicates that the
      server supports that particular authentication mechanism.  All
      such names are, by definition, part of this specification.  For
      example, the authorization capability for an experimental
      "blurdybloop" authenticator would be "AUTH=XBLURDYBLOOP" and not
      "XAUTH=BLURDYBLOOP" or "XAUTH=XBLURDYBLOOP".

      Other capability names refer to extensions, revisions, or
      amendments to this specification.  See the documentation of the
      CAPABILITY response for additional information.  No capabilities,
      beyond the base IMAP4rev1 set defined in this specification, are
      enabled without explicit client action to invoke the capability.

      Client and server implementations MUST implement the STARTTLS,
      LOGINDISABLED, and AUTH=PLAIN (described in [IMAP-TLS])
      capabilities.  See the Security Considerations section for
      important information.

      See the section entitled "Client Commands -
      Experimental/Expansion" for information about the form of site or
      implementation-specific capabilities.

   Example:    C: abcd CAPABILITY
               S: * CAPABILITY IMAP4rev1 STARTTLS AUTH=GSSAPI
               LOGINDISABLED
               S: abcd OK CAPABILITY completed
               C: efgh STARTTLS
               S: efgh OK STARTLS completed
               <TLS negotiation, further commands are under [TLS] layer>
               C: ijkl CAPABILITY
               S: * CAPABILITY IMAP4rev1 AUTH=GSSAPI AUTH=PLAIN
               S: ijkl OK CAPABILITY completed

6.1.2.  NOOP Command

   Arguments:  none

   Responses:  no specific responses for this command (but see below)

   Result:     OK - noop completed
               BAD - command unknown or arguments invalid

      The NOOP command always succeeds.  It does nothing.

      Since any command can return a status update as untagged data, the
      NOOP command can be used as a periodic poll for new messages or
      message status updates during a period of inactivity (this is the
      preferred method to do this).  The NOOP command can also be used
      to reset any inactivity autologout timer on the server.

   Example:    C: a002 NOOP
               S: a002 OK NOOP completed
                  . . .
               C: a047 NOOP
               S: * 22 EXPUNGE
               S: * 23 EXISTS
               S: * 3 RECENT
               S: * 14 FETCH (FLAGS (\Seen \Deleted))
               S: a047 OK NOOP completed

6.1.3.  LOGOUT Command

   Arguments:  none

   Responses:  REQUIRED untagged response: BYE

   Result:     OK - logout completed
               BAD - command unknown or arguments invalid

      The LOGOUT command informs the server that the client is done with
      the connection.  The server MUST send a BYE untagged response
      before the (tagged) OK response, and then close the network
      connection.

   Example:    C: A023 LOGOUT
               S: * BYE IMAP4rev1 Server logging out
               S: A023 OK LOGOUT completed
               (Server and client then close the connection)

6.2.    Client Commands - Not Authenticated State

   In the not authenticated state, the AUTHENTICATE or LOGIN command
   establishes authentication and enters the authenticated state.  The
   AUTHENTICATE command provides a general mechanism for a variety of
   authentication techniques, privacy protection, and integrity
   checking; whereas the LOGIN command uses a traditional user name and
   plaintext password pair and has no means of establishing privacy
   protection or integrity checking.

   The STARTTLS command is an alternate form of establishing session
   privacy protection and integrity checking, but does not establish
   authentication or enter the authenticated state.

   Server implementations MAY allow access to certain mailboxes without
   establishing authentication.  This can be done by means of the
   ANONYMOUS [SASL] authenticator described in [ANONYMOUS].  An older
   convention is a LOGIN command using the userid "anonymous"; in this
   case, a password is required although the server may choose to accept
   any password.  The restrictions placed on anonymous users are
   implementation-dependent.

   Once authenticated (including as anonymous), it is not possible to
   re-enter not authenticated state.

   In addition to the universal commands (CAPABILITY, NOOP, and LOGOUT),
   the following commands are valid in the not authenticated state:
   STARTTLS, AUTHENTICATE and LOGIN.  See the Security Considerations
   section for important information about these commands.

6.2.1.  STARTTLS Command

   Arguments:  none

   Responses:  no specific response for this command

   Result:     OK - starttls completed, begin TLS negotiation
               BAD - command unknown or arguments invalid

      A [TLS] negotiation begins immediately after the CRLF at the end
      of the tagged OK response from the server.  Once a client issues a
      STARTTLS command, it MUST NOT issue further commands until a
      server response is seen and the [TLS] negotiation is complete.

      The server remains in the non-authenticated state, even if client
      credentials are supplied during the [TLS] negotiation.  This does
      not preclude an authentication mechanism such as EXTERNAL (defined
      in [SASL]) from using client identity determined by the [TLS]
      negotiation.

      Once [TLS] has been started, the client MUST discard cached
      information about server capabilities and SHOULD re-issue the
      CAPABILITY command.  This is necessary to protect against man-in-
      the-middle attacks which alter the capabilities list prior to
      STARTTLS.  The server MAY advertise different capabilities after
      STARTTLS.

   Example:    C: a001 CAPABILITY
               S: * CAPABILITY IMAP4rev1 STARTTLS LOGINDISABLED
               S: a001 OK CAPABILITY completed
               C: a002 STARTTLS
               S: a002 OK Begin TLS negotiation now
               <TLS negotiation, further commands are under [TLS] layer>
               C: a003 CAPABILITY
               S: * CAPABILITY IMAP4rev1 AUTH=PLAIN
               S: a003 OK CAPABILITY completed
               C: a004 LOGIN joe password
               S: a004 OK LOGIN completed

6.2.2.  AUTHENTICATE Command

   Arguments:  authentication mechanism name

   Responses:  continuation data can be requested

   Result:     OK - authenticate completed, now in authenticated state
               NO - authenticate failure: unsupported authentication
                    mechanism, credentials rejected
               BAD - command unknown or arguments invalid,
                    authentication exchange cancelled

      The AUTHENTICATE command indicates a [SASL] authentication
      mechanism to the server.  If the server supports the requested
      authentication mechanism, it performs an authentication protocol
      exchange to authenticate and identify the client.  It MAY also
      negotiate an OPTIONAL security layer for subsequent protocol
      interactions.  If the requested authentication mechanism is not
      supported, the server SHOULD reject the AUTHENTICATE command by
      sending a tagged NO response.

      The AUTHENTICATE command does not support the optional "initial
      response" feature of [SASL].  Section 5.1 of [SASL] specifies how
      to handle an authentication mechanism which uses an initial
      response.

      The service name specified by this protocol's profile of [SASL] is
      "imap".

      The authentication protocol exchange consists of a series of
      server challenges and client responses that are specific to the
      authentication mechanism.  A server challenge consists of a
      command continuation request response with the "+" token followed
      by a BASE64 encoded string.  The client response consists of a
      single line consisting of a BASE64 encoded string.  If the client
      wishes to cancel an authentication exchange, it issues a line
      consisting of a single "*".  If the server receives such a
      response, it MUST reject the AUTHENTICATE command by sending a
      tagged BAD response.

      If a security layer is negotiated through the [SASL]
      authentication exchange, it takes effect immediately following the
      CRLF that concludes the authentication exchange for the client,
      and the CRLF of the tagged OK response for the server.

      While client and server implementations MUST implement the
      AUTHENTICATE command itself, it is not required to implement any
      authentication mechanisms other than the PLAIN mechanism described

      in [IMAP-TLS].  Also, an authentication mechanism is not required
      to support any security layers.

           Note: a server implementation MUST implement a
           configuration in which it does NOT permit any plaintext
           password mechanisms, unless either the STARTTLS command
           has been negotiated or some other mechanism that
           protects the session from password snooping has been
           provided.  Server sites SHOULD NOT use any configuration
           which permits a plaintext password mechanism without
           such a protection mechanism against password snooping.
           Client and server implementations SHOULD implement
           additional [SASL] mechanisms that do not use plaintext
           passwords, such the GSSAPI mechanism described in [SASL]
           and/or the [DIGEST-MD5] mechanism.

      Servers and clients can support multiple authentication
      mechanisms.  The server SHOULD list its supported authentication
      mechanisms in the response to the CAPABILITY command so that the
      client knows which authentication mechanisms to use.

      A server MAY include a CAPABILITY response code in the tagged OK
      response of a successful AUTHENTICATE command in order to send
      capabilities automatically.  It is unnecessary for a client to
      send a separate CAPABILITY command if it recognizes these
      automatic capabilities.  This should only be done if a security
      layer was not negotiated by the AUTHENTICATE command, because the
      tagged OK response as part of an AUTHENTICATE command is not
      protected by encryption/integrity checking.  [SASL] requires the
      client to re-issue a CAPABILITY command in this case.

      If an AUTHENTICATE command fails with a NO response, the client
      MAY try another authentication mechanism by issuing another
      AUTHENTICATE command.  It MAY also attempt to authenticate by
      using the LOGIN command (see section 6.2.3 for more detail).  In
      other words, the client MAY request authentication types in
      decreasing order of preference, with the LOGIN command as a last
      resort.

      The authorization identity passed from the client to the server
      during the authentication exchange is interpreted by the server as
      the user name whose privileges the client is requesting.

   Example:    S: * OK IMAP4rev1 Server
               C: A001 AUTHENTICATE GSSAPI
               S: +
               C: YIIB+wYJKoZIhvcSAQICAQBuggHqMIIB5qADAgEFoQMCAQ6iBw
                  MFACAAAACjggEmYYIBIjCCAR6gAwIBBaESGxB1Lndhc2hpbmd0
                  b24uZWR1oi0wK6ADAgEDoSQwIhsEaW1hcBsac2hpdmFtcy5jYW
                  Mud2FzaGluZ3Rvbi5lZHWjgdMwgdCgAwIBAaEDAgEDooHDBIHA
                  cS1GSa5b+fXnPZNmXB9SjL8Ollj2SKyb+3S0iXMljen/jNkpJX
                  AleKTz6BQPzj8duz8EtoOuNfKgweViyn/9B9bccy1uuAE2HI0y
                  C/PHXNNU9ZrBziJ8Lm0tTNc98kUpjXnHZhsMcz5Mx2GR6dGknb
                  I0iaGcRerMUsWOuBmKKKRmVMMdR9T3EZdpqsBd7jZCNMWotjhi
                  vd5zovQlFqQ2Wjc2+y46vKP/iXxWIuQJuDiisyXF0Y8+5GTpAL
                  pHDc1/pIGmMIGjoAMCAQGigZsEgZg2on5mSuxoDHEA1w9bcW9n
                  FdFxDKpdrQhVGVRDIzcCMCTzvUboqb5KjY1NJKJsfjRQiBYBdE
                  NKfzK+g5DlV8nrw81uOcP8NOQCLR5XkoMHC0Dr/80ziQzbNqhx
                  O6652Npft0LQwJvenwDI13YxpwOdMXzkWZN/XrEqOWp6GCgXTB
                  vCyLWLlWnbaUkZdEYbKHBPjd8t/1x5Yg==
               S: + YGgGCSqGSIb3EgECAgIAb1kwV6ADAgEFoQMCAQ+iSzBJoAMC
                  AQGiQgRAtHTEuOP2BXb9sBYFR4SJlDZxmg39IxmRBOhXRKdDA0
                  uHTCOT9Bq3OsUTXUlk0CsFLoa8j+gvGDlgHuqzWHPSQg==
               C:
               S: + YDMGCSqGSIb3EgECAgIBAAD/////6jcyG4GE3KkTzBeBiVHe
                  ceP2CWY0SR0fAQAgAAQEBAQ=
               C: YDMGCSqGSIb3EgECAgIBAAD/////3LQBHXTpFfZgrejpLlLImP
                  wkhbfa2QteAQAgAG1yYwE=
               S: A001 OK GSSAPI authentication successful

        Note: The line breaks within server challenges and client
        responses are for editorial clarity and are not in real
        authenticators.

6.2.3.  LOGIN Command

   Arguments:  user name
               password

   Responses:  no specific responses for this command

   Result:     OK - login completed, now in authenticated state
               NO - login failure: user name or password rejected
               BAD - command unknown or arguments invalid

      The LOGIN command identifies the client to the server and carries
      the plaintext password authenticating this user.

      A server MAY include a CAPABILITY response code in the tagged OK
      response to a successful LOGIN command in order to send
      capabilities automatically.  It is unnecessary for a client to
      send a separate CAPABILITY command if it recognizes these
      automatic capabilities.

   Example:    C: a001 LOGIN SMITH SESAME
               S: a001 OK LOGIN completed

        Note: Use of the LOGIN command over an insecure network
        (such as the Internet) is a security risk, because anyone
        monitoring network traffic can obtain plaintext passwords.
        The LOGIN command SHOULD NOT be used except as a last
        resort, and it is recommended that client implementations
        have a means to disable any automatic use of the LOGIN
        command.

        Unless either the STARTTLS command has been negotiated or
        some other mechanism that protects the session from
        password snooping has been provided, a server
        implementation MUST implement a configuration in which it
        advertises the LOGINDISABLED capability and does NOT permit
        the LOGIN command.  Server sites SHOULD NOT use any
        configuration which permits the LOGIN command without such
        a protection mechanism against password snooping.  A client
        implementation MUST NOT send a LOGIN command if the
        LOGINDISABLED capability is advertised.

6.3.    Client Commands - Authenticated State

   In the authenticated state, commands that manipulate mailboxes as
   atomic entities are permitted.  Of these commands, the SELECT and
   EXAMINE commands will select a mailbox for access and enter the
   selected state.

   In addition to the universal commands (CAPABILITY, NOOP, and LOGOUT),
   the following commands are valid in the authenticated state: SELECT,
   EXAMINE, CREATE, DELETE, RENAME, SUBSCRIBE, UNSUBSCRIBE, LIST, LSUB,
   STATUS, and APPEND.

6.3.1.  SELECT Command

   Arguments:  mailbox name

   Responses:  REQUIRED untagged responses: FLAGS, EXISTS, RECENT
               REQUIRED OK untagged responses:  UNSEEN,  PERMANENTFLAGS,
               UIDNEXT, UIDVALIDITY

   Result:     OK - select completed, now in selected state
               NO - select failure, now in authenticated state: no
                    such mailbox, can't access mailbox
               BAD - command unknown or arguments invalid

      The SELECT command selects a mailbox so that messages in the
      mailbox can be accessed.  Before returning an OK to the client,
      the server MUST send the following untagged data to the client.
      Note that earlier versions of this protocol only required the
      FLAGS, EXISTS, and RECENT untagged data; consequently, client
      implementations SHOULD implement default behavior for missing data
      as discussed with the individual item.

         FLAGS       Defined flags in the mailbox.  See the description
                     of the FLAGS response for more detail.

         <n> EXISTS  The number of messages in the mailbox.  See the
                     description of the EXISTS response for more detail.

         <n> RECENT  The number of messages with the \Recent flag set.
                     See the description of the RECENT response for more
                     detail.

         OK [UNSEEN <n>]
                     The message sequence number of the first unseen
                     message in the mailbox.  If this is missing, the
                     client can not make any assumptions about the first
                     unseen message in the mailbox, and needs to issue a
                     SEARCH command if it wants to find it.

         OK [PERMANENTFLAGS (<list of flags>)]
                     A list of message flags that the client can change
                     permanently.  If this is missing, the client should
                     assume that all flags can be changed permanently.

         OK [UIDNEXT <n>]
                     The next unique identifier value.  Refer to section
                     2.3.1.1 for more information.  If this is missing,
                     the client can not make any assumptions about the
                     next unique identifier value.

         OK [UIDVALIDITY <n>]
                     The unique identifier validity value.  Refer to
                     section 2.3.1.1 for more information.  If this is
                     missing, the server does not support unique
                     identifiers.

      Only one mailbox can be selected at a time in a connection;
      simultaneous access to multiple mailboxes requires multiple
      connections.  The SELECT command automatically deselects any
      currently selected mailbox before attempting the new selection.
      Consequently, if a mailbox is selected and a SELECT command that
      fails is attempted, no mailbox is selected.

      If the client is permitted to modify the mailbox, the server
      SHOULD prefix the text of the tagged OK response with the
      "[READ-WRITE]" response code.

      If the client is not permitted to modify the mailbox but is
      permitted read access, the mailbox is selected as read-only, and
      the server MUST prefix the text of the tagged OK response to
      SELECT with the "[READ-ONLY]" response code.  Read-only access
      through SELECT differs from the EXAMINE command in that certain
      read-only mailboxes MAY permit the change of permanent state on a
      per-user (as opposed to global) basis.  Netnews messages marked in
      a server-based .newsrc file are an example of such per-user
      permanent state that can be modified with read-only mailboxes.

   Example:    C: A142 SELECT INBOX
               S: * 172 EXISTS
               S: * 1 RECENT
               S: * OK [UNSEEN 12] Message 12 is first unseen
               S: * OK [UIDVALIDITY 3857529045] UIDs valid
               S: * OK [UIDNEXT 4392] Predicted next UID
               S: * FLAGS (\Answered \Flagged \Deleted \Seen \Draft)
               S: * OK [PERMANENTFLAGS (\Deleted \Seen \*)] Limited
               S: A142 OK [READ-WRITE] SELECT completed

6.3.2.  EXAMINE Command

   Arguments:  mailbox name

   Responses:  REQUIRED untagged responses: FLAGS, EXISTS, RECENT
               REQUIRED OK untagged responses:  UNSEEN,  PERMANENTFLAGS,
               UIDNEXT, UIDVALIDITY

   Result:     OK - examine completed, now in selected state
               NO - examine failure, now in authenticated state: no
                    such mailbox, can't access mailbox
               BAD - command unknown or arguments invalid

      The EXAMINE command is identical to SELECT and returns the same
      output; however, the selected mailbox is identified as read-only.
      No changes to the permanent state of the mailbox, including
      per-user state, are permitted; in particular, EXAMINE MUST NOT
      cause messages to lose the \Recent flag.

      The text of the tagged OK response to the EXAMINE command MUST
      begin with the "[READ-ONLY]" response code.

   Example:    C: A932 EXAMINE blurdybloop
               S: * 17 EXISTS
               S: * 2 RECENT
               S: * OK [UNSEEN 8] Message 8 is first unseen
               S: * OK [UIDVALIDITY 3857529045] UIDs valid
               S: * OK [UIDNEXT 4392] Predicted next UID
               S: * FLAGS (\Answered \Flagged \Deleted \Seen \Draft)
               S: * OK [PERMANENTFLAGS ()] No permanent flags permitted
               S: A932 OK [READ-ONLY] EXAMINE completed

6.3.3.  CREATE Command

   Arguments:  mailbox name

   Responses:  no specific responses for this command

   Result:     OK - create completed
               NO - create failure: can't create mailbox with that name
               BAD - command unknown or arguments invalid

      The CREATE command creates a mailbox with the given name.  An OK
      response is returned only if a new mailbox with that name has been
      created.  It is an error to attempt to create INBOX or a mailbox
      with a name that refers to an extant mailbox.  Any error in
      creation will return a tagged NO response.

      If the mailbox name is suffixed with the server's hierarchy
      separator character (as returned from the server by a LIST
      command), this is a declaration that the client intends to create
      mailbox names under this name in the hierarchy.  Server
      implementations that do not require this declaration MUST ignore
      the declaration.  In any case, the name created is without the
      trailing hierarchy delimiter.

      If the server's hierarchy separator character appears elsewhere in
      the name, the server SHOULD create any superior hierarchical names
      that are needed for the CREATE command to be successfully
      completed.  In other words, an attempt to create "foo/bar/zap" on
      a server in which "/" is the hierarchy separator character SHOULD
      create foo/ and foo/bar/ if they do not already exist.

      If a new mailbox is created with the same name as a mailbox which
      was deleted, its unique identifiers MUST be greater than any
      unique identifiers used in the previous incarnation of the mailbox
      UNLESS the new incarnation has a different unique identifier
      validity value.  See the description of the UID command for more
      detail.

   Example:    C: A003 CREATE owatagusiam/
               S: A003 OK CREATE completed
               C: A004 CREATE owatagusiam/blurdybloop
               S: A004 OK CREATE completed

        Note: The interpretation of this example depends on whether
        "/" was returned as the hierarchy separator from LIST.  If
        "/" is the hierarchy separator, a new level of hierarchy
        named "owatagusiam" with a member called "blurdybloop" is
        created.  Otherwise, two mailboxes at the same hierarchy
        level are created.

6.3.4.  DELETE Command

   Arguments:  mailbox name

   Responses:  no specific responses for this command

   Result:     OK - delete completed
               NO - delete failure: can't delete mailbox with that name
               BAD - command unknown or arguments invalid

      The DELETE command permanently removes the mailbox with the given
      name.  A tagged OK response is returned only if the mailbox has
      been deleted.  It is an error to attempt to delete INBOX or a
      mailbox name that does not exist.

      The DELETE command MUST NOT remove inferior hierarchical names.
      For example, if a mailbox "foo" has an inferior "foo.bar"
      (assuming "." is the hierarchy delimiter character), removing
      "foo" MUST NOT remove "foo.bar".  It is an error to attempt to
      delete a name that has inferior hierarchical names and also has
      the \Noselect mailbox name attribute (see the description of the
      LIST response for more details).

      It is permitted to delete a name that has inferior hierarchical
      names and does not have the \Noselect mailbox name attribute.  In
      this case, all messages in that mailbox are removed, and the name
      will acquire the \Noselect mailbox name attribute.

      The value of the highest-used unique identifier of the deleted
      mailbox MUST be preserved so that a new mailbox created with the
      same name will not reuse the identifiers of the former
      incarnation, UNLESS the new incarnation has a different unique
      identifier validity value.  See the description of the UID command
      for more detail.

   Examples:   C: A682 LIST "" *
               S: * LIST () "/" blurdybloop
               S: * LIST (\Noselect) "/" foo
               S: * LIST () "/" foo/bar
               S: A682 OK LIST completed
               C: A683 DELETE blurdybloop
               S: A683 OK DELETE completed
               C: A684 DELETE foo
               S: A684 NO Name "foo" has inferior hierarchical names
               C: A685 DELETE foo/bar
               S: A685 OK DELETE Completed
               C: A686 LIST "" *
               S: * LIST (\Noselect) "/" foo
               S: A686 OK LIST completed
               C: A687 DELETE foo
               S: A687 OK DELETE Completed

               C: A82 LIST "" *
               S: * LIST () "." blurdybloop
               S: * LIST () "." foo
               S: * LIST () "." foo.bar
               S: A82 OK LIST completed
               C: A83 DELETE blurdybloop
               S: A83 OK DELETE completed
               C: A84 DELETE foo
               S: A84 OK DELETE Completed
               C: A85 LIST "" *
               S: * LIST () "." foo.bar
               S: A85 OK LIST completed
               C: A86 LIST "" %
               S: * LIST (\Noselect) "." foo
               S: A86 OK LIST completed

6.3.5.  RENAME Command

   Arguments:  existing mailbox name
               new mailbox name

   Responses:  no specific responses for this command

   Result:     OK - rename completed
               NO - rename failure: can't rename mailbox with that name,
                    can't rename to mailbox with that name
               BAD - command unknown or arguments invalid

      The RENAME command changes the name of a mailbox.  A tagged OK
      response is returned only if the mailbox has been renamed.  It is
      an error to attempt to rename from a mailbox name that does not
      exist or to a mailbox name that already exists.  Any error in
      renaming will return a tagged NO response.

      If the name has inferior hierarchical names, then the inferior
      hierarchical names MUST also be renamed.  For example, a rename of
      "foo" to "zap" will rename "foo/bar" (assuming "/" is the
      hierarchy delimiter character) to "zap/bar".

      If the server's hierarchy separator character appears in the name,
      the server SHOULD create any superior hierarchical names that are
      needed for the RENAME command to complete successfully.  In other
      words, an attempt to rename "foo/bar/zap" to baz/rag/zowie on a
      server in which "/" is the hierarchy separator character SHOULD
      create baz/ and baz/rag/ if they do not already exist.

      The value of the highest-used unique identifier of the old mailbox
      name MUST be preserved so that a new mailbox created with the same
      name will not reuse the identifiers of the former incarnation,
      UNLESS the new incarnation has a different unique identifier
      validity value.  See the description of the UID command for more
      detail.

      Renaming INBOX is permitted, and has special behavior.  It moves
      all messages in INBOX to a new mailbox with the given name,
      leaving INBOX empty.  If the server implementation supports
      inferior hierarchical names of INBOX, these are unaffected by a
      rename of INBOX.

   Examples:   C: A682 LIST "" *
               S: * LIST () "/" blurdybloop
               S: * LIST (\Noselect) "/" foo
               S: * LIST () "/" foo/bar
               S: A682 OK LIST completed
               C: A683 RENAME blurdybloop sarasoop
               S: A683 OK RENAME completed
               C: A684 RENAME foo zowie
               S: A684 OK RENAME Completed
               C: A685 LIST "" *
               S: * LIST () "/" sarasoop
               S: * LIST (\Noselect) "/" zowie
               S: * LIST () "/" zowie/bar
               S: A685 OK LIST completed

               C: Z432 LIST "" *
               S: * LIST () "." INBOX
               S: * LIST () "." INBOX.bar
               S: Z432 OK LIST completed
               C: Z433 RENAME INBOX old-mail
               S: Z433 OK RENAME completed
               C: Z434 LIST "" *
               S: * LIST () "." INBOX
               S: * LIST () "." INBOX.bar
               S: * LIST () "." old-mail
               S: Z434 OK LIST completed

6.3.6.  SUBSCRIBE Command

   Arguments:  mailbox

   Responses:  no specific responses for this command

   Result:     OK - subscribe completed
               NO - subscribe failure: can't subscribe to that name
               BAD - command unknown or arguments invalid

      The SUBSCRIBE command adds the specified mailbox name to the
      server's set of "active" or "subscribed" mailboxes as returned by
      the LSUB command.  This command returns a tagged OK response only
      if the subscription is successful.

      A server MAY validate the mailbox argument to SUBSCRIBE to verify
      that it exists.  However, it MUST NOT unilaterally remove an
      existing mailbox name from the subscription list even if a mailbox
      by that name no longer exists.

           Note: This requirement is because a server site can
           choose to routinely remove a mailbox with a well-known
           name (e.g., "system-alerts") after its contents expire,
           with the intention of recreating it when new contents
           are appropriate.

   Example:    C: A002 SUBSCRIBE #news.comp.mail.mime
               S: A002 OK SUBSCRIBE completed

6.3.7.  UNSUBSCRIBE Command

   Arguments:  mailbox name

   Responses:  no specific responses for this command

   Result:     OK - unsubscribe completed
               NO - unsubscribe failure: can't unsubscribe that name
               BAD - command unknown or arguments invalid

      The UNSUBSCRIBE command removes the specified mailbox name from
      the server's set of "active" or "subscribed" mailboxes as returned
      by the LSUB command.  This command returns a tagged OK response
      only if the unsubscription is successful.

   Example:    C: A002 UNSUBSCRIBE #news.comp.mail.mime
               S: A002 OK UNSUBSCRIBE completed

6.3.8.  LIST Command

   Arguments:  reference name
               mailbox name with possible wildcards

   Responses:  untagged responses: LIST

   Result:     OK - list completed
               NO - list failure: can't list that reference or name
               BAD - command unknown or arguments invalid

      The LIST command returns a subset of names from the complete set
      of all names available to the client.  Zero or more untagged LIST
      replies are returned, containing the name attributes, hierarchy
      delimiter, and name; see the description of the LIST reply for
      more detail.

      The LIST command SHOULD return its data quickly, without undue
      delay.  For example, it SHOULD NOT go to excess trouble to
      calculate the \Marked or \Unmarked status or perform other
      processing; if each name requires 1 second of processing, then a
      list of 1200 names would take 20 minutes!

      An empty ("" string) reference name argument indicates that the
      mailbox name is interpreted as by SELECT.  The returned mailbox
      names MUST match the supplied mailbox name pattern.  A non-empty
      reference name argument is the name of a mailbox or a level of
      mailbox hierarchy, and indicates the context in which the mailbox
      name is interpreted.

      An empty ("" string) mailbox name argument is a special request to
      return the hierarchy delimiter and the root name of the name given
      in the reference.  The value returned as the root MAY be the empty
      string if the reference is non-rooted or is an empty string.  In
      all cases, a hierarchy delimiter (or NIL if there is no hierarchy)
      is returned.  This permits a client to get the hierarchy delimiter
      (or find out that the mailbox names are flat) even when no
      mailboxes by that name currently exist.

      The reference and mailbox name arguments are interpreted into a
      canonical form that represents an unambiguous left-to-right
      hierarchy.  The returned mailbox names will be in the interpreted
      form.

           Note: The interpretation of the reference argument is
           implementation-defined.  It depends upon whether the
           server implementation has a concept of the "current
           working directory" and leading "break out characters",
           which override the current working directory.

           For example, on a server which exports a UNIX or NT
           filesystem, the reference argument contains the current
           working directory, and the mailbox name argument would
           contain the name as interpreted in the current working
           directory.

           If a server implementation has no concept of break out
           characters, the canonical form is normally the reference
           name appended with the mailbox name.  Note that if the
           server implements the namespace convention (section
           5.1.2), "#" is a break out character and must be treated
           as such.

           If the reference argument is not a level of mailbox
           hierarchy (that is, it is a \NoInferiors name), and/or
           the reference argument does not end with the hierarchy
           delimiter, it is implementation-dependent how this is
           interpreted.  For example, a reference of "foo/bar" and
           mailbox name of "rag/baz" could be interpreted as
           "foo/bar/rag/baz", "foo/barrag/baz", or "foo/rag/baz".
           A client SHOULD NOT use such a reference argument except
           at the explicit request of the user.  A hierarchical
           browser MUST NOT make any assumptions about server
           interpretation of the reference unless the reference is
           a level of mailbox hierarchy AND ends with the hierarchy
           delimiter.

      Any part of the reference argument that is included in the
      interpreted form SHOULD prefix the interpreted form.  It SHOULD
      also be in the same form as the reference name argument.  This
      rule permits the client to determine if the returned mailbox name
      is in the context of the reference argument, or if something about
      the mailbox argument overrode the reference argument.  Without
      this rule, the client would have to have knowledge of the server's
      naming semantics including what characters are "breakouts" that
      override a naming context.

           For example, here are some examples of how references
           and mailbox names might be interpreted on a UNIX-based
           server:

               Reference     Mailbox Name  Interpretation
               ------------  ------------  --------------
               ~smith/Mail/  foo.*         ~smith/Mail/foo.*
               archive/      %             archive/%
               #news.        comp.mail.*   #news.comp.mail.*
               ~smith/Mail/  /usr/doc/foo  /usr/doc/foo
               archive/      ~fred/Mail/*  ~fred/Mail/*

           The first three examples demonstrate interpretations in
           the context of the reference argument.  Note that
           "~smith/Mail" SHOULD NOT be transformed into something
           like "/u2/users/smith/Mail", or it would be impossible
           for the client to determine that the interpretation was
           in the context of the reference.

      The character "*" is a wildcard, and matches zero or more
      characters at this position.  The character "%" is similar to "*",
      but it does not match a hierarchy delimiter.  If the "%" wildcard
      is the last character of a mailbox name argument, matching levels
      of hierarchy are also returned.  If these levels of hierarchy are
      not also selectable mailboxes, they are returned with the
      \Noselect mailbox name attribute (see the description of the LIST
      response for more details).

      Server implementations are permitted to "hide" otherwise
      accessible mailboxes from the wildcard characters, by preventing
      certain characters or names from matching a wildcard in certain
      situations.  For example, a UNIX-based server might restrict the
      interpretation of "*" so that an initial "/" character does not
      match.

      The special name INBOX is included in the output from LIST, if
      INBOX is supported by this server for this user and if the
      uppercase string "INBOX" matches the interpreted reference and
      mailbox name arguments with wildcards as described above.  The
      criteria for omitting INBOX is whether SELECT INBOX will return
      failure; it is not relevant whether the user's real INBOX resides
      on this or some other server.

   Example:    C: A101 LIST "" ""
               S: * LIST (\Noselect) "/" ""
               S: A101 OK LIST Completed
               C: A102 LIST #news.comp.mail.misc ""
               S: * LIST (\Noselect) "." #news.
               S: A102 OK LIST Completed
               C: A103 LIST /usr/staff/jones ""
               S: * LIST (\Noselect) "/" /
               S: A103 OK LIST Completed
               C: A202 LIST ~/Mail/ %
               S: * LIST (\Noselect) "/" ~/Mail/foo
               S: * LIST () "/" ~/Mail/meetings
               S: A202 OK LIST completed

6.3.9.  LSUB Command

   Arguments:  reference name
               mailbox name with possible wildcards

   Responses:  untagged responses: LSUB

   Result:     OK - lsub completed
               NO - lsub failure: can't list that reference or name
               BAD - command unknown or arguments invalid

      The LSUB command returns a subset of names from the set of names
      that the user has declared as being "active" or "subscribed".
      Zero or more untagged LSUB replies are returned.  The arguments to
      LSUB are in the same form as those for LIST.

      The returned untagged LSUB response MAY contain different mailbox
      flags from a LIST untagged response.  If this should happen, the
      flags in the untagged LIST are considered more authoritative.

      A special situation occurs when using LSUB with the % wildcard.
      Consider what happens if "foo/bar" (with a hierarchy delimiter of
      "/") is subscribed but "foo" is not.  A "%" wildcard to LSUB must
      return foo, not foo/bar, in the LSUB response, and it MUST be
      flagged with the \Noselect attribute.

      The server MUST NOT unilaterally remove an existing mailbox name
      from the subscription list even if a mailbox by that name no
      longer exists.

   Example:    C: A002 LSUB "#news." "comp.mail.*"
               S: * LSUB () "." #news.comp.mail.mime
               S: * LSUB () "." #news.comp.mail.misc
               S: A002 OK LSUB completed
               C: A003 LSUB "#news." "comp.%"
               S: * LSUB (\NoSelect) "." #news.comp.mail
               S: A003 OK LSUB completed

6.3.10. STATUS Command

   Arguments:  mailbox name
               status data item names

   Responses:  untagged responses: STATUS

   Result:     OK - status completed
               NO - status failure: no status for that name
               BAD - command unknown or arguments invalid

      The STATUS command requests the status of the indicated mailbox.
      It does not change the currently selected mailbox, nor does it
      affect the state of any messages in the queried mailbox (in
      particular, STATUS MUST NOT cause messages to lose the \Recent
      flag).

      The STATUS command provides an alternative to opening a second
      IMAP4rev1 connection and doing an EXAMINE command on a mailbox to
      query that mailbox's status without deselecting the current
      mailbox in the first IMAP4rev1 connection.

      Unlike the LIST command, the STATUS command is not guaranteed to
      be fast in its response.  Under certain circumstances, it can be
      quite slow.  In some implementations, the server is obliged to
      open the mailbox read-only internally to obtain certain status
      information.  Also unlike the LIST command, the STATUS command
      does not accept wildcards.

           Note: The STATUS command is intended to access the
           status of mailboxes other than the currently selected
           mailbox.  Because the STATUS command can cause the
           mailbox to be opened internally, and because this
           information is available by other means on the selected
           mailbox, the STATUS command SHOULD NOT be used on the
           currently selected mailbox.

           The STATUS command MUST NOT be used as a "check for new
           messages in the selected mailbox" operation (refer to
           sections 7, 7.3.1, and 7.3.2 for more information about
           the proper method for new message checking).

           Because the STATUS command is not guaranteed to be fast
           in its results, clients SHOULD NOT expect to be able to
           issue many consecutive STATUS commands and obtain
           reasonable performance.

      The currently defined status data items that can be requested are:

      MESSAGES
         The number of messages in the mailbox.

      RECENT
         The number of messages with the \Recent flag set.

      UIDNEXT
         The next unique identifier value of the mailbox.  Refer to
         section 2.3.1.1 for more information.

      UIDVALIDITY
         The unique identifier validity value of the mailbox.  Refer to
         section 2.3.1.1 for more information.

      UNSEEN
         The number of messages which do not have the \Seen flag set.

   Example:    C: A042 STATUS blurdybloop (UIDNEXT MESSAGES)
               S: * STATUS blurdybloop (MESSAGES 231 UIDNEXT 44292)
               S: A042 OK STATUS completed

6.3.11. APPEND Command

   Arguments:  mailbox name
               OPTIONAL flag parenthesized list
               OPTIONAL date/time string
               message literal

   Responses:  no specific responses for this command

   Result:     OK - append completed
               NO - append error: can't append to that mailbox, error
                    in flags or date/time or message text
               BAD - command unknown or arguments invalid

      The APPEND command appends the literal argument as a new message
      to the end of the specified destination mailbox.  This argument
      SHOULD be in the format of an [RFC-2822] message.  8-bit
      characters are permitted in the message.  A server implementation
      that is unable to preserve 8-bit data properly MUST be able to
      reversibly convert 8-bit APPEND data to 7-bit using a [MIME-IMB]
      content transfer encoding.

           Note: There MAY be exceptions, e.g., draft messages, in
           which required [RFC-2822] header lines are omitted in
           the message literal argument to APPEND.  The full
           implications of doing so MUST be understood and
           carefully weighed.

      If a flag parenthesized list is specified, the flags SHOULD be set
      in the resulting message; otherwise, the flag list of the
      resulting message is set to empty by default.  In either case, the
      Recent flag is also set.

      If a date-time is specified, the internal date SHOULD be set in
      the resulting message; otherwise, the internal date of the
      resulting message is set to the current date and time by default.

      If the append is unsuccessful for any reason, the mailbox MUST be
      restored to its state before the APPEND attempt; no partial
      appending is permitted.

      If the destination mailbox does not exist, a server MUST return an
      error, and MUST NOT automatically create the mailbox.  Unless it
      is certain that the destination mailbox can not be created, the
      server MUST send the response code "[TRYCREATE]" as the prefix of
      the text of the tagged NO response.  This gives a hint to the
      client that it can attempt a CREATE command and retry the APPEND
      if the CREATE is successful.

      If the mailbox is currently selected, the normal new message
      actions SHOULD occur.  Specifically, the server SHOULD notify the
      client immediately via an untagged EXISTS response.  If the server
      does not do so, the client MAY issue a NOOP command (or failing
      that, a CHECK command) after one or more APPEND commands.

   Example:    C: A003 APPEND saved-messages (\Seen) {310}
               S: + Ready for literal data
               C: Date: Mon, 7 Feb 1994 21:52:25 -0800 (PST)
               C: From: Fred Foobar <foobar@Blurdybloop.COM>
               C: Subject: afternoon meeting
               C: To: mooch@owatagu.siam.edu
               C: Message-Id: <B27397-0100000@Blurdybloop.COM>
               C: MIME-Version: 1.0
               C: Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
               C:
               C: Hello Joe, do you think we can meet at 3:30 tomorrow?
               C:
               S: A003 OK APPEND completed

        Note: The APPEND command is not used for message delivery,
        because it does not provide a mechanism to transfer [SMTP]
        envelope information.

6.4.    Client Commands - Selected State

   In the selected state, commands that manipulate messages in a mailbox
   are permitted.

   In addition to the universal commands (CAPABILITY, NOOP, and LOGOUT),
   and the authenticated state commands (SELECT, EXAMINE, CREATE,
   DELETE, RENAME, SUBSCRIBE, UNSUBSCRIBE, LIST, LSUB, STATUS, and
   APPEND), the following commands are valid in the selected state:
   CHECK, CLOSE, EXPUNGE, SEARCH, FETCH, STORE, COPY, and UID.

6.4.1.  CHECK Command

   Arguments:  none

   Responses:  no specific responses for this command

   Result:     OK - check completed
               BAD - command unknown or arguments invalid

      The CHECK command requests a checkpoint of the currently selected
      mailbox.  A checkpoint refers to any implementation-dependent
      housekeeping associated with the mailbox (e.g., resolving the
      server's in-memory state of the mailbox with the state on its

      disk) that is not normally executed as part of each command.  A
      checkpoint MAY take a non-instantaneous amount of real time to
      complete.  If a server implementation has no such housekeeping
      considerations, CHECK is equivalent to NOOP.

      There is no guarantee that an EXISTS untagged response will happen
      as a result of CHECK.  NOOP, not CHECK, SHOULD be used for new
      message polling.

   Example:    C: FXXZ CHECK
               S: FXXZ OK CHECK Completed

6.4.2.  CLOSE Command

   Arguments:  none

   Responses:  no specific responses for this command

   Result:     OK - close completed, now in authenticated state
               BAD - command unknown or arguments invalid

      The CLOSE command permanently removes all messages that have the
      \Deleted flag set from the currently selected mailbox, and returns
      to the authenticated state from the selected state.  No untagged
      EXPUNGE responses are sent.

      No messages are removed, and no error is given, if the mailbox is
      selected by an EXAMINE command or is otherwise selected read-only.

      Even if a mailbox is selected, a SELECT, EXAMINE, or LOGOUT
      command MAY be issued without previously issuing a CLOSE command.
      The SELECT, EXAMINE, and LOGOUT commands implicitly close the
      currently selected mailbox without doing an expunge.  However,
      when many messages are deleted, a CLOSE-LOGOUT or CLOSE-SELECT
      sequence is considerably faster than an EXPUNGE-LOGOUT or
      EXPUNGE-SELECT because no untagged EXPUNGE responses (which the
      client would probably ignore) are sent.

   Example:    C: A341 CLOSE
               S: A341 OK CLOSE completed

6.4.3.  EXPUNGE Command

   Arguments:  none

   Responses:  untagged responses: EXPUNGE

   Result:     OK - expunge completed
               NO - expunge failure: can't expunge (e.g., permission
                    denied)
               BAD - command unknown or arguments invalid

      The EXPUNGE command permanently removes all messages that have the
      \Deleted flag set from the currently selected mailbox.  Before
      returning an OK to the client, an untagged EXPUNGE response is
      sent for each message that is removed.

   Example:    C: A202 EXPUNGE
               S: * 3 EXPUNGE
               S: * 3 EXPUNGE
               S: * 5 EXPUNGE
               S: * 8 EXPUNGE
               S: A202 OK EXPUNGE completed

        Note: In this example, messages 3, 4, 7, and 11 had the
        \Deleted flag set.  See the description of the EXPUNGE
        response for further explanation.

6.4.4.  SEARCH Command

   Arguments:  OPTIONAL [CHARSET] specification
               searching criteria (one or more)

   Responses:  REQUIRED untagged response: SEARCH

   Result:     OK - search completed
               NO - search error: can't search that [CHARSET] or
                    criteria
               BAD - command unknown or arguments invalid

      The SEARCH command searches the mailbox for messages that match
      the given searching criteria.  Searching criteria consist of one
      or more search keys.  The untagged SEARCH response from the server
      contains a listing of message sequence numbers corresponding to
      those messages that match the searching criteria.

      When multiple keys are specified, the result is the intersection
      (AND function) of all the messages that match those keys.  For
      example, the criteria DELETED FROM "SMITH" SINCE 1-Feb-1994 refers
      to all deleted messages from Smith that were placed in the mailbox
      since February 1, 1994.  A search key can also be a parenthesized
      list of one or more search keys (e.g., for use with the OR and NOT
      keys).

      Server implementations MAY exclude [MIME-IMB] body parts with
      terminal content media types other than TEXT and MESSAGE from
      consideration in SEARCH matching.

      The OPTIONAL [CHARSET] specification consists of the word
      "CHARSET" followed by a registered [CHARSET].  It indicates the
      [CHARSET] of the strings that appear in the search criteria.
      [MIME-IMB] content transfer encodings, and [MIME-HDRS] strings in
      [RFC-2822]/[MIME-IMB] headers, MUST be decoded before comparing
      text in a [CHARSET] other than US-ASCII.  US-ASCII MUST be
      supported; other [CHARSET]s MAY be supported.

      If the server does not support the specified [CHARSET], it MUST
      return a tagged NO response (not a BAD).  This response SHOULD
      contain the BADCHARSET response code, which MAY list the
      [CHARSET]s supported by the server.

      In all search keys that use strings, a message matches the key if
      the string is a substring of the field.  The matching is
      case-insensitive.

      The defined search keys are as follows.  Refer to the Formal
      Syntax section for the precise syntactic definitions of the
      arguments.

      <sequence set>
         Messages with message sequence numbers corresponding to the
         specified message sequence number set.

      ALL
         All messages in the mailbox; the default initial key for
         ANDing.

      ANSWERED
         Messages with the \Answered flag set.

      BCC <string>
         Messages that contain the specified string in the envelope
         structure's BCC field.

      BEFORE <date>
         Messages whose internal date (disregarding time and timezone)
         is earlier than the specified date.

      BODY <string>
         Messages that contain the specified string in the body of the
         message.

      CC <string>
         Messages that contain the specified string in the envelope
         structure's CC field.

      DELETED
         Messages with the \Deleted flag set.

      DRAFT
         Messages with the \Draft flag set.

      FLAGGED
         Messages with the \Flagged flag set.

      FROM <string>
         Messages that contain the specified string in the envelope
         structure's FROM field.

      HEADER <field-name> <string>
         Messages that have a header with the specified field-name (as
         defined in [RFC-2822]) and that contains the specified string
         in the text of the header (what comes after the colon).  If the
         string to search is zero-length, this matches all messages that
         have a header line with the specified field-name regardless of
         the contents.

      KEYWORD <flag>
         Messages with the specified keyword flag set.

      LARGER <n>
         Messages with an [RFC-2822] size larger than the specified
         number of octets.

      NEW
         Messages that have the \Recent flag set but not the \Seen flag.
         This is functionally equivalent to "(RECENT UNSEEN)".

      NOT <search-key>
         Messages that do not match the specified search key.

      OLD
         Messages that do not have the \Recent flag set.  This is
         functionally equivalent to "NOT RECENT" (as opposed to "NOT
         NEW").

      ON <date>
         Messages whose internal date (disregarding time and timezone)
         is within the specified date.

      OR <search-key1> <search-key2>
         Messages that match either search key.

      RECENT
         Messages that have the \Recent flag set.

      SEEN
         Messages that have the \Seen flag set.

      SENTBEFORE <date>
         Messages whose [RFC-2822] Date: header (disregarding time and
         timezone) is earlier than the specified date.

      SENTON <date>
         Messages whose [RFC-2822] Date: header (disregarding time and
         timezone) is within the specified date.

      SENTSINCE <date>
         Messages whose [RFC-2822] Date: header (disregarding time and
         timezone) is within or later than the specified date.

      SINCE <date>
         Messages whose internal date (disregarding time and timezone)
         is within or later than the specified date.

      SMALLER <n>
         Messages with an [RFC-2822] size smaller than the specified
         number of octets.

      SUBJECT <string>
         Messages that contain the specified string in the envelope
         structure's SUBJECT field.

      TEXT <string>
         Messages that contain the specified string in the header or
         body of the message.

      TO <string>
         Messages that contain the specified string in the envelope
         structure's TO field.

      UID <sequence set>
         Messages with unique identifiers corresponding to the specified
         unique identifier set.  Sequence set ranges are permitted.

      UNANSWERED
         Messages that do not have the \Answered flag set.

      UNDELETED
         Messages that do not have the \Deleted flag set.

      UNDRAFT
         Messages that do not have the \Draft flag set.

      UNFLAGGED
         Messages that do not have the \Flagged flag set.

      UNKEYWORD <flag>
         Messages that do not have the specified keyword flag set.

      UNSEEN
         Messages that do not have the \Seen flag set.

   Example:    C: A282 SEARCH FLAGGED SINCE 1-Feb-1994 NOT FROM "Smith"
               S: * SEARCH 2 84 882
               S: A282 OK SEARCH completed
               C: A283 SEARCH TEXT "string not in mailbox"
               S: * SEARCH
               S: A283 OK SEARCH completed
               C: A284 SEARCH CHARSET UTF-8 TEXT {6}
               C: XXXXXX
               S: * SEARCH 43
               S: A284 OK SEARCH completed

        Note: Since this document is restricted to 7-bit ASCII
        text, it is not possible to show actual UTF-8 data.  The
        "XXXXXX" is a placeholder for what would be 6 octets of
        8-bit data in an actual transaction.

6.4.5.  FETCH Command

   Arguments:  sequence set
               message data item names or macro

   Responses:  untagged responses: FETCH

   Result:     OK - fetch completed
               NO - fetch error: can't fetch that data
               BAD - command unknown or arguments invalid

      The FETCH command retrieves data associated with a message in the
      mailbox.  The data items to be fetched can be either a single atom
      or a parenthesized list.

      Most data items, identified in the formal syntax under the
      msg-att-static rule, are static and MUST NOT change for any
      particular message.  Other data items, identified in the formal
      syntax under the msg-att-dynamic rule, MAY change, either as a
      result of a STORE command or due to external events.

           For example, if a client receives an ENVELOPE for a
           message when it already knows the envelope, it can
           safely ignore the newly transmitted envelope.

      There are three macros which specify commonly-used sets of data
      items, and can be used instead of data items.  A macro must be
      used by itself, and not in conjunction with other macros or data
      items.

      ALL
         Macro equivalent to: (FLAGS INTERNALDATE RFC822.SIZE ENVELOPE)

      FAST
         Macro equivalent to: (FLAGS INTERNALDATE RFC822.SIZE)

      FULL
         Macro equivalent to: (FLAGS INTERNALDATE RFC822.SIZE ENVELOPE
         BODY)

      The currently defined data items that can be fetched are:

      BODY
         Non-extensible form of BODYSTRUCTURE.

      BODY[<section>]<<partial>>
         The text of a particular body section.  The section
         specification is a set of zero or more part specifiers
         delimited by periods.  A part specifier is either a part number
         or one of the following: HEADER, HEADER.FIELDS,
         HEADER.FIELDS.NOT, MIME, and TEXT.  An empty section
         specification refers to the entire message, including the
         header.

         Every message has at least one part number.  Non-[MIME-IMB]
         messages, and non-multipart [MIME-IMB] messages with no
         encapsulated message, only have a part 1.

         Multipart messages are assigned consecutive part numbers, as
         they occur in the message.  If a particular part is of type
         message or multipart, its parts MUST be indicated by a period
         followed by the part number within that nested multipart part.

         A part of type MESSAGE/RFC822 also has nested part numbers,
         referring to parts of the MESSAGE part's body.

         The HEADER, HEADER.FIELDS, HEADER.FIELDS.NOT, and TEXT part
         specifiers can be the sole part specifier or can be prefixed by
         one or more numeric part specifiers, provided that the numeric
         part specifier refers to a part of type MESSAGE/RFC822.  The
         MIME part specifier MUST be prefixed by one or more numeric
         part specifiers.

         The HEADER, HEADER.FIELDS, and HEADER.FIELDS.NOT part
         specifiers refer to the [RFC-2822] header of the message or of
         an encapsulated [MIME-IMT] MESSAGE/RFC822 message.
         HEADER.FIELDS and HEADER.FIELDS.NOT are followed by a list of
         field-name (as defined in [RFC-2822]) names, and return a

         subset of the header.  The subset returned by HEADER.FIELDS
         contains only those header fields with a field-name that
         matches one of the names in the list; similarly, the subset
         returned by HEADER.FIELDS.NOT contains only the header fields
         with a non-matching field-name.  The field-matching is
         case-insensitive but otherwise exact.  Subsetting does not
         exclude the [RFC-2822] delimiting blank line between the header
         and the body; the blank line is included in all header fetches,
         except in the case of a message which has no body and no blank
         line.

         The MIME part specifier refers to the [MIME-IMB] header for
         this part.

         The TEXT part specifier refers to the text body of the message,
         omitting the [RFC-2822] header.

            Here is an example of a complex message with some of its
            part specifiers:

       HEADER     ([RFC-2822] header of the message)
       TEXT       ([RFC-2822] text body of the message) MULTIPART/MIXED
       1          TEXT/PLAIN
       2          APPLICATION/OCTET-STREAM
       3          MESSAGE/RFC822
       3.HEADER   ([RFC-2822] header of the message)
       3.TEXT     ([RFC-2822] text body of the message) MULTIPART/MIXED
       3.1        TEXT/PLAIN
       3.2        APPLICATION/OCTET-STREAM
       4          MULTIPART/MIXED
       4.1        IMAGE/GIF
       4.1.MIME   ([MIME-IMB] header for the IMAGE/GIF)
       4.2        MESSAGE/RFC822
       4.2.HEADER ([RFC-2822] header of the message)
       4.2.TEXT   ([RFC-2822] text body of the message) MULTIPART/MIXED
       4.2.1      TEXT/PLAIN
       4.2.2      MULTIPART/ALTERNATIVE
       4.2.2.1    TEXT/PLAIN
       4.2.2.2    TEXT/RICHTEXT

         It is possible to fetch a substring of the designated text.
         This is done by appending an open angle bracket ("<"), the
         octet position of the first desired octet, a period, the
         maximum number of octets desired, and a close angle bracket
         (">") to the part specifier.  If the starting octet is beyond
         the end of the text, an empty string is returned.

         Any partial fetch that attempts to read beyond the end of the
         text is truncated as appropriate.  A partial fetch that starts
         at octet 0 is returned as a partial fetch, even if this
         truncation happened.

            Note: This means that BODY[]<0.2048> of a 1500-octet message
            will return BODY[]<0> with a literal of size 1500, not
            BODY[].

            Note: A substring fetch of a HEADER.FIELDS or
            HEADER.FIELDS.NOT part specifier is calculated after
            subsetting the header.

         The \Seen flag is implicitly set; if this causes the flags to
         change, they SHOULD be included as part of the FETCH responses.

      BODY.PEEK[<section>]<<partial>>
         An alternate form of BODY[<section>] that does not implicitly
         set the \Seen flag.

      BODYSTRUCTURE
         The [MIME-IMB] body structure of the message.  This is computed
         by the server by parsing the [MIME-IMB] header fields in the
         [RFC-2822] header and [MIME-IMB] headers.

      ENVELOPE
         The envelope structure of the message.  This is computed by the
         server by parsing the [RFC-2822] header into the component
         parts, defaulting various fields as necessary.

      FLAGS
         The flags that are set for this message.

      INTERNALDATE
         The internal date of the message.

      RFC822
         Functionally equivalent to BODY[], differing in the syntax of
         the resulting untagged FETCH data (RFC822 is returned).

      RFC822.HEADER
         Functionally equivalent to BODY.PEEK[HEADER], differing in the
         syntax of the resulting untagged FETCH data (RFC822.HEADER is
         returned).

      RFC822.SIZE
         The [RFC-2822] size of the message.

      RFC822.TEXT
         Functionally equivalent to BODY[TEXT], differing in the syntax
         of the resulting untagged FETCH data (RFC822.TEXT is returned).

      UID
         The unique identifier for the message.

   Example:    C: A654 FETCH 2:4 (FLAGS BODY[HEADER.FIELDS (DATE FROM)])
               S: * 2 FETCH ....
               S: * 3 FETCH ....
               S: * 4 FETCH ....
               S: A654 OK FETCH completed

6.4.6.  STORE Command

   Arguments:  sequence set
               message data item name
               value for message data item

   Responses:  untagged responses: FETCH

   Result:     OK - store completed
               NO - store error: can't store that data
               BAD - command unknown or arguments invalid

      The STORE command alters data associated with a message in the
      mailbox.  Normally, STORE will return the updated value of the
      data with an untagged FETCH response.  A suffix of ".SILENT" in
      the data item name prevents the untagged FETCH, and the server
      SHOULD assume that the client has determined the updated value
      itself or does not care about the updated value.

           Note: Regardless of whether or not the ".SILENT" suffix
           was used, the server SHOULD send an untagged FETCH
           response if a change to a message's flags from an
           external source is observed.  The intent is that the
           status of the flags is determinate without a race
           condition.

      The currently defined data items that can be stored are:

      FLAGS <flag list>
         Replace the flags for the message (other than \Recent) with the
         argument.  The new value of the flags is returned as if a FETCH
         of those flags was done.

      FLAGS.SILENT <flag list>
         Equivalent to FLAGS, but without returning a new value.

      +FLAGS <flag list>
         Add the argument to the flags for the message.  The new value
         of the flags is returned as if a FETCH of those flags was done.

      +FLAGS.SILENT <flag list>
         Equivalent to +FLAGS, but without returning a new value.

      -FLAGS <flag list>
         Remove the argument from the flags for the message.  The new
         value of the flags is returned as if a FETCH of those flags was
         done.

      -FLAGS.SILENT <flag list>
         Equivalent to -FLAGS, but without returning a new value.

   Example:    C: A003 STORE 2:4 +FLAGS (\Deleted)
               S: * 2 FETCH (FLAGS (\Deleted \Seen))
               S: * 3 FETCH (FLAGS (\Deleted))
               S: * 4 FETCH (FLAGS (\Deleted \Flagged \Seen))
               S: A003 OK STORE completed

6.4.7.  COPY Command

   Arguments:  sequence set
               mailbox name

   Responses:  no specific responses for this command

   Result:     OK - copy completed
               NO - copy error: can't copy those messages or to that
                    name
               BAD - command unknown or arguments invalid

      The COPY command copies the specified message(s) to the end of the
      specified destination mailbox.  The flags and internal date of the
      message(s) SHOULD be preserved, and the Recent flag SHOULD be set,
      in the copy.

      If the destination mailbox does not exist, a server SHOULD return
      an error.  It SHOULD NOT automatically create the mailbox.  Unless
      it is certain that the destination mailbox can not be created, the
      server MUST send the response code "[TRYCREATE]" as the prefix of
      the text of the tagged NO response.  This gives a hint to the
      client that it can attempt a CREATE command and retry the COPY if
      the CREATE is successful.

      If the COPY command is unsuccessful for any reason, server
      implementations MUST restore the destination mailbox to its state
      before the COPY attempt.

   Example:    C: A003 COPY 2:4 MEETING
               S: A003 OK COPY completed

6.4.8.  UID Command

   Arguments:  command name
               command arguments

   Responses:  untagged responses: FETCH, SEARCH

   Result:     OK - UID command completed
               NO - UID command error
               BAD - command unknown or arguments invalid

      The UID command has two forms.  In the first form, it takes as its
      arguments a COPY, FETCH, or STORE command with arguments
      appropriate for the associated command.  However, the numbers in
      the sequence set argument are unique identifiers instead of
      message sequence numbers.  Sequence set ranges are permitted, but
      there is no guarantee that unique identifiers will be contiguous.

      A non-existent unique identifier is ignored without any error
      message generated.  Thus, it is possible for a UID FETCH command
      to return an OK without any data or a UID COPY or UID STORE to
      return an OK without performing any operations.

      In the second form, the UID command takes a SEARCH command with
      SEARCH command arguments.  The interpretation of the arguments is
      the same as with SEARCH; however, the numbers returned in a SEARCH
      response for a UID SEARCH command are unique identifiers instead

      of message sequence numbers.  For example, the command UID SEARCH
      1:100 UID 443:557 returns the unique identifiers corresponding to
      the intersection of two sequence sets, the message sequence number
      range 1:100 and the UID range 443:557.

           Note: in the above example, the UID range 443:557
           appears.  The same comment about a non-existent unique
           identifier being ignored without any error message also
           applies here.  Hence, even if neither UID 443 or 557
           exist, this range is valid and would include an existing
           UID 495.

           Also note that a UID range of 559:* always includes the
           UID of the last message in the mailbox, even if 559 is
           higher than any assigned UID value.  This is because the
           contents of a range are independent of the order of the
           range endpoints.  Thus, any UID range with * as one of
           the endpoints indicates at least one message (the
           message with the highest numbered UID), unless the
           mailbox is empty.

      The number after the "*" in an untagged FETCH response is always a
      message sequence number, not a unique identifier, even for a UID
      command response.  However, server implementations MUST implicitly
      include the UID message data item as part of any FETCH response
      caused by a UID command, regardless of whether a UID was specified
      as a message data item to the FETCH.

      Note: The rule about including the UID message data item as part
      of a FETCH response primarily applies to the UID FETCH and UID
      STORE commands, including a UID FETCH command that does not
      include UID as a message data item.  Although it is unlikely that
      the other UID commands will cause an untagged FETCH, this rule
      applies to these commands as well.

   Example:    C: A999 UID FETCH 4827313:4828442 FLAGS
               S: * 23 FETCH (FLAGS (\Seen) UID 4827313)
               S: * 24 FETCH (FLAGS (\Seen) UID 4827943)
               S: * 25 FETCH (FLAGS (\Seen) UID 4828442)
               S: A999 OK UID FETCH completed

6.5.    Client Commands - Experimental/Expansion

6.5.1.  X<atom> Command

   Arguments:  implementation defined

   Responses:  implementation defined

   Result:     OK - command completed
               NO - failure
               BAD - command unknown or arguments invalid

      Any command prefixed with an X is an experimental command.
      Commands which are not part of this specification, a standard or
      standards-track revision of this specification, or an
      IESG-approved experimental protocol, MUST use the X prefix.

      Any added untagged responses issued by an experimental command
      MUST also be prefixed with an X.  Server implementations MUST NOT
      send any such untagged responses, unless the client requested it
      by issuing the associated experimental command.

   Example:    C: a441 CAPABILITY
               S: * CAPABILITY IMAP4rev1 XPIG-LATIN
               S: a441 OK CAPABILITY completed
               C: A442 XPIG-LATIN
               S: * XPIG-LATIN ow-nay eaking-spay ig-pay atin-lay
               S: A442 OK XPIG-LATIN ompleted-cay

7.      Server Responses

   Server responses are in three forms: status responses, server data,
   and command continuation request.  The information contained in a
   server response, identified by "Contents:" in the response
   descriptions below, is described by function, not by syntax.  The
   precise syntax of server responses is described in the Formal Syntax
   section.

   The client MUST be prepared to accept any response at all times.

   Status responses can be tagged or untagged.  Tagged status responses
   indicate the completion result (OK, NO, or BAD status) of a client
   command, and have a tag matching the command.

   Some status responses, and all server data, are untagged.  An
   untagged response is indicated by the token "*" instead of a tag.
   Untagged status responses indicate server greeting, or server status

   that does not indicate the completion of a command (for example, an
   impending system shutdown alert).  For historical reasons, untagged
   server data responses are also called "unsolicited data", although
   strictly speaking, only unilateral server data is truly
   "unsolicited".

   Certain server data MUST be recorded by the client when it is
   received; this is noted in the description of that data.  Such data
   conveys critical information which affects the interpretation of all
   subsequent commands and responses (e.g., updates reflecting the
   creation or destruction of messages).

   Other server data SHOULD be recorded for later reference; if the
   client does not need to record the data, or if recording the data has
   no obvious purpose (e.g., a SEARCH response when no SEARCH command is
   in progress), the data SHOULD be ignored.

   An example of unilateral untagged server data occurs when the IMAP
   connection is in the selected state.  In the selected state, the
   server checks the mailbox for new messages as part of command
   execution.  Normally, this is part of the execution of every command;
   hence, a NOOP command suffices to check for new messages.  If new
   messages are found, the server sends untagged EXISTS and RECENT
   responses reflecting the new size of the mailbox.  Server
   implementations that offer multiple simultaneous access to the same
   mailbox SHOULD also send appropriate unilateral untagged FETCH and
   EXPUNGE responses if another agent changes the state of any message
   flags or expunges any messages.

   Command continuation request responses use the token "+" instead of a
   tag.  These responses are sent by the server to indicate acceptance
   of an incomplete client command and readiness for the remainder of
   the command.

7.1.    Server Responses - Status Responses

   Status responses are OK, NO, BAD, PREAUTH and BYE.  OK, NO, and BAD
   can be tagged or untagged.  PREAUTH and BYE are always untagged.

   Status responses MAY include an OPTIONAL "response code".  A response
   code consists of data inside square brackets in the form of an atom,
   possibly followed by a space and arguments.  The response code
   contains additional information or status codes for client software
   beyond the OK/NO/BAD condition, and are defined when there is a
   specific action that a client can take based upon the additional
   information.

   The currently defined response codes are:

      ALERT

         The human-readable text contains a special alert that MUST be
         presented to the user in a fashion that calls the user's
         attention to the message.

      BADCHARSET

         Optionally followed by a parenthesized list of charsets.  A
         SEARCH failed because the given charset is not supported by
         this implementation.  If the optional list of charsets is
         given, this lists the charsets that are supported by this
         implementation.

      CAPABILITY

         Followed by a list of capabilities.  This can appear in the
         initial OK or PREAUTH response to transmit an initial
         capabilities list.  This makes it unnecessary for a client to
         send a separate CAPABILITY command if it recognizes this
         response.

      PARSE

         The human-readable text represents an error in parsing the
         [RFC-2822] header or [MIME-IMB] headers of a message in the
         mailbox.

      PERMANENTFLAGS

         Followed by a parenthesized list of flags, indicates which of
         the known flags the client can change permanently.  Any flags
         that are in the FLAGS untagged response, but not the
         PERMANENTFLAGS list, can not be set permanently.  If the client
         attempts to STORE a flag that is not in the PERMANENTFLAGS
         list, the server will either ignore the change or store the
         state change for the remainder of the current session only.
         The PERMANENTFLAGS list can also include the special flag \*,
         which indicates that it is possible to create new keywords by
         attempting to store those flags in the mailbox.

      READ-ONLY

         The mailbox is selected read-only, or its access while selected
         has changed from read-write to read-only.

      READ-WRITE

         The mailbox is selected read-write, or its access while
         selected has changed from read-only to read-write.

      TRYCREATE

         An APPEND or COPY attempt is failing because the target mailbox
         does not exist (as opposed to some other reason).  This is a
         hint to the client that the operation can succeed if the
         mailbox is first created by the CREATE command.

      UIDNEXT

         Followed by a decimal number, indicates the next unique
         identifier value.  Refer to section 2.3.1.1 for more
         information.

      UIDVALIDITY

         Followed by a decimal number, indicates the unique identifier
         validity value.  Refer to section 2.3.1.1 for more information.

      UNSEEN

         Followed by a decimal number, indicates the number of the first
         message without the \Seen flag set.

      Additional response codes defined by particular client or server
      implementations SHOULD be prefixed with an "X" until they are
      added to a revision of this protocol.  Client implementations
      SHOULD ignore response codes that they do not recognize.

7.1.1.  OK Response

   Contents:   OPTIONAL response code
               human-readable text

      The OK response indicates an information message from the server.
      When tagged, it indicates successful completion of the associated
      command.  The human-readable text MAY be presented to the user as
      an information message.  The untagged form indicates an

      information-only message; the nature of the information MAY be
      indicated by a response code.

      The untagged form is also used as one of three possible greetings
      at connection startup.  It indicates that the connection is not
      yet authenticated and that a LOGIN command is needed.

   Example:    S: * OK IMAP4rev1 server ready
               C: A001 LOGIN fred blurdybloop
               S: * OK [ALERT] System shutdown in 10 minutes
               S: A001 OK LOGIN Completed

7.1.2.  NO Response

   Contents:   OPTIONAL response code
               human-readable text

      The NO response indicates an operational error message from the
      server.  When tagged, it indicates unsuccessful completion of the
      associated command.  The untagged form indicates a warning; the
      command can still complete successfully.  The human-readable text
      describes the condition.

   Example:    C: A222 COPY 1:2 owatagusiam
               S: * NO Disk is 98% full, please delete unnecessary data
               S: A222 OK COPY completed
               C: A223 COPY 3:200 blurdybloop
               S: * NO Disk is 98% full, please delete unnecessary data
               S: * NO Disk is 99% full, please delete unnecessary data
               S: A223 NO COPY failed: disk is full

7.1.3.  BAD Response

   Contents:   OPTIONAL response code
               human-readable text

      The BAD response indicates an error message from the server.  When
      tagged, it reports a protocol-level error in the client's command;
      the tag indicates the command that caused the error.  The untagged
      form indicates a protocol-level error for which the associated
      command can not be determined; it can also indicate an internal
      server failure.  The human-readable text describes the condition.

   Example:    C: ...very long command line...
               S: * BAD Command line too long
               C: ...empty line...
               S: * BAD Empty command line
               C: A443 EXPUNGE
               S: * BAD Disk crash, attempting salvage to a new disk!
               S: * OK Salvage successful, no data lost
               S: A443 OK Expunge completed

7.1.4.  PREAUTH Response

   Contents:   OPTIONAL response code
               human-readable text

      The PREAUTH response is always untagged, and is one of three
      possible greetings at connection startup.  It indicates that the
      connection has already been authenticated by external means; thus
      no LOGIN command is needed.

   Example:    S: * PREAUTH IMAP4rev1 server logged in as Smith

7.1.5.  BYE Response

   Contents:   OPTIONAL response code
               human-readable text

      The BYE response is always untagged, and indicates that the server
      is about to close the connection.  The human-readable text MAY be
      displayed to the user in a status report by the client.  The BYE
      response is sent under one of four conditions:

         1) as part of a normal logout sequence.  The server will close
            the connection after sending the tagged OK response to the
            LOGOUT command.

         2) as a panic shutdown announcement.  The server closes the
            connection immediately.

         3) as an announcement of an inactivity autologout.  The server
            closes the connection immediately.

         4) as one of three possible greetings at connection startup,
            indicating that the server is not willing to accept a
            connection from this client.  The server closes the
            connection immediately.

      The difference between a BYE that occurs as part of a normal
      LOGOUT sequence (the first case) and a BYE that occurs because of
      a failure (the other three cases) is that the connection closes
      immediately in the failure case.  In all cases the client SHOULD
      continue to read response data from the server until the
      connection is closed; this will ensure that any pending untagged
      or completion responses are read and processed.

   Example:    S: * BYE Autologout; idle for too long

7.2.    Server Responses - Server and Mailbox Status

   These responses are always untagged.  This is how server and mailbox
   status data are transmitted from the server to the client.  Many of
   these responses typically result from a command with the same name.

7.2.1.  CAPABILITY Response

   Contents:   capability listing

      The CAPABILITY response occurs as a result of a CAPABILITY
      command.  The capability listing contains a space-separated
      listing of capability names that the server supports.  The
      capability listing MUST include the atom "IMAP4rev1".

      In addition, client and server implementations MUST implement the
      STARTTLS, LOGINDISABLED, and AUTH=PLAIN (described in [IMAP-TLS])
      capabilities.  See the Security Considerations section for
      important information.

      A capability name which begins with "AUTH=" indicates that the
      server supports that particular authentication mechanism.

      The LOGINDISABLED capability indicates that the LOGIN command is
      disabled, and that the server will respond with a tagged NO
      response to any attempt to use the LOGIN command even if the user
      name and password are valid.  An IMAP client MUST NOT issue the
      LOGIN command if the server advertises the LOGINDISABLED
      capability.

      Other capability names indicate that the server supports an
      extension, revision, or amendment to the IMAP4rev1 protocol.
      Server responses MUST conform to this document until the client
      issues a command that uses the associated capability.

      Capability names MUST either begin with "X" or be standard or
      standards-track IMAP4rev1 extensions, revisions, or amendments
      registered with IANA.  A server MUST NOT offer unregistered or

      non-standard capability names, unless such names are prefixed with
      an "X".

      Client implementations SHOULD NOT require any capability name
      other than "IMAP4rev1", and MUST ignore any unknown capability
      names.

      A server MAY send capabilities automatically, by using the
      CAPABILITY response code in the initial PREAUTH or OK responses,
      and by sending an updated CAPABILITY response code in the tagged
      OK response as part of a successful authentication.  It is
      unnecessary for a client to send a separate CAPABILITY command if
      it recognizes these automatic capabilities.

   Example:    S: * CAPABILITY IMAP4rev1 STARTTLS AUTH=GSSAPI XPIG-LATIN

7.2.2.  LIST Response

   Contents:   name attributes
               hierarchy delimiter
               name

      The LIST response occurs as a result of a LIST command.  It
      returns a single name that matches the LIST specification.  There
      can be multiple LIST responses for a single LIST command.

      Four name attributes are defined:

      \Noinferiors
         It is not possible for any child levels of hierarchy to exist
         under this name; no child levels exist now and none can be
         created in the future.

      \Noselect
         It is not possible to use this name as a selectable mailbox.

      \Marked
         The mailbox has been marked "interesting" by the server; the
         mailbox probably contains messages that have been added since
         the last time the mailbox was selected.

      \Unmarked
         The mailbox does not contain any additional messages since the
         last time the mailbox was selected.

      If it is not feasible for the server to determine whether or not
      the mailbox is "interesting", or if the name is a \Noselect name,
      the server SHOULD NOT send either \Marked or \Unmarked.

      The hierarchy delimiter is a character used to delimit levels of
      hierarchy in a mailbox name.  A client can use it to create child
      mailboxes, and to search higher or lower levels of naming
      hierarchy.  All children of a top-level hierarchy node MUST use
      the same separator character.  A NIL hierarchy delimiter means
      that no hierarchy exists; the name is a "flat" name.

      The name represents an unambiguous left-to-right hierarchy, and
      MUST be valid for use as a reference in LIST and LSUB commands.
      Unless \Noselect is indicated, the name MUST also be valid as an
      argument for commands, such as SELECT, that accept mailbox names.

   Example:    S: * LIST (\Noselect) "/" ~/Mail/foo

7.2.3.  LSUB Response

   Contents:   name attributes
               hierarchy delimiter
               name

      The LSUB response occurs as a result of an LSUB command.  It
      returns a single name that matches the LSUB specification.  There
      can be multiple LSUB responses for a single LSUB command.  The
      data is identical in format to the LIST response.

   Example:    S: * LSUB () "." #news.comp.mail.misc

7.2.4   STATUS Response

   Contents:   name
               status parenthesized list

      The STATUS response occurs as a result of an STATUS command.  It
      returns the mailbox name that matches the STATUS specification and
      the requested mailbox status information.

   Example:    S: * STATUS blurdybloop (MESSAGES 231 UIDNEXT 44292)

7.2.5.  SEARCH Response

   Contents:   zero or more numbers

      The SEARCH response occurs as a result of a SEARCH or UID SEARCH
      command.  The number(s) refer to those messages that match the
      search criteria.  For SEARCH, these are message sequence numbers;
      for UID SEARCH, these are unique identifiers.  Each number is
      delimited by a space.

   Example:    S: * SEARCH 2 3 6

7.2.6.  FLAGS Response

   Contents:   flag parenthesized list

      The FLAGS response occurs as a result of a SELECT or EXAMINE
      command.  The flag parenthesized list identifies the flags (at a
      minimum, the system-defined flags) that are applicable for this
      mailbox.  Flags other than the system flags can also exist,
      depending on server implementation.

      The update from the FLAGS response MUST be recorded by the client.

   Example:    S: * FLAGS (\Answered \Flagged \Deleted \Seen \Draft)

7.3.    Server Responses - Mailbox Size

   These responses are always untagged.  This is how changes in the size
   of the mailbox are transmitted from the server to the client.
   Immediately following the "*" token is a number that represents a
   message count.

7.3.1.  EXISTS Response

   Contents:   none

      The EXISTS response reports the number of messages in the mailbox.
      This response occurs as a result of a SELECT or EXAMINE command,
      and if the size of the mailbox changes (e.g., new messages).

      The update from the EXISTS response MUST be recorded by the
      client.

   Example:    S: * 23 EXISTS

7.3.2.  RECENT Response

   Contents:   none

      The RECENT response reports the number of messages with the
      \Recent flag set.  This response occurs as a result of a SELECT or
      EXAMINE command, and if the size of the mailbox changes (e.g., new
      messages).

           Note: It is not guaranteed that the message sequence
           numbers of recent messages will be a contiguous range of
           the highest n messages in the mailbox (where n is the
           value reported by the RECENT response).  Examples of
           situations in which this is not the case are: multiple
           clients having the same mailbox open (the first session
           to be notified will see it as recent, others will
           probably see it as non-recent), and when the mailbox is
           re-ordered by a non-IMAP agent.

           The only reliable way to identify recent messages is to
           look at message flags to see which have the \Recent flag
           set, or to do a SEARCH RECENT.

      The update from the RECENT response MUST be recorded by the
      client.

   Example:    S: * 5 RECENT

7.4.    Server Responses - Message Status

   These responses are always untagged.  This is how message data are
   transmitted from the server to the client, often as a result of a
   command with the same name.  Immediately following the "*" token is a
   number that represents a message sequence number.

7.4.1.  EXPUNGE Response

   Contents:   none

      The EXPUNGE response reports that the specified message sequence
      number has been permanently removed from the mailbox.  The message
      sequence number for each successive message in the mailbox is
      immediately decremented by 1, and this decrement is reflected in
      message sequence numbers in subsequent responses (including other
      untagged EXPUNGE responses).

      The EXPUNGE response also decrements the number of messages in the
      mailbox; it is not necessary to send an EXISTS response with the
      new value.

      As a result of the immediate decrement rule, message sequence
      numbers that appear in a set of successive EXPUNGE responses
      depend upon whether the messages are removed starting from lower
      numbers to higher numbers, or from higher numbers to lower
      numbers.  For example, if the last 5 messages in a 9-message
      mailbox are expunged, a "lower to higher" server will send five
      untagged EXPUNGE responses for message sequence number 5, whereas
      a "higher to lower server" will send successive untagged EXPUNGE
      responses for message sequence numbers 9, 8, 7, 6, and 5.

      An EXPUNGE response MUST NOT be sent when no command is in
      progress, nor while responding to a FETCH, STORE, or SEARCH
      command.  This rule is necessary to prevent a loss of
      synchronization of message sequence numbers between client and
      server.  A command is not "in progress" until the complete command
      has been received; in particular, a command is not "in progress"
      during the negotiation of command continuation.

           Note: UID FETCH, UID STORE, and UID SEARCH are different
           commands from FETCH, STORE, and SEARCH.  An EXPUNGE
           response MAY be sent during a UID command.

      The update from the EXPUNGE response MUST be recorded by the
      client.

   Example:    S: * 44 EXPUNGE

7.4.2.  FETCH Response

   Contents:   message data

      The FETCH response returns data about a message to the client.
      The data are pairs of data item names and their values in
      parentheses.  This response occurs as the result of a FETCH or
      STORE command, as well as by unilateral server decision (e.g.,
      flag updates).

      The current data items are:

      BODY
         A form of BODYSTRUCTURE without extension data.

      BODY[<section>]<<origin octet>>
         A string expressing the body contents of the specified section.
         The string SHOULD be interpreted by the client according to the
         content transfer encoding, body type, and subtype.

         If the origin octet is specified, this string is a substring of
         the entire body contents, starting at that origin octet.  This
         means that BODY[]<0> MAY be truncated, but BODY[] is NEVER
         truncated.

            Note: The origin octet facility MUST NOT be used by a server
            in a FETCH response unless the client specifically requested
            it by means of a FETCH of a BODY[<section>]<<partial>> data
            item.

         8-bit textual data is permitted if a [CHARSET] identifier is
         part of the body parameter parenthesized list for this section.
         Note that headers (part specifiers HEADER or MIME, or the
         header portion of a MESSAGE/RFC822 part), MUST be 7-bit; 8-bit
         characters are not permitted in headers.  Note also that the
         [RFC-2822] delimiting blank line between the header and the
         body is not affected by header line subsetting; the blank line
         is always included as part of header data, except in the case
         of a message which has no body and no blank line.

         Non-textual data such as binary data MUST be transfer encoded
         into a textual form, such as BASE64, prior to being sent to the
         client.  To derive the original binary data, the client MUST
         decode the transfer encoded string.

      BODYSTRUCTURE
         A parenthesized list that describes the [MIME-IMB] body
         structure of a message.  This is computed by the server by
         parsing the [MIME-IMB] header fields, defaulting various fields
         as necessary.

         For example, a simple text message of 48 lines and 2279 octets
         can have a body structure of: ("TEXT" "PLAIN" ("CHARSET"
         "US-ASCII") NIL NIL "7BIT" 2279 48)

         Multiple parts are indicated by parenthesis nesting.  Instead
         of a body type as the first element of the parenthesized list,
         there is a sequence of one or more nested body structures.  The
         second element of the parenthesized list is the multipart
         subtype (mixed, digest, parallel, alternative, etc.).

         For example, a two part message consisting of a text and a
         BASE64-encoded text attachment can have a body structure of:
         (("TEXT" "PLAIN" ("CHARSET" "US-ASCII") NIL NIL "7BIT" 1152
         23)("TEXT" "PLAIN" ("CHARSET" "US-ASCII" "NAME" "cc.diff")
         "<960723163407.20117h@cac.washington.edu>" "Compiler diff"
         "BASE64" 4554 73) "MIXED")

         Extension data follows the multipart subtype.  Extension data
         is never returned with the BODY fetch, but can be returned with
         a BODYSTRUCTURE fetch.  Extension data, if present, MUST be in
         the defined order.  The extension data of a multipart body part
         are in the following order:

         body parameter parenthesized list
            A parenthesized list of attribute/value pairs [e.g., ("foo"
            "bar" "baz" "rag") where "bar" is the value of "foo", and
            "rag" is the value of "baz"] as defined in [MIME-IMB].

         body disposition
            A parenthesized list, consisting of a disposition type
            string, followed by a parenthesized list of disposition
            attribute/value pairs as defined in [DISPOSITION].

         body language
            A string or parenthesized list giving the body language
            value as defined in [LANGUAGE-TAGS].

         body location
            A string list giving the body content URI as defined in
            [LOCATION].

         Any following extension data are not yet defined in this
         version of the protocol.  Such extension data can consist of
         zero or more NILs, strings, numbers, or potentially nested
         parenthesized lists of such data.  Client implementations that
         do a BODYSTRUCTURE fetch MUST be prepared to accept such
         extension data.  Server implementations MUST NOT send such
         extension data until it has been defined by a revision of this
         protocol.

         The basic fields of a non-multipart body part are in the
         following order:

         body type
            A string giving the content media type name as defined in
            [MIME-IMB].

         body subtype
            A string giving the content subtype name as defined in
            [MIME-IMB].

         body parameter parenthesized list
            A parenthesized list of attribute/value pairs [e.g., ("foo"
            "bar" "baz" "rag") where "bar" is the value of "foo" and
            "rag" is the value of "baz"] as defined in [MIME-IMB].

         body id
            A string giving the content id as defined in [MIME-IMB].

         body description
            A string giving the content description as defined in
            [MIME-IMB].

         body encoding
            A string giving the content transfer encoding as defined in
            [MIME-IMB].

         body size
            A number giving the size of the body in octets.  Note that
            this size is the size in its transfer encoding and not the
            resulting size after any decoding.

         A body type of type MESSAGE and subtype RFC822 contains,
         immediately after the basic fields, the envelope structure,
         body structure, and size in text lines of the encapsulated
         message.

         A body type of type TEXT contains, immediately after the basic
         fields, the size of the body in text lines.  Note that this
         size is the size in its content transfer encoding and not the
         resulting size after any decoding.

         Extension data follows the basic fields and the type-specific
         fields listed above.  Extension data is never returned with the
         BODY fetch, but can be returned with a BODYSTRUCTURE fetch.
         Extension data, if present, MUST be in the defined order.

         The extension data of a non-multipart body part are in the
         following order:

         body MD5
            A string giving the body MD5 value as defined in [MD5].

         body disposition
            A parenthesized list with the same content and function as
            the body disposition for a multipart body part.

         body language
            A string or parenthesized list giving the body language
            value as defined in [LANGUAGE-TAGS].

         body location
            A string list giving the body content URI as defined in
            [LOCATION].

         Any following extension data are not yet defined in this
         version of the protocol, and would be as described above under
         multipart extension data.

      ENVELOPE
         A parenthesized list that describes the envelope structure of a
         message.  This is computed by the server by parsing the
         [RFC-2822] header into the component parts, defaulting various
         fields as necessary.

         The fields of the envelope structure are in the following
         order: date, subject, from, sender, reply-to, to, cc, bcc,
         in-reply-to, and message-id.  The date, subject, in-reply-to,
         and message-id fields are strings.  The from, sender, reply-to,
         to, cc, and bcc fields are parenthesized lists of address
         structures.

         An address structure is a parenthesized list that describes an
         electronic mail address.  The fields of an address structure
         are in the following order: personal name, [SMTP]
         at-domain-list (source route), mailbox name, and host name.

         [RFC-2822] group syntax is indicated by a special form of
         address structure in which the host name field is NIL.  If the
         mailbox name field is also NIL, this is an end of group marker
         (semi-colon in RFC 822 syntax).  If the mailbox name field is
         non-NIL, this is a start of group marker, and the mailbox name
         field holds the group name phrase.

         If the Date, Subject, In-Reply-To, and Message-ID header lines
         are absent in the [RFC-2822] header, the corresponding member
         of the envelope is NIL; if these header lines are present but
         empty the corresponding member of the envelope is the empty
         string.

            Note: some servers may return a NIL envelope member in the
            "present but empty" case.  Clients SHOULD treat NIL and
            empty string as identical.

            Note: [RFC-2822] requires that all messages have a valid
            Date header.  Therefore, the date member in the envelope can
            not be NIL or the empty string.

            Note: [RFC-2822] requires that the In-Reply-To and
            Message-ID headers, if present, have non-empty content.
            Therefore, the in-reply-to and message-id members in the
            envelope can not be the empty string.

         If the From, To, cc, and bcc header lines are absent in the
         [RFC-2822] header, or are present but empty, the corresponding
         member of the envelope is NIL.

         If the Sender or Reply-To lines are absent in the [RFC-2822]
         header, or are present but empty, the server sets the
         corresponding member of the envelope to be the same value as
         the from member (the client is not expected to know to do
         this).

            Note: [RFC-2822] requires that all messages have a valid
            From header.  Therefore, the from, sender, and reply-to
            members in the envelope can not be NIL.

      FLAGS
         A parenthesized list of flags that are set for this message.

      INTERNALDATE
         A string representing the internal date of the message.

      RFC822
         Equivalent to BODY[].

      RFC822.HEADER
         Equivalent to BODY[HEADER].  Note that this did not result in
         \Seen being set, because RFC822.HEADER response data occurs as
         a result of a FETCH of RFC822.HEADER.  BODY[HEADER] response
         data occurs as a result of a FETCH of BODY[HEADER] (which sets
         \Seen) or BODY.PEEK[HEADER] (which does not set \Seen).

      RFC822.SIZE
         A number expressing the [RFC-2822] size of the message.

      RFC822.TEXT
         Equivalent to BODY[TEXT].

      UID
         A number expressing the unique identifier of the message.

   Example:    S: * 23 FETCH (FLAGS (\Seen) RFC822.SIZE 44827)

7.5.    Server Responses - Command Continuation Request

   The command continuation request response is indicated by a "+" token
   instead of a tag.  This form of response indicates that the server is
   ready to accept the continuation of a command from the client.  The
   remainder of this response is a line of text.

   This response is used in the AUTHENTICATE command to transmit server
   data to the client, and request additional client data.  This
   response is also used if an argument to any command is a literal.

   The client is not permitted to send the octets of the literal unless
   the server indicates that it is expected.  This permits the server to
   process commands and reject errors on a line-by-line basis.  The
   remainder of the command, including the CRLF that terminates a
   command, follows the octets of the literal.  If there are any
   additional command arguments, the literal octets are followed by a
   space and those arguments.

   Example:    C: A001 LOGIN {11}
               S: + Ready for additional command text
               C: FRED FOOBAR {7}
               S: + Ready for additional command text
               C: fat man
               S: A001 OK LOGIN completed
               C: A044 BLURDYBLOOP {102856}
               S: A044 BAD No such command as "BLURDYBLOOP"

8.      Sample IMAP4rev1 connection

   The following is a transcript of an IMAP4rev1 connection.  A long
   line in this sample is broken for editorial clarity.

S:   * OK IMAP4rev1 Service Ready
C:   a001 login mrc secret
S:   a001 OK LOGIN completed
C:   a002 select inbox
S:   * 18 EXISTS
S:   * FLAGS (\Answered \Flagged \Deleted \Seen \Draft)
S:   * 2 RECENT
S:   * OK [UNSEEN 17] Message 17 is the first unseen message
S:   * OK [UIDVALIDITY 3857529045] UIDs valid
S:   a002 OK [READ-WRITE] SELECT completed
C:   a003 fetch 12 full
S:   * 12 FETCH (FLAGS (\Seen) INTERNALDATE "17-Jul-1996 02:44:25 -0700"
      RFC822.SIZE 4286 ENVELOPE ("Wed, 17 Jul 1996 02:23:25 -0700 (PDT)"
      "IMAP4rev1 WG mtg summary and minutes"
      (("Terry Gray" NIL "gray" "cac.washington.edu"))
      (("Terry Gray" NIL "gray" "cac.washington.edu"))
      (("Terry Gray" NIL "gray" "cac.washington.edu"))
      ((NIL NIL "imap" "cac.washington.edu"))
      ((NIL NIL "minutes" "CNRI.Reston.VA.US")
      ("John Klensin" NIL "KLENSIN" "MIT.EDU")) NIL NIL
      "<B27397-0100000@cac.washington.edu>")
       BODY ("TEXT" "PLAIN" ("CHARSET" "US-ASCII") NIL NIL "7BIT" 3028
       92))
S:    a003 OK FETCH completed
C:    a004 fetch 12 body[header]
S:    * 12 FETCH (BODY[HEADER] {342}
S:    Date: Wed, 17 Jul 1996 02:23:25 -0700 (PDT)
S:    From: Terry Gray <gray@cac.washington.edu>
S:    Subject: IMAP4rev1 WG mtg summary and minutes
S:    To: imap@cac.washington.edu
S:    cc: minutes@CNRI.Reston.VA.US, John Klensin <KLENSIN@MIT.EDU>
S:    Message-Id: <B27397-0100000@cac.washington.edu>
S:    MIME-Version: 1.0
S:    Content-Type: TEXT/PLAIN; CHARSET=US-ASCII
S:
S:    )
S:    a004 OK FETCH completed
C:    a005 store 12 +flags \deleted
S:    * 12 FETCH (FLAGS (\Seen \Deleted))
S:    a005 OK +FLAGS completed
C:    a006 logout
S:    * BYE IMAP4rev1 server terminating connection
S:    a006 OK LOGOUT completed

9.      Formal Syntax

   The following syntax specification uses the Augmented Backus-Naur
   Form (ABNF) notation as specified in [ABNF].

   In the case of alternative or optional rules in which a later rule
   overlaps an earlier rule, the rule which is listed earlier MUST take
   priority.  For example, "\Seen" when parsed as a flag is the \Seen
   flag name and not a flag-extension, even though "\Seen" can be parsed
   as a flag-extension.  Some, but not all, instances of this rule are
   noted below.

        Note: [ABNF] rules MUST be followed strictly; in
        particular:

        (1) Except as noted otherwise, all alphabetic characters
        are case-insensitive.  The use of upper or lower case
        characters to define token strings is for editorial clarity
        only.  Implementations MUST accept these strings in a
        case-insensitive fashion.

        (2) In all cases, SP refers to exactly one space.  It is
        NOT permitted to substitute TAB, insert additional spaces,
        or otherwise treat SP as being equivalent to LWSP.

        (3) The ASCII NUL character, %x00, MUST NOT be used at any
        time.

address         = "(" addr-name SP addr-adl SP addr-mailbox SP
                  addr-host ")"

addr-adl        = nstring
                    ; Holds route from [RFC-2822] route-addr if
                    ; non-NIL

addr-host       = nstring
                    ; NIL indicates [RFC-2822] group syntax.
                    ; Otherwise, holds [RFC-2822] domain name

addr-mailbox    = nstring
                    ; NIL indicates end of [RFC-2822] group; if
                    ; non-NIL and addr-host is NIL, holds
                    ; [RFC-2822] group name.
                    ; Otherwise, holds [RFC-2822] local-part
                    ; after removing [RFC-2822] quoting

addr-name       = nstring
                    ; If non-NIL, holds phrase from [RFC-2822]
                    ; mailbox after removing [RFC-2822] quoting

append          = "APPEND" SP mailbox [SP flag-list] [SP date-time] SP
                  literal

astring         = 1*ASTRING-CHAR / string

ASTRING-CHAR   = ATOM-CHAR / resp-specials

atom            = 1*ATOM-CHAR

ATOM-CHAR       = <any CHAR except atom-specials>

atom-specials   = "(" / ")" / "{" / SP / CTL / list-wildcards /
                  quoted-specials / resp-specials

authenticate    = "AUTHENTICATE" SP auth-type *(CRLF base64)

auth-type       = atom
                    ; Defined by [SASL]

base64          = *(4base64-char) [base64-terminal]

base64-char     = ALPHA / DIGIT / "+" / "/"
                    ; Case-sensitive

base64-terminal = (2base64-char "==") / (3base64-char "=")

body            = "(" (body-type-1part / body-type-mpart) ")"

body-extension  = nstring / number /
                   "(" body-extension *(SP body-extension) ")"
                    ; Future expansion.  Client implementations
                    ; MUST accept body-extension fields.  Server
                    ; implementations MUST NOT generate
                    ; body-extension fields except as defined by
                    ; future standard or standards-track
                    ; revisions of this specification.

body-ext-1part  = body-fld-md5 [SP body-fld-dsp [SP body-fld-lang
                  [SP body-fld-loc *(SP body-extension)]]]
                    ; MUST NOT be returned on non-extensible
                    ; "BODY" fetch

body-ext-mpart  = body-fld-param [SP body-fld-dsp [SP body-fld-lang
                  [SP body-fld-loc *(SP body-extension)]]]
                    ; MUST NOT be returned on non-extensible
                    ; "BODY" fetch

body-fields     = body-fld-param SP body-fld-id SP body-fld-desc SP
                  body-fld-enc SP body-fld-octets

body-fld-desc   = nstring

body-fld-dsp    = "(" string SP body-fld-param ")" / nil

body-fld-enc    = (DQUOTE ("7BIT" / "8BIT" / "BINARY" / "BASE64"/
                  "QUOTED-PRINTABLE") DQUOTE) / string

body-fld-id     = nstring

body-fld-lang   = nstring / "(" string *(SP string) ")"

body-fld-loc    = nstring

body-fld-lines  = number

body-fld-md5    = nstring

body-fld-octets = number

body-fld-param  = "(" string SP string *(SP string SP string) ")" / nil

body-type-1part = (body-type-basic / body-type-msg / body-type-text)
                  [SP body-ext-1part]

body-type-basic = media-basic SP body-fields
                    ; MESSAGE subtype MUST NOT be "RFC822"

body-type-mpart = 1*body SP media-subtype
                  [SP body-ext-mpart]

body-type-msg   = media-message SP body-fields SP envelope
                  SP body SP body-fld-lines

body-type-text  = media-text SP body-fields SP body-fld-lines

capability      = ("AUTH=" auth-type) / atom
                    ; New capabilities MUST begin with "X" or be
                    ; registered with IANA as standard or
                    ; standards-track

capability-data = "CAPABILITY" *(SP capability) SP "IMAP4rev1"
                  *(SP capability)
                    ; Servers MUST implement the STARTTLS, AUTH=PLAIN,
                    ; and LOGINDISABLED capabilities
                    ; Servers which offer RFC 1730 compatibility MUST
                    ; list "IMAP4" as the first capability.

CHAR8           = %x01-ff
                    ; any OCTET except NUL, %x00

command         = tag SP (command-any / command-auth / command-nonauth /
                  command-select) CRLF
                    ; Modal based on state

command-any     = "CAPABILITY" / "LOGOUT" / "NOOP" / x-command
                    ; Valid in all states

command-auth    = append / create / delete / examine / list / lsub /
                  rename / select / status / subscribe / unsubscribe
                    ; Valid only in Authenticated or Selected state

command-nonauth = login / authenticate / "STARTTLS"
                    ; Valid only when in Not Authenticated state

command-select  = "CHECK" / "CLOSE" / "EXPUNGE" / copy / fetch / store /
                  uid / search
                    ; Valid only when in Selected state

continue-req    = "+" SP (resp-text / base64) CRLF

copy            = "COPY" SP sequence-set SP mailbox

create          = "CREATE" SP mailbox
                    ; Use of INBOX gives a NO error

date            = date-text / DQUOTE date-text DQUOTE

date-day        = 1*2DIGIT
                    ; Day of month

date-day-fixed  = (SP DIGIT) / 2DIGIT
                    ; Fixed-format version of date-day

date-month      = "Jan" / "Feb" / "Mar" / "Apr" / "May" / "Jun" /
                  "Jul" / "Aug" / "Sep" / "Oct" / "Nov" / "Dec"

date-text       = date-day "-" date-month "-" date-year

date-year       = 4DIGIT

date-time       = DQUOTE date-day-fixed "-" date-month "-" date-year
                  SP time SP zone DQUOTE

delete          = "DELETE" SP mailbox
                    ; Use of INBOX gives a NO error

digit-nz        = %x31-39
                    ; 1-9

envelope        = "(" env-date SP env-subject SP env-from SP
                  env-sender SP env-reply-to SP env-to SP env-cc SP
                  env-bcc SP env-in-reply-to SP env-message-id ")"

env-bcc         = "(" 1*address ")" / nil

env-cc          = "(" 1*address ")" / nil

env-date        = nstring

env-from        = "(" 1*address ")" / nil

env-in-reply-to = nstring

env-message-id  = nstring

env-reply-to    = "(" 1*address ")" / nil

env-sender      = "(" 1*address ")" / nil

env-subject     = nstring

env-to          = "(" 1*address ")" / nil

examine         = "EXAMINE" SP mailbox

fetch           = "FETCH" SP sequence-set SP ("ALL" / "FULL" / "FAST" /
                  fetch-att / "(" fetch-att *(SP fetch-att) ")")

fetch-att       = "ENVELOPE" / "FLAGS" / "INTERNALDATE" /
                  "RFC822" [".HEADER" / ".SIZE" / ".TEXT"] /
                  "BODY" ["STRUCTURE"] / "UID" /
                  "BODY" section ["<" number "." nz-number ">"] /
                  "BODY.PEEK" section ["<" number "." nz-number ">"]

flag            = "\Answered" / "\Flagged" / "\Deleted" /
                  "\Seen" / "\Draft" / flag-keyword / flag-extension
                    ; Does not include "\Recent"

flag-extension  = "\" atom
                    ; Future expansion.  Client implementations
                    ; MUST accept flag-extension flags.  Server
                    ; implementations MUST NOT generate
                    ; flag-extension flags except as defined by
                    ; future standard or standards-track
                    ; revisions of this specification.

flag-fetch      = flag / "\Recent"

flag-keyword    = atom

flag-list       = "(" [flag *(SP flag)] ")"

flag-perm       = flag / "\*"

greeting        = "*" SP (resp-cond-auth / resp-cond-bye) CRLF

header-fld-name = astring

header-list     = "(" header-fld-name *(SP header-fld-name) ")"

list            = "LIST" SP mailbox SP list-mailbox

list-mailbox    = 1*list-char / string

list-char       = ATOM-CHAR / list-wildcards / resp-specials

list-wildcards  = "%" / "*"

literal         = "{" number "}" CRLF *CHAR8
                    ; Number represents the number of CHAR8s

login           = "LOGIN" SP userid SP password

lsub            = "LSUB" SP mailbox SP list-mailbox

mailbox         = "INBOX" / astring
                    ; INBOX is case-insensitive.  All case variants of
                    ; INBOX (e.g., "iNbOx") MUST be interpreted as INBOX
                    ; not as an astring.  An astring which consists of
                    ; the case-insensitive sequence "I" "N" "B" "O" "X"
                    ; is considered to be INBOX and not an astring.
                    ;  Refer to section 5.1 for further
                    ; semantic details of mailbox names.

mailbox-data    =  "FLAGS" SP flag-list / "LIST" SP mailbox-list /
                   "LSUB" SP mailbox-list / "SEARCH" *(SP nz-number) /
                   "STATUS" SP mailbox SP "(" [status-att-list] ")" /
                   number SP "EXISTS" / number SP "RECENT"

mailbox-list    = "(" [mbx-list-flags] ")" SP
                   (DQUOTE QUOTED-CHAR DQUOTE / nil) SP mailbox

mbx-list-flags  = *(mbx-list-oflag SP) mbx-list-sflag
                  *(SP mbx-list-oflag) /
                  mbx-list-oflag *(SP mbx-list-oflag)

mbx-list-oflag  = "\Noinferiors" / flag-extension
                    ; Other flags; multiple possible per LIST response

mbx-list-sflag  = "\Noselect" / "\Marked" / "\Unmarked"
                    ; Selectability flags; only one per LIST response

media-basic     = ((DQUOTE ("APPLICATION" / "AUDIO" / "IMAGE" /
                  "MESSAGE" / "VIDEO") DQUOTE) / string) SP
                  media-subtype
                    ; Defined in [MIME-IMT]

media-message   = DQUOTE "MESSAGE" DQUOTE SP DQUOTE "RFC822" DQUOTE
                    ; Defined in [MIME-IMT]

media-subtype   = string
                    ; Defined in [MIME-IMT]

media-text      = DQUOTE "TEXT" DQUOTE SP media-subtype
                    ; Defined in [MIME-IMT]

message-data    = nz-number SP ("EXPUNGE" / ("FETCH" SP msg-att))

msg-att         = "(" (msg-att-dynamic / msg-att-static)
                   *(SP (msg-att-dynamic / msg-att-static)) ")"

msg-att-dynamic = "FLAGS" SP "(" [flag-fetch *(SP flag-fetch)] ")"
                    ; MAY change for a message

msg-att-static  = "ENVELOPE" SP envelope / "INTERNALDATE" SP date-time /
                  "RFC822" [".HEADER" / ".TEXT"] SP nstring /
                  "RFC822.SIZE" SP number /
                  "BODY" ["STRUCTURE"] SP body /
                  "BODY" section ["<" number ">"] SP nstring /
                  "UID" SP uniqueid
                    ; MUST NOT change for a message

nil             = "NIL"

nstring         = string / nil

number          = 1*DIGIT
                    ; Unsigned 32-bit integer
                    ; (0 <= n < 4,294,967,296)

nz-number       = digit-nz *DIGIT
                    ; Non-zero unsigned 32-bit integer
                    ; (0 < n < 4,294,967,296)

password        = astring

quoted          = DQUOTE *QUOTED-CHAR DQUOTE

QUOTED-CHAR     = <any TEXT-CHAR except quoted-specials> /
                  "\" quoted-specials

quoted-specials = DQUOTE / "\"

rename          = "RENAME" SP mailbox SP mailbox
                    ; Use of INBOX as a destination gives a NO error

response        = *(continue-req / response-data) response-done

response-data   = "*" SP (resp-cond-state / resp-cond-bye /
                  mailbox-data / message-data / capability-data) CRLF

response-done   = response-tagged / response-fatal

response-fatal  = "*" SP resp-cond-bye CRLF
                    ; Server closes connection immediately

response-tagged = tag SP resp-cond-state CRLF

resp-cond-auth  = ("OK" / "PREAUTH") SP resp-text
                    ; Authentication condition

resp-cond-bye   = "BYE" SP resp-text

resp-cond-state = ("OK" / "NO" / "BAD") SP resp-text
                    ; Status condition

resp-specials   = "]"

resp-text       = ["[" resp-text-code "]" SP] text

resp-text-code  = "ALERT" /
                  "BADCHARSET" [SP "(" astring *(SP astring) ")" ] /
                  capability-data / "PARSE" /
                  "PERMANENTFLAGS" SP "("
                  [flag-perm *(SP flag-perm)] ")" /
                  "READ-ONLY" / "READ-WRITE" / "TRYCREATE" /
                  "UIDNEXT" SP nz-number / "UIDVALIDITY" SP nz-number /
                  "UNSEEN" SP nz-number /
                  atom [SP 1*<any TEXT-CHAR except "]">]

search          = "SEARCH" [SP "CHARSET" SP astring] 1*(SP search-key)
                    ; CHARSET argument to MUST be registered with IANA

search-key      = "ALL" / "ANSWERED" / "BCC" SP astring /
                  "BEFORE" SP date / "BODY" SP astring /
                  "CC" SP astring / "DELETED" / "FLAGGED" /
                  "FROM" SP astring / "KEYWORD" SP flag-keyword /
                  "NEW" / "OLD" / "ON" SP date / "RECENT" / "SEEN" /
                  "SINCE" SP date / "SUBJECT" SP astring /
                  "TEXT" SP astring / "TO" SP astring /
                  "UNANSWERED" / "UNDELETED" / "UNFLAGGED" /
                  "UNKEYWORD" SP flag-keyword / "UNSEEN" /
                    ; Above this line were in [IMAP2]
                  "DRAFT" / "HEADER" SP header-fld-name SP astring /
                  "LARGER" SP number / "NOT" SP search-key /
                  "OR" SP search-key SP search-key /
                  "SENTBEFORE" SP date / "SENTON" SP date /
                  "SENTSINCE" SP date / "SMALLER" SP number /
                  "UID" SP sequence-set / "UNDRAFT" / sequence-set /
                  "(" search-key *(SP search-key) ")"

section         = "[" [section-spec] "]"

section-msgtext = "HEADER" / "HEADER.FIELDS" [".NOT"] SP header-list /
                  "TEXT"
                    ; top-level or MESSAGE/RFC822 part

section-part    = nz-number *("." nz-number)
                    ; body part nesting

section-spec    = section-msgtext / (section-part ["." section-text])

section-text    = section-msgtext / "MIME"
                    ; text other than actual body part (headers, etc.)

select          = "SELECT" SP mailbox

seq-number      = nz-number / "*"
                    ; message sequence number (COPY, FETCH, STORE
                    ; commands) or unique identifier (UID COPY,
                    ; UID FETCH, UID STORE commands).
                    ; * represents the largest number in use.  In
                    ; the case of message sequence numbers, it is
                    ; the number of messages in a non-empty mailbox.
                    ; In the case of unique identifiers, it is the
                    ; unique identifier of the last message in the
                    ; mailbox or, if the mailbox is empty, the
                    ; mailbox's current UIDNEXT value.
                    ; The server should respond with a tagged BAD
                    ; response to a command that uses a message
                    ; sequence number greater than the number of
                    ; messages in the selected mailbox.  This
                    ; includes "*" if the selected mailbox is empty.

seq-range       = seq-number ":" seq-number
                    ; two seq-number values and all values between
                    ; these two regardless of order.
                    ; Example: 2:4 and 4:2 are equivalent and indicate
                    ; values 2, 3, and 4.
                    ; Example: a unique identifier sequence range of
                    ; 3291:* includes the UID of the last message in
                    ; the mailbox, even if that value is less than 3291.

sequence-set    = (seq-number / seq-range) *("," sequence-set)
                    ; set of seq-number values, regardless of order.
                    ; Servers MAY coalesce overlaps and/or execute the
                    ; sequence in any order.
                    ; Example: a message sequence number set of
                    ; 2,4:7,9,12:* for a mailbox with 15 messages is
                    ; equivalent to 2,4,5,6,7,9,12,13,14,15
                    ; Example: a message sequence number set of *:4,5:7
                    ; for a mailbox with 10 messages is equivalent to
                    ; 10,9,8,7,6,5,4,5,6,7 and MAY be reordered and
                    ; overlap coalesced to be 4,5,6,7,8,9,10.

status          = "STATUS" SP mailbox SP
                  "(" status-att *(SP status-att) ")"

status-att      = "MESSAGES" / "RECENT" / "UIDNEXT" / "UIDVALIDITY" /
                  "UNSEEN"

status-att-list =  status-att SP number *(SP status-att SP number)

store           = "STORE" SP sequence-set SP store-att-flags

store-att-flags = (["+" / "-"] "FLAGS" [".SILENT"]) SP
                  (flag-list / (flag *(SP flag)))

string          = quoted / literal

subscribe       = "SUBSCRIBE" SP mailbox

tag             = 1*<any ASTRING-CHAR except "+">

text            = 1*TEXT-CHAR

TEXT-CHAR       = <any CHAR except CR and LF>

time            = 2DIGIT ":" 2DIGIT ":" 2DIGIT
                    ; Hours minutes seconds

uid             = "UID" SP (copy / fetch / search / store)
                    ; Unique identifiers used instead of message
                    ; sequence numbers

uniqueid        = nz-number
                    ; Strictly ascending

unsubscribe     = "UNSUBSCRIBE" SP mailbox

userid          = astring

x-command       = "X" atom <experimental command arguments>

zone            = ("+" / "-") 4DIGIT
                    ; Signed four-digit value of hhmm representing
                    ; hours and minutes east of Greenwich (that is,
                    ; the amount that the given time differs from
                    ; Universal Time).  Subtracting the timezone
                    ; from the given time will give the UT form.
                    ; The Universal Time zone is "+0000".

10.     Author's Note

   This document is a revision or rewrite of earlier documents, and
   supercedes the protocol specification in those documents: RFC 2060,
   RFC 1730, unpublished IMAP2bis.TXT document, RFC 1176, and RFC 1064.

11.     Security Considerations

   IMAP4rev1 protocol transactions, including electronic mail data, are
   sent in the clear over the network unless protection from snooping is
   negotiated.  This can be accomplished either by the use of STARTTLS,
   negotiated privacy protection in the AUTHENTICATE command, or some
   other protection mechanism.

11.1.   STARTTLS Security Considerations

   The specification of the STARTTLS command and LOGINDISABLED
   capability in this document replaces that in [IMAP-TLS].  [IMAP-TLS]
   remains normative for the PLAIN [SASL] authenticator.

   IMAP client and server implementations MUST implement the
   TLS_RSA_WITH_RC4_128_MD5 [TLS] cipher suite, and SHOULD implement the
   TLS_DHE_DSS_WITH_3DES_EDE_CBC_SHA [TLS] cipher suite.  This is
   important as it assures that any two compliant implementations can be
   configured to interoperate.  All other cipher suites are OPTIONAL.
   Note that this is a change from section 2.1 of [IMAP-TLS].

   During the [TLS] negotiation, the client MUST check its understanding
   of the server hostname against the server's identity as presented in
   the server Certificate message, in order to prevent man-in-the-middle
   attacks.  If the match fails, the client SHOULD either ask for
   explicit user confirmation, or terminate the connection and indicate
   that the server's identity is suspect.  Matching is performed
   according to these rules:

        The client MUST use the server hostname it used to open the
        connection as the value to compare against the server name
        as expressed in the server certificate.  The client MUST
        NOT use any form of the server hostname derived from an
        insecure remote source (e.g., insecure DNS lookup).  CNAME
        canonicalization is not done.

        If a subjectAltName extension of type dNSName is present in
        the certificate, it SHOULD be used as the source of the
        server's identity.

        Matching is case-insensitive.

        A "*" wildcard character MAY be used as the left-most name
        component in the certificate.  For example, *.example.com
        would match a.example.com, foo.example.com, etc. but would
        not match example.com.

        If the certificate contains multiple names (e.g., more than
        one dNSName field), then a match with any one of the fields
        is considered acceptable.

   Both the client and server MUST check the result of the STARTTLS
   command and subsequent [TLS] negotiation to see whether acceptable
   authentication or privacy was achieved.

11.2.   Other Security Considerations

   A server error message for an AUTHENTICATE command which fails due to
   invalid credentials SHOULD NOT detail why the credentials are
   invalid.

   Use of the LOGIN command sends passwords in the clear.  This can be
   avoided by using the AUTHENTICATE command with a [SASL] mechanism
   that does not use plaintext passwords, by first negotiating
   encryption via STARTTLS or some other protection mechanism.

   A server implementation MUST implement a configuration that, at the
   time of authentication, requires:
      (1) The STARTTLS command has been negotiated.
   OR
      (2) Some other mechanism that protects the session from password
      snooping has been provided.
   OR
      (3) The following measures are in place:
         (a) The LOGINDISABLED capability is advertised, and [SASL]
         mechanisms (such as PLAIN) using plaintext passwords are NOT
         advertised in the CAPABILITY list.
      AND
         (b) The LOGIN command returns an error even if the password is
         correct.
      AND
         (c) The AUTHENTICATE command returns an error with all [SASL]
         mechanisms that use plaintext passwords, even if the password
         is correct.

   A server error message for a failing LOGIN command SHOULD NOT specify
   that the user name, as opposed to the password, is invalid.

   A server SHOULD have mechanisms in place to limit or delay failed
   AUTHENTICATE/LOGIN attempts.

   Additional security considerations are discussed in the section
   discussing the AUTHENTICATE and LOGIN commands.

12.     IANA Considerations

   IMAP4 capabilities are registered by publishing a standards track or
   IESG approved experimental RFC.  The registry is currently located
   at:

        http://www.iana.org/assignments/imap4-capabilities

   As this specification revises the STARTTLS and LOGINDISABLED
   extensions previously defined in [IMAP-TLS], the registry will be
   updated accordingly.

Appendices

A.      Normative References

   The following documents contain definitions or specifications that
   are necessary to understand this document properly:
   [ABNF]                Crocker, D. and P. Overell, "Augmented BNF for
                         Syntax Specifications: ABNF", RFC 2234,
                         November 1997.

   [ANONYMOUS]           Newman, C., "Anonymous SASL Mechanism", RFC
                         2245, November 1997.

   [CHARSET]             Freed, N. and J. Postel, "IANA Character Set
                         Registration Procedures", RFC 2978, October
                         2000.

   [DIGEST-MD5]          Leach, P. and C. Newman, "Using Digest
                         Authentication as a SASL Mechanism", RFC 2831,
                         May 2000.

   [DISPOSITION]         Troost, R., Dorner, S. and K. Moore,
                         "Communicating Presentation Information in
                         Internet Messages: The Content-Disposition
                         Header", RFC 2183, August 1997.

   [IMAP-TLS]            Newman, C., "Using TLS with IMAP, POP3 and
                         ACAP", RFC 2595, June 1999.

   [KEYWORDS]            Bradner, S., "Key words for use in RFCs to
                         Indicate Requirement Levels", BCP 14, RFC 2119,
                         March 1997.

   [LANGUAGE-TAGS]       Alvestrand, H., "Tags for the Identification of
                         Languages", BCP 47, RFC 3066, January 2001.

   [LOCATION]            Palme, J., Hopmann, A. and N. Shelness, "MIME
                         Encapsulation of Aggregate Documents, such as
                         HTML (MHTML)", RFC 2557, March 1999.

   [MD5]                 Myers, J. and M. Rose, "The Content-MD5 Header
                         Field", RFC 1864, October 1995.

   [MIME-HDRS]           Moore, K., "MIME (Multipurpose Internet Mail
                         Extensions) Part Three: Message Header
                         Extensions for Non-ASCII Text", RFC 2047,
                         November 1996.

   [MIME-IMB]            Freed, N. and N. Borenstein, "MIME
                         (Multipurpose Internet Mail Extensions) Part
                         One: Format of Internet Message Bodies", RFC
                         2045, November 1996.

   [MIME-IMT]            Freed, N. and N. Borenstein, "MIME
                         (Multipurpose Internet Mail Extensions) Part
                         Two: Media Types", RFC 2046, November 1996.

   [RFC-2822]            Resnick, P., "Internet Message Format", RFC
                         2822, April 2001.

   [SASL]                Myers, J., "Simple Authentication and Security
                         Layer (SASL)", RFC 2222, October 1997.

   [TLS]                 Dierks, T. and C. Allen, "The TLS Protocol
                         Version 1.0", RFC 2246, January 1999.

   [UTF-7]               Goldsmith, D. and M. Davis, "UTF-7: A Mail-Safe
                         Transformation Format of Unicode", RFC 2152,
                         May 1997.

   The following documents describe quality-of-implementation issues
   that should be carefully considered when implementing this protocol:

   [IMAP-IMPLEMENTATION] Leiba, B., "IMAP Implementation
                         Recommendations", RFC 2683, September 1999.

   [IMAP-MULTIACCESS]    Gahrns, M., "IMAP4 Multi-Accessed Mailbox
                         Practice", RFC 2180, July 1997.

A.1     Informative References

   The following documents describe related protocols:

   [IMAP-DISC]           Austein, R., "Synchronization Operations for
                         Disconnected IMAP4 Clients", Work in Progress.

   [IMAP-MODEL]          Crispin, M., "Distributed Electronic Mail
                         Models in IMAP4", RFC 1733, December 1994.

   [ACAP]                Newman, C. and J. Myers, "ACAP -- Application
                         Configuration Access Protocol", RFC 2244,
                         November 1997.

   [SMTP]                Klensin, J., "Simple Mail Transfer Protocol",
                         STD 10, RFC 2821, April 2001.

   The following documents are historical or describe historical aspects
   of this protocol:

   [IMAP-COMPAT]         Crispin, M., "IMAP4 Compatibility with
                         IMAP2bis", RFC 2061, December 1996.

   [IMAP-HISTORICAL]     Crispin, M., "IMAP4 Compatibility with IMAP2
                         and IMAP2bis", RFC 1732, December 1994.

   [IMAP-OBSOLETE]       Crispin, M., "Internet Message Access Protocol
                         - Obsolete Syntax", RFC 2062, December 1996.

   [IMAP2]               Crispin, M., "Interactive Mail Access Protocol
                         - Version 2", RFC 1176, August 1990.

   [RFC-822]             Crocker, D., "Standard for the Format of ARPA
                         Internet Text Messages", STD 11, RFC 822,
                         August 1982.

   [RFC-821]             Postel, J., "Simple Mail Transfer Protocol",
                         STD 10, RFC 821, August 1982.

B.      Changes from RFC 2060

   1) Clarify description of unique identifiers and their semantics.

   2) Fix the SELECT description to clarify that UIDVALIDITY is required
   in the SELECT and EXAMINE responses.

   3) Added an example of a failing search.

   4) Correct store-att-flags: "#flag" should be "1#flag".

   5) Made search and section rules clearer.

   6) Correct the STORE example.

   7) Correct "BASE645" misspelling.

   8) Remove extraneous close parenthesis in example of two-part message
   with text and BASE64 attachment.

   9) Remove obsolete "MAILBOX" response from mailbox-data.

   10) A spurious "<" in the rule for mailbox-data was removed.

   11) Add CRLF to continue-req.

   12) Specifically exclude "]" from the atom in resp-text-code.

   13) Clarify that clients and servers should adhere strictly to the
   protocol syntax.

   14) Emphasize in 5.2 that EXISTS can not be used to shrink a mailbox.

   15) Add NEWNAME to resp-text-code.

   16) Clarify that the empty string, not NIL, is used as arguments to
   LIST.

   17) Clarify that NIL can be returned as a hierarchy delimiter for the
   empty string mailbox name argument if the mailbox namespace is flat.

   18) Clarify that addr-mailbox and addr-name have RFC-2822 quoting
   removed.

   19) Update UTF-7 reference.

   20) Fix example in 6.3.11.

   21) Clarify that non-existent UIDs are ignored.

   22) Update DISPOSITION reference.

   23) Expand state diagram.

   24) Clarify that partial fetch responses are only returned in
   response to a partial fetch command.

   25) Add UIDNEXT response code.  Correct UIDVALIDITY definition
   reference.

   26) Further clarification of "can" vs. "MAY".

   27) Reference RFC-2119.

   28) Clarify that superfluous shifts are not permitted in modified
   UTF-7.

   29) Clarify that there are no implicit shifts in modified UTF-7.

   30) Clarify that "INBOX" in a mailbox name is always INBOX, even if
   it is given as a string.

   31) Add missing open parenthesis in media-basic grammar rule.

   32) Correct attribute syntax in mailbox-data.

   33) Add UIDNEXT to EXAMINE responses.

   34) Clarify UNSEEN, PERMANENTFLAGS, UIDVALIDITY, and UIDNEXT
   responses in SELECT and EXAMINE.  They are required now, but weren't
   in older versions.

   35) Update references with RFC numbers.

   36) Flush text-mime2.

   37) Clarify that modified UTF-7 names must be case-sensitive and that
   violating the convention should be avoided.

   38) Correct UID FETCH example.

   39) Clarify UID FETCH, UID STORE, and UID SEARCH vs. untagged EXPUNGE
   responses.

   40) Clarify the use of the word "convention".

   41) Clarify that a command is not "in progress" until it has been
   fully received (specifically, that a command is not "in progress"
   during command continuation negotiation).

   42) Clarify envelope defaulting.

   43) Clarify that SP means one and only one space character.

   44) Forbid silly states in LIST response.

   45) Clarify that the ENVELOPE, INTERNALDATE, RFC822*, BODY*, and UID
   for a message is static.

   46) Add BADCHARSET response code.

   47) Update formal syntax to [ABNF] conventions.

   48) Clarify trailing hierarchy delimiter in CREATE semantics.

   49) Clarify that the "blank line" is the [RFC-2822] delimiting blank
   line.

   50) Clarify that RENAME should also create hierarchy as needed for
   the command to complete.

   51) Fix body-ext-mpart to not require language if disposition
   present.

   52) Clarify the RFC822.HEADER response.

   53) Correct missing space after charset astring in search.

   54) Correct missing quote for BADCHARSET in resp-text-code.

   55) Clarify that ALL, FAST, and FULL preclude any other data items
   appearing.

   56) Clarify semantics of reference argument in LIST.

   57) Clarify that a null string for SEARCH HEADER X-FOO means any
   message with a header line with a field-name of X-FOO regardless of
   the text of the header.

   58) Specifically reserve 8-bit mailbox names for future use as UTF-8.

   59) It is not an error for the client to store a flag that is not in
   the PERMANENTFLAGS list; however, the server will either ignore the
   change or make the change in the session only.

   60) Correct/clarify the text regarding superfluous shifts.

   61) Correct typographic errors in the "Changes" section.

   62) Clarify that STATUS must not be used to check for new messages in
   the selected mailbox

   63) Clarify LSUB behavior with "%" wildcard.

   64) Change AUTHORIZATION to AUTHENTICATE in section 7.5.

   65) Clarify description of multipart body type.

   66) Clarify that STORE FLAGS does not affect \Recent.

   67) Change "west" to "east" in description of timezone.

   68) Clarify that commands which break command pipelining must wait
   for a completion result response.

   69) Clarify that EXAMINE does not affect \Recent.

   70) Make description of MIME structure consistent.

   71) Clarify that date searches disregard the time and timezone of the
   INTERNALDATE or Date: header.  In other words, "ON 13-APR-2000" means
   messages with an INTERNALDATE text which starts with "13-APR-2000",
   even if timezone differential from the local timezone is sufficient
   to move that INTERNALDATE into the previous or next day.

   72) Clarify that the header fetches don't add a blank line if one
   isn't in the [RFC-2822] message.

   73) Clarify (in discussion of UIDs) that messages are immutable.

   74) Add an example of CHARSET searching.

   75) Clarify in SEARCH that keywords are a type of flag.

   76) Clarify the mandatory nature of the SELECT data responses.

   77) Add optional CAPABILITY response code in the initial OK or
   PREAUTH.

   78) Add note that server can send an untagged CAPABILITY command as
   part of the responses to AUTHENTICATE and LOGIN.

   79) Remove statement about it being unnecessary to issue a CAPABILITY
   command more than once in a connection.  That statement is no longer
   true.

   80) Clarify that untagged EXPUNGE decrements the number of messages
   in the mailbox.

   81) Fix definition of "body" (concatenation has tighter binding than
   alternation).

   82) Add a new "Special Notes to Implementors" section with reference
   to [IMAP-IMPLEMENTATION].

   83) Clarify that an untagged CAPABILITY response to an AUTHENTICATE
   command should only be done if a security layer was not negotiated.

   84) Change the definition of atom to exclude "]".  Update astring to
   include "]" for compatibility with the past.  Remove resp-text-atom.

   85) Remove NEWNAME.  It can't work because mailbox names can be
   literals and can include "]".  Functionality can be addressed via
   referrals.

   86) Move modified UTF-7 rationale in order to have more logical
   paragraph flow.

   87) Clarify UID uniqueness guarantees with the use of MUST.

   88) Note that clients should read response data until the connection
   is closed instead of immediately closing on a BYE.

   89) Change RFC-822 references to RFC-2822.

   90) Clarify that RFC-2822 should be followed instead of RFC-822.

   91) Change recommendation of optional automatic capabilities in LOGIN
   and AUTHENTICATE to use the CAPABILITY response code in the tagged
   OK.  This is more interoperable than an unsolicited untagged
   CAPABILITY response.

   92) STARTTLS and AUTH=PLAIN are mandatory to implement; add
   recommendations for other [SASL] mechanisms.

   93) Clarify that a "connection" (as opposed to "server" or "command")
   is in one of the four states.

   94) Clarify that a failed or rejected command does not change state.

   95) Split references between normative and informative.

   96) Discuss authentication failure issues in security section.

   97) Clarify that a data item is not necessarily of only one data
   type.

   98) Clarify that sequence ranges are independent of order.

   99) Change an example to clarify that superfluous shifts in
   Modified-UTF7 can not be fixed just by omitting the shift.  The
   entire string must be recalculated.

   100) Change Envelope Structure definition since [RFC-2822] uses
   "envelope" to refer to the [SMTP] envelope and not the envelope data
   that appears in the [RFC-2822] header.

   101) Expand on RFC822.HEADER response data vs. BODY[HEADER].

   102) Clarify Logout state semantics, change ASCII art.

   103) Security changes to comply with IESG requirements.

   104) Add definition for body URI.

   105) Break sequence range definition into three rules, with rewritten
   descriptions for each.

   106) Move STARTTLS and LOGINDISABLED here from [IMAP-TLS].

   107) Add IANA Considerations section.

   108) Clarify valid client assumptions for new message UIDs vs.
   UIDNEXT.

   109) Clarify that changes to permanentflags affect concurrent
   sessions as well as subsequent sessions.

   110) Clarify that authenticated state can be entered by the CLOSE
   command.

   111) Emphasize that SELECT and EXAMINE are the exceptions to the rule
   that a failing command does not change state.

   112) Clarify that newly-appended messages have the Recent flag set.

   113) Clarify that newly-copied messages SHOULD have the Recent flag
   set.

   114) Clarify that UID commands always return the UID in FETCH
   responses.

C.      Key Word Index

       +FLAGS <flag list> (store command data item) ...............   59
       +FLAGS.SILENT <flag list> (store command data item) ........   59
       -FLAGS <flag list> (store command data item) ...............   59
       -FLAGS.SILENT <flag list> (store command data item) ........   59
       ALERT (response code) ......................................   64
       ALL (fetch item) ...........................................   55
       ALL (search key) ...........................................   50
       ANSWERED (search key) ......................................   50
       APPEND (command) ...........................................   45
       AUTHENTICATE (command) .....................................   27
       BAD (response) .............................................   66
       BADCHARSET (response code) .................................   64
       BCC <string> (search key) ..................................   51
       BEFORE <date> (search key) .................................   51
       BODY (fetch item) ..........................................   55
       BODY (fetch result) ........................................   73
       BODY <string> (search key) .................................   51

       BODY.PEEK[<section>]<<partial>> (fetch item) ...............   57
       BODYSTRUCTURE (fetch item) .................................   57
       BODYSTRUCTURE (fetch result) ...............................   74
       BODY[<section>]<<origin octet>> (fetch result) .............   74
       BODY[<section>]<<partial>> (fetch item) ....................   55
       BYE (response) .............................................   67
       Body Structure (message attribute) .........................   12
       CAPABILITY (command) .......................................   24
       CAPABILITY (response code) .................................   64
       CAPABILITY (response) ......................................   68
       CC <string> (search key) ...................................   51
       CHECK (command) ............................................   47
       CLOSE (command) ............................................   48
       COPY (command) .............................................   59
       CREATE (command) ...........................................   34
       DELETE (command) ...........................................   35
       DELETED (search key) .......................................   51
       DRAFT (search key) .........................................   51
       ENVELOPE (fetch item) ......................................   57
       ENVELOPE (fetch result) ....................................   77
       EXAMINE (command) ..........................................   33
       EXISTS (response) ..........................................   71
       EXPUNGE (command) ..........................................   48
       EXPUNGE (response) .........................................   72
       Envelope Structure (message attribute) .....................   12
       FAST (fetch item) ..........................................   55
       FETCH (command) ............................................   54
       FETCH (response) ...........................................   73
       FLAGGED (search key) .......................................   51
       FLAGS (fetch item) .........................................   57
       FLAGS (fetch result) .......................................   78
       FLAGS (response) ...........................................   71
       FLAGS <flag list> (store command data item) ................   59
       FLAGS.SILENT <flag list> (store command data item) .........   59
       FROM <string> (search key) .................................   51
       FULL (fetch item) ..........................................   55
       Flags (message attribute) ..................................   11
       HEADER (part specifier) ....................................   55
       HEADER <field-name> <string> (search key) ..................   51
       HEADER.FIELDS <header-list> (part specifier) ...............   55
       HEADER.FIELDS.NOT <header-list> (part specifier) ...........   55
       INTERNALDATE (fetch item) ..................................   57
       INTERNALDATE (fetch result) ................................   78
       Internal Date (message attribute) ..........................   12
       KEYWORD <flag> (search key) ................................   51
       Keyword (type of flag) .....................................   11
       LARGER <n> (search key) ....................................   51
       LIST (command) .............................................   40

       LIST (response) ............................................   69
       LOGIN (command) ............................................   30
       LOGOUT (command) ...........................................   25
       LSUB (command) .............................................   43
       LSUB (response) ............................................   70
       MAY (specification requirement term) .......................    4
       MESSAGES (status item) .....................................   45
       MIME (part specifier) ......................................   56
       MUST (specification requirement term) ......................    4
       MUST NOT (specification requirement term) ..................    4
       Message Sequence Number (message attribute) ................   10
       NEW (search key) ...........................................   51
       NO (response) ..............................................   66
       NOOP (command) .............................................   25
       NOT <search-key> (search key) ..............................   52
       OK (response) ..............................................   65
       OLD (search key) ...........................................   52
       ON <date> (search key) .....................................   52
       OPTIONAL (specification requirement term) ..................    4
       OR <search-key1> <search-key2> (search key) ................   52
       PARSE (response code) ......................................   64
       PERMANENTFLAGS (response code) .............................   64
       PREAUTH (response) .........................................   67
       Permanent Flag (class of flag) .............................   12
       READ-ONLY (response code) ..................................   65
       READ-WRITE (response code) .................................   65
       RECENT (response) ..........................................   72
       RECENT (search key) ........................................   52
       RECENT (status item) .......................................   45
       RENAME (command) ...........................................   37
       REQUIRED (specification requirement term) ..................    4
       RFC822 (fetch item) ........................................   57
       RFC822 (fetch result) ......................................   78
       RFC822.HEADER (fetch item) .................................   57
       RFC822.HEADER (fetch result) ...............................   78
       RFC822.SIZE (fetch item) ...................................   57
       RFC822.SIZE (fetch result) .................................   78
       RFC822.TEXT (fetch item) ...................................   58
       RFC822.TEXT (fetch result) .................................   79
       SEARCH (command) ...........................................   49
       SEARCH (response) ..........................................   71
       SEEN (search key) ..........................................   52
       SELECT (command) ...........................................   31
       SENTBEFORE <date> (search key) .............................   52
       SENTON <date> (search key) .................................   52
       SENTSINCE <date> (search key) ..............................   52
       SHOULD (specification requirement term) ....................    4
       SHOULD NOT (specification requirement term) ................    4

       SINCE <date> (search key) ..................................   52
       SMALLER <n> (search key) ...................................   52
       STARTTLS (command) .........................................   27
       STATUS (command) ...........................................   44
       STATUS (response) ..........................................   70
       STORE (command) ............................................   58
       SUBJECT <string> (search key) ..............................   53
       SUBSCRIBE (command) ........................................   38
       Session Flag (class of flag) ...............................   12
       System Flag (type of flag) .................................   11
       TEXT (part specifier) ......................................   56
       TEXT <string> (search key) .................................   53
       TO <string> (search key) ...................................   53
       TRYCREATE (response code) ..................................   65
       UID (command) ..............................................   60
       UID (fetch item) ...........................................   58
       UID (fetch result) .........................................   79
       UID <sequence set> (search key) ............................   53
       UIDNEXT (response code) ....................................   65
       UIDNEXT (status item) ......................................   45
       UIDVALIDITY (response code) ................................   65
       UIDVALIDITY (status item) ..................................   45
       UNANSWERED (search key) ....................................   53
       UNDELETED (search key) .....................................   53
       UNDRAFT (search key) .......................................   53
       UNFLAGGED (search key) .....................................   53
       UNKEYWORD <flag> (search key) ..............................   53
       UNSEEN (response code) .....................................   65
       UNSEEN (search key) ........................................   53
       UNSEEN (status item) .......................................   45
       UNSUBSCRIBE (command) ......................................   39
       Unique Identifier (UID) (message attribute) ................    8
       X<atom> (command) ..........................................   62
       [RFC-2822] Size (message attribute) ........................   12
       \Answered (system flag) ....................................   11
       \Deleted (system flag) .....................................   11
       \Draft (system flag) .......................................   11
       \Flagged (system flag) .....................................   11
       \Marked (mailbox name attribute) ...........................   69
       \Noinferiors (mailbox name attribute) ......................   69
       \Noselect (mailbox name attribute) .........................   69
       \Recent (system flag) ......................................   11
       \Seen (system flag) ........................................   11
       \Unmarked (mailbox name attribute) .........................   69

Author's Address

   Mark R. Crispin
   Networks and Distributed Computing
   University of Washington
   4545 15th Avenue NE
   Seattle, WA  98105-4527

   Phone: (206) 543-5762

   EMail: MRC@CAC.Washington.EDU

Full Copyright Statement

   Copyright (C) The Internet Society (2003).  All Rights Reserved.

   This document and translations of it may be copied and furnished to
   others, and derivative works that comment on or otherwise explain it
   or assist in its implementation may be prepared, copied, published
   and distributed, in whole or in part, without restriction of any
   kind, provided that the above copyright notice and this paragraph are
   included on all such copies and derivative works.  However, this
   document itself may not be modified in any way, such as by removing
   the copyright notice or references to the Internet Society or other
   Internet organizations, except as needed for the purpose of
   developing Internet standards in which case the procedures for
   copyrights defined in the Internet Standards process must be
   followed, or as required to translate it into languages other than
   English.

   The limited permissions granted above are perpetual and will not be
   revoked by the Internet Society or its successors or assigns.  v This
   document and the information contained herein is provided on an "AS
   IS" basis and THE INTERNET SOCIETY AND THE INTERNET ENGINEERING TASK
   FORCE DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT
   LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL
   NOT INFRINGE ANY RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY
   OR FITNESS FOR A PARTICULAR PURPOSE.

Acknowledgement

   Funding for the RFC Editor function is currently provided by the
   Internet Society.

=============================================================================================
우와와 길다...imap4 라고 불리우는 Internet Message Access Protocol Version4rev1 에 관한 이야기이다.
사실 imap은 좀 복잡해서 연관되어진 rfc문서가 더 있고, 개발을 하다보면 조금은 생소한 UTF-7을 보게 되어진다.

마이그레이션 툴을 짤때도 맨 처음 UTF-7을 어디서 인코딩/디코딩을 해야할지 몰라서 한참동안 고민을 했었다. (답은 이 RFC문서를 참고로 하여 겨우 찾았다..=_=;;;;)

사용방법이 광범위하여 각 솔루션마다의 특징을 지니고 있지만, 기본 규약은 어딜가나 지켜야하는것이 RFC문서이다..이 RFC문서를 모른다면 개발해봤자 욕만 들어먹을뿐..;

저작자 표시 비영리 변경 금지
신고

'02 .administration > vi. Mail' 카테고리의 다른 글

mstone (mailclient test & stress tool)  (0) 2011.02.22
RFC 3501 (imap4 protocol)  (0) 2009.08.31
RFC 1939 (pop3 protocol)  (0) 2009.08.31
SMTP protocol (RFC 821) - (ESMTP는 RFC 1821)  (0) 2009.08.10
Posted by 신머루군
   RFC 821

                     SIMPLE MAIL TRANSFER PROTOCOL

                           Jonathan B. Postel

                              August 1982

                     Information Sciences Institute
                   University of Southern California
                           4676 Admiralty Way
                   Marina del Rey, California  90291

                             (213) 822-1511

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

                           TABLE OF CONTENTS

   1.  INTRODUCTION .................................................. 1

   2.  THE SMTP MODEL ................................................ 2

   3.  THE SMTP PROCEDURE ............................................ 4

      3.1.  Mail ..................................................... 4
      3.2.  Forwarding ............................................... 7
      3.3.  Verifying and Expanding .................................. 8
      3.4.  Sending and Mailing ..................................... 11
      3.5.  Opening and Closing ..................................... 13
      3.6.  Relaying ................................................ 14
      3.7.  Domains ................................................. 17
      3.8.  Changing Roles .......................................... 18

   4.  THE SMTP SPECIFICATIONS ...................................... 19

      4.1.  SMTP Commands ........................................... 19
      4.1.1.  Command Semantics ..................................... 19
      4.1.2.  Command Syntax ........................................ 27
      4.2.  SMTP Replies ............................................ 34
      4.2.1.  Reply Codes by Function Group ......................... 35
      4.2.2.  Reply Codes in Numeric Order .......................... 36
      4.3.  Sequencing of Commands and Replies ...................... 37
      4.4.  State Diagrams .......................................... 39
      4.5.  Details ................................................. 41
      4.5.1.  Minimum Implementation ................................ 41
      4.5.2.  Transparency .......................................... 41
      4.5.3.  Sizes ................................................. 42

   APPENDIX A:  TCP ................................................. 44
   APPENDIX B:  NCP ................................................. 45
   APPENDIX C:  NITS ................................................ 46
   APPENDIX D:  X.25 ................................................ 47
   APPENDIX E:  Theory of Reply Codes ............................... 48
   APPENDIX F:  Scenarios ........................................... 51

   GLOSSARY ......................................................... 64

   REFERENCES ....................................................... 67

Network Working Group                                          J. Postel
Request for Comments: DRAFT                                          ISI
Replaces: RFC 788, 780, 772                                  August 1982

                     SIMPLE MAIL TRANSFER PROTOCOL

1.  INTRODUCTION

   The objective of Simple Mail Transfer Protocol (SMTP) is to transfer
   mail reliably and efficiently.

   SMTP is independent of the particular transmission subsystem and
   requires only a reliable ordered data stream channel.  Appendices A,
   B, C, and D describe the use of SMTP with various transport services.
   A Glossary provides the definitions of terms as used in this
   document.

   An important feature of SMTP is its capability to relay mail across
   transport service environments.  A transport service provides an
   interprocess communication environment (IPCE).  An IPCE may cover one
   network, several networks, or a subset of a network.  It is important
   to realize that transport systems (or IPCEs) are not one-to-one with
   networks.  A process can communicate directly with another process
   through any mutually known IPCE.  Mail is an application or use of
   interprocess communication.  Mail can be communicated between
   processes in different IPCEs by relaying through a process connected
   to two (or more) IPCEs.  More specifically, mail can be relayed
   between hosts on different transport systems by a host on both
   transport systems.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

2.  THE SMTP MODEL

   The SMTP design is based on the following model of communication:  as
   the result of a user mail request, the sender-SMTP establishes a
   two-way transmission channel to a receiver-SMTP.  The receiver-SMTP
   may be either the ultimate destination or an intermediate.  SMTP
   commands are generated by the sender-SMTP and sent to the
   receiver-SMTP.  SMTP replies are sent from the receiver-SMTP to the
   sender-SMTP in response to the commands.

   Once the transmission channel is established, the SMTP-sender sends a
   MAIL command indicating the sender of the mail.  If the SMTP-receiver
   can accept mail it responds with an OK reply.  The SMTP-sender then
   sends a RCPT command identifying a recipient of the mail.  If the
   SMTP-receiver can accept mail for that recipient it responds with an
   OK reply; if not, it responds with a reply rejecting that recipient
   (but not the whole mail transaction).  The SMTP-sender and
   SMTP-receiver may negotiate several recipients.  When the recipients
   have been negotiated the SMTP-sender sends the mail data, terminating
   with a special sequence.  If the SMTP-receiver successfully processes
   the mail data it responds with an OK reply.  The dialog is purposely
   lock-step, one-at-a-time.

     -------------------------------------------------------------

               +----------+                +----------+
   +------+    |          |                |          |
   | User |<-->|          |      SMTP      |          |
   +------+    |  Sender- |Commands/Replies| Receiver-|
   +------+    |   SMTP   |<-------------->|    SMTP  |    +------+
   | File |<-->|          |    and Mail    |          |<-->| File |
   |System|    |          |                |          |    |System|
   +------+    +----------+                +----------+    +------+

                Sender-SMTP                Receiver-SMTP

                           Model for SMTP Use

                                Figure 1

     -------------------------------------------------------------

   The SMTP provides mechanisms for the transmission of mail; directly
   from the sending user's host to the receiving user's host when the

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   two host are connected to the same transport service, or via one or
   more relay SMTP-servers when the source and destination hosts are not
   connected to the same transport service.

   To be able to provide the relay capability the SMTP-server must be
   supplied with the name of the ultimate destination host as well as
   the destination mailbox name.

   The argument to the MAIL command is a reverse-path, which specifies
   who the mail is from.  The argument to the RCPT command is a
   forward-path, which specifies who the mail is to.  The forward-path
   is a source route, while the reverse-path is a return route (which
   may be used to return a message to the sender when an error occurs
   with a relayed message).

   When the same message is sent to multiple recipients the SMTP
   encourages the transmission of only one copy of the data for all the
   recipients at the same destination host.

   The mail commands and replies have a rigid syntax.  Replies also have
   a numeric code.  In the following, examples appear which use actual
   commands and replies.  The complete lists of commands and replies
   appears in Section 4 on specifications.

   Commands and replies are not case sensitive.  That is, a command or
   reply word may be upper case, lower case, or any mixture of upper and
   lower case.  Note that this is not true of mailbox user names.  For
   some hosts the user name is case sensitive, and SMTP implementations
   must take case to preserve the case of user names as they appear in
   mailbox arguments.  Host names are not case sensitive.

   Commands and replies are composed of characters from the ASCII
   character set [1].  When the transport service provides an 8-bit byte
   (octet) transmission channel, each 7-bit character is transmitted
   right justified in an octet with the high order bit cleared to zero.

   When specifying the general form of a command or reply, an argument
   (or special symbol) will be denoted by a meta-linguistic variable (or
   constant), for example, "<string>" or "<reverse-path>".  Here the
   angle brackets indicate these are meta-linguistic variables.
   However, some arguments use the angle brackets literally.  For
   example, an actual reverse-path is enclosed in angle brackets, i.e.,
   "<John.Smith@USC-ISI.ARPA>" is an instance of <reverse-path> (the
   angle brackets are actually transmitted in the command or reply).

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

3.  THE SMTP PROCEDURES

   This section presents the procedures used in SMTP in several parts.
   First comes the basic mail procedure defined as a mail transaction.
   Following this are descriptions of forwarding mail, verifying mailbox
   names and expanding mailing lists, sending to terminals instead of or
   in combination with mailboxes, and the opening and closing exchanges.
   At the end of this section are comments on relaying, a note on mail
   domains, and a discussion of changing roles.  Throughout this section
   are examples of partial command and reply sequences, several complete
   scenarios are presented in Appendix F.

   3.1.  MAIL

      There are three steps to SMTP mail transactions.  The transaction
      is started with a MAIL command which gives the sender
      identification.  A series of one or more RCPT commands follows
      giving the receiver information.  Then a DATA command gives the
      mail data.  And finally, the end of mail data indicator confirms
      the transaction.

         The first step in the procedure is the MAIL command.  The
         <reverse-path> contains the source mailbox.

            MAIL <SP> FROM:<reverse-path> <CRLF>

         This command tells the SMTP-receiver that a new mail
         transaction is starting and to reset all its state tables and
         buffers, including any recipients or mail data.  It gives the
         reverse-path which can be used to report errors.  If accepted,
         the receiver-SMTP returns a 250 OK reply.

         The <reverse-path> can contain more than just a mailbox.  The
         <reverse-path> is a reverse source routing list of hosts and
         source mailbox.  The first host in the <reverse-path> should be
         the host sending this command.

         The second step in the procedure is the RCPT command.

            RCPT <SP> TO:<forward-path> <CRLF>

         This command gives a forward-path identifying one recipient.
         If accepted, the receiver-SMTP returns a 250 OK reply, and
         stores the forward-path.  If the recipient is unknown the
         receiver-SMTP returns a 550 Failure reply.  This second step of
         the procedure can be repeated any number of times.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

         The <forward-path> can contain more than just a mailbox.  The
         <forward-path> is a source routing list of hosts and the
         destination mailbox.  The first host in the <forward-path>
         should be the host receiving this command.

         The third step in the procedure is the DATA command.

            DATA <CRLF>

         If accepted, the receiver-SMTP returns a 354 Intermediate reply
         and considers all succeeding lines to be the message text.
         When the end of text is received and stored the SMTP-receiver
         sends a 250 OK reply.

         Since the mail data is sent on the transmission channel the end
         of the mail data must be indicated so that the command and
         reply dialog can be resumed.  SMTP indicates the end of the
         mail data by sending a line containing only a period.  A
         transparency procedure is used to prevent this from interfering
         with the user's text (see Section 4.5.2).

            Please note that the mail data includes the memo header
            items such as Date, Subject, To, Cc, From [2].

         The end of mail data indicator also confirms the mail
         transaction and tells the receiver-SMTP to now process the
         stored recipients and mail data.  If accepted, the
         receiver-SMTP returns a 250 OK reply.  The DATA command should
         fail only if the mail transaction was incomplete (for example,
         no recipients), or if resources are not available.

      The above procedure is an example of a mail transaction.  These
      commands must be used only in the order discussed above.
      Example 1 (below) illustrates the use of these commands in a mail
      transaction.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      -------------------------------------------------------------

                     Example of the SMTP Procedure

         This SMTP example shows mail sent by Smith at host Alpha.ARPA,
         to Jones, Green, and Brown at host Beta.ARPA.  Here we assume
         that host Alpha contacts host Beta directly.

            S: MAIL FROM:<Smith@Alpha.ARPA>
            R: 250 OK

            S: RCPT TO:<Jones@Beta.ARPA>
            R: 250 OK

            S: RCPT TO:<Green@Beta.ARPA>
            R: 550 No such user here

            S: RCPT TO:<Brown@Beta.ARPA>
            R: 250 OK

            S: DATA
            R: 354 Start mail input; end with <CRLF>.<CRLF>
            S: Blah blah blah...
            S: ...etc. etc. etc.
            S: <CRLF>.<CRLF>
            R: 250 OK

         The mail has now been accepted for Jones and Brown.  Green did
         not have a mailbox at host Beta.

                               Example 1

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   3.2.  FORWARDING

      There are some cases where the destination information in the
      <forward-path> is incorrect, but the receiver-SMTP knows the
      correct destination.  In such cases, one of the following replies
      should be used to allow the sender to contact the correct
      destination.

         251 User not local; will forward to <forward-path>

            This reply indicates that the receiver-SMTP knows the user's
            mailbox is on another host and indicates the correct
            forward-path to use in the future.  Note that either the
            host or user or both may be different.  The receiver takes
            responsibility for delivering the message.

         551 User not local; please try <forward-path>

            This reply indicates that the receiver-SMTP knows the user's
            mailbox is on another host and indicates the correct
            forward-path to use.  Note that either the host or user or
            both may be different.  The receiver refuses to accept mail
            for this user, and the sender must either redirect the mail
            according to the information provided or return an error
            response to the originating user.

      Example 2 illustrates the use of these responses.

      -------------------------------------------------------------

                         Example of Forwarding

      Either

      S: RCPT TO:<Postel@USC-ISI.ARPA>
      R: 251 User not local; will forward to <Postel@USC-ISIF.ARPA>

      Or

      S: RCPT TO:<Paul@USC-ISIB.ARPA>
      R: 551 User not local; please try <Mockapetris@USC-ISIF.ARPA>

                               Example 2

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   3.3.  VERIFYING AND EXPANDING

      SMTP provides as additional features, commands to verify a user
      name or expand a mailing list.  This is done with the VRFY and
      EXPN commands, which have character string arguments.  For the
      VRFY command, the string is a user name, and the response may
      include the full name of the user and must include the mailbox of
      the user.  For the EXPN command, the string identifies a mailing
      list, and the multiline response may include the full name of the
      users and must give the mailboxes on the mailing list.

      "User name" is a fuzzy term and used purposely.  If a host
      implements the VRFY or EXPN commands then at least local mailboxes
      must be recognized as "user names".  If a host chooses to
      recognize other strings as "user names" that is allowed.

      In some hosts the distinction between a mailing list and an alias
      for a single mailbox is a bit fuzzy, since a common data structure
      may hold both types of entries, and it is possible to have mailing
      lists of one mailbox.  If a request is made to verify a mailing
      list a positive response can be given if on receipt of a message
      so addressed it will be delivered to everyone on the list,
      otherwise an error should be reported (e.g., "550 That is a
      mailing list, not a user").  If a request is made to expand a user
      name a positive response can be formed by returning a list
      containing one name, or an error can be reported (e.g., "550 That
      is a user name, not a mailing list").

      In the case of a multiline reply (normal for EXPN) exactly one
      mailbox is to be specified on each line of the reply.  In the case
      of an ambiguous request, for example, "VRFY Smith", where there
      are two Smith's the response must be "553 User ambiguous".

      The case of verifying a user name is straightforward as shown in
      example 3.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

      -------------------------------------------------------------

                    Example of Verifying a User Name

         Either

            S: VRFY Smith
            R: 250 Fred Smith <Smith@USC-ISIF.ARPA>

         Or

            S: VRFY Smith
            R: 251 User not local; will forward to <Smith@USC-ISIQ.ARPA>

         Or

            S: VRFY Jones
            R: 550 String does not match anything.

         Or

            S: VRFY Jones
            R: 551 User not local; please try <Jones@USC-ISIQ.ARPA>

         Or

            S: VRFY Gourzenkyinplatz
            R: 553 User ambiguous.

                               Example 3

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      The case of expanding a mailbox list requires a multiline reply as
      shown in example 4.

      -------------------------------------------------------------

                  Example of Expanding a Mailing List

         Either

            S: EXPN Example-People
            R: 250-Jon Postel <Postel@USC-ISIF.ARPA>
            R: 250-Fred Fonebone <Fonebone@USC-ISIQ.ARPA>
            R: 250-Sam Q. Smith <SQSmith@USC-ISIQ.ARPA>
            R: 250-Quincy Smith <@USC-ISIF.ARPA:Q-Smith@ISI-VAXA.ARPA>
            R: 250-<joe@foo-unix.ARPA>
            R: 250 <xyz@bar-unix.ARPA>

         Or

            S: EXPN Executive-Washroom-List
            R: 550 Access Denied to You.

                               Example 4

      -------------------------------------------------------------

      The character string arguments of the VRFY and EXPN commands
      cannot be further restricted due to the variety of implementations
      of the user name and mailbox list concepts.  On some systems it
      may be appropriate for the argument of the EXPN command to be a
      file name for a file containing a mailing list, but again there is
      a variety of file naming conventions in the Internet.

      The VRFY and EXPN commands are not included in the minimum
      implementation (Section 4.5.1), and are not required to work
      across relays when they are implemented.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   3.4.  SENDING AND MAILING

      The main purpose of SMTP is to deliver messages to user's
      mailboxes.  A very similar service provided by some hosts is to
      deliver messages to user's terminals (provided the user is active
      on the host).  The delivery to the user's mailbox is called
      "mailing", the delivery to the user's terminal is called
      "sending".  Because in many hosts the implementation of sending is
      nearly identical to the implementation of mailing these two
      functions are combined in SMTP.  However the sending commands are
      not included in the required minimum implementation
      (Section 4.5.1).  Users should have the ability to control the
      writing of messages on their terminals.  Most hosts permit the
      users to accept or refuse such messages.

      The following three command are defined to support the sending
      options.  These are used in the mail transaction instead of the
      MAIL command and inform the receiver-SMTP of the special semantics
      of this transaction:

         SEND <SP> FROM:<reverse-path> <CRLF>

            The SEND command requires that the mail data be delivered to
            the user's terminal.  If the user is not active (or not
            accepting terminal messages) on the host a 450 reply may
            returned to a RCPT command.  The mail transaction is
            successful if the message is delivered the terminal.

         SOML <SP> FROM:<reverse-path> <CRLF>

            The Send Or MaiL command requires that the mail data be
            delivered to the user's terminal if the user is active (and
            accepting terminal messages) on the host.  If the user is
            not active (or not accepting terminal messages) then the
            mail data is entered into the user's mailbox.  The mail
            transaction is successful if the message is delivered either
            to the terminal or the mailbox.

         SAML <SP> FROM:<reverse-path> <CRLF>

            The Send And MaiL command requires that the mail data be
            delivered to the user's terminal if the user is active (and
            accepting terminal messages) on the host.  In any case the
            mail data is entered into the user's mailbox.  The mail
            transaction is successful if the message is delivered the
            mailbox.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      The same reply codes that are used for the MAIL commands are used
      for these commands.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   3.5.  OPENING AND CLOSING

      At the time the transmission channel is opened there is an
      exchange to ensure that the hosts are communicating with the hosts
      they think they are.

      The following two commands are used in transmission channel
      opening and closing:

         HELO <SP> <domain> <CRLF>

         QUIT <CRLF>

      In the HELO command the host sending the command identifies
      itself; the command may be interpreted as saying "Hello, I am
      <domain>".

      -------------------------------------------------------------

                     Example of Connection Opening

         R: 220 BBN-UNIX.ARPA Simple Mail Transfer Service Ready
         S: HELO USC-ISIF.ARPA
         R: 250 BBN-UNIX.ARPA

                               Example 5

      -------------------------------------------------------------

      -------------------------------------------------------------

                     Example of Connection Closing

         S: QUIT
         R: 221 BBN-UNIX.ARPA Service closing transmission channel

                               Example 6

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   3.6.  RELAYING

      The forward-path may be a source route of the form
      "@ONE,@TWO:JOE@THREE", where ONE, TWO, and THREE are hosts.  This
      form is used to emphasize the distinction between an address and a
      route.  The mailbox is an absolute address, and the route is
      information about how to get there.  The two concepts should not
      be confused.

      Conceptually the elements of the forward-path are moved to the
      reverse-path as the message is relayed from one server-SMTP to
      another.  The reverse-path is a reverse source route, (i.e., a
      source route from the current location of the message to the
      originator of the message).  When a server-SMTP deletes its
      identifier from the forward-path and inserts it into the
      reverse-path, it must use the name it is known by in the
      environment it is sending into, not the environment the mail came
      from, in case the server-SMTP is known by different names in
      different environments.

      If when the message arrives at an SMTP the first element of the
      forward-path is not the identifier of that SMTP the element is not
      deleted from the forward-path and is used to determine the next
      SMTP to send the message to.  In any case, the SMTP adds its own
      identifier to the reverse-path.

      Using source routing the receiver-SMTP receives mail to be relayed
      to another server-SMTP  The receiver-SMTP may accept or reject the
      task of relaying the mail in the same way it accepts or rejects
      mail for a local user.  The receiver-SMTP transforms the command
      arguments by moving its own identifier from the forward-path to
      the beginning of the reverse-path.  The receiver-SMTP then becomes
      a sender-SMTP, establishes a transmission channel to the next SMTP
      in the forward-path, and sends it the mail.

      The first host in the reverse-path should be the host sending the
      SMTP commands, and the first host in the forward-path should be
      the host receiving the SMTP commands.

      Notice that the forward-path and reverse-path appear in the SMTP
      commands and replies, but not necessarily in the message.  That
      is, there is no need for these paths and especially this syntax to
      appear in the "To:" , "From:", "CC:", etc. fields of the message
      header.

      If a server-SMTP has accepted the task of relaying the mail and

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

      later finds that the forward-path is incorrect or that the mail
      cannot be delivered for whatever reason, then it must construct an
      "undeliverable mail" notification message and send it to the
      originator of the undeliverable mail (as indicated by the
      reverse-path).

      This notification message must be from the server-SMTP at this
      host.  Of course, server-SMTPs should not send notification
      messages about problems with notification messages.  One way to
      prevent loops in error reporting is to specify a null reverse-path
      in the MAIL command of a notification message.  When such a
      message is relayed it is permissible to leave the reverse-path
      null.  A MAIL command with a null reverse-path appears as follows:

         MAIL FROM:<>

      An undeliverable mail notification message is shown in example 7.
      This notification is in response to a message originated by JOE at
      HOSTW and sent via HOSTX to HOSTY with instructions to relay it on
      to HOSTZ.  What we see in the example is the transaction between
      HOSTY and HOSTX, which is the first step in the return of the
      notification message.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      -------------------------------------------------------------

            Example Undeliverable Mail Notification Message

         S: MAIL FROM:<>
         R: 250 ok
         S: RCPT TO:<@HOSTX.ARPA:JOE@HOSTW.ARPA>
         R: 250 ok
         S: DATA
         R: 354 send the mail data, end with .
         S: Date: 23 Oct 81 11:22:33
         S: From: SMTP@HOSTY.ARPA
         S: To: JOE@HOSTW.ARPA
         S: Subject: Mail System Problem
         S:
         S:   Sorry JOE, your message to SAM@HOSTZ.ARPA lost.
         S:   HOSTZ.ARPA said this:
         S:    "550 No Such User"
         S: .
         R: 250 ok

                               Example 7

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   3.7.  DOMAINS

      Domains are a recently introduced concept in the ARPA Internet
      mail system.  The use of domains changes the address space from a
      flat global space of simple character string host names to a
      hierarchically structured rooted tree of global addresses.  The
      host name is replaced by a domain and host designator which is a
      sequence of domain element strings separated by periods with the
      understanding that the domain elements are ordered from the most
      specific to the most general.

      For example, "USC-ISIF.ARPA", "Fred.Cambridge.UK", and
      "PC7.LCS.MIT.ARPA" might be host-and-domain identifiers.

      Whenever domain names are used in SMTP only the official names are
      used, the use of nicknames or aliases is not allowed.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   3.8.  CHANGING ROLES

      The TURN command may be used to reverse the roles of the two
      programs communicating over the transmission channel.

      If program-A is currently the sender-SMTP and it sends the TURN
      command and receives an ok reply (250) then program-A becomes the
      receiver-SMTP.

      If program-B is currently the receiver-SMTP and it receives the
      TURN command and sends an ok reply (250) then program-B becomes
      the sender-SMTP.

      To refuse to change roles the receiver sends the 502 reply.

      Please note that this command is optional.  It would not normally
      be used in situations where the transmission channel is TCP.
      However, when the cost of establishing the transmission channel is
      high, this command may be quite useful.  For example, this command
      may be useful in supporting be mail exchange using the public
      switched telephone system as a transmission channel, especially if
      some hosts poll other hosts for mail exchanges.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

4.  THE SMTP SPECIFICATIONS

   4.1.  SMTP COMMANDS

      4.1.1.  COMMAND SEMANTICS

         The SMTP commands define the mail transfer or the mail system
         function requested by the user.  SMTP commands are character
         strings terminated by <CRLF>.  The command codes themselves are
         alphabetic characters terminated by <SP> if parameters follow
         and <CRLF> otherwise.  The syntax of mailboxes must conform to
         receiver site conventions.  The SMTP commands are discussed
         below.  The SMTP replies are discussed in the Section 4.2.

         A mail transaction involves several data objects which are
         communicated as arguments to different commands.  The
         reverse-path is the argument of the MAIL command, the
         forward-path is the argument of the RCPT command, and the mail
         data is the argument of the DATA command.  These arguments or
         data objects must be transmitted and held pending the
         confirmation communicated by the end of mail data indication
         which finalizes the transaction.  The model for this is that
         distinct buffers are provided to hold the types of data
         objects, that is, there is a reverse-path buffer, a
         forward-path buffer, and a mail data buffer.  Specific commands
         cause information to be appended to a specific buffer, or cause
         one or more buffers to be cleared.

         HELLO (HELO)

            This command is used to identify the sender-SMTP to the
            receiver-SMTP.  The argument field contains the host name of
            the sender-SMTP.

            The receiver-SMTP identifies itself to the sender-SMTP in
            the connection greeting reply, and in the response to this
            command.

            This command and an OK reply to it confirm that both the
            sender-SMTP and the receiver-SMTP are in the initial state,
            that is, there is no transaction in progress and all state
            tables and buffers are cleared.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         MAIL (MAIL)

            This command is used to initiate a mail transaction in which
            the mail data is delivered to one or more mailboxes.  The
            argument field contains a reverse-path.

            The reverse-path consists of an optional list of hosts and
            the sender mailbox.  When the list of hosts is present, it
            is a "reverse" source route and indicates that the mail was
            relayed through each host on the list (the first host in the
            list was the most recent relay).  This list is used as a
            source route to return non-delivery notices to the sender.
            As each relay host adds itself to the beginning of the list,
            it must use its name as known in the IPCE to which it is
            relaying the mail rather than the IPCE from which the mail
            came (if they are different).  In some types of error
            reporting messages (for example, undeliverable mail
            notifications) the reverse-path may be null (see Example 7).

            This command clears the reverse-path buffer, the
            forward-path buffer, and the mail data buffer; and inserts
            the reverse-path information from this command into the
            reverse-path buffer.

         RECIPIENT (RCPT)

            This command is used to identify an individual recipient of
            the mail data; multiple recipients are specified by multiple
            use of this command.

            The forward-path consists of an optional list of hosts and a
            required destination mailbox.  When the list of hosts is
            present, it is a source route and indicates that the mail
            must be relayed to the next host on the list.  If the
            receiver-SMTP does not implement the relay function it may
            user the same reply it would for an unknown local user
            (550).

            When mail is relayed, the relay host must remove itself from
            the beginning forward-path and put itself at the beginning
            of the reverse-path.  When mail reaches its ultimate
            destination (the forward-path contains only a destination
            mailbox), the receiver-SMTP inserts it into the destination
            mailbox in accordance with its host mail conventions.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

               For example, mail received at relay host A with arguments

                  FROM:<USERX@HOSTY.ARPA>
                  TO:<@HOSTA.ARPA,@HOSTB.ARPA:USERC@HOSTD.ARPA>

               will be relayed on to host B with arguments

                  FROM:<@HOSTA.ARPA:USERX@HOSTY.ARPA>
                  TO:<@HOSTB.ARPA:USERC@HOSTD.ARPA>.

            This command causes its forward-path argument to be appended
            to the forward-path buffer.

         DATA (DATA)

            The receiver treats the lines following the command as mail
            data from the sender.  This command causes the mail data
            from this command to be appended to the mail data buffer.
            The mail data may contain any of the 128 ASCII character
            codes.

            The mail data is terminated by a line containing only a
            period, that is the character sequence "<CRLF>.<CRLF>" (see
            Section 4.5.2 on Transparency).  This is the end of mail
            data indication.

            The end of mail data indication requires that the receiver
            must now process the stored mail transaction information.
            This processing consumes the information in the reverse-path
            buffer, the forward-path buffer, and the mail data buffer,
            and on the completion of this command these buffers are
            cleared.  If the processing is successful the receiver must
            send an OK reply.  If the processing fails completely the
            receiver must send a failure reply.

            When the receiver-SMTP accepts a message either for relaying
            or for final delivery it inserts at the beginning of the
            mail data a time stamp line.  The time stamp line indicates
            the identity of the host that sent the message, and the
            identity of the host that received the message (and is
            inserting this time stamp), and the date and time the
            message was received.  Relayed messages will have multiple
            time stamp lines.

            When the receiver-SMTP makes the "final delivery" of a
            message it inserts at the beginning of the mail data a

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

            return path line.  The return path line preserves the
            information in the <reverse-path> from the MAIL command.
            Here, final delivery means the message leaves the SMTP
            world.  Normally, this would mean it has been delivered to
            the destination user, but in some cases it may be further
            processed and transmitted by another mail system.

               It is possible for the mailbox in the return path be
               different from the actual sender's mailbox, for example,
               if error responses are to be delivered a special error
               handling mailbox rather than the message senders.

            The preceding two paragraphs imply that the final mail data
            will begin with a  return path line, followed by one or more
            time stamp lines.  These lines will be followed by the mail
            data header and body [2].  See Example 8.

            Special mention is needed of the response and further action
            required when the processing following the end of mail data
            indication is partially successful.  This could arise if
            after accepting several recipients and the mail data, the
            receiver-SMTP finds that the mail data can be successfully
            delivered to some of the recipients, but it cannot be to
            others (for example, due to mailbox space allocation
            problems).  In such a situation, the response to the DATA
            command must be an OK reply.  But, the receiver-SMTP must
            compose and send an "undeliverable mail" notification
            message to the originator of the message.  Either a single
            notification which lists all of the recipients that failed
            to get the message, or separate notification messages must
            be sent for each failed recipient (see Example 7).  All
            undeliverable mail notification messages are sent using the
            MAIL command (even if they result from processing a SEND,
            SOML, or SAML command).

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

     -------------------------------------------------------------

            Example of Return Path and Received Time Stamps

      Return-Path: <@GHI.ARPA,@DEF.ARPA,@ABC.ARPA:JOE@ABC.ARPA>  
      Received: from GHI.ARPA by JKL.ARPA ; 27 Oct 81 15:27:39 PST
      Received: from DEF.ARPA by GHI.ARPA ; 27 Oct 81 15:15:13 PST
      Received: from ABC.ARPA by DEF.ARPA ; 27 Oct 81 15:01:59 PST
      Date: 27 Oct 81 15:01:01 PST                               
      From: JOE@ABC.ARPA                                         
      Subject: Improved Mailing System Installed                 
      To: SAM@JKL.ARPA                                           

      This is to inform you that ...                             

                               Example 8

     -------------------------------------------------------------

         SEND (SEND)

            This command is used to initiate a mail transaction in which
            the mail data is delivered to one or more terminals.  The
            argument field contains a reverse-path.  This command is
            successful if the message is delivered to a terminal.

            The reverse-path consists of an optional list of hosts and
            the sender mailbox.  When the list of hosts is present, it
            is a "reverse" source route and indicates that the mail was
            relayed through each host on the list (the first host in the
            list was the most recent relay).  This list is used as a
            source route to return non-delivery notices to the sender.
            As each relay host adds itself to the beginning of the list,
            it must use its name as known in the IPCE to which it is
            relaying the mail rather than the IPCE from which the mail
            came (if they are different).

            This command clears the reverse-path buffer, the
            forward-path buffer, and the mail data buffer; and inserts
            the reverse-path information from this command into the
            reverse-path buffer.

         SEND OR MAIL (SOML)

            This command is used to initiate a mail transaction in which
            the mail data is delivered to one or more terminals or

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

            mailboxes. For each recipient the mail data is delivered to
            the recipient's terminal if the recipient is active on the
            host (and accepting terminal messages), otherwise to the
            recipient's mailbox.  The argument field contains a
            reverse-path.  This command is successful if the message is
            delivered to a terminal or the mailbox.

            The reverse-path consists of an optional list of hosts and
            the sender mailbox.  When the list of hosts is present, it
            is a "reverse" source route and indicates that the mail was
            relayed through each host on the list (the first host in the
            list was the most recent relay).  This list is used as a
            source route to return non-delivery notices to the sender.
            As each relay host adds itself to the beginning of the list,
            it must use its name as known in the IPCE to which it is
            relaying the mail rather than the IPCE from which the mail
            came (if they are different).

            This command clears the reverse-path buffer, the
            forward-path buffer, and the mail data buffer; and inserts
            the reverse-path information from this command into the
            reverse-path buffer.

         SEND AND MAIL (SAML)

            This command is used to initiate a mail transaction in which
            the mail data is delivered to one or more terminals and
            mailboxes. For each recipient the mail data is delivered to
            the recipient's terminal if the recipient is active on the
            host (and accepting terminal messages), and for all
            recipients to the recipient's mailbox.  The argument field
            contains a reverse-path.  This command is successful if the
            message is delivered to the mailbox.

            The reverse-path consists of an optional list of hosts and
            the sender mailbox.  When the list of hosts is present, it
            is a "reverse" source route and indicates that the mail was
            relayed through each host on the list (the first host in the
            list was the most recent relay).  This list is used as a
            source route to return non-delivery notices to the sender.
            As each relay host adds itself to the beginning of the list,
            it must use its name as known in the IPCE to which it is
            relaying the mail rather than the IPCE from which the mail
            came (if they are different).

            This command clears the reverse-path buffer, the

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

            forward-path buffer, and the mail data buffer; and inserts
            the reverse-path information from this command into the
            reverse-path buffer.

         RESET (RSET)

            This command specifies that the current mail transaction is
            to be aborted.  Any stored sender, recipients, and mail data
            must be discarded, and all buffers and state tables cleared.
            The receiver must send an OK reply.

         VERIFY (VRFY)

            This command asks the receiver to confirm that the argument
            identifies a user.  If it is a user name, the full name of
            the user (if known) and the fully specified mailbox are
            returned.

            This command has no effect on any of the reverse-path
            buffer, the forward-path buffer, or the mail data buffer.

         EXPAND (EXPN)

            This command asks the receiver to confirm that the argument
            identifies a mailing list, and if so, to return the
            membership of that list.  The full name of the users (if
            known) and the fully specified mailboxes are returned in a
            multiline reply.

            This command has no effect on any of the reverse-path
            buffer, the forward-path buffer, or the mail data buffer.

         HELP (HELP)

            This command causes the receiver to send helpful information
            to the sender of the HELP command.  The command may take an
            argument (e.g., any command name) and return more specific
            information as a response.

            This command has no effect on any of the reverse-path
            buffer, the forward-path buffer, or the mail data buffer.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         NOOP (NOOP)

            This command does not affect any parameters or previously
            entered commands.  It specifies no action other than that
            the receiver send an OK reply.

            This command has no effect on any of the reverse-path
            buffer, the forward-path buffer, or the mail data buffer.

         QUIT (QUIT)

            This command specifies that the receiver must send an OK
            reply, and then close the transmission channel.

            The receiver should not close the transmission channel until
            it receives and replies to a QUIT command (even if there was
            an error).  The sender should not close the transmission
            channel until it send a QUIT command and receives the reply
            (even if there was an error response to a previous command).
            If the connection is closed prematurely the receiver should
            act as if a RSET command had been received (canceling any
            pending transaction, but not undoing any previously
            completed transaction), the sender should act as if the
            command or transaction in progress had received a temporary
            error (4xx).

         TURN (TURN)

            This command specifies that the receiver must either (1)
            send an OK reply and then take on the role of the
            sender-SMTP, or (2) send a refusal reply and retain the role
            of the receiver-SMTP.

            If program-A is currently the sender-SMTP and it sends the
            TURN command and receives an OK reply (250) then program-A
            becomes the receiver-SMTP.  Program-A is then in the initial
            state as if the transmission channel just opened, and it
            then sends the 220 service ready greeting.

            If program-B is currently the receiver-SMTP and it receives
            the TURN command and sends an OK reply (250) then program-B
            becomes the sender-SMTP.  Program-B is then in the initial
            state as if the transmission channel just opened, and it
            then expects to receive the 220 service ready greeting.

            To refuse to change roles the receiver sends the 502 reply.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

         There are restrictions on the order in which these command may
         be used.

            The first command in a session must be the HELO command.
            The HELO command may be used later in a session as well.  If
            the HELO command argument is not acceptable a 501 failure
            reply must be returned and the receiver-SMTP must stay in
            the same state.

            The NOOP, HELP, EXPN, and VRFY commands can be used at any
            time during a session.

            The MAIL, SEND, SOML, or SAML commands begin a mail
            transaction.  Once started a mail transaction consists of
            one of the transaction beginning commands, one or more RCPT
            commands, and a DATA command, in that order.  A mail
            transaction may be aborted by the RSET command.  There may
            be zero or more transactions in a session.

            If the transaction beginning command argument is not
            acceptable a 501 failure reply must be returned and the
            receiver-SMTP must stay in the same state.  If the commands
            in a transaction are out of order a 503 failure reply must
            be returned and the receiver-SMTP must stay in the same
            state.

            The last command in a session must be the QUIT command.  The
            QUIT command can not be used at any other time in a session.

      4.1.2.  COMMAND SYNTAX

         The commands consist of a command code followed by an argument
         field.  Command codes are four alphabetic characters.  Upper
         and lower case alphabetic characters are to be treated
         identically.  Thus, any of the following may represent the mail
         command:

            MAIL    Mail    mail    MaIl    mAIl

         This also applies to any symbols representing parameter values,
         such as "TO" or "to" for the forward-path.  Command codes and
         the argument fields are separated by one or more spaces.
         However, within the reverse-path and forward-path arguments
         case is important.  In particular, in some hosts the user
         "smith" is different from the user "Smith".

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         The argument field consists of a variable length character
         string ending with the character sequence <CRLF>.  The receiver
         is to take no action until this sequence is received.

         Square brackets denote an optional argument field.  If the
         option is not taken, the appropriate default is implied.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

         The following are the SMTP commands:

            HELO <SP> <domain> <CRLF>

            MAIL <SP> FROM:<reverse-path> <CRLF>

            RCPT <SP> TO:<forward-path> <CRLF>

            DATA <CRLF>

            RSET <CRLF>

            SEND <SP> FROM:<reverse-path> <CRLF>

            SOML <SP> FROM:<reverse-path> <CRLF>

            SAML <SP> FROM:<reverse-path> <CRLF>

            VRFY <SP> <string> <CRLF>

            EXPN <SP> <string> <CRLF>

            HELP [<SP> <string>] <CRLF>

            NOOP <CRLF>

            QUIT <CRLF>

            TURN <CRLF>

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         The syntax of the above argument fields (using BNF notation
         where applicable) is given below.  The "..." notation indicates
         that a field may be repeated one or more times.

            <reverse-path> ::= <path>

            <forward-path> ::= <path>

            <path> ::= "<" [ <a-d-l> ":" ] <mailbox> ">"

            <a-d-l> ::= <at-domain> | <at-domain> "," <a-d-l>

            <at-domain> ::= "@" <domain>

            <domain> ::=  <element> | <element> "." <domain>

            <element> ::= <name> | "#" <number> | "[" <dotnum> "]"

            <mailbox> ::= <local-part> "@" <domain>

            <local-part> ::= <dot-string> | <quoted-string>

            <name> ::= <a> <ldh-str> <let-dig>

            <ldh-str> ::= <let-dig-hyp> | <let-dig-hyp> <ldh-str>

            <let-dig> ::= <a> | <d>

            <let-dig-hyp> ::= <a> | <d> | "-"

            <dot-string> ::= <string> | <string> "." <dot-string>

            <string> ::= <char> | <char> <string>

            <quoted-string> ::=  """ <qtext> """

            <qtext> ::=  "\" <x> | "\" <x> <qtext> | <q> | <q> <qtext>

            <char> ::= <c> | "\" <x>

            <dotnum> ::= <snum> "." <snum> "." <snum> "." <snum>

            <number> ::= <d> | <d> <number>

            <CRLF> ::= <CR> <LF>

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

            <CR> ::= the carriage return character (ASCII code 13)

            <LF> ::= the line feed character (ASCII code 10)

            <SP> ::= the space character (ASCII code 32)

            <snum> ::= one, two, or three digits representing a decimal
                      integer value in the range 0 through 255

            <a> ::= any one of the 52 alphabetic characters A through Z
                      in upper case and a through z in lower case

            <c> ::= any one of the 128 ASCII characters, but not any
                      <special> or <SP>

            <d> ::= any one of the ten digits 0 through 9

            <q> ::= any one of the 128 ASCII characters except <CR>,
                      <LF>, quote ("), or backslash (\)

            <x> ::= any one of the 128 ASCII characters (no exceptions)

            <special> ::= "<" | ">" | "(" | ")" | "[" | "]" | "\" | "."
                      | "," | ";" | ":" | "@"  """ | the control
                      characters (ASCII codes 0 through 31 inclusive and
                      127)

         Note that the backslash, "\", is a quote character, which is
         used to indicate that the next character is to be used
         literally (instead of its normal interpretation).  For example,
         "Joe\,Smith" could be used to indicate a single nine character
         user field with comma being the fourth character of the field.

         Hosts are generally known by names which are translated to
         addresses in each host.  Note that the name elements of domains
         are the official names -- no use of nicknames or aliases is
         allowed.

         Sometimes a host is not known to the translation function and
         communication is blocked.  To bypass this barrier two numeric
         forms are also allowed for host "names".  One form is a decimal
         integer prefixed by a pound sign, "#", which indicates the
         number is the address of the host.  Another form is four small
         decimal integers separated by dots and enclosed by brackets,
         e.g., "[123.255.37.2]", which indicates a 32-bit ARPA Internet
         Address in four 8-bit fields.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         The time stamp line and the return path line are formally
         defined as follows:

         <return-path-line> ::= "Return-Path:" <SP><reverse-path><CRLF>

         <time-stamp-line> ::= "Received:" <SP> <stamp> <CRLF>

            <stamp> ::= <from-domain> <by-domain> <opt-info> ";"
                      <daytime>

            <from-domain> ::= "FROM" <SP> <domain> <SP>

            <by-domain> ::= "BY" <SP> <domain> <SP>

            <opt-info> ::= [<via>] [<with>] [<id>] [<for>]

            <via> ::= "VIA" <SP> <link> <SP>

            <with> ::= "WITH" <SP> <protocol> <SP>

            <id> ::= "ID" <SP> <string> <SP>

            <for> ::= "FOR" <SP> <path> <SP>

            <link> ::= The standard names for links are registered with
                      the Network Information Center.

            <protocol> ::= The standard names for protocols are
                      registered with the Network Information Center.

            <daytime> ::= <SP> <date> <SP> <time>

            <date> ::= <dd> <SP> <mon> <SP> <yy>

            <time> ::= <hh> ":" <mm> ":" <ss> <SP> <zone>

            <dd> ::= the one or two decimal integer day of the month in
                      the range 1 to 31.

            <mon> ::= "JAN" | "FEB" | "MAR" | "APR" | "MAY" | "JUN" |
                      "JUL" | "AUG" | "SEP" | "OCT" | "NOV" | "DEC"

            <yy> ::= the two decimal integer year of the century in the
                      range 00 to 99.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

            <hh> ::= the two decimal integer hour of the day in the
                      range 00 to 24.

            <mm> ::= the two decimal integer minute of the hour in the
                      range 00 to 59.

            <ss> ::= the two decimal integer second of the minute in the
                      range 00 to 59.

            <zone> ::= "UT" for Universal Time (the default) or other
                      time zone designator (as in [2]).

     -------------------------------------------------------------

                          Return Path Example

         Return-Path: <@CHARLIE.ARPA,@BAKER.ARPA:JOE@ABLE.ARPA>

                               Example 9

     -------------------------------------------------------------

     -------------------------------------------------------------

                        Time Stamp Line Example

      Received: FROM ABC.ARPA BY XYZ.ARPA ; 22 OCT 81 09:23:59 PDT

         Received: from ABC.ARPA by XYZ.ARPA via TELENET with X25
                   id M12345 for Smith@PDQ.ARPA ; 22 OCT 81 09:23:59 PDT

                               Example 10

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   4.2.  SMTP REPLIES

      Replies to SMTP commands are devised to ensure the synchronization
      of requests and actions in the process of mail transfer, and to
      guarantee that the sender-SMTP always knows the state of the
      receiver-SMTP.  Every command must generate exactly one reply.

         The details of the command-reply sequence are made explicit in
         Section 5.3 on Sequencing and Section 5.4 State Diagrams.

      An SMTP reply consists of a three digit number (transmitted as
      three alphanumeric characters) followed by some text.  The number
      is intended for use by automata to determine what state to enter
      next; the text is meant for the human user.  It is intended that
      the three digits contain enough encoded information that the
      sender-SMTP need not examine the text and may either discard it or
      pass it on to the user, as appropriate.  In particular, the text
      may be receiver-dependent and context dependent, so there are
      likely to be varying texts for each reply code.  A discussion of
      the theory of reply codes is given in Appendix E.  Formally, a
      reply is defined to be the sequence:  a three-digit code, <SP>,
      one line of text, and <CRLF>, or a multiline reply (as defined in
      Appendix E).  Only the EXPN and HELP commands are expected to
      result in multiline replies in normal circumstances, however
      multiline replies are allowed for any command.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

      4.2.1.  REPLY CODES BY FUNCTION GROUPS

         500 Syntax error, command unrecognized
            [This may include errors such as command line too long]
         501 Syntax error in parameters or arguments
         502 Command not implemented
         503 Bad sequence of commands
         504 Command parameter not implemented

         211 System status, or system help reply
         214 Help message
            [Information on how to use the receiver or the meaning of a
            particular non-standard command; this reply is useful only
            to the human user]

         220 <domain> Service ready
         221 <domain> Service closing transmission channel
         421 <domain> Service not available,
             closing transmission channel
            [This may be a reply to any command if the service knows it
            must shut down]

         250 Requested mail action okay, completed
         251 User not local; will forward to <forward-path>
         450 Requested mail action not taken: mailbox unavailable
            [E.g., mailbox busy]
         550 Requested action not taken: mailbox unavailable
            [E.g., mailbox not found, no access]
         451 Requested action aborted: error in processing
         551 User not local; please try <forward-path>
         452 Requested action not taken: insufficient system storage
         552 Requested mail action aborted: exceeded storage allocation
         553 Requested action not taken: mailbox name not allowed
            [E.g., mailbox syntax incorrect]
         354 Start mail input; end with <CRLF>.<CRLF>
         554 Transaction failed

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      4.2.2.  NUMERIC ORDER LIST OF REPLY CODES

         211 System status, or system help reply
         214 Help message
            [Information on how to use the receiver or the meaning of a
            particular non-standard command; this reply is useful only
            to the human user]
         220 <domain> Service ready
         221 <domain> Service closing transmission channel
         250 Requested mail action okay, completed
         251 User not local; will forward to <forward-path>

         354 Start mail input; end with <CRLF>.<CRLF>

         421 <domain> Service not available,
             closing transmission channel
            [This may be a reply to any command if the service knows it
            must shut down]
         450 Requested mail action not taken: mailbox unavailable
            [E.g., mailbox busy]
         451 Requested action aborted: local error in processing
         452 Requested action not taken: insufficient system storage

         500 Syntax error, command unrecognized
            [This may include errors such as command line too long]
         501 Syntax error in parameters or arguments
         502 Command not implemented
         503 Bad sequence of commands
         504 Command parameter not implemented
         550 Requested action not taken: mailbox unavailable
            [E.g., mailbox not found, no access]
         551 User not local; please try <forward-path>
         552 Requested mail action aborted: exceeded storage allocation
         553 Requested action not taken: mailbox name not allowed
            [E.g., mailbox syntax incorrect]
         554 Transaction failed

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   4.3.  SEQUENCING OF COMMANDS AND REPLIES

      The communication between the sender and receiver is intended to
      be an alternating dialogue, controlled by the sender.  As such,
      the sender issues a command and the receiver responds with a
      reply.  The sender must wait for this response before sending
      further commands.

      One important reply is the connection greeting.  Normally, a
      receiver will send a 220 "Service ready" reply when the connection
      is completed.  The sender should wait for this greeting message
      before sending any commands.

         Note: all the greeting type replies have the official name of
         the server host as the first word following the reply code.

            For example,

               220 <SP> USC-ISIF.ARPA <SP> Service ready <CRLF>

      The table below lists alternative success and failure replies for
      each command.  These must be strictly adhered to; a receiver may
      substitute text in the replies, but the meaning and action implied
      by the code numbers and by the specific command reply sequence
      cannot be altered.

      COMMAND-REPLY SEQUENCES

         Each command is listed with its possible replies.  The prefixes
         used before the possible replies are "P" for preliminary (not
         used in SMTP), "I" for intermediate, "S" for success, "F" for
         failure, and "E" for error.  The 421 reply (service not
         available, closing transmission channel) may be given to any
         command if the SMTP-receiver knows it must shut down.  This
         listing forms the basis for the State Diagrams in Section 4.4.

            CONNECTION ESTABLISHMENT
               S: 220
               F: 421
            HELO
               S: 250
               E: 500, 501, 504, 421
            MAIL
               S: 250
               F: 552, 451, 452
               E: 500, 501, 421

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

            RCPT
               S: 250, 251
               F: 550, 551, 552, 553, 450, 451, 452
               E: 500, 501, 503, 421
            DATA
               I: 354 -> data -> S: 250
                                 F: 552, 554, 451, 452
               F: 451, 554
               E: 500, 501, 503, 421
            RSET
               S: 250
               E: 500, 501, 504, 421
            SEND
               S: 250
               F: 552, 451, 452
               E: 500, 501, 502, 421
            SOML
               S: 250
               F: 552, 451, 452
               E: 500, 501, 502, 421
            SAML
               S: 250
               F: 552, 451, 452
               E: 500, 501, 502, 421
            VRFY
               S: 250, 251
               F: 550, 551, 553
               E: 500, 501, 502, 504, 421
            EXPN
               S: 250
               F: 550
               E: 500, 501, 502, 504, 421
            HELP
               S: 211, 214
               E: 500, 501, 502, 504, 421
            NOOP
               S: 250
               E: 500, 421
            QUIT
               S: 221
               E: 500
            TURN
               S: 250
               F: 502
               E: 500, 503

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   4.4.  STATE DIAGRAMS

      Following are state diagrams for a simple-minded SMTP
      implementation.  Only the first digit of the reply codes is used.
      There is one state diagram for each group of SMTP commands.  The
      command groupings were determined by constructing a model for each
      command and then collecting together the commands with
      structurally identical models.

      For each command there are three possible outcomes:  "success"
      (S), "failure" (F), and "error" (E). In the state diagrams below
      we use the symbol B for "begin", and the symbol W for "wait for
      reply".

      First, the diagram that represents most of the SMTP commands:

                                  1,3    +---+
                             ----------->| E |
                            |            +---+
                            |
         +---+    cmd    +---+    2      +---+
         | B |---------->| W |---------->| S |
         +---+           +---+           +---+
                            |
                            |     4,5    +---+
                             ----------->| F |
                                         +---+

         This diagram models the commands:

            HELO, MAIL, RCPT, RSET, SEND, SOML, SAML, VRFY, EXPN, HELP,
            NOOP, QUIT, TURN.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      A more complex diagram models the DATA command:

         +---+   DATA    +---+ 1,2                 +---+
         | B |---------->| W |-------------------->| E |
         +---+           +---+        ------------>+---+
                         3| |4,5     |
                          | |        |
            --------------   -----   |
           |                      |  |             +---+
           |               ----------     -------->| S |
           |              |       |      |         +---+
           |              |  ------------
           |              | |     |
           V           1,3| |2    |
         +---+   data    +---+     --------------->+---+
         |   |---------->| W |                     | F |
         +---+           +---+-------------------->+---+
                              4,5

         Note that the "data" here is a series of lines sent from the
         sender to the receiver with no response expected until the last
         line is sent.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   4.5.  DETAILS

      4.5.1.  MINIMUM IMPLEMENTATION

         In order to make SMTP workable, the following minimum
         implementation is required for all receivers:

            COMMANDS -- HELO
                        MAIL
                        RCPT
                        DATA
                        RSET
                        NOOP
                        QUIT

      4.5.2.  TRANSPARENCY

         Without some provision for data transparency the character
         sequence "<CRLF>.<CRLF>" ends the mail text and cannot be sent
         by the user.  In general, users are not aware of such
         "forbidden" sequences.  To allow all user composed text to be
         transmitted transparently the following procedures are used.

            1. Before sending a line of mail text the sender-SMTP checks
            the first character of the line.  If it is a period, one
            additional period is inserted at the beginning of the line.

            2. When a line of mail text is received by the receiver-SMTP
            it checks the line.  If the line is composed of a single
            period it is the end of mail.  If the first character is a
            period and there are other characters on the line, the first
            character is deleted.

         The mail data may contain any of the 128 ASCII characters.  All
         characters are to be delivered to the recipient's mailbox
         including format effectors and other control characters.  If
         the transmission channel provides an 8-bit byte (octets) data
         stream, the 7-bit ASCII codes are transmitted right justified
         in the octets with the high order bits cleared to zero.

            In some systems it may be necessary to transform the data as
            it is received and stored.  This may be necessary for hosts
            that use a different character set than ASCII as their local
            character set, or that store data in records rather than

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

            strings.  If such transforms are necessary, they must be
            reversible -- especially if such transforms are applied to
            mail being relayed.

      4.5.3.  SIZES

         There are several objects that have required minimum maximum
         sizes.  That is, every implementation must be able to receive
         objects of at least these sizes, but must not send objects
         larger than these sizes.

          ****************************************************
          *                                                  *
          *  TO THE MAXIMUM EXTENT POSSIBLE, IMPLEMENTATION  *
          *  TECHNIQUES WHICH IMPOSE NO LIMITS ON THE LENGTH *
          *  OF THESE OBJECTS SHOULD BE USED.                *
          *                                                  *
          ****************************************************

            user

               The maximum total length of a user name is 64 characters.

            domain

               The maximum total length of a domain name or number is 64
               characters.

            path

               The maximum total length of a reverse-path or
               forward-path is 256 characters (including the punctuation
               and element separators).

            command line

               The maximum total length of a command line including the
               command word and the <CRLF> is 512 characters.

            reply line

               The maximum total length of a reply line including the
               reply code and the <CRLF> is 512 characters.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

            text line

               The maximum total length of a text line including the
               <CRLF> is 1000 characters (but not counting the leading
               dot duplicated for transparency).

            recipients buffer

               The maximum total number of recipients that must be
               buffered is 100 recipients.

          ****************************************************
          *                                                  *
          *  TO THE MAXIMUM EXTENT POSSIBLE, IMPLEMENTATION  *
          *  TECHNIQUES WHICH IMPOSE NO LIMITS ON THE LENGTH *
          *  OF THESE OBJECTS SHOULD BE USED.                *
          *                                                  *
          ****************************************************

         Errors due to exceeding these limits may be reported by using
         the reply codes, for example:

            500 Line too long.

            501 Path too long

            552 Too many recipients.

            552 Too much mail data.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

APPENDIX A

   TCP Transport service

      The Transmission Control Protocol [3] is used in the ARPA
      Internet, and in any network following the US DoD standards for
      internetwork protocols.

      Connection Establishment

         The SMTP transmission channel is a TCP connection established
         between the sender process port U and the receiver process port
         L.  This single full duplex connection is used as the
         transmission channel.  This protocol is assigned the service
         port 25 (31 octal), that is L=25.

      Data Transfer

         The TCP connection supports the transmission of 8-bit bytes.
         The SMTP data is 7-bit ASCII characters.  Each character is
         transmitted as an 8-bit byte with the high-order bit cleared to
         zero.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

APPENDIX B

   NCP Transport service

      The ARPANET Host-to-Host Protocol [4] (implemented by the Network
      Control Program) may be used in the ARPANET.

      Connection Establishment

         The SMTP transmission channel is established via NCP between
         the sender process socket U and receiver process socket L.  The
         Initial Connection Protocol [5] is followed resulting in a pair
         of simplex connections.  This pair of connections is used as
         the transmission channel.  This protocol is assigned the
         contact socket 25 (31 octal), that is L=25.

      Data Transfer

         The NCP data connections are established in 8-bit byte mode.
         The SMTP data is 7-bit ASCII characters.  Each character is
         transmitted as an 8-bit byte with the high-order bit cleared to
         zero.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

APPENDIX C

   NITS

      The Network Independent Transport Service [6] may be used.

      Connection Establishment

         The SMTP transmission channel is established via NITS between
         the sender process and receiver process.  The sender process
         executes the CONNECT primitive, and the waiting receiver
         process executes the ACCEPT primitive.

      Data Transfer

         The NITS connection supports the transmission of 8-bit bytes.
         The SMTP data is 7-bit ASCII characters.  Each character is
         transmitted as an 8-bit byte with the high-order bit cleared to
         zero.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

APPENDIX D

   X.25 Transport service

      It may be possible to use the X.25 service [7] as provided by the
      Public Data Networks directly, however, it is suggested that a
      reliable end-to-end protocol such as TCP be used on top of X.25
      connections.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

APPENDIX E

   Theory of Reply Codes

      The three digits of the reply each have a special significance.
      The first digit denotes whether the response is good, bad or
      incomplete.  An unsophisticated sender-SMTP will be able to
      determine its next action (proceed as planned, redo, retrench,
      etc.) by simply examining this first digit.  A sender-SMTP that
      wants to know approximately what kind of error occurred (e.g.,
      mail system error, command syntax error) may examine the second
      digit, reserving the third digit for the finest gradation of
      information.

         There are five values for the first digit of the reply code:

            1yz   Positive Preliminary reply

               The command has been accepted, but the requested action
               is being held in abeyance, pending confirmation of the
               information in this reply.  The sender-SMTP should send
               another command specifying whether to continue or abort
               the action.

                  [Note: SMTP does not have any commands that allow this
                  type of reply, and so does not have the continue or
                  abort commands.]

            2yz   Positive Completion reply

               The requested action has been successfully completed.  A
               new request may be initiated.

            3yz   Positive Intermediate reply

               The command has been accepted, but the requested action
               is being held in abeyance, pending receipt of further
               information.  The sender-SMTP should send another command
               specifying this information.  This reply is used in
               command sequence groups.

            4yz   Transient Negative Completion reply

               The command was not accepted and the requested action did
               not occur.  However, the error condition is temporary and
               the action may be requested again.  The sender should

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

               return to the beginning of the command sequence (if any).
               It is difficult to assign a meaning to "transient" when
               two different sites (receiver- and sender- SMTPs) must
               agree on the interpretation.  Each reply in this category
               might have a different time value, but the sender-SMTP is
               encouraged to try again.  A rule of thumb to determine if
               a reply fits into the 4yz or the 5yz category (see below)
               is that replies are 4yz if they can be repeated without
               any change in command form or in properties of the sender
               or receiver.  (E.g., the command is repeated identically
               and the receiver does not put up a new implementation.)

            5yz   Permanent Negative Completion reply

               The command was not accepted and the requested action did
               not occur.  The sender-SMTP is discouraged from repeating
               the exact request (in the same sequence).  Even some
               "permanent" error conditions can be corrected, so the
               human user may want to direct the sender-SMTP to
               reinitiate the command sequence by direct action at some
               point in the future (e.g., after the spelling has been
               changed, or the user has altered the account status).

         The second digit encodes responses in specific categories:

            x0z   Syntax -- These replies refer to syntax errors,
                  syntactically correct commands that don't fit any
                  functional category, and unimplemented or superfluous
                  commands.

            x1z   Information --  These are replies to requests for
                  information, such as status or help.

            x2z   Connections -- These are replies referring to the
                  transmission channel.

            x3z   Unspecified as yet.

            x4z   Unspecified as yet.

            x5z   Mail system -- These replies indicate the status of
                  the receiver mail system vis-a-vis the requested
                  transfer or other mail system action.

         The third digit gives a finer gradation of meaning in each
         category specified by the second digit.  The list of replies

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         illustrates this.  Each reply text is recommended rather than
         mandatory, and may even change according to the command with
         which it is associated.  On the other hand, the reply codes
         must strictly follow the specifications in this section.
         Receiver implementations should not invent new codes for
         slightly different situations from the ones described here, but
         rather adapt codes already defined.

         For example, a command such as NOOP whose successful execution
         does not offer the sender-SMTP any new information will return
         a 250 reply.  The response is 502 when the command requests an
         unimplemented non-site-specific action.  A refinement of that
         is the 504 reply for a command that is implemented, but that
         requests an unimplemented parameter.

      The reply text may be longer than a single line; in these cases
      the complete text must be marked so the sender-SMTP knows when it
      can stop reading the reply.  This requires a special format to
      indicate a multiple line reply.

         The format for multiline replies requires that every line,
         except the last, begin with the reply code, followed
         immediately by a hyphen, "-" (also known as minus), followed by
         text.  The last line will begin with the reply code, followed
         immediately by <SP>, optionally some text, and <CRLF>.

            For example:
                                123-First line
                                123-Second line
                                123-234 text beginning with numbers
                                123 The last line

         In many cases the sender-SMTP then simply needs to search for
         the reply code followed by <SP> at the beginning of a line, and
         ignore all preceding lines.  In a few cases, there is important
         data for the sender in the reply "text".  The sender will know
         these cases from the current context.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

APPENDIX F

   Scenarios

      This section presents complete scenarios of several types of SMTP
      sessions.

   A Typical SMTP Transaction Scenario

      This SMTP example shows mail sent by Smith at host USC-ISIF, to
      Jones, Green, and Brown at host BBN-UNIX.  Here we assume that
      host USC-ISIF contacts host BBN-UNIX directly.  The mail is
      accepted for Jones and Brown.  Green does not have a mailbox at
      host BBN-UNIX.

      -------------------------------------------------------------

         R: 220 BBN-UNIX.ARPA Simple Mail Transfer Service Ready
         S: HELO USC-ISIF.ARPA
         R: 250 BBN-UNIX.ARPA

         S: MAIL FROM:<Smith@USC-ISIF.ARPA>
         R: 250 OK

         S: RCPT TO:<Jones@BBN-UNIX.ARPA>
         R: 250 OK

         S: RCPT TO:<Green@BBN-UNIX.ARPA>
         R: 550 No such user here

         S: RCPT TO:<Brown@BBN-UNIX.ARPA>
         R: 250 OK

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: QUIT
         R: 221 BBN-UNIX.ARPA Service closing transmission channel

                               Scenario 1

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   Aborted SMTP Transaction Scenario

      -------------------------------------------------------------

         R: 220 MIT-Multics.ARPA Simple Mail Transfer Service Ready
         S: HELO ISI-VAXA.ARPA
         R: 250 MIT-Multics.ARPA

         S: MAIL FROM:<Smith@ISI-VAXA.ARPA>
         R: 250 OK

         S: RCPT TO:<Jones@MIT-Multics.ARPA>
         R: 250 OK

         S: RCPT TO:<Green@MIT-Multics.ARPA>
         R: 550 No such user here

         S: RSET
         R: 250 OK

         S: QUIT
         R: 221 MIT-Multics.ARPA Service closing transmission channel

                               Scenario 2

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   Relayed Mail Scenario

      -------------------------------------------------------------

         Step 1  --  Source Host to Relay Host

            R: 220 USC-ISIE.ARPA Simple Mail Transfer Service Ready
            S: HELO MIT-AI.ARPA
            R: 250 USC-ISIE.ARPA

            S: MAIL FROM:<JQP@MIT-AI.ARPA>
            R: 250 OK

            S: RCPT TO:<@USC-ISIE.ARPA:Jones@BBN-VAX.ARPA>
            R: 250 OK

            S: DATA
            R: 354 Start mail input; end with <CRLF>.<CRLF>
            S: Date: 2 Nov 81 22:33:44
            S: From: John Q. Public <JQP@MIT-AI.ARPA>
            S: Subject:  The Next Meeting of the Board
            S: To: Jones@BBN-Vax.ARPA
            S:
            S: Bill:
            S: The next meeting of the board of directors will be
            S: on Tuesday.
            S:                                              John.
            S: .
            R: 250 OK

            S: QUIT
            R: 221 USC-ISIE.ARPA Service closing transmission channel

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         Step 2  --  Relay Host to Destination Host

            R: 220 BBN-VAX.ARPA Simple Mail Transfer Service Ready
            S: HELO USC-ISIE.ARPA
            R: 250 BBN-VAX.ARPA

            S: MAIL FROM:<@USC-ISIE.ARPA:JQP@MIT-AI.ARPA>
            R: 250 OK

            S: RCPT TO:<Jones@BBN-VAX.ARPA>
            R: 250 OK

            S: DATA
            R: 354 Start mail input; end with <CRLF>.<CRLF>
            S: Received: from MIT-AI.ARPA by USC-ISIE.ARPA ;
               2 Nov 81 22:40:10 UT
            S: Date: 2 Nov 81 22:33:44
            S: From: John Q. Public <JQP@MIT-AI.ARPA>
            S: Subject:  The Next Meeting of the Board
            S: To: Jones@BBN-Vax.ARPA
            S:
            S: Bill:
            S: The next meeting of the board of directors will be
            S: on Tuesday.
            S:                                              John.
            S: .
            R: 250 OK

            S: QUIT
            R: 221 USC-ISIE.ARPA Service closing transmission channel

                               Scenario 3

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   Verifying and Sending Scenario

      -------------------------------------------------------------

         R: 220 SU-SCORE.ARPA Simple Mail Transfer Service Ready
         S: HELO MIT-MC.ARPA
         R: 250 SU-SCORE.ARPA

         S: VRFY Crispin
         R: 250 Mark Crispin <Admin.MRC@SU-SCORE.ARPA>

         S: SEND FROM:<EAK@MIT-MC.ARPA>
         R: 250 OK

         S: RCPT TO:<Admin.MRC@SU-SCORE.ARPA>
         R: 250 OK

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: QUIT
         R: 221 SU-SCORE.ARPA Service closing transmission channel

                               Scenario 4

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   Sending and Mailing Scenarios

      First the user's name is verified, then  an attempt is made to
      send to the user's terminal.  When that fails, the messages is
      mailed to the user's mailbox.

      -------------------------------------------------------------

         R: 220 SU-SCORE.ARPA Simple Mail Transfer Service Ready
         S: HELO MIT-MC.ARPA
         R: 250 SU-SCORE.ARPA

         S: VRFY Crispin
         R: 250 Mark Crispin <Admin.MRC@SU-SCORE.ARPA>

         S: SEND FROM:<EAK@MIT-MC.ARPA>
         R: 250 OK

         S: RCPT TO:<Admin.MRC@SU-SCORE.ARPA>
         R: 450 User not active now

         S: RSET
         R: 250 OK

         S: MAIL FROM:<EAK@MIT-MC.ARPA>
         R: 250 OK

         S: RCPT TO:<Admin.MRC@SU-SCORE.ARPA>
         R: 250 OK

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: QUIT
         R: 221 SU-SCORE.ARPA Service closing transmission channel

                               Scenario 5

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

      Doing the preceding scenario more efficiently.

      -------------------------------------------------------------

         R: 220 SU-SCORE.ARPA Simple Mail Transfer Service Ready
         S: HELO MIT-MC.ARPA
         R: 250 SU-SCORE.ARPA

         S: VRFY Crispin
         R: 250 Mark Crispin <Admin.MRC@SU-SCORE.ARPA>

         S: SOML FROM:<EAK@MIT-MC.ARPA>
         R: 250 OK

         S: RCPT TO:<Admin.MRC@SU-SCORE.ARPA>
         R: 250 User not active now, so will do mail.

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: QUIT
         R: 221 SU-SCORE.ARPA Service closing transmission channel

                               Scenario 6

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   Mailing List Scenario

      First each of two mailing lists are expanded in separate sessions
      with different hosts.  Then the message is sent to everyone that
      appeared on either list (but no duplicates) via a relay host.

      -------------------------------------------------------------

         Step 1  --  Expanding the First List

            R: 220 MIT-AI.ARPA Simple Mail Transfer Service Ready
            S: HELO SU-SCORE.ARPA
            R: 250 MIT-AI.ARPA

            S: EXPN Example-People
            R: 250-<ABC@MIT-MC.ARPA>
            R: 250-Fred Fonebone <Fonebone@USC-ISIQ.ARPA>
            R: 250-Xenon Y. Zither <XYZ@MIT-AI.ARPA>
            R: 250-Quincy Smith <@USC-ISIF.ARPA:Q-Smith@ISI-VAXA.ARPA>
            R: 250-<joe@foo-unix.ARPA>
            R: 250 <xyz@bar-unix.ARPA>

            S: QUIT
            R: 221 MIT-AI.ARPA Service closing transmission channel

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

         Step 2  --  Expanding the Second List

            R: 220 MIT-MC.ARPA Simple Mail Transfer Service Ready
            S: HELO SU-SCORE.ARPA
            R: 250 MIT-MC.ARPA

            S: EXPN Interested-Parties
            R: 250-Al Calico <ABC@MIT-MC.ARPA>
            R: 250-<XYZ@MIT-AI.ARPA>
            R: 250-Quincy Smith <@USC-ISIF.ARPA:Q-Smith@ISI-VAXA.ARPA>
            R: 250-<fred@BBN-UNIX.ARPA>
            R: 250 <xyz@bar-unix.ARPA>

            S: QUIT
            R: 221 MIT-MC.ARPA Service closing transmission channel

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

         Step 3  --  Mailing to All via a Relay Host

            R: 220 USC-ISIE.ARPA Simple Mail Transfer Service Ready
            S: HELO SU-SCORE.ARPA
            R: 250 USC-ISIE.ARPA

            S: MAIL FROM:<Account.Person@SU-SCORE.ARPA>
            R: 250 OK
            S: RCPT TO:<@USC-ISIE.ARPA:ABC@MIT-MC.ARPA>
            R: 250 OK
            S: RCPT TO:<@USC-ISIE.ARPA:Fonebone@USC-ISIQA.ARPA>
            R: 250 OK
            S: RCPT TO:<@USC-ISIE.ARPA:XYZ@MIT-AI.ARPA>
            R: 250 OK
            S: RCPT
                TO:<@USC-ISIE.ARPA,@USC-ISIF.ARPA:Q-Smith@ISI-VAXA.ARPA>
            R: 250 OK
            S: RCPT TO:<@USC-ISIE.ARPA:joe@FOO-UNIX.ARPA>
            R: 250 OK
            S: RCPT TO:<@USC-ISIE.ARPA:xyz@BAR-UNIX.ARPA>
            R: 250 OK
            S: RCPT TO:<@USC-ISIE.ARPA:fred@BBN-UNIX.ARPA>
            R: 250 OK

            S: DATA
            R: 354 Start mail input; end with <CRLF>.<CRLF>
            S: Blah blah blah...
            S: ...etc. etc. etc.
            S: .
            R: 250 OK

            S: QUIT
            R: 221 USC-ISIE.ARPA Service closing transmission channel

                               Scenario 7

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   Forwarding Scenarios

      -------------------------------------------------------------

         R: 220 USC-ISIF.ARPA Simple Mail Transfer Service Ready
         S: HELO LBL-UNIX.ARPA
         R: 250 USC-ISIF.ARPA

         S: MAIL FROM:<mo@LBL-UNIX.ARPA>
         R: 250 OK

         S: RCPT TO:<fred@USC-ISIF.ARPA>
         R: 251 User not local; will forward to <Jones@USC-ISI.ARPA>

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: QUIT
         R: 221 USC-ISIF.ARPA Service closing transmission channel

                               Scenario 8

      -------------------------------------------------------------

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

      -------------------------------------------------------------

         Step 1  --  Trying the Mailbox at the First Host

            R: 220 USC-ISIF.ARPA Simple Mail Transfer Service Ready
            S: HELO LBL-UNIX.ARPA
            R: 250 USC-ISIF.ARPA

            S: MAIL FROM:<mo@LBL-UNIX.ARPA>
            R: 250 OK

            S: RCPT TO:<fred@USC-ISIF.ARPA>
            R: 251 User not local; will forward to <Jones@USC-ISI.ARPA>

            S: RSET
            R: 250 OK

            S: QUIT
            R: 221 USC-ISIF.ARPA Service closing transmission channel

         Step 2  --  Delivering the Mail at the Second Host

            R: 220 USC-ISI.ARPA Simple Mail Transfer Service Ready
            S: HELO LBL-UNIX.ARPA
            R: 250 USC-ISI.ARPA

            S: MAIL FROM:<mo@LBL-UNIX.ARPA>
            R: 250 OK

            S: RCPT TO:<Jones@USC-ISI.ARPA>
            R: OK

            S: DATA
            R: 354 Start mail input; end with <CRLF>.<CRLF>
            S: Blah blah blah...
            S: ...etc. etc. etc.
            S: .
            R: 250 OK

            S: QUIT
            R: 221 USC-ISI.ARPA Service closing transmission channel

                               Scenario 9

      -------------------------------------------------------------

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   Too Many Recipients Scenario

      -------------------------------------------------------------

         R: 220 BERKELEY.ARPA Simple Mail Transfer Service Ready
         S: HELO USC-ISIF.ARPA
         R: 250 BERKELEY.ARPA

         S: MAIL FROM:<Postel@USC-ISIF.ARPA>
         R: 250 OK

         S: RCPT TO:<fabry@BERKELEY.ARPA>
         R: 250 OK

         S: RCPT TO:<eric@BERKELEY.ARPA>
         R: 552 Recipient storage full, try again in another transaction

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: MAIL FROM:<Postel@USC-ISIF.ARPA>
         R: 250 OK

         S: RCPT TO:<eric@BERKELEY.ARPA>
         R: 250 OK

         S: DATA
         R: 354 Start mail input; end with <CRLF>.<CRLF>
         S: Blah blah blah...
         S: ...etc. etc. etc.
         S: .
         R: 250 OK

         S: QUIT
         R: 221 BERKELEY.ARPA Service closing transmission channel

                              Scenario 10

      -------------------------------------------------------------

      Note that a real implementation must handle many recipients as
      specified in Section 4.5.3.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

GLOSSARY

   ASCII

      American Standard Code for Information Interchange [1].

   command

      A request for a mail service action sent by the sender-SMTP to the
      receiver-SMTP.

   domain

      The hierarchially structured global character string address of a
      host computer in the mail system.

   end of mail data indication

      A special sequence of characters that indicates the end of the
      mail data.  In particular, the five characters carriage return,
      line feed, period, carriage return, line feed, in that order.

   host

      A computer in the internetwork environment on which mailboxes or
      SMTP processes reside.

   line

      A a sequence of ASCII characters ending with a <CRLF>.

   mail data

      A sequence of ASCII characters of arbitrary length, which conforms
      to the standard set in the Standard for the Format of ARPA
      Internet Text Messages (RFC 822 [2]).

   mailbox

      A character string (address) which identifies a user to whom mail
      is to be sent.  Mailbox normally consists of the host and user
      specifications.  The standard mailbox naming convention is defined
      to be "user@domain".  Additionally, the "container" in which mail
      is stored.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

   receiver-SMTP process

      A process which transfers mail in cooperation with a sender-SMTP
      process.  It waits for a connection to be established via the
      transport service.  It receives SMTP commands from the
      sender-SMTP, sends replies, and performs the specified operations.

   reply

      A reply is an acknowledgment (positive or negative) sent from
      receiver to sender via the transmission channel in response to a
      command.  The general form of a reply is a completion code
      (including error codes) followed by a text string.  The codes are
      for use by programs and the text is usually intended for human
      users.

   sender-SMTP process

      A process which transfers mail in cooperation with a receiver-SMTP
      process.  A local language may be used in the user interface
      command/reply dialogue.  The sender-SMTP initiates the transport
      service connection.  It initiates SMTP commands, receives replies,
      and governs the transfer of mail.

   session

      The set of exchanges that occur while the transmission channel is
      open.

   transaction

      The set of exchanges required for one message to be transmitted
      for one or more recipients.

   transmission channel

      A full-duplex communication path between a sender-SMTP and a
      receiver-SMTP for the exchange of commands, replies, and mail
      text.

   transport service

      Any reliable stream-oriented data communication services.  For
      example, NCP, TCP, NITS.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   user

      A human being (or a process on behalf of a human being) wishing to
      obtain mail transfer service.  In addition, a recipient of
      computer mail.

   word

      A sequence of printing characters.

   <CRLF>

      The characters carriage return and line feed (in that order).

   <SP>

      The space character.

RFC 821                                                      August 1982
                                           Simple Mail Transfer Protocol

REFERENCES

   [1]  ASCII

      ASCII, "USA Code for Information Interchange", United States of
      America Standards Institute, X3.4, 1968.  Also in:  Feinler, E.
      and J. Postel, eds., "ARPANET Protocol Handbook", NIC 7104, for
      the Defense Communications Agency by SRI International, Menlo
      Park, California, Revised January 1978.

   [2]  RFC 822

      Crocker, D., "Standard for the Format of ARPA Internet Text
      Messages," RFC 822, Department of Electrical Engineering,
      University of Delaware, August 1982.

   [3]  TCP

      Postel, J., ed., "Transmission Control Protocol - DARPA Internet
      Program Protocol Specification", RFC 793, USC/Information Sciences
      Institute, NTIS AD Number A111091, September 1981.  Also in:
      Feinler, E. and J. Postel, eds., "Internet Protocol Transition
      Workbook", SRI International, Menlo Park, California, March 1982.

   [4]  NCP

      McKenzie,A., "Host/Host Protocol for the ARPA Network", NIC 8246,
      January 1972.  Also in:  Feinler, E. and J. Postel, eds., "ARPANET
      Protocol Handbook", NIC 7104, for the Defense Communications
      Agency by SRI International, Menlo Park, California, Revised
      January 1978.

   [5]  Initial Connection Protocol

      Postel, J., "Official Initial Connection Protocol", NIC 7101,
      11 June 1971.  Also in:  Feinler, E. and J. Postel, eds., "ARPANET
      Protocol Handbook", NIC 7104, for the Defense Communications
      Agency by SRI International, Menlo Park, California, Revised
      January 1978.

   [6]  NITS

      PSS/SG3, "A Network Independent Transport Service", Study Group 3,
      The Post Office PSS Users Group, February 1980.  Available from
      the DCPU, National Physical Laboratory, Teddington, UK.

August 1982                                                      RFC 821
Simple Mail Transfer Protocol                                          

   [7]  X.25

      CCITT, "Recommendation X.25 - Interface Between Data Terminal
      Equipment (DTE) and Data Circuit-terminating Equipment (DCE) for
      Terminals Operating in the Packet Mode on Public Data Networks,"
      CCITT Orange Book, Vol. VIII.2, International Telephone and
      Telegraph Consultative Committee, Geneva, 1976.

저작자 표시 비영리 변경 금지
신고

'02 .administration > vi. Mail' 카테고리의 다른 글

mstone (mailclient test & stress tool)  (0) 2011.02.22
RFC 3501 (imap4 protocol)  (0) 2009.08.31
RFC 1939 (pop3 protocol)  (0) 2009.08.31
SMTP protocol (RFC 821) - (ESMTP는 RFC 1821)  (0) 2009.08.10
Posted by 신머루군

티스토리 툴바