1
0
mirror of https://github.com/samba-team/samba.git synced 2024-12-25 23:21:54 +03:00
samba-mirror/python/samba/tests/strings.py
Jelmer Vernooij 87afc3aee1 Move python modules from source4/scripting/python/ to python/.
Reviewed-by: Andrew Bartlett <abartlet@samba.org>

Autobuild-User(master): Andrew Bartlett <abartlet@samba.org>
Autobuild-Date(master): Sat Mar  2 03:57:34 CET 2013 on sn-devel-104
2013-03-02 03:57:34 +01:00

104 lines
4.1 KiB
Python

# subunit test cases for Samba string functions.
# Copyright (C) 2003 by Martin Pool <mbp@samba.org>
# Copyright (C) 2011 Andrew Bartlett
#
# This program is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License as published by
# the Free Software Foundation; either version 3 of the License, or
# (at your option) any later version.
#
# This program is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License
# along with this program. If not, see <http://www.gnu.org/licenses/>.
#
# XXX: All this code assumes that the Unix character set is UTF-8,
# which is the most common setting. I guess it would be better to
# force it to that value while running the tests. I'm not sure of the
# best way to do that yet.
#
# -- mbp
from unicodenames import *
import samba.tests
from samba import strcasecmp_m, strstr_m
def signum(a):
if a < 0:
return -1
elif a > 0:
return +1
else:
return 0
class strcasecmp_m_Tests(samba.tests.TestCase):
"""String comparisons in simple ASCII and unicode"""
def test_strcasecmp_m(self):
# A, B, strcasecmp(A, B)
cases = [('hello', 'hello', 0),
('hello', 'goodbye', +1),
('goodbye', 'hello', -1),
('hell', 'hello', -1),
('', '', 0),
('a', '', +1),
('', 'a', -1),
('a', 'A', 0),
('aa', 'aA', 0),
('Aa', 'aa', 0),
('longstring ' * 100, 'longstring ' * 100, 0),
('longstring ' * 100, 'longstring ' * 100 + 'a', -1),
('longstring ' * 100 + 'a', 'longstring ' * 100, +1),
(KATAKANA_LETTER_A, KATAKANA_LETTER_A, 0),
(KATAKANA_LETTER_A, 'a', 1),
]
for a, b, expect in cases:
self.assertEquals(signum(strcasecmp_m(a.encode('utf-8'),
b.encode('utf-8'))),
expect)
class strstr_m_Tests(samba.tests.TestCase):
"""strstr_m tests in simple ASCII and unicode strings"""
def test_strstr_m(self):
# A, B, strstr_m(A, B)
cases = [('hello', 'hello', 'hello'),
('hello', 'goodbye', None),
('goodbye', 'hello', None),
('hell', 'hello', None),
('hello', 'hell', 'hello'),
('', '', ''),
('a', '', 'a'),
('', 'a', None),
('a', 'A', None),
('aa', 'aA', None),
('Aa', 'aa', None),
('%v foo', '%v', '%v foo'),
('foo %v foo', '%v', '%v foo'),
('foo %v', '%v', '%v'),
('longstring ' * 100, 'longstring ' * 99, 'longstring ' * 100),
('longstring ' * 99, 'longstring ' * 100, None),
('longstring a' * 99, 'longstring ' * 100 + 'a', None),
('longstring ' * 100 + 'a', 'longstring ' * 100, 'longstring ' * 100 + 'a'),
(KATAKANA_LETTER_A, KATAKANA_LETTER_A + 'bcd', None),
(KATAKANA_LETTER_A + 'bcde', KATAKANA_LETTER_A + 'bcd', KATAKANA_LETTER_A + 'bcde'),
('d'+KATAKANA_LETTER_A + 'bcd', KATAKANA_LETTER_A + 'bcd', KATAKANA_LETTER_A + 'bcd'),
('d'+KATAKANA_LETTER_A + 'bd', KATAKANA_LETTER_A + 'bcd', None),
('e'+KATAKANA_LETTER_A + 'bcdf', KATAKANA_LETTER_A + 'bcd', KATAKANA_LETTER_A + 'bcdf'),
(KATAKANA_LETTER_A, KATAKANA_LETTER_A + 'bcd', None),
(KATAKANA_LETTER_A*3, 'a', None),
]
for a, b, expect in cases:
if expect is not None:
expect = expect.encode('utf-8')
self.assertEquals(strstr_m(a.encode('utf-8'),
b.encode('utf-8')),
expect)