Network Working Group                                      N. Borenstein
Request for Comments: 1524                                      Bellcore
Category: Informational                                   September 1993


                 A User Agent Configuration Mechanism
                For Multimedia Mail Format Information

Status of This Memo

  This memo provides information for the Internet community.  It does
  not specify an Internet standard.  Distribution of this memo is
  unlimited.

Abstract

  This memo suggests a file format to be used to inform multiple mail
  reading user agent programs about the locally-installed facilities
  for handling mail in various formats.  The mechanism is explicitly
  designed to work with mail systems based Internet mail as defined by
  RFC's 821 (STD 10), 822 (STD 11), 934, 1049 (STD 11), 1113, and the
  Multipurpose Internet Mail Extensions, known as MIME.  However, with
  some extensions it could probably be made to work for X.400-based
  mail systems as well.  The format and mechanism are proposed in a
  manner that is generally operating-system independent.  However,
  certain implementation details will inevitably reflect operating
  system differences, some of which will have to be handled in a
  uniform manner for each operating system.  This memo makes such
  situations explicit, and, in an appendix, suggests a standard
  behavior under the UNIX operating system.

Introduction

  The electronic mail world is in the midst of a transition from
  single-part text-only mail to multi-part, multi-media mail.  In
  support of this transition, various extensions to RFC 821 and RFC 822
  have been proposed and/or adopted, notably including MIME [RFC-1521].
  Various parties have demonstrated extremely high-functionality
  multimedia mail, but the problem of mail interchange between
  different user agents has been severe.  In general, only text
  messages have been shared between user agents that were not
  explicitly designed to work together.  This limitation is not
  compatible with a smooth transition to a multi-media mail world.

  One approach to this transition is to modify diverse sets of mail
  reading user agents so that, when they need to display mail of an
  unfamiliar (non-text) type, they consult an external file for
  information on how to display that file.  That file might say, for



Borenstein                                                      [Page 1]

RFC 1524             Multimedia Mail Configuration        September 1993


  example, that if the content-type of a message is "foo" it can be
  displayed to the user via the "displayfoo" program.

  This approach means that, with a one-time modification, a wide
  variety of mail reading programs can be given the ability to display
  a wide variety of types of message.  Moreover, extending the set of
  media types supported at a site becomes a simple matter of installing
  a binary and adding a single line to a configuration file.  Crucial
  to this scheme, however, is that all of the user agents agree on a
  common representation and source for the configuration file.  This
  memo proposes such a common representation.

Location of Configuration Information

  Each user agent must clearly obtain the configuration information
  from a common location, if the same information is to be used to
  configure all user agents.  However, individual users should be able
  to override or augment a site's configuration.  The configuration
  information should therefore be obtained from a designated set of
  locations.  The overall configuration will be obtained through the
  virtual concatenation of several individual configuration files known
  as mailcap files.  The configuration information will be obtained
  from the FIRST matching entry in a mailcap file, where "matching"
  depends on both a matching content-type specification, an entry
  containing sufficient information for the purposes of the application
  doing the searching, and the success of any test in the "test="
  field, if present.

  The precise location of the mailcap files is operating-system
  dependent.  A standard location for UNIX is specified in Appendix A.

Overall Format of a Mailcap File

  Each mailcap file consists of a set of entries that describe the
  proper handling of one media type at the local site.

  For example, one line might tell how to display a message in Group
  III fax format.  A mailcap file consists of a sequence of such
  individual entries, separated by newlines (according to the operating
  system's newline conventions). Blank lines and lines that start with
  the "#" character (ASCII 35) are considered comments, and are
  ignored.  Long entries may be continued on multiple lines if each
  non-terminal line ends with a backslash character ('\', ASCII 92), in
  which case the multiple lines are to be treated as a single mailcap
  entry.  Note that for such "continued" lines, the backslash must be
  the last character on the line to be continued.





Borenstein                                                      [Page 2]

RFC 1524             Multimedia Mail Configuration        September 1993


  Thus the overall format of a mailcap file is given, in the modified
  BNF of RFC 822, as:

        Mailcap-File = *Mailcap-Line

        Mailcap-Line = Comment / Mailcap-Entry

        Comment = NEWLINE  /  "#" *CHAR NEWLINE

        NEWLINE = <newline as defined by OS convention>

  Note that the above specification implies that comments must appear
  on lines all to themselves, with a "#" character as the first
  character on each comment line.

Format of a Mailcap Entry

  Each mailcap entry consists of a number of fields, separated by
  semi-colons.  The first two fields are required, and must occur in
  the specified order.  The remaining fields are optional, and may
  appear in any order.

  The first field is the content-type, which indicates the type of data
  this mailcap entry describes how to handle.  It is to be matched
  against the type/subtype specification in the "Content-Type" header
  field of an Internet mail message.  If the subtype is specified as
  "*", it is intended to match all subtypes of the named content-type.

  The second field, view-command, is a specification of how the message
  or body part can be viewed at the local site.  Although the syntax of
  this field is fully specified, the semantics of program execution are
  necessarily somewhat operating system dependent.  UNIX semantics are
  given in Appendix A.

  The optional fields, which may be given in any order, are as follows:

  -- The "compose" field may be used to specify a program that can be
     used to compose a new body or body part in the given format.  Its
     intended use is to support mail composing agents that support the
     composition of multiple types of mail using external composing
     agents.  As with the view-command, the semantics of program
     execution are operating system dependent, with UNIX semantics
     specified in Appendix A.  The result of the composing program may
     be data that is not yet suitable for mail transport -- that is, a
     Content-Transfer-Encoding may need to be applied to the data.

  -- The "composetyped" field is similar to the "compose" field, but is
     to be used when the composing program needs to specify the



Borenstein                                                      [Page 3]

RFC 1524             Multimedia Mail Configuration        September 1993


     Content-type header field to be applied to the composed data.  The
     "compose" field is simpler, and is preferred for use with existing
     (non-mail-oriented) programs for composing data in a given format.
     The "composetyped" field is necessary when the Content-type
     information must include auxilliary parameters, and the
     composition program must then know enough about mail formats to
     produce output that includes the mail type information.

  -- The "edit" field may be used to specify a program that can be used
     to edit a body or body part in the given format.  In many cases,
     it may be identical in content to the "compose" field, and shares
     the operating-system dependent semantics for program execution.

  -- The "print" field may be used to specify a program that can be
     used to print a message or body part in the given format.  As with
     the view-command, the semantics of program execution are operating
     system dependent, with UNIX semantics specified in Appendix A.

  -- The "test" field may be used to test some external condition
     (e.g., the machine architecture, or the window system in use) to
     determine whether or not the mailcap line applies.  It specifies a
     program to be run to test some condition.  The semantics of
     execution and of the value returned by the test program are
     operating system dependent, with UNIX semantics specified in
     Appendix A.  If the test fails, a subsequent mailcap entry should
     be sought.  Multiple test fields are not permitted -- since a test
     can call a program, it can already be arbitrarily complex.

  -- The "needsterminal" field indicates that the view-command must be
     run on an interactive terminal.  This is needed to inform window-
     oriented user agents that an interactive terminal is needed.  (The
     decision is not left exclusively to the view-command because in
     some circumstances it may not be possible for such programs to
     tell whether or not they are on interactive terminals.)  The
     needsterminal command should be assumed to apply to the compose
     and edit commands, too, if they exist.  Note that this is NOT a
     test -- it is a requirement for the environment in which the
     program will be executed, and should typically cause the creation
     of a terminal window when not executed on either a real terminal
     or a terminal window.

  -- The "copiousoutput" field indicates that the output from the
     view-command will be an extended stream of output, and is to be
     interpreted as advice to the UA (User Agent mail-reading program)
     that the output should be either paged or made scrollable. Note
     that it is probably a mistake if needsterminal and copiousoutput
     are both specified.




Borenstein                                                      [Page 4]

RFC 1524             Multimedia Mail Configuration        September 1993


  -- The "description" field simply provides a textual description,
     optionally quoted, that describes the type of data, to be used
     optionally by mail readers that wish to describe the data before
     offering to display it.

  -- The "textualnewlines" field, if set to any non-zero value,
     indicates that this type of data is line-oriented and that, if
     encoded in base64, all newlines should be converted to canonical
     form (CRLF) before encoding, and will be in that form after
     decoding.  In general, this field is needed only if there is
     line-oriented data of some type other than text/* or non-line-
     oriented data that is a subtype of text.

  -- The "x11-bitmap" field names a file, in X11 bitmap (xbm) format,
     which points to an appropriate icon to be used to visually denote
     the presence of this kind of data.

  -- The "nametemplate" field gives a file name format, in which %s
     will be replaced by a short unique string to give the name of the
     temporary file to be passed to the viewing command.  This is only
     expected to be relevant in environments where filename extensions
     are meaningful, e.g., one coulld specify that a GIF file being
     passed to a gif viewer should have a name eding in ".gif" by using
     "nametemplate=%s.gif".

  Any other fields beginning with "x-" may be included for local or
  mailer-specific extensions of this format.  Implementations should
  simply ignore all such unrecognized fields to permit such extensions,
  some of which might be standardized in a future version of this
  document.

  Some of the fields above, such as "needsterminal", apply to the
  actions of the view-command, edit-command, and compose-command,
  alike.  In some unusual cases, this may not be desirable, but
  differentiation can be accomplished via separate mailcap entries,
  taking advantage of the fact that subsequent mailcap entries are
  searched if an earlier mailcap entry does not provide enough
  information:

      application/postscript; ps-to-terminal %s;\ needsterminal
      application/postscript; ps-to-terminal %s; \compose=idraw %s

  In RFC 822 modified BNF, the following grammar describes a mailcap
  entry:







Borenstein                                                      [Page 5]

RFC 1524             Multimedia Mail Configuration        September 1993


        Mailcap-Entry = typefield ; view-command
                            [";" 1#field]

        typefield = propertype / implicit-wild

        propertype = type "/" wildsubtype

        implicitwild = type

        wildsubtype = subtype / "*"

        view-command = mtext

        mtext = *mchar

        mchar = schar / qchar

        schar = * <any CHAR except ";","\", and CTLS>

        qchar = "\" CHAR ; may quote any char

        field = flag / namedfield

        namedfield = fieldname "=" mtext

        flag = "needsterminal"   ; All these literals are to
             / "copiousoutput"   ; be interpreted as
             / x-token           ; case-insensitive

        fieldname =    / "compose"      ;Also all of these
                       / "composetyped" ;are case-insensitive.
                       / "print"
                       / "edit"
                       / "test"
                       / "x11-bitmap"
                       / "textualnewlines"
                       / "description"
                       / x-token

  Note that "type", "subtype", and "x-token" are defined in MIME.  Note
  also that while the definition of "schar" includes the percent sign,
  "%", this character has a special meaning in at least the UNIX
  semantics, and will therefore need to be quoted as a qchar to be used
  literally.







Borenstein                                                      [Page 6]

RFC 1524             Multimedia Mail Configuration        September 1993


Acknowledgements

  The author wishes to thank Malcolm Bjorn Gillies, Dan Heller, Olle
  Jaernefors, Keith Moore, Luc Rooijakkers, and the other members of
  the IETF task force on mail extensions for their comments on earlier
  versions of this draft.  If other acknowledgements were neglected,
  please let me know, as it was surely accidental.

Security Considerations

  Security issues are not discussed in this memo.  However, the use of
  the mechanisms described in this memo can make it easier for
  implementations to slip into the kind of security problems discussed
  in the MIME document.  Implementors and mailcap administrators should
  be aware of these security considerations, and in particular should
  exercise caution in the choice of programs to be listed in a mailcap
  file for automatic execution.

Author's Address

  Nathaniel S. Borenstein
  MRE 2D-296, Bellcore
  445 South St.
  Morristown, NJ 07962-1910

  EMail: [email protected]
  Phone: +1 201 829 4270
  Fax:  +1 201 829 7019























Borenstein                                                      [Page 7]

RFC 1524             Multimedia Mail Configuration        September 1993


Appendix A:  Implementation Details for UNIX

  Although this memo fully specifies a syntax for "mailcap" files, the
  semantics of the mailcap file are of necessity operating-system
  dependent in four respects.  In order to clarify the intent, and to
  promote a standard usage, this appendix proposes a UNIX semantics for
  these four cases.  If a mailcap mechanism is implemented on non-UNIX
  systems, similar semantic decisions should be made and published.

Location of the Mailcap File(s)

  For UNIX, a path search of mailcap files is specified.  The default
  path search is specified as including at least the following:

  $HOME/.mailcap:/etc/mailcap:/usr/etc/mailcap:/usr/local/etc/mailcap

  However, this path may itself be overridden by a path specified by
  the MAILCAPS environment variable.

Semantics of executable commands

  Several portions of a mailcap entry specify commands to be executed.
  In particular, the mandatory second fie ld, the view-command, takes a
  command to be executed, as do the optional print, edit, test, and
  compose fields.

  On a UNIX system, such commands will each be a full shell command
  line, including the path name for a program and its arguments.
  (Because of differences in shells and the implementation and behavior
  of the same shell from one system to another, it is specified that
  the command line be intended as input to the Bourne shell, i.e., that
  it is implicitly preceded by "/bin/sh -c " on the command line.)

  The two characters "%s", if used, will be replaced by the name of a
  file for the actual mail body data.  In the case of the edit adn
  view-command, the body part will be passed to this command as
  standard input unless one or more instances of "%s" appear in the
  view-command, in which case %s will be replaced by the name of a file
  containing the body part, a file which may have to be created before
  the view-command program is executed.  (Such files cannot be presumed
  to continue to exist after the view-command program exits.  Thus a
  view-command that wishes to exit and continue processing in the
  background should take care to save the data first.)  In the case of
  the compose and composetyped commands, %s should be replaced by the
  name of a file to which the composed data should be written by the
  programs named in the compose or composedtyped commands.  Thus, the
  calling program will look in that file later in order to retrieve the
  composed data. If %s does not appear in the compose or composetyped



Borenstein                                                      [Page 8]

RFC 1524             Multimedia Mail Configuration        September 1993


  commands, then the composed data will be assumed to be written by the
  composing programs to standard output.

  Furthermore, any occurrence of "%t" will be replaced by the content-
  type and subtype specification.  (That is, if the content-type is
  "text/plain", then %t will be replaced by "text/plain".)  A literal %
  character may be quoted as \%.  Finally, named parameters from the
  Content-type field may be placed in the command execution line using
  "%{" followed by the parameter name and a closing "}" character.  The
  entire parameter should appear as a single command line argument,
  regardless of embedded spaces.  Thus, if the message has a Content-
  type line of:

        Content-type:  multipart/mixed; boundary=42

  and the mailcap file has a line of:

        multipart/*; /usr/local/bin/showmulti \
          %t %{boundary}

  then the equivalent  of  the  following  command  should  be
  executed:

       /usr/local/bin/showmulti multipart/mixed 42

  If the content-type is "multipart" (any subtype), then the two
  characters "%n" will be replaced by an integer giving the number of
  sub-parts within the multipart entity.  Also, the two characters "%F"
  will be replaced by a set of arguments, twice as many arguments as
  the number of sub-parts, consisting of alternating content-types and
  file names for each part in turn.  Thus if multipart entity has three
  parts, "%F" will be replaced by the equivalent of "content-type1
  file-name1 content-type2 file-name2 content-type3 file-name3".

Semantics of the "test" field

  The "test" field specifies a program to be used to test whether or
  not the current mailcap line applies.  This can be used, for example,
  to have a mailcap line that only applies if the X window system is
  running, or if the user is running on a SPARCstation with a
  /dev/audio.  The value of the "test" field is a program to run to
  test such a condition.  The precise program to run and arguments to
  give it are determined as specified in the previous section.  The
  test program should return an exit code of zero if the condition is
  true, and a non-zero code otherwise.






Borenstein                                                      [Page 9]

RFC 1524             Multimedia Mail Configuration        September 1993


Semantics of the "compose" field

  On UNIX, the composing program is expected to produce a data stream
  for such a body part as its standard output.  The program will be
  executed with the command line arguments determined as specified
  above.  The data returned via its standard output will be given a
  Content-Type field that has no supplementary parameters.  For
  example, the following mailcap entry:

       audio/basic; /usr/local/bin/showaudio %t
         compose = /usr/local/bin/recordaudio

  would result in tagging the data composed by the "recordaudio"
  program as:

       Content-Type: audio/basic

  If this is unacceptable -- for example, in the case of multipart mail
  a "boundary" parameter is required -- then the "compose" field cannot
  be used.  Instead, the "composetyped" field should be used in the
  mailcap file.

Semantics of the "composetyped" field

  The "composetyped" filed is much like the "compose" field, except
  that it names a composition program that produces, not raw data, but
  data that includes a MIME-conformant type specification.  The program
  will be executed with the command line arguments determined as
  specified above.  The data returned via its standard output must
  begin with a Content-Type header, followed optionally by other
  Content-* headers, and then by a blank line and the data.  For
  example, the following mailcap entry:

       multipart/mixed; /usr/local/bin/showmulti %t \
         %{boundary}; \
         composetyped = /usr/local/bin/makemulti

  would result in executing the "makemulti" program, which would be
  expected to begin its output with a line of the form:

       Content-Type: multipart/mixed; boundary=foobar

  Note that a composition program need not encode binary data in base64
  or quoted-printable. It remains the responsibility of the software
  calling the composition program to encode such data as necessary.
  However, if a composing program does encode data, which is not
  encouraged, it should announce that fact using a Content-Transfer-
  Encoding header in the standard manner defined by MIME.  Because such



Borenstein                                                     [Page 10]

RFC 1524             Multimedia Mail Configuration        September 1993


  encodings must be announced by such a header, they are an option only
  for composetyped programs, not for compose programs.

Appendix B: Sample Mailcap File

  The following is an example of a mailcap file for UNIX that
  demonstrates most of the syntax above.  It contains explanatory
  comments where necessary.

        # Mailcap file for Bellcore lab 214.
        #
        # The next line sends "richtext" to the richtext
        program
        text/richtext; richtext %s; copiousoutput
        #
        # Next, basic u-law audio
        audio/*; showaudio; test=/usr/local/bin/hasaudio
        #
        # Next, use the xview program to handle several image
        formats
        image/*; xview %s; test=/usr/local/bin/RunningX
        #
        # The ATOMICMAIL interpreter uses curses, so needs a
        terminal
        application/atomicmail; /usr/local/bin/atomicmail %s; \
            needsterminal
        #
        # The next line handles Andrew format,
        #   if ez and ezview are installed
        x-be2; /usr/andrew/bin/ezview %s; \
           print=/usr/andrew/bin/ezprint %s ; \
           compose=/usr/andrew/bin/ez -d %s \;
           edit=/usr/andrew/bin/ez -d %s; \;
           copiousoutput
        #
        # The next silly example demonstrates the use of
        quoting
        application/*; echo "This is \"%t\" but \
           is 50 \% Greek to me" \; cat %s; copiousoutput












Borenstein                                                     [Page 11]

RFC 1524             Multimedia Mail Configuration        September 1993


Appendix C:  A Note on Format Translation

  It has been suggested that another function of a mailcap-like
  mechanism might be to specify the locally available tools for
  document format translation.  For example, the file could designate a
  program for translating from format A to format B, another for
  translating from format B to format C, and finally a mechanism for
  displaying format C.  Although this mechanism would be somewhat
  richer than the current mailcap file, and might conceivably also have
  utility at the message transport layer, it significantly complicates
  the processing effort necessary for a user agent that simply wants to
  display a message in format A.  Using the current, simpler, mailcap
  scheme, a single line could tell such a user agent to display A-
  format mail using a pipeline of translators and the C-format viewer.
  This memo resists the temptation to complicate the necessary
  processing for a user agent to accomplish this task.  Using the
  mailcap format defined here, it is only necessary to find the correct
  single line in a mailcap file, and to execute the command given in
  that line.

References

    [RFC-822] Crocker, D., "Standard for the format of ARPA Internet
    text messages", STD 11, RFC 822, UDEL, August 1982.

    [RFC-1521] Borenstein, N., and N.  Freed, "MIME (Multipurpose
    Internet Mail Extensions) Part One: Mechanisms for Specifying and
    Describing the Format of Internet Message Bodies", RFC 1521,
    Bellcore, Innosoft, September 1993.






















Borenstein                                                     [Page 12]