shutil.rmtree(tmpdir)
sys.exit(0)
-from twisted.trial import unittest
-from mock import patch
+
import os, sys, locale
+from twisted.trial import unittest
+
from allmydata.test.common_util import ReallyEqualMixin
from allmydata.util import encodingutil, fileutil
from allmydata.util.encodingutil import argv_to_unicode, unicode_to_url, \
unicode_to_output, quote_output, quote_path, quote_local_unicode_path, \
unicode_platform, listdir_unicode, FilenameEncodingError, get_io_encoding, \
- get_filesystem_encoding, _reload
+ get_filesystem_encoding, to_str, from_utf8_or_none, _reload
from allmydata.dirnode import normalize
from twisted.python import usage
+
+class MockStdout(object):
+ pass
+
class EncodingUtilErrors(ReallyEqualMixin, unittest.TestCase):
+ def test_get_io_encoding(self):
+ mock_stdout = MockStdout()
+ self.patch(sys, 'stdout', mock_stdout)
- @patch('sys.stdout')
- def test_get_io_encoding(self, mock_stdout):
mock_stdout.encoding = 'UTF-8'
_reload()
self.failUnlessReallyEqual(get_io_encoding(), 'utf-8')
else:
self.failUnlessRaises(AssertionError, _reload)
- @patch('locale.getpreferredencoding')
- def test_get_io_encoding_not_from_stdout(self, mock_locale_getpreferredencoding):
- locale # hush pyflakes
- mock_locale_getpreferredencoding.return_value = 'koi8-r'
+ def test_get_io_encoding_not_from_stdout(self):
+ preferredencoding = 'koi8-r'
+ def call_locale_getpreferredencoding():
+ return preferredencoding
+ self.patch(locale, 'getpreferredencoding', call_locale_getpreferredencoding)
+ mock_stdout = MockStdout()
+ self.patch(sys, 'stdout', mock_stdout)
- class DummyStdout:
- pass
- old_stdout = sys.stdout
- sys.stdout = DummyStdout()
- try:
- expected = sys.platform == "win32" and 'utf-8' or 'koi8-r'
- _reload()
- self.failUnlessReallyEqual(get_io_encoding(), expected)
+ expected = sys.platform == "win32" and 'utf-8' or 'koi8-r'
+ _reload()
+ self.failUnlessReallyEqual(get_io_encoding(), expected)
- sys.stdout.encoding = None
- _reload()
- self.failUnlessReallyEqual(get_io_encoding(), expected)
+ mock_stdout.encoding = None
+ _reload()
+ self.failUnlessReallyEqual(get_io_encoding(), expected)
- mock_locale_getpreferredencoding.return_value = None
- _reload()
- self.failUnlessReallyEqual(get_io_encoding(), 'utf-8')
- finally:
- sys.stdout = old_stdout
+ preferredencoding = None
+ _reload()
+ self.failUnlessReallyEqual(get_io_encoding(), 'utf-8')
def test_argv_to_unicode(self):
encodingutil.io_encoding = 'utf-8'
encodingutil.io_encoding = 'koi8-r'
self.failUnlessRaises(UnicodeEncodeError, unicode_to_output, lumiere_nfc)
- @patch('os.listdir')
- def test_no_unicode_normalization(self, mock):
+ def test_no_unicode_normalization(self):
# Pretend to run on a Unicode platform.
- # We normalized to NFC in 1.7beta, but we now don't.
- orig_platform = sys.platform
- try:
- sys.platform = 'darwin'
- mock.return_value = [Artonwall_nfd]
- _reload()
- self.failUnlessReallyEqual(listdir_unicode(u'/dummy'), [Artonwall_nfd])
- finally:
- sys.platform = orig_platform
+ # listdir_unicode normalized to NFC in 1.7beta, but now doesn't.
+
+ def call_os_listdir(path):
+ return [Artonwall_nfd]
+ self.patch(os, 'listdir', call_os_listdir)
+ self.patch(sys, 'platform', 'darwin')
+
+ _reload()
+ self.failUnlessReallyEqual(listdir_unicode(u'/dummy'), [Artonwall_nfd])
+
# The following tests apply only to platforms that don't store filenames as
# Unicode entities on the filesystem.
sys.platform = self.original_platform
_reload()
- @patch('sys.getfilesystemencoding')
- @patch('os.listdir')
- def test_listdir_unicode(self, mock_listdir, mock_getfilesystemencoding):
+ def test_listdir_unicode(self):
# What happens if latin1-encoded filenames are encountered on an UTF-8
# filesystem?
- mock_listdir.return_value = [
- lumiere_nfc.encode('utf-8'),
- lumiere_nfc.encode('latin1')]
+ def call_os_listdir(path):
+ return [
+ lumiere_nfc.encode('utf-8'),
+ lumiere_nfc.encode('latin1')
+ ]
+ self.patch(os, 'listdir', call_os_listdir)
+
+ sys_filesystemencoding = 'utf-8'
+ def call_sys_getfilesystemencoding():
+ return sys_filesystemencoding
+ self.patch(sys, 'getfilesystemencoding', call_sys_getfilesystemencoding)
- mock_getfilesystemencoding.return_value = 'utf-8'
_reload()
self.failUnlessRaises(FilenameEncodingError,
listdir_unicode,
# We're trying to list a directory whose name cannot be represented in
# the filesystem encoding. This should fail.
- mock_getfilesystemencoding.return_value = 'ascii'
+ sys_filesystemencoding = 'ascii'
_reload()
self.failUnlessRaises(FilenameEncodingError,
listdir_unicode,
sys.platform = self.original_platform
_reload()
- @patch('sys.stdout')
- def test_argv_to_unicode(self, mock):
+ def test_argv_to_unicode(self):
if 'argv' not in dir(self):
return
- mock.encoding = self.io_encoding
+ mock_stdout = MockStdout()
+ mock_stdout.encoding = self.io_encoding
+ self.patch(sys, 'stdout', mock_stdout)
+
argu = lumiere_nfc
argv = self.argv
_reload()
def test_unicode_to_url(self):
self.failUnless(unicode_to_url(lumiere_nfc), "lumi\xc3\xa8re")
- @patch('sys.stdout')
- def test_unicode_to_output(self, mock):
+ def test_unicode_to_output(self):
if 'argv' not in dir(self):
return
- mock.encoding = self.io_encoding
+ mock_stdout = MockStdout()
+ mock_stdout.encoding = self.io_encoding
+ self.patch(sys, 'stdout', mock_stdout)
+
_reload()
self.failUnlessReallyEqual(unicode_to_output(lumiere_nfc), self.argv)
_reload()
self.failUnlessReallyEqual(unicode_platform(), matrix[self.platform])
- @patch('sys.getfilesystemencoding')
- @patch('os.listdir')
- def test_listdir_unicode(self, mock_listdir, mock_getfilesystemencoding):
+ def test_listdir_unicode(self):
if 'dirlist' not in dir(self):
return
"that we are testing for the benefit of a different platform."
% (self.filesystem_encoding,))
- mock_listdir.return_value = self.dirlist
- mock_getfilesystemencoding.return_value = self.filesystem_encoding
+ def call_os_listdir(path):
+ return self.dirlist
+ self.patch(os, 'listdir', call_os_listdir)
+
+ def call_sys_getfilesystemencoding():
+ return self.filesystem_encoding
+ self.patch(sys, 'getfilesystemencoding', call_sys_getfilesystemencoding)
_reload()
filenames = listdir_unicode(u'/dummy')
def test_quote_path(self):
self.failUnlessReallyEqual(quote_path([u'foo', u'bar']), "'foo/bar'")
self.failUnlessReallyEqual(quote_path([u'foo', u'bar'], quotemarks=True), "'foo/bar'")
+ self.failUnlessReallyEqual(quote_path([u'foo', u'bar'], quotemarks=False), "foo/bar")
self.failUnlessReallyEqual(quote_path([u'foo', u'\nbar']), '"foo/\\x0abar"')
self.failUnlessReallyEqual(quote_path([u'foo', u'\nbar'], quotemarks=True), '"foo/\\x0abar"')
+ self.failUnlessReallyEqual(quote_path([u'foo', u'\nbar'], quotemarks=False), '"foo/\\x0abar"')
+
+ def win32_other(win32, other):
+ return win32 if sys.platform == "win32" else other
self.failUnlessReallyEqual(quote_local_unicode_path(u"\\\\?\\C:\\foo"),
- "'C:\\foo'" if sys.platform == "win32" else "'\\\\?\\C:\\foo'")
+ win32_other("'C:\\foo'", "'\\\\?\\C:\\foo'"))
+ self.failUnlessReallyEqual(quote_local_unicode_path(u"\\\\?\\C:\\foo", quotemarks=True),
+ win32_other("'C:\\foo'", "'\\\\?\\C:\\foo'"))
+ self.failUnlessReallyEqual(quote_local_unicode_path(u"\\\\?\\C:\\foo", quotemarks=False),
+ win32_other("C:\\foo", "\\\\?\\C:\\foo"))
self.failUnlessReallyEqual(quote_local_unicode_path(u"\\\\?\\UNC\\foo\\bar"),
- "'\\\\foo\\bar'" if sys.platform == "win32" else "'\\\\?\\UNC\\foo\\bar'")
+ win32_other("'\\\\foo\\bar'", "'\\\\?\\UNC\\foo\\bar'"))
+ self.failUnlessReallyEqual(quote_local_unicode_path(u"\\\\?\\UNC\\foo\\bar", quotemarks=True),
+ win32_other("'\\\\foo\\bar'", "'\\\\?\\UNC\\foo\\bar'"))
+ self.failUnlessReallyEqual(quote_local_unicode_path(u"\\\\?\\UNC\\foo\\bar", quotemarks=False),
+ win32_other("\\\\foo\\bar", "\\\\?\\UNC\\foo\\bar"))
class UbuntuKarmicUTF8(EncodingUtil, unittest.TestCase):
filesystem_encoding = '646'
io_encoding = '646'
# Oops, I cannot write filenames containing non-ascii characters
+
+
+class TestToFromStr(ReallyEqualMixin, unittest.TestCase):
+ def test_to_str(self):
+ self.failUnlessReallyEqual(to_str("foo"), "foo")
+ self.failUnlessReallyEqual(to_str("lumi\xc3\xa8re"), "lumi\xc3\xa8re")
+ self.failUnlessReallyEqual(to_str("\xFF"), "\xFF") # passes through invalid UTF-8 -- is this what we want?
+ self.failUnlessReallyEqual(to_str(u"lumi\u00E8re"), "lumi\xc3\xa8re")
+ self.failUnlessReallyEqual(to_str(None), None)
+
+ def test_from_utf8_or_none(self):
+ self.failUnlessRaises(AssertionError, from_utf8_or_none, u"foo")
+ self.failUnlessReallyEqual(from_utf8_or_none("lumi\xc3\xa8re"), u"lumi\u00E8re")
+ self.failUnlessReallyEqual(from_utf8_or_none(None), None)
+ self.failUnlessRaises(UnicodeDecodeError, from_utf8_or_none, "\xFF")
projects / tahoe-lafs / tahoe-lafs.git / blobdiff