# pm-jamime-decode.rc -- decode MIME body contents; quoted-printable, base64
#
#   File id
#
#       Copyright (C) 1997-2010 Jari Aalto
#
#       This program is free software; you can redistribute it and/or
#       modify it under the terms of the GNU General Public License as
#       published by the Free Software Foundation; either version 2 of the
#       License, or (at your option) any later version
#
#       This program is distributed in the hope that it will be useful, but
#       WITHOUT ANY WARRANTY; without even the implied warranty of
#       MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
#       General Public License for more details at
#       <http://www.gnu.org/copyleft/gpl.html>.
#
#   Documentation
#
#       The original father of the decoding scheme used here was
#       presented by Peter Galbraith <galbraith@mixing.qc.dfo.ca> in
#       procmail mailing list somewhere at the end of 1997.
#
#       This subroutine supposes that the header has MIME header
#       Content-Type: text/plain and performs quoted-printable or
#       base64 decoding on the whole message. Note, that if you
#       receive messages that have many mime attachments, then this
#       recipe is not suitable for it.
#
#       Procmail is *not* designed to handle mime attachments and this
#       recipe only applies to whole _body_.
#
#       The `pm-jamime-*.rc' is really stretching the limits and any
#       serious work should be delegated to other tools. Alternatives:
#
#       o   A Perl MIME module which will allow you to manipulate MIME
#           body parts rather elegantly. See
#           http://www.perl.com/CPAN-local/authors/Eryq/ for MIME-tools.
#       o   mimedecode at ftp://ftp.dde.dk/pub/mimedecode.c
#
#   Notes
#
#       Perl or Python is not used, because both are CPU intensive. It
#       would be too expansive for accounts or environments receiving
#       hundreds of mails per day (like from several mailing lists).
#
#       RFC 2047 gives possiblity to use MIME iso-8859-1 extensions
#       for mail headers.
#
#           Subject: Re: [PIC]: RSA =?iso-8859-1?Q?encryption=B7=B7?=
#           Subject: =?iso-8859-1?Q?=5BEE=5D:TV_&_video_IC=B4s_!!?=
#
#       There is also base64 possibility (although rare):
#
#           Subject: =?iso-8859-1?B?zvLi5fI6ICAgICAgTVBMQUIzLjQw?=
#
#       In worst possible case there is even multiple ISO encoded
#       strings in subject. Yes, this is valid, the continued line
#       includes spaces at front to keep it with original just like
#       in `Received:' headers. This subroutine will not touch headers
#       that have multiple ISO tags - procmail is too limited for that.
#
#           Subject: AW: Re: AW: neue =3D?ISO-8859-1?Q?M=3DF6glichkeiten_=3D28was_=3D=C4hn?=3D
#              =3D?ISO-8859-1?Q?lichkeiten_von_=3DDCbungen=3D29?=3D
#
#   Required settings
#
#       Variable `PMSRC' must point to source directory of procmail code.
#       This subroutine will include
#
#       o   pm-javar.rc, pm-jamime.rc
#       o   Programs `$MIME_BIN', `$MIME_BIN_QP' and `$MIME_BIN_64'
#           must have been installed (see pm-javar.rc).
#
#   Call arguments (variables to set before calling)
#
#       o   `JA_MIME_DECODE_TREAT_SUBJECT', default "yes". Decode
#           Subject header by removing mime.
#       o   `JA_MIME_DECODE_TREAT_FROM', default "no". Decode
#           From header by removing mime.
#       o   `JA_MIME_DECODE_TREAT_BODY', default "no". Decode
#           body of message by removing quoted-printable from a
#           message that contains only one part. Messages with multiple
#           parts are not handled.
#
#   Return values
#
#       o    `PM_JAMIME_COMPLEX_SUBJECT' is set to "yes".
#            This flag is set to indicate that some other program
#            should handle the message. If Subject header contains ISO
#            encoding several times, it cannot be handled by this module.
#
#   Examples
#
#       Instead of testing the existence of text/plain in the body,
#       you can force decoding by settings JA_MIME_DECODE_REGEXP to
#       ".*".
#
#           RC_MIME_DECODE = $PMSRC/pm-jamime-decode.rc
#
#           :0
#           * condition
#           {
#               JA_MIME_DECODE_REGEXP = ".*"
#           }
#
#           INCLUDERC = $RC_MIME_DECODE     # call subroutine.
#
#
#   Change Log (none)


dummy = "
========================================================================
pm-jamime-decode.rc: init:
"

:0
* !  WSPC ?? [ ]
{
    INCLUDERC = $PMSRC/pm-javar.rc
}

:0
* !  MIME_VER ?? [0-9]
{
    INCLUDERC = $PMSRC/pm-jamime.rc
}

# .......................................................... &public ...
# User configurable sections

JA_MIME_DECODE_TREAT_SUBJECT = "yes"   # Set to "no"  to disable
JA_MIME_DECODE_TREAT_FROM    = "no"    # Set to "yes" to enable
JA_MIME_DECODE_TREAT_BODY    = "no"    # Set to "yes" to enable

JA_MIME_DECODE_REGEXP = ${JA_MIME_DECODE_REGEXP:-\
"^Content-Type: *text/plain"}

# ........................................................... &do-it ...
#   Run conversion if it was quoted printable.
#   Also reflect correct MIME header

JA_MIME_DECODE_HEADER = ${JA_MIME_DECODE_HEADER:-\
"X-Mime-Header-Decoded"}

dummy = "pm-jamime-decode.rc: handle quoted printable"

PM_JAMIME_COMPLEX_SUBJECT       # Return value if too complex subject

jamimeHandledSubject = "no"

#  Kill variables
jamimeSubject
jamimeSubjectRest

:0
*  JA_MIME_DECODE_TREAT_SUBJECT ?? yes
* ^Subject: \/.*
{
    jamimeSubject = $MATCH
    jamimeISO     = "\?iso-8859-[1-9]\?[QB]\?.+\?="

    :0
    *$ jamimeSubject ?? $jamimeISO()\/.*
    {
        #  ".*" is actually minimum match. See pm-doc.sf.net
        jamimeSubjectRest = $MATCH
    }

    :0
    *$ jamimeSubjectRest ?? $jamimeISO()
    {
        #  This Subejct line contains ISO encoding several times.
        #  This is out of our league. Can't hanle it, so quit.

        jamimeSubject
        JAMIME_COMPLEX_SUBJECT = "yes"
    }
}

:0
* ! jamimeSubject ?? ^^^^
{
    str     = $jamimeSubject
    decoder = $MIME_BIN_QP
    type    = "quoted-printable"

    :0
    * ^Subject:.*\?B\?
    {
        type    = "base64"
        decoder = $MIME_BIN_64
    }

   #   NOTE "?" is not wildcard "as is" in sed.

    clean =                                                         \
    `   echo "$str"                                                 \
        | $SED                                                      \
          -e 's/^\(.*\)?\(iso\|ISO\)-\{0,1\}8859-[1-9]?[QB]?\(.*[^?]\)?=/\1\2/g'   \
          -e 's/=0D//g'                                             \
          -e 's/=0A//g'                                             \
          -e 's/=B7//g'                                             \
          -e 's/_/ /g'                                              \
        | $decoder
    `

    handledSubject = "yes"

    :0 fhw
    * ! clean ?? ^^^^
    | $FORMAIL                                                      \
        -i "Subject: $clean"                                        \
        -I "${JA_MIME_DECODE_HEADER}-Subject: $type"
}

#  Too bad Procmail does not have subroutines. This recipe is
#  identical to "Subject" check above.
#
#  Note, that 'From' header is different from the 'Subject' header
#  in respect to the ISO encoding. Look closely where '?=' ends:
#
#    Subject: =?ISO-8859-1?Q?=C4hnlichkeiten_von_=DCbungen?=
#    From: =?ISO-8859-1?Q?Holger_Hoffst=E4tte?= <holger@example.com>

jamimeHandledFrom = "no"

:0
*  JA_MIME_DECODE_TREAT_FROM ?? yes
*$ ^From: +()\/=\?iso-8859-[1-9]\?[QB]\?.+\?=.*
*$ ^From: +()\/.*=\?.+[>]
{
    str     = $MATCH
    decoder = $MIME_BIN_QP
    type    = "quoted-printable"

    :0
    * ^Subject:.*\?B\?
    {
        type    = "base64"
        decoder = $MIME_BIN_64
    }

    clean =                                                         \
    `   echo "$str"                                                 \
        | $SED                                                      \
          -e 's/=?\(iso\|ISO\)-\{0,1\}8859-[1-9]?[QB]?\(.*[^?]\)?=/\2/g'   \
          -e 's/=0D//g'                                             \
          -e 's/=0A//g'                                             \
          -e 's/=B7//g'                                             \
        | $decoder
    `

    jamimeHandledFrom = "yes"

    #  Make sure there is <send@example.com> before replacing anything

    :0 fhw
    * clean ?? [>]
    | $FORMAIL							    \
	-i "From: $clean"					    \
	-i "${JA_MIME_DECODE_HEADER}-From: $type"
}

# Touch only real mime messages and text/plain
# $MIME_BIN_QP does not handle separate MIME sections

:0
*    JA_MIME_DECODE_TREAT_BODY  ?? yes
*$   $JA_MIME_DECODE_REGEXP
{
    :0 fbw
    * ^Content-Transfer-Encoding: *quoted-printable
    | $MIME_BIN_QP

        :0 A fhw
        | $FORMAIL -I "Content-Transfer-Encoding: 8bit"

    :0 fbw
    * ^Content-Transfer-Encoding: *base64
    | $MIME_BIN_64

        :0 A fhw
        | $FORMAIL -I "Content-Transfer-Encoding: 8bit"
}

dummy = "pm-jamime-decode.rc: end:"

# End of file pm-jamime-decode.rc