mirror of
https://github.com/samba-team/samba.git
synced 2024-12-23 17:34:34 +03:00
aecb2b779b
This made Python 2's print behave like Python 3's print(). In some cases, where we had: from __future__ import print_function """Intended module documentation...""" this will have the side effect of making the intended module documentation work as the actual module documentation (i.e. becoming __doc__), because it is once again the first statement in the module. Signed-off-by: Douglas Bagnall <douglas.bagnall@catalyst.net.nz> Reviewed-by: Andrew Bartlett <abartlet@samba.org>
380 lines
15 KiB
Python
380 lines
15 KiB
Python
#!/usr/bin/env python3
|
||
# -*- coding: utf-8 -*-
|
||
# Originally based on ./sam.py
|
||
from unicodedata import normalize
|
||
import locale
|
||
locale.setlocale(locale.LC_ALL, ('en_US', 'UTF-8'))
|
||
|
||
import optparse
|
||
import sys
|
||
import os
|
||
import re
|
||
|
||
sys.path.insert(0, "bin/python")
|
||
import samba
|
||
from samba.tests.subunitrun import SubunitOptions, TestProgram
|
||
from samba.common import cmp
|
||
from functools import cmp_to_key
|
||
import samba.getopt as options
|
||
|
||
from samba.auth import system_session
|
||
import ldb
|
||
from samba.samdb import SamDB
|
||
|
||
parser = optparse.OptionParser("sort.py [options] <host>")
|
||
sambaopts = options.SambaOptions(parser)
|
||
parser.add_option_group(sambaopts)
|
||
parser.add_option_group(options.VersionOptions(parser))
|
||
# use command line creds if available
|
||
credopts = options.CredentialsOptions(parser)
|
||
parser.add_option_group(credopts)
|
||
subunitopts = SubunitOptions(parser)
|
||
parser.add_option_group(subunitopts)
|
||
|
||
parser.add_option('--elements', type='int', default=33,
|
||
help="use this many elements in the tests")
|
||
|
||
opts, args = parser.parse_args()
|
||
|
||
if len(args) < 1:
|
||
parser.print_usage()
|
||
sys.exit(1)
|
||
|
||
datadir = os.getenv("DATA_DIR", None)
|
||
if not datadir:
|
||
print("Please specify the location of the sort expected results with env variable DATA_DIR")
|
||
sys.exit(1)
|
||
|
||
host = os.getenv("SERVER", None)
|
||
if not host:
|
||
print("Please specify the host with env variable SERVER")
|
||
sys.exit(1)
|
||
|
||
lp = sambaopts.get_loadparm()
|
||
creds = credopts.get_credentials(lp)
|
||
|
||
|
||
def norm(x):
|
||
if not isinstance(x, str):
|
||
x = x.decode('utf8')
|
||
return normalize('NFKC', x).upper()
|
||
|
||
|
||
# Python, Windows, and Samba all sort the following sequence in
|
||
# drastically different ways. The order here is what you get from
|
||
# Windows2012R2.
|
||
FIENDISH_TESTS = [' ', ' e', '\t-\t', '\n\t\t', '!@#!@#!', '¼', '¹', '1',
|
||
'1/4', '1⁄4', '1\xe2\x81\x845', '3', 'abc', 'fo\x00od',
|
||
|
||
# Here we also had '\x00food', but that seems to sort
|
||
# non-deterministically on Windows vis-a-vis 'fo\x00od'.
|
||
|
||
'kōkako', 'ŋđ¼³ŧ “«đð', 'ŋđ¼³ŧ“«đð',
|
||
'sorttest', 'sorttēst11,', 'śorttest2', 'śoRttest2',
|
||
'ś-o-r-t-t-e-s-t-2', 'soRTTēst2,', 'ṡorttest4', 'ṡorttesT4',
|
||
'sörttest-5', 'sÖrttest-5', 'so-rttest7,', '桑巴']
|
||
|
||
|
||
class BaseSortTests(samba.tests.TestCase):
|
||
avoid_tricky_sort = False
|
||
maxDiff = 2000
|
||
|
||
def create_user(self, i, n, prefix='sorttest', suffix='', attrs=None,
|
||
tricky=False):
|
||
name = "%s%d%s" % (prefix, i, suffix)
|
||
user = {
|
||
'cn': name,
|
||
"objectclass": "user",
|
||
'givenName': "abcdefghijklmnopqrstuvwxyz"[i % 26],
|
||
"roomNumber": "%sb\x00c" % (n - i),
|
||
# with python3 re.sub(r'[^\w,.]', repl, string) doesn't
|
||
# work as expected with unicode as value for carLicense
|
||
"carLicense": "XXXXXXXXX" if self.avoid_tricky_sort else "后来经",
|
||
"employeeNumber": "%s%sx" % (abs(i * (99 - i)), '\n' * (i & 255)),
|
||
"accountExpires": "%s" % (10 ** 9 + 1000000 * i),
|
||
"msTSExpireDate4": "19%02d0101010000.0Z" % (i % 100),
|
||
"flags": str(i * (n - i)),
|
||
"serialNumber": "abc %s%s%s" % ('AaBb |-/'[i & 7],
|
||
' 3z}'[i & 3],
|
||
'"@'[i & 1],),
|
||
"comment": "Favourite colour is %d" % (n % (i + 1)),
|
||
}
|
||
|
||
if self.avoid_tricky_sort:
|
||
# We are not even going to try passing tests that assume
|
||
# some kind of Unicode awareness.
|
||
for k, v in user.items():
|
||
user[k] = re.sub(r'[^\w,.]', 'X', v)
|
||
else:
|
||
# Add some even trickier ones!
|
||
fiendish_index = i % len(FIENDISH_TESTS)
|
||
user.update({
|
||
# Sort doesn't look past a NUL byte.
|
||
"photo": "\x00%d" % (n - i),
|
||
"audio": "%sn octet string %s%s ♫♬\x00lalala" % ('Aa'[i & 1],
|
||
chr(i & 255),
|
||
i),
|
||
"displayNamePrintable": "%d\x00%c" % (i, i & 255),
|
||
"adminDisplayName": "%d\x00b" % (n - i),
|
||
"title": "%d%sb" % (n - i, '\x00' * i),
|
||
|
||
# Names that vary only in case. Windows returns
|
||
# equivalent addresses in the order they were put
|
||
# in ('a st', 'A st',...). We don't check that.
|
||
"street": "%s st" % (chr(65 | (i & 14) | ((i & 1) * 32))),
|
||
|
||
"streetAddress": FIENDISH_TESTS[fiendish_index],
|
||
"postalAddress": FIENDISH_TESTS[-fiendish_index],
|
||
})
|
||
|
||
if attrs is not None:
|
||
user.update(attrs)
|
||
|
||
user['dn'] = "cn=%s,%s" % (user['cn'], self.ou)
|
||
|
||
self.users.append(user)
|
||
self.ldb.add(user)
|
||
return user
|
||
|
||
def setUp(self):
|
||
super(BaseSortTests, self).setUp()
|
||
self.ldb = SamDB(host, credentials=creds,
|
||
session_info=system_session(lp), lp=lp)
|
||
|
||
self.base_dn = self.ldb.domain_dn()
|
||
self.ou = "ou=sort,%s" % self.base_dn
|
||
if False:
|
||
try:
|
||
self.ldb.delete(self.ou, ['tree_delete:1'])
|
||
except ldb.LdbError as e:
|
||
print("tried deleting %s, got error %s" % (self.ou, e))
|
||
|
||
self.ldb.add({
|
||
"dn": self.ou,
|
||
"objectclass": "organizationalUnit"})
|
||
self.users = []
|
||
n = opts.elements
|
||
for i in range(n):
|
||
self.create_user(i, n)
|
||
|
||
attrs = set(self.users[0].keys()) - set([
|
||
'objectclass', 'dn'])
|
||
self.binary_sorted_keys = attrs.intersection(['audio',
|
||
'photo',
|
||
"msTSExpireDate4",
|
||
'serialNumber',
|
||
"displayNamePrintable"])
|
||
|
||
self.numeric_sorted_keys = attrs.intersection(['flags',
|
||
'accountExpires'])
|
||
|
||
self.timestamp_keys = attrs.intersection(['msTSExpireDate4'])
|
||
|
||
self.int64_keys = set(['accountExpires'])
|
||
|
||
self.locale_sorted_keys = [x for x in attrs if
|
||
x not in (self.binary_sorted_keys |
|
||
self.numeric_sorted_keys)]
|
||
|
||
self.expected_results = {}
|
||
self.expected_results_binary = {}
|
||
|
||
for k in self.binary_sorted_keys:
|
||
forward = sorted((x[k] for x in self.users))
|
||
reverse = list(reversed(forward))
|
||
self.expected_results_binary[k] = (forward, reverse)
|
||
|
||
# FYI: Expected result data was generated from the old
|
||
# code that was manually sorting (while executing with
|
||
# python2)
|
||
# The resulting data was injected into the data file with
|
||
# code similar to:
|
||
#
|
||
# for k in self.expected_results:
|
||
# f.write("%s = %s\n" % (k, repr(self.expected_results[k][0])))
|
||
|
||
f = open(self.results_file, "r")
|
||
for line in f:
|
||
if len(line.split('=', 1)) == 2:
|
||
key = line.split('=', 1)[0].strip()
|
||
value = line.split('=', 1)[1].strip()
|
||
if value.startswith('['):
|
||
import ast
|
||
fwd_list = ast.literal_eval(value)
|
||
rev_list = list(reversed(fwd_list))
|
||
self.expected_results[key] = (fwd_list, rev_list)
|
||
f.close()
|
||
def tearDown(self):
|
||
super(BaseSortTests, self).tearDown()
|
||
self.ldb.delete(self.ou, ['tree_delete:1'])
|
||
|
||
def _test_server_sort_default(self):
|
||
attrs = self.locale_sorted_keys
|
||
|
||
for attr in attrs:
|
||
for rev in (0, 1):
|
||
res = self.ldb.search(self.ou,
|
||
scope=ldb.SCOPE_ONELEVEL, attrs=[attr],
|
||
controls=["server_sort:1:%d:%s" %
|
||
(rev, attr)])
|
||
self.assertEqual(len(res), len(self.users))
|
||
|
||
expected_order = self.expected_results[attr][rev]
|
||
received_order = [norm(x[attr][0]) for x in res]
|
||
if expected_order != received_order:
|
||
print(attr, ['forward', 'reverse'][rev])
|
||
print("expected", expected_order)
|
||
print("received", received_order)
|
||
print("unnormalised:", [x[attr][0] for x in res])
|
||
print("unnormalised: «%s»" % '» «'.join(str(x[attr][0])
|
||
for x in res))
|
||
self.assertEqual(expected_order, received_order)
|
||
|
||
def _test_server_sort_binary(self):
|
||
for attr in self.binary_sorted_keys:
|
||
for rev in (0, 1):
|
||
res = self.ldb.search(self.ou,
|
||
scope=ldb.SCOPE_ONELEVEL, attrs=[attr],
|
||
controls=["server_sort:1:%d:%s" %
|
||
(rev, attr)])
|
||
|
||
self.assertEqual(len(res), len(self.users))
|
||
expected_order = self.expected_results_binary[attr][rev]
|
||
received_order = [str(x[attr][0]) for x in res]
|
||
if expected_order != received_order:
|
||
print(attr)
|
||
print(expected_order)
|
||
print(received_order)
|
||
self.assertEqual(expected_order, received_order)
|
||
|
||
def _test_server_sort_us_english(self):
|
||
# Windows doesn't support many matching rules, but does allow
|
||
# the locale specific sorts -- if it has the locale installed.
|
||
# The most reliable locale is the default US English, which
|
||
# won't change the sort order.
|
||
|
||
for lang, oid in [('en_US', '1.2.840.113556.1.4.1499'),
|
||
]:
|
||
|
||
for attr in self.locale_sorted_keys:
|
||
for rev in (0, 1):
|
||
res = self.ldb.search(self.ou,
|
||
scope=ldb.SCOPE_ONELEVEL,
|
||
attrs=[attr],
|
||
controls=["server_sort:1:%d:%s:%s" %
|
||
(rev, attr, oid)])
|
||
|
||
self.assertTrue(len(res) == len(self.users))
|
||
expected_order = self.expected_results[attr][rev]
|
||
received_order = [norm(x[attr][0]) for x in res]
|
||
if expected_order != received_order:
|
||
print(attr, lang)
|
||
print(['forward', 'reverse'][rev])
|
||
print("expected: ", expected_order)
|
||
print("received: ", received_order)
|
||
print("unnormalised:", [x[attr][0] for x in res])
|
||
print("unnormalised: «%s»" % '» «'.join(str(x[attr][0])
|
||
for x in res))
|
||
|
||
self.assertEqual(expected_order, received_order)
|
||
|
||
def _test_server_sort_different_attr(self):
|
||
|
||
def cmp_locale(a, b):
|
||
return locale.strcoll(a[0], b[0])
|
||
|
||
def cmp_binary(a, b):
|
||
return cmp(a[0], b[0])
|
||
|
||
def cmp_numeric(a, b):
|
||
return cmp(int(a[0]), int(b[0]))
|
||
|
||
# For testing simplicity, the attributes in here need to be
|
||
# unique for each user. Otherwise there are multiple possible
|
||
# valid answers.
|
||
sort_functions = {'cn': cmp_binary,
|
||
"employeeNumber": cmp_locale,
|
||
"accountExpires": cmp_numeric,
|
||
"msTSExpireDate4": cmp_binary}
|
||
attrs = list(sort_functions.keys())
|
||
attr_pairs = zip(attrs, attrs[1:] + attrs[:1])
|
||
|
||
for sort_attr, result_attr in attr_pairs:
|
||
forward = sorted(((norm(x[sort_attr]), norm(x[result_attr]))
|
||
for x in self.users),
|
||
key=cmp_to_key(sort_functions[sort_attr]))
|
||
reverse = list(reversed(forward))
|
||
|
||
for rev in (0, 1):
|
||
res = self.ldb.search(self.ou,
|
||
scope=ldb.SCOPE_ONELEVEL,
|
||
attrs=[result_attr],
|
||
controls=["server_sort:1:%d:%s" %
|
||
(rev, sort_attr)])
|
||
self.assertEqual(len(res), len(self.users))
|
||
pairs = (forward, reverse)[rev]
|
||
|
||
expected_order = [x[1] for x in pairs]
|
||
received_order = [norm(x[result_attr][0]) for x in res]
|
||
|
||
if expected_order != received_order:
|
||
print(sort_attr, result_attr, ['forward', 'reverse'][rev])
|
||
print("expected", expected_order)
|
||
print("received", received_order)
|
||
print("unnormalised:", [x[result_attr][0] for x in res])
|
||
print("unnormalised: «%s»" % '» «'.join(str(x[result_attr][0])
|
||
for x in res))
|
||
print("pairs:", pairs)
|
||
# There are bugs in Windows that we don't want (or
|
||
# know how) to replicate regarding timestamp sorting.
|
||
# Let's remind ourselves.
|
||
if result_attr == "msTSExpireDate4":
|
||
print('-' * 72)
|
||
print("This test fails against Windows with the "
|
||
"default number of elements (33).")
|
||
print("Try with --elements=27 (or similar).")
|
||
print('-' * 72)
|
||
|
||
self.assertEqual(expected_order, received_order)
|
||
for x in res:
|
||
if sort_attr in x:
|
||
self.fail('the search for %s should not return %s' %
|
||
(result_attr, sort_attr))
|
||
|
||
|
||
class SimpleSortTests(BaseSortTests):
|
||
avoid_tricky_sort = True
|
||
results_file = os.path.join(datadir, "simplesort.expected")
|
||
def test_server_sort_different_attr(self):
|
||
self._test_server_sort_different_attr()
|
||
|
||
def test_server_sort_default(self):
|
||
self._test_server_sort_default()
|
||
|
||
def test_server_sort_binary(self):
|
||
self._test_server_sort_binary()
|
||
|
||
def test_server_sort_us_english(self):
|
||
self._test_server_sort_us_english()
|
||
|
||
|
||
class UnicodeSortTests(BaseSortTests):
|
||
avoid_tricky_sort = False
|
||
results_file = os.path.join(datadir, "unicodesort.expected")
|
||
|
||
def test_server_sort_default(self):
|
||
self._test_server_sort_default()
|
||
|
||
def test_server_sort_us_english(self):
|
||
self._test_server_sort_us_english()
|
||
|
||
def test_server_sort_different_attr(self):
|
||
self._test_server_sort_different_attr()
|
||
|
||
|
||
if "://" not in host:
|
||
if os.path.isfile(host):
|
||
host = "tdb://%s" % host
|
||
else:
|
||
host = "ldap://%s" % host
|