test_email.py

来自「mallet是自然语言处理、机器学习领域的一个开源项目。」· Python 代码 · 共 1,770 行 · 第 1/5 页

# Copyright (C) 2001,2002,2003 Python Software Foundation
# email package unit tests

import os
import sys
import time
import base64
import difflib
import unittest
import warnings
from cStringIO import StringIO
from types import StringType, ListType

import email

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
from email.Message import Message
from email.MIMEAudio import MIMEAudio
from email.MIMEText import MIMEText
from email.MIMEImage import MIMEImage
from email.MIMEBase import MIMEBase
from email.MIMEMessage import MIMEMessage
from email.MIMEMultipart import MIMEMultipart
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.test_support import findfile, run_unittest
from email.test import __file__ as landmark


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

# We don't care about DeprecationWarnings
warnings.filterwarnings('ignore', '', DeprecationWarning, __name__)

try:
    True, False
except NameError:
    True = 1
    False = 0



def openfile(filename, mode='r'):
    path = os.path.join(os.path.dirname(landmark), 'data', filename)
    return open(path, mode)



# Base test class
class TestEmailBase(unittest.TestCase):
    if hasattr(difflib, 'ndiff'):
        # Python 2.2 and beyond
        def ndiffAssertEqual(self, first, second):
            """Like failUnlessEqual except use ndiff for readable output."""
            if first <> second:
                sfirst = str(first)
                ssecond = str(second)
                diff = difflib.ndiff(sfirst.splitlines(), ssecond.splitlines())
                fp = StringIO()
                print >> fp, NL, NL.join(diff)
                raise self.failureException, fp.getvalue()
    else:
        # Python 2.1
        ndiffAssertEqual = unittest.TestCase.assertEqual

    def _msgobj(self, filename, strict=False):
        fp = openfile(findfile(filename))
        try:
            msg = email.message_from_file(fp, strict=strict)
        finally:
            fp.close()
        return msg



# 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'), ['[e-mail]', '[e-mail]', '[e-mail]'])
        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_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_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_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_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),
           'This is a 7bit encoded message.\n')
        # Subpart 2 is quopri
        eq(msg.get_payload(1).get_payload(decode=True),
           '\xa1This is a Quoted Printable encoded message!\n')
        # Subpart 3 is base64
        eq(msg.get_payload(2).get_payload(decode=True),
           'This is a Base64 encoded message.')
        # Subpart 4 has no Content-Transfer-Encoding: header.
        eq(msg.get_payload(3).get_payload(decode=True),
           '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), 'hello world')
        # Now try some bogus data
        msg.set_payload('foo')
        eq(msg.get_payload(decode=True), 'foo')

    def test_decoded_generator(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_07.txt')
        fp = openfile('msg_17.txt')
        try:
            text = fp.read()
        finally:
            fp.close()
        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.failUnless('from' in msg)
        self.failUnless('From' in msg)
        self.failUnless('FROM' in msg)
        self.failUnless('to' in msg)
        self.failUnless('To' in msg)
        self.failUnless('TO' in msg)

    def test_as_string(self):
        eq = self.assertEqual
        msg = self._msgobj('msg_01.txt')
        fp = openfile('msg_01.txt')
        try:
            text = fp.read()
        finally:
            fp.close()
        eq(text, msg.as_string())
        fullrepr = str(msg)
        lines = fullrepr.split('\n')
        self.failUnless(lines[0].startswith('From '))
        eq(text, NL.join(lines[1:]))

    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')])

    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.

    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')

    def test_has_key(self):
        msg = email.message_from_string('Header: exists')
        self.failUnless(msg.has_key('header'))
        self.failUnless(msg.has_key('Header'))
        self.failUnless(msg.has_key('HEADER'))
        self.failIf(msg.has_key('headeri'))

    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'),