/Lib/test/test_email/test_email.py

# Copyright (C) 2001-2010 Python Software Foundation
# Contact: email-sig@python.org
# email package unit tests

import re
import time
import base64
import unittest
import textwrap

from io import StringIO, BytesIO
from itertools import chain
from random import choice
from socket import getfqdn
from threading import Thread

import email
import email.policy

from email.charset import Charset
from email.header import Header, decode_header, make_header
from email.parser import Parser, HeaderParser
from email.generator import Generator, DecodedGenerator, BytesGenerator
from email.message import Message
from email.mime.application import MIMEApplication
from email.mime.audio import MIMEAudio
from email.mime.text import MIMEText
from email.mime.image import MIMEImage
from email.mime.base import MIMEBase
from email.mime.message import MIMEMessage
from email.mime.multipart import MIMEMultipart
from email.mime.nonmultipart import MIMENonMultipart
from email import utils
from email import errors
from email import encoders
from email import iterators
from email import base64mime
from email import quoprimime

from test.support import unlink, start_threads
from test.test_email import openfile, TestEmailBase

# These imports are documented to work, but we are testing them using a
# different path, so we import them here just to make sure they are importable.
from email.parser import FeedParser, BytesFeedParser

NL = '\n'
EMPTYSTRING = ''
SPACE = ' '


# Test various aspects of the Message class's API
class TestMessageAPI(TestEmailBase):
    def test_get_all(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_20.txt')
        eq(msg.get_all('cc'), ['ccc@zzz.org', 'ddd@zzz.org', 'eee@zzz.org'])
        eq(msg.get_all('xx', 'n/a'), 'n/a')

    def test_getset_charset(self):
        eq = self.assertEqual
        msg = Message()
        eq(msg.get_charset(), None)
        charset = Charset('iso-8859-1')
        msg.set_charset(charset)
        eq(msg['mime-version'], '1.0')
        eq(msg.get_content_type(), 'text/plain')
        eq(msg['content-type'], 'text/plain; charset="iso-8859-1"')
        eq(msg.get_param('charset'), 'iso-8859-1')
        eq(msg['content-transfer-encoding'], 'quoted-printable')
        eq(msg.get_charset().input_charset, 'iso-8859-1')
        # Remove the charset
        msg.set_charset(None)
        eq(msg.get_charset(), None)
        eq(msg['content-type'], 'text/plain')
        # Try adding a charset when there's already MIME headers present
        msg = Message()
        msg['MIME-Version'] = '2.0'
        msg['Content-Type'] = 'text/x-weird'
        msg['Content-Transfer-Encoding'] = 'quinted-puntable'
        msg.set_charset(charset)
        eq(msg['mime-version'], '2.0')
        eq(msg['content-type'], 'text/x-weird; charset="iso-8859-1"')
        eq(msg['content-transfer-encoding'], 'quinted-puntable')

    def test_set_charset_from_string(self):
        eq = self.assertEqual
        msg = Message()
        msg.set_charset('us-ascii')
        eq(msg.get_charset().input_charset, 'us-ascii')
        eq(msg['content-type'], 'text/plain; charset="us-ascii"')

    def test_set_payload_with_charset(self):
        msg = Message()
        charset = Charset('iso-8859-1')
        msg.set_payload('This is a string payload', charset)
        self.assertEqual(msg.get_charset().input_charset, 'iso-8859-1')

    def test_set_payload_with_8bit_data_and_charset(self):
        data = b'\xd0\x90\xd0\x91\xd0\x92'
        charset = Charset('utf-8')
        msg = Message()
        msg.set_payload(data, charset)
        self.assertEqual(msg['content-transfer-encoding'], 'base64')
        self.assertEqual(msg.get_payload(decode=True), data)
        self.assertEqual(msg.get_payload(), '0JDQkdCS\n')

    def test_set_payload_with_non_ascii_and_charset_body_encoding_none(self):
        data = b'\xd0\x90\xd0\x91\xd0\x92'
        charset = Charset('utf-8')
        charset.body_encoding = None # Disable base64 encoding
        msg = Message()
        msg.set_payload(data.decode('utf-8'), charset)
        self.assertEqual(msg['content-transfer-encoding'], '8bit')
        self.assertEqual(msg.get_payload(decode=True), data)

    def test_set_payload_with_8bit_data_and_charset_body_encoding_none(self):
        data = b'\xd0\x90\xd0\x91\xd0\x92'
        charset = Charset('utf-8')
        charset.body_encoding = None # Disable base64 encoding
        msg = Message()
        msg.set_payload(data, charset)
        self.assertEqual(msg['content-transfer-encoding'], '8bit')
        self.assertEqual(msg.get_payload(decode=True), data)

    def test_set_payload_to_list(self):
        msg = Message()
        msg.set_payload([])
        self.assertEqual(msg.get_payload(), [])

    def test_attach_when_payload_is_string(self):
        msg = Message()
        msg['Content-Type'] = 'multipart/mixed'
        msg.set_payload('string payload')
        sub_msg = MIMEMessage(Message())
        self.assertRaisesRegex(TypeError, "[Aa]ttach.*non-multipart",
                               msg.attach, sub_msg)

    def test_get_charsets(self):
        eq = self.assertEqual

        msg = self._msgobj('msg_08.txt')
        charsets = msg.get_charsets()
        eq(charsets, [None, 'us-ascii', 'iso-8859-1', 'iso-8859-2', 'koi8-r'])

        msg = self._msgobj('msg_09.txt')
        charsets = msg.get_charsets('dingbat')
        eq(charsets, ['dingbat', 'us-ascii', 'iso-8859-1', 'dingbat',
                      'koi8-r'])

        msg = self._msgobj('msg_12.txt')
        charsets = msg.get_charsets()
        eq(charsets, [None, 'us-ascii', 'iso-8859-1', None, 'iso-8859-2',
                      'iso-8859-3', 'us-ascii', 'koi8-r'])

    def test_get_filename(self):
        eq = self.assertEqual

        msg = self._msgobj('msg_04.txt')
        filenames = [p.get_filename() for p in msg.get_payload()]
        eq(filenames, ['msg.txt', 'msg.txt'])

        msg = self._msgobj('msg_07.txt')
        subpart = msg.get_payload(1)
        eq(subpart.get_filename(), 'dingusfish.gif')

    def test_get_filename_with_name_parameter(self):
        eq = self.assertEqual

        msg = self._msgobj('msg_44.txt')
        filenames = [p.get_filename() for p in msg.get_payload()]
        eq(filenames, ['msg.txt', 'msg.txt'])

    def test_get_boundary(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_07.txt')
        # No quotes!
        eq(msg.get_boundary(), 'BOUNDARY')

    def test_set_boundary(self):
        eq = self.assertEqual
        # This one has no existing boundary parameter, but the Content-Type:
        # header appears fifth.
        msg = self._msgobj('msg_01.txt')
        msg.set_boundary('BOUNDARY')
        header, value = msg.items()[4]
        eq(header.lower(), 'content-type')
        eq(value, 'text/plain; charset="us-ascii"; boundary="BOUNDARY"')
        # This one has a Content-Type: header, with a boundary, stuck in the
        # middle of its headers.  Make sure the order is preserved; it should
        # be fifth.
        msg = self._msgobj('msg_04.txt')
        msg.set_boundary('BOUNDARY')
        header, value = msg.items()[4]
        eq(header.lower(), 'content-type')
        eq(value, 'multipart/mixed; boundary="BOUNDARY"')
        # And this one has no Content-Type: header at all.
        msg = self._msgobj('msg_03.txt')
        self.assertRaises(errors.HeaderParseError,
                          msg.set_boundary, 'BOUNDARY')

    def test_make_boundary(self):
        msg = MIMEMultipart('form-data')
        # Note that when the boundary gets created is an implementation
        # detail and might change.
        self.assertEqual(msg.items()[0][1], 'multipart/form-data')
        # Trigger creation of boundary
        msg.as_string()
        self.assertEqual(msg.items()[0][1][:33],
                        'multipart/form-data; boundary="==')
        # XXX: there ought to be tests of the uniqueness of the boundary, too.

    def test_message_rfc822_only(self):
        # Issue 7970: message/rfc822 not in multipart parsed by
        # HeaderParser caused an exception when flattened.
        with openfile('msg_46.txt') as fp:
            msgdata = fp.read()
        parser = HeaderParser()
        msg = parser.parsestr(msgdata)
        out = StringIO()
        gen = Generator(out, True, 0)
        gen.flatten(msg, False)
        self.assertEqual(out.getvalue(), msgdata)

    def test_byte_message_rfc822_only(self):
        # Make sure new bytes header parser also passes this.
        with openfile('msg_46.txt') as fp:
            msgdata = fp.read().encode('ascii')
        parser = email.parser.BytesHeaderParser()
        msg = parser.parsebytes(msgdata)
        out = BytesIO()
        gen = email.generator.BytesGenerator(out)
        gen.flatten(msg)
        self.assertEqual(out.getvalue(), msgdata)

    def test_get_decoded_payload(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_10.txt')
        # The outer message is a multipart
        eq(msg.get_payload(decode=True), None)
        # Subpart 1 is 7bit encoded
        eq(msg.get_payload(0).get_payload(decode=True),
           b'This is a 7bit encoded message.\n')
        # Subpart 2 is quopri
        eq(msg.get_payload(1).get_payload(decode=True),
           b'\xa1This is a Quoted Printable encoded message!\n')
        # Subpart 3 is base64
        eq(msg.get_payload(2).get_payload(decode=True),
           b'This is a Base64 encoded message.')
        # Subpart 4 is base64 with a trailing newline, which
        # used to be stripped (issue 7143).
        eq(msg.get_payload(3).get_payload(decode=True),
           b'This is a Base64 encoded message.\n')
        # Subpart 5 has no Content-Transfer-Encoding: header.
        eq(msg.get_payload(4).get_payload(decode=True),
           b'This has no Content-Transfer-Encoding: header.\n')

    def test_get_decoded_uu_payload(self):
        eq = self.assertEqual
        msg = Message()
        msg.set_payload('begin 666 -\n+:&5L;&\\@=V]R;&0 \n \nend\n')
        for cte in ('x-uuencode', 'uuencode', 'uue', 'x-uue'):
            msg['content-transfer-encoding'] = cte
            eq(msg.get_payload(decode=True), b'hello world')
        # Now try some bogus data
        msg.set_payload('foo')
        eq(msg.get_payload(decode=True), b'foo')

    def test_get_payload_n_raises_on_non_multipart(self):
        msg = Message()
        self.assertRaises(TypeError, msg.get_payload, 1)

    def test_decoded_generator(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_07.txt')
        with openfile('msg_17.txt') as fp:
            text = fp.read()
        s = StringIO()
        g = DecodedGenerator(s)
        g.flatten(msg)
        eq(s.getvalue(), text)

    def test__contains__(self):
        msg = Message()
        msg['From'] = 'Me'
        msg['to'] = 'You'
        # Check for case insensitivity
        self.assertIn('from', msg)
        self.assertIn('From', msg)
        self.assertIn('FROM', msg)
        self.assertIn('to', msg)
        self.assertIn('To', msg)
        self.assertIn('TO', msg)

    def test_as_string(self):
        msg = self._msgobj('msg_01.txt')
        with openfile('msg_01.txt') as fp:
            text = fp.read()
        self.assertEqual(text, str(msg))
        fullrepr = msg.as_string(unixfrom=True)
        lines = fullrepr.split('\n')
        self.assertTrue(lines[0].startswith('From '))
        self.assertEqual(text, NL.join(lines[1:]))

    def test_as_string_policy(self):
        msg = self._msgobj('msg_01.txt')
        newpolicy = msg.policy.clone(linesep='\r\n')
        fullrepr = msg.as_string(policy=newpolicy)
        s = StringIO()
        g = Generator(s, policy=newpolicy)
        g.flatten(msg)
        self.assertEqual(fullrepr, s.getvalue())

    def test_as_bytes(self):
        msg = self._msgobj('msg_01.txt')
        with openfile('msg_01.txt') as fp:
            data = fp.read().encode('ascii')
        self.assertEqual(data, bytes(msg))
        fullrepr = msg.as_bytes(unixfrom=True)
        lines = fullrepr.split(b'\n')
        self.assertTrue(lines[0].startswith(b'From '))
        self.assertEqual(data, b'\n'.join(lines[1:]))

    def test_as_bytes_policy(self):
        msg = self._msgobj('msg_01.txt')
        newpolicy = msg.policy.clone(linesep='\r\n')
        fullrepr = msg.as_bytes(policy=newpolicy)
        s = BytesIO()
        g = BytesGenerator(s,policy=newpolicy)
        g.flatten(msg)
        self.assertEqual(fullrepr, s.getvalue())

    # test_headerregistry.TestContentTypeHeader.bad_params
    def test_bad_param(self):
        msg = email.message_from_string("Content-Type: blarg; baz; boo\n")
        self.assertEqual(msg.get_param('baz'), '')

    def test_missing_filename(self):
        msg = email.message_from_string("From: foo\n")
        self.assertEqual(msg.get_filename(), None)

    def test_bogus_filename(self):
        msg = email.message_from_string(
        "Content-Disposition: blarg; filename\n")
        self.assertEqual(msg.get_filename(), '')

    def test_missing_boundary(self):
        msg = email.message_from_string("From: foo\n")
        self.assertEqual(msg.get_boundary(), None)

    def test_get_params(self):
        eq = self.assertEqual
        msg = email.message_from_string(
            'X-Header: foo=one; bar=two; baz=three\n')
        eq(msg.get_params(header='x-header'),
           [('foo', 'one'), ('bar', 'two'), ('baz', 'three')])
        msg = email.message_from_string(
            'X-Header: foo; bar=one; baz=two\n')
        eq(msg.get_params(header='x-header'),
           [('foo', ''), ('bar', 'one'), ('baz', 'two')])
        eq(msg.get_params(), None)
        msg = email.message_from_string(
            'X-Header: foo; bar="one"; baz=two\n')
        eq(msg.get_params(header='x-header'),
           [('foo', ''), ('bar', 'one'), ('baz', 'two')])

    # test_headerregistry.TestContentTypeHeader.spaces_around_param_equals
    def test_get_param_liberal(self):
        msg = Message()
        msg['Content-Type'] = 'Content-Type: Multipart/mixed; boundary = "CPIMSSMTPC06p5f3tG"'
        self.assertEqual(msg.get_param('boundary'), 'CPIMSSMTPC06p5f3tG')

    def test_get_param(self):
        eq = self.assertEqual
        msg = email.message_from_string(
            "X-Header: foo=one; bar=two; baz=three\n")
        eq(msg.get_param('bar', header='x-header'), 'two')
        eq(msg.get_param('quuz', header='x-header'), None)
        eq(msg.get_param('quuz'), None)
        msg = email.message_from_string(
            'X-Header: foo; bar="one"; baz=two\n')
        eq(msg.get_param('foo', header='x-header'), '')
        eq(msg.get_param('bar', header='x-header'), 'one')
        eq(msg.get_param('baz', header='x-header'), 'two')
        # XXX: We are not RFC-2045 compliant!  We cannot parse:
        # msg["Content-Type"] = 'text/plain; weird="hey; dolly? [you] @ <\\"home\\">?"'
        # msg.get_param("weird")
        # yet.

    # test_headerregistry.TestContentTypeHeader.spaces_around_semis
    def test_get_param_funky_continuation_lines(self):
        msg = self._msgobj('msg_22.txt')
        self.assertEqual(msg.get_payload(1).get_param('name'), 'wibble.JPG')

    # test_headerregistry.TestContentTypeHeader.semis_inside_quotes
    def test_get_param_with_semis_in_quotes(self):
        msg = email.message_from_string(
            'Content-Type: image/pjpeg; name="Jim&amp;&amp;Jill"\n')
        self.assertEqual(msg.get_param('name'), 'Jim&amp;&amp;Jill')
        self.assertEqual(msg.get_param('name', unquote=False),
                         '"Jim&amp;&amp;Jill"')

    # test_headerregistry.TestContentTypeHeader.quotes_inside_rfc2231_value
    def test_get_param_with_quotes(self):
        msg = email.message_from_string(
            'Content-Type: foo; bar*0="baz\\"foobar"; bar*1="\\"baz"')
        self.assertEqual(msg.get_param('bar'), 'baz"foobar"baz')
        msg = email.message_from_string(
            "Content-Type: foo; bar*0=\"baz\\\"foobar\"; bar*1=\"\\\"baz\"")
        self.assertEqual(msg.get_param('bar'), 'baz"foobar"baz')

    def test_field_containment(self):
        msg = email.message_from_string('Header: exists')
        self.assertIn('header', msg)
        self.assertIn('Header', msg)
        self.assertIn('HEADER', msg)
        self.assertNotIn('headerx', msg)

    def test_set_param(self):
        eq = self.assertEqual
        msg = Message()
        msg.set_param('charset', 'iso-2022-jp')
        eq(msg.get_param('charset'), 'iso-2022-jp')
        msg.set_param('importance', 'high value')
        eq(msg.get_param('importance'), 'high value')
        eq(msg.get_param('importance', unquote=False), '"high value"')
        eq(msg.get_params(), [('text/plain', ''),
                              ('charset', 'iso-2022-jp'),
                              ('importance', 'high value')])
        eq(msg.get_params(unquote=False), [('text/plain', ''),
                                       ('charset', '"iso-2022-jp"'),
                                       ('importance', '"high value"')])
        msg.set_param('charset', 'iso-9999-xx', header='X-Jimmy')
        eq(msg.get_param('charset', header='X-Jimmy'), 'iso-9999-xx')

    def test_del_param(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_05.txt')
        eq(msg.get_params(),
           [('multipart/report', ''), ('report-type', 'delivery-status'),
            ('boundary', 'D1690A7AC1.996856090/mail.example.com')])
        old_val = msg.get_param("report-type")
        msg.del_param("report-type")
        eq(msg.get_params(),
           [('multipart/report', ''),
            ('boundary', 'D1690A7AC1.996856090/mail.example.com')])
        msg.set_param("report-type", old_val)
        eq(msg.get_params(),
           [('multipart/report', ''),
            ('boundary', 'D1690A7AC1.996856090/mail.example.com'),
            ('report-type', old_val)])

    def test_del_param_on_other_header(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'attachment', filename='bud.gif')
        msg.del_param('filename', 'content-disposition')
        self.assertEqual(msg['content-disposition'], 'attachment')

    def test_del_param_on_nonexistent_header(self):
        msg = Message()
        # Deleting param on empty msg should not raise exception.
        msg.del_param('filename', 'content-disposition')

    def test_del_nonexistent_param(self):
        msg = Message()
        msg.add_header('Content-Type', 'text/plain', charset='utf-8')
        existing_header = msg['Content-Type']
        msg.del_param('foobar', header='Content-Type')
        self.assertEqual(msg['Content-Type'], existing_header)

    def test_set_type(self):
        eq = self.assertEqual
        msg = Message()
        self.assertRaises(ValueError, msg.set_type, 'text')
        msg.set_type('text/plain')
        eq(msg['content-type'], 'text/plain')
        msg.set_param('charset', 'us-ascii')
        eq(msg['content-type'], 'text/plain; charset="us-ascii"')
        msg.set_type('text/html')
        eq(msg['content-type'], 'text/html; charset="us-ascii"')

    def test_set_type_on_other_header(self):
        msg = Message()
        msg['X-Content-Type'] = 'text/plain'
        msg.set_type('application/octet-stream', 'X-Content-Type')
        self.assertEqual(msg['x-content-type'], 'application/octet-stream')

    def test_get_content_type_missing(self):
        msg = Message()
        self.assertEqual(msg.get_content_type(), 'text/plain')

    def test_get_content_type_missing_with_default_type(self):
        msg = Message()
        msg.set_default_type('message/rfc822')
        self.assertEqual(msg.get_content_type(), 'message/rfc822')

    def test_get_content_type_from_message_implicit(self):
        msg = self._msgobj('msg_30.txt')
        self.assertEqual(msg.get_payload(0).get_content_type(),
                         'message/rfc822')

    def test_get_content_type_from_message_explicit(self):
        msg = self._msgobj('msg_28.txt')
        self.assertEqual(msg.get_payload(0).get_content_type(),
                         'message/rfc822')

    def test_get_content_type_from_message_text_plain_implicit(self):
        msg = self._msgobj('msg_03.txt')
        self.assertEqual(msg.get_content_type(), 'text/plain')

    def test_get_content_type_from_message_text_plain_explicit(self):
        msg = self._msgobj('msg_01.txt')
        self.assertEqual(msg.get_content_type(), 'text/plain')

    def test_get_content_maintype_missing(self):
        msg = Message()
        self.assertEqual(msg.get_content_maintype(), 'text')

    def test_get_content_maintype_missing_with_default_type(self):
        msg = Message()
        msg.set_default_type('message/rfc822')
        self.assertEqual(msg.get_content_maintype(), 'message')

    def test_get_content_maintype_from_message_implicit(self):
        msg = self._msgobj('msg_30.txt')
        self.assertEqual(msg.get_payload(0).get_content_maintype(), 'message')

    def test_get_content_maintype_from_message_explicit(self):
        msg = self._msgobj('msg_28.txt')
        self.assertEqual(msg.get_payload(0).get_content_maintype(), 'message')

    def test_get_content_maintype_from_message_text_plain_implicit(self):
        msg = self._msgobj('msg_03.txt')
        self.assertEqual(msg.get_content_maintype(), 'text')

    def test_get_content_maintype_from_message_text_plain_explicit(self):
        msg = self._msgobj('msg_01.txt')
        self.assertEqual(msg.get_content_maintype(), 'text')

    def test_get_content_subtype_missing(self):
        msg = Message()
        self.assertEqual(msg.get_content_subtype(), 'plain')

    def test_get_content_subtype_missing_with_default_type(self):
        msg = Message()
        msg.set_default_type('message/rfc822')
        self.assertEqual(msg.get_content_subtype(), 'rfc822')

    def test_get_content_subtype_from_message_implicit(self):
        msg = self._msgobj('msg_30.txt')
        self.assertEqual(msg.get_payload(0).get_content_subtype(), 'rfc822')

    def test_get_content_subtype_from_message_explicit(self):
        msg = self._msgobj('msg_28.txt')
        self.assertEqual(msg.get_payload(0).get_content_subtype(), 'rfc822')

    def test_get_content_subtype_from_message_text_plain_implicit(self):
        msg = self._msgobj('msg_03.txt')
        self.assertEqual(msg.get_content_subtype(), 'plain')

    def test_get_content_subtype_from_message_text_plain_explicit(self):
        msg = self._msgobj('msg_01.txt')
        self.assertEqual(msg.get_content_subtype(), 'plain')

    def test_get_content_maintype_error(self):
        msg = Message()
        msg['Content-Type'] = 'no-slash-in-this-string'
        self.assertEqual(msg.get_content_maintype(), 'text')

    def test_get_content_subtype_error(self):
        msg = Message()
        msg['Content-Type'] = 'no-slash-in-this-string'
        self.assertEqual(msg.get_content_subtype(), 'plain')

    def test_replace_header(self):
        eq = self.assertEqual
        msg = Message()
        msg.add_header('First', 'One')
        msg.add_header('Second', 'Two')
        msg.add_header('Third', 'Three')
        eq(msg.keys(), ['First', 'Second', 'Third'])
        eq(msg.values(), ['One', 'Two', 'Three'])
        msg.replace_header('Second', 'Twenty')
        eq(msg.keys(), ['First', 'Second', 'Third'])
        eq(msg.values(), ['One', 'Twenty', 'Three'])
        msg.add_header('First', 'Eleven')
        msg.replace_header('First', 'One Hundred')
        eq(msg.keys(), ['First', 'Second', 'Third', 'First'])
        eq(msg.values(), ['One Hundred', 'Twenty', 'Three', 'Eleven'])
        self.assertRaises(KeyError, msg.replace_header, 'Fourth', 'Missing')

    def test_get_content_disposition(self):
        msg = Message()
        self.assertIsNone(msg.get_content_disposition())
        msg.add_header('Content-Disposition', 'attachment',
                       filename='random.avi')
        self.assertEqual(msg.get_content_disposition(), 'attachment')
        msg.replace_header('Content-Disposition', 'inline')
        self.assertEqual(msg.get_content_disposition(), 'inline')
        msg.replace_header('Content-Disposition', 'InlinE')
        self.assertEqual(msg.get_content_disposition(), 'inline')

    # test_defect_handling:test_invalid_chars_in_base64_payload
    def test_broken_base64_payload(self):
        x = 'AwDp0P7//y6LwKEAcPa/6Q=9'
        msg = Message()
        msg['content-type'] = 'audio/x-midi'
        msg['content-transfer-encoding'] = 'base64'
        msg.set_payload(x)
        self.assertEqual(msg.get_payload(decode=True),
                         (b'\x03\x00\xe9\xd0\xfe\xff\xff.\x8b\xc0'
                          b'\xa1\x00p\xf6\xbf\xe9\x0f'))
        self.assertIsInstance(msg.defects[0],
                              errors.InvalidBase64CharactersDefect)

    def test_broken_unicode_payload(self):
        # This test improves coverage but is not a compliance test.
        # The behavior in this situation is currently undefined by the API.
        x = 'this is a br\xf6ken thing to do'
        msg = Message()
        msg['content-type'] = 'text/plain'
        msg['content-transfer-encoding'] = '8bit'
        msg.set_payload(x)
        self.assertEqual(msg.get_payload(decode=True),
                         bytes(x, 'raw-unicode-escape'))

    def test_questionable_bytes_payload(self):
        # This test improves coverage but is not a compliance test,
        # since it involves poking inside the black box.
        x = 'this is a quéstionable thing to do'.encode('utf-8')
        msg = Message()
        msg['content-type'] = 'text/plain; charset="utf-8"'
        msg['content-transfer-encoding'] = '8bit'
        msg._payload = x
        self.assertEqual(msg.get_payload(decode=True), x)

    # Issue 1078919
    def test_ascii_add_header(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'attachment',
                       filename='bud.gif')
        self.assertEqual('attachment; filename="bud.gif"',
            msg['Content-Disposition'])

    def test_noascii_add_header(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'attachment',
            filename="Fußballer.ppt")
        self.assertEqual(
            'attachment; filename*=utf-8\'\'Fu%C3%9Fballer.ppt',
            msg['Content-Disposition'])

    def test_nonascii_add_header_via_triple(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'attachment',
            filename=('iso-8859-1', '', 'Fußballer.ppt'))
        self.assertEqual(
            'attachment; filename*=iso-8859-1\'\'Fu%DFballer.ppt',
            msg['Content-Disposition'])

    def test_ascii_add_header_with_tspecial(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'attachment',
            filename="windows [filename].ppt")
        self.assertEqual(
            'attachment; filename="windows [filename].ppt"',
            msg['Content-Disposition'])

    def test_nonascii_add_header_with_tspecial(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'attachment',
            filename="Fußballer [filename].ppt")
        self.assertEqual(
            "attachment; filename*=utf-8''Fu%C3%9Fballer%20%5Bfilename%5D.ppt",
            msg['Content-Disposition'])

    def test_binary_quopri_payload(self):
        for charset in ('latin-1', 'ascii'):
            msg = Message()
            msg['content-type'] = 'text/plain; charset=%s' % charset
            msg['content-transfer-encoding'] = 'quoted-printable'
            msg.set_payload(b'foo=e6=96=87bar')
            self.assertEqual(
                msg.get_payload(decode=True),
                b'foo\xe6\x96\x87bar',
                'get_payload returns wrong result with charset %s.' % charset)

    def test_binary_base64_payload(self):
        for charset in ('latin-1', 'ascii'):
            msg = Message()
            msg['content-type'] = 'text/plain; charset=%s' % charset
            msg['content-transfer-encoding'] = 'base64'
            msg.set_payload(b'Zm9v5paHYmFy')
            self.assertEqual(
                msg.get_payload(decode=True),
                b'foo\xe6\x96\x87bar',
                'get_payload returns wrong result with charset %s.' % charset)

    def test_binary_uuencode_payload(self):
        for charset in ('latin-1', 'ascii'):
            for encoding in ('x-uuencode', 'uuencode', 'uue', 'x-uue'):
                msg = Message()
                msg['content-type'] = 'text/plain; charset=%s' % charset
                msg['content-transfer-encoding'] = encoding
                msg.set_payload(b"begin 666 -\n)9F]OYI:'8F%R\n \nend\n")
                self.assertEqual(
                    msg.get_payload(decode=True),
                    b'foo\xe6\x96\x87bar',
                    str(('get_payload returns wrong result ',
                         'with charset {0} and encoding {1}.')).\
                        format(charset, encoding))

    def test_add_header_with_name_only_param(self):
        msg = Message()
        msg.add_header('Content-Disposition', 'inline', foo_bar=None)
        self.assertEqual("inline; foo-bar", msg['Content-Disposition'])

    def test_add_header_with_no_value(self):
        msg = Message()
        msg.add_header('X-Status', None)
        self.assertEqual('', msg['X-Status'])

    # Issue 5871: reject an attempt to embed a header inside a header value
    # (header injection attack).
    def test_embedded_header_via_Header_rejected(self):
        msg = Message()
        msg['Dummy'] = Header('dummy\nX-Injected-Header: test')
        self.assertRaises(errors.HeaderParseError, msg.as_string)

    def test_embedded_header_via_string_rejected(self):
        msg = Message()
        msg['Dummy'] = 'dummy\nX-Injected-Header: test'
        self.assertRaises(errors.HeaderParseError, msg.as_string)

    def test_unicode_header_defaults_to_utf8_encoding(self):
        # Issue 14291
        m = MIMEText('abc\n')
        m['Subject'] = 'É test'
        self.assertEqual(str(m),textwrap.dedent("""\
            Content-Type: text/plain; charset="us-ascii"
            MIME-Version: 1.0
            Content-Transfer-Encoding: 7bit
            Subject: =?utf-8?q?=C3=89_test?=

            abc
            """))

    def test_unicode_body_defaults_to_utf8_encoding(self):
        # Issue 14291
        m = MIMEText('É testabc\n')
        self.assertEqual(str(m),textwrap.dedent("""\
            Content-Type: text/plain; charset="utf-8"
            MIME-Version: 1.0
            Content-Transfer-Encoding: base64

            w4kgdGVzdGFiYwo=
            """))


# Test the email.encoders module
class TestEncoders(unittest.TestCase):

    def test_EncodersEncode_base64(self):
        with openfile('PyBanner048.gif', 'rb') as fp:
            bindata = fp.read()
        mimed = email.mime.image.MIMEImage(bindata)
        base64ed = mimed.get_payload()
        # the transfer-encoded body lines should all be <=76 characters
        lines = base64ed.split('\n')
        self.assertLessEqual(max([ len(x) for x in lines ]), 76)

    def test_encode_empty_payload(self):
        eq = self.assertEqual
        msg = Message()
        msg.set_charset('us-ascii')
        eq(msg['content-transfer-encoding'], '7bit')

    def test_default_cte(self):
        eq = self.assertEqual
        # 7bit data and the default us-ascii _charset
        msg = MIMEText('hello world')
        eq(msg['content-transfer-encoding'], '7bit')
        # Similar, but with 8bit data
        msg = MIMEText('hello \xf8 world')
        eq(msg['content-transfer-encoding'], 'base64')
        # And now with a different charset
        msg = MIMEText('hello \xf8 world', _charset='iso-8859-1')
        eq(msg['content-transfer-encoding'], 'quoted-printable')

    def test_encode7or8bit(self):
        # Make sure a charset whose input character set is 8bit but
        # whose output character set is 7bit gets a transfer-encoding
        # of 7bit.
        eq = self.assertEqual
        msg = MIMEText('文\n', _charset='euc-jp')
        eq(msg['content-transfer-encoding'], '7bit')
        eq(msg.as_string(), textwrap.dedent("""\
            MIME-Version: 1.0
            Content-Type: text/plain; charset="iso-2022-jp"
            Content-Transfer-Encoding: 7bit

            \x1b$BJ8\x1b(B
            """))

    def test_qp_encode_latin1(self):
        msg = MIMEText('\xe1\xf6\n', 'text', 'ISO-8859-1')
        self.assertEqual(str(msg), textwrap.dedent("""\
            MIME-Version: 1.0
            Content-Type: text/text; charset="iso-8859-1"
            Content-Transfer-Encoding: quoted-printable

            =E1=F6
            """))

    def test_qp_encode_non_latin1(self):
        # Issue 16948
        msg = MIMEText('\u017c\n', 'text', 'ISO-8859-2')
        self.assertEqual(str(msg), textwrap.dedent("""\
            MIME-Version: 1.0
            Content-Type: text/text; charset="iso-8859-2"
            Content-Transfer-Encoding: quoted-printable

            =BF
            """))


# Test long header wrapping
class TestLongHeaders(TestEmailBase):

    maxDiff = None

    def test_split_long_continuation(self):
        eq = self.ndiffAssertEqual
        msg = email.message_from_string("""\
Subject: bug demonstration
\t12345678911234567892123456789312345678941234567895123456789612345678971234567898112345678911234567892123456789112345678911234567892123456789
\tmore text

test
""")
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        eq(sfp.getvalue(), """\
Subject: bug demonstration
\t12345678911234567892123456789312345678941234567895123456789612345678971234567898112345678911234567892123456789112345678911234567892123456789
\tmore text

test
""")

    def test_another_long_almost_unsplittable_header(self):
        eq = self.ndiffAssertEqual
        hstr = """\
bug demonstration
\t12345678911234567892123456789312345678941234567895123456789612345678971234567898112345678911234567892123456789112345678911234567892123456789
\tmore text"""
        h = Header(hstr, continuation_ws='\t')
        eq(h.encode(), """\
bug demonstration
\t12345678911234567892123456789312345678941234567895123456789612345678971234567898112345678911234567892123456789112345678911234567892123456789
\tmore text""")
        h = Header(hstr.replace('\t', ' '))
        eq(h.encode(), """\
bug demonstration
 12345678911234567892123456789312345678941234567895123456789612345678971234567898112345678911234567892123456789112345678911234567892123456789
 more text""")

    def test_long_nonstring(self):
        eq = self.ndiffAssertEqual
        g = Charset("iso-8859-1")
        cz = Charset("iso-8859-2")
        utf8 = Charset("utf-8")
        g_head = (b'Die Mieter treten hier ein werden mit einem Foerderband '
                  b'komfortabel den Korridor entlang, an s\xfcdl\xfcndischen '
                  b'Wandgem\xe4lden vorbei, gegen die rotierenden Klingen '
                  b'bef\xf6rdert. ')
        cz_head = (b'Finan\xe8ni metropole se hroutily pod tlakem jejich '
                   b'd\xf9vtipu.. ')
        utf8_head = ('\u6b63\u78ba\u306b\u8a00\u3046\u3068\u7ffb\u8a33\u306f'
                     '\u3055\u308c\u3066\u3044\u307e\u305b\u3093\u3002\u4e00'
                     '\u90e8\u306f\u30c9\u30a4\u30c4\u8a9e\u3067\u3059\u304c'
                     '\u3001\u3042\u3068\u306f\u3067\u305f\u3089\u3081\u3067'
                     '\u3059\u3002\u5b9f\u969b\u306b\u306f\u300cWenn ist das '
                     'Nunstuck git und Slotermeyer? Ja! Beiherhund das Oder '
                     'die Flipperwaldt gersput.\u300d\u3068\u8a00\u3063\u3066'
                     '\u3044\u307e\u3059\u3002')
        h = Header(g_head, g, header_name='Subject')
        h.append(cz_head, cz)
        h.append(utf8_head, utf8)
        msg = Message()
        msg['Subject'] = h
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        eq(sfp.getvalue(), """\
Subject: =?iso-8859-1?q?Die_Mieter_treten_hier_ein_werden_mit_einem_Foerderb?=
 =?iso-8859-1?q?and_komfortabel_den_Korridor_entlang=2C_an_s=FCdl=FCndischen?=
 =?iso-8859-1?q?_Wandgem=E4lden_vorbei=2C_gegen_die_rotierenden_Klingen_bef?=
 =?iso-8859-1?q?=F6rdert=2E_?= =?iso-8859-2?q?Finan=E8ni_metropole_se_hrouti?=
 =?iso-8859-2?q?ly_pod_tlakem_jejich_d=F9vtipu=2E=2E_?= =?utf-8?b?5q2j56K6?=
 =?utf-8?b?44Gr6KiA44GG44Go57+76Kiz44Gv44GV44KM44Gm44GE44G+44Gb44KT44CC5LiA?=
 =?utf-8?b?6YOo44Gv44OJ44Kk44OE6Kqe44Gn44GZ44GM44CB44GC44Go44Gv44Gn44Gf44KJ?=
 =?utf-8?b?44KB44Gn44GZ44CC5a6f6Zqb44Gr44Gv44CMV2VubiBpc3QgZGFzIE51bnN0dWNr?=
 =?utf-8?b?IGdpdCB1bmQgU2xvdGVybWV5ZXI/IEphISBCZWloZXJodW5kIGRhcyBPZGVyIGRp?=
 =?utf-8?b?ZSBGbGlwcGVyd2FsZHQgZ2Vyc3B1dC7jgI3jgajoqIDjgaPjgabjgYTjgb7jgZk=?=
 =?utf-8?b?44CC?=

""")
        eq(h.encode(maxlinelen=76), """\
=?iso-8859-1?q?Die_Mieter_treten_hier_ein_werden_mit_einem_Foerde?=
 =?iso-8859-1?q?rband_komfortabel_den_Korridor_entlang=2C_an_s=FCdl=FCndis?=
 =?iso-8859-1?q?chen_Wandgem=E4lden_vorbei=2C_gegen_die_rotierenden_Klinge?=
 =?iso-8859-1?q?n_bef=F6rdert=2E_?= =?iso-8859-2?q?Finan=E8ni_metropole_se?=
 =?iso-8859-2?q?_hroutily_pod_tlakem_jejich_d=F9vtipu=2E=2E_?=
 =?utf-8?b?5q2j56K644Gr6KiA44GG44Go57+76Kiz44Gv44GV44KM44Gm44GE44G+44Gb?=
 =?utf-8?b?44KT44CC5LiA6YOo44Gv44OJ44Kk44OE6Kqe44Gn44GZ44GM44CB44GC44Go?=
 =?utf-8?b?44Gv44Gn44Gf44KJ44KB44Gn44GZ44CC5a6f6Zqb44Gr44Gv44CMV2VubiBp?=
 =?utf-8?b?c3QgZGFzIE51bnN0dWNrIGdpdCB1bmQgU2xvdGVybWV5ZXI/IEphISBCZWlo?=
 =?utf-8?b?ZXJodW5kIGRhcyBPZGVyIGRpZSBGbGlwcGVyd2FsZHQgZ2Vyc3B1dC7jgI0=?=
 =?utf-8?b?44Go6KiA44Gj44Gm44GE44G+44GZ44CC?=""")

    def test_long_header_encode(self):
        eq = self.ndiffAssertEqual
        h = Header('wasnipoop; giraffes="very-long-necked-animals"; '
                   'spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"',
                   header_name='X-Foobar-Spoink-Defrobnit')
        eq(h.encode(), '''\
wasnipoop; giraffes="very-long-necked-animals";
 spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"''')

    def test_long_header_encode_with_tab_continuation_is_just_a_hint(self):
        eq = self.ndiffAssertEqual
        h = Header('wasnipoop; giraffes="very-long-necked-animals"; '
                   'spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"',
                   header_name='X-Foobar-Spoink-Defrobnit',
                   continuation_ws='\t')
        eq(h.encode(), '''\
wasnipoop; giraffes="very-long-necked-animals";
 spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"''')

    def test_long_header_encode_with_tab_continuation(self):
        eq = self.ndiffAssertEqual
        h = Header('wasnipoop; giraffes="very-long-necked-animals";\t'
                   'spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"',
                   header_name='X-Foobar-Spoink-Defrobnit',
                   continuation_ws='\t')
        eq(h.encode(), '''\
wasnipoop; giraffes="very-long-necked-animals";
\tspooge="yummy"; hippos="gargantuan"; marshmallows="gooey"''')

    def test_header_encode_with_different_output_charset(self):
        h = Header('文', 'euc-jp')
        self.assertEqual(h.encode(), "=?iso-2022-jp?b?GyRCSjgbKEI=?=")

    def test_long_header_encode_with_different_output_charset(self):
        h = Header(b'test-ja \xa4\xd8\xc5\xea\xb9\xc6\xa4\xb5\xa4\xec\xa4'
            b'\xbf\xa5\xe1\xa1\xbc\xa5\xeb\xa4\xcf\xbb\xca\xb2\xf1\xbc\xd4'
            b'\xa4\xce\xbe\xb5\xc7\xa7\xa4\xf2\xc2\xd4\xa4\xc3\xa4\xc6\xa4'
            b'\xa4\xa4\xde\xa4\xb9'.decode('euc-jp'), 'euc-jp')
        res = """\
=?iso-2022-jp?b?dGVzdC1qYSAbJEIkWEVqOUYkNSRsJD8lYSE8JWskTztKMnE8VCROPjUbKEI=?=
 =?iso-2022-jp?b?GyRCRyckckJUJEMkRiQkJF4kORsoQg==?="""
        self.assertEqual(h.encode(), res)

    def test_header_splitter(self):
        eq = self.ndiffAssertEqual
        msg = MIMEText('')
        # It'd be great if we could use add_header() here, but that doesn't
        # guarantee an order of the parameters.
        msg['X-Foobar-Spoink-Defrobnit'] = (
            'wasnipoop; giraffes="very-long-necked-animals"; '
            'spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"')
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        eq(sfp.getvalue(), '''\
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
X-Foobar-Spoink-Defrobnit: wasnipoop; giraffes="very-long-necked-animals";
 spooge="yummy"; hippos="gargantuan"; marshmallows="gooey"

''')

    def test_no_semis_header_splitter(self):
        eq = self.ndiffAssertEqual
        msg = Message()
        msg['From'] = 'test@dom.ain'
        msg['References'] = SPACE.join('<%d@dom.ain>' % i for i in range(10))
        msg.set_payload('Test')
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        eq(sfp.getvalue(), """\
From: test@dom.ain
References: <0@dom.ain> <1@dom.ain> <2@dom.ain> <3@dom.ain> <4@dom.ain>
 <5@dom.ain> <6@dom.ain> <7@dom.ain> <8@dom.ain> <9@dom.ain>

Test""")

    def test_last_split_chunk_does_not_fit(self):
        eq = self.ndiffAssertEqual
        h = Header('Subject: the first part of this is short, but_the_second'
            '_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line'
            '_all_by_itself')
        eq(h.encode(), """\
Subject: the first part of this is short,
 but_the_second_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself""")

    def test_splittable_leading_char_followed_by_overlong_unsplitable(self):
        eq = self.ndiffAssertEqual
        h = Header(', but_the_second'
            '_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line'
            '_all_by_itself')
        eq(h.encode(), """\
,
 but_the_second_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself""")

    def test_multiple_splittable_leading_char_followed_by_overlong_unsplitable(self):
        eq = self.ndiffAssertEqual
        h = Header(', , but_the_second'
            '_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line'
            '_all_by_itself')
        eq(h.encode(), """\
, ,
 but_the_second_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself""")

    def test_trailing_splitable_on_overlong_unsplitable(self):
        eq = self.ndiffAssertEqual
        h = Header('this_part_does_not_fit_within_maxlinelen_and_thus_should_'
            'be_on_a_line_all_by_itself;')
        eq(h.encode(), "this_part_does_not_fit_within_maxlinelen_and_thus_should_"
            "be_on_a_line_all_by_itself;")

    def test_trailing_splitable_on_overlong_unsplitable_with_leading_splitable(self):
        eq = self.ndiffAssertEqual
        h = Header('; '
            'this_part_does_not_fit_within_maxlinelen_and_thus_should_'
            'be_on_a_line_all_by_itself; ')
        eq(h.encode(), """\
;
 this_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself; """)

    def test_long_header_with_multiple_sequential_split_chars(self):
        eq = self.ndiffAssertEqual
        h = Header('This is a long line that has two whitespaces  in a row.  '
            'This used to cause truncation of the header when folded')
        eq(h.encode(), """\
This is a long line that has two whitespaces  in a row.  This used to cause
 truncation of the header when folded""")

    def test_splitter_split_on_punctuation_only_if_fws_with_header(self):
        eq = self.ndiffAssertEqual
        h = Header('thisverylongheaderhas;semicolons;and,commas,but'
            'they;arenotlegal;fold,points')
        eq(h.encode(), "thisverylongheaderhas;semicolons;and,commas,butthey;"
                        "arenotlegal;fold,points")

    def test_leading_splittable_in_the_middle_just_before_overlong_last_part(self):
        eq = self.ndiffAssertEqual
        h = Header('this is a  test where we need to have more than one line '
            'before; our final line that is just too big to fit;; '
            'this_part_does_not_fit_within_maxlinelen_and_thus_should_'
            'be_on_a_line_all_by_itself;')
        eq(h.encode(), """\
this is a  test where we need to have more than one line before;
 our final line that is just too big to fit;;
 this_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself;""")

    def test_overlong_last_part_followed_by_split_point(self):
        eq = self.ndiffAssertEqual
        h = Header('this_part_does_not_fit_within_maxlinelen_and_thus_should_'
            'be_on_a_line_all_by_itself ')
        eq(h.encode(), "this_part_does_not_fit_within_maxlinelen_and_thus_"
                        "should_be_on_a_line_all_by_itself ")

    def test_multiline_with_overlong_parts_separated_by_two_split_points(self):
        eq = self.ndiffAssertEqual
        h = Header('this_is_a__test_where_we_need_to_have_more_than_one_line_'
            'before_our_final_line_; ; '
            'this_part_does_not_fit_within_maxlinelen_and_thus_should_'
            'be_on_a_line_all_by_itself; ')
        eq(h.encode(), """\
this_is_a__test_where_we_need_to_have_more_than_one_line_before_our_final_line_;
 ;
 this_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself; """)

    def test_multiline_with_overlong_last_part_followed_by_split_point(self):
        eq = self.ndiffAssertEqual
        h = Header('this is a test where we need to have more than one line '
            'before our final line; ; '
            'this_part_does_not_fit_within_maxlinelen_and_thus_should_'
            'be_on_a_line_all_by_itself; ')
        eq(h.encode(), """\
this is a test where we need to have more than one line before our final line;
 ;
 this_part_does_not_fit_within_maxlinelen_and_thus_should_be_on_a_line_all_by_itself; """)

    def test_long_header_with_whitespace_runs(self):
        eq = self.ndiffAssertEqual
        msg = Message()
        msg['From'] = 'test@dom.ain'
        msg['References'] = SPACE.join(['<foo@dom.ain>  '] * 10)
        msg.set_payload('Test')
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        eq(sfp.getvalue(), """\
From: test@dom.ain
References: <foo@dom.ain>   <foo@dom.ain>   <foo@dom.ain>   <foo@dom.ain>
   <foo@dom.ain>   <foo@dom.ain>   <foo@dom.ain>   <foo@dom.ain>
   <foo@dom.ain>   <foo@dom.ain>\x20\x20

Test""")

    def test_long_run_with_semi_header_splitter(self):
        eq = self.ndiffAssertEqual
        msg = Message()
        msg['From'] = 'test@dom.ain'
        msg['References'] = SPACE.join(['<foo@dom.ain>'] * 10) + '; abc'
        msg.set_payload('Test')
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        eq(sfp.getvalue(), """\
From: test@dom.ain
References: <foo@dom.ain> <foo@dom.ain> <foo@dom.ain> <foo@dom.ain>
 <foo@dom.ain> <foo@dom.ain> <foo@dom.ain> <foo@dom.ain> <foo@dom.ain>
 <foo@dom.ain>; abc

Test""")

    def test_splitter_split_on_punctuation_only_if_fws(self):
        eq = self.ndiffAssertEqual
        msg = Message()
        msg['From'] = 'test@dom.ain'
        msg['References'] = ('thisverylongheaderhas;semicolons;and,commas,but'
            'they;arenotlegal;fold,points')
        msg.set_payload('Test')
        sfp = StringIO()
        g = Generator(sfp)
        g.flatten(msg)
        # XXX the space after the header should not be there.
        eq(sfp.getvalue(), """\
From: test@dom.ain
References:\x20
 thisverylongheaderhas;semicolons;and,commas,butthey;arenotlegal;fold,points

Test""")

    def test_no_split_long_header(self):
        eq = self.ndiffAssertEqual
        hstr = 'References: ' + 'x' * 80
        h = Header(hstr)
        # These come on two lines because Headers are really field value
        # classes and don't really know about their field names.
        eq(h.encode(), """\
References:
 xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx""")
        h = Header('x' * 80)
        eq(h.encode(), 'xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx')

    def test_splitting_multiple_long_lines(self):
        eq = self.ndiffAssertEqual
        hstr = """\
from babylon.socal-raves.org (localhost [127.0.0.1]); by babylon.socal-raves.org (Postfix) with ESMTP id B570E51B81; for <mailman-admin@babylon.socal-raves.org>; Sat, 2 Feb 2002 17:00:06 -0800 (PST)
\tfrom babylon.socal-raves.org (localhost [127.0.0.1]); by babylon.socal-raves.org (Postfix) with ESMTP id B570E51B81; for <mailman-admin@babylon.socal-raves.org>; Sat, 2 Feb 2002 17:00:06 -0800 (PST)
\tfrom babylon.socal-raves.org (localhost [127.0.0.1]); by babylon.socal-raves.org (Postfix) with ESMTP id B570E51B81; for <mailman-admin@babylon.socal-raves.org>; Sat, 2 Feb 2002 17:00:06 -0800 (PST)
"""
        h = Header(hstr, continuation_ws='\t')
        eq(h.encode(), """\
from babylon.socal-raves.org (localhost [127.0.0.1]);
 by babylon.socal-raves.org (Postfix) with ESMTP id B570E51B81;
 for <mailman-admin@babylon.socal-raves.org>;
 Sat, 2 Feb 2002 17:00:06 -0800 (PST)
\tfrom babylon.socal-raves.org (localhost [127.0.0.1]);
 by babylon.socal-raves.org (Postfix) with ESMTP id B570E51B81;
 for <mailman-admin@babylon.socal-raves.org>;
 Sat, 2 Feb 2002 17:00:06 -0800 (PST)
\tfrom babylon.socal-raves.org (localhost [127.0.0.1]);
 by babylon.socal-raves.org (Postfix) with ESMTP id B570E51B81;
 for <mailman-admin@babylon.socal-raves.org>;
 Sat, 2 Feb 2002 17:00:06 -0800 (PST)""")

    def test_splitting_first_line_only_is_long(self):
        eq = self.ndiffAssertEqual
        hstr = """\
from modemcable093.139-201-24.que.mc.videotron.ca ([24.201.139.93] helo=cthulhu.gerg.ca)
\tby kronos.mems-exchange.org with esmtp (Exim 4.05)
\tid 17k4h5-00034i-00
\tfor test@mems-exchange.org; Wed, 28 Aug 2002 11:25:20 -0400"""
        h = Header(hstr, maxlinelen=78, header_name='Receiv