1
# Copyright (C) 2005-2011 Canonical Ltd
3
# This program is free software; you can redistribute it and/or modify
4
# it under the terms of the GNU General Public License as published by
5
# the Free Software Foundation; either version 2 of the License, or
6
# (at your option) any later version.
8
# This program is distributed in the hope that it will be useful,
9
# but WITHOUT ANY WARRANTY; without even the implied warranty of
10
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
11
# GNU General Public License for more details.
13
# You should have received a copy of the GNU General Public License
14
# along with this program; if not, write to the Free Software
15
# Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA
17
"""Testing framework extensions"""
19
# NOTE: Some classes in here use camelCaseNaming() rather than
20
# underscore_naming(). That's for consistency with unittest; it's not the
21
# general style of bzrlib. Please continue that consistency when adding e.g.
22
# new assertFoo() methods.
27
from cStringIO import StringIO
50
# nb: check this before importing anything else from within it
51
_testtools_version = getattr(testtools, '__version__', ())
52
if _testtools_version < (0, 9, 5):
53
raise ImportError("need at least testtools 0.9.5: %s is %r"
54
% (testtools.__file__, _testtools_version))
55
from testtools import content
62
commands as _mod_commands,
72
plugin as _mod_plugin,
79
transport as _mod_transport,
85
# lsprof not available
87
from bzrlib.smart import client, request
88
from bzrlib.transport import (
92
from bzrlib.symbol_versioning import (
96
from bzrlib.tests import (
102
from bzrlib.ui import NullProgressView
103
from bzrlib.ui.text import TextUIFactory
104
from bzrlib.tests.features import _CompatabilityThunkFeature
106
# Mark this python module as being part of the implementation
107
# of unittest: this gives us better tracebacks where the last
108
# shown frame is the test code, not our assertXYZ.
111
default_transport = test_server.LocalURLServer
114
_unitialized_attr = object()
115
"""A sentinel needed to act as a default value in a method signature."""
118
# Subunit result codes, defined here to prevent a hard dependency on subunit.
122
# These are intentionally brought into this namespace. That way plugins, etc
123
# can just "from bzrlib.tests import TestCase, TestLoader, etc"
124
TestSuite = TestUtil.TestSuite
125
TestLoader = TestUtil.TestLoader
127
# Tests should run in a clean and clearly defined environment. The goal is to
128
# keep them isolated from the running environment as mush as possible. The test
129
# framework ensures the variables defined below are set (or deleted if the
130
# value is None) before a test is run and reset to their original value after
131
# the test is run. Generally if some code depends on an environment variable,
132
# the tests should start without this variable in the environment. There are a
133
# few exceptions but you shouldn't violate this rule lightly.
137
# bzr now uses the Win32 API and doesn't rely on APPDATA, but the
138
# tests do check our impls match APPDATA
139
'BZR_EDITOR': None, # test_msgeditor manipulates this variable
143
'BZREMAIL': None, # may still be present in the environment
144
'EMAIL': 'jrandom@example.com', # set EMAIL as bzr does not guess
145
'BZR_PROGRESS_BAR': None,
146
# This should trap leaks to ~/.bzr.log. This occurs when tests use TestCase
147
# as a base class instead of TestCaseInTempDir. Tests inheriting from
148
# TestCase should not use disk resources, BZR_LOG is one.
149
'BZR_LOG': '/you-should-use-TestCaseInTempDir-if-you-need-a-log-file',
150
'BZR_PLUGIN_PATH': None,
151
'BZR_DISABLE_PLUGINS': None,
152
'BZR_PLUGINS_AT': None,
153
'BZR_CONCURRENCY': None,
154
# Make sure that any text ui tests are consistent regardless of
155
# the environment the test case is run in; you may want tests that
156
# test other combinations. 'dumb' is a reasonable guess for tests
157
# going to a pipe or a StringIO.
163
'SSH_AUTH_SOCK': None,
173
# Nobody cares about ftp_proxy, FTP_PROXY AFAIK. So far at
174
# least. If you do (care), please update this comment
178
'BZR_REMOTE_PATH': None,
179
# Generally speaking, we don't want apport reporting on crashes in
180
# the test envirnoment unless we're specifically testing apport,
181
# so that it doesn't leak into the real system environment. We
182
# use an env var so it propagates to subprocesses.
183
'APPORT_DISABLE': '1',
187
def override_os_environ(test, env=None):
188
"""Modify os.environ keeping a copy.
190
:param test: A test instance
192
:param env: A dict containing variable definitions to be installed
195
env = isolated_environ
196
test._original_os_environ = dict([(var, value)
197
for var, value in os.environ.iteritems()])
198
for var, value in env.iteritems():
199
osutils.set_or_unset_env(var, value)
200
if var not in test._original_os_environ:
201
# The var is new, add it with a value of None, so
202
# restore_os_environ will delete it
203
test._original_os_environ[var] = None
206
def restore_os_environ(test):
207
"""Restore os.environ to its original state.
209
:param test: A test instance previously passed to override_os_environ.
211
for var, value in test._original_os_environ.iteritems():
212
# Restore the original value (or delete it if the value has been set to
213
# None in override_os_environ).
214
osutils.set_or_unset_env(var, value)
217
def _clear__type_equality_funcs(test):
218
"""Cleanup bound methods stored on TestCase instances
220
Clear the dict breaking a few (mostly) harmless cycles in the affected
221
unittests released with Python 2.6 and initial Python 2.7 versions.
223
For a few revisions between Python 2.7.1 and Python 2.7.2 that annoyingly
224
shipped in Oneiric, an object with no clear method was used, hence the
225
extra complications, see bug 809048 for details.
227
type_equality_funcs = getattr(test, "_type_equality_funcs", None)
228
if type_equality_funcs is not None:
229
tef_clear = getattr(type_equality_funcs, "clear", None)
230
if tef_clear is None:
231
tef_instance_dict = getattr(type_equality_funcs, "__dict__", None)
232
if tef_instance_dict is not None:
233
tef_clear = tef_instance_dict.clear
234
if tef_clear is not None:
238
class ExtendedTestResult(testtools.TextTestResult):
239
"""Accepts, reports and accumulates the results of running tests.
241
Compared to the unittest version this class adds support for
242
profiling, benchmarking, stopping as soon as a test fails, and
243
skipping tests. There are further-specialized subclasses for
244
different types of display.
246
When a test finishes, in whatever way, it calls one of the addSuccess,
247
addFailure or addError classes. These in turn may redirect to a more
248
specific case for the special test results supported by our extended
251
Note that just one of these objects is fed the results from many tests.
256
def __init__(self, stream, descriptions, verbosity,
260
"""Construct new TestResult.
262
:param bench_history: Optionally, a writable file object to accumulate
265
testtools.TextTestResult.__init__(self, stream)
266
if bench_history is not None:
267
from bzrlib.version import _get_bzr_source_tree
268
src_tree = _get_bzr_source_tree()
271
revision_id = src_tree.get_parent_ids()[0]
273
# XXX: if this is a brand new tree, do the same as if there
277
# XXX: If there's no branch, what should we do?
279
bench_history.write("--date %s %s\n" % (time.time(), revision_id))
280
self._bench_history = bench_history
281
self.ui = ui.ui_factory
284
self.failure_count = 0
285
self.known_failure_count = 0
287
self.not_applicable_count = 0
288
self.unsupported = {}
290
self._overall_start_time = time.time()
291
self._strict = strict
292
self._first_thread_leaker_id = None
293
self._tests_leaking_threads_count = 0
294
self._traceback_from_test = None
296
def stopTestRun(self):
299
stopTime = time.time()
300
timeTaken = stopTime - self.startTime
301
# GZ 2010-07-19: Seems testtools has no printErrors method, and though
302
# the parent class method is similar have to duplicate
303
self._show_list('ERROR', self.errors)
304
self._show_list('FAIL', self.failures)
305
self.stream.write(self.sep2)
306
self.stream.write("%s %d test%s in %.3fs\n\n" % (actionTaken,
307
run, run != 1 and "s" or "", timeTaken))
308
if not self.wasSuccessful():
309
self.stream.write("FAILED (")
310
failed, errored = map(len, (self.failures, self.errors))
312
self.stream.write("failures=%d" % failed)
314
if failed: self.stream.write(", ")
315
self.stream.write("errors=%d" % errored)
316
if self.known_failure_count:
317
if failed or errored: self.stream.write(", ")
318
self.stream.write("known_failure_count=%d" %
319
self.known_failure_count)
320
self.stream.write(")\n")
322
if self.known_failure_count:
323
self.stream.write("OK (known_failures=%d)\n" %
324
self.known_failure_count)
326
self.stream.write("OK\n")
327
if self.skip_count > 0:
328
skipped = self.skip_count
329
self.stream.write('%d test%s skipped\n' %
330
(skipped, skipped != 1 and "s" or ""))
332
for feature, count in sorted(self.unsupported.items()):
333
self.stream.write("Missing feature '%s' skipped %d tests.\n" %
336
ok = self.wasStrictlySuccessful()
338
ok = self.wasSuccessful()
339
if self._first_thread_leaker_id:
341
'%s is leaking threads among %d leaking tests.\n' % (
342
self._first_thread_leaker_id,
343
self._tests_leaking_threads_count))
344
# We don't report the main thread as an active one.
346
'%d non-main threads were left active in the end.\n'
347
% (len(self._active_threads) - 1))
349
def getDescription(self, test):
352
def _extractBenchmarkTime(self, testCase, details=None):
353
"""Add a benchmark time for the current test case."""
354
if details and 'benchtime' in details:
355
return float(''.join(details['benchtime'].iter_bytes()))
356
return getattr(testCase, "_benchtime", None)
358
def _elapsedTestTimeString(self):
359
"""Return a time string for the overall time the current test has taken."""
360
return self._formatTime(self._delta_to_float(
361
self._now() - self._start_datetime))
363
def _testTimeString(self, testCase):
364
benchmark_time = self._extractBenchmarkTime(testCase)
365
if benchmark_time is not None:
366
return self._formatTime(benchmark_time) + "*"
368
return self._elapsedTestTimeString()
370
def _formatTime(self, seconds):
371
"""Format seconds as milliseconds with leading spaces."""
372
# some benchmarks can take thousands of seconds to run, so we need 8
374
return "%8dms" % (1000 * seconds)
376
def _shortened_test_description(self, test):
378
what = re.sub(r'^bzrlib\.tests\.', '', what)
381
# GZ 2010-10-04: Cloned tests may end up harmlessly calling this method
382
# multiple times in a row, because the handler is added for
383
# each test but the container list is shared between cases.
384
# See lp:498869 lp:625574 and lp:637725 for background.
385
def _record_traceback_from_test(self, exc_info):
386
"""Store the traceback from passed exc_info tuple till"""
387
self._traceback_from_test = exc_info[2]
389
def startTest(self, test):
390
super(ExtendedTestResult, self).startTest(test)
394
self.report_test_start(test)
395
test.number = self.count
396
self._recordTestStartTime()
397
# Make testtools cases give us the real traceback on failure
398
addOnException = getattr(test, "addOnException", None)
399
if addOnException is not None:
400
addOnException(self._record_traceback_from_test)
401
# Only check for thread leaks on bzrlib derived test cases
402
if isinstance(test, TestCase):
403
test.addCleanup(self._check_leaked_threads, test)
405
def stopTest(self, test):
406
super(ExtendedTestResult, self).stopTest(test)
407
# Manually break cycles, means touching various private things but hey
408
getDetails = getattr(test, "getDetails", None)
409
if getDetails is not None:
411
_clear__type_equality_funcs(test)
412
self._traceback_from_test = None
414
def startTests(self):
415
self.report_tests_starting()
416
self._active_threads = threading.enumerate()
418
def _check_leaked_threads(self, test):
419
"""See if any threads have leaked since last call
421
A sample of live threads is stored in the _active_threads attribute,
422
when this method runs it compares the current live threads and any not
423
in the previous sample are treated as having leaked.
425
now_active_threads = set(threading.enumerate())
426
threads_leaked = now_active_threads.difference(self._active_threads)
428
self._report_thread_leak(test, threads_leaked, now_active_threads)
429
self._tests_leaking_threads_count += 1
430
if self._first_thread_leaker_id is None:
431
self._first_thread_leaker_id = test.id()
432
self._active_threads = now_active_threads
434
def _recordTestStartTime(self):
435
"""Record that a test has started."""
436
self._start_datetime = self._now()
438
def addError(self, test, err):
439
"""Tell result that test finished with an error.
441
Called from the TestCase run() method when the test
442
fails with an unexpected error.
444
self._post_mortem(self._traceback_from_test)
445
super(ExtendedTestResult, self).addError(test, err)
446
self.error_count += 1
447
self.report_error(test, err)
451
def addFailure(self, test, err):
452
"""Tell result that test failed.
454
Called from the TestCase run() method when the test
455
fails because e.g. an assert() method failed.
457
self._post_mortem(self._traceback_from_test)
458
super(ExtendedTestResult, self).addFailure(test, err)
459
self.failure_count += 1
460
self.report_failure(test, err)
464
def addSuccess(self, test, details=None):
465
"""Tell result that test completed successfully.
467
Called from the TestCase run()
469
if self._bench_history is not None:
470
benchmark_time = self._extractBenchmarkTime(test, details)
471
if benchmark_time is not None:
472
self._bench_history.write("%s %s\n" % (
473
self._formatTime(benchmark_time),
475
self.report_success(test)
476
super(ExtendedTestResult, self).addSuccess(test)
477
test._log_contents = ''
479
def addExpectedFailure(self, test, err):
480
self.known_failure_count += 1
481
self.report_known_failure(test, err)
483
def addUnexpectedSuccess(self, test, details=None):
484
"""Tell result the test unexpectedly passed, counting as a failure
486
When the minimum version of testtools required becomes 0.9.8 this
487
can be updated to use the new handling there.
489
super(ExtendedTestResult, self).addFailure(test, details=details)
490
self.failure_count += 1
491
self.report_unexpected_success(test,
492
"".join(details["reason"].iter_text()))
496
def addNotSupported(self, test, feature):
497
"""The test will not be run because of a missing feature.
499
# this can be called in two different ways: it may be that the
500
# test started running, and then raised (through requireFeature)
501
# UnavailableFeature. Alternatively this method can be called
502
# while probing for features before running the test code proper; in
503
# that case we will see startTest and stopTest, but the test will
504
# never actually run.
505
self.unsupported.setdefault(str(feature), 0)
506
self.unsupported[str(feature)] += 1
507
self.report_unsupported(test, feature)
509
def addSkip(self, test, reason):
510
"""A test has not run for 'reason'."""
512
self.report_skip(test, reason)
514
def addNotApplicable(self, test, reason):
515
self.not_applicable_count += 1
516
self.report_not_applicable(test, reason)
518
def _count_stored_tests(self):
519
"""Count of tests instances kept alive due to not succeeding"""
520
return self.error_count + self.failure_count + self.known_failure_count
522
def _post_mortem(self, tb=None):
523
"""Start a PDB post mortem session."""
524
if os.environ.get('BZR_TEST_PDB', None):
528
def progress(self, offset, whence):
529
"""The test is adjusting the count of tests to run."""
530
if whence == SUBUNIT_SEEK_SET:
531
self.num_tests = offset
532
elif whence == SUBUNIT_SEEK_CUR:
533
self.num_tests += offset
535
raise errors.BzrError("Unknown whence %r" % whence)
537
def report_tests_starting(self):
538
"""Display information before the test run begins"""
539
if getattr(sys, 'frozen', None) is None:
540
bzr_path = osutils.realpath(sys.argv[0])
542
bzr_path = sys.executable
544
'bzr selftest: %s\n' % (bzr_path,))
547
bzrlib.__path__[0],))
549
' bzr-%s python-%s %s\n' % (
550
bzrlib.version_string,
551
bzrlib._format_version_tuple(sys.version_info),
552
platform.platform(aliased=1),
554
self.stream.write('\n')
556
def report_test_start(self, test):
557
"""Display information on the test just about to be run"""
559
def _report_thread_leak(self, test, leaked_threads, active_threads):
560
"""Display information on a test that leaked one or more threads"""
561
# GZ 2010-09-09: A leak summary reported separately from the general
562
# thread debugging would be nice. Tests under subunit
563
# need something not using stream, perhaps adding a
564
# testtools details object would be fitting.
565
if 'threads' in selftest_debug_flags:
566
self.stream.write('%s is leaking, active is now %d\n' %
567
(test.id(), len(active_threads)))
569
def startTestRun(self):
570
self.startTime = time.time()
572
def report_success(self, test):
575
def wasStrictlySuccessful(self):
576
if self.unsupported or self.known_failure_count:
578
return self.wasSuccessful()
581
class TextTestResult(ExtendedTestResult):
582
"""Displays progress and results of tests in text form"""
584
def __init__(self, stream, descriptions, verbosity,
589
ExtendedTestResult.__init__(self, stream, descriptions, verbosity,
590
bench_history, strict)
591
# We no longer pass them around, but just rely on the UIFactory stack
594
warnings.warn("Passing pb to TextTestResult is deprecated")
595
self.pb = self.ui.nested_progress_bar()
596
self.pb.show_pct = False
597
self.pb.show_spinner = False
598
self.pb.show_eta = False,
599
self.pb.show_count = False
600
self.pb.show_bar = False
601
self.pb.update_latency = 0
602
self.pb.show_transport_activity = False
604
def stopTestRun(self):
605
# called when the tests that are going to run have run
608
super(TextTestResult, self).stopTestRun()
610
def report_tests_starting(self):
611
super(TextTestResult, self).report_tests_starting()
612
self.pb.update('[test 0/%d] Starting' % (self.num_tests))
614
def _progress_prefix_text(self):
615
# the longer this text, the less space we have to show the test
617
a = '[%d' % self.count # total that have been run
618
# tests skipped as known not to be relevant are not important enough
620
## if self.skip_count:
621
## a += ', %d skip' % self.skip_count
622
## if self.known_failure_count:
623
## a += '+%dX' % self.known_failure_count
625
a +='/%d' % self.num_tests
627
runtime = time.time() - self._overall_start_time
629
a += '%dm%ds' % (runtime / 60, runtime % 60)
632
total_fail_count = self.error_count + self.failure_count
634
a += ', %d failed' % total_fail_count
635
# if self.unsupported:
636
# a += ', %d missing' % len(self.unsupported)
640
def report_test_start(self, test):
642
self._progress_prefix_text()
644
+ self._shortened_test_description(test))
646
def _test_description(self, test):
647
return self._shortened_test_description(test)
649
def report_error(self, test, err):
650
self.stream.write('ERROR: %s\n %s\n' % (
651
self._test_description(test),
655
def report_failure(self, test, err):
656
self.stream.write('FAIL: %s\n %s\n' % (
657
self._test_description(test),
661
def report_known_failure(self, test, err):
664
def report_unexpected_success(self, test, reason):
665
self.stream.write('FAIL: %s\n %s: %s\n' % (
666
self._test_description(test),
667
"Unexpected success. Should have failed",
671
def report_skip(self, test, reason):
674
def report_not_applicable(self, test, reason):
677
def report_unsupported(self, test, feature):
678
"""test cannot be run because feature is missing."""
681
class VerboseTestResult(ExtendedTestResult):
682
"""Produce long output, with one line per test run plus times"""
684
def _ellipsize_to_right(self, a_string, final_width):
685
"""Truncate and pad a string, keeping the right hand side"""
686
if len(a_string) > final_width:
687
result = '...' + a_string[3-final_width:]
690
return result.ljust(final_width)
692
def report_tests_starting(self):
693
self.stream.write('running %d tests...\n' % self.num_tests)
694
super(VerboseTestResult, self).report_tests_starting()
696
def report_test_start(self, test):
697
name = self._shortened_test_description(test)
698
width = osutils.terminal_width()
699
if width is not None:
700
# width needs space for 6 char status, plus 1 for slash, plus an
701
# 11-char time string, plus a trailing blank
702
# when NUMBERED_DIRS: plus 5 chars on test number, plus 1 char on
704
self.stream.write(self._ellipsize_to_right(name, width-18))
706
self.stream.write(name)
709
def _error_summary(self, err):
711
return '%s%s' % (indent, err[1])
713
def report_error(self, test, err):
714
self.stream.write('ERROR %s\n%s\n'
715
% (self._testTimeString(test),
716
self._error_summary(err)))
718
def report_failure(self, test, err):
719
self.stream.write(' FAIL %s\n%s\n'
720
% (self._testTimeString(test),
721
self._error_summary(err)))
723
def report_known_failure(self, test, err):
724
self.stream.write('XFAIL %s\n%s\n'
725
% (self._testTimeString(test),
726
self._error_summary(err)))
728
def report_unexpected_success(self, test, reason):
729
self.stream.write(' FAIL %s\n%s: %s\n'
730
% (self._testTimeString(test),
731
"Unexpected success. Should have failed",
734
def report_success(self, test):
735
self.stream.write(' OK %s\n' % self._testTimeString(test))
736
for bench_called, stats in getattr(test, '_benchcalls', []):
737
self.stream.write('LSProf output for %s(%s, %s)\n' % bench_called)
738
stats.pprint(file=self.stream)
739
# flush the stream so that we get smooth output. This verbose mode is
740
# used to show the output in PQM.
743
def report_skip(self, test, reason):
744
self.stream.write(' SKIP %s\n%s\n'
745
% (self._testTimeString(test), reason))
747
def report_not_applicable(self, test, reason):
748
self.stream.write(' N/A %s\n %s\n'
749
% (self._testTimeString(test), reason))
751
def report_unsupported(self, test, feature):
752
"""test cannot be run because feature is missing."""
753
self.stream.write("NODEP %s\n The feature '%s' is not available.\n"
754
%(self._testTimeString(test), feature))
757
class TextTestRunner(object):
758
stop_on_failure = False
766
result_decorators=None,
768
"""Create a TextTestRunner.
770
:param result_decorators: An optional list of decorators to apply
771
to the result object being used by the runner. Decorators are
772
applied left to right - the first element in the list is the
775
# stream may know claim to know to write unicode strings, but in older
776
# pythons this goes sufficiently wrong that it is a bad idea. (
777
# specifically a built in file with encoding 'UTF-8' will still try
778
# to encode using ascii.
779
new_encoding = osutils.get_terminal_encoding()
780
codec = codecs.lookup(new_encoding)
781
if type(codec) is tuple:
785
encode = codec.encode
786
# GZ 2010-09-08: Really we don't want to be writing arbitrary bytes,
787
# so should swap to the plain codecs.StreamWriter
788
stream = osutils.UnicodeOrBytesToBytesWriter(encode, stream,
790
stream.encoding = new_encoding
792
self.descriptions = descriptions
793
self.verbosity = verbosity
794
self._bench_history = bench_history
795
self._strict = strict
796
self._result_decorators = result_decorators or []
799
"Run the given test case or test suite."
800
if self.verbosity == 1:
801
result_class = TextTestResult
802
elif self.verbosity >= 2:
803
result_class = VerboseTestResult
804
original_result = result_class(self.stream,
807
bench_history=self._bench_history,
810
# Signal to result objects that look at stop early policy to stop,
811
original_result.stop_early = self.stop_on_failure
812
result = original_result
813
for decorator in self._result_decorators:
814
result = decorator(result)
815
result.stop_early = self.stop_on_failure
816
result.startTestRun()
821
# higher level code uses our extended protocol to determine
822
# what exit code to give.
823
return original_result
826
def iter_suite_tests(suite):
827
"""Return all tests in a suite, recursing through nested suites"""
828
if isinstance(suite, unittest.TestCase):
830
elif isinstance(suite, unittest.TestSuite):
832
for r in iter_suite_tests(item):
835
raise Exception('unknown type %r for object %r'
836
% (type(suite), suite))
839
TestSkipped = testtools.testcase.TestSkipped
842
class TestNotApplicable(TestSkipped):
843
"""A test is not applicable to the situation where it was run.
845
This is only normally raised by parameterized tests, if they find that
846
the instance they're constructed upon does not support one aspect
851
# traceback._some_str fails to format exceptions that have the default
852
# __str__ which does an implicit ascii conversion. However, repr() on those
853
# objects works, for all that its not quite what the doctor may have ordered.
854
def _clever_some_str(value):
859
return repr(value).replace('\\n', '\n')
861
return '<unprintable %s object>' % type(value).__name__
863
traceback._some_str = _clever_some_str
866
# deprecated - use self.knownFailure(), or self.expectFailure.
867
KnownFailure = testtools.testcase._ExpectedFailure
870
class UnavailableFeature(Exception):
871
"""A feature required for this test was not available.
873
This can be considered a specialised form of SkippedTest.
875
The feature should be used to construct the exception.
879
class StringIOWrapper(object):
880
"""A wrapper around cStringIO which just adds an encoding attribute.
882
Internally we can check sys.stdout to see what the output encoding
883
should be. However, cStringIO has no encoding attribute that we can
884
set. So we wrap it instead.
889
def __init__(self, s=None):
891
self.__dict__['_cstring'] = StringIO(s)
893
self.__dict__['_cstring'] = StringIO()
895
def __getattr__(self, name, getattr=getattr):
896
return getattr(self.__dict__['_cstring'], name)
898
def __setattr__(self, name, val):
899
if name == 'encoding':
900
self.__dict__['encoding'] = val
902
return setattr(self._cstring, name, val)
905
class TestUIFactory(TextUIFactory):
906
"""A UI Factory for testing.
908
Hide the progress bar but emit note()s.
910
Allows get_password to be tested without real tty attached.
912
See also CannedInputUIFactory which lets you provide programmatic input in
915
# TODO: Capture progress events at the model level and allow them to be
916
# observed by tests that care.
918
# XXX: Should probably unify more with CannedInputUIFactory or a
919
# particular configuration of TextUIFactory, or otherwise have a clearer
920
# idea of how they're supposed to be different.
921
# See https://bugs.launchpad.net/bzr/+bug/408213
923
def __init__(self, stdout=None, stderr=None, stdin=None):
924
if stdin is not None:
925
# We use a StringIOWrapper to be able to test various
926
# encodings, but the user is still responsible to
927
# encode the string and to set the encoding attribute
928
# of StringIOWrapper.
929
stdin = StringIOWrapper(stdin)
930
super(TestUIFactory, self).__init__(stdin, stdout, stderr)
932
def get_non_echoed_password(self):
933
"""Get password from stdin without trying to handle the echo mode"""
934
password = self.stdin.readline()
937
if password[-1] == '\n':
938
password = password[:-1]
941
def make_progress_view(self):
942
return NullProgressView()
945
def isolated_doctest_setUp(test):
946
override_os_environ(test)
949
def isolated_doctest_tearDown(test):
950
restore_os_environ(test)
953
def IsolatedDocTestSuite(*args, **kwargs):
954
"""Overrides doctest.DocTestSuite to handle isolation.
956
The method is really a factory and users are expected to use it as such.
959
kwargs['setUp'] = isolated_doctest_setUp
960
kwargs['tearDown'] = isolated_doctest_tearDown
961
return doctest.DocTestSuite(*args, **kwargs)
964
class TestCase(testtools.TestCase):
965
"""Base class for bzr unit tests.
967
Tests that need access to disk resources should subclass
968
TestCaseInTempDir not TestCase.
970
Error and debug log messages are redirected from their usual
971
location into a temporary file, the contents of which can be
972
retrieved by _get_log(). We use a real OS file, not an in-memory object,
973
so that it can also capture file IO. When the test completes this file
974
is read into memory and removed from disk.
976
There are also convenience functions to invoke bzr's command-line
977
routine, and to build and check bzr trees.
979
In addition to the usual method of overriding tearDown(), this class also
980
allows subclasses to register cleanup functions via addCleanup, which are
981
run in order as the object is torn down. It's less likely this will be
982
accidentally overlooked.
986
# record lsprof data when performing benchmark calls.
987
_gather_lsprof_in_benchmarks = False
989
def __init__(self, methodName='testMethod'):
990
super(TestCase, self).__init__(methodName)
991
self._directory_isolation = True
992
self.exception_handlers.insert(0,
993
(UnavailableFeature, self._do_unsupported_or_skip))
994
self.exception_handlers.insert(0,
995
(TestNotApplicable, self._do_not_applicable))
998
super(TestCase, self).setUp()
1000
timeout = config.GlobalStack().get('selftest.timeout')
1002
timeout_fixture = fixtures.TimeoutFixture(timeout)
1003
timeout_fixture.setUp()
1004
self.addCleanup(timeout_fixture.cleanUp)
1006
for feature in getattr(self, '_test_needs_features', []):
1007
self.requireFeature(feature)
1008
self._cleanEnvironment()
1010
if bzrlib.global_state is not None:
1011
self.overrideAttr(bzrlib.global_state, 'cmdline_overrides',
1012
config.CommandLineStore())
1015
self._startLogFile()
1016
self._benchcalls = []
1017
self._benchtime = None
1019
self._track_transports()
1021
self._clear_debug_flags()
1022
# Isolate global verbosity level, to make sure it's reproducible
1023
# between tests. We should get rid of this altogether: bug 656694. --
1025
self.overrideAttr(bzrlib.trace, '_verbosity_level', 0)
1026
# Isolate config option expansion until its default value for bzrlib is
1027
# settled on or a the FIXME associated with _get_expand_default_value
1028
# is addressed -- vila 20110219
1029
self.overrideAttr(config, '_expand_default_value', None)
1030
self._log_files = set()
1031
# Each key in the ``_counters`` dict holds a value for a different
1032
# counter. When the test ends, addDetail() should be used to output the
1033
# counter values. This happens in install_counter_hook().
1035
if 'config_stats' in selftest_debug_flags:
1036
self._install_config_stats_hooks()
1037
# Do not use i18n for tests (unless the test reverses this)
1043
# The sys preserved stdin/stdout should allow blackbox tests debugging
1044
pdb.Pdb(stdin=sys.__stdin__, stdout=sys.__stdout__
1045
).set_trace(sys._getframe().f_back)
1047
def discardDetail(self, name):
1048
"""Extend the addDetail, getDetails api so we can remove a detail.
1050
eg. bzr always adds the 'log' detail at startup, but we don't want to
1051
include it for skipped, xfail, etc tests.
1053
It is safe to call this for a detail that doesn't exist, in case this
1054
gets called multiple times.
1056
# We cheat. details is stored in __details which means we shouldn't
1057
# touch it. but getDetails() returns the dict directly, so we can
1059
details = self.getDetails()
1063
def install_counter_hook(self, hooks, name, counter_name=None):
1064
"""Install a counting hook.
1066
Any hook can be counted as long as it doesn't need to return a value.
1068
:param hooks: Where the hook should be installed.
1070
:param name: The hook name that will be counted.
1072
:param counter_name: The counter identifier in ``_counters``, defaults
1075
_counters = self._counters # Avoid closing over self
1076
if counter_name is None:
1078
if _counters.has_key(counter_name):
1079
raise AssertionError('%s is already used as a counter name'
1081
_counters[counter_name] = 0
1082
self.addDetail(counter_name, content.Content(content.UTF8_TEXT,
1083
lambda: ['%d' % (_counters[counter_name],)]))
1084
def increment_counter(*args, **kwargs):
1085
_counters[counter_name] += 1
1086
label = 'count %s calls' % (counter_name,)
1087
hooks.install_named_hook(name, increment_counter, label)
1088
self.addCleanup(hooks.uninstall_named_hook, name, label)
1090
def _install_config_stats_hooks(self):
1091
"""Install config hooks to count hook calls.
1094
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1095
self.install_counter_hook(config.ConfigHooks, hook_name,
1096
'config.%s' % (hook_name,))
1098
# The OldConfigHooks are private and need special handling to protect
1099
# against recursive tests (tests that run other tests), so we just do
1100
# manually what registering them into _builtin_known_hooks will provide
1102
self.overrideAttr(config, 'OldConfigHooks', config._OldConfigHooks())
1103
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1104
self.install_counter_hook(config.OldConfigHooks, hook_name,
1105
'old_config.%s' % (hook_name,))
1107
def _clear_debug_flags(self):
1108
"""Prevent externally set debug flags affecting tests.
1110
Tests that want to use debug flags can just set them in the
1111
debug_flags set during setup/teardown.
1113
# Start with a copy of the current debug flags we can safely modify.
1114
self.overrideAttr(debug, 'debug_flags', set(debug.debug_flags))
1115
if 'allow_debug' not in selftest_debug_flags:
1116
debug.debug_flags.clear()
1117
if 'disable_lock_checks' not in selftest_debug_flags:
1118
debug.debug_flags.add('strict_locks')
1120
def _clear_hooks(self):
1121
# prevent hooks affecting tests
1122
known_hooks = hooks.known_hooks
1123
self._preserved_hooks = {}
1124
for key, (parent, name) in known_hooks.iter_parent_objects():
1125
current_hooks = getattr(parent, name)
1126
self._preserved_hooks[parent] = (name, current_hooks)
1127
self._preserved_lazy_hooks = hooks._lazy_hooks
1128
hooks._lazy_hooks = {}
1129
self.addCleanup(self._restoreHooks)
1130
for key, (parent, name) in known_hooks.iter_parent_objects():
1131
factory = known_hooks.get(key)
1132
setattr(parent, name, factory())
1133
# this hook should always be installed
1134
request._install_hook()
1136
def disable_directory_isolation(self):
1137
"""Turn off directory isolation checks."""
1138
self._directory_isolation = False
1140
def enable_directory_isolation(self):
1141
"""Enable directory isolation checks."""
1142
self._directory_isolation = True
1144
def _silenceUI(self):
1145
"""Turn off UI for duration of test"""
1146
# by default the UI is off; tests can turn it on if they want it.
1147
self.overrideAttr(ui, 'ui_factory', ui.SilentUIFactory())
1149
def _check_locks(self):
1150
"""Check that all lock take/release actions have been paired."""
1151
# We always check for mismatched locks. If a mismatch is found, we
1152
# fail unless -Edisable_lock_checks is supplied to selftest, in which
1153
# case we just print a warning.
1155
acquired_locks = [lock for action, lock in self._lock_actions
1156
if action == 'acquired']
1157
released_locks = [lock for action, lock in self._lock_actions
1158
if action == 'released']
1159
broken_locks = [lock for action, lock in self._lock_actions
1160
if action == 'broken']
1161
# trivially, given the tests for lock acquistion and release, if we
1162
# have as many in each list, it should be ok. Some lock tests also
1163
# break some locks on purpose and should be taken into account by
1164
# considering that breaking a lock is just a dirty way of releasing it.
1165
if len(acquired_locks) != (len(released_locks) + len(broken_locks)):
1167
'Different number of acquired and '
1168
'released or broken locks.\n'
1172
(acquired_locks, released_locks, broken_locks))
1173
if not self._lock_check_thorough:
1174
# Rather than fail, just warn
1175
print "Broken test %s: %s" % (self, message)
1179
def _track_locks(self):
1180
"""Track lock activity during tests."""
1181
self._lock_actions = []
1182
if 'disable_lock_checks' in selftest_debug_flags:
1183
self._lock_check_thorough = False
1185
self._lock_check_thorough = True
1187
self.addCleanup(self._check_locks)
1188
_mod_lock.Lock.hooks.install_named_hook('lock_acquired',
1189
self._lock_acquired, None)
1190
_mod_lock.Lock.hooks.install_named_hook('lock_released',
1191
self._lock_released, None)
1192
_mod_lock.Lock.hooks.install_named_hook('lock_broken',
1193
self._lock_broken, None)
1195
def _lock_acquired(self, result):
1196
self._lock_actions.append(('acquired', result))
1198
def _lock_released(self, result):
1199
self._lock_actions.append(('released', result))
1201
def _lock_broken(self, result):
1202
self._lock_actions.append(('broken', result))
1204
def permit_dir(self, name):
1205
"""Permit a directory to be used by this test. See permit_url."""
1206
name_transport = _mod_transport.get_transport_from_path(name)
1207
self.permit_url(name)
1208
self.permit_url(name_transport.base)
1210
def permit_url(self, url):
1211
"""Declare that url is an ok url to use in this test.
1213
Do this for memory transports, temporary test directory etc.
1215
Do not do this for the current working directory, /tmp, or any other
1216
preexisting non isolated url.
1218
if not url.endswith('/'):
1220
self._bzr_selftest_roots.append(url)
1222
def permit_source_tree_branch_repo(self):
1223
"""Permit the source tree bzr is running from to be opened.
1225
Some code such as bzrlib.version attempts to read from the bzr branch
1226
that bzr is executing from (if any). This method permits that directory
1227
to be used in the test suite.
1229
path = self.get_source_path()
1230
self.record_directory_isolation()
1233
workingtree.WorkingTree.open(path)
1234
except (errors.NotBranchError, errors.NoWorkingTree):
1235
raise TestSkipped('Needs a working tree of bzr sources')
1237
self.enable_directory_isolation()
1239
def _preopen_isolate_transport(self, transport):
1240
"""Check that all transport openings are done in the test work area."""
1241
while isinstance(transport, pathfilter.PathFilteringTransport):
1242
# Unwrap pathfiltered transports
1243
transport = transport.server.backing_transport.clone(
1244
transport._filter('.'))
1245
url = transport.base
1246
# ReadonlySmartTCPServer_for_testing decorates the backing transport
1247
# urls it is given by prepending readonly+. This is appropriate as the
1248
# client shouldn't know that the server is readonly (or not readonly).
1249
# We could register all servers twice, with readonly+ prepending, but
1250
# that makes for a long list; this is about the same but easier to
1252
if url.startswith('readonly+'):
1253
url = url[len('readonly+'):]
1254
self._preopen_isolate_url(url)
1256
def _preopen_isolate_url(self, url):
1257
if not self._directory_isolation:
1259
if self._directory_isolation == 'record':
1260
self._bzr_selftest_roots.append(url)
1262
# This prevents all transports, including e.g. sftp ones backed on disk
1263
# from working unless they are explicitly granted permission. We then
1264
# depend on the code that sets up test transports to check that they are
1265
# appropriately isolated and enable their use by calling
1266
# self.permit_transport()
1267
if not osutils.is_inside_any(self._bzr_selftest_roots, url):
1268
raise errors.BzrError("Attempt to escape test isolation: %r %r"
1269
% (url, self._bzr_selftest_roots))
1271
def record_directory_isolation(self):
1272
"""Gather accessed directories to permit later access.
1274
This is used for tests that access the branch bzr is running from.
1276
self._directory_isolation = "record"
1278
def start_server(self, transport_server, backing_server=None):
1279
"""Start transport_server for this test.
1281
This starts the server, registers a cleanup for it and permits the
1282
server's urls to be used.
1284
if backing_server is None:
1285
transport_server.start_server()
1287
transport_server.start_server(backing_server)
1288
self.addCleanup(transport_server.stop_server)
1289
# Obtain a real transport because if the server supplies a password, it
1290
# will be hidden from the base on the client side.
1291
t = _mod_transport.get_transport_from_url(transport_server.get_url())
1292
# Some transport servers effectively chroot the backing transport;
1293
# others like SFTPServer don't - users of the transport can walk up the
1294
# transport to read the entire backing transport. This wouldn't matter
1295
# except that the workdir tests are given - and that they expect the
1296
# server's url to point at - is one directory under the safety net. So
1297
# Branch operations into the transport will attempt to walk up one
1298
# directory. Chrooting all servers would avoid this but also mean that
1299
# we wouldn't be testing directly against non-root urls. Alternatively
1300
# getting the test framework to start the server with a backing server
1301
# at the actual safety net directory would work too, but this then
1302
# means that the self.get_url/self.get_transport methods would need
1303
# to transform all their results. On balance its cleaner to handle it
1304
# here, and permit a higher url when we have one of these transports.
1305
if t.base.endswith('/work/'):
1306
# we have safety net/test root/work
1307
t = t.clone('../..')
1308
elif isinstance(transport_server,
1309
test_server.SmartTCPServer_for_testing):
1310
# The smart server adds a path similar to work, which is traversed
1311
# up from by the client. But the server is chrooted - the actual
1312
# backing transport is not escaped from, and VFS requests to the
1313
# root will error (because they try to escape the chroot).
1315
while t2.base != t.base:
1318
self.permit_url(t.base)
1320
def _track_transports(self):
1321
"""Install checks for transport usage."""
1322
# TestCase has no safe place it can write to.
1323
self._bzr_selftest_roots = []
1324
# Currently the easiest way to be sure that nothing is going on is to
1325
# hook into bzr dir opening. This leaves a small window of error for
1326
# transport tests, but they are well known, and we can improve on this
1328
bzrdir.BzrDir.hooks.install_named_hook("pre_open",
1329
self._preopen_isolate_transport, "Check bzr directories are safe.")
1331
def _ndiff_strings(self, a, b):
1332
"""Return ndiff between two strings containing lines.
1334
A trailing newline is added if missing to make the strings
1336
if b and b[-1] != '\n':
1338
if a and a[-1] != '\n':
1340
difflines = difflib.ndiff(a.splitlines(True),
1342
linejunk=lambda x: False,
1343
charjunk=lambda x: False)
1344
return ''.join(difflines)
1346
def assertEqual(self, a, b, message=''):
1350
except UnicodeError, e:
1351
# If we can't compare without getting a UnicodeError, then
1352
# obviously they are different
1353
trace.mutter('UnicodeError: %s', e)
1356
raise AssertionError("%snot equal:\na = %s\nb = %s\n"
1358
pprint.pformat(a), pprint.pformat(b)))
1360
assertEquals = assertEqual
1362
def assertEqualDiff(self, a, b, message=None):
1363
"""Assert two texts are equal, if not raise an exception.
1365
This is intended for use with multi-line strings where it can
1366
be hard to find the differences by eye.
1368
# TODO: perhaps override assertEquals to call this for strings?
1372
message = "texts not equal:\n"
1374
message = 'first string is missing a final newline.\n'
1376
message = 'second string is missing a final newline.\n'
1377
raise AssertionError(message +
1378
self._ndiff_strings(a, b))
1380
def assertEqualMode(self, mode, mode_test):
1381
self.assertEqual(mode, mode_test,
1382
'mode mismatch %o != %o' % (mode, mode_test))
1384
def assertEqualStat(self, expected, actual):
1385
"""assert that expected and actual are the same stat result.
1387
:param expected: A stat result.
1388
:param actual: A stat result.
1389
:raises AssertionError: If the expected and actual stat values differ
1390
other than by atime.
1392
self.assertEqual(expected.st_size, actual.st_size,
1393
'st_size did not match')
1394
self.assertEqual(expected.st_mtime, actual.st_mtime,
1395
'st_mtime did not match')
1396
self.assertEqual(expected.st_ctime, actual.st_ctime,
1397
'st_ctime did not match')
1398
if sys.platform == 'win32':
1399
# On Win32 both 'dev' and 'ino' cannot be trusted. In python2.4 it
1400
# is 'dev' that varies, in python 2.5 (6?) it is st_ino that is
1401
# odd. We just force it to always be 0 to avoid any problems.
1402
self.assertEqual(0, expected.st_dev)
1403
self.assertEqual(0, actual.st_dev)
1404
self.assertEqual(0, expected.st_ino)
1405
self.assertEqual(0, actual.st_ino)
1407
self.assertEqual(expected.st_dev, actual.st_dev,
1408
'st_dev did not match')
1409
self.assertEqual(expected.st_ino, actual.st_ino,
1410
'st_ino did not match')
1411
self.assertEqual(expected.st_mode, actual.st_mode,
1412
'st_mode did not match')
1414
def assertLength(self, length, obj_with_len):
1415
"""Assert that obj_with_len is of length length."""
1416
if len(obj_with_len) != length:
1417
self.fail("Incorrect length: wanted %d, got %d for %r" % (
1418
length, len(obj_with_len), obj_with_len))
1420
def assertLogsError(self, exception_class, func, *args, **kwargs):
1421
"""Assert that `func(*args, **kwargs)` quietly logs a specific error.
1424
orig_log_exception_quietly = trace.log_exception_quietly
1427
orig_log_exception_quietly()
1428
captured.append(sys.exc_info()[1])
1429
trace.log_exception_quietly = capture
1430
func(*args, **kwargs)
1432
trace.log_exception_quietly = orig_log_exception_quietly
1433
self.assertLength(1, captured)
1435
self.assertIsInstance(err, exception_class)
1438
def assertPositive(self, val):
1439
"""Assert that val is greater than 0."""
1440
self.assertTrue(val > 0, 'expected a positive value, but got %s' % val)
1442
def assertNegative(self, val):
1443
"""Assert that val is less than 0."""
1444
self.assertTrue(val < 0, 'expected a negative value, but got %s' % val)
1446
def assertStartsWith(self, s, prefix):
1447
if not s.startswith(prefix):
1448
raise AssertionError('string %r does not start with %r' % (s, prefix))
1450
def assertEndsWith(self, s, suffix):
1451
"""Asserts that s ends with suffix."""
1452
if not s.endswith(suffix):
1453
raise AssertionError('string %r does not end with %r' % (s, suffix))
1455
def assertContainsRe(self, haystack, needle_re, flags=0):
1456
"""Assert that a contains something matching a regular expression."""
1457
if not re.search(needle_re, haystack, flags):
1458
if '\n' in haystack or len(haystack) > 60:
1459
# a long string, format it in a more readable way
1460
raise AssertionError(
1461
'pattern "%s" not found in\n"""\\\n%s"""\n'
1462
% (needle_re, haystack))
1464
raise AssertionError('pattern "%s" not found in "%s"'
1465
% (needle_re, haystack))
1467
def assertNotContainsRe(self, haystack, needle_re, flags=0):
1468
"""Assert that a does not match a regular expression"""
1469
if re.search(needle_re, haystack, flags):
1470
raise AssertionError('pattern "%s" found in "%s"'
1471
% (needle_re, haystack))
1473
def assertContainsString(self, haystack, needle):
1474
if haystack.find(needle) == -1:
1475
self.fail("string %r not found in '''%s'''" % (needle, haystack))
1477
def assertNotContainsString(self, haystack, needle):
1478
if haystack.find(needle) != -1:
1479
self.fail("string %r found in '''%s'''" % (needle, haystack))
1481
def assertSubset(self, sublist, superlist):
1482
"""Assert that every entry in sublist is present in superlist."""
1483
missing = set(sublist) - set(superlist)
1484
if len(missing) > 0:
1485
raise AssertionError("value(s) %r not present in container %r" %
1486
(missing, superlist))
1488
def assertListRaises(self, excClass, func, *args, **kwargs):
1489
"""Fail unless excClass is raised when the iterator from func is used.
1491
Many functions can return generators this makes sure
1492
to wrap them in a list() call to make sure the whole generator
1493
is run, and that the proper exception is raised.
1496
list(func(*args, **kwargs))
1500
if getattr(excClass,'__name__', None) is not None:
1501
excName = excClass.__name__
1503
excName = str(excClass)
1504
raise self.failureException, "%s not raised" % excName
1506
def assertRaises(self, excClass, callableObj, *args, **kwargs):
1507
"""Assert that a callable raises a particular exception.
1509
:param excClass: As for the except statement, this may be either an
1510
exception class, or a tuple of classes.
1511
:param callableObj: A callable, will be passed ``*args`` and
1514
Returns the exception so that you can examine it.
1517
callableObj(*args, **kwargs)
1521
if getattr(excClass,'__name__', None) is not None:
1522
excName = excClass.__name__
1525
excName = str(excClass)
1526
raise self.failureException, "%s not raised" % excName
1528
def assertIs(self, left, right, message=None):
1529
if not (left is right):
1530
if message is not None:
1531
raise AssertionError(message)
1533
raise AssertionError("%r is not %r." % (left, right))
1535
def assertIsNot(self, left, right, message=None):
1537
if message is not None:
1538
raise AssertionError(message)
1540
raise AssertionError("%r is %r." % (left, right))
1542
def assertTransportMode(self, transport, path, mode):
1543
"""Fail if a path does not have mode "mode".
1545
If modes are not supported on this transport, the assertion is ignored.
1547
if not transport._can_roundtrip_unix_modebits():
1549
path_stat = transport.stat(path)
1550
actual_mode = stat.S_IMODE(path_stat.st_mode)
1551
self.assertEqual(mode, actual_mode,
1552
'mode of %r incorrect (%s != %s)'
1553
% (path, oct(mode), oct(actual_mode)))
1555
def assertIsSameRealPath(self, path1, path2):
1556
"""Fail if path1 and path2 points to different files"""
1557
self.assertEqual(osutils.realpath(path1),
1558
osutils.realpath(path2),
1559
"apparent paths:\na = %s\nb = %s\n," % (path1, path2))
1561
def assertIsInstance(self, obj, kls, msg=None):
1562
"""Fail if obj is not an instance of kls
1564
:param msg: Supplementary message to show if the assertion fails.
1566
if not isinstance(obj, kls):
1567
m = "%r is an instance of %s rather than %s" % (
1568
obj, obj.__class__, kls)
1573
def assertFileEqual(self, content, path):
1574
"""Fail if path does not contain 'content'."""
1575
self.assertPathExists(path)
1576
f = file(path, 'rb')
1581
self.assertEqualDiff(content, s)
1583
def assertDocstring(self, expected_docstring, obj):
1584
"""Fail if obj does not have expected_docstring"""
1586
# With -OO the docstring should be None instead
1587
self.assertIs(obj.__doc__, None)
1589
self.assertEqual(expected_docstring, obj.__doc__)
1591
@symbol_versioning.deprecated_method(symbol_versioning.deprecated_in((2, 4)))
1592
def failUnlessExists(self, path):
1593
return self.assertPathExists(path)
1595
def assertPathExists(self, path):
1596
"""Fail unless path or paths, which may be abs or relative, exist."""
1597
if not isinstance(path, basestring):
1599
self.assertPathExists(p)
1601
self.assertTrue(osutils.lexists(path),
1602
path + " does not exist")
1604
@symbol_versioning.deprecated_method(symbol_versioning.deprecated_in((2, 4)))
1605
def failIfExists(self, path):
1606
return self.assertPathDoesNotExist(path)
1608
def assertPathDoesNotExist(self, path):
1609
"""Fail if path or paths, which may be abs or relative, exist."""
1610
if not isinstance(path, basestring):
1612
self.assertPathDoesNotExist(p)
1614
self.assertFalse(osutils.lexists(path),
1617
def _capture_deprecation_warnings(self, a_callable, *args, **kwargs):
1618
"""A helper for callDeprecated and applyDeprecated.
1620
:param a_callable: A callable to call.
1621
:param args: The positional arguments for the callable
1622
:param kwargs: The keyword arguments for the callable
1623
:return: A tuple (warnings, result). result is the result of calling
1624
a_callable(``*args``, ``**kwargs``).
1627
def capture_warnings(msg, cls=None, stacklevel=None):
1628
# we've hooked into a deprecation specific callpath,
1629
# only deprecations should getting sent via it.
1630
self.assertEqual(cls, DeprecationWarning)
1631
local_warnings.append(msg)
1632
original_warning_method = symbol_versioning.warn
1633
symbol_versioning.set_warning_method(capture_warnings)
1635
result = a_callable(*args, **kwargs)
1637
symbol_versioning.set_warning_method(original_warning_method)
1638
return (local_warnings, result)
1640
def applyDeprecated(self, deprecation_format, a_callable, *args, **kwargs):
1641
"""Call a deprecated callable without warning the user.
1643
Note that this only captures warnings raised by symbol_versioning.warn,
1644
not other callers that go direct to the warning module.
1646
To test that a deprecated method raises an error, do something like
1647
this (remember that both assertRaises and applyDeprecated delays *args
1648
and **kwargs passing)::
1650
self.assertRaises(errors.ReservedId,
1651
self.applyDeprecated,
1652
deprecated_in((1, 5, 0)),
1656
:param deprecation_format: The deprecation format that the callable
1657
should have been deprecated with. This is the same type as the
1658
parameter to deprecated_method/deprecated_function. If the
1659
callable is not deprecated with this format, an assertion error
1661
:param a_callable: A callable to call. This may be a bound method or
1662
a regular function. It will be called with ``*args`` and
1664
:param args: The positional arguments for the callable
1665
:param kwargs: The keyword arguments for the callable
1666
:return: The result of a_callable(``*args``, ``**kwargs``)
1668
call_warnings, result = self._capture_deprecation_warnings(a_callable,
1670
expected_first_warning = symbol_versioning.deprecation_string(
1671
a_callable, deprecation_format)
1672
if len(call_warnings) == 0:
1673
self.fail("No deprecation warning generated by call to %s" %
1675
self.assertEqual(expected_first_warning, call_warnings[0])
1678
def callCatchWarnings(self, fn, *args, **kw):
1679
"""Call a callable that raises python warnings.
1681
The caller's responsible for examining the returned warnings.
1683
If the callable raises an exception, the exception is not
1684
caught and propagates up to the caller. In that case, the list
1685
of warnings is not available.
1687
:returns: ([warning_object, ...], fn_result)
1689
# XXX: This is not perfect, because it completely overrides the
1690
# warnings filters, and some code may depend on suppressing particular
1691
# warnings. It's the easiest way to insulate ourselves from -Werror,
1692
# though. -- Andrew, 20071062
1694
def _catcher(message, category, filename, lineno, file=None, line=None):
1695
# despite the name, 'message' is normally(?) a Warning subclass
1697
wlist.append(message)
1698
saved_showwarning = warnings.showwarning
1699
saved_filters = warnings.filters
1701
warnings.showwarning = _catcher
1702
warnings.filters = []
1703
result = fn(*args, **kw)
1705
warnings.showwarning = saved_showwarning
1706
warnings.filters = saved_filters
1707
return wlist, result
1709
def callDeprecated(self, expected, callable, *args, **kwargs):
1710
"""Assert that a callable is deprecated in a particular way.
1712
This is a very precise test for unusual requirements. The
1713
applyDeprecated helper function is probably more suited for most tests
1714
as it allows you to simply specify the deprecation format being used
1715
and will ensure that that is issued for the function being called.
1717
Note that this only captures warnings raised by symbol_versioning.warn,
1718
not other callers that go direct to the warning module. To catch
1719
general warnings, use callCatchWarnings.
1721
:param expected: a list of the deprecation warnings expected, in order
1722
:param callable: The callable to call
1723
:param args: The positional arguments for the callable
1724
:param kwargs: The keyword arguments for the callable
1726
call_warnings, result = self._capture_deprecation_warnings(callable,
1728
self.assertEqual(expected, call_warnings)
1731
def _startLogFile(self):
1732
"""Setup a in-memory target for bzr and testcase log messages"""
1733
pseudo_log_file = StringIO()
1734
def _get_log_contents_for_weird_testtools_api():
1735
return [pseudo_log_file.getvalue().decode(
1736
"utf-8", "replace").encode("utf-8")]
1737
self.addDetail("log", content.Content(content.ContentType("text",
1738
"plain", {"charset": "utf8"}),
1739
_get_log_contents_for_weird_testtools_api))
1740
self._log_file = pseudo_log_file
1741
self._log_memento = trace.push_log_file(self._log_file)
1742
self.addCleanup(self._finishLogFile)
1744
def _finishLogFile(self):
1745
"""Flush and dereference the in-memory log for this testcase"""
1746
if trace._trace_file:
1747
# flush the log file, to get all content
1748
trace._trace_file.flush()
1749
trace.pop_log_file(self._log_memento)
1750
# The logging module now tracks references for cleanup so discard ours
1751
del self._log_memento
1753
def thisFailsStrictLockCheck(self):
1754
"""It is known that this test would fail with -Dstrict_locks.
1756
By default, all tests are run with strict lock checking unless
1757
-Edisable_lock_checks is supplied. However there are some tests which
1758
we know fail strict locks at this point that have not been fixed.
1759
They should call this function to disable the strict checking.
1761
This should be used sparingly, it is much better to fix the locking
1762
issues rather than papering over the problem by calling this function.
1764
debug.debug_flags.discard('strict_locks')
1766
def overrideAttr(self, obj, attr_name, new=_unitialized_attr):
1767
"""Overrides an object attribute restoring it after the test.
1769
:note: This should be used with discretion; you should think about
1770
whether it's better to make the code testable without monkey-patching.
1772
:param obj: The object that will be mutated.
1774
:param attr_name: The attribute name we want to preserve/override in
1777
:param new: The optional value we want to set the attribute to.
1779
:returns: The actual attr value.
1781
value = getattr(obj, attr_name)
1782
# The actual value is captured by the call below
1783
self.addCleanup(setattr, obj, attr_name, value)
1784
if new is not _unitialized_attr:
1785
setattr(obj, attr_name, new)
1788
def overrideEnv(self, name, new):
1789
"""Set an environment variable, and reset it after the test.
1791
:param name: The environment variable name.
1793
:param new: The value to set the variable to. If None, the
1794
variable is deleted from the environment.
1796
:returns: The actual variable value.
1798
value = osutils.set_or_unset_env(name, new)
1799
self.addCleanup(osutils.set_or_unset_env, name, value)
1802
def recordCalls(self, obj, attr_name):
1803
"""Monkeypatch in a wrapper that will record calls.
1805
The monkeypatch is automatically removed when the test concludes.
1807
:param obj: The namespace holding the reference to be replaced;
1808
typically a module, class, or object.
1809
:param attr_name: A string for the name of the attribute to
1811
:returns: A list that will be extended with one item every time the
1812
function is called, with a tuple of (args, kwargs).
1816
def decorator(*args, **kwargs):
1817
calls.append((args, kwargs))
1818
return orig(*args, **kwargs)
1819
orig = self.overrideAttr(obj, attr_name, decorator)
1822
def _cleanEnvironment(self):
1823
for name, value in isolated_environ.iteritems():
1824
self.overrideEnv(name, value)
1826
def _restoreHooks(self):
1827
for klass, (name, hooks) in self._preserved_hooks.items():
1828
setattr(klass, name, hooks)
1829
self._preserved_hooks.clear()
1830
bzrlib.hooks._lazy_hooks = self._preserved_lazy_hooks
1831
self._preserved_lazy_hooks.clear()
1833
def knownFailure(self, reason):
1834
"""Declare that this test fails for a known reason
1836
Tests that are known to fail should generally be using expectedFailure
1837
with an appropriate reverse assertion if a change could cause the test
1838
to start passing. Conversely if the test has no immediate prospect of
1839
succeeding then using skip is more suitable.
1841
When this method is called while an exception is being handled, that
1842
traceback will be used, otherwise a new exception will be thrown to
1843
provide one but won't be reported.
1845
self._add_reason(reason)
1847
exc_info = sys.exc_info()
1848
if exc_info != (None, None, None):
1849
self._report_traceback(exc_info)
1852
raise self.failureException(reason)
1853
except self.failureException:
1854
exc_info = sys.exc_info()
1855
# GZ 02-08-2011: Maybe cleanup this err.exc_info attribute too?
1856
raise testtools.testcase._ExpectedFailure(exc_info)
1860
def _suppress_log(self):
1861
"""Remove the log info from details."""
1862
self.discardDetail('log')
1864
def _do_skip(self, result, reason):
1865
self._suppress_log()
1866
addSkip = getattr(result, 'addSkip', None)
1867
if not callable(addSkip):
1868
result.addSuccess(result)
1870
addSkip(self, reason)
1873
def _do_known_failure(self, result, e):
1874
self._suppress_log()
1875
err = sys.exc_info()
1876
addExpectedFailure = getattr(result, 'addExpectedFailure', None)
1877
if addExpectedFailure is not None:
1878
addExpectedFailure(self, err)
1880
result.addSuccess(self)
1883
def _do_not_applicable(self, result, e):
1885
reason = 'No reason given'
1888
self._suppress_log ()
1889
addNotApplicable = getattr(result, 'addNotApplicable', None)
1890
if addNotApplicable is not None:
1891
result.addNotApplicable(self, reason)
1893
self._do_skip(result, reason)
1896
def _report_skip(self, result, err):
1897
"""Override the default _report_skip.
1899
We want to strip the 'log' detail. If we waint until _do_skip, it has
1900
already been formatted into the 'reason' string, and we can't pull it
1903
self._suppress_log()
1904
super(TestCase, self)._report_skip(self, result, err)
1907
def _report_expected_failure(self, result, err):
1910
See _report_skip for motivation.
1912
self._suppress_log()
1913
super(TestCase, self)._report_expected_failure(self, result, err)
1916
def _do_unsupported_or_skip(self, result, e):
1918
self._suppress_log()
1919
addNotSupported = getattr(result, 'addNotSupported', None)
1920
if addNotSupported is not None:
1921
result.addNotSupported(self, reason)
1923
self._do_skip(result, reason)
1925
def time(self, callable, *args, **kwargs):
1926
"""Run callable and accrue the time it takes to the benchmark time.
1928
If lsprofiling is enabled (i.e. by --lsprof-time to bzr selftest) then
1929
this will cause lsprofile statistics to be gathered and stored in
1932
if self._benchtime is None:
1933
self.addDetail('benchtime', content.Content(content.ContentType(
1934
"text", "plain"), lambda:[str(self._benchtime)]))
1938
if not self._gather_lsprof_in_benchmarks:
1939
return callable(*args, **kwargs)
1941
# record this benchmark
1942
ret, stats = bzrlib.lsprof.profile(callable, *args, **kwargs)
1944
self._benchcalls.append(((callable, args, kwargs), stats))
1947
self._benchtime += time.time() - start
1949
def log(self, *args):
1953
"""Get a unicode string containing the log from bzrlib.trace.
1955
Undecodable characters are replaced.
1957
return u"".join(self.getDetails()['log'].iter_text())
1959
def requireFeature(self, feature):
1960
"""This test requires a specific feature is available.
1962
:raises UnavailableFeature: When feature is not available.
1964
if not feature.available():
1965
raise UnavailableFeature(feature)
1967
def _run_bzr_autosplit(self, args, retcode, encoding, stdin,
1969
"""Run bazaar command line, splitting up a string command line."""
1970
if isinstance(args, basestring):
1971
# shlex don't understand unicode strings,
1972
# so args should be plain string (bialix 20070906)
1973
args = list(shlex.split(str(args)))
1974
return self._run_bzr_core(args, retcode=retcode,
1975
encoding=encoding, stdin=stdin, working_dir=working_dir,
1978
def _run_bzr_core(self, args, retcode, encoding, stdin,
1980
# Clear chk_map page cache, because the contents are likely to mask
1982
chk_map.clear_cache()
1983
if encoding is None:
1984
encoding = osutils.get_user_encoding()
1985
stdout = StringIOWrapper()
1986
stderr = StringIOWrapper()
1987
stdout.encoding = encoding
1988
stderr.encoding = encoding
1990
self.log('run bzr: %r', args)
1991
# FIXME: don't call into logging here
1992
handler = logging.StreamHandler(stderr)
1993
handler.setLevel(logging.INFO)
1994
logger = logging.getLogger('')
1995
logger.addHandler(handler)
1996
old_ui_factory = ui.ui_factory
1997
ui.ui_factory = TestUIFactory(stdin=stdin, stdout=stdout, stderr=stderr)
2000
if working_dir is not None:
2001
cwd = osutils.getcwd()
2002
os.chdir(working_dir)
2006
result = self.apply_redirected(
2007
ui.ui_factory.stdin,
2009
_mod_commands.run_bzr_catch_user_errors,
2011
except KeyboardInterrupt:
2012
# Reraise KeyboardInterrupt with contents of redirected stdout
2013
# and stderr as arguments, for tests which are interested in
2014
# stdout and stderr and are expecting the exception.
2015
out = stdout.getvalue()
2016
err = stderr.getvalue()
2018
self.log('output:\n%r', out)
2020
self.log('errors:\n%r', err)
2021
raise KeyboardInterrupt(out, err)
2023
logger.removeHandler(handler)
2024
ui.ui_factory = old_ui_factory
2028
out = stdout.getvalue()
2029
err = stderr.getvalue()
2031
self.log('output:\n%r', out)
2033
self.log('errors:\n%r', err)
2034
if retcode is not None:
2035
self.assertEquals(retcode, result,
2036
message='Unexpected return code')
2037
return result, out, err
2039
def run_bzr(self, args, retcode=0, encoding=None, stdin=None,
2040
working_dir=None, error_regexes=[], output_encoding=None):
2041
"""Invoke bzr, as if it were run from the command line.
2043
The argument list should not include the bzr program name - the
2044
first argument is normally the bzr command. Arguments may be
2045
passed in three ways:
2047
1- A list of strings, eg ["commit", "a"]. This is recommended
2048
when the command contains whitespace or metacharacters, or
2049
is built up at run time.
2051
2- A single string, eg "add a". This is the most convenient
2052
for hardcoded commands.
2054
This runs bzr through the interface that catches and reports
2055
errors, and with logging set to something approximating the
2056
default, so that error reporting can be checked.
2058
This should be the main method for tests that want to exercise the
2059
overall behavior of the bzr application (rather than a unit test
2060
or a functional test of the library.)
2062
This sends the stdout/stderr results into the test's log,
2063
where it may be useful for debugging. See also run_captured.
2065
:keyword stdin: A string to be used as stdin for the command.
2066
:keyword retcode: The status code the command should return;
2068
:keyword working_dir: The directory to run the command in
2069
:keyword error_regexes: A list of expected error messages. If
2070
specified they must be seen in the error output of the command.
2072
retcode, out, err = self._run_bzr_autosplit(
2077
working_dir=working_dir,
2079
self.assertIsInstance(error_regexes, (list, tuple))
2080
for regex in error_regexes:
2081
self.assertContainsRe(err, regex)
2084
def run_bzr_error(self, error_regexes, *args, **kwargs):
2085
"""Run bzr, and check that stderr contains the supplied regexes
2087
:param error_regexes: Sequence of regular expressions which
2088
must each be found in the error output. The relative ordering
2090
:param args: command-line arguments for bzr
2091
:param kwargs: Keyword arguments which are interpreted by run_bzr
2092
This function changes the default value of retcode to be 3,
2093
since in most cases this is run when you expect bzr to fail.
2095
:return: (out, err) The actual output of running the command (in case
2096
you want to do more inspection)
2100
# Make sure that commit is failing because there is nothing to do
2101
self.run_bzr_error(['no changes to commit'],
2102
['commit', '-m', 'my commit comment'])
2103
# Make sure --strict is handling an unknown file, rather than
2104
# giving us the 'nothing to do' error
2105
self.build_tree(['unknown'])
2106
self.run_bzr_error(['Commit refused because there are unknown files'],
2107
['commit', --strict', '-m', 'my commit comment'])
2109
kwargs.setdefault('retcode', 3)
2110
kwargs['error_regexes'] = error_regexes
2111
out, err = self.run_bzr(*args, **kwargs)
2114
def run_bzr_subprocess(self, *args, **kwargs):
2115
"""Run bzr in a subprocess for testing.
2117
This starts a new Python interpreter and runs bzr in there.
2118
This should only be used for tests that have a justifiable need for
2119
this isolation: e.g. they are testing startup time, or signal
2120
handling, or early startup code, etc. Subprocess code can't be
2121
profiled or debugged so easily.
2123
:keyword retcode: The status code that is expected. Defaults to 0. If
2124
None is supplied, the status code is not checked.
2125
:keyword env_changes: A dictionary which lists changes to environment
2126
variables. A value of None will unset the env variable.
2127
The values must be strings. The change will only occur in the
2128
child, so you don't need to fix the environment after running.
2129
:keyword universal_newlines: Convert CRLF => LF
2130
:keyword allow_plugins: By default the subprocess is run with
2131
--no-plugins to ensure test reproducibility. Also, it is possible
2132
for system-wide plugins to create unexpected output on stderr,
2133
which can cause unnecessary test failures.
2135
env_changes = kwargs.get('env_changes', {})
2136
working_dir = kwargs.get('working_dir', None)
2137
allow_plugins = kwargs.get('allow_plugins', False)
2139
if isinstance(args[0], list):
2141
elif isinstance(args[0], basestring):
2142
args = list(shlex.split(args[0]))
2144
raise ValueError("passing varargs to run_bzr_subprocess")
2145
process = self.start_bzr_subprocess(args, env_changes=env_changes,
2146
working_dir=working_dir,
2147
allow_plugins=allow_plugins)
2148
# We distinguish between retcode=None and retcode not passed.
2149
supplied_retcode = kwargs.get('retcode', 0)
2150
return self.finish_bzr_subprocess(process, retcode=supplied_retcode,
2151
universal_newlines=kwargs.get('universal_newlines', False),
2154
def start_bzr_subprocess(self, process_args, env_changes=None,
2155
skip_if_plan_to_signal=False,
2157
allow_plugins=False, stderr=subprocess.PIPE):
2158
"""Start bzr in a subprocess for testing.
2160
This starts a new Python interpreter and runs bzr in there.
2161
This should only be used for tests that have a justifiable need for
2162
this isolation: e.g. they are testing startup time, or signal
2163
handling, or early startup code, etc. Subprocess code can't be
2164
profiled or debugged so easily.
2166
:param process_args: a list of arguments to pass to the bzr executable,
2167
for example ``['--version']``.
2168
:param env_changes: A dictionary which lists changes to environment
2169
variables. A value of None will unset the env variable.
2170
The values must be strings. The change will only occur in the
2171
child, so you don't need to fix the environment after running.
2172
:param skip_if_plan_to_signal: raise TestSkipped when true and system
2173
doesn't support signalling subprocesses.
2174
:param allow_plugins: If False (default) pass --no-plugins to bzr.
2175
:param stderr: file to use for the subprocess's stderr. Valid values
2176
are those valid for the stderr argument of `subprocess.Popen`.
2177
Default value is ``subprocess.PIPE``.
2179
:returns: Popen object for the started process.
2181
if skip_if_plan_to_signal:
2182
if os.name != "posix":
2183
raise TestSkipped("Sending signals not supported")
2185
if env_changes is None:
2189
def cleanup_environment():
2190
for env_var, value in env_changes.iteritems():
2191
old_env[env_var] = osutils.set_or_unset_env(env_var, value)
2193
def restore_environment():
2194
for env_var, value in old_env.iteritems():
2195
osutils.set_or_unset_env(env_var, value)
2197
bzr_path = self.get_bzr_path()
2200
if working_dir is not None:
2201
cwd = osutils.getcwd()
2202
os.chdir(working_dir)
2205
# win32 subprocess doesn't support preexec_fn
2206
# so we will avoid using it on all platforms, just to
2207
# make sure the code path is used, and we don't break on win32
2208
cleanup_environment()
2209
# Include the subprocess's log file in the test details, in case
2210
# the test fails due to an error in the subprocess.
2211
self._add_subprocess_log(trace._get_bzr_log_filename())
2212
command = [sys.executable]
2213
# frozen executables don't need the path to bzr
2214
if getattr(sys, "frozen", None) is None:
2215
command.append(bzr_path)
2216
if not allow_plugins:
2217
command.append('--no-plugins')
2218
command.extend(process_args)
2219
process = self._popen(command, stdin=subprocess.PIPE,
2220
stdout=subprocess.PIPE,
2223
restore_environment()
2229
def _add_subprocess_log(self, log_file_path):
2230
if len(self._log_files) == 0:
2231
# Register an addCleanup func. We do this on the first call to
2232
# _add_subprocess_log rather than in TestCase.setUp so that this
2233
# addCleanup is registered after any cleanups for tempdirs that
2234
# subclasses might create, which will probably remove the log file
2236
self.addCleanup(self._subprocess_log_cleanup)
2237
# self._log_files is a set, so if a log file is reused we won't grab it
2239
self._log_files.add(log_file_path)
2241
def _subprocess_log_cleanup(self):
2242
for count, log_file_path in enumerate(self._log_files):
2243
# We use buffer_now=True to avoid holding the file open beyond
2244
# the life of this function, which might interfere with e.g.
2245
# cleaning tempdirs on Windows.
2246
# XXX: Testtools 0.9.5 doesn't have the content_from_file helper
2247
#detail_content = content.content_from_file(
2248
# log_file_path, buffer_now=True)
2249
with open(log_file_path, 'rb') as log_file:
2250
log_file_bytes = log_file.read()
2251
detail_content = content.Content(content.ContentType("text",
2252
"plain", {"charset": "utf8"}), lambda: [log_file_bytes])
2253
self.addDetail("start_bzr_subprocess-log-%d" % (count,),
2256
def _popen(self, *args, **kwargs):
2257
"""Place a call to Popen.
2259
Allows tests to override this method to intercept the calls made to
2260
Popen for introspection.
2262
return subprocess.Popen(*args, **kwargs)
2264
def get_source_path(self):
2265
"""Return the path of the directory containing bzrlib."""
2266
return os.path.dirname(os.path.dirname(bzrlib.__file__))
2268
def get_bzr_path(self):
2269
"""Return the path of the 'bzr' executable for this test suite."""
2270
bzr_path = os.path.join(self.get_source_path(), "bzr")
2271
if not os.path.isfile(bzr_path):
2272
# We are probably installed. Assume sys.argv is the right file
2273
bzr_path = sys.argv[0]
2276
def finish_bzr_subprocess(self, process, retcode=0, send_signal=None,
2277
universal_newlines=False, process_args=None):
2278
"""Finish the execution of process.
2280
:param process: the Popen object returned from start_bzr_subprocess.
2281
:param retcode: The status code that is expected. Defaults to 0. If
2282
None is supplied, the status code is not checked.
2283
:param send_signal: an optional signal to send to the process.
2284
:param universal_newlines: Convert CRLF => LF
2285
:returns: (stdout, stderr)
2287
if send_signal is not None:
2288
os.kill(process.pid, send_signal)
2289
out, err = process.communicate()
2291
if universal_newlines:
2292
out = out.replace('\r\n', '\n')
2293
err = err.replace('\r\n', '\n')
2295
if retcode is not None and retcode != process.returncode:
2296
if process_args is None:
2297
process_args = "(unknown args)"
2298
trace.mutter('Output of bzr %s:\n%s', process_args, out)
2299
trace.mutter('Error for bzr %s:\n%s', process_args, err)
2300
self.fail('Command bzr %s failed with retcode %s != %s'
2301
% (process_args, retcode, process.returncode))
2304
def check_tree_shape(self, tree, shape):
2305
"""Compare a tree to a list of expected names.
2307
Fail if they are not precisely equal.
2310
shape = list(shape) # copy
2311
for path, ie in tree.iter_entries_by_dir():
2312
name = path.replace('\\', '/')
2313
if ie.kind == 'directory':
2316
pass # ignore root entry
2322
self.fail("expected paths not found in inventory: %r" % shape)
2324
self.fail("unexpected paths found in inventory: %r" % extras)
2326
def apply_redirected(self, stdin=None, stdout=None, stderr=None,
2327
a_callable=None, *args, **kwargs):
2328
"""Call callable with redirected std io pipes.
2330
Returns the return code."""
2331
if not callable(a_callable):
2332
raise ValueError("a_callable must be callable.")
2334
stdin = StringIO("")
2336
if getattr(self, "_log_file", None) is not None:
2337
stdout = self._log_file
2341
if getattr(self, "_log_file", None is not None):
2342
stderr = self._log_file
2345
real_stdin = sys.stdin
2346
real_stdout = sys.stdout
2347
real_stderr = sys.stderr
2352
return a_callable(*args, **kwargs)
2354
sys.stdout = real_stdout
2355
sys.stderr = real_stderr
2356
sys.stdin = real_stdin
2358
def reduceLockdirTimeout(self):
2359
"""Reduce the default lock timeout for the duration of the test, so that
2360
if LockContention occurs during a test, it does so quickly.
2362
Tests that expect to provoke LockContention errors should call this.
2364
self.overrideAttr(lockdir, '_DEFAULT_TIMEOUT_SECONDS', 0)
2366
def make_utf8_encoded_stringio(self, encoding_type=None):
2367
"""Return a StringIOWrapper instance, that will encode Unicode
2370
if encoding_type is None:
2371
encoding_type = 'strict'
2373
output_encoding = 'utf-8'
2374
sio = codecs.getwriter(output_encoding)(sio, errors=encoding_type)
2375
sio.encoding = output_encoding
2378
def disable_verb(self, verb):
2379
"""Disable a smart server verb for one test."""
2380
from bzrlib.smart import request
2381
request_handlers = request.request_handlers
2382
orig_method = request_handlers.get(verb)
2383
orig_info = request_handlers.get_info(verb)
2384
request_handlers.remove(verb)
2385
self.addCleanup(request_handlers.register, verb, orig_method,
2389
class CapturedCall(object):
2390
"""A helper for capturing smart server calls for easy debug analysis."""
2392
def __init__(self, params, prefix_length):
2393
"""Capture the call with params and skip prefix_length stack frames."""
2396
# The last 5 frames are the __init__, the hook frame, and 3 smart
2397
# client frames. Beyond this we could get more clever, but this is good
2399
stack = traceback.extract_stack()[prefix_length:-5]
2400
self.stack = ''.join(traceback.format_list(stack))
2403
return self.call.method
2406
return self.call.method
2412
class TestCaseWithMemoryTransport(TestCase):
2413
"""Common test class for tests that do not need disk resources.
2415
Tests that need disk resources should derive from TestCaseInTempDir
2416
orTestCaseWithTransport.
2418
TestCaseWithMemoryTransport sets the TEST_ROOT variable for all bzr tests.
2420
For TestCaseWithMemoryTransport the ``test_home_dir`` is set to the name of
2421
a directory which does not exist. This serves to help ensure test isolation
2422
is preserved. ``test_dir`` is set to the TEST_ROOT, as is cwd, because they
2423
must exist. However, TestCaseWithMemoryTransport does not offer local file
2424
defaults for the transport in tests, nor does it obey the command line
2425
override, so tests that accidentally write to the common directory should
2428
:cvar TEST_ROOT: Directory containing all temporary directories, plus a
2429
``.bzr`` directory that stops us ascending higher into the filesystem.
2435
def __init__(self, methodName='runTest'):
2436
# allow test parameterization after test construction and before test
2437
# execution. Variables that the parameterizer sets need to be
2438
# ones that are not set by setUp, or setUp will trash them.
2439
super(TestCaseWithMemoryTransport, self).__init__(methodName)
2440
self.vfs_transport_factory = default_transport
2441
self.transport_server = None
2442
self.transport_readonly_server = None
2443
self.__vfs_server = None
2445
def get_transport(self, relpath=None):
2446
"""Return a writeable transport.
2448
This transport is for the test scratch space relative to
2451
:param relpath: a path relative to the base url.
2453
t = _mod_transport.get_transport_from_url(self.get_url(relpath))
2454
self.assertFalse(t.is_readonly())
2457
def get_readonly_transport(self, relpath=None):
2458
"""Return a readonly transport for the test scratch space
2460
This can be used to test that operations which should only need
2461
readonly access in fact do not try to write.
2463
:param relpath: a path relative to the base url.
2465
t = _mod_transport.get_transport_from_url(
2466
self.get_readonly_url(relpath))
2467
self.assertTrue(t.is_readonly())
2470
def create_transport_readonly_server(self):
2471
"""Create a transport server from class defined at init.
2473
This is mostly a hook for daughter classes.
2475
return self.transport_readonly_server()
2477
def get_readonly_server(self):
2478
"""Get the server instance for the readonly transport
2480
This is useful for some tests with specific servers to do diagnostics.
2482
if self.__readonly_server is None:
2483
if self.transport_readonly_server is None:
2484
# readonly decorator requested
2485
self.__readonly_server = test_server.ReadonlyServer()
2487
# explicit readonly transport.
2488
self.__readonly_server = self.create_transport_readonly_server()
2489
self.start_server(self.__readonly_server,
2490
self.get_vfs_only_server())
2491
return self.__readonly_server
2493
def get_readonly_url(self, relpath=None):
2494
"""Get a URL for the readonly transport.
2496
This will either be backed by '.' or a decorator to the transport
2497
used by self.get_url()
2498
relpath provides for clients to get a path relative to the base url.
2499
These should only be downwards relative, not upwards.
2501
base = self.get_readonly_server().get_url()
2502
return self._adjust_url(base, relpath)
2504
def get_vfs_only_server(self):
2505
"""Get the vfs only read/write server instance.
2507
This is useful for some tests with specific servers that need
2510
For TestCaseWithMemoryTransport this is always a MemoryServer, and there
2511
is no means to override it.
2513
if self.__vfs_server is None:
2514
self.__vfs_server = memory.MemoryServer()
2515
self.start_server(self.__vfs_server)
2516
return self.__vfs_server
2518
def get_server(self):
2519
"""Get the read/write server instance.
2521
This is useful for some tests with specific servers that need
2524
This is built from the self.transport_server factory. If that is None,
2525
then the self.get_vfs_server is returned.
2527
if self.__server is None:
2528
if (self.transport_server is None or self.transport_server is
2529
self.vfs_transport_factory):
2530
self.__server = self.get_vfs_only_server()
2532
# bring up a decorated means of access to the vfs only server.
2533
self.__server = self.transport_server()
2534
self.start_server(self.__server, self.get_vfs_only_server())
2535
return self.__server
2537
def _adjust_url(self, base, relpath):
2538
"""Get a URL (or maybe a path) for the readwrite transport.
2540
This will either be backed by '.' or to an equivalent non-file based
2542
relpath provides for clients to get a path relative to the base url.
2543
These should only be downwards relative, not upwards.
2545
if relpath is not None and relpath != '.':
2546
if not base.endswith('/'):
2548
# XXX: Really base should be a url; we did after all call
2549
# get_url()! But sometimes it's just a path (from
2550
# LocalAbspathServer), and it'd be wrong to append urlescaped data
2551
# to a non-escaped local path.
2552
if base.startswith('./') or base.startswith('/'):
2555
base += urlutils.escape(relpath)
2558
def get_url(self, relpath=None):
2559
"""Get a URL (or maybe a path) for the readwrite transport.
2561
This will either be backed by '.' or to an equivalent non-file based
2563
relpath provides for clients to get a path relative to the base url.
2564
These should only be downwards relative, not upwards.
2566
base = self.get_server().get_url()
2567
return self._adjust_url(base, relpath)
2569
def get_vfs_only_url(self, relpath=None):
2570
"""Get a URL (or maybe a path for the plain old vfs transport.
2572
This will never be a smart protocol. It always has all the
2573
capabilities of the local filesystem, but it might actually be a
2574
MemoryTransport or some other similar virtual filesystem.
2576
This is the backing transport (if any) of the server returned by
2577
get_url and get_readonly_url.
2579
:param relpath: provides for clients to get a path relative to the base
2580
url. These should only be downwards relative, not upwards.
2583
base = self.get_vfs_only_server().get_url()
2584
return self._adjust_url(base, relpath)
2586
def _create_safety_net(self):
2587
"""Make a fake bzr directory.
2589
This prevents any tests propagating up onto the TEST_ROOT directory's
2592
root = TestCaseWithMemoryTransport.TEST_ROOT
2593
# Make sure we get a readable and accessible home for .bzr.log
2594
# and/or config files, and not fallback to weird defaults (see
2595
# http://pad.lv/825027).
2596
self.assertIs(None, os.environ.get('BZR_HOME', None))
2597
os.environ['BZR_HOME'] = root
2598
wt = bzrdir.BzrDir.create_standalone_workingtree(root)
2599
del os.environ['BZR_HOME']
2600
# Hack for speed: remember the raw bytes of the dirstate file so that
2601
# we don't need to re-open the wt to check it hasn't changed.
2602
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE = (
2603
wt.control_transport.get_bytes('dirstate'))
2605
def _check_safety_net(self):
2606
"""Check that the safety .bzr directory have not been touched.
2608
_make_test_root have created a .bzr directory to prevent tests from
2609
propagating. This method ensures than a test did not leaked.
2611
root = TestCaseWithMemoryTransport.TEST_ROOT
2612
t = _mod_transport.get_transport_from_path(root)
2613
self.permit_url(t.base)
2614
if (t.get_bytes('.bzr/checkout/dirstate') !=
2615
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE):
2616
# The current test have modified the /bzr directory, we need to
2617
# recreate a new one or all the followng tests will fail.
2618
# If you need to inspect its content uncomment the following line
2619
# import pdb; pdb.set_trace()
2620
_rmtree_temp_dir(root + '/.bzr', test_id=self.id())
2621
self._create_safety_net()
2622
raise AssertionError('%s/.bzr should not be modified' % root)
2624
def _make_test_root(self):
2625
if TestCaseWithMemoryTransport.TEST_ROOT is None:
2626
# Watch out for tricky test dir (on OSX /tmp -> /private/tmp)
2627
root = osutils.realpath(osutils.mkdtemp(prefix='testbzr-',
2629
TestCaseWithMemoryTransport.TEST_ROOT = root
2631
self._create_safety_net()
2633
# The same directory is used by all tests, and we're not
2634
# specifically told when all tests are finished. This will do.
2635
atexit.register(_rmtree_temp_dir, root)
2637
self.permit_dir(TestCaseWithMemoryTransport.TEST_ROOT)
2638
self.addCleanup(self._check_safety_net)
2640
def makeAndChdirToTestDir(self):
2641
"""Create a temporary directories for this one test.
2643
This must set self.test_home_dir and self.test_dir and chdir to
2646
For TestCaseWithMemoryTransport we chdir to the TEST_ROOT for this test.
2648
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2649
self.test_dir = TestCaseWithMemoryTransport.TEST_ROOT
2650
self.test_home_dir = self.test_dir + "/MemoryTransportMissingHomeDir"
2651
self.permit_dir(self.test_dir)
2653
def make_branch(self, relpath, format=None):
2654
"""Create a branch on the transport at relpath."""
2655
repo = self.make_repository(relpath, format=format)
2656
return repo.bzrdir.create_branch(append_revisions_only=False)
2658
def get_default_format(self):
2661
def resolve_format(self, format):
2662
"""Resolve an object to a ControlDir format object.
2664
The initial format object can either already be
2665
a ControlDirFormat, None (for the default format),
2666
or a string with the name of the control dir format.
2668
:param format: Object to resolve
2669
:return A ControlDirFormat instance
2672
format = self.get_default_format()
2673
if isinstance(format, basestring):
2674
format = bzrdir.format_registry.make_bzrdir(format)
2677
def make_bzrdir(self, relpath, format=None):
2679
# might be a relative or absolute path
2680
maybe_a_url = self.get_url(relpath)
2681
segments = maybe_a_url.rsplit('/', 1)
2682
t = _mod_transport.get_transport(maybe_a_url)
2683
if len(segments) > 1 and segments[-1] not in ('', '.'):
2685
format = self.resolve_format(format)
2686
return format.initialize_on_transport(t)
2687
except errors.UninitializableFormat:
2688
raise TestSkipped("Format %s is not initializable." % format)
2690
def make_repository(self, relpath, shared=None, format=None):
2691
"""Create a repository on our default transport at relpath.
2693
Note that relpath must be a relative path, not a full url.
2695
# FIXME: If you create a remoterepository this returns the underlying
2696
# real format, which is incorrect. Actually we should make sure that
2697
# RemoteBzrDir returns a RemoteRepository.
2698
# maybe mbp 20070410
2699
made_control = self.make_bzrdir(relpath, format=format)
2700
return made_control.create_repository(shared=shared)
2702
def make_smart_server(self, path, backing_server=None):
2703
if backing_server is None:
2704
backing_server = self.get_server()
2705
smart_server = test_server.SmartTCPServer_for_testing()
2706
self.start_server(smart_server, backing_server)
2707
remote_transport = _mod_transport.get_transport_from_url(smart_server.get_url()
2709
return remote_transport
2711
def make_branch_and_memory_tree(self, relpath, format=None):
2712
"""Create a branch on the default transport and a MemoryTree for it."""
2713
b = self.make_branch(relpath, format=format)
2714
return memorytree.MemoryTree.create_on_branch(b)
2716
def make_branch_builder(self, relpath, format=None):
2717
branch = self.make_branch(relpath, format=format)
2718
return branchbuilder.BranchBuilder(branch=branch)
2720
def overrideEnvironmentForTesting(self):
2721
test_home_dir = self.test_home_dir
2722
if isinstance(test_home_dir, unicode):
2723
test_home_dir = test_home_dir.encode(sys.getfilesystemencoding())
2724
self.overrideEnv('HOME', test_home_dir)
2725
self.overrideEnv('BZR_HOME', test_home_dir)
2728
super(TestCaseWithMemoryTransport, self).setUp()
2729
# Ensure that ConnectedTransport doesn't leak sockets
2730
def get_transport_from_url_with_cleanup(*args, **kwargs):
2731
t = orig_get_transport_from_url(*args, **kwargs)
2732
if isinstance(t, _mod_transport.ConnectedTransport):
2733
self.addCleanup(t.disconnect)
2736
orig_get_transport_from_url = self.overrideAttr(
2737
_mod_transport, 'get_transport_from_url',
2738
get_transport_from_url_with_cleanup)
2739
self._make_test_root()
2740
self.addCleanup(os.chdir, os.getcwdu())
2741
self.makeAndChdirToTestDir()
2742
self.overrideEnvironmentForTesting()
2743
self.__readonly_server = None
2744
self.__server = None
2745
self.reduceLockdirTimeout()
2747
def setup_smart_server_with_call_log(self):
2748
"""Sets up a smart server as the transport server with a call log."""
2749
self.transport_server = test_server.SmartTCPServer_for_testing
2750
self.hpss_calls = []
2752
# Skip the current stack down to the caller of
2753
# setup_smart_server_with_call_log
2754
prefix_length = len(traceback.extract_stack()) - 2
2755
def capture_hpss_call(params):
2756
self.hpss_calls.append(
2757
CapturedCall(params, prefix_length))
2758
client._SmartClient.hooks.install_named_hook(
2759
'call', capture_hpss_call, None)
2761
def reset_smart_call_log(self):
2762
self.hpss_calls = []
2765
class TestCaseInTempDir(TestCaseWithMemoryTransport):
2766
"""Derived class that runs a test within a temporary directory.
2768
This is useful for tests that need to create a branch, etc.
2770
The directory is created in a slightly complex way: for each
2771
Python invocation, a new temporary top-level directory is created.
2772
All test cases create their own directory within that. If the
2773
tests complete successfully, the directory is removed.
2775
:ivar test_base_dir: The path of the top-level directory for this
2776
test, which contains a home directory and a work directory.
2778
:ivar test_home_dir: An initially empty directory under test_base_dir
2779
which is used as $HOME for this test.
2781
:ivar test_dir: A directory under test_base_dir used as the current
2782
directory when the test proper is run.
2785
OVERRIDE_PYTHON = 'python'
2788
super(TestCaseInTempDir, self).setUp()
2789
# Remove the protection set in isolated_environ, we have a proper
2790
# access to disk resources now.
2791
self.overrideEnv('BZR_LOG', None)
2793
def check_file_contents(self, filename, expect):
2794
self.log("check contents of file %s" % filename)
2800
if contents != expect:
2801
self.log("expected: %r" % expect)
2802
self.log("actually: %r" % contents)
2803
self.fail("contents of %s not as expected" % filename)
2805
def _getTestDirPrefix(self):
2806
# create a directory within the top level test directory
2807
if sys.platform in ('win32', 'cygwin'):
2808
name_prefix = re.sub('[<>*=+",:;_/\\-]', '_', self.id())
2809
# windows is likely to have path-length limits so use a short name
2810
name_prefix = name_prefix[-30:]
2812
name_prefix = re.sub('[/]', '_', self.id())
2815
def makeAndChdirToTestDir(self):
2816
"""See TestCaseWithMemoryTransport.makeAndChdirToTestDir().
2818
For TestCaseInTempDir we create a temporary directory based on the test
2819
name and then create two subdirs - test and home under it.
2821
name_prefix = osutils.pathjoin(TestCaseWithMemoryTransport.TEST_ROOT,
2822
self._getTestDirPrefix())
2824
for i in range(100):
2825
if os.path.exists(name):
2826
name = name_prefix + '_' + str(i)
2828
# now create test and home directories within this dir
2829
self.test_base_dir = name
2830
self.addCleanup(self.deleteTestDir)
2831
os.mkdir(self.test_base_dir)
2833
self.permit_dir(self.test_base_dir)
2834
# 'sprouting' and 'init' of a branch both walk up the tree to find
2835
# stacking policy to honour; create a bzr dir with an unshared
2836
# repository (but not a branch - our code would be trying to escape
2837
# then!) to stop them, and permit it to be read.
2838
# control = bzrdir.BzrDir.create(self.test_base_dir)
2839
# control.create_repository()
2840
self.test_home_dir = self.test_base_dir + '/home'
2841
os.mkdir(self.test_home_dir)
2842
self.test_dir = self.test_base_dir + '/work'
2843
os.mkdir(self.test_dir)
2844
os.chdir(self.test_dir)
2845
# put name of test inside
2846
f = file(self.test_base_dir + '/name', 'w')
2852
def deleteTestDir(self):
2853
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2854
_rmtree_temp_dir(self.test_base_dir, test_id=self.id())
2856
def build_tree(self, shape, line_endings='binary', transport=None):
2857
"""Build a test tree according to a pattern.
2859
shape is a sequence of file specifications. If the final
2860
character is '/', a directory is created.
2862
This assumes that all the elements in the tree being built are new.
2864
This doesn't add anything to a branch.
2866
:type shape: list or tuple.
2867
:param line_endings: Either 'binary' or 'native'
2868
in binary mode, exact contents are written in native mode, the
2869
line endings match the default platform endings.
2870
:param transport: A transport to write to, for building trees on VFS's.
2871
If the transport is readonly or None, "." is opened automatically.
2874
if type(shape) not in (list, tuple):
2875
raise AssertionError("Parameter 'shape' should be "
2876
"a list or a tuple. Got %r instead" % (shape,))
2877
# It's OK to just create them using forward slashes on windows.
2878
if transport is None or transport.is_readonly():
2879
transport = _mod_transport.get_transport_from_path(".")
2881
self.assertIsInstance(name, basestring)
2883
transport.mkdir(urlutils.escape(name[:-1]))
2885
if line_endings == 'binary':
2887
elif line_endings == 'native':
2890
raise errors.BzrError(
2891
'Invalid line ending request %r' % line_endings)
2892
content = "contents of %s%s" % (name.encode('utf-8'), end)
2893
transport.put_bytes_non_atomic(urlutils.escape(name), content)
2895
build_tree_contents = staticmethod(treeshape.build_tree_contents)
2897
def assertInWorkingTree(self, path, root_path='.', tree=None):
2898
"""Assert whether path or paths are in the WorkingTree"""
2900
tree = workingtree.WorkingTree.open(root_path)
2901
if not isinstance(path, basestring):
2903
self.assertInWorkingTree(p, tree=tree)
2905
self.assertIsNot(tree.path2id(path), None,
2906
path+' not in working tree.')
2908
def assertNotInWorkingTree(self, path, root_path='.', tree=None):
2909
"""Assert whether path or paths are not in the WorkingTree"""
2911
tree = workingtree.WorkingTree.open(root_path)
2912
if not isinstance(path, basestring):
2914
self.assertNotInWorkingTree(p,tree=tree)
2916
self.assertIs(tree.path2id(path), None, path+' in working tree.')
2919
class TestCaseWithTransport(TestCaseInTempDir):
2920
"""A test case that provides get_url and get_readonly_url facilities.
2922
These back onto two transport servers, one for readonly access and one for
2925
If no explicit class is provided for readonly access, a
2926
ReadonlyTransportDecorator is used instead which allows the use of non disk
2927
based read write transports.
2929
If an explicit class is provided for readonly access, that server and the
2930
readwrite one must both define get_url() as resolving to os.getcwd().
2933
def get_vfs_only_server(self):
2934
"""See TestCaseWithMemoryTransport.
2936
This is useful for some tests with specific servers that need
2939
if self.__vfs_server is None:
2940
self.__vfs_server = self.vfs_transport_factory()
2941
self.start_server(self.__vfs_server)
2942
return self.__vfs_server
2944
def make_branch_and_tree(self, relpath, format=None):
2945
"""Create a branch on the transport and a tree locally.
2947
If the transport is not a LocalTransport, the Tree can't be created on
2948
the transport. In that case if the vfs_transport_factory is
2949
LocalURLServer the working tree is created in the local
2950
directory backing the transport, and the returned tree's branch and
2951
repository will also be accessed locally. Otherwise a lightweight
2952
checkout is created and returned.
2954
We do this because we can't physically create a tree in the local
2955
path, with a branch reference to the transport_factory url, and
2956
a branch + repository in the vfs_transport, unless the vfs_transport
2957
namespace is distinct from the local disk - the two branch objects
2958
would collide. While we could construct a tree with its branch object
2959
pointing at the transport_factory transport in memory, reopening it
2960
would behaving unexpectedly, and has in the past caused testing bugs
2961
when we tried to do it that way.
2963
:param format: The BzrDirFormat.
2964
:returns: the WorkingTree.
2966
# TODO: always use the local disk path for the working tree,
2967
# this obviously requires a format that supports branch references
2968
# so check for that by checking bzrdir.BzrDirFormat.get_default_format()
2970
format = self.resolve_format(format=format)
2971
if not format.supports_workingtrees:
2972
b = self.make_branch(relpath+'.branch', format=format)
2973
return b.create_checkout(relpath, lightweight=True)
2974
b = self.make_branch(relpath, format=format)
2976
return b.bzrdir.create_workingtree()
2977
except errors.NotLocalUrl:
2978
# We can only make working trees locally at the moment. If the
2979
# transport can't support them, then we keep the non-disk-backed
2980
# branch and create a local checkout.
2981
if self.vfs_transport_factory is test_server.LocalURLServer:
2982
# the branch is colocated on disk, we cannot create a checkout.
2983
# hopefully callers will expect this.
2984
local_controldir= bzrdir.BzrDir.open(self.get_vfs_only_url(relpath))
2985
wt = local_controldir.create_workingtree()
2986
if wt.branch._format != b._format:
2988
# Make sure that assigning to wt._branch fixes wt.branch,
2989
# in case the implementation details of workingtree objects
2991
self.assertIs(b, wt.branch)
2994
return b.create_checkout(relpath, lightweight=True)
2996
def assertIsDirectory(self, relpath, transport):
2997
"""Assert that relpath within transport is a directory.
2999
This may not be possible on all transports; in that case it propagates
3000
a TransportNotPossible.
3003
mode = transport.stat(relpath).st_mode
3004
except errors.NoSuchFile:
3005
self.fail("path %s is not a directory; no such file"
3007
if not stat.S_ISDIR(mode):
3008
self.fail("path %s is not a directory; has mode %#o"
3011
def assertTreesEqual(self, left, right):
3012
"""Check that left and right have the same content and properties."""
3013
# we use a tree delta to check for equality of the content, and we
3014
# manually check for equality of other things such as the parents list.
3015
self.assertEqual(left.get_parent_ids(), right.get_parent_ids())
3016
differences = left.changes_from(right)
3017
self.assertFalse(differences.has_changed(),
3018
"Trees %r and %r are different: %r" % (left, right, differences))
3021
super(TestCaseWithTransport, self).setUp()
3022
self.__vfs_server = None
3024
def disable_missing_extensions_warning(self):
3025
"""Some tests expect a precise stderr content.
3027
There is no point in forcing them to duplicate the extension related
3030
config.GlobalConfig().set_user_option('ignore_missing_extensions', True)
3033
class ChrootedTestCase(TestCaseWithTransport):
3034
"""A support class that provides readonly urls outside the local namespace.
3036
This is done by checking if self.transport_server is a MemoryServer. if it
3037
is then we are chrooted already, if it is not then an HttpServer is used
3040
TODO RBC 20060127: make this an option to TestCaseWithTransport so it can
3041
be used without needed to redo it when a different
3042
subclass is in use ?
3046
from bzrlib.tests import http_server
3047
super(ChrootedTestCase, self).setUp()
3048
if not self.vfs_transport_factory == memory.MemoryServer:
3049
self.transport_readonly_server = http_server.HttpServer
3052
def condition_id_re(pattern):
3053
"""Create a condition filter which performs a re check on a test's id.
3055
:param pattern: A regular expression string.
3056
:return: A callable that returns True if the re matches.
3058
filter_re = re.compile(pattern, 0)
3059
def condition(test):
3061
return filter_re.search(test_id)
3065
def condition_isinstance(klass_or_klass_list):
3066
"""Create a condition filter which returns isinstance(param, klass).
3068
:return: A callable which when called with one parameter obj return the
3069
result of isinstance(obj, klass_or_klass_list).
3072
return isinstance(obj, klass_or_klass_list)
3076
def condition_id_in_list(id_list):
3077
"""Create a condition filter which verify that test's id in a list.
3079
:param id_list: A TestIdList object.
3080
:return: A callable that returns True if the test's id appears in the list.
3082
def condition(test):
3083
return id_list.includes(test.id())
3087
def condition_id_startswith(starts):
3088
"""Create a condition filter verifying that test's id starts with a string.
3090
:param starts: A list of string.
3091
:return: A callable that returns True if the test's id starts with one of
3094
def condition(test):
3095
for start in starts:
3096
if test.id().startswith(start):
3102
def exclude_tests_by_condition(suite, condition):
3103
"""Create a test suite which excludes some tests from suite.
3105
:param suite: The suite to get tests from.
3106
:param condition: A callable whose result evaluates True when called with a
3107
test case which should be excluded from the result.
3108
:return: A suite which contains the tests found in suite that fail
3112
for test in iter_suite_tests(suite):
3113
if not condition(test):
3115
return TestUtil.TestSuite(result)
3118
def filter_suite_by_condition(suite, condition):
3119
"""Create a test suite by filtering another one.
3121
:param suite: The source suite.
3122
:param condition: A callable whose result evaluates True when called with a
3123
test case which should be included in the result.
3124
:return: A suite which contains the tests found in suite that pass
3128
for test in iter_suite_tests(suite):
3131
return TestUtil.TestSuite(result)
3134
def filter_suite_by_re(suite, pattern):
3135
"""Create a test suite by filtering another one.
3137
:param suite: the source suite
3138
:param pattern: pattern that names must match
3139
:returns: the newly created suite
3141
condition = condition_id_re(pattern)
3142
result_suite = filter_suite_by_condition(suite, condition)
3146
def filter_suite_by_id_list(suite, test_id_list):
3147
"""Create a test suite by filtering another one.
3149
:param suite: The source suite.
3150
:param test_id_list: A list of the test ids to keep as strings.
3151
:returns: the newly created suite
3153
condition = condition_id_in_list(test_id_list)
3154
result_suite = filter_suite_by_condition(suite, condition)
3158
def filter_suite_by_id_startswith(suite, start):
3159
"""Create a test suite by filtering another one.
3161
:param suite: The source suite.
3162
:param start: A list of string the test id must start with one of.
3163
:returns: the newly created suite
3165
condition = condition_id_startswith(start)
3166
result_suite = filter_suite_by_condition(suite, condition)
3170
def exclude_tests_by_re(suite, pattern):
3171
"""Create a test suite which excludes some tests from suite.
3173
:param suite: The suite to get tests from.
3174
:param pattern: A regular expression string. Test ids that match this
3175
pattern will be excluded from the result.
3176
:return: A TestSuite that contains all the tests from suite without the
3177
tests that matched pattern. The order of tests is the same as it was in
3180
return exclude_tests_by_condition(suite, condition_id_re(pattern))
3183
def preserve_input(something):
3184
"""A helper for performing test suite transformation chains.
3186
:param something: Anything you want to preserve.
3192
def randomize_suite(suite):
3193
"""Return a new TestSuite with suite's tests in random order.
3195
The tests in the input suite are flattened into a single suite in order to
3196
accomplish this. Any nested TestSuites are removed to provide global
3199
tests = list(iter_suite_tests(suite))
3200
random.shuffle(tests)
3201
return TestUtil.TestSuite(tests)
3204
def split_suite_by_condition(suite, condition):
3205
"""Split a test suite into two by a condition.
3207
:param suite: The suite to split.
3208
:param condition: The condition to match on. Tests that match this
3209
condition are returned in the first test suite, ones that do not match
3210
are in the second suite.
3211
:return: A tuple of two test suites, where the first contains tests from
3212
suite matching the condition, and the second contains the remainder
3213
from suite. The order within each output suite is the same as it was in
3218
for test in iter_suite_tests(suite):
3220
matched.append(test)
3222
did_not_match.append(test)
3223
return TestUtil.TestSuite(matched), TestUtil.TestSuite(did_not_match)
3226
def split_suite_by_re(suite, pattern):
3227
"""Split a test suite into two by a regular expression.
3229
:param suite: The suite to split.
3230
:param pattern: A regular expression string. Test ids that match this
3231
pattern will be in the first test suite returned, and the others in the
3232
second test suite returned.
3233
:return: A tuple of two test suites, where the first contains tests from
3234
suite matching pattern, and the second contains the remainder from
3235
suite. The order within each output suite is the same as it was in
3238
return split_suite_by_condition(suite, condition_id_re(pattern))
3241
def run_suite(suite, name='test', verbose=False, pattern=".*",
3242
stop_on_failure=False,
3243
transport=None, lsprof_timed=None, bench_history=None,
3244
matching_tests_first=None,
3247
exclude_pattern=None,
3250
suite_decorators=None,
3252
result_decorators=None,
3254
"""Run a test suite for bzr selftest.
3256
:param runner_class: The class of runner to use. Must support the
3257
constructor arguments passed by run_suite which are more than standard
3259
:return: A boolean indicating success.
3261
TestCase._gather_lsprof_in_benchmarks = lsprof_timed
3266
if runner_class is None:
3267
runner_class = TextTestRunner
3270
runner = runner_class(stream=stream,
3272
verbosity=verbosity,
3273
bench_history=bench_history,
3275
result_decorators=result_decorators,
3277
runner.stop_on_failure=stop_on_failure
3278
if isinstance(suite, unittest.TestSuite):
3279
# Empty out _tests list of passed suite and populate new TestSuite
3280
suite._tests[:], suite = [], TestSuite(suite)
3281
# built in decorator factories:
3283
random_order(random_seed, runner),
3284
exclude_tests(exclude_pattern),
3286
if matching_tests_first:
3287
decorators.append(tests_first(pattern))
3289
decorators.append(filter_tests(pattern))
3290
if suite_decorators:
3291
decorators.extend(suite_decorators)
3292
# tell the result object how many tests will be running: (except if
3293
# --parallel=fork is being used. Robert said he will provide a better
3294
# progress design later -- vila 20090817)
3295
if fork_decorator not in decorators:
3296
decorators.append(CountingDecorator)
3297
for decorator in decorators:
3298
suite = decorator(suite)
3300
# Done after test suite decoration to allow randomisation etc
3301
# to take effect, though that is of marginal benefit.
3303
stream.write("Listing tests only ...\n")
3304
for t in iter_suite_tests(suite):
3305
stream.write("%s\n" % (t.id()))
3307
result = runner.run(suite)
3309
return result.wasStrictlySuccessful()
3311
return result.wasSuccessful()
3314
# A registry where get() returns a suite decorator.
3315
parallel_registry = registry.Registry()
3318
def fork_decorator(suite):
3319
if getattr(os, "fork", None) is None:
3320
raise errors.BzrCommandError("platform does not support fork,"
3321
" try --parallel=subprocess instead.")
3322
concurrency = osutils.local_concurrency()
3323
if concurrency == 1:
3325
from testtools import ConcurrentTestSuite
3326
return ConcurrentTestSuite(suite, fork_for_tests)
3327
parallel_registry.register('fork', fork_decorator)
3330
def subprocess_decorator(suite):
3331
concurrency = osutils.local_concurrency()
3332
if concurrency == 1:
3334
from testtools import ConcurrentTestSuite
3335
return ConcurrentTestSuite(suite, reinvoke_for_tests)
3336
parallel_registry.register('subprocess', subprocess_decorator)
3339
def exclude_tests(exclude_pattern):
3340
"""Return a test suite decorator that excludes tests."""
3341
if exclude_pattern is None:
3342
return identity_decorator
3343
def decorator(suite):
3344
return ExcludeDecorator(suite, exclude_pattern)
3348
def filter_tests(pattern):
3350
return identity_decorator
3351
def decorator(suite):
3352
return FilterTestsDecorator(suite, pattern)
3356
def random_order(random_seed, runner):
3357
"""Return a test suite decorator factory for randomising tests order.
3359
:param random_seed: now, a string which casts to a long, or a long.
3360
:param runner: A test runner with a stream attribute to report on.
3362
if random_seed is None:
3363
return identity_decorator
3364
def decorator(suite):
3365
return RandomDecorator(suite, random_seed, runner.stream)
3369
def tests_first(pattern):
3371
return identity_decorator
3372
def decorator(suite):
3373
return TestFirstDecorator(suite, pattern)
3377
def identity_decorator(suite):
3382
class TestDecorator(TestUtil.TestSuite):
3383
"""A decorator for TestCase/TestSuite objects.
3385
Contains rather than flattening suite passed on construction
3388
def __init__(self, suite=None):
3389
super(TestDecorator, self).__init__()
3390
if suite is not None:
3393
# Don't need subclass run method with suite emptying
3394
run = unittest.TestSuite.run
3397
class CountingDecorator(TestDecorator):
3398
"""A decorator which calls result.progress(self.countTestCases)."""
3400
def run(self, result):
3401
progress_method = getattr(result, 'progress', None)
3402
if callable(progress_method):
3403
progress_method(self.countTestCases(), SUBUNIT_SEEK_SET)
3404
return super(CountingDecorator, self).run(result)
3407
class ExcludeDecorator(TestDecorator):
3408
"""A decorator which excludes test matching an exclude pattern."""
3410
def __init__(self, suite, exclude_pattern):
3411
super(ExcludeDecorator, self).__init__(
3412
exclude_tests_by_re(suite, exclude_pattern))
3415
class FilterTestsDecorator(TestDecorator):
3416
"""A decorator which filters tests to those matching a pattern."""
3418
def __init__(self, suite, pattern):
3419
super(FilterTestsDecorator, self).__init__(
3420
filter_suite_by_re(suite, pattern))
3423
class RandomDecorator(TestDecorator):
3424
"""A decorator which randomises the order of its tests."""
3426
def __init__(self, suite, random_seed, stream):
3427
random_seed = self.actual_seed(random_seed)
3428
stream.write("Randomizing test order using seed %s\n\n" %
3430
# Initialise the random number generator.
3431
random.seed(random_seed)
3432
super(RandomDecorator, self).__init__(randomize_suite(suite))
3435
def actual_seed(seed):
3437
# We convert the seed to a long to make it reuseable across
3438
# invocations (because the user can reenter it).
3439
return long(time.time())
3441
# Convert the seed to a long if we can
3444
except (TypeError, ValueError):
3449
class TestFirstDecorator(TestDecorator):
3450
"""A decorator which moves named tests to the front."""
3452
def __init__(self, suite, pattern):
3453
super(TestFirstDecorator, self).__init__()
3454
self.addTests(split_suite_by_re(suite, pattern))
3457
def partition_tests(suite, count):
3458
"""Partition suite into count lists of tests."""
3459
# This just assigns tests in a round-robin fashion. On one hand this
3460
# splits up blocks of related tests that might run faster if they shared
3461
# resources, but on the other it avoids assigning blocks of slow tests to
3462
# just one partition. So the slowest partition shouldn't be much slower
3464
partitions = [list() for i in range(count)]
3465
tests = iter_suite_tests(suite)
3466
for partition, test in itertools.izip(itertools.cycle(partitions), tests):
3467
partition.append(test)
3471
def workaround_zealous_crypto_random():
3472
"""Crypto.Random want to help us being secure, but we don't care here.
3474
This workaround some test failure related to the sftp server. Once paramiko
3475
stop using the controversial API in Crypto.Random, we may get rid of it.
3478
from Crypto.Random import atfork
3484
def fork_for_tests(suite):
3485
"""Take suite and start up one runner per CPU by forking()
3487
:return: An iterable of TestCase-like objects which can each have
3488
run(result) called on them to feed tests to result.
3490
concurrency = osutils.local_concurrency()
3492
from subunit import ProtocolTestCase
3493
from subunit.test_results import AutoTimingTestResultDecorator
3494
class TestInOtherProcess(ProtocolTestCase):
3495
# Should be in subunit, I think. RBC.
3496
def __init__(self, stream, pid):
3497
ProtocolTestCase.__init__(self, stream)
3500
def run(self, result):
3502
ProtocolTestCase.run(self, result)
3504
pid, status = os.waitpid(self.pid, 0)
3505
# GZ 2011-10-18: If status is nonzero, should report to the result
3506
# that something went wrong.
3508
test_blocks = partition_tests(suite, concurrency)
3509
# Clear the tests from the original suite so it doesn't keep them alive
3510
suite._tests[:] = []
3511
for process_tests in test_blocks:
3512
process_suite = TestUtil.TestSuite(process_tests)
3513
# Also clear each split list so new suite has only reference
3514
process_tests[:] = []
3515
c2pread, c2pwrite = os.pipe()
3519
stream = os.fdopen(c2pwrite, 'wb', 1)
3520
workaround_zealous_crypto_random()
3522
# Leave stderr and stdout open so we can see test noise
3523
# Close stdin so that the child goes away if it decides to
3524
# read from stdin (otherwise its a roulette to see what
3525
# child actually gets keystrokes for pdb etc).
3527
subunit_result = AutoTimingTestResultDecorator(
3528
SubUnitBzrProtocolClient(stream))
3529
process_suite.run(subunit_result)
3531
# Try and report traceback on stream, but exit with error even
3532
# if stream couldn't be created or something else goes wrong.
3533
# The traceback is formatted to a string and written in one go
3534
# to avoid interleaving lines from multiple failing children.
3536
stream.write(traceback.format_exc())
3542
stream = os.fdopen(c2pread, 'rb', 1)
3543
test = TestInOtherProcess(stream, pid)
3548
def reinvoke_for_tests(suite):
3549
"""Take suite and start up one runner per CPU using subprocess().
3551
:return: An iterable of TestCase-like objects which can each have
3552
run(result) called on them to feed tests to result.
3554
concurrency = osutils.local_concurrency()
3556
from subunit import ProtocolTestCase
3557
class TestInSubprocess(ProtocolTestCase):
3558
def __init__(self, process, name):
3559
ProtocolTestCase.__init__(self, process.stdout)
3560
self.process = process
3561
self.process.stdin.close()
3564
def run(self, result):
3566
ProtocolTestCase.run(self, result)
3569
os.unlink(self.name)
3570
# print "pid %d finished" % finished_process
3571
test_blocks = partition_tests(suite, concurrency)
3572
for process_tests in test_blocks:
3573
# ugly; currently reimplement rather than reuses TestCase methods.
3574
bzr_path = os.path.dirname(os.path.dirname(bzrlib.__file__))+'/bzr'
3575
if not os.path.isfile(bzr_path):
3576
# We are probably installed. Assume sys.argv is the right file
3577
bzr_path = sys.argv[0]
3578
bzr_path = [bzr_path]
3579
if sys.platform == "win32":
3580
# if we're on windows, we can't execute the bzr script directly
3581
bzr_path = [sys.executable] + bzr_path
3582
fd, test_list_file_name = tempfile.mkstemp()
3583
test_list_file = os.fdopen(fd, 'wb', 1)
3584
for test in process_tests:
3585
test_list_file.write(test.id() + '\n')
3586
test_list_file.close()
3588
argv = bzr_path + ['selftest', '--load-list', test_list_file_name,
3590
if '--no-plugins' in sys.argv:
3591
argv.append('--no-plugins')
3592
# stderr=subprocess.STDOUT would be ideal, but until we prevent
3593
# noise on stderr it can interrupt the subunit protocol.
3594
process = subprocess.Popen(argv, stdin=subprocess.PIPE,
3595
stdout=subprocess.PIPE,
3596
stderr=subprocess.PIPE,
3598
test = TestInSubprocess(process, test_list_file_name)
3601
os.unlink(test_list_file_name)
3606
class ProfileResult(testtools.ExtendedToOriginalDecorator):
3607
"""Generate profiling data for all activity between start and success.
3609
The profile data is appended to the test's _benchcalls attribute and can
3610
be accessed by the forwarded-to TestResult.
3612
While it might be cleaner do accumulate this in stopTest, addSuccess is
3613
where our existing output support for lsprof is, and this class aims to
3614
fit in with that: while it could be moved it's not necessary to accomplish
3615
test profiling, nor would it be dramatically cleaner.
3618
def startTest(self, test):
3619
self.profiler = bzrlib.lsprof.BzrProfiler()
3620
# Prevent deadlocks in tests that use lsprof: those tests will
3622
bzrlib.lsprof.BzrProfiler.profiler_block = 0
3623
self.profiler.start()
3624
testtools.ExtendedToOriginalDecorator.startTest(self, test)
3626
def addSuccess(self, test):
3627
stats = self.profiler.stop()
3629
calls = test._benchcalls
3630
except AttributeError:
3631
test._benchcalls = []
3632
calls = test._benchcalls
3633
calls.append(((test.id(), "", ""), stats))
3634
testtools.ExtendedToOriginalDecorator.addSuccess(self, test)
3636
def stopTest(self, test):
3637
testtools.ExtendedToOriginalDecorator.stopTest(self, test)
3638
self.profiler = None
3641
# Controlled by "bzr selftest -E=..." option
3642
# Currently supported:
3643
# -Eallow_debug Will no longer clear debug.debug_flags() so it
3644
# preserves any flags supplied at the command line.
3645
# -Edisable_lock_checks Turns errors in mismatched locks into simple prints
3646
# rather than failing tests. And no longer raise
3647
# LockContention when fctnl locks are not being used
3648
# with proper exclusion rules.
3649
# -Ethreads Will display thread ident at creation/join time to
3650
# help track thread leaks
3651
# -Euncollected_cases Display the identity of any test cases that weren't
3652
# deallocated after being completed.
3653
# -Econfig_stats Will collect statistics using addDetail
3654
selftest_debug_flags = set()
3657
def selftest(verbose=False, pattern=".*", stop_on_failure=True,
3659
test_suite_factory=None,
3662
matching_tests_first=None,
3665
exclude_pattern=None,
3671
suite_decorators=None,
3675
"""Run the whole test suite under the enhanced runner"""
3676
# XXX: Very ugly way to do this...
3677
# Disable warning about old formats because we don't want it to disturb
3678
# any blackbox tests.
3679
from bzrlib import repository
3680
repository._deprecation_warning_done = True
3682
global default_transport
3683
if transport is None:
3684
transport = default_transport
3685
old_transport = default_transport
3686
default_transport = transport
3687
global selftest_debug_flags
3688
old_debug_flags = selftest_debug_flags
3689
if debug_flags is not None:
3690
selftest_debug_flags = set(debug_flags)
3692
if load_list is None:
3695
keep_only = load_test_id_list(load_list)
3697
starting_with = [test_prefix_alias_registry.resolve_alias(start)
3698
for start in starting_with]
3699
if test_suite_factory is None:
3700
# Reduce loading time by loading modules based on the starting_with
3702
suite = test_suite(keep_only, starting_with)
3704
suite = test_suite_factory()
3706
# But always filter as requested.
3707
suite = filter_suite_by_id_startswith(suite, starting_with)
3708
result_decorators = []
3710
result_decorators.append(ProfileResult)
3711
return run_suite(suite, 'testbzr', verbose=verbose, pattern=pattern,
3712
stop_on_failure=stop_on_failure,
3713
transport=transport,
3714
lsprof_timed=lsprof_timed,
3715
bench_history=bench_history,
3716
matching_tests_first=matching_tests_first,
3717
list_only=list_only,
3718
random_seed=random_seed,
3719
exclude_pattern=exclude_pattern,
3721
runner_class=runner_class,
3722
suite_decorators=suite_decorators,
3724
result_decorators=result_decorators,
3727
default_transport = old_transport
3728
selftest_debug_flags = old_debug_flags
3731
def load_test_id_list(file_name):
3732
"""Load a test id list from a text file.
3734
The format is one test id by line. No special care is taken to impose
3735
strict rules, these test ids are used to filter the test suite so a test id
3736
that do not match an existing test will do no harm. This allows user to add
3737
comments, leave blank lines, etc.
3741
ftest = open(file_name, 'rt')
3743
if e.errno != errno.ENOENT:
3746
raise errors.NoSuchFile(file_name)
3748
for test_name in ftest.readlines():
3749
test_list.append(test_name.strip())
3754
def suite_matches_id_list(test_suite, id_list):
3755
"""Warns about tests not appearing or appearing more than once.
3757
:param test_suite: A TestSuite object.
3758
:param test_id_list: The list of test ids that should be found in
3761
:return: (absents, duplicates) absents is a list containing the test found
3762
in id_list but not in test_suite, duplicates is a list containing the
3763
test found multiple times in test_suite.
3765
When using a prefined test id list, it may occurs that some tests do not
3766
exist anymore or that some tests use the same id. This function warns the
3767
tester about potential problems in his workflow (test lists are volatile)
3768
or in the test suite itself (using the same id for several tests does not
3769
help to localize defects).
3771
# Build a dict counting id occurrences
3773
for test in iter_suite_tests(test_suite):
3775
tests[id] = tests.get(id, 0) + 1
3780
occurs = tests.get(id, 0)
3782
not_found.append(id)
3784
duplicates.append(id)
3786
return not_found, duplicates
3789
class TestIdList(object):
3790
"""Test id list to filter a test suite.
3792
Relying on the assumption that test ids are built as:
3793
<module>[.<class>.<method>][(<param>+)], <module> being in python dotted
3794
notation, this class offers methods to :
3795
- avoid building a test suite for modules not refered to in the test list,
3796
- keep only the tests listed from the module test suite.
3799
def __init__(self, test_id_list):
3800
# When a test suite needs to be filtered against us we compare test ids
3801
# for equality, so a simple dict offers a quick and simple solution.
3802
self.tests = dict().fromkeys(test_id_list, True)
3804
# While unittest.TestCase have ids like:
3805
# <module>.<class>.<method>[(<param+)],
3806
# doctest.DocTestCase can have ids like:
3809
# <module>.<function>
3810
# <module>.<class>.<method>
3812
# Since we can't predict a test class from its name only, we settle on
3813
# a simple constraint: a test id always begins with its module name.
3816
for test_id in test_id_list:
3817
parts = test_id.split('.')
3818
mod_name = parts.pop(0)
3819
modules[mod_name] = True
3821
mod_name += '.' + part
3822
modules[mod_name] = True
3823
self.modules = modules
3825
def refers_to(self, module_name):
3826
"""Is there tests for the module or one of its sub modules."""
3827
return self.modules.has_key(module_name)
3829
def includes(self, test_id):
3830
return self.tests.has_key(test_id)
3833
class TestPrefixAliasRegistry(registry.Registry):
3834
"""A registry for test prefix aliases.
3836
This helps implement shorcuts for the --starting-with selftest
3837
option. Overriding existing prefixes is not allowed but not fatal (a
3838
warning will be emitted).
3841
def register(self, key, obj, help=None, info=None,
3842
override_existing=False):
3843
"""See Registry.register.
3845
Trying to override an existing alias causes a warning to be emitted,
3846
not a fatal execption.
3849
super(TestPrefixAliasRegistry, self).register(
3850
key, obj, help=help, info=info, override_existing=False)
3852
actual = self.get(key)
3854
'Test prefix alias %s is already used for %s, ignoring %s'
3855
% (key, actual, obj))
3857
def resolve_alias(self, id_start):
3858
"""Replace the alias by the prefix in the given string.
3860
Using an unknown prefix is an error to help catching typos.
3862
parts = id_start.split('.')
3864
parts[0] = self.get(parts[0])
3866
raise errors.BzrCommandError(
3867
'%s is not a known test prefix alias' % parts[0])
3868
return '.'.join(parts)
3871
test_prefix_alias_registry = TestPrefixAliasRegistry()
3872
"""Registry of test prefix aliases."""
3875
# This alias allows to detect typos ('bzrlin.') by making all valid test ids
3876
# appear prefixed ('bzrlib.' is "replaced" by 'bzrlib.').
3877
test_prefix_alias_registry.register('bzrlib', 'bzrlib')
3879
# Obvious highest levels prefixes, feel free to add your own via a plugin
3880
test_prefix_alias_registry.register('bd', 'bzrlib.doc')
3881
test_prefix_alias_registry.register('bu', 'bzrlib.utils')
3882
test_prefix_alias_registry.register('bt', 'bzrlib.tests')
3883
test_prefix_alias_registry.register('bb', 'bzrlib.tests.blackbox')
3884
test_prefix_alias_registry.register('bp', 'bzrlib.plugins')
3887
def _test_suite_testmod_names():
3888
"""Return the standard list of test module names to test."""
3891
'bzrlib.tests.blackbox',
3892
'bzrlib.tests.commands',
3893
'bzrlib.tests.doc_generate',
3894
'bzrlib.tests.per_branch',
3895
'bzrlib.tests.per_bzrdir',
3896
'bzrlib.tests.per_controldir',
3897
'bzrlib.tests.per_controldir_colo',
3898
'bzrlib.tests.per_foreign_vcs',
3899
'bzrlib.tests.per_interrepository',
3900
'bzrlib.tests.per_intertree',
3901
'bzrlib.tests.per_inventory',
3902
'bzrlib.tests.per_interbranch',
3903
'bzrlib.tests.per_lock',
3904
'bzrlib.tests.per_merger',
3905
'bzrlib.tests.per_transport',
3906
'bzrlib.tests.per_tree',
3907
'bzrlib.tests.per_pack_repository',
3908
'bzrlib.tests.per_repository',
3909
'bzrlib.tests.per_repository_chk',
3910
'bzrlib.tests.per_repository_reference',
3911
'bzrlib.tests.per_repository_vf',
3912
'bzrlib.tests.per_uifactory',
3913
'bzrlib.tests.per_versionedfile',
3914
'bzrlib.tests.per_workingtree',
3915
'bzrlib.tests.test__annotator',
3916
'bzrlib.tests.test__bencode',
3917
'bzrlib.tests.test__btree_serializer',
3918
'bzrlib.tests.test__chk_map',
3919
'bzrlib.tests.test__dirstate_helpers',
3920
'bzrlib.tests.test__groupcompress',
3921
'bzrlib.tests.test__known_graph',
3922
'bzrlib.tests.test__rio',
3923
'bzrlib.tests.test__simple_set',
3924
'bzrlib.tests.test__static_tuple',
3925
'bzrlib.tests.test__walkdirs_win32',
3926
'bzrlib.tests.test_ancestry',
3927
'bzrlib.tests.test_annotate',
3928
'bzrlib.tests.test_api',
3929
'bzrlib.tests.test_atomicfile',
3930
'bzrlib.tests.test_bad_files',
3931
'bzrlib.tests.test_bisect_multi',
3932
'bzrlib.tests.test_branch',
3933
'bzrlib.tests.test_branchbuilder',
3934
'bzrlib.tests.test_btree_index',
3935
'bzrlib.tests.test_bugtracker',
3936
'bzrlib.tests.test_bundle',
3937
'bzrlib.tests.test_bzrdir',
3938
'bzrlib.tests.test__chunks_to_lines',
3939
'bzrlib.tests.test_cache_utf8',
3940
'bzrlib.tests.test_chk_map',
3941
'bzrlib.tests.test_chk_serializer',
3942
'bzrlib.tests.test_chunk_writer',
3943
'bzrlib.tests.test_clean_tree',
3944
'bzrlib.tests.test_cleanup',
3945
'bzrlib.tests.test_cmdline',
3946
'bzrlib.tests.test_commands',
3947
'bzrlib.tests.test_commit',
3948
'bzrlib.tests.test_commit_merge',
3949
'bzrlib.tests.test_config',
3950
'bzrlib.tests.test_conflicts',
3951
'bzrlib.tests.test_controldir',
3952
'bzrlib.tests.test_counted_lock',
3953
'bzrlib.tests.test_crash',
3954
'bzrlib.tests.test_decorators',
3955
'bzrlib.tests.test_delta',
3956
'bzrlib.tests.test_debug',
3957
'bzrlib.tests.test_diff',
3958
'bzrlib.tests.test_directory_service',
3959
'bzrlib.tests.test_dirstate',
3960
'bzrlib.tests.test_email_message',
3961
'bzrlib.tests.test_eol_filters',
3962
'bzrlib.tests.test_errors',
3963
'bzrlib.tests.test_estimate_compressed_size',
3964
'bzrlib.tests.test_export',
3965
'bzrlib.tests.test_export_pot',
3966
'bzrlib.tests.test_extract',
3967
'bzrlib.tests.test_features',
3968
'bzrlib.tests.test_fetch',
3969
'bzrlib.tests.test_fixtures',
3970
'bzrlib.tests.test_fifo_cache',
3971
'bzrlib.tests.test_filters',
3972
'bzrlib.tests.test_filter_tree',
3973
'bzrlib.tests.test_ftp_transport',
3974
'bzrlib.tests.test_foreign',
3975
'bzrlib.tests.test_generate_docs',
3976
'bzrlib.tests.test_generate_ids',
3977
'bzrlib.tests.test_globbing',
3978
'bzrlib.tests.test_gpg',
3979
'bzrlib.tests.test_graph',
3980
'bzrlib.tests.test_groupcompress',
3981
'bzrlib.tests.test_hashcache',
3982
'bzrlib.tests.test_help',
3983
'bzrlib.tests.test_hooks',
3984
'bzrlib.tests.test_http',
3985
'bzrlib.tests.test_http_response',
3986
'bzrlib.tests.test_https_ca_bundle',
3987
'bzrlib.tests.test_i18n',
3988
'bzrlib.tests.test_identitymap',
3989
'bzrlib.tests.test_ignores',
3990
'bzrlib.tests.test_index',
3991
'bzrlib.tests.test_import_tariff',
3992
'bzrlib.tests.test_info',
3993
'bzrlib.tests.test_inv',
3994
'bzrlib.tests.test_inventory_delta',
3995
'bzrlib.tests.test_knit',
3996
'bzrlib.tests.test_lazy_import',
3997
'bzrlib.tests.test_lazy_regex',
3998
'bzrlib.tests.test_library_state',
3999
'bzrlib.tests.test_lock',
4000
'bzrlib.tests.test_lockable_files',
4001
'bzrlib.tests.test_lockdir',
4002
'bzrlib.tests.test_log',
4003
'bzrlib.tests.test_lru_cache',
4004
'bzrlib.tests.test_lsprof',
4005
'bzrlib.tests.test_mail_client',
4006
'bzrlib.tests.test_matchers',
4007
'bzrlib.tests.test_memorytree',
4008
'bzrlib.tests.test_merge',
4009
'bzrlib.tests.test_merge3',
4010
'bzrlib.tests.test_merge_core',
4011
'bzrlib.tests.test_merge_directive',
4012
'bzrlib.tests.test_mergetools',
4013
'bzrlib.tests.test_missing',
4014
'bzrlib.tests.test_msgeditor',
4015
'bzrlib.tests.test_multiparent',
4016
'bzrlib.tests.test_mutabletree',
4017
'bzrlib.tests.test_nonascii',
4018
'bzrlib.tests.test_options',
4019
'bzrlib.tests.test_osutils',
4020
'bzrlib.tests.test_osutils_encodings',
4021
'bzrlib.tests.test_pack',
4022
'bzrlib.tests.test_patch',
4023
'bzrlib.tests.test_patches',
4024
'bzrlib.tests.test_permissions',
4025
'bzrlib.tests.test_plugins',
4026
'bzrlib.tests.test_progress',
4027
'bzrlib.tests.test_pyutils',
4028
'bzrlib.tests.test_read_bundle',
4029
'bzrlib.tests.test_reconcile',
4030
'bzrlib.tests.test_reconfigure',
4031
'bzrlib.tests.test_registry',
4032
'bzrlib.tests.test_remote',
4033
'bzrlib.tests.test_rename_map',
4034
'bzrlib.tests.test_repository',
4035
'bzrlib.tests.test_revert',
4036
'bzrlib.tests.test_revision',
4037
'bzrlib.tests.test_revisionspec',
4038
'bzrlib.tests.test_revisiontree',
4039
'bzrlib.tests.test_rio',
4040
'bzrlib.tests.test_rules',
4041
'bzrlib.tests.test_sampler',
4042
'bzrlib.tests.test_scenarios',
4043
'bzrlib.tests.test_script',
4044
'bzrlib.tests.test_selftest',
4045
'bzrlib.tests.test_serializer',
4046
'bzrlib.tests.test_setup',
4047
'bzrlib.tests.test_sftp_transport',
4048
'bzrlib.tests.test_shelf',
4049
'bzrlib.tests.test_shelf_ui',
4050
'bzrlib.tests.test_smart',
4051
'bzrlib.tests.test_smart_add',
4052
'bzrlib.tests.test_smart_request',
4053
'bzrlib.tests.test_smart_signals',
4054
'bzrlib.tests.test_smart_transport',
4055
'bzrlib.tests.test_smtp_connection',
4056
'bzrlib.tests.test_source',
4057
'bzrlib.tests.test_ssh_transport',
4058
'bzrlib.tests.test_status',
4059
'bzrlib.tests.test_store',
4060
'bzrlib.tests.test_strace',
4061
'bzrlib.tests.test_subsume',
4062
'bzrlib.tests.test_switch',
4063
'bzrlib.tests.test_symbol_versioning',
4064
'bzrlib.tests.test_tag',
4065
'bzrlib.tests.test_test_server',
4066
'bzrlib.tests.test_testament',
4067
'bzrlib.tests.test_textfile',
4068
'bzrlib.tests.test_textmerge',
4069
'bzrlib.tests.test_cethread',
4070
'bzrlib.tests.test_timestamp',
4071
'bzrlib.tests.test_trace',
4072
'bzrlib.tests.test_transactions',
4073
'bzrlib.tests.test_transform',
4074
'bzrlib.tests.test_transport',
4075
'bzrlib.tests.test_transport_log',
4076
'bzrlib.tests.test_tree',
4077
'bzrlib.tests.test_treebuilder',
4078
'bzrlib.tests.test_treeshape',
4079
'bzrlib.tests.test_tsort',
4080
'bzrlib.tests.test_tuned_gzip',
4081
'bzrlib.tests.test_ui',
4082
'bzrlib.tests.test_uncommit',
4083
'bzrlib.tests.test_upgrade',
4084
'bzrlib.tests.test_upgrade_stacked',
4085
'bzrlib.tests.test_urlutils',
4086
'bzrlib.tests.test_utextwrap',
4087
'bzrlib.tests.test_version',
4088
'bzrlib.tests.test_version_info',
4089
'bzrlib.tests.test_versionedfile',
4090
'bzrlib.tests.test_weave',
4091
'bzrlib.tests.test_whitebox',
4092
'bzrlib.tests.test_win32utils',
4093
'bzrlib.tests.test_workingtree',
4094
'bzrlib.tests.test_workingtree_4',
4095
'bzrlib.tests.test_wsgi',
4096
'bzrlib.tests.test_xml',
4100
def _test_suite_modules_to_doctest():
4101
"""Return the list of modules to doctest."""
4103
# GZ 2009-03-31: No docstrings with -OO so there's nothing to doctest
4107
'bzrlib.branchbuilder',
4108
'bzrlib.decorators',
4110
'bzrlib.iterablefile',
4115
'bzrlib.symbol_versioning',
4117
'bzrlib.tests.fixtures',
4119
'bzrlib.transport.http',
4120
'bzrlib.version_info_formats.format_custom',
4124
def test_suite(keep_only=None, starting_with=None):
4125
"""Build and return TestSuite for the whole of bzrlib.
4127
:param keep_only: A list of test ids limiting the suite returned.
4129
:param starting_with: An id limiting the suite returned to the tests
4132
This function can be replaced if you need to change the default test
4133
suite on a global basis, but it is not encouraged.
4136
loader = TestUtil.TestLoader()
4138
if keep_only is not None:
4139
id_filter = TestIdList(keep_only)
4141
# We take precedence over keep_only because *at loading time* using
4142
# both options means we will load less tests for the same final result.
4143
def interesting_module(name):
4144
for start in starting_with:
4146
# Either the module name starts with the specified string
4147
name.startswith(start)
4148
# or it may contain tests starting with the specified string
4149
or start.startswith(name)
4153
loader = TestUtil.FilteredByModuleTestLoader(interesting_module)
4155
elif keep_only is not None:
4156
loader = TestUtil.FilteredByModuleTestLoader(id_filter.refers_to)
4157
def interesting_module(name):
4158
return id_filter.refers_to(name)
4161
loader = TestUtil.TestLoader()
4162
def interesting_module(name):
4163
# No filtering, all modules are interesting
4166
suite = loader.suiteClass()
4168
# modules building their suite with loadTestsFromModuleNames
4169
suite.addTest(loader.loadTestsFromModuleNames(_test_suite_testmod_names()))
4171
for mod in _test_suite_modules_to_doctest():
4172
if not interesting_module(mod):
4173
# No tests to keep here, move along
4176
# note that this really does mean "report only" -- doctest
4177
# still runs the rest of the examples
4178
doc_suite = IsolatedDocTestSuite(
4179
mod, optionflags=doctest.REPORT_ONLY_FIRST_FAILURE)
4180
except ValueError, e:
4181
print '**failed to get doctest for: %s\n%s' % (mod, e)
4183
if len(doc_suite._tests) == 0:
4184
raise errors.BzrError("no doctests found in %s" % (mod,))
4185
suite.addTest(doc_suite)
4187
default_encoding = sys.getdefaultencoding()
4188
for name, plugin in _mod_plugin.plugins().items():
4189
if not interesting_module(plugin.module.__name__):
4191
plugin_suite = plugin.test_suite()
4192
# We used to catch ImportError here and turn it into just a warning,
4193
# but really if you don't have --no-plugins this should be a failure.
4194
# mbp 20080213 - see http://bugs.launchpad.net/bugs/189771
4195
if plugin_suite is None:
4196
plugin_suite = plugin.load_plugin_tests(loader)
4197
if plugin_suite is not None:
4198
suite.addTest(plugin_suite)
4199
if default_encoding != sys.getdefaultencoding():
4201
'Plugin "%s" tried to reset default encoding to: %s', name,
4202
sys.getdefaultencoding())
4204
sys.setdefaultencoding(default_encoding)
4206
if keep_only is not None:
4207
# Now that the referred modules have loaded their tests, keep only the
4209
suite = filter_suite_by_id_list(suite, id_filter)
4210
# Do some sanity checks on the id_list filtering
4211
not_found, duplicates = suite_matches_id_list(suite, keep_only)
4213
# The tester has used both keep_only and starting_with, so he is
4214
# already aware that some tests are excluded from the list, there
4215
# is no need to tell him which.
4218
# Some tests mentioned in the list are not in the test suite. The
4219
# list may be out of date, report to the tester.
4220
for id in not_found:
4221
trace.warning('"%s" not found in the test suite', id)
4222
for id in duplicates:
4223
trace.warning('"%s" is used as an id by several tests', id)
4228
def multiply_scenarios(*scenarios):
4229
"""Multiply two or more iterables of scenarios.
4231
It is safe to pass scenario generators or iterators.
4233
:returns: A list of compound scenarios: the cross-product of all
4234
scenarios, with the names concatenated and the parameters
4237
return reduce(_multiply_two_scenarios, map(list, scenarios))
4240
def _multiply_two_scenarios(scenarios_left, scenarios_right):
4241
"""Multiply two sets of scenarios.
4243
:returns: the cartesian product of the two sets of scenarios, that is
4244
a scenario for every possible combination of a left scenario and a
4248
('%s,%s' % (left_name, right_name),
4249
dict(left_dict.items() + right_dict.items()))
4250
for left_name, left_dict in scenarios_left
4251
for right_name, right_dict in scenarios_right]
4254
def multiply_tests(tests, scenarios, result):
4255
"""Multiply tests_list by scenarios into result.
4257
This is the core workhorse for test parameterisation.
4259
Typically the load_tests() method for a per-implementation test suite will
4260
call multiply_tests and return the result.
4262
:param tests: The tests to parameterise.
4263
:param scenarios: The scenarios to apply: pairs of (scenario_name,
4264
scenario_param_dict).
4265
:param result: A TestSuite to add created tests to.
4267
This returns the passed in result TestSuite with the cross product of all
4268
the tests repeated once for each scenario. Each test is adapted by adding
4269
the scenario name at the end of its id(), and updating the test object's
4270
__dict__ with the scenario_param_dict.
4272
>>> import bzrlib.tests.test_sampler
4273
>>> r = multiply_tests(
4274
... bzrlib.tests.test_sampler.DemoTest('test_nothing'),
4275
... [('one', dict(param=1)),
4276
... ('two', dict(param=2))],
4277
... TestUtil.TestSuite())
4278
>>> tests = list(iter_suite_tests(r))
4282
'bzrlib.tests.test_sampler.DemoTest.test_nothing(one)'
4288
for test in iter_suite_tests(tests):
4289
apply_scenarios(test, scenarios, result)
4293
def apply_scenarios(test, scenarios, result):
4294
"""Apply the scenarios in scenarios to test and add to result.
4296
:param test: The test to apply scenarios to.
4297
:param scenarios: An iterable of scenarios to apply to test.
4299
:seealso: apply_scenario
4301
for scenario in scenarios:
4302
result.addTest(apply_scenario(test, scenario))
4306
def apply_scenario(test, scenario):
4307
"""Copy test and apply scenario to it.
4309
:param test: A test to adapt.
4310
:param scenario: A tuple describing the scenarion.
4311
The first element of the tuple is the new test id.
4312
The second element is a dict containing attributes to set on the
4314
:return: The adapted test.
4316
new_id = "%s(%s)" % (test.id(), scenario[0])
4317
new_test = clone_test(test, new_id)
4318
for name, value in scenario[1].items():
4319
setattr(new_test, name, value)
4323
def clone_test(test, new_id):
4324
"""Clone a test giving it a new id.
4326
:param test: The test to clone.
4327
:param new_id: The id to assign to it.
4328
:return: The new test.
4330
new_test = copy.copy(test)
4331
new_test.id = lambda: new_id
4332
# XXX: Workaround <https://bugs.launchpad.net/testtools/+bug/637725>, which
4333
# causes cloned tests to share the 'details' dict. This makes it hard to
4334
# read the test output for parameterized tests, because tracebacks will be
4335
# associated with irrelevant tests.
4337
details = new_test._TestCase__details
4338
except AttributeError:
4339
# must be a different version of testtools than expected. Do nothing.
4342
# Reset the '__details' dict.
4343
new_test._TestCase__details = {}
4347
def permute_tests_for_extension(standard_tests, loader, py_module_name,
4349
"""Helper for permutating tests against an extension module.
4351
This is meant to be used inside a modules 'load_tests()' function. It will
4352
create 2 scenarios, and cause all tests in the 'standard_tests' to be run
4353
against both implementations. Setting 'test.module' to the appropriate
4354
module. See bzrlib.tests.test__chk_map.load_tests as an example.
4356
:param standard_tests: A test suite to permute
4357
:param loader: A TestLoader
4358
:param py_module_name: The python path to a python module that can always
4359
be loaded, and will be considered the 'python' implementation. (eg
4360
'bzrlib._chk_map_py')
4361
:param ext_module_name: The python path to an extension module. If the
4362
module cannot be loaded, a single test will be added, which notes that
4363
the module is not available. If it can be loaded, all standard_tests
4364
will be run against that module.
4365
:return: (suite, feature) suite is a test-suite that has all the permuted
4366
tests. feature is the Feature object that can be used to determine if
4367
the module is available.
4370
from bzrlib.tests.features import ModuleAvailableFeature
4371
py_module = pyutils.get_named_object(py_module_name)
4373
('python', {'module': py_module}),
4375
suite = loader.suiteClass()
4376
feature = ModuleAvailableFeature(ext_module_name)
4377
if feature.available():
4378
scenarios.append(('C', {'module': feature.module}))
4380
# the compiled module isn't available, so we add a failing test
4381
class FailWithoutFeature(TestCase):
4382
def test_fail(self):
4383
self.requireFeature(feature)
4384
suite.addTest(loader.loadTestsFromTestCase(FailWithoutFeature))
4385
result = multiply_tests(standard_tests, scenarios, suite)
4386
return result, feature
4389
def _rmtree_temp_dir(dirname, test_id=None):
4390
# If LANG=C we probably have created some bogus paths
4391
# which rmtree(unicode) will fail to delete
4392
# so make sure we are using rmtree(str) to delete everything
4393
# except on win32, where rmtree(str) will fail
4394
# since it doesn't have the property of byte-stream paths
4395
# (they are either ascii or mbcs)
4396
if sys.platform == 'win32':
4397
# make sure we are using the unicode win32 api
4398
dirname = unicode(dirname)
4400
dirname = dirname.encode(sys.getfilesystemencoding())
4402
osutils.rmtree(dirname)
4404
# We don't want to fail here because some useful display will be lost
4405
# otherwise. Polluting the tmp dir is bad, but not giving all the
4406
# possible info to the test runner is even worse.
4408
ui.ui_factory.clear_term()
4409
sys.stderr.write('\nWhile running: %s\n' % (test_id,))
4410
# Ugly, but the last thing we want here is fail, so bear with it.
4411
printable_e = str(e).decode(osutils.get_user_encoding(), 'replace'
4412
).encode('ascii', 'replace')
4413
sys.stderr.write('Unable to remove testing dir %s\n%s'
4414
% (os.path.basename(dirname), printable_e))
4417
def probe_unicode_in_user_encoding():
4418
"""Try to encode several unicode strings to use in unicode-aware tests.
4419
Return first successfull match.
4421
:return: (unicode value, encoded plain string value) or (None, None)
4423
possible_vals = [u'm\xb5', u'\xe1', u'\u0410']
4424
for uni_val in possible_vals:
4426
str_val = uni_val.encode(osutils.get_user_encoding())
4427
except UnicodeEncodeError:
4428
# Try a different character
4431
return uni_val, str_val
4435
def probe_bad_non_ascii(encoding):
4436
"""Try to find [bad] character with code [128..255]
4437
that cannot be decoded to unicode in some encoding.
4438
Return None if all non-ascii characters is valid
4441
for i in xrange(128, 256):
4444
char.decode(encoding)
4445
except UnicodeDecodeError:
4450
# Only define SubUnitBzrRunner if subunit is available.
4452
from subunit import TestProtocolClient
4453
from subunit.test_results import AutoTimingTestResultDecorator
4454
class SubUnitBzrProtocolClient(TestProtocolClient):
4456
def stopTest(self, test):
4457
super(SubUnitBzrProtocolClient, self).stopTest(test)
4458
_clear__type_equality_funcs(test)
4460
def addSuccess(self, test, details=None):
4461
# The subunit client always includes the details in the subunit
4462
# stream, but we don't want to include it in ours.
4463
if details is not None and 'log' in details:
4465
return super(SubUnitBzrProtocolClient, self).addSuccess(
4468
class SubUnitBzrRunner(TextTestRunner):
4469
def run(self, test):
4470
result = AutoTimingTestResultDecorator(
4471
SubUnitBzrProtocolClient(self.stream))
4478
# API compatibility for old plugins; see bug 892622.
4481
'HTTPServerFeature',
4482
'ModuleAvailableFeature',
4483
'HTTPSServerFeature', 'SymlinkFeature', 'HardlinkFeature',
4484
'OsFifoFeature', 'UnicodeFilenameFeature',
4485
'ByteStringNamedFilesystem', 'UTF8Filesystem',
4486
'BreakinFeature', 'CaseInsCasePresFilenameFeature',
4487
'CaseInsensitiveFilesystemFeature', 'case_sensitive_filesystem_feature',
4488
'posix_permissions_feature',
4490
globals()[name] = _CompatabilityThunkFeature(
4491
symbol_versioning.deprecated_in((2, 5, 0)),
4492
'bzrlib.tests', name,
4493
name, 'bzrlib.tests.features')
4496
for (old_name, new_name) in [
4497
('UnicodeFilename', 'UnicodeFilenameFeature'),
4499
globals()[name] = _CompatabilityThunkFeature(
4500
symbol_versioning.deprecated_in((2, 5, 0)),
4501
'bzrlib.tests', old_name,
4502
new_name, 'bzrlib.tests.features')
added
removed
bzrlib/tests/__init__.py
