Doc/lib/email.tex

% Copyright (C) 2001,2002 Python Software Foundation
% Author: barry@zope.com (Barry Warsaw)

\section{\module{email} ---
	 An email and MIME handling package}

\declaremodule{standard}{email}
\modulesynopsis{Package supporting the parsing, manipulating, and
    generating email messages, including MIME documents.}
\moduleauthor{Barry A. Warsaw}{barry@zope.com}
\sectionauthor{Barry A. Warsaw}{barry@zope.com}

\versionadded{2.2}

The \module{email} package is a library for managing email messages,
including MIME and other \rfc{2822}-based message documents.  It
subsumes most of the functionality in several older standard modules
such as \refmodule{rfc822}, \refmodule{mimetools},
\refmodule{multifile}, and other non-standard packages such as
\module{mimecntl}.  It is specifically \emph{not} designed to do any
sending of email messages to SMTP (\rfc{2821}) servers; that is the
function of the \refmodule{smtplib} module.  The \module{email}
package attempts to be as RFC-compliant as possible, supporting in
addition to \rfc{2822}, such MIME-related RFCs as
\rfc{2045}-\rfc{2047}, and \rfc{2231}.

The primary distinguishing feature of the \module{email} package is
that it splits the parsing and generating of email messages from the
internal \emph{object model} representation of email.  Applications
using the \module{email} package deal primarily with objects; you can
add sub-objects to messages, remove sub-objects from messages,
completely re-arrange the contents, etc.  There is a separate parser
and a separate generator which handles the transformation from flat
text to the object model, and then back to flat text again.  There
are also handy subclasses for some common MIME object types, and a few
miscellaneous utilities that help with such common tasks as extracting
and parsing message field values, creating RFC-compliant dates, etc.

The following sections describe the functionality of the
\module{email} package.  The ordering follows a progression that
should be common in applications: an email message is read as flat
text from a file or other source, the text is parsed to produce the
object structure of the email message, this structure is manipulated,
and finally rendered back into flat text.

It is perfectly feasible to create the object structure out of whole
cloth --- i.e. completely from scratch.  From there, a similar
progression can be taken as above.

Also included are detailed specifications of all the classes and
modules that the \module{email} package provides, the exception
classes you might encounter while using the \module{email} package,
some auxiliary utilities, and a few examples.  For users of the older
\module{mimelib} package, or previous versions of the \module{email}
package, a section on differences and porting is provided.

\begin{seealso}
    \seemodule{smtplib}{SMTP protocol client}
\end{seealso}

\subsection{Representing an email message}
\input{emailmessage}

\subsection{Parsing email messages}
\input{emailparser}

\subsection{Generating MIME documents}
\input{emailgenerator}

\subsection{Creating email and MIME objects from scratch}
\input{emailmimebase}

\subsection{Internationalized headers}
\input{emailheaders}

\subsection{Representing character sets}
\input{emailcharsets}

\subsection{Encoders}
\input{emailencoders}

\subsection{Exception classes}
\input{emailexc}

\subsection{Miscellaneous utilities}
\input{emailutil}

\subsection{Iterators}
\input{emailiter}

\subsection{Differences from \module{email} v1 (up to Python 2.2.1)}

Version 1 of the \module{email} package was bundled with Python
releases up to Python 2.2.1.  Version 2 was developed for the Python
2.3 release, and backported to Python 2.2.2.  It was also available as
a separate distutils based package.  \module{email} version 2 is
almost entirely backward compatible with version 1, with the
following differences:

\begin{itemize}
\item The \module{email.Header} and \module{email.Charset} modules
      have been added.
\item The pickle format for \class{Message} instances has changed.
      Since this was never (and still isn't) formally defined, this
      isn't considered a backward incompatibility.  However if your
      application pickles and unpickles \class{Message} instances, be
      aware that in \module{email} version 2, \class{Message}
      instances now have private variables \var{_charset} and
      \var{_default_type}.
\item Several methods in the \class{Message} class have been
      deprecated, or their signatures changed.  Also, many new methods
      have been added.  See the documentation for the \class{Message}
      class for details.  The changes should be completely backward
      compatible.
\item The object structure has changed in the face of
      \mimetype{message/rfc822} content types.  In \module{email}
      version 1, such a type would be represented by a scalar payload,
      i.e. the container message's \method{is_multipart()} returned
      false, \method{get_payload()} was not a list object, but a single
      \class{Message} instance.

      This structure was inconsistent with the rest of the package, so
      the object representation for \mimetype{message/rfc822} content
      types was changed.  In \module{email} version 2, the container
      \emph{does} return \code{True} from \method{is_multipart()}, and
      \method{get_payload()} returns a list containing a single
      \class{Message} item.

      Note that this is one place that backward compatibility could
      not be completely maintained.  However, if you're already
      testing the return type of \method{get_payload()}, you should be
      fine.  You just need to make sure your code doesn't do a
      \method{set_payload()} with a \class{Message} instance on a
      container with a content type of \mimetype{message/rfc822}.
\item The \class{Parser} constructor's \var{strict} argument was
      added, and its \method{parse()} and \method{parsestr()} methods
      grew a \var{headersonly} argument.  The \var{strict} flag was
      also added to functions \function{email.message_from_file()}
      and \function{email.message_from_string()}.
\item \method{Generator.__call__()} is deprecated; use
      \method{Generator.flatten()} instead.  The \class{Generator}
      class has also grown the \method{clone()} method.
\item The \class{DecodedGenerator} class in the
      \module{email.Generator} module was added.
\item The intermediate base classes \class{MIMENonMultipart} and
      \class{MIMEMultipart} have been added, and interposed in the
      class hierarchy for most of the other MIME-related derived
      classes.
\item The \var{_encoder} argument to the \class{MIMEText} constructor
      has been deprecated.  Encoding  now happens implicitly based
      on the \var{_charset} argument.
\item The following functions in the \module{email.Utils} module have
      been deprecated: \function{dump_address_pairs()},
      \function{decode()}, and \function{encode()}.  The following
      functions have been added to the module:
      \function{make_msgid()}, \function{decode_rfc2231()},
      \function{encode_rfc2231()}, and \function{decode_params()}.
\item The non-public function \function{email.Iterators._structure()}
      was added.
\end{itemize}

\subsection{Differences from \module{mimelib}}

The \module{email} package was originally prototyped as a separate
library called
\ulink{\module{mimelib}}{http://mimelib.sf.net/}.
Changes have been made so that
method names are more consistent, and some methods or modules have
either been added or removed.  The semantics of some of the methods
have also changed.  For the most part, any functionality available in
\module{mimelib} is still available in the \refmodule{email} package,
albeit often in a different way.  Backward compatibility between
the \module{mimelib} package and the \module{email} package was not a
priority.

Here is a brief description of the differences between the
\module{mimelib} and the \refmodule{email} packages, along with hints on
how to port your applications.

Of course, the most visible difference between the two packages is
that the package name has been changed to \refmodule{email}.  In
addition, the top-level package has the following differences:

\begin{itemize}
\item \function{messageFromString()} has been renamed to
      \function{message_from_string()}.
\item \function{messageFromFile()} has been renamed to
      \function{message_from_file()}.
\end{itemize}

The \class{Message} class has the following differences:

\begin{itemize}
\item The method \method{asString()} was renamed to \method{as_string()}.
\item The method \method{ismultipart()} was renamed to
      \method{is_multipart()}.
\item The \method{get_payload()} method has grown a \var{decode}
      optional argument.
\item The method \method{getall()} was renamed to \method{get_all()}.
\item The method \method{addheader()} was renamed to \method{add_header()}.
\item The method \method{gettype()} was renamed to \method{get_type()}.
\item The method\method{getmaintype()} was renamed to
      \method{get_main_type()}.
\item The method \method{getsubtype()} was renamed to
      \method{get_subtype()}.
\item The method \method{getparams()} was renamed to
      \method{get_params()}.
      Also, whereas \method{getparams()} returned a list of strings,
      \method{get_params()} returns a list of 2-tuples, effectively
      the key/value pairs of the parameters, split on the \character{=}
      sign.
\item The method \method{getparam()} was renamed to \method{get_param()}.
\item The method \method{getcharsets()} was renamed to
      \method{get_charsets()}.
\item The method \method{getfilename()} was renamed to
      \method{get_filename()}.
\item The method \method{getboundary()} was renamed to
      \method{get_boundary()}.
\item The method \method{setboundary()} was renamed to
      \method{set_boundary()}.
\item The method \method{getdecodedpayload()} was removed.  To get
      similar functionality, pass the value 1 to the \var{decode} flag
      of the {get_payload()} method.
\item The method \method{getpayloadastext()} was removed.  Similar
      functionality
      is supported by the \class{DecodedGenerator} class in the
      \refmodule{email.Generator} module.
\item The method \method{getbodyastext()} was removed.  You can get
      similar functionality by creating an iterator with
      \function{typed_subpart_iterator()} in the
      \refmodule{email.Iterators} module.
\end{itemize}

The \class{Parser} class has no differences in its public interface.
It does have some additional smarts to recognize
\mimetype{message/delivery-status} type messages, which it represents as
a \class{Message} instance containing separate \class{Message}
subparts for each header block in the delivery status
notification\footnote{Delivery Status Notifications (DSN) are defined
in \rfc{1894}.}.

The \class{Generator} class has no differences in its public
interface.  There is a new class in the \refmodule{email.Generator}
module though, called \class{DecodedGenerator} which provides most of
the functionality previously available in the
\method{Message.getpayloadastext()} method.

The following modules and classes have been changed:

\begin{itemize}
\item The \class{MIMEBase} class constructor arguments \var{_major}
      and \var{_minor} have changed to \var{_maintype} and
      \var{_subtype} respectively.
\item The \code{Image} class/module has been renamed to
      \code{MIMEImage}.  The \var{_minor} argument has been renamed to
      \var{_subtype}.
\item The \code{Text} class/module has been renamed to
      \code{MIMEText}.  The \var{_minor} argument has been renamed to
      \var{_subtype}.
\item The \code{MessageRFC822} class/module has been renamed to
      \code{MIMEMessage}.  Note that an earlier version of
      \module{mimelib} called this class/module \code{RFC822}, but
      that clashed with the Python standard library module
      \refmodule{rfc822} on some case-insensitive file systems.

      Also, the \class{MIMEMessage} class now represents any kind of
      MIME message with main type \mimetype{message}.  It takes an
      optional argument \var{_subtype} which is used to set the MIME
      subtype.  \var{_subtype} defaults to \mimetype{rfc822}.
\end{itemize}

\module{mimelib} provided some utility functions in its
\module{address} and \module{date} modules.  All of these functions
have been moved to the \refmodule{email.Utils} module.

The \code{MsgReader} class/module has been removed.  Its functionality
is most closely supported in the \function{body_line_iterator()}
function in the \refmodule{email.Iterators} module.

\subsection{Examples}

Here are a few examples of how to use the \module{email} package to
read, write, and send simple email messages, as well as more complex
MIME messages.

First, let's see how to create and send a simple text message:

\begin{verbatim}
# Import smtplib for the actual sending function
import smtplib

# Here are the email pacakge modules we'll need
from email import Encoders
from email.MIMEText import MIMEText

# Open a plain text file for reading
fp = open(textfile)
# Create a text/plain message, using Quoted-Printable encoding for non-ASCII
# characters.
msg = MIMEText(fp.read(), _encoder=Encoders.encode_quopri)
fp.close()

# me == the sender's email address
# you == the recipient's email address
msg['Subject'] = 'The contents of %s' % textfile
msg['From'] = me
msg['To'] = you

# Send the message via our own SMTP server.  Use msg.as_string() with
# unixfrom=0 so as not to confuse SMTP.
s = smtplib.SMTP()
s.connect()
s.sendmail(me, [you], msg.as_string(0))
s.close()
\end{verbatim}

Here's an example of how to send a MIME message containing a bunch of
family pictures:

\begin{verbatim}
# Import smtplib for the actual sending function
import smtplib

# Here are the email pacakge modules we'll need
from email.MIMEImage import MIMEImage
from email.MIMEBase import MIMEBase

COMMASPACE = ', '

# Create the container (outer) email message.
# me == the sender's email address
# family = the list of all recipients' email addresses
msg = MIMEBase('multipart', 'mixed')
msg['Subject'] = 'Our family reunion'
msg['From'] = me
msg['To'] = COMMASPACE.join(family)
msg.preamble = 'Our family reunion'
# Guarantees the message ends in a newline
msg.epilogue = ''

# Assume we know that the image files are all in PNG format
for file in pngfiles:
    # Open the files in binary mode.  Let the MIMEIMage class automatically
    # guess the specific image type.
    fp = open(file, 'rb')
    img = MIMEImage(fp.read())
    fp.close()
    msg.attach(img)

# Send the email via our own SMTP server.
s = smtplib.SMTP()
s.connect()
s.sendmail(me, family, msg.as_string(unixfrom=0))
s.close()
\end{verbatim}

Here's an example\footnote{Thanks to Matthew Dixon Cowles for the
original inspiration and examples.} of how to send the entire contents
of a directory as an email message:

\begin{verbatim}
#!/usr/bin/env python

"""Send the contents of a directory as a MIME message.

Usage: dirmail [options] from to [to ...]*

Options:
    -h / --help
        Print this message and exit.

    -d directory
    --directory=directory
        Mail the contents of the specified directory, otherwise use the
        current directory.  Only the regular files in the directory are sent,
        and we don't recurse to subdirectories.

`from' is the email address of the sender of the message.

`to' is the email address of the recipient of the message, and multiple
recipients may be given.

The email is sent by forwarding to your local SMTP server, which then does the
normal delivery process.  Your local machine must be running an SMTP server.
"""

import sys
import os
import getopt
import smtplib
# For guessing MIME type based on file name extension
import mimetypes

from email import Encoders
from email.Message import Message
from email.MIMEAudio import MIMEAudio
from email.MIMEBase import MIMEBase
from email.MIMEImage import MIMEImage
from email.MIMEText import MIMEText

COMMASPACE = ', '


def usage(code, msg=''):
    print >> sys.stderr, __doc__
    if msg:
        print >> sys.stderr, msg
    sys.exit(code)


def main():
    try:
        opts, args = getopt.getopt(sys.argv[1:], 'hd:', ['help', 'directory='])
    except getopt.error, msg:
        usage(1, msg)

    dir = os.curdir
    for opt, arg in opts:
        if opt in ('-h', '--help'):
            usage(0)
        elif opt in ('-d', '--directory'):
            dir = arg

    if len(args) < 2:
        usage(1)

    sender = args[0]
    recips = args[1:]
    
    # Create the enclosing (outer) message
    outer = MIMEBase('multipart', 'mixed')
    outer['Subject'] = 'Contents of directory %s' % os.path.abspath(dir)
    outer['To'] = COMMASPACE.join(recips)
    outer['From'] = sender
    outer.preamble = 'You will not see this in a MIME-aware mail reader.\n'
    # To guarantee the message ends with a newline
    outer.epilogue = ''

    for filename in os.listdir(dir):
        path = os.path.join(dir, filename)
        if not os.path.isfile(path):
            continue
        # Guess the Content-Type: based on the file's extension.  Encoding
        # will be ignored, although we should check for simple things like
        # gzip'd or compressed files
        ctype, encoding = mimetypes.guess_type(path)
        if ctype is None or encoding is not None:
            # No guess could be made, or the file is encoded (compressed), so
            # use a generic bag-of-bits type.
            ctype = 'application/octet-stream'
        maintype, subtype = ctype.split('/', 1)
        if maintype == 'text':
            fp = open(path)
            # Note: we should handle calculating the charset
            msg = MIMEText(fp.read(), _subtype=subtype)
            fp.close()
        elif maintype == 'image':
            fp = open(path, 'rb')
            msg = MIMEImage(fp.read(), _subtype=subtype)
            fp.close()
        elif maintype == 'audio':
            fp = open(path, 'rb')
            msg = MIMEAudio(fp.read(), _subtype=subtype)
            fp.close()
        else:
            fp = open(path, 'rb')
            msg = MIMEBase(maintype, subtype)
            msg.add_payload(fp.read())
            fp.close()
            # Encode the payload using Base64
            Encoders.encode_base64(msg)
        # Set the filename parameter
        msg.add_header('Content-Disposition', 'attachment', filename=filename)
        outer.attach(msg)

    fp = open('/tmp/debug.pck', 'w')
    import cPickle
    cPickle.dump(outer, fp)
    fp.close()
    # Now send the message
    s = smtplib.SMTP()
    s.connect()
    s.sendmail(sender, recips, outer.as_string(0))
    s.close()


if __name__ == '__main__':
    main()
\end{verbatim}

And finally, here's an example of how to unpack a MIME message like
the one above, into a directory of files:

\begin{verbatim}
#!/usr/bin/env python

"""Unpack a MIME message into a directory of files.

Usage: unpackmail [options] msgfile

Options:
    -h / --help
        Print this message and exit.

    -d directory
    --directory=directory
        Unpack the MIME message into the named directory, which will be
        created if it doesn't already exist.

msgfile is the path to the file containing the MIME message.
"""

import sys
import os
import getopt
import errno
import mimetypes
import email


def usage(code, msg=''):
    print >> sys.stderr, __doc__
    if msg:
        print >> sys.stderr, msg
    sys.exit(code)


def main():
    try:
        opts, args = getopt.getopt(sys.argv[1:], 'hd:', ['help', 'directory='])
    except getopt.error, msg:
        usage(1, msg)

    dir = os.curdir
    for opt, arg in opts:
        if opt in ('-h', '--help'):
            usage(0)
        elif opt in ('-d', '--directory'):
            dir = arg

    try:
        msgfile = args[0]
    except IndexError:
        usage(1)

    try:
        os.mkdir(dir)
    except OSError, e:
        # Ignore directory exists error
        if e.errno <> errno.EEXIST: raise

    fp = open(msgfile)
    msg = email.message_from_file(fp)
    fp.close()

    counter = 1
    for part in msg.walk():
        # multipart/* are just containers
        if part.get_main_type() == 'multipart':
            continue
        # Applications should really sanitize the given filename so that an
        # email message can't be used to overwrite important files
        filename = part.get_filename()
        if not filename:
            ext = mimetypes.guess_extension(part.get_type())
            if not ext:
                # Use a generic bag-of-bits extension
                ext = '.bin'
            filename = 'part-%03d%s' % (counter, ext)
        counter += 1
        fp = open(os.path.join(dir, filename), 'wb')
        fp.write(part.get_payload(decode=1))
        fp.close()


if __name__ == '__main__':
    main()
\end{verbatim}