1
# Copyright (C) 2005-2013, 2015, 2016 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 breezy. Please continue that consistency when adding e.g.
22
# new assertFoo() methods.
57
# nb: check this before importing anything else from within it
58
_testtools_version = getattr(testtools, '__version__', ())
59
if _testtools_version < (0, 9, 5):
60
raise ImportError("need at least testtools 0.9.5: %s is %r"
61
% (testtools.__file__, _testtools_version))
62
from testtools import content
68
commands as _mod_commands,
77
plugin as _mod_plugin,
84
transport as _mod_transport,
87
from breezy.bzr import (
93
# lsprof not available
95
from ..bzr.smart import client, request
96
from ..transport import (
100
from ..tests import (
108
# Mark this python module as being part of the implementation
109
# of unittest: this gives us better tracebacks where the last
110
# shown frame is the test code, not our assertXYZ.
113
default_transport = test_server.LocalURLServer
116
_unitialized_attr = object()
117
"""A sentinel needed to act as a default value in a method signature."""
120
# Subunit result codes, defined here to prevent a hard dependency on subunit.
124
# These are intentionally brought into this namespace. That way plugins, etc
125
# can just "from breezy.tests import TestCase, TestLoader, etc"
126
TestSuite = TestUtil.TestSuite
127
TestLoader = TestUtil.TestLoader
129
# Tests should run in a clean and clearly defined environment. The goal is to
130
# keep them isolated from the running environment as mush as possible. The test
131
# framework ensures the variables defined below are set (or deleted if the
132
# value is None) before a test is run and reset to their original value after
133
# the test is run. Generally if some code depends on an environment variable,
134
# the tests should start without this variable in the environment. There are a
135
# few exceptions but you shouldn't violate this rule lightly.
140
'XDG_CONFIG_HOME': None,
141
# brz now uses the Win32 API and doesn't rely on APPDATA, but the
142
# tests do check our impls match APPDATA
143
'BRZ_EDITOR': None, # test_msgeditor manipulates this variable
147
'BZREMAIL': None, # may still be present in the environment
148
'EMAIL': 'jrandom@example.com', # set EMAIL as brz does not guess
149
'BRZ_PROGRESS_BAR': None,
150
# Trap leaks to $XDG_CACHE_HOME/breezy/brz.log. This occurs when tests use
151
# TestCase as a base class instead of TestCaseInTempDir. Tests inheriting
152
# from TestCase should not use disk resources, BRZ_LOG is one.
153
'BRZ_LOG': '/you-should-use-TestCaseInTempDir-if-you-need-a-log-file',
154
'BRZ_PLUGIN_PATH': '-site',
155
'BRZ_DISABLE_PLUGINS': None,
156
'BRZ_PLUGINS_AT': None,
157
'BRZ_CONCURRENCY': None,
158
# Make sure that any text ui tests are consistent regardless of
159
# the environment the test case is run in; you may want tests that
160
# test other combinations. 'dumb' is a reasonable guess for tests
161
# going to a pipe or a BytesIO.
167
'SSH_AUTH_SOCK': None,
177
'BZR_REMOTE_PATH': None,
178
# Generally speaking, we don't want apport reporting on crashes in
179
# the test envirnoment unless we're specifically testing apport,
180
# so that it doesn't leak into the real system environment. We
181
# use an env var so it propagates to subprocesses.
182
'APPORT_DISABLE': '1',
186
def override_os_environ(test, env=None):
187
"""Modify os.environ keeping a copy.
189
:param test: A test instance
191
:param env: A dict containing variable definitions to be installed
194
env = isolated_environ
195
test._original_os_environ = dict(**os.environ)
197
osutils.set_or_unset_env(var, env[var])
198
if var not in test._original_os_environ:
199
# The var is new, add it with a value of None, so
200
# restore_os_environ will delete it
201
test._original_os_environ[var] = None
204
def restore_os_environ(test):
205
"""Restore os.environ to its original state.
207
:param test: A test instance previously passed to override_os_environ.
209
for var, value in test._original_os_environ.items():
210
# Restore the original value (or delete it if the value has been set to
211
# None in override_os_environ).
212
osutils.set_or_unset_env(var, value)
215
def _clear__type_equality_funcs(test):
216
"""Cleanup bound methods stored on TestCase instances
218
Clear the dict breaking a few (mostly) harmless cycles in the affected
219
unittests released with Python 2.6 and initial Python 2.7 versions.
221
For a few revisions between Python 2.7.1 and Python 2.7.2 that annoyingly
222
shipped in Oneiric, an object with no clear method was used, hence the
223
extra complications, see bug 809048 for details.
225
type_equality_funcs = getattr(test, "_type_equality_funcs", None)
226
if type_equality_funcs is not None:
227
tef_clear = getattr(type_equality_funcs, "clear", None)
228
if tef_clear is None:
229
tef_instance_dict = getattr(type_equality_funcs, "__dict__", None)
230
if tef_instance_dict is not None:
231
tef_clear = tef_instance_dict.clear
232
if tef_clear is not None:
236
class ExtendedTestResult(testtools.TextTestResult):
237
"""Accepts, reports and accumulates the results of running tests.
239
Compared to the unittest version this class adds support for
240
profiling, benchmarking, stopping as soon as a test fails, and
241
skipping tests. There are further-specialized subclasses for
242
different types of display.
244
When a test finishes, in whatever way, it calls one of the addSuccess,
245
addFailure or addError methods. These in turn may redirect to a more
246
specific case for the special test results supported by our extended
249
Note that just one of these objects is fed the results from many tests.
254
def __init__(self, stream, descriptions, verbosity,
258
"""Construct new TestResult.
260
:param bench_history: Optionally, a writable file object to accumulate
263
testtools.TextTestResult.__init__(self, stream)
264
if bench_history is not None:
265
from breezy.version import _get_bzr_source_tree
266
src_tree = _get_bzr_source_tree()
269
revision_id = src_tree.get_parent_ids()[0]
271
# XXX: if this is a brand new tree, do the same as if there
275
# XXX: If there's no branch, what should we do?
277
bench_history.write("--date %s %s\n" % (time.time(), revision_id))
278
self._bench_history = bench_history
279
self.ui = ui.ui_factory
282
self.failure_count = 0
283
self.known_failure_count = 0
285
self.not_applicable_count = 0
286
self.unsupported = {}
288
self._overall_start_time = time.time()
289
self._strict = strict
290
self._first_thread_leaker_id = None
291
self._tests_leaking_threads_count = 0
292
self._traceback_from_test = None
294
def stopTestRun(self):
297
stopTime = time.time()
298
timeTaken = stopTime - self.startTime
299
# GZ 2010-07-19: Seems testtools has no printErrors method, and though
300
# the parent class method is similar have to duplicate
301
self._show_list('ERROR', self.errors)
302
self._show_list('FAIL', self.failures)
303
self.stream.write(self.sep2)
304
self.stream.write("%s %d test%s in %.3fs\n\n" % (
305
actionTaken, run, run != 1 and "s" or "", timeTaken))
306
if not self.wasSuccessful():
307
self.stream.write("FAILED (")
308
failed, errored = map(len, (self.failures, self.errors))
310
self.stream.write("failures=%d" % failed)
313
self.stream.write(", ")
314
self.stream.write("errors=%d" % errored)
315
if self.known_failure_count:
316
if failed or errored:
317
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 _delta_to_float(self, a_timedelta, precision):
359
# This calls ceiling to ensure that the most pessimistic view of time
360
# taken is shown (rather than leaving it to the Python %f operator
361
# to decide whether to round/floor/ceiling. This was added when we
362
# had pyp3 test failures that suggest a floor was happening.
363
shift = 10 ** precision
365
(a_timedelta.days * 86400.0 + a_timedelta.seconds +
366
a_timedelta.microseconds / 1000000.0) * shift) / shift
368
def _elapsedTestTimeString(self):
369
"""Return time string for overall time the current test has taken."""
370
return self._formatTime(self._delta_to_float(
371
self._now() - self._start_datetime, 3))
373
def _testTimeString(self, testCase):
374
benchmark_time = self._extractBenchmarkTime(testCase)
375
if benchmark_time is not None:
376
return self._formatTime(benchmark_time) + "*"
378
return self._elapsedTestTimeString()
380
def _formatTime(self, seconds):
381
"""Format seconds as milliseconds with leading spaces."""
382
# some benchmarks can take thousands of seconds to run, so we need 8
384
return "%8dms" % (1000 * seconds)
386
def _shortened_test_description(self, test):
388
what = re.sub(r'^breezy\.tests\.', '', what)
391
# GZ 2010-10-04: Cloned tests may end up harmlessly calling this method
392
# multiple times in a row, because the handler is added for
393
# each test but the container list is shared between cases.
394
# See lp:498869 lp:625574 and lp:637725 for background.
395
def _record_traceback_from_test(self, exc_info):
396
"""Store the traceback from passed exc_info tuple till"""
397
self._traceback_from_test = exc_info[2]
399
def startTest(self, test):
400
super(ExtendedTestResult, self).startTest(test)
404
self.report_test_start(test)
405
test.number = self.count
406
self._recordTestStartTime()
407
# Make testtools cases give us the real traceback on failure
408
addOnException = getattr(test, "addOnException", None)
409
if addOnException is not None:
410
addOnException(self._record_traceback_from_test)
411
# Only check for thread leaks on breezy derived test cases
412
if isinstance(test, TestCase):
413
test.addCleanup(self._check_leaked_threads, test)
415
def stopTest(self, test):
416
super(ExtendedTestResult, self).stopTest(test)
417
# Manually break cycles, means touching various private things but hey
418
getDetails = getattr(test, "getDetails", None)
419
if getDetails is not None:
421
_clear__type_equality_funcs(test)
422
self._traceback_from_test = None
424
def startTests(self):
425
self.report_tests_starting()
426
self._active_threads = threading.enumerate()
428
def _check_leaked_threads(self, test):
429
"""See if any threads have leaked since last call
431
A sample of live threads is stored in the _active_threads attribute,
432
when this method runs it compares the current live threads and any not
433
in the previous sample are treated as having leaked.
435
now_active_threads = set(threading.enumerate())
436
threads_leaked = now_active_threads.difference(self._active_threads)
438
self._report_thread_leak(test, threads_leaked, now_active_threads)
439
self._tests_leaking_threads_count += 1
440
if self._first_thread_leaker_id is None:
441
self._first_thread_leaker_id = test.id()
442
self._active_threads = now_active_threads
444
def _recordTestStartTime(self):
445
"""Record that a test has started."""
446
self._start_datetime = self._now()
448
def addError(self, test, err):
449
"""Tell result that test finished with an error.
451
Called from the TestCase run() method when the test
452
fails with an unexpected error.
454
self._post_mortem(self._traceback_from_test or err[2])
455
super(ExtendedTestResult, self).addError(test, err)
456
self.error_count += 1
457
self.report_error(test, err)
461
def addFailure(self, test, err):
462
"""Tell result that test failed.
464
Called from the TestCase run() method when the test
465
fails because e.g. an assert() method failed.
467
self._post_mortem(self._traceback_from_test or err[2])
468
super(ExtendedTestResult, self).addFailure(test, err)
469
self.failure_count += 1
470
self.report_failure(test, err)
474
def addSuccess(self, test, details=None):
475
"""Tell result that test completed successfully.
477
Called from the TestCase run()
479
if self._bench_history is not None:
480
benchmark_time = self._extractBenchmarkTime(test, details)
481
if benchmark_time is not None:
482
self._bench_history.write("%s %s\n" % (
483
self._formatTime(benchmark_time),
485
self.report_success(test)
486
super(ExtendedTestResult, self).addSuccess(test)
487
test._log_contents = ''
489
def addExpectedFailure(self, test, err):
490
self.known_failure_count += 1
491
self.report_known_failure(test, err)
493
def addUnexpectedSuccess(self, test, details=None):
494
"""Tell result the test unexpectedly passed, counting as a failure
496
When the minimum version of testtools required becomes 0.9.8 this
497
can be updated to use the new handling there.
499
super(ExtendedTestResult, self).addFailure(test, details=details)
500
self.failure_count += 1
501
self.report_unexpected_success(test,
502
"".join(details["reason"].iter_text()))
506
def addNotSupported(self, test, feature):
507
"""The test will not be run because of a missing feature.
509
# this can be called in two different ways: it may be that the
510
# test started running, and then raised (through requireFeature)
511
# UnavailableFeature. Alternatively this method can be called
512
# while probing for features before running the test code proper; in
513
# that case we will see startTest and stopTest, but the test will
514
# never actually run.
515
self.unsupported.setdefault(str(feature), 0)
516
self.unsupported[str(feature)] += 1
517
self.report_unsupported(test, feature)
519
def addSkip(self, test, reason):
520
"""A test has not run for 'reason'."""
522
self.report_skip(test, reason)
524
def addNotApplicable(self, test, reason):
525
self.not_applicable_count += 1
526
self.report_not_applicable(test, reason)
528
def _count_stored_tests(self):
529
"""Count of tests instances kept alive due to not succeeding"""
530
return self.error_count + self.failure_count + self.known_failure_count
532
def _post_mortem(self, tb=None):
533
"""Start a PDB post mortem session."""
534
if os.environ.get('BRZ_TEST_PDB', None):
538
def progress(self, offset, whence):
539
"""The test is adjusting the count of tests to run."""
540
if whence == SUBUNIT_SEEK_SET:
541
self.num_tests = offset
542
elif whence == SUBUNIT_SEEK_CUR:
543
self.num_tests += offset
545
raise errors.BzrError("Unknown whence %r" % whence)
547
def report_tests_starting(self):
548
"""Display information before the test run begins"""
549
if getattr(sys, 'frozen', None) is None:
550
bzr_path = osutils.realpath(sys.argv[0])
552
bzr_path = sys.executable
554
'brz selftest: %s\n' % (bzr_path,))
557
breezy.__path__[0],))
559
' bzr-%s python-%s %s\n' % (
560
breezy.version_string,
561
breezy._format_version_tuple(sys.version_info),
562
platform.platform(aliased=1),
564
self.stream.write('\n')
566
def report_test_start(self, test):
567
"""Display information on the test just about to be run"""
569
def _report_thread_leak(self, test, leaked_threads, active_threads):
570
"""Display information on a test that leaked one or more threads"""
571
# GZ 2010-09-09: A leak summary reported separately from the general
572
# thread debugging would be nice. Tests under subunit
573
# need something not using stream, perhaps adding a
574
# testtools details object would be fitting.
575
if 'threads' in selftest_debug_flags:
576
self.stream.write('%s is leaking, active is now %d\n' %
577
(test.id(), len(active_threads)))
579
def startTestRun(self):
580
self.startTime = time.time()
582
def report_success(self, test):
585
def wasStrictlySuccessful(self):
586
if self.unsupported or self.known_failure_count:
588
return self.wasSuccessful()
591
class TextTestResult(ExtendedTestResult):
592
"""Displays progress and results of tests in text form"""
594
def __init__(self, stream, descriptions, verbosity,
598
ExtendedTestResult.__init__(self, stream, descriptions, verbosity,
599
bench_history, strict)
600
self.pb = self.ui.nested_progress_bar()
601
self.pb.show_pct = False
602
self.pb.show_spinner = False
603
self.pb.show_eta = False,
604
self.pb.show_count = False
605
self.pb.show_bar = False
606
self.pb.update_latency = 0
607
self.pb.show_transport_activity = False
609
def stopTestRun(self):
610
# called when the tests that are going to run have run
613
super(TextTestResult, self).stopTestRun()
615
def report_tests_starting(self):
616
super(TextTestResult, self).report_tests_starting()
617
self.pb.update('[test 0/%d] Starting' % (self.num_tests))
619
def _progress_prefix_text(self):
620
# the longer this text, the less space we have to show the test
622
a = '[%d' % self.count # total that have been run
623
# tests skipped as known not to be relevant are not important enough
625
# if self.skip_count:
626
# a += ', %d skip' % self.skip_count
627
# if self.known_failure_count:
628
# a += '+%dX' % self.known_failure_count
630
a += '/%d' % self.num_tests
632
runtime = time.time() - self._overall_start_time
634
a += '%dm%ds' % (runtime / 60, runtime % 60)
637
total_fail_count = self.error_count + self.failure_count
639
a += ', %d failed' % total_fail_count
640
# if self.unsupported:
641
# a += ', %d missing' % len(self.unsupported)
645
def report_test_start(self, test):
647
self._progress_prefix_text() +
649
self._shortened_test_description(test))
651
def _test_description(self, test):
652
return self._shortened_test_description(test)
654
def report_error(self, test, err):
655
self.stream.write('ERROR: %s\n %s\n' % (
656
self._test_description(test),
660
def report_failure(self, test, err):
661
self.stream.write('FAIL: %s\n %s\n' % (
662
self._test_description(test),
666
def report_known_failure(self, test, err):
669
def report_unexpected_success(self, test, reason):
670
self.stream.write('FAIL: %s\n %s: %s\n' % (
671
self._test_description(test),
672
"Unexpected success. Should have failed",
676
def report_skip(self, test, reason):
679
def report_not_applicable(self, test, reason):
682
def report_unsupported(self, test, feature):
683
"""test cannot be run because feature is missing."""
686
class VerboseTestResult(ExtendedTestResult):
687
"""Produce long output, with one line per test run plus times"""
689
def _ellipsize_to_right(self, a_string, final_width):
690
"""Truncate and pad a string, keeping the right hand side"""
691
if len(a_string) > final_width:
692
result = '...' + a_string[3 - final_width:]
695
return result.ljust(final_width)
697
def report_tests_starting(self):
698
self.stream.write('running %d tests...\n' % self.num_tests)
699
super(VerboseTestResult, self).report_tests_starting()
701
def report_test_start(self, test):
702
name = self._shortened_test_description(test)
703
width = osutils.terminal_width()
704
if width is not None:
705
# width needs space for 6 char status, plus 1 for slash, plus an
706
# 11-char time string, plus a trailing blank
707
# when NUMBERED_DIRS: plus 5 chars on test number, plus 1 char on
709
self.stream.write(self._ellipsize_to_right(name, width - 18))
711
self.stream.write(name)
714
def _error_summary(self, err):
716
return '%s%s' % (indent, err[1])
718
def report_error(self, test, err):
719
self.stream.write('ERROR %s\n%s\n'
720
% (self._testTimeString(test),
721
self._error_summary(err)))
723
def report_failure(self, test, err):
724
self.stream.write(' FAIL %s\n%s\n'
725
% (self._testTimeString(test),
726
self._error_summary(err)))
728
def report_known_failure(self, test, err):
729
self.stream.write('XFAIL %s\n%s\n'
730
% (self._testTimeString(test),
731
self._error_summary(err)))
733
def report_unexpected_success(self, test, reason):
734
self.stream.write(' FAIL %s\n%s: %s\n'
735
% (self._testTimeString(test),
736
"Unexpected success. Should have failed",
739
def report_success(self, test):
740
self.stream.write(' OK %s\n' % self._testTimeString(test))
741
for bench_called, stats in getattr(test, '_benchcalls', []):
742
self.stream.write('LSProf output for %s(%s, %s)\n' % bench_called)
743
stats.pprint(file=self.stream)
744
# flush the stream so that we get smooth output. This verbose mode is
745
# used to show the output in PQM.
748
def report_skip(self, test, reason):
749
self.stream.write(' SKIP %s\n%s\n'
750
% (self._testTimeString(test), reason))
752
def report_not_applicable(self, test, reason):
753
self.stream.write(' N/A %s\n %s\n'
754
% (self._testTimeString(test), reason))
756
def report_unsupported(self, test, feature):
757
"""test cannot be run because feature is missing."""
758
self.stream.write("NODEP %s\n The feature '%s' is not available.\n"
759
% (self._testTimeString(test), feature))
762
class TextTestRunner(object):
763
stop_on_failure = False
771
result_decorators=None,
773
"""Create a TextTestRunner.
775
:param result_decorators: An optional list of decorators to apply
776
to the result object being used by the runner. Decorators are
777
applied left to right - the first element in the list is the
780
# stream may know claim to know to write unicode strings, but in older
781
# pythons this goes sufficiently wrong that it is a bad idea. (
782
# specifically a built in file with encoding 'UTF-8' will still try
783
# to encode using ascii.
784
new_encoding = osutils.get_terminal_encoding()
785
codec = codecs.lookup(new_encoding)
786
encode = codec.encode
787
# GZ 2010-09-08: Really we don't want to be writing arbitrary bytes,
788
# so should swap to the plain codecs.StreamWriter
789
stream = osutils.UnicodeOrBytesToBytesWriter(encode, stream,
791
stream.encoding = new_encoding
793
self.descriptions = descriptions
794
self.verbosity = verbosity
795
self._bench_history = bench_history
796
self._strict = strict
797
self._result_decorators = result_decorators or []
800
"Run the given test case or test suite."
801
if self.verbosity == 1:
802
result_class = TextTestResult
803
elif self.verbosity >= 2:
804
result_class = VerboseTestResult
805
original_result = result_class(self.stream,
808
bench_history=self._bench_history,
811
# Signal to result objects that look at stop early policy to stop,
812
original_result.stop_early = self.stop_on_failure
813
result = original_result
814
for decorator in self._result_decorators:
815
result = decorator(result)
816
result.stop_early = self.stop_on_failure
817
result.startTestRun()
822
# higher level code uses our extended protocol to determine
823
# what exit code to give.
824
return original_result
827
def iter_suite_tests(suite):
828
"""Return all tests in a suite, recursing through nested suites"""
829
if isinstance(suite, unittest.TestCase):
831
elif isinstance(suite, unittest.TestSuite):
833
for r in iter_suite_tests(item):
836
raise Exception('unknown type %r for object %r'
837
% (type(suite), suite))
840
TestSkipped = testtools.testcase.TestSkipped
843
class TestNotApplicable(TestSkipped):
844
"""A test is not applicable to the situation where it was run.
846
This is only normally raised by parameterized tests, if they find that
847
the instance they're constructed upon does not support one aspect
852
# traceback._some_str fails to format exceptions that have the default
853
# __str__ which does an implicit ascii conversion. However, repr() on those
854
# objects works, for all that its not quite what the doctor may have ordered.
855
def _clever_some_str(value):
858
except BaseException:
860
return repr(value).replace('\\n', '\n')
861
except BaseException:
862
return '<unprintable %s object>' % type(value).__name__
865
traceback._some_str = _clever_some_str
868
# deprecated - use self.knownFailure(), or self.expectFailure.
869
KnownFailure = testtools.testcase._ExpectedFailure
872
class UnavailableFeature(Exception):
873
"""A feature required for this test was not available.
875
This can be considered a specialised form of SkippedTest.
877
The feature should be used to construct the exception.
881
class StringIOWrapper(ui_testing.BytesIOWithEncoding):
883
@symbol_versioning.deprecated_method(
884
symbol_versioning.deprecated_in((3, 0)))
885
def __init__(self, s=None):
886
super(StringIOWrapper, self).__init__(s)
889
TestUIFactory = ui_testing.TestUIFactory
892
def isolated_doctest_setUp(test):
893
override_os_environ(test)
894
osutils.set_or_unset_env('BRZ_HOME', '/nonexistent')
895
test._orig_ui_factory = ui.ui_factory
896
ui.ui_factory = ui.SilentUIFactory()
899
def isolated_doctest_tearDown(test):
900
restore_os_environ(test)
901
ui.ui_factory = test._orig_ui_factory
904
def IsolatedDocTestSuite(*args, **kwargs):
905
"""Overrides doctest.DocTestSuite to handle isolation.
907
The method is really a factory and users are expected to use it as such.
910
kwargs['setUp'] = isolated_doctest_setUp
911
kwargs['tearDown'] = isolated_doctest_tearDown
912
return doctest.DocTestSuite(*args, **kwargs)
915
class TestCase(testtools.TestCase):
916
"""Base class for brz unit tests.
918
Tests that need access to disk resources should subclass
919
TestCaseInTempDir not TestCase.
921
Error and debug log messages are redirected from their usual
922
location into a temporary file, the contents of which can be
923
retrieved by _get_log(). We use a real OS file, not an in-memory object,
924
so that it can also capture file IO. When the test completes this file
925
is read into memory and removed from disk.
927
There are also convenience functions to invoke bzr's command-line
928
routine, and to build and check brz trees.
930
In addition to the usual method of overriding tearDown(), this class also
931
allows subclasses to register cleanup functions via addCleanup, which are
932
run in order as the object is torn down. It's less likely this will be
933
accidentally overlooked.
937
# record lsprof data when performing benchmark calls.
938
_gather_lsprof_in_benchmarks = False
940
def __init__(self, methodName='testMethod'):
941
super(TestCase, self).__init__(methodName)
942
self._directory_isolation = True
943
self.exception_handlers.insert(
944
0, (UnavailableFeature, self._do_unsupported_or_skip))
945
self.exception_handlers.insert(
946
0, (TestNotApplicable, self._do_not_applicable))
949
super(TestCase, self).setUp()
951
# At this point we're still accessing the config files in $BRZ_HOME (as
952
# set by the user running selftest).
953
timeout = config.GlobalStack().get('selftest.timeout')
955
timeout_fixture = fixtures.TimeoutFixture(timeout)
956
timeout_fixture.setUp()
957
self.addCleanup(timeout_fixture.cleanUp)
959
for feature in getattr(self, '_test_needs_features', []):
960
self.requireFeature(feature)
961
self._cleanEnvironment()
963
self.overrideAttr(breezy.get_global_state(), 'cmdline_overrides',
964
config.CommandLineStore())
968
self._benchcalls = []
969
self._benchtime = None
971
self._track_transports()
973
self._clear_debug_flags()
974
# Isolate global verbosity level, to make sure it's reproducible
975
# between tests. We should get rid of this altogether: bug 656694. --
977
self.overrideAttr(breezy.trace, '_verbosity_level', 0)
978
self._log_files = set()
979
# Each key in the ``_counters`` dict holds a value for a different
980
# counter. When the test ends, addDetail() should be used to output the
981
# counter values. This happens in install_counter_hook().
983
if 'config_stats' in selftest_debug_flags:
984
self._install_config_stats_hooks()
985
# Do not use i18n for tests (unless the test reverses this)
991
# The sys preserved stdin/stdout should allow blackbox tests debugging
992
pdb.Pdb(stdin=sys.__stdin__, stdout=sys.__stdout__
993
).set_trace(sys._getframe().f_back)
995
def discardDetail(self, name):
996
"""Extend the addDetail, getDetails api so we can remove a detail.
998
eg. brz always adds the 'log' detail at startup, but we don't want to
999
include it for skipped, xfail, etc tests.
1001
It is safe to call this for a detail that doesn't exist, in case this
1002
gets called multiple times.
1004
# We cheat. details is stored in __details which means we shouldn't
1005
# touch it. but getDetails() returns the dict directly, so we can
1007
details = self.getDetails()
1011
def install_counter_hook(self, hooks, name, counter_name=None):
1012
"""Install a counting hook.
1014
Any hook can be counted as long as it doesn't need to return a value.
1016
:param hooks: Where the hook should be installed.
1018
:param name: The hook name that will be counted.
1020
:param counter_name: The counter identifier in ``_counters``, defaults
1023
_counters = self._counters # Avoid closing over self
1024
if counter_name is None:
1026
if counter_name in _counters:
1027
raise AssertionError('%s is already used as a counter name'
1029
_counters[counter_name] = 0
1030
self.addDetail(counter_name, content.Content(
1032
lambda: [b'%d' % (_counters[counter_name],)]))
1034
def increment_counter(*args, **kwargs):
1035
_counters[counter_name] += 1
1036
label = 'count %s calls' % (counter_name,)
1037
hooks.install_named_hook(name, increment_counter, label)
1038
self.addCleanup(hooks.uninstall_named_hook, name, label)
1040
def _install_config_stats_hooks(self):
1041
"""Install config hooks to count hook calls.
1044
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1045
self.install_counter_hook(config.ConfigHooks, hook_name,
1046
'config.%s' % (hook_name,))
1048
# The OldConfigHooks are private and need special handling to protect
1049
# against recursive tests (tests that run other tests), so we just do
1050
# manually what registering them into _builtin_known_hooks will provide
1052
self.overrideAttr(config, 'OldConfigHooks', config._OldConfigHooks())
1053
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1054
self.install_counter_hook(config.OldConfigHooks, hook_name,
1055
'old_config.%s' % (hook_name,))
1057
def _clear_debug_flags(self):
1058
"""Prevent externally set debug flags affecting tests.
1060
Tests that want to use debug flags can just set them in the
1061
debug_flags set during setup/teardown.
1063
# Start with a copy of the current debug flags we can safely modify.
1064
self.overrideAttr(debug, 'debug_flags', set(debug.debug_flags))
1065
if 'allow_debug' not in selftest_debug_flags:
1066
debug.debug_flags.clear()
1067
if 'disable_lock_checks' not in selftest_debug_flags:
1068
debug.debug_flags.add('strict_locks')
1070
def _clear_hooks(self):
1071
# prevent hooks affecting tests
1072
known_hooks = hooks.known_hooks
1073
self._preserved_hooks = {}
1074
for key, (parent, name) in known_hooks.iter_parent_objects():
1075
current_hooks = getattr(parent, name)
1076
self._preserved_hooks[parent] = (name, current_hooks)
1077
self._preserved_lazy_hooks = hooks._lazy_hooks
1078
hooks._lazy_hooks = {}
1079
self.addCleanup(self._restoreHooks)
1080
for key, (parent, name) in known_hooks.iter_parent_objects():
1081
factory = known_hooks.get(key)
1082
setattr(parent, name, factory())
1083
# this hook should always be installed
1084
request._install_hook()
1086
def disable_directory_isolation(self):
1087
"""Turn off directory isolation checks."""
1088
self._directory_isolation = False
1090
def enable_directory_isolation(self):
1091
"""Enable directory isolation checks."""
1092
self._directory_isolation = True
1094
def _silenceUI(self):
1095
"""Turn off UI for duration of test"""
1096
# by default the UI is off; tests can turn it on if they want it.
1097
self.overrideAttr(ui, 'ui_factory', ui.SilentUIFactory())
1099
def _check_locks(self):
1100
"""Check that all lock take/release actions have been paired."""
1101
# We always check for mismatched locks. If a mismatch is found, we
1102
# fail unless -Edisable_lock_checks is supplied to selftest, in which
1103
# case we just print a warning.
1105
acquired_locks = [lock for action, lock in self._lock_actions
1106
if action == 'acquired']
1107
released_locks = [lock for action, lock in self._lock_actions
1108
if action == 'released']
1109
broken_locks = [lock for action, lock in self._lock_actions
1110
if action == 'broken']
1111
# trivially, given the tests for lock acquistion and release, if we
1112
# have as many in each list, it should be ok. Some lock tests also
1113
# break some locks on purpose and should be taken into account by
1114
# considering that breaking a lock is just a dirty way of releasing it.
1115
if len(acquired_locks) != (len(released_locks) + len(broken_locks)):
1117
'Different number of acquired and '
1118
'released or broken locks.\n'
1122
(acquired_locks, released_locks, broken_locks))
1123
if not self._lock_check_thorough:
1124
# Rather than fail, just warn
1125
print("Broken test %s: %s" % (self, message))
1129
def _track_locks(self):
1130
"""Track lock activity during tests."""
1131
self._lock_actions = []
1132
if 'disable_lock_checks' in selftest_debug_flags:
1133
self._lock_check_thorough = False
1135
self._lock_check_thorough = True
1137
self.addCleanup(self._check_locks)
1138
_mod_lock.Lock.hooks.install_named_hook('lock_acquired',
1139
self._lock_acquired, None)
1140
_mod_lock.Lock.hooks.install_named_hook('lock_released',
1141
self._lock_released, None)
1142
_mod_lock.Lock.hooks.install_named_hook('lock_broken',
1143
self._lock_broken, None)
1145
def _lock_acquired(self, result):
1146
self._lock_actions.append(('acquired', result))
1148
def _lock_released(self, result):
1149
self._lock_actions.append(('released', result))
1151
def _lock_broken(self, result):
1152
self._lock_actions.append(('broken', result))
1154
def permit_dir(self, name):
1155
"""Permit a directory to be used by this test. See permit_url."""
1156
name_transport = _mod_transport.get_transport_from_path(name)
1157
self.permit_url(name)
1158
self.permit_url(name_transport.base)
1160
def permit_url(self, url):
1161
"""Declare that url is an ok url to use in this test.
1163
Do this for memory transports, temporary test directory etc.
1165
Do not do this for the current working directory, /tmp, or any other
1166
preexisting non isolated url.
1168
if not url.endswith('/'):
1170
self._bzr_selftest_roots.append(url)
1172
def permit_source_tree_branch_repo(self):
1173
"""Permit the source tree brz is running from to be opened.
1175
Some code such as breezy.version attempts to read from the brz branch
1176
that brz is executing from (if any). This method permits that directory
1177
to be used in the test suite.
1179
path = self.get_source_path()
1180
self.record_directory_isolation()
1183
workingtree.WorkingTree.open(path)
1184
except (errors.NotBranchError, errors.NoWorkingTree):
1185
raise TestSkipped('Needs a working tree of brz sources')
1187
self.enable_directory_isolation()
1189
def _preopen_isolate_transport(self, transport):
1190
"""Check that all transport openings are done in the test work area."""
1191
while isinstance(transport, pathfilter.PathFilteringTransport):
1192
# Unwrap pathfiltered transports
1193
transport = transport.server.backing_transport.clone(
1194
transport._filter('.'))
1195
url = transport.base
1196
# ReadonlySmartTCPServer_for_testing decorates the backing transport
1197
# urls it is given by prepending readonly+. This is appropriate as the
1198
# client shouldn't know that the server is readonly (or not readonly).
1199
# We could register all servers twice, with readonly+ prepending, but
1200
# that makes for a long list; this is about the same but easier to
1202
if url.startswith('readonly+'):
1203
url = url[len('readonly+'):]
1204
self._preopen_isolate_url(url)
1206
def _preopen_isolate_url(self, url):
1207
if not self._directory_isolation:
1209
if self._directory_isolation == 'record':
1210
self._bzr_selftest_roots.append(url)
1212
# This prevents all transports, including e.g. sftp ones backed on disk
1213
# from working unless they are explicitly granted permission. We then
1214
# depend on the code that sets up test transports to check that they
1215
# are appropriately isolated and enable their use by calling
1216
# self.permit_transport()
1217
if not osutils.is_inside_any(self._bzr_selftest_roots, url):
1218
raise errors.BzrError("Attempt to escape test isolation: %r %r"
1219
% (url, self._bzr_selftest_roots))
1221
def record_directory_isolation(self):
1222
"""Gather accessed directories to permit later access.
1224
This is used for tests that access the branch brz is running from.
1226
self._directory_isolation = "record"
1228
def start_server(self, transport_server, backing_server=None):
1229
"""Start transport_server for this test.
1231
This starts the server, registers a cleanup for it and permits the
1232
server's urls to be used.
1234
if backing_server is None:
1235
transport_server.start_server()
1237
transport_server.start_server(backing_server)
1238
self.addCleanup(transport_server.stop_server)
1239
# Obtain a real transport because if the server supplies a password, it
1240
# will be hidden from the base on the client side.
1241
t = _mod_transport.get_transport_from_url(transport_server.get_url())
1242
# Some transport servers effectively chroot the backing transport;
1243
# others like SFTPServer don't - users of the transport can walk up the
1244
# transport to read the entire backing transport. This wouldn't matter
1245
# except that the workdir tests are given - and that they expect the
1246
# server's url to point at - is one directory under the safety net. So
1247
# Branch operations into the transport will attempt to walk up one
1248
# directory. Chrooting all servers would avoid this but also mean that
1249
# we wouldn't be testing directly against non-root urls. Alternatively
1250
# getting the test framework to start the server with a backing server
1251
# at the actual safety net directory would work too, but this then
1252
# means that the self.get_url/self.get_transport methods would need
1253
# to transform all their results. On balance its cleaner to handle it
1254
# here, and permit a higher url when we have one of these transports.
1255
if t.base.endswith('/work/'):
1256
# we have safety net/test root/work
1257
t = t.clone('../..')
1258
elif isinstance(transport_server,
1259
test_server.SmartTCPServer_for_testing):
1260
# The smart server adds a path similar to work, which is traversed
1261
# up from by the client. But the server is chrooted - the actual
1262
# backing transport is not escaped from, and VFS requests to the
1263
# root will error (because they try to escape the chroot).
1265
while t2.base != t.base:
1268
self.permit_url(t.base)
1270
def _track_transports(self):
1271
"""Install checks for transport usage."""
1272
# TestCase has no safe place it can write to.
1273
self._bzr_selftest_roots = []
1274
# Currently the easiest way to be sure that nothing is going on is to
1275
# hook into brz dir opening. This leaves a small window of error for
1276
# transport tests, but they are well known, and we can improve on this
1278
controldir.ControlDir.hooks.install_named_hook(
1279
"pre_open", self._preopen_isolate_transport,
1280
"Check brz directories are safe.")
1282
def _ndiff_strings(self, a, b):
1283
"""Return ndiff between two strings containing lines.
1285
A trailing newline is added if missing to make the strings
1287
if b and not b.endswith('\n'):
1289
if a and not a.endswith('\n'):
1291
difflines = difflib.ndiff(a.splitlines(True),
1293
linejunk=lambda x: False,
1294
charjunk=lambda x: False)
1295
return ''.join(difflines)
1297
def assertEqual(self, a, b, message=''):
1301
except UnicodeError as e:
1302
# If we can't compare without getting a UnicodeError, then
1303
# obviously they are different
1304
trace.mutter('UnicodeError: %s', e)
1307
raise AssertionError("%snot equal:\na = %s\nb = %s\n"
1309
pprint.pformat(a), pprint.pformat(b)))
1311
# FIXME: This is deprecated in unittest2 but plugins may still use it so we
1312
# need a deprecation period for them -- vila 2016-02-01
1313
assertEquals = assertEqual
1315
def assertEqualDiff(self, a, b, message=None):
1316
"""Assert two texts are equal, if not raise an exception.
1318
This is intended for use with multi-line strings where it can
1319
be hard to find the differences by eye.
1321
# TODO: perhaps override assertEqual to call this for strings?
1325
message = "texts not equal:\n"
1326
if a + ('\n' if isinstance(a, str) else b'\n') == b:
1327
message = 'first string is missing a final newline.\n'
1328
if a == b + ('\n' if isinstance(b, str) else b'\n'):
1329
message = 'second string is missing a final newline.\n'
1330
raise AssertionError(message
1331
+ self._ndiff_strings(
1332
a if isinstance(a, str) else a.decode(),
1333
b if isinstance(b, str) else b.decode()))
1335
def assertEqualMode(self, mode, mode_test):
1336
self.assertEqual(mode, mode_test,
1337
'mode mismatch %o != %o' % (mode, mode_test))
1339
def assertEqualStat(self, expected, actual):
1340
"""assert that expected and actual are the same stat result.
1342
:param expected: A stat result.
1343
:param actual: A stat result.
1344
:raises AssertionError: If the expected and actual stat values differ
1345
other than by atime.
1347
self.assertEqual(expected.st_size, actual.st_size,
1348
'st_size did not match')
1349
self.assertEqual(expected.st_mtime, actual.st_mtime,
1350
'st_mtime did not match')
1351
self.assertEqual(expected.st_ctime, actual.st_ctime,
1352
'st_ctime did not match')
1353
if sys.platform == 'win32':
1354
# On Win32 both 'dev' and 'ino' cannot be trusted. In python2.4 it
1355
# is 'dev' that varies, in python 2.5 (6?) it is st_ino that is
1356
# odd. We just force it to always be 0 to avoid any problems.
1357
self.assertEqual(0, expected.st_dev)
1358
self.assertEqual(0, actual.st_dev)
1359
self.assertEqual(0, expected.st_ino)
1360
self.assertEqual(0, actual.st_ino)
1362
self.assertEqual(expected.st_dev, actual.st_dev,
1363
'st_dev did not match')
1364
self.assertEqual(expected.st_ino, actual.st_ino,
1365
'st_ino did not match')
1366
self.assertEqual(expected.st_mode, actual.st_mode,
1367
'st_mode did not match')
1369
def assertLength(self, length, obj_with_len):
1370
"""Assert that obj_with_len is of length length."""
1371
if len(obj_with_len) != length:
1372
self.fail("Incorrect length: wanted %d, got %d for %r" % (
1373
length, len(obj_with_len), obj_with_len))
1375
def assertLogsError(self, exception_class, func, *args, **kwargs):
1376
"""Assert that `func(*args, **kwargs)` quietly logs a specific error.
1379
orig_log_exception_quietly = trace.log_exception_quietly
1382
orig_log_exception_quietly()
1383
captured.append(sys.exc_info()[1])
1384
trace.log_exception_quietly = capture
1385
func(*args, **kwargs)
1387
trace.log_exception_quietly = orig_log_exception_quietly
1388
self.assertLength(1, captured)
1390
self.assertIsInstance(err, exception_class)
1393
def assertPositive(self, val):
1394
"""Assert that val is greater than 0."""
1395
self.assertTrue(val > 0, 'expected a positive value, but got %s' % val)
1397
def assertNegative(self, val):
1398
"""Assert that val is less than 0."""
1399
self.assertTrue(val < 0, 'expected a negative value, but got %s' % val)
1401
def assertStartsWith(self, s, prefix):
1402
if not s.startswith(prefix):
1403
raise AssertionError(
1404
'string %r does not start with %r' % (s, prefix))
1406
def assertEndsWith(self, s, suffix):
1407
"""Asserts that s ends with suffix."""
1408
if not s.endswith(suffix):
1409
raise AssertionError(
1410
'string %r does not end with %r' % (s, suffix))
1412
def assertContainsRe(self, haystack, needle_re, flags=0):
1413
"""Assert that a contains something matching a regular expression."""
1414
if not re.search(needle_re, haystack, flags):
1415
if (('\n' if isinstance(haystack, str) else b'\n') in haystack or
1416
len(haystack) > 60):
1417
# a long string, format it in a more readable way
1418
raise AssertionError(
1419
'pattern "%s" not found in\n"""\\\n%s"""\n'
1420
% (needle_re, haystack))
1422
raise AssertionError('pattern "%s" not found in "%s"'
1423
% (needle_re, haystack))
1425
def assertNotContainsRe(self, haystack, needle_re, flags=0):
1426
"""Assert that a does not match a regular expression"""
1427
if re.search(needle_re, haystack, flags):
1428
raise AssertionError('pattern "%s" found in "%s"'
1429
% (needle_re, haystack))
1431
def assertContainsString(self, haystack, needle):
1432
if haystack.find(needle) == -1:
1433
self.fail("string %r not found in '''%s'''" % (needle, haystack))
1435
def assertNotContainsString(self, haystack, needle):
1436
if haystack.find(needle) != -1:
1437
self.fail("string %r found in '''%s'''" % (needle, haystack))
1439
def assertSubset(self, sublist, superlist):
1440
"""Assert that every entry in sublist is present in superlist."""
1441
missing = set(sublist) - set(superlist)
1442
if len(missing) > 0:
1443
raise AssertionError("value(s) %r not present in container %r" %
1444
(missing, superlist))
1446
def assertListRaises(self, excClass, func, *args, **kwargs):
1447
"""Fail unless excClass is raised when the iterator from func is used.
1449
Many functions can return generators this makes sure
1450
to wrap them in a list() call to make sure the whole generator
1451
is run, and that the proper exception is raised.
1454
list(func(*args, **kwargs))
1455
except excClass as e:
1458
if getattr(excClass, '__name__', None) is not None:
1459
excName = excClass.__name__
1461
excName = str(excClass)
1462
raise self.failureException("%s not raised" % excName)
1464
def assertRaises(self, excClass, callableObj, *args, **kwargs):
1465
"""Assert that a callable raises a particular exception.
1467
:param excClass: As for the except statement, this may be either an
1468
exception class, or a tuple of classes.
1469
:param callableObj: A callable, will be passed ``*args`` and
1472
Returns the exception so that you can examine it.
1475
callableObj(*args, **kwargs)
1476
except excClass as e:
1479
if getattr(excClass, '__name__', None) is not None:
1480
excName = excClass.__name__
1483
excName = str(excClass)
1484
raise self.failureException("%s not raised" % excName)
1486
def assertIs(self, left, right, message=None):
1487
if not (left is right):
1488
if message is not None:
1489
raise AssertionError(message)
1491
raise AssertionError("%r is not %r." % (left, right))
1493
def assertIsNot(self, left, right, message=None):
1495
if message is not None:
1496
raise AssertionError(message)
1498
raise AssertionError("%r is %r." % (left, right))
1500
def assertTransportMode(self, transport, path, mode):
1501
"""Fail if a path does not have mode "mode".
1503
If modes are not supported on this transport, the assertion is ignored.
1505
if not transport._can_roundtrip_unix_modebits():
1507
path_stat = transport.stat(path)
1508
actual_mode = stat.S_IMODE(path_stat.st_mode)
1509
self.assertEqual(mode, actual_mode,
1510
'mode of %r incorrect (%s != %s)'
1511
% (path, oct(mode), oct(actual_mode)))
1513
def assertIsSameRealPath(self, path1, path2):
1514
"""Fail if path1 and path2 points to different files"""
1515
self.assertEqual(osutils.realpath(path1),
1516
osutils.realpath(path2),
1517
"apparent paths:\na = %s\nb = %s\n," % (path1, path2))
1519
def assertIsInstance(self, obj, kls, msg=None):
1520
"""Fail if obj is not an instance of kls
1522
:param msg: Supplementary message to show if the assertion fails.
1524
if not isinstance(obj, kls):
1525
m = "%r is an instance of %s rather than %s" % (
1526
obj, obj.__class__, kls)
1531
def assertFileEqual(self, content, path):
1532
"""Fail if path does not contain 'content'."""
1533
self.assertPathExists(path)
1535
mode = 'r' + ('b' if isinstance(content, bytes) else '')
1536
with open(path, mode) as f:
1538
self.assertEqualDiff(content, s)
1540
def assertDocstring(self, expected_docstring, obj):
1541
"""Fail if obj does not have expected_docstring"""
1543
# With -OO the docstring should be None instead
1544
self.assertIs(obj.__doc__, None)
1546
self.assertEqual(expected_docstring, obj.__doc__)
1548
def assertPathExists(self, path):
1549
"""Fail unless path or paths, which may be abs or relative, exist."""
1550
# TODO(jelmer): Clean this up for pad.lv/1696545
1551
if not isinstance(path, (bytes, str)):
1553
self.assertPathExists(p)
1555
self.assertTrue(osutils.lexists(path),
1556
path + " does not exist")
1558
def assertPathDoesNotExist(self, path):
1559
"""Fail if path or paths, which may be abs or relative, exist."""
1560
if not isinstance(path, (str, str)):
1562
self.assertPathDoesNotExist(p)
1564
self.assertFalse(osutils.lexists(path),
1567
def _capture_deprecation_warnings(self, a_callable, *args, **kwargs):
1568
"""A helper for callDeprecated and applyDeprecated.
1570
:param a_callable: A callable to call.
1571
:param args: The positional arguments for the callable
1572
:param kwargs: The keyword arguments for the callable
1573
:return: A tuple (warnings, result). result is the result of calling
1574
a_callable(``*args``, ``**kwargs``).
1578
def capture_warnings(msg, cls=None, stacklevel=None):
1579
# we've hooked into a deprecation specific callpath,
1580
# only deprecations should getting sent via it.
1581
self.assertEqual(cls, DeprecationWarning)
1582
local_warnings.append(msg)
1583
original_warning_method = symbol_versioning.warn
1584
symbol_versioning.set_warning_method(capture_warnings)
1586
result = a_callable(*args, **kwargs)
1588
symbol_versioning.set_warning_method(original_warning_method)
1589
return (local_warnings, result)
1591
def applyDeprecated(self, deprecation_format, a_callable, *args, **kwargs):
1592
"""Call a deprecated callable without warning the user.
1594
Note that this only captures warnings raised by symbol_versioning.warn,
1595
not other callers that go direct to the warning module.
1597
To test that a deprecated method raises an error, do something like
1598
this (remember that both assertRaises and applyDeprecated delays *args
1599
and **kwargs passing)::
1601
self.assertRaises(errors.ReservedId,
1602
self.applyDeprecated,
1603
deprecated_in((1, 5, 0)),
1607
:param deprecation_format: The deprecation format that the callable
1608
should have been deprecated with. This is the same type as the
1609
parameter to deprecated_method/deprecated_function. If the
1610
callable is not deprecated with this format, an assertion error
1612
:param a_callable: A callable to call. This may be a bound method or
1613
a regular function. It will be called with ``*args`` and
1615
:param args: The positional arguments for the callable
1616
:param kwargs: The keyword arguments for the callable
1617
:return: The result of a_callable(``*args``, ``**kwargs``)
1619
call_warnings, result = self._capture_deprecation_warnings(
1620
a_callable, *args, **kwargs)
1621
expected_first_warning = symbol_versioning.deprecation_string(
1622
a_callable, deprecation_format)
1623
if len(call_warnings) == 0:
1624
self.fail("No deprecation warning generated by call to %s" %
1626
self.assertEqual(expected_first_warning, call_warnings[0])
1629
def callCatchWarnings(self, fn, *args, **kw):
1630
"""Call a callable that raises python warnings.
1632
The caller's responsible for examining the returned warnings.
1634
If the callable raises an exception, the exception is not
1635
caught and propagates up to the caller. In that case, the list
1636
of warnings is not available.
1638
:returns: ([warning_object, ...], fn_result)
1640
# XXX: This is not perfect, because it completely overrides the
1641
# warnings filters, and some code may depend on suppressing particular
1642
# warnings. It's the easiest way to insulate ourselves from -Werror,
1643
# though. -- Andrew, 20071062
1646
def _catcher(message, category, filename, lineno, file=None, line=None):
1647
# despite the name, 'message' is normally(?) a Warning subclass
1649
wlist.append(message)
1650
saved_showwarning = warnings.showwarning
1651
saved_filters = warnings.filters
1653
warnings.showwarning = _catcher
1654
warnings.filters = []
1655
result = fn(*args, **kw)
1657
warnings.showwarning = saved_showwarning
1658
warnings.filters = saved_filters
1659
return wlist, result
1661
def callDeprecated(self, expected, callable, *args, **kwargs):
1662
"""Assert that a callable is deprecated in a particular way.
1664
This is a very precise test for unusual requirements. The
1665
applyDeprecated helper function is probably more suited for most tests
1666
as it allows you to simply specify the deprecation format being used
1667
and will ensure that that is issued for the function being called.
1669
Note that this only captures warnings raised by symbol_versioning.warn,
1670
not other callers that go direct to the warning module. To catch
1671
general warnings, use callCatchWarnings.
1673
:param expected: a list of the deprecation warnings expected, in order
1674
:param callable: The callable to call
1675
:param args: The positional arguments for the callable
1676
:param kwargs: The keyword arguments for the callable
1678
call_warnings, result = self._capture_deprecation_warnings(
1679
callable, *args, **kwargs)
1680
self.assertEqual(expected, call_warnings)
1683
def _startLogFile(self):
1684
"""Setup a in-memory target for bzr and testcase log messages"""
1685
pseudo_log_file = BytesIO()
1687
def _get_log_contents_for_weird_testtools_api():
1688
return [pseudo_log_file.getvalue().decode(
1689
"utf-8", "replace").encode("utf-8")]
1691
"log", content.Content(
1692
content.ContentType("text", "plain", {"charset": "utf8"}),
1693
_get_log_contents_for_weird_testtools_api))
1694
self._log_file = pseudo_log_file
1695
self._log_memento = trace.push_log_file(self._log_file)
1696
self.addCleanup(self._finishLogFile)
1698
def _finishLogFile(self):
1699
"""Flush and dereference the in-memory log for this testcase"""
1700
if trace._trace_file:
1701
# flush the log file, to get all content
1702
trace._trace_file.flush()
1703
trace.pop_log_file(self._log_memento)
1704
# The logging module now tracks references for cleanup so discard ours
1705
del self._log_memento
1707
def thisFailsStrictLockCheck(self):
1708
"""It is known that this test would fail with -Dstrict_locks.
1710
By default, all tests are run with strict lock checking unless
1711
-Edisable_lock_checks is supplied. However there are some tests which
1712
we know fail strict locks at this point that have not been fixed.
1713
They should call this function to disable the strict checking.
1715
This should be used sparingly, it is much better to fix the locking
1716
issues rather than papering over the problem by calling this function.
1718
debug.debug_flags.discard('strict_locks')
1720
def overrideAttr(self, obj, attr_name, new=_unitialized_attr):
1721
"""Overrides an object attribute restoring it after the test.
1723
:note: This should be used with discretion; you should think about
1724
whether it's better to make the code testable without monkey-patching.
1726
:param obj: The object that will be mutated.
1728
:param attr_name: The attribute name we want to preserve/override in
1731
:param new: The optional value we want to set the attribute to.
1733
:returns: The actual attr value.
1735
# The actual value is captured by the call below
1736
value = getattr(obj, attr_name, _unitialized_attr)
1737
if value is _unitialized_attr:
1738
# When the test completes, the attribute should not exist, but if
1739
# we aren't setting a value, we don't need to do anything.
1740
if new is not _unitialized_attr:
1741
self.addCleanup(delattr, obj, attr_name)
1743
self.addCleanup(setattr, obj, attr_name, value)
1744
if new is not _unitialized_attr:
1745
setattr(obj, attr_name, new)
1748
def overrideEnv(self, name, new):
1749
"""Set an environment variable, and reset it after the test.
1751
:param name: The environment variable name.
1753
:param new: The value to set the variable to. If None, the
1754
variable is deleted from the environment.
1756
:returns: The actual variable value.
1758
value = osutils.set_or_unset_env(name, new)
1759
self.addCleanup(osutils.set_or_unset_env, name, value)
1762
def recordCalls(self, obj, attr_name):
1763
"""Monkeypatch in a wrapper that will record calls.
1765
The monkeypatch is automatically removed when the test concludes.
1767
:param obj: The namespace holding the reference to be replaced;
1768
typically a module, class, or object.
1769
:param attr_name: A string for the name of the attribute to patch.
1770
:returns: A list that will be extended with one item every time the
1771
function is called, with a tuple of (args, kwargs).
1775
def decorator(*args, **kwargs):
1776
calls.append((args, kwargs))
1777
return orig(*args, **kwargs)
1778
orig = self.overrideAttr(obj, attr_name, decorator)
1781
def _cleanEnvironment(self):
1782
for name, value in isolated_environ.items():
1783
self.overrideEnv(name, value)
1785
def _restoreHooks(self):
1786
for klass, (name, hooks) in self._preserved_hooks.items():
1787
setattr(klass, name, hooks)
1788
self._preserved_hooks.clear()
1789
breezy.hooks._lazy_hooks = self._preserved_lazy_hooks
1790
self._preserved_lazy_hooks.clear()
1792
def knownFailure(self, reason):
1793
"""Declare that this test fails for a known reason
1795
Tests that are known to fail should generally be using expectedFailure
1796
with an appropriate reverse assertion if a change could cause the test
1797
to start passing. Conversely if the test has no immediate prospect of
1798
succeeding then using skip is more suitable.
1800
When this method is called while an exception is being handled, that
1801
traceback will be used, otherwise a new exception will be thrown to
1802
provide one but won't be reported.
1804
self._add_reason(reason)
1806
exc_info = sys.exc_info()
1807
if exc_info != (None, None, None):
1808
self._report_traceback(exc_info)
1811
raise self.failureException(reason)
1812
except self.failureException:
1813
exc_info = sys.exc_info()
1814
# GZ 02-08-2011: Maybe cleanup this err.exc_info attribute too?
1815
raise testtools.testcase._ExpectedFailure(exc_info)
1819
def _suppress_log(self):
1820
"""Remove the log info from details."""
1821
self.discardDetail('log')
1823
def _do_skip(self, result, reason):
1824
self._suppress_log()
1825
addSkip = getattr(result, 'addSkip', None)
1826
if not callable(addSkip):
1827
result.addSuccess(result)
1829
addSkip(self, str(reason))
1832
def _do_known_failure(self, result, e):
1833
self._suppress_log()
1834
err = sys.exc_info()
1835
addExpectedFailure = getattr(result, 'addExpectedFailure', None)
1836
if addExpectedFailure is not None:
1837
addExpectedFailure(self, err)
1839
result.addSuccess(self)
1842
def _do_not_applicable(self, result, e):
1844
reason = 'No reason given'
1847
self._suppress_log()
1848
addNotApplicable = getattr(result, 'addNotApplicable', None)
1849
if addNotApplicable is not None:
1850
result.addNotApplicable(self, reason)
1852
self._do_skip(result, reason)
1855
def _report_skip(self, result, err):
1856
"""Override the default _report_skip.
1858
We want to strip the 'log' detail. If we waint until _do_skip, it has
1859
already been formatted into the 'reason' string, and we can't pull it
1862
self._suppress_log()
1863
super(TestCase, self)._report_skip(self, result, err)
1866
def _report_expected_failure(self, result, err):
1869
See _report_skip for motivation.
1871
self._suppress_log()
1872
super(TestCase, self)._report_expected_failure(self, result, err)
1875
def _do_unsupported_or_skip(self, result, e):
1877
self._suppress_log()
1878
addNotSupported = getattr(result, 'addNotSupported', None)
1879
if addNotSupported is not None:
1880
result.addNotSupported(self, reason)
1882
self._do_skip(result, reason)
1884
def time(self, callable, *args, **kwargs):
1885
"""Run callable and accrue the time it takes to the benchmark time.
1887
If lsprofiling is enabled (i.e. by --lsprof-time to brz selftest) then
1888
this will cause lsprofile statistics to be gathered and stored in
1891
if self._benchtime is None:
1892
self.addDetail('benchtime', content.Content(
1894
lambda: [str(self._benchtime).encode('utf-8')]))
1898
if not self._gather_lsprof_in_benchmarks:
1899
return callable(*args, **kwargs)
1901
# record this benchmark
1902
ret, stats = breezy.lsprof.profile(callable, *args, **kwargs)
1904
self._benchcalls.append(((callable, args, kwargs), stats))
1907
self._benchtime += time.time() - start
1909
def log(self, *args):
1913
"""Get a unicode string containing the log from breezy.trace.
1915
Undecodable characters are replaced.
1917
return u"".join(self.getDetails()['log'].iter_text())
1919
def requireFeature(self, feature):
1920
"""This test requires a specific feature is available.
1922
:raises UnavailableFeature: When feature is not available.
1924
if not feature.available():
1925
raise UnavailableFeature(feature)
1927
def _run_bzr_core(self, args, encoding, stdin, stdout, stderr,
1929
# Clear chk_map page cache, because the contents are likely to mask
1931
chk_map.clear_cache()
1933
self.log('run brz: %r', args)
1935
self._last_cmd_stdout = stdout
1936
self._last_cmd_stderr = stderr
1938
old_ui_factory = ui.ui_factory
1939
ui.ui_factory = ui_testing.TestUIFactory(
1941
stdout=self._last_cmd_stdout,
1942
stderr=self._last_cmd_stderr)
1945
if working_dir is not None:
1946
cwd = osutils.getcwd()
1947
os.chdir(working_dir)
1951
result = self.apply_redirected(
1952
ui.ui_factory.stdin,
1954
_mod_commands.run_bzr_catch_user_errors,
1957
ui.ui_factory = old_ui_factory
1963
def run_bzr_raw(self, args, retcode=0, stdin=None, encoding=None,
1964
working_dir=None, error_regexes=[]):
1965
"""Invoke brz, as if it were run from the command line.
1967
The argument list should not include the brz program name - the
1968
first argument is normally the brz command. Arguments may be
1969
passed in three ways:
1971
1- A list of strings, eg ["commit", "a"]. This is recommended
1972
when the command contains whitespace or metacharacters, or
1973
is built up at run time.
1975
2- A single string, eg "add a". This is the most convenient
1976
for hardcoded commands.
1978
This runs brz through the interface that catches and reports
1979
errors, and with logging set to something approximating the
1980
default, so that error reporting can be checked.
1982
This should be the main method for tests that want to exercise the
1983
overall behavior of the brz application (rather than a unit test
1984
or a functional test of the library.)
1986
This sends the stdout/stderr results into the test's log,
1987
where it may be useful for debugging. See also run_captured.
1989
:keyword stdin: A string to be used as stdin for the command.
1990
:keyword retcode: The status code the command should return;
1992
:keyword working_dir: The directory to run the command in
1993
:keyword error_regexes: A list of expected error messages. If
1994
specified they must be seen in the error output of the command.
1996
if isinstance(args, str):
1997
args = shlex.split(args)
1999
if encoding is None:
2000
encoding = osutils.get_user_encoding()
2004
wrapped_stdout = TextIOWrapper(stdout, encoding)
2005
wrapped_stderr = TextIOWrapper(stderr, encoding)
2006
handler = logging.StreamHandler(wrapped_stderr)
2007
handler.setLevel(logging.INFO)
2009
logger = logging.getLogger('')
2010
logger.addHandler(handler)
2012
result = self._run_bzr_core(
2013
args, encoding=encoding, stdin=stdin, stdout=wrapped_stdout,
2014
stderr=wrapped_stderr, working_dir=working_dir,
2017
logger.removeHandler(handler)
2019
wrapped_stdout.flush()
2020
wrapped_stderr.flush()
2022
out = stdout.getvalue()
2023
err = stderr.getvalue()
2025
self.log('output:\n%r', out)
2027
self.log('errors:\n%r', err)
2028
if retcode is not None:
2029
self.assertEqual(retcode, result,
2030
message='Unexpected return code')
2031
self.assertIsInstance(error_regexes, (list, tuple))
2032
for regex in error_regexes:
2033
self.assertContainsRe(err, regex)
2036
def run_bzr(self, args, retcode=0, stdin=None, encoding=None,
2037
working_dir=None, error_regexes=[]):
2038
"""Invoke brz, as if it were run from the command line.
2040
The argument list should not include the brz program name - the
2041
first argument is normally the brz command. Arguments may be
2042
passed in three ways:
2044
1- A list of strings, eg ["commit", "a"]. This is recommended
2045
when the command contains whitespace or metacharacters, or
2046
is built up at run time.
2048
2- A single string, eg "add a". This is the most convenient
2049
for hardcoded commands.
2051
This runs brz through the interface that catches and reports
2052
errors, and with logging set to something approximating the
2053
default, so that error reporting can be checked.
2055
This should be the main method for tests that want to exercise the
2056
overall behavior of the brz application (rather than a unit test
2057
or a functional test of the library.)
2059
This sends the stdout/stderr results into the test's log,
2060
where it may be useful for debugging. See also run_captured.
2062
:keyword stdin: A string to be used as stdin for the command.
2063
:keyword retcode: The status code the command should return;
2065
:keyword working_dir: The directory to run the command in
2066
:keyword error_regexes: A list of expected error messages. If
2067
specified they must be seen in the error output of the command.
2069
if isinstance(args, str):
2070
args = shlex.split(args)
2072
if encoding is None:
2073
encoding = osutils.get_user_encoding()
2075
stdout = ui_testing.StringIOWithEncoding()
2076
stderr = ui_testing.StringIOWithEncoding()
2077
stdout.encoding = stderr.encoding = encoding
2078
handler = logging.StreamHandler(stream=stderr)
2079
handler.setLevel(logging.INFO)
2081
logger = logging.getLogger('')
2082
logger.addHandler(handler)
2085
result = self._run_bzr_core(
2086
args, encoding=encoding, stdin=stdin, stdout=stdout,
2087
stderr=stderr, working_dir=working_dir)
2089
logger.removeHandler(handler)
2091
out = stdout.getvalue()
2092
err = stderr.getvalue()
2094
self.log('output:\n%r', out)
2096
self.log('errors:\n%r', err)
2097
if retcode is not None:
2098
self.assertEqual(retcode, result,
2099
message='Unexpected return code')
2100
self.assertIsInstance(error_regexes, (list, tuple))
2101
for regex in error_regexes:
2102
self.assertContainsRe(err, regex)
2105
def run_bzr_error(self, error_regexes, *args, **kwargs):
2106
"""Run brz, and check that stderr contains the supplied regexes
2108
:param error_regexes: Sequence of regular expressions which
2109
must each be found in the error output. The relative ordering
2111
:param args: command-line arguments for brz
2112
:param kwargs: Keyword arguments which are interpreted by run_brz
2113
This function changes the default value of retcode to be 3,
2114
since in most cases this is run when you expect brz to fail.
2116
:return: (out, err) The actual output of running the command (in case
2117
you want to do more inspection)
2121
# Make sure that commit is failing because there is nothing to do
2122
self.run_bzr_error(['no changes to commit'],
2123
['commit', '-m', 'my commit comment'])
2124
# Make sure --strict is handling an unknown file, rather than
2125
# giving us the 'nothing to do' error
2126
self.build_tree(['unknown'])
2128
['Commit refused because there are unknown files'],
2129
['commit', --strict', '-m', 'my commit comment'])
2131
kwargs.setdefault('retcode', 3)
2132
kwargs['error_regexes'] = error_regexes
2133
out, err = self.run_bzr(*args, **kwargs)
2136
def run_bzr_subprocess(self, *args, **kwargs):
2137
"""Run brz in a subprocess for testing.
2139
This starts a new Python interpreter and runs brz in there.
2140
This should only be used for tests that have a justifiable need for
2141
this isolation: e.g. they are testing startup time, or signal
2142
handling, or early startup code, etc. Subprocess code can't be
2143
profiled or debugged so easily.
2145
:keyword retcode: The status code that is expected. Defaults to 0. If
2146
None is supplied, the status code is not checked.
2147
:keyword env_changes: A dictionary which lists changes to environment
2148
variables. A value of None will unset the env variable.
2149
The values must be strings. The change will only occur in the
2150
child, so you don't need to fix the environment after running.
2151
:keyword universal_newlines: Convert CRLF => LF
2152
:keyword allow_plugins: By default the subprocess is run with
2153
--no-plugins to ensure test reproducibility. Also, it is possible
2154
for system-wide plugins to create unexpected output on stderr,
2155
which can cause unnecessary test failures.
2157
env_changes = kwargs.get('env_changes', {})
2158
working_dir = kwargs.get('working_dir', None)
2159
allow_plugins = kwargs.get('allow_plugins', False)
2161
if isinstance(args[0], list):
2163
elif isinstance(args[0], str):
2164
args = list(shlex.split(args[0]))
2166
raise ValueError("passing varargs to run_bzr_subprocess")
2167
process = self.start_bzr_subprocess(args, env_changes=env_changes,
2168
working_dir=working_dir,
2169
allow_plugins=allow_plugins)
2170
# We distinguish between retcode=None and retcode not passed.
2171
supplied_retcode = kwargs.get('retcode', 0)
2172
return self.finish_bzr_subprocess(process, retcode=supplied_retcode,
2173
universal_newlines=kwargs.get(
2174
'universal_newlines', False),
2177
def start_bzr_subprocess(self, process_args, env_changes=None,
2178
skip_if_plan_to_signal=False,
2180
allow_plugins=False, stderr=subprocess.PIPE):
2181
"""Start brz in a subprocess for testing.
2183
This starts a new Python interpreter and runs brz in there.
2184
This should only be used for tests that have a justifiable need for
2185
this isolation: e.g. they are testing startup time, or signal
2186
handling, or early startup code, etc. Subprocess code can't be
2187
profiled or debugged so easily.
2189
:param process_args: a list of arguments to pass to the brz executable,
2190
for example ``['--version']``.
2191
:param env_changes: A dictionary which lists changes to environment
2192
variables. A value of None will unset the env variable.
2193
The values must be strings. The change will only occur in the
2194
child, so you don't need to fix the environment after running.
2195
:param skip_if_plan_to_signal: raise TestSkipped when true and system
2196
doesn't support signalling subprocesses.
2197
:param allow_plugins: If False (default) pass --no-plugins to brz.
2198
:param stderr: file to use for the subprocess's stderr. Valid values
2199
are those valid for the stderr argument of `subprocess.Popen`.
2200
Default value is ``subprocess.PIPE``.
2202
:returns: Popen object for the started process.
2204
if skip_if_plan_to_signal:
2205
if os.name != "posix":
2206
raise TestSkipped("Sending signals not supported")
2208
if env_changes is None:
2210
# Because $HOME is set to a tempdir for the context of a test, modules
2211
# installed in the user dir will not be found unless $PYTHONUSERBASE
2212
# gets set to the computed directory of this parent process.
2213
if site.USER_BASE is not None:
2214
env_changes["PYTHONUSERBASE"] = site.USER_BASE
2217
def cleanup_environment():
2218
for env_var, value in env_changes.items():
2219
old_env[env_var] = osutils.set_or_unset_env(env_var, value)
2221
def restore_environment():
2222
for env_var, value in old_env.items():
2223
osutils.set_or_unset_env(env_var, value)
2225
bzr_path = self.get_brz_path()
2228
if working_dir is not None:
2229
cwd = osutils.getcwd()
2230
os.chdir(working_dir)
2233
# win32 subprocess doesn't support preexec_fn
2234
# so we will avoid using it on all platforms, just to
2235
# make sure the code path is used, and we don't break on win32
2236
cleanup_environment()
2237
# Include the subprocess's log file in the test details, in case
2238
# the test fails due to an error in the subprocess.
2239
self._add_subprocess_log(trace._get_brz_log_filename())
2240
command = [sys.executable]
2241
# frozen executables don't need the path to bzr
2242
if getattr(sys, "frozen", None) is None:
2243
command.append(bzr_path)
2244
if not allow_plugins:
2245
command.append('--no-plugins')
2246
command.extend(process_args)
2247
process = self._popen(command, stdin=subprocess.PIPE,
2248
stdout=subprocess.PIPE,
2249
stderr=stderr, bufsize=0)
2251
restore_environment()
2257
def _add_subprocess_log(self, log_file_path):
2258
if len(self._log_files) == 0:
2259
# Register an addCleanup func. We do this on the first call to
2260
# _add_subprocess_log rather than in TestCase.setUp so that this
2261
# addCleanup is registered after any cleanups for tempdirs that
2262
# subclasses might create, which will probably remove the log file
2264
self.addCleanup(self._subprocess_log_cleanup)
2265
# self._log_files is a set, so if a log file is reused we won't grab it
2267
self._log_files.add(log_file_path)
2269
def _subprocess_log_cleanup(self):
2270
for count, log_file_path in enumerate(self._log_files):
2271
# We use buffer_now=True to avoid holding the file open beyond
2272
# the life of this function, which might interfere with e.g.
2273
# cleaning tempdirs on Windows.
2274
# XXX: Testtools 0.9.5 doesn't have the content_from_file helper
2275
# detail_content = content.content_from_file(
2276
# log_file_path, buffer_now=True)
2277
with open(log_file_path, 'rb') as log_file:
2278
log_file_bytes = log_file.read()
2279
detail_content = content.Content(
2280
content.ContentType("text", "plain", {"charset": "utf8"}),
2281
lambda: [log_file_bytes])
2282
self.addDetail("start_bzr_subprocess-log-%d" % (count,),
2285
def _popen(self, *args, **kwargs):
2286
"""Place a call to Popen.
2288
Allows tests to override this method to intercept the calls made to
2289
Popen for introspection.
2291
return subprocess.Popen(*args, **kwargs)
2293
def get_source_path(self):
2294
"""Return the path of the directory containing breezy."""
2295
return os.path.dirname(os.path.dirname(breezy.__file__))
2297
def get_brz_path(self):
2298
"""Return the path of the 'brz' executable for this test suite."""
2299
brz_path = os.path.join(self.get_source_path(), "brz")
2300
if not os.path.isfile(brz_path):
2301
# We are probably installed. Assume sys.argv is the right file
2302
brz_path = sys.argv[0]
2305
def finish_bzr_subprocess(self, process, retcode=0, send_signal=None,
2306
universal_newlines=False, process_args=None):
2307
"""Finish the execution of process.
2309
:param process: the Popen object returned from start_bzr_subprocess.
2310
:param retcode: The status code that is expected. Defaults to 0. If
2311
None is supplied, the status code is not checked.
2312
:param send_signal: an optional signal to send to the process.
2313
:param universal_newlines: Convert CRLF => LF
2314
:returns: (stdout, stderr)
2316
if send_signal is not None:
2317
os.kill(process.pid, send_signal)
2318
out, err = process.communicate()
2320
if universal_newlines:
2321
out = out.replace(b'\r\n', b'\n')
2322
err = err.replace(b'\r\n', b'\n')
2324
if retcode is not None and retcode != process.returncode:
2325
if process_args is None:
2326
process_args = "(unknown args)"
2327
trace.mutter('Output of brz %r:\n%s', process_args, out)
2328
trace.mutter('Error for brz %r:\n%s', process_args, err)
2329
self.fail('Command brz %r failed with retcode %d != %d'
2330
% (process_args, retcode, process.returncode))
2333
def check_tree_shape(self, tree, shape):
2334
"""Compare a tree to a list of expected names.
2336
Fail if they are not precisely equal.
2339
shape = list(shape) # copy
2340
for path, ie in tree.iter_entries_by_dir():
2341
name = path.replace('\\', '/')
2342
if ie.kind == 'directory':
2345
pass # ignore root entry
2351
self.fail("expected paths not found in inventory: %r" % shape)
2353
self.fail("unexpected paths found in inventory: %r" % extras)
2355
def apply_redirected(self, stdin=None, stdout=None, stderr=None,
2356
a_callable=None, *args, **kwargs):
2357
"""Call callable with redirected std io pipes.
2359
Returns the return code."""
2360
if not callable(a_callable):
2361
raise ValueError("a_callable must be callable.")
2363
stdin = BytesIO(b"")
2365
if getattr(self, "_log_file", None) is not None:
2366
stdout = self._log_file
2370
if getattr(self, "_log_file", None is not None):
2371
stderr = self._log_file
2374
real_stdin = sys.stdin
2375
real_stdout = sys.stdout
2376
real_stderr = sys.stderr
2381
return a_callable(*args, **kwargs)
2383
sys.stdout = real_stdout
2384
sys.stderr = real_stderr
2385
sys.stdin = real_stdin
2387
def reduceLockdirTimeout(self):
2388
"""Reduce the default lock timeout for the duration of the test, so that
2389
if LockContention occurs during a test, it does so quickly.
2391
Tests that expect to provoke LockContention errors should call this.
2393
self.overrideAttr(lockdir, '_DEFAULT_TIMEOUT_SECONDS', 0)
2395
def make_utf8_encoded_stringio(self, encoding_type=None):
2396
"""Return a wrapped BytesIO, that will encode text input to UTF-8."""
2397
if encoding_type is None:
2398
encoding_type = 'strict'
2400
output_encoding = 'utf-8'
2401
sio = codecs.getwriter(output_encoding)(bio, errors=encoding_type)
2402
sio.encoding = output_encoding
2405
def disable_verb(self, verb):
2406
"""Disable a smart server verb for one test."""
2407
from breezy.bzr.smart import request
2408
request_handlers = request.request_handlers
2409
orig_method = request_handlers.get(verb)
2410
orig_info = request_handlers.get_info(verb)
2411
request_handlers.remove(verb)
2412
self.addCleanup(request_handlers.register, verb, orig_method,
2419
class CapturedCall(object):
2420
"""A helper for capturing smart server calls for easy debug analysis."""
2422
def __init__(self, params, prefix_length):
2423
"""Capture the call with params and skip prefix_length stack frames."""
2426
# The last 5 frames are the __init__, the hook frame, and 3 smart
2427
# client frames. Beyond this we could get more clever, but this is good
2429
stack = traceback.extract_stack()[prefix_length:-5]
2430
self.stack = ''.join(traceback.format_list(stack))
2433
return self.call.method.decode('utf-8')
2436
return self.call.method.decode('utf-8')
2442
class TestCaseWithMemoryTransport(TestCase):
2443
"""Common test class for tests that do not need disk resources.
2445
Tests that need disk resources should derive from TestCaseInTempDir
2446
orTestCaseWithTransport.
2448
TestCaseWithMemoryTransport sets the TEST_ROOT variable for all brz tests.
2450
For TestCaseWithMemoryTransport the ``test_home_dir`` is set to the name of
2451
a directory which does not exist. This serves to help ensure test isolation
2452
is preserved. ``test_dir`` is set to the TEST_ROOT, as is cwd, because they
2453
must exist. However, TestCaseWithMemoryTransport does not offer local file
2454
defaults for the transport in tests, nor does it obey the command line
2455
override, so tests that accidentally write to the common directory should
2458
:cvar TEST_ROOT: Directory containing all temporary directories, plus a
2459
``.bzr`` directory that stops us ascending higher into the filesystem.
2465
def __init__(self, methodName='runTest'):
2466
# allow test parameterization after test construction and before test
2467
# execution. Variables that the parameterizer sets need to be
2468
# ones that are not set by setUp, or setUp will trash them.
2469
super(TestCaseWithMemoryTransport, self).__init__(methodName)
2470
self.vfs_transport_factory = default_transport
2471
self.transport_server = None
2472
self.transport_readonly_server = None
2473
self.__vfs_server = None
2476
super(TestCaseWithMemoryTransport, self).setUp()
2478
def _add_disconnect_cleanup(transport):
2479
"""Schedule disconnection of given transport at test cleanup
2481
This needs to happen for all connected transports or leaks occur.
2483
Note reconnections may mean we call disconnect multiple times per
2484
transport which is suboptimal but seems harmless.
2486
self.addCleanup(transport.disconnect)
2488
_mod_transport.Transport.hooks.install_named_hook(
2489
'post_connect', _add_disconnect_cleanup, None)
2491
self._make_test_root()
2492
self.addCleanup(os.chdir, osutils.getcwd())
2493
self.makeAndChdirToTestDir()
2494
self.overrideEnvironmentForTesting()
2495
self.__readonly_server = None
2496
self.__server = None
2497
self.reduceLockdirTimeout()
2498
# Each test may use its own config files even if the local config files
2499
# don't actually exist. They'll rightly fail if they try to create them
2501
self.overrideAttr(config, '_shared_stores', {})
2503
def get_transport(self, relpath=None):
2504
"""Return a writeable transport.
2506
This transport is for the test scratch space relative to
2509
:param relpath: a path relative to the base url.
2511
t = _mod_transport.get_transport_from_url(self.get_url(relpath))
2512
self.assertFalse(t.is_readonly())
2515
def get_readonly_transport(self, relpath=None):
2516
"""Return a readonly transport for the test scratch space
2518
This can be used to test that operations which should only need
2519
readonly access in fact do not try to write.
2521
:param relpath: a path relative to the base url.
2523
t = _mod_transport.get_transport_from_url(
2524
self.get_readonly_url(relpath))
2525
self.assertTrue(t.is_readonly())
2528
def create_transport_readonly_server(self):
2529
"""Create a transport server from class defined at init.
2531
This is mostly a hook for daughter classes.
2533
return self.transport_readonly_server()
2535
def get_readonly_server(self):
2536
"""Get the server instance for the readonly transport
2538
This is useful for some tests with specific servers to do diagnostics.
2540
if self.__readonly_server is None:
2541
if self.transport_readonly_server is None:
2542
# readonly decorator requested
2543
self.__readonly_server = test_server.ReadonlyServer()
2545
# explicit readonly transport.
2546
self.__readonly_server = (
2547
self.create_transport_readonly_server())
2548
self.start_server(self.__readonly_server,
2549
self.get_vfs_only_server())
2550
return self.__readonly_server
2552
def get_readonly_url(self, relpath=None):
2553
"""Get a URL for the readonly transport.
2555
This will either be backed by '.' or a decorator to the transport
2556
used by self.get_url()
2557
relpath provides for clients to get a path relative to the base url.
2558
These should only be downwards relative, not upwards.
2560
base = self.get_readonly_server().get_url()
2561
return self._adjust_url(base, relpath)
2563
def get_vfs_only_server(self):
2564
"""Get the vfs only read/write server instance.
2566
This is useful for some tests with specific servers that need
2569
For TestCaseWithMemoryTransport this is always a MemoryServer, and there
2570
is no means to override it.
2572
if self.__vfs_server is None:
2573
self.__vfs_server = memory.MemoryServer()
2574
self.start_server(self.__vfs_server)
2575
return self.__vfs_server
2577
def get_server(self):
2578
"""Get the read/write server instance.
2580
This is useful for some tests with specific servers that need
2583
This is built from the self.transport_server factory. If that is None,
2584
then the self.get_vfs_server is returned.
2586
if self.__server is None:
2587
if (self.transport_server is None or self.transport_server is
2588
self.vfs_transport_factory):
2589
self.__server = self.get_vfs_only_server()
2591
# bring up a decorated means of access to the vfs only server.
2592
self.__server = self.transport_server()
2593
self.start_server(self.__server, self.get_vfs_only_server())
2594
return self.__server
2596
def _adjust_url(self, base, relpath):
2597
"""Get a URL (or maybe a path) for the readwrite transport.
2599
This will either be backed by '.' or to an equivalent non-file based
2601
relpath provides for clients to get a path relative to the base url.
2602
These should only be downwards relative, not upwards.
2604
if relpath is not None and relpath != '.':
2605
if not base.endswith('/'):
2607
# XXX: Really base should be a url; we did after all call
2608
# get_url()! But sometimes it's just a path (from
2609
# LocalAbspathServer), and it'd be wrong to append urlescaped data
2610
# to a non-escaped local path.
2611
if base.startswith('./') or base.startswith('/'):
2614
base += urlutils.escape(relpath)
2617
def get_url(self, relpath=None):
2618
"""Get a URL (or maybe a path) for the readwrite transport.
2620
This will either be backed by '.' or to an equivalent non-file based
2622
relpath provides for clients to get a path relative to the base url.
2623
These should only be downwards relative, not upwards.
2625
base = self.get_server().get_url()
2626
return self._adjust_url(base, relpath)
2628
def get_vfs_only_url(self, relpath=None):
2629
"""Get a URL (or maybe a path for the plain old vfs transport.
2631
This will never be a smart protocol. It always has all the
2632
capabilities of the local filesystem, but it might actually be a
2633
MemoryTransport or some other similar virtual filesystem.
2635
This is the backing transport (if any) of the server returned by
2636
get_url and get_readonly_url.
2638
:param relpath: provides for clients to get a path relative to the base
2639
url. These should only be downwards relative, not upwards.
2642
base = self.get_vfs_only_server().get_url()
2643
return self._adjust_url(base, relpath)
2645
def _create_safety_net(self):
2646
"""Make a fake bzr directory.
2648
This prevents any tests propagating up onto the TEST_ROOT directory's
2651
root = TestCaseWithMemoryTransport.TEST_ROOT
2653
# Make sure we get a readable and accessible home for brz.log
2654
# and/or config files, and not fallback to weird defaults (see
2655
# http://pad.lv/825027).
2656
self.assertIs(None, os.environ.get('BRZ_HOME', None))
2657
os.environ['BRZ_HOME'] = root
2658
from breezy.bzr.bzrdir import BzrDirMetaFormat1
2659
wt = controldir.ControlDir.create_standalone_workingtree(
2660
root, format=BzrDirMetaFormat1())
2661
del os.environ['BRZ_HOME']
2662
except Exception as e:
2663
self.fail("Fail to initialize the safety net: %r\n" % (e,))
2664
# Hack for speed: remember the raw bytes of the dirstate file so that
2665
# we don't need to re-open the wt to check it hasn't changed.
2666
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE = (
2667
wt.control_transport.get_bytes('dirstate'))
2669
def _check_safety_net(self):
2670
"""Check that the safety .bzr directory have not been touched.
2672
_make_test_root have created a .bzr directory to prevent tests from
2673
propagating. This method ensures than a test did not leaked.
2675
root = TestCaseWithMemoryTransport.TEST_ROOT
2676
t = _mod_transport.get_transport_from_path(root)
2677
self.permit_url(t.base)
2678
if (t.get_bytes('.bzr/checkout/dirstate') !=
2679
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE):
2680
# The current test have modified the /bzr directory, we need to
2681
# recreate a new one or all the followng tests will fail.
2682
# If you need to inspect its content uncomment the following line
2683
# import pdb; pdb.set_trace()
2684
_rmtree_temp_dir(root + '/.bzr', test_id=self.id())
2685
self._create_safety_net()
2686
raise AssertionError('%s/.bzr should not be modified' % root)
2688
def _make_test_root(self):
2689
if TestCaseWithMemoryTransport.TEST_ROOT is None:
2690
# Watch out for tricky test dir (on OSX /tmp -> /private/tmp)
2691
root = osutils.realpath(osutils.mkdtemp(prefix='testbzr-',
2693
TestCaseWithMemoryTransport.TEST_ROOT = root
2695
self._create_safety_net()
2697
# The same directory is used by all tests, and we're not
2698
# specifically told when all tests are finished. This will do.
2699
atexit.register(_rmtree_temp_dir, root)
2701
self.permit_dir(TestCaseWithMemoryTransport.TEST_ROOT)
2702
self.addCleanup(self._check_safety_net)
2704
def makeAndChdirToTestDir(self):
2705
"""Create a temporary directories for this one test.
2707
This must set self.test_home_dir and self.test_dir and chdir to
2710
For TestCaseWithMemoryTransport we chdir to the TEST_ROOT for this
2713
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2714
self.test_dir = TestCaseWithMemoryTransport.TEST_ROOT
2715
self.test_home_dir = self.test_dir + "/MemoryTransportMissingHomeDir"
2716
self.permit_dir(self.test_dir)
2718
def make_branch(self, relpath, format=None, name=None):
2719
"""Create a branch on the transport at relpath."""
2720
repo = self.make_repository(relpath, format=format)
2721
return repo.controldir.create_branch(append_revisions_only=False, name=name)
2723
def get_default_format(self):
2726
def resolve_format(self, format):
2727
"""Resolve an object to a ControlDir format object.
2729
The initial format object can either already be
2730
a ControlDirFormat, None (for the default format),
2731
or a string with the name of the control dir format.
2733
:param format: Object to resolve
2734
:return A ControlDirFormat instance
2737
format = self.get_default_format()
2738
if isinstance(format, str):
2739
format = controldir.format_registry.make_controldir(format)
2742
def make_controldir(self, relpath, format=None):
2744
# might be a relative or absolute path
2745
maybe_a_url = self.get_url(relpath)
2746
segments = maybe_a_url.rsplit('/', 1)
2747
t = _mod_transport.get_transport(maybe_a_url)
2748
if len(segments) > 1 and segments[-1] not in ('', '.'):
2750
format = self.resolve_format(format)
2751
return format.initialize_on_transport(t)
2752
except errors.UninitializableFormat:
2753
raise TestSkipped("Format %s is not initializable." % format)
2755
def make_repository(self, relpath, shared=None, format=None):
2756
"""Create a repository on our default transport at relpath.
2758
Note that relpath must be a relative path, not a full url.
2760
# FIXME: If you create a remoterepository this returns the underlying
2761
# real format, which is incorrect. Actually we should make sure that
2762
# RemoteBzrDir returns a RemoteRepository.
2763
# maybe mbp 20070410
2764
made_control = self.make_controldir(relpath, format=format)
2765
return made_control.create_repository(shared=shared)
2767
def make_smart_server(self, path, backing_server=None):
2768
if backing_server is None:
2769
backing_server = self.get_server()
2770
smart_server = test_server.SmartTCPServer_for_testing()
2771
self.start_server(smart_server, backing_server)
2772
remote_transport = _mod_transport.get_transport_from_url(smart_server.get_url()
2774
return remote_transport
2776
def make_branch_and_memory_tree(self, relpath, format=None):
2777
"""Create a branch on the default transport and a MemoryTree for it."""
2778
b = self.make_branch(relpath, format=format)
2779
return b.create_memorytree()
2781
def make_branch_builder(self, relpath, format=None):
2782
branch = self.make_branch(relpath, format=format)
2783
return branchbuilder.BranchBuilder(branch=branch)
2785
def overrideEnvironmentForTesting(self):
2786
test_home_dir = self.test_home_dir
2787
self.overrideEnv('HOME', test_home_dir)
2788
self.overrideEnv('BRZ_HOME', test_home_dir)
2789
self.overrideEnv('GNUPGHOME', os.path.join(test_home_dir, '.gnupg'))
2791
def setup_smart_server_with_call_log(self):
2792
"""Sets up a smart server as the transport server with a call log."""
2793
self.transport_server = test_server.SmartTCPServer_for_testing
2794
self.hpss_connections = []
2795
self.hpss_calls = []
2797
# Skip the current stack down to the caller of
2798
# setup_smart_server_with_call_log
2799
prefix_length = len(traceback.extract_stack()) - 2
2801
def capture_hpss_call(params):
2802
self.hpss_calls.append(
2803
CapturedCall(params, prefix_length))
2805
def capture_connect(transport):
2806
self.hpss_connections.append(transport)
2807
client._SmartClient.hooks.install_named_hook(
2808
'call', capture_hpss_call, None)
2809
_mod_transport.Transport.hooks.install_named_hook(
2810
'post_connect', capture_connect, None)
2812
def reset_smart_call_log(self):
2813
self.hpss_calls = []
2814
self.hpss_connections = []
2817
class TestCaseInTempDir(TestCaseWithMemoryTransport):
2818
"""Derived class that runs a test within a temporary directory.
2820
This is useful for tests that need to create a branch, etc.
2822
The directory is created in a slightly complex way: for each
2823
Python invocation, a new temporary top-level directory is created.
2824
All test cases create their own directory within that. If the
2825
tests complete successfully, the directory is removed.
2827
:ivar test_base_dir: The path of the top-level directory for this
2828
test, which contains a home directory and a work directory.
2830
:ivar test_home_dir: An initially empty directory under test_base_dir
2831
which is used as $HOME for this test.
2833
:ivar test_dir: A directory under test_base_dir used as the current
2834
directory when the test proper is run.
2837
OVERRIDE_PYTHON = 'python'
2840
super(TestCaseInTempDir, self).setUp()
2841
# Remove the protection set in isolated_environ, we have a proper
2842
# access to disk resources now.
2843
self.overrideEnv('BRZ_LOG', None)
2845
def check_file_contents(self, filename, expect):
2846
self.log("check contents of file %s" % filename)
2847
with open(filename, 'rb') as f:
2849
if contents != expect:
2850
self.log("expected: %r" % expect)
2851
self.log("actually: %r" % contents)
2852
self.fail("contents of %s not as expected" % filename)
2854
def _getTestDirPrefix(self):
2855
# create a directory within the top level test directory
2856
if sys.platform in ('win32', 'cygwin'):
2857
name_prefix = re.sub('[<>*=+",:;_/\\-]', '_', self.id())
2858
# windows is likely to have path-length limits so use a short name
2859
name_prefix = name_prefix[-30:]
2861
name_prefix = re.sub('[/]', '_', self.id())
2864
def makeAndChdirToTestDir(self):
2865
"""See TestCaseWithMemoryTransport.makeAndChdirToTestDir().
2867
For TestCaseInTempDir we create a temporary directory based on the test
2868
name and then create two subdirs - test and home under it.
2870
name_prefix = osutils.pathjoin(TestCaseWithMemoryTransport.TEST_ROOT,
2871
self._getTestDirPrefix())
2873
for i in range(100):
2874
if os.path.exists(name):
2875
name = name_prefix + '_' + str(i)
2877
# now create test and home directories within this dir
2878
self.test_base_dir = name
2879
self.addCleanup(self.deleteTestDir)
2880
os.mkdir(self.test_base_dir)
2882
self.permit_dir(self.test_base_dir)
2883
# 'sprouting' and 'init' of a branch both walk up the tree to find
2884
# stacking policy to honour; create a bzr dir with an unshared
2885
# repository (but not a branch - our code would be trying to escape
2886
# then!) to stop them, and permit it to be read.
2887
# control = controldir.ControlDir.create(self.test_base_dir)
2888
# control.create_repository()
2889
self.test_home_dir = self.test_base_dir + '/home'
2890
os.mkdir(self.test_home_dir)
2891
self.test_dir = self.test_base_dir + '/work'
2892
os.mkdir(self.test_dir)
2893
os.chdir(self.test_dir)
2894
# put name of test inside
2895
with open(self.test_base_dir + '/name', 'w') as f:
2898
def deleteTestDir(self):
2899
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2900
_rmtree_temp_dir(self.test_base_dir, test_id=self.id())
2902
def build_tree(self, shape, line_endings='binary', transport=None):
2903
"""Build a test tree according to a pattern.
2905
shape is a sequence of file specifications. If the final
2906
character is '/', a directory is created.
2908
This assumes that all the elements in the tree being built are new.
2910
This doesn't add anything to a branch.
2912
:type shape: list or tuple.
2913
:param line_endings: Either 'binary' or 'native'
2914
in binary mode, exact contents are written in native mode, the
2915
line endings match the default platform endings.
2916
:param transport: A transport to write to, for building trees on VFS's.
2917
If the transport is readonly or None, "." is opened automatically.
2920
if type(shape) not in (list, tuple):
2921
raise AssertionError("Parameter 'shape' should be "
2922
"a list or a tuple. Got %r instead" % (shape,))
2923
# It's OK to just create them using forward slashes on windows.
2924
if transport is None or transport.is_readonly():
2925
transport = _mod_transport.get_transport_from_path(".")
2927
self.assertIsInstance(name, str)
2929
transport.mkdir(urlutils.escape(name[:-1]))
2931
if line_endings == 'binary':
2933
elif line_endings == 'native':
2934
end = os.linesep.encode('ascii')
2936
raise errors.BzrError(
2937
'Invalid line ending request %r' % line_endings)
2938
content = b"contents of %s%s" % (name.encode('utf-8'), end)
2939
transport.put_bytes_non_atomic(urlutils.escape(name), content)
2941
build_tree_contents = staticmethod(treeshape.build_tree_contents)
2943
def assertInWorkingTree(self, path, root_path='.', tree=None):
2944
"""Assert whether path or paths are in the WorkingTree"""
2946
tree = workingtree.WorkingTree.open(root_path)
2947
if not isinstance(path, str):
2949
self.assertInWorkingTree(p, tree=tree)
2951
self.assertTrue(tree.is_versioned(path),
2952
path + ' not in working tree.')
2954
def assertNotInWorkingTree(self, path, root_path='.', tree=None):
2955
"""Assert whether path or paths are not in the WorkingTree"""
2957
tree = workingtree.WorkingTree.open(root_path)
2958
if not isinstance(path, str):
2960
self.assertNotInWorkingTree(p, tree=tree)
2962
self.assertFalse(tree.is_versioned(
2963
path), path + ' in working tree.')
2966
class TestCaseWithTransport(TestCaseInTempDir):
2967
"""A test case that provides get_url and get_readonly_url facilities.
2969
These back onto two transport servers, one for readonly access and one for
2972
If no explicit class is provided for readonly access, a
2973
ReadonlyTransportDecorator is used instead which allows the use of non disk
2974
based read write transports.
2976
If an explicit class is provided for readonly access, that server and the
2977
readwrite one must both define get_url() as resolving to os.getcwd().
2981
super(TestCaseWithTransport, self).setUp()
2982
self.__vfs_server = None
2984
def get_vfs_only_server(self):
2985
"""See TestCaseWithMemoryTransport.
2987
This is useful for some tests with specific servers that need
2990
if self.__vfs_server is None:
2991
self.__vfs_server = self.vfs_transport_factory()
2992
self.start_server(self.__vfs_server)
2993
return self.__vfs_server
2995
def make_branch_and_tree(self, relpath, format=None):
2996
"""Create a branch on the transport and a tree locally.
2998
If the transport is not a LocalTransport, the Tree can't be created on
2999
the transport. In that case if the vfs_transport_factory is
3000
LocalURLServer the working tree is created in the local
3001
directory backing the transport, and the returned tree's branch and
3002
repository will also be accessed locally. Otherwise a lightweight
3003
checkout is created and returned.
3005
We do this because we can't physically create a tree in the local
3006
path, with a branch reference to the transport_factory url, and
3007
a branch + repository in the vfs_transport, unless the vfs_transport
3008
namespace is distinct from the local disk - the two branch objects
3009
would collide. While we could construct a tree with its branch object
3010
pointing at the transport_factory transport in memory, reopening it
3011
would behaving unexpectedly, and has in the past caused testing bugs
3012
when we tried to do it that way.
3014
:param format: The BzrDirFormat.
3015
:returns: the WorkingTree.
3017
# TODO: always use the local disk path for the working tree,
3018
# this obviously requires a format that supports branch references
3019
# so check for that by checking bzrdir.BzrDirFormat.get_default_format()
3021
format = self.resolve_format(format=format)
3022
if not format.supports_workingtrees:
3023
b = self.make_branch(relpath + '.branch', format=format)
3024
return b.create_checkout(relpath, lightweight=True)
3025
b = self.make_branch(relpath, format=format)
3027
return b.controldir.create_workingtree()
3028
except errors.NotLocalUrl:
3029
# We can only make working trees locally at the moment. If the
3030
# transport can't support them, then we keep the non-disk-backed
3031
# branch and create a local checkout.
3032
if self.vfs_transport_factory is test_server.LocalURLServer:
3033
# the branch is colocated on disk, we cannot create a checkout.
3034
# hopefully callers will expect this.
3035
local_controldir = controldir.ControlDir.open(
3036
self.get_vfs_only_url(relpath))
3037
wt = local_controldir.create_workingtree()
3038
if wt.branch._format != b._format:
3040
# Make sure that assigning to wt._branch fixes wt.branch,
3041
# in case the implementation details of workingtree objects
3043
self.assertIs(b, wt.branch)
3046
return b.create_checkout(relpath, lightweight=True)
3048
def assertIsDirectory(self, relpath, transport):
3049
"""Assert that relpath within transport is a directory.
3051
This may not be possible on all transports; in that case it propagates
3052
a TransportNotPossible.
3055
mode = transport.stat(relpath).st_mode
3056
except errors.NoSuchFile:
3057
self.fail("path %s is not a directory; no such file"
3059
if not stat.S_ISDIR(mode):
3060
self.fail("path %s is not a directory; has mode %#o"
3063
def assertTreesEqual(self, left, right):
3064
"""Check that left and right have the same content and properties."""
3065
# we use a tree delta to check for equality of the content, and we
3066
# manually check for equality of other things such as the parents list.
3067
self.assertEqual(left.get_parent_ids(), right.get_parent_ids())
3068
differences = left.changes_from(right)
3069
self.assertFalse(differences.has_changed(),
3070
"Trees %r and %r are different: %r" % (left, right, differences))
3072
def disable_missing_extensions_warning(self):
3073
"""Some tests expect a precise stderr content.
3075
There is no point in forcing them to duplicate the extension related
3078
config.GlobalConfig().set_user_option(
3079
'suppress_warnings', 'missing_extensions')
3082
class ChrootedTestCase(TestCaseWithTransport):
3083
"""A support class that provides readonly urls outside the local namespace.
3085
This is done by checking if self.transport_server is a MemoryServer. if it
3086
is then we are chrooted already, if it is not then an HttpServer is used
3089
TODO RBC 20060127: make this an option to TestCaseWithTransport so it can
3090
be used without needed to redo it when a different
3091
subclass is in use ?
3095
from breezy.tests import http_server
3096
super(ChrootedTestCase, self).setUp()
3097
if not self.vfs_transport_factory == memory.MemoryServer:
3098
self.transport_readonly_server = http_server.HttpServer
3101
def condition_id_re(pattern):
3102
"""Create a condition filter which performs a re check on a test's id.
3104
:param pattern: A regular expression string.
3105
:return: A callable that returns True if the re matches.
3107
filter_re = re.compile(pattern, 0)
3109
def condition(test):
3111
return filter_re.search(test_id)
3115
def condition_isinstance(klass_or_klass_list):
3116
"""Create a condition filter which returns isinstance(param, klass).
3118
:return: A callable which when called with one parameter obj return the
3119
result of isinstance(obj, klass_or_klass_list).
3122
return isinstance(obj, klass_or_klass_list)
3126
def condition_id_in_list(id_list):
3127
"""Create a condition filter which verify that test's id in a list.
3129
:param id_list: A TestIdList object.
3130
:return: A callable that returns True if the test's id appears in the list.
3132
def condition(test):
3133
return id_list.includes(test.id())
3137
def condition_id_startswith(starts):
3138
"""Create a condition filter verifying that test's id starts with a string.
3140
:param starts: A list of string.
3141
:return: A callable that returns True if the test's id starts with one of
3144
def condition(test):
3145
for start in starts:
3146
if test.id().startswith(start):
3152
def exclude_tests_by_condition(suite, condition):
3153
"""Create a test suite which excludes some tests from suite.
3155
:param suite: The suite to get tests from.
3156
:param condition: A callable whose result evaluates True when called with a
3157
test case which should be excluded from the result.
3158
:return: A suite which contains the tests found in suite that fail
3162
for test in iter_suite_tests(suite):
3163
if not condition(test):
3165
return TestUtil.TestSuite(result)
3168
def filter_suite_by_condition(suite, condition):
3169
"""Create a test suite by filtering another one.
3171
:param suite: The source suite.
3172
:param condition: A callable whose result evaluates True when called with a
3173
test case which should be included in the result.
3174
:return: A suite which contains the tests found in suite that pass
3178
for test in iter_suite_tests(suite):
3181
return TestUtil.TestSuite(result)
3184
def filter_suite_by_re(suite, pattern):
3185
"""Create a test suite by filtering another one.
3187
:param suite: the source suite
3188
:param pattern: pattern that names must match
3189
:returns: the newly created suite
3191
condition = condition_id_re(pattern)
3192
result_suite = filter_suite_by_condition(suite, condition)
3196
def filter_suite_by_id_list(suite, test_id_list):
3197
"""Create a test suite by filtering another one.
3199
:param suite: The source suite.
3200
:param test_id_list: A list of the test ids to keep as strings.
3201
:returns: the newly created suite
3203
condition = condition_id_in_list(test_id_list)
3204
result_suite = filter_suite_by_condition(suite, condition)
3208
def filter_suite_by_id_startswith(suite, start):
3209
"""Create a test suite by filtering another one.
3211
:param suite: The source suite.
3212
:param start: A list of string the test id must start with one of.
3213
:returns: the newly created suite
3215
condition = condition_id_startswith(start)
3216
result_suite = filter_suite_by_condition(suite, condition)
3220
def exclude_tests_by_re(suite, pattern):
3221
"""Create a test suite which excludes some tests from suite.
3223
:param suite: The suite to get tests from.
3224
:param pattern: A regular expression string. Test ids that match this
3225
pattern will be excluded from the result.
3226
:return: A TestSuite that contains all the tests from suite without the
3227
tests that matched pattern. The order of tests is the same as it was in
3230
return exclude_tests_by_condition(suite, condition_id_re(pattern))
3233
def preserve_input(something):
3234
"""A helper for performing test suite transformation chains.
3236
:param something: Anything you want to preserve.
3242
def randomize_suite(suite):
3243
"""Return a new TestSuite with suite's tests in random order.
3245
The tests in the input suite are flattened into a single suite in order to
3246
accomplish this. Any nested TestSuites are removed to provide global
3249
tests = list(iter_suite_tests(suite))
3250
random.shuffle(tests)
3251
return TestUtil.TestSuite(tests)
3254
def split_suite_by_condition(suite, condition):
3255
"""Split a test suite into two by a condition.
3257
:param suite: The suite to split.
3258
:param condition: The condition to match on. Tests that match this
3259
condition are returned in the first test suite, ones that do not match
3260
are in the second suite.
3261
:return: A tuple of two test suites, where the first contains tests from
3262
suite matching the condition, and the second contains the remainder
3263
from suite. The order within each output suite is the same as it was in
3268
for test in iter_suite_tests(suite):
3270
matched.append(test)
3272
did_not_match.append(test)
3273
return TestUtil.TestSuite(matched), TestUtil.TestSuite(did_not_match)
3276
def split_suite_by_re(suite, pattern):
3277
"""Split a test suite into two by a regular expression.
3279
:param suite: The suite to split.
3280
:param pattern: A regular expression string. Test ids that match this
3281
pattern will be in the first test suite returned, and the others in the
3282
second test suite returned.
3283
:return: A tuple of two test suites, where the first contains tests from
3284
suite matching pattern, and the second contains the remainder from
3285
suite. The order within each output suite is the same as it was in
3288
return split_suite_by_condition(suite, condition_id_re(pattern))
3291
def run_suite(suite, name='test', verbose=False, pattern=".*",
3292
stop_on_failure=False,
3293
transport=None, lsprof_timed=None, bench_history=None,
3294
matching_tests_first=None,
3297
exclude_pattern=None,
3300
suite_decorators=None,
3302
result_decorators=None,
3304
"""Run a test suite for brz selftest.
3306
:param runner_class: The class of runner to use. Must support the
3307
constructor arguments passed by run_suite which are more than standard
3309
:return: A boolean indicating success.
3311
TestCase._gather_lsprof_in_benchmarks = lsprof_timed
3316
if runner_class is None:
3317
runner_class = TextTestRunner
3320
runner = runner_class(stream=stream,
3322
verbosity=verbosity,
3323
bench_history=bench_history,
3325
result_decorators=result_decorators,
3327
runner.stop_on_failure = stop_on_failure
3328
if isinstance(suite, unittest.TestSuite):
3329
# Empty out _tests list of passed suite and populate new TestSuite
3330
suite._tests[:], suite = [], TestSuite(suite)
3331
# built in decorator factories:
3333
random_order(random_seed, runner),
3334
exclude_tests(exclude_pattern),
3336
if matching_tests_first:
3337
decorators.append(tests_first(pattern))
3339
decorators.append(filter_tests(pattern))
3340
if suite_decorators:
3341
decorators.extend(suite_decorators)
3342
# tell the result object how many tests will be running: (except if
3343
# --parallel=fork is being used. Robert said he will provide a better
3344
# progress design later -- vila 20090817)
3345
if fork_decorator not in decorators:
3346
decorators.append(CountingDecorator)
3347
for decorator in decorators:
3348
suite = decorator(suite)
3350
# Done after test suite decoration to allow randomisation etc
3351
# to take effect, though that is of marginal benefit.
3353
stream.write("Listing tests only ...\n")
3354
if getattr(runner, 'list', None) is not None:
3357
for t in iter_suite_tests(suite):
3358
stream.write("%s\n" % (t.id()))
3360
result = runner.run(suite)
3361
if strict and getattr(result, 'wasStrictlySuccessful', False):
3362
return result.wasStrictlySuccessful()
3364
return result.wasSuccessful()
3367
# A registry where get() returns a suite decorator.
3368
parallel_registry = registry.Registry()
3371
def fork_decorator(suite):
3372
if getattr(os, "fork", None) is None:
3373
raise errors.CommandError("platform does not support fork,"
3374
" try --parallel=subprocess instead.")
3375
concurrency = osutils.local_concurrency()
3376
if concurrency == 1:
3378
from testtools import ConcurrentTestSuite
3379
return ConcurrentTestSuite(suite, fork_for_tests)
3382
parallel_registry.register('fork', fork_decorator)
3385
def subprocess_decorator(suite):
3386
concurrency = osutils.local_concurrency()
3387
if concurrency == 1:
3389
from testtools import ConcurrentTestSuite
3390
return ConcurrentTestSuite(suite, reinvoke_for_tests)
3393
parallel_registry.register('subprocess', subprocess_decorator)
3396
def exclude_tests(exclude_pattern):
3397
"""Return a test suite decorator that excludes tests."""
3398
if exclude_pattern is None:
3399
return identity_decorator
3401
def decorator(suite):
3402
return ExcludeDecorator(suite, exclude_pattern)
3406
def filter_tests(pattern):
3408
return identity_decorator
3410
def decorator(suite):
3411
return FilterTestsDecorator(suite, pattern)
3415
def random_order(random_seed, runner):
3416
"""Return a test suite decorator factory for randomising tests order.
3418
:param random_seed: now, a string which casts to an integer, or an integer.
3419
:param runner: A test runner with a stream attribute to report on.
3421
if random_seed is None:
3422
return identity_decorator
3424
def decorator(suite):
3425
return RandomDecorator(suite, random_seed, runner.stream)
3429
def tests_first(pattern):
3431
return identity_decorator
3433
def decorator(suite):
3434
return TestFirstDecorator(suite, pattern)
3438
def identity_decorator(suite):
3443
class TestDecorator(TestUtil.TestSuite):
3444
"""A decorator for TestCase/TestSuite objects.
3446
Contains rather than flattening suite passed on construction
3449
def __init__(self, suite=None):
3450
super(TestDecorator, self).__init__()
3451
if suite is not None:
3454
# Don't need subclass run method with suite emptying
3455
run = unittest.TestSuite.run
3458
class CountingDecorator(TestDecorator):
3459
"""A decorator which calls result.progress(self.countTestCases)."""
3461
def run(self, result):
3462
progress_method = getattr(result, 'progress', None)
3463
if callable(progress_method):
3464
progress_method(self.countTestCases(), SUBUNIT_SEEK_SET)
3465
return super(CountingDecorator, self).run(result)
3468
class ExcludeDecorator(TestDecorator):
3469
"""A decorator which excludes test matching an exclude pattern."""
3471
def __init__(self, suite, exclude_pattern):
3472
super(ExcludeDecorator, self).__init__(
3473
exclude_tests_by_re(suite, exclude_pattern))
3476
class FilterTestsDecorator(TestDecorator):
3477
"""A decorator which filters tests to those matching a pattern."""
3479
def __init__(self, suite, pattern):
3480
super(FilterTestsDecorator, self).__init__(
3481
filter_suite_by_re(suite, pattern))
3484
class RandomDecorator(TestDecorator):
3485
"""A decorator which randomises the order of its tests."""
3487
def __init__(self, suite, random_seed, stream):
3488
random_seed = self.actual_seed(random_seed)
3489
stream.write("Randomizing test order using seed %s\n\n" %
3491
# Initialise the random number generator.
3492
random.seed(random_seed)
3493
super(RandomDecorator, self).__init__(randomize_suite(suite))
3496
def actual_seed(seed):
3498
# We convert the seed to an integer to make it reuseable across
3499
# invocations (because the user can reenter it).
3500
return int(time.time())
3502
# Convert the seed to an integer if we can
3505
except (TypeError, ValueError):
3510
class TestFirstDecorator(TestDecorator):
3511
"""A decorator which moves named tests to the front."""
3513
def __init__(self, suite, pattern):
3514
super(TestFirstDecorator, self).__init__()
3515
self.addTests(split_suite_by_re(suite, pattern))
3518
def partition_tests(suite, count):
3519
"""Partition suite into count lists of tests."""
3520
# This just assigns tests in a round-robin fashion. On one hand this
3521
# splits up blocks of related tests that might run faster if they shared
3522
# resources, but on the other it avoids assigning blocks of slow tests to
3523
# just one partition. So the slowest partition shouldn't be much slower
3525
partitions = [list() for i in range(count)]
3526
tests = iter_suite_tests(suite)
3527
for partition, test in zip(itertools.cycle(partitions), tests):
3528
partition.append(test)
3532
def workaround_zealous_crypto_random():
3533
"""Crypto.Random want to help us being secure, but we don't care here.
3535
This workaround some test failure related to the sftp server. Once paramiko
3536
stop using the controversial API in Crypto.Random, we may get rid of it.
3539
from Crypto.Random import atfork
3545
def fork_for_tests(suite):
3546
"""Take suite and start up one runner per CPU by forking()
3548
:return: An iterable of TestCase-like objects which can each have
3549
run(result) called on them to feed tests to result.
3551
concurrency = osutils.local_concurrency()
3553
from subunit import ProtocolTestCase
3554
from subunit.test_results import AutoTimingTestResultDecorator
3556
class TestInOtherProcess(ProtocolTestCase):
3557
# Should be in subunit, I think. RBC.
3558
def __init__(self, stream, pid):
3559
ProtocolTestCase.__init__(self, stream)
3562
def run(self, result):
3564
ProtocolTestCase.run(self, result)
3566
pid, status = os.waitpid(self.pid, 0)
3567
# GZ 2011-10-18: If status is nonzero, should report to the result
3568
# that something went wrong.
3570
test_blocks = partition_tests(suite, concurrency)
3571
# Clear the tests from the original suite so it doesn't keep them alive
3572
suite._tests[:] = []
3573
for process_tests in test_blocks:
3574
process_suite = TestUtil.TestSuite(process_tests)
3575
# Also clear each split list so new suite has only reference
3576
process_tests[:] = []
3577
c2pread, c2pwrite = os.pipe()
3581
stream = os.fdopen(c2pwrite, 'wb', 0)
3582
workaround_zealous_crypto_random()
3588
coverage.process_startup()
3590
# Leave stderr and stdout open so we can see test noise
3591
# Close stdin so that the child goes away if it decides to
3592
# read from stdin (otherwise its a roulette to see what
3593
# child actually gets keystrokes for pdb etc).
3595
subunit_result = AutoTimingTestResultDecorator(
3596
SubUnitBzrProtocolClientv1(stream))
3597
process_suite.run(subunit_result)
3599
# Try and report traceback on stream, but exit with error even
3600
# if stream couldn't be created or something else goes wrong.
3601
# The traceback is formatted to a string and written in one go
3602
# to avoid interleaving lines from multiple failing children.
3603
tb = traceback.format_exc()
3604
if isinstance(tb, str):
3605
tb = tb.encode('utf-8')
3614
stream = os.fdopen(c2pread, 'rb', 0)
3615
test = TestInOtherProcess(stream, pid)
3620
def reinvoke_for_tests(suite):
3621
"""Take suite and start up one runner per CPU using subprocess().
3623
:return: An iterable of TestCase-like objects which can each have
3624
run(result) called on them to feed tests to result.
3626
concurrency = osutils.local_concurrency()
3628
from subunit import ProtocolTestCase
3630
class TestInSubprocess(ProtocolTestCase):
3631
def __init__(self, process, name):
3632
ProtocolTestCase.__init__(self, process.stdout)
3633
self.process = process
3634
self.process.stdin.close()
3637
def run(self, result):
3639
ProtocolTestCase.run(self, result)
3642
os.unlink(self.name)
3643
# print "pid %d finished" % finished_process
3644
test_blocks = partition_tests(suite, concurrency)
3645
for process_tests in test_blocks:
3646
# ugly; currently reimplement rather than reuses TestCase methods.
3647
bzr_path = os.path.dirname(os.path.dirname(breezy.__file__)) + '/bzr'
3648
if not os.path.isfile(bzr_path):
3649
# We are probably installed. Assume sys.argv is the right file
3650
bzr_path = sys.argv[0]
3651
bzr_path = [bzr_path]
3652
if sys.platform == "win32":
3653
# if we're on windows, we can't execute the bzr script directly
3654
bzr_path = [sys.executable] + bzr_path
3655
fd, test_list_file_name = tempfile.mkstemp()
3656
test_list_file = os.fdopen(fd, 'wb', 1)
3657
for test in process_tests:
3658
test_list_file.write(test.id() + '\n')
3659
test_list_file.close()
3661
argv = bzr_path + ['selftest', '--load-list', test_list_file_name,
3663
if '--no-plugins' in sys.argv:
3664
argv.append('--no-plugins')
3665
# stderr=subprocess.STDOUT would be ideal, but until we prevent
3666
# noise on stderr it can interrupt the subunit protocol.
3667
process = subprocess.Popen(argv, stdin=subprocess.PIPE,
3668
stdout=subprocess.PIPE,
3669
stderr=subprocess.PIPE,
3671
test = TestInSubprocess(process, test_list_file_name)
3674
os.unlink(test_list_file_name)
3679
class ProfileResult(testtools.ExtendedToOriginalDecorator):
3680
"""Generate profiling data for all activity between start and success.
3682
The profile data is appended to the test's _benchcalls attribute and can
3683
be accessed by the forwarded-to TestResult.
3685
While it might be cleaner do accumulate this in stopTest, addSuccess is
3686
where our existing output support for lsprof is, and this class aims to
3687
fit in with that: while it could be moved it's not necessary to accomplish
3688
test profiling, nor would it be dramatically cleaner.
3691
def startTest(self, test):
3692
self.profiler = breezy.lsprof.BzrProfiler()
3693
# Prevent deadlocks in tests that use lsprof: those tests will
3695
breezy.lsprof.BzrProfiler.profiler_block = 0
3696
self.profiler.start()
3697
testtools.ExtendedToOriginalDecorator.startTest(self, test)
3699
def addSuccess(self, test):
3700
stats = self.profiler.stop()
3702
calls = test._benchcalls
3703
except AttributeError:
3704
test._benchcalls = []
3705
calls = test._benchcalls
3706
calls.append(((test.id(), "", ""), stats))
3707
testtools.ExtendedToOriginalDecorator.addSuccess(self, test)
3709
def stopTest(self, test):
3710
testtools.ExtendedToOriginalDecorator.stopTest(self, test)
3711
self.profiler = None
3714
# Controlled by "brz selftest -E=..." option
3715
# Currently supported:
3716
# -Eallow_debug Will no longer clear debug.debug_flags() so it
3717
# preserves any flags supplied at the command line.
3718
# -Edisable_lock_checks Turns errors in mismatched locks into simple prints
3719
# rather than failing tests. And no longer raise
3720
# LockContention when fctnl locks are not being used
3721
# with proper exclusion rules.
3722
# -Ethreads Will display thread ident at creation/join time to
3723
# help track thread leaks
3724
# -Euncollected_cases Display the identity of any test cases that weren't
3725
# deallocated after being completed.
3726
# -Econfig_stats Will collect statistics using addDetail
3727
selftest_debug_flags = set()
3730
def selftest(verbose=False, pattern=".*", stop_on_failure=True,
3732
test_suite_factory=None,
3735
matching_tests_first=None,
3738
exclude_pattern=None,
3744
suite_decorators=None,
3748
"""Run the whole test suite under the enhanced runner"""
3749
# XXX: Very ugly way to do this...
3750
# Disable warning about old formats because we don't want it to disturb
3751
# any blackbox tests.
3752
from breezy import repository
3753
repository._deprecation_warning_done = True
3755
global default_transport
3756
if transport is None:
3757
transport = default_transport
3758
old_transport = default_transport
3759
default_transport = transport
3760
global selftest_debug_flags
3761
old_debug_flags = selftest_debug_flags
3762
if debug_flags is not None:
3763
selftest_debug_flags = set(debug_flags)
3765
if load_list is None:
3768
keep_only = load_test_id_list(load_list)
3770
starting_with = [test_prefix_alias_registry.resolve_alias(start)
3771
for start in starting_with]
3772
# Always consider 'unittest' an interesting name so that failed
3773
# suites wrapped as test cases appear in the output.
3774
starting_with.append('unittest')
3775
if test_suite_factory is None:
3776
# Reduce loading time by loading modules based on the starting_with
3778
suite = test_suite(keep_only, starting_with)
3780
suite = test_suite_factory()
3782
# But always filter as requested.
3783
suite = filter_suite_by_id_startswith(suite, starting_with)
3784
result_decorators = []
3786
result_decorators.append(ProfileResult)
3787
return run_suite(suite, 'testbzr', verbose=verbose, pattern=pattern,
3788
stop_on_failure=stop_on_failure,
3789
transport=transport,
3790
lsprof_timed=lsprof_timed,
3791
bench_history=bench_history,
3792
matching_tests_first=matching_tests_first,
3793
list_only=list_only,
3794
random_seed=random_seed,
3795
exclude_pattern=exclude_pattern,
3797
runner_class=runner_class,
3798
suite_decorators=suite_decorators,
3800
result_decorators=result_decorators,
3803
default_transport = old_transport
3804
selftest_debug_flags = old_debug_flags
3807
def load_test_id_list(file_name):
3808
"""Load a test id list from a text file.
3810
The format is one test id by line. No special care is taken to impose
3811
strict rules, these test ids are used to filter the test suite so a test id
3812
that do not match an existing test will do no harm. This allows user to add
3813
comments, leave blank lines, etc.
3817
ftest = open(file_name, 'rt')
3818
except IOError as e:
3819
if e.errno != errno.ENOENT:
3822
raise errors.NoSuchFile(file_name)
3824
for test_name in ftest.readlines():
3825
test_list.append(test_name.strip())
3830
def suite_matches_id_list(test_suite, id_list):
3831
"""Warns about tests not appearing or appearing more than once.
3833
:param test_suite: A TestSuite object.
3834
:param test_id_list: The list of test ids that should be found in
3837
:return: (absents, duplicates) absents is a list containing the test found
3838
in id_list but not in test_suite, duplicates is a list containing the
3839
tests found multiple times in test_suite.
3841
When using a prefined test id list, it may occurs that some tests do not
3842
exist anymore or that some tests use the same id. This function warns the
3843
tester about potential problems in his workflow (test lists are volatile)
3844
or in the test suite itself (using the same id for several tests does not
3845
help to localize defects).
3847
# Build a dict counting id occurrences
3849
for test in iter_suite_tests(test_suite):
3851
tests[id] = tests.get(id, 0) + 1
3856
occurs = tests.get(id, 0)
3858
not_found.append(id)
3860
duplicates.append(id)
3862
return not_found, duplicates
3865
class TestIdList(object):
3866
"""Test id list to filter a test suite.
3868
Relying on the assumption that test ids are built as:
3869
<module>[.<class>.<method>][(<param>+)], <module> being in python dotted
3870
notation, this class offers methods to :
3871
- avoid building a test suite for modules not refered to in the test list,
3872
- keep only the tests listed from the module test suite.
3875
def __init__(self, test_id_list):
3876
# When a test suite needs to be filtered against us we compare test ids
3877
# for equality, so a simple dict offers a quick and simple solution.
3878
self.tests = dict().fromkeys(test_id_list, True)
3880
# While unittest.TestCase have ids like:
3881
# <module>.<class>.<method>[(<param+)],
3882
# doctest.DocTestCase can have ids like:
3885
# <module>.<function>
3886
# <module>.<class>.<method>
3888
# Since we can't predict a test class from its name only, we settle on
3889
# a simple constraint: a test id always begins with its module name.
3892
for test_id in test_id_list:
3893
parts = test_id.split('.')
3894
mod_name = parts.pop(0)
3895
modules[mod_name] = True
3897
mod_name += '.' + part
3898
modules[mod_name] = True
3899
self.modules = modules
3901
def refers_to(self, module_name):
3902
"""Is there tests for the module or one of its sub modules."""
3903
return module_name in self.modules
3905
def includes(self, test_id):
3906
return test_id in self.tests
3909
class TestPrefixAliasRegistry(registry.Registry):
3910
"""A registry for test prefix aliases.
3912
This helps implement shorcuts for the --starting-with selftest
3913
option. Overriding existing prefixes is not allowed but not fatal (a
3914
warning will be emitted).
3917
def register(self, key, obj, help=None, info=None,
3918
override_existing=False):
3919
"""See Registry.register.
3921
Trying to override an existing alias causes a warning to be emitted,
3922
not a fatal execption.
3925
super(TestPrefixAliasRegistry, self).register(
3926
key, obj, help=help, info=info, override_existing=False)
3928
actual = self.get(key)
3930
'Test prefix alias %s is already used for %s, ignoring %s'
3931
% (key, actual, obj))
3933
def resolve_alias(self, id_start):
3934
"""Replace the alias by the prefix in the given string.
3936
Using an unknown prefix is an error to help catching typos.
3938
parts = id_start.split('.')
3940
parts[0] = self.get(parts[0])
3942
raise errors.CommandError(
3943
'%s is not a known test prefix alias' % parts[0])
3944
return '.'.join(parts)
3947
test_prefix_alias_registry = TestPrefixAliasRegistry()
3948
"""Registry of test prefix aliases."""
3951
# This alias allows to detect typos ('bzrlin.') by making all valid test ids
3952
# appear prefixed ('breezy.' is "replaced" by 'breezy.').
3953
test_prefix_alias_registry.register('breezy', 'breezy')
3955
# Obvious highest levels prefixes, feel free to add your own via a plugin
3956
test_prefix_alias_registry.register('bd', 'breezy.doc')
3957
test_prefix_alias_registry.register('bu', 'breezy.utils')
3958
test_prefix_alias_registry.register('bt', 'breezy.tests')
3959
test_prefix_alias_registry.register('bgt', 'breezy.git.tests')
3960
test_prefix_alias_registry.register('bbt', 'breezy.bzr.tests')
3961
test_prefix_alias_registry.register('bb', 'breezy.tests.blackbox')
3962
test_prefix_alias_registry.register('bp', 'breezy.plugins')
3965
def _test_suite_testmod_names():
3966
"""Return the standard list of test module names to test."""
3970
'breezy.tests.blackbox',
3971
'breezy.tests.commands',
3972
'breezy.tests.per_branch',
3973
'breezy.tests.per_controldir',
3974
'breezy.tests.per_controldir_colo',
3975
'breezy.tests.per_foreign_vcs',
3976
'breezy.tests.per_interrepository',
3977
'breezy.tests.per_intertree',
3978
'breezy.tests.per_interbranch',
3979
'breezy.tests.per_lock',
3980
'breezy.tests.per_merger',
3981
'breezy.tests.per_transport',
3982
'breezy.tests.per_tree',
3983
'breezy.tests.per_repository',
3984
'breezy.tests.per_repository_reference',
3985
'breezy.tests.per_uifactory',
3986
'breezy.tests.per_workingtree',
3987
'breezy.tests.test__annotator',
3988
'breezy.tests.test__bencode',
3989
'breezy.tests.test__known_graph',
3990
'breezy.tests.test__simple_set',
3991
'breezy.tests.test__static_tuple',
3992
'breezy.tests.test__walkdirs_win32',
3993
'breezy.tests.test_ancestry',
3994
'breezy.tests.test_annotate',
3995
'breezy.tests.test_atomicfile',
3996
'breezy.tests.test_bad_files',
3997
'breezy.tests.test_bisect',
3998
'breezy.tests.test_bisect_multi',
3999
'breezy.tests.test_branch',
4000
'breezy.tests.test_branchbuilder',
4001
'breezy.tests.test_bugtracker',
4002
'breezy.tests.test__chunks_to_lines',
4003
'breezy.tests.test_cache_utf8',
4004
'breezy.tests.test_chunk_writer',
4005
'breezy.tests.test_clean_tree',
4006
'breezy.tests.test_cmdline',
4007
'breezy.tests.test_commands',
4008
'breezy.tests.test_commit',
4009
'breezy.tests.test_commit_merge',
4010
'breezy.tests.test_config',
4011
'breezy.tests.test_bedding',
4012
'breezy.tests.test_conflicts',
4013
'breezy.tests.test_controldir',
4014
'breezy.tests.test_counted_lock',
4015
'breezy.tests.test_crash',
4016
'breezy.tests.test_decorators',
4017
'breezy.tests.test_delta',
4018
'breezy.tests.test_debug',
4019
'breezy.tests.test_diff',
4020
'breezy.tests.test_directory_service',
4021
'breezy.tests.test_dirty_tracker',
4022
'breezy.tests.test_email_message',
4023
'breezy.tests.test_eol_filters',
4024
'breezy.tests.test_errors',
4025
'breezy.tests.test_estimate_compressed_size',
4026
'breezy.tests.test_export',
4027
'breezy.tests.test_export_pot',
4028
'breezy.tests.test_extract',
4029
'breezy.tests.test_features',
4030
'breezy.tests.test_fetch',
4031
'breezy.tests.test_fetch_ghosts',
4032
'breezy.tests.test_fixtures',
4033
'breezy.tests.test_fifo_cache',
4034
'breezy.tests.test_filters',
4035
'breezy.tests.test_filter_tree',
4036
'breezy.tests.test_foreign',
4037
'breezy.tests.test_generate_docs',
4038
'breezy.tests.test_globbing',
4039
'breezy.tests.test_gpg',
4040
'breezy.tests.test_graph',
4041
'breezy.tests.test_grep',
4042
'breezy.tests.test_hashcache',
4043
'breezy.tests.test_help',
4044
'breezy.tests.test_hooks',
4045
'breezy.tests.test_http',
4046
'breezy.tests.test_http_response',
4047
'breezy.tests.test_https_ca_bundle',
4048
'breezy.tests.test_https_urllib',
4049
'breezy.tests.test_i18n',
4050
'breezy.tests.test_identitymap',
4051
'breezy.tests.test_ignores',
4052
'breezy.tests.test_import_tariff',
4053
'breezy.tests.test_info',
4054
'breezy.tests.test_lazy_import',
4055
'breezy.tests.test_lazy_regex',
4056
'breezy.tests.test_library_state',
4057
'breezy.tests.test_location',
4058
'breezy.tests.test_lock',
4059
'breezy.tests.test_lockable_files',
4060
'breezy.tests.test_lockdir',
4061
'breezy.tests.test_log',
4062
'breezy.tests.test_lru_cache',
4063
'breezy.tests.test_lsprof',
4064
'breezy.tests.test_mail_client',
4065
'breezy.tests.test_matchers',
4066
'breezy.tests.test_memorybranch',
4067
'breezy.tests.test_memorytree',
4068
'breezy.tests.test_merge',
4069
'breezy.tests.test_merge3',
4070
'breezy.tests.test_mergeable',
4071
'breezy.tests.test_merge_core',
4072
'breezy.tests.test_merge_directive',
4073
'breezy.tests.test_mergetools',
4074
'breezy.tests.test_missing',
4075
'breezy.tests.test_msgeditor',
4076
'breezy.tests.test_multiparent',
4077
'breezy.tests.test_multiwalker',
4078
'breezy.tests.test_mutabletree',
4079
'breezy.tests.test_nonascii',
4080
'breezy.tests.test_options',
4081
'breezy.tests.test_osutils',
4082
'breezy.tests.test_osutils_encodings',
4083
'breezy.tests.test_patch',
4084
'breezy.tests.test_patches',
4085
'breezy.tests.test_permissions',
4086
'breezy.tests.test_plugins',
4087
'breezy.tests.test_progress',
4088
'breezy.tests.test_propose',
4089
'breezy.tests.test_pyutils',
4090
'breezy.tests.test_reconcile',
4091
'breezy.tests.test_reconfigure',
4092
'breezy.tests.test_registry',
4093
'breezy.tests.test_rename_map',
4094
'breezy.tests.test_revert',
4095
'breezy.tests.test_revision',
4096
'breezy.tests.test_revisionspec',
4097
'breezy.tests.test_revisiontree',
4098
'breezy.tests.test_rio',
4099
'breezy.tests.test__rio',
4100
'breezy.tests.test_rules',
4101
'breezy.tests.test_url_policy_open',
4102
'breezy.tests.test_sampler',
4103
'breezy.tests.test_scenarios',
4104
'breezy.tests.test_script',
4105
'breezy.tests.test_selftest',
4106
'breezy.tests.test_setup',
4107
'breezy.tests.test_sftp_transport',
4108
'breezy.tests.test_shelf',
4109
'breezy.tests.test_shelf_ui',
4110
'breezy.tests.test_smart_add',
4111
'breezy.tests.test_smtp_connection',
4112
'breezy.tests.test_source',
4113
'breezy.tests.test_ssh_transport',
4114
'breezy.tests.test_status',
4115
'breezy.tests.test_strace',
4116
'breezy.tests.test_subsume',
4117
'breezy.tests.test_switch',
4118
'breezy.tests.test_symbol_versioning',
4119
'breezy.tests.test_tag',
4120
'breezy.tests.test_test_server',
4121
'breezy.tests.test_textfile',
4122
'breezy.tests.test_textmerge',
4123
'breezy.tests.test_cethread',
4124
'breezy.tests.test_timestamp',
4125
'breezy.tests.test_trace',
4126
'breezy.tests.test_transactions',
4127
'breezy.tests.test_transform',
4128
'breezy.tests.test_transport',
4129
'breezy.tests.test_transport_log',
4130
'breezy.tests.test_tree',
4131
'breezy.tests.test_treebuilder',
4132
'breezy.tests.test_treeshape',
4133
'breezy.tests.test_tsort',
4134
'breezy.tests.test_tuned_gzip',
4135
'breezy.tests.test_ui',
4136
'breezy.tests.test_uncommit',
4137
'breezy.tests.test_upgrade',
4138
'breezy.tests.test_upgrade_stacked',
4139
'breezy.tests.test_upstream_import',
4140
'breezy.tests.test_urlutils',
4141
'breezy.tests.test_utextwrap',
4142
'breezy.tests.test_version',
4143
'breezy.tests.test_version_info',
4144
'breezy.tests.test_views',
4145
'breezy.tests.test_whitebox',
4146
'breezy.tests.test_win32utils',
4147
'breezy.tests.test_workspace',
4148
'breezy.tests.test_workingtree',
4149
'breezy.tests.test_wsgi',
4153
def _test_suite_modules_to_doctest():
4154
"""Return the list of modules to doctest."""
4156
# GZ 2009-03-31: No docstrings with -OO so there's nothing to doctest
4160
'breezy.branchbuilder',
4161
'breezy.bzr.inventory',
4162
'breezy.decorators',
4163
'breezy.iterablefile',
4168
'breezy.symbol_versioning',
4170
'breezy.tests.fixtures',
4172
'breezy.transport.http',
4173
'breezy.version_info_formats.format_custom',
4177
def test_suite(keep_only=None, starting_with=None):
4178
"""Build and return TestSuite for the whole of breezy.
4180
:param keep_only: A list of test ids limiting the suite returned.
4182
:param starting_with: An id limiting the suite returned to the tests
4185
This function can be replaced if you need to change the default test
4186
suite on a global basis, but it is not encouraged.
4189
loader = TestUtil.TestLoader()
4191
if keep_only is not None:
4192
id_filter = TestIdList(keep_only)
4194
# We take precedence over keep_only because *at loading time* using
4195
# both options means we will load less tests for the same final result.
4196
def interesting_module(name):
4197
for start in starting_with:
4198
# Either the module name starts with the specified string
4199
# or it may contain tests starting with the specified string
4200
if name.startswith(start) or start.startswith(name):
4203
loader = TestUtil.FilteredByModuleTestLoader(interesting_module)
4205
elif keep_only is not None:
4206
loader = TestUtil.FilteredByModuleTestLoader(id_filter.refers_to)
4208
def interesting_module(name):
4209
return id_filter.refers_to(name)
4212
loader = TestUtil.TestLoader()
4214
def interesting_module(name):
4215
# No filtering, all modules are interesting
4218
suite = loader.suiteClass()
4220
# modules building their suite with loadTestsFromModuleNames
4221
suite.addTest(loader.loadTestsFromModuleNames(_test_suite_testmod_names()))
4223
suite.addTest(loader.loadTestsFromModuleNames(['breezy.doc']))
4225
for mod in _test_suite_modules_to_doctest():
4226
if not interesting_module(mod):
4227
# No tests to keep here, move along
4230
# note that this really does mean "report only" -- doctest
4231
# still runs the rest of the examples
4232
doc_suite = IsolatedDocTestSuite(
4233
mod, optionflags=doctest.REPORT_ONLY_FIRST_FAILURE)
4234
except ValueError as e:
4235
print('**failed to get doctest for: %s\n%s' % (mod, e))
4237
if len(doc_suite._tests) == 0:
4238
raise errors.BzrError("no doctests found in %s" % (mod,))
4239
suite.addTest(doc_suite)
4241
default_encoding = sys.getdefaultencoding()
4242
for name, plugin in _mod_plugin.plugins().items():
4243
if not interesting_module(plugin.module.__name__):
4245
plugin_suite = plugin.test_suite()
4246
# We used to catch ImportError here and turn it into just a warning,
4247
# but really if you don't have --no-plugins this should be a failure.
4248
# mbp 20080213 - see http://bugs.launchpad.net/bugs/189771
4249
if plugin_suite is None:
4250
plugin_suite = plugin.load_plugin_tests(loader)
4251
if plugin_suite is not None:
4252
suite.addTest(plugin_suite)
4253
if default_encoding != sys.getdefaultencoding():
4255
'Plugin "%s" tried to reset default encoding to: %s', name,
4256
sys.getdefaultencoding())
4258
sys.setdefaultencoding(default_encoding)
4260
if keep_only is not None:
4261
# Now that the referred modules have loaded their tests, keep only the
4263
suite = filter_suite_by_id_list(suite, id_filter)
4264
# Do some sanity checks on the id_list filtering
4265
not_found, duplicates = suite_matches_id_list(suite, keep_only)
4267
# The tester has used both keep_only and starting_with, so he is
4268
# already aware that some tests are excluded from the list, there
4269
# is no need to tell him which.
4272
# Some tests mentioned in the list are not in the test suite. The
4273
# list may be out of date, report to the tester.
4274
for id in not_found:
4275
trace.warning('"%s" not found in the test suite', id)
4276
for id in duplicates:
4277
trace.warning('"%s" is used as an id by several tests', id)
4282
def multiply_scenarios(*scenarios):
4283
"""Multiply two or more iterables of scenarios.
4285
It is safe to pass scenario generators or iterators.
4287
:returns: A list of compound scenarios: the cross-product of all
4288
scenarios, with the names concatenated and the parameters
4291
return functools.reduce(_multiply_two_scenarios, map(list, scenarios))
4294
def _multiply_two_scenarios(scenarios_left, scenarios_right):
4295
"""Multiply two sets of scenarios.
4297
:returns: the cartesian product of the two sets of scenarios, that is
4298
a scenario for every possible combination of a left scenario and a
4302
('%s,%s' % (left_name, right_name),
4303
dict(left_dict, **right_dict))
4304
for left_name, left_dict in scenarios_left
4305
for right_name, right_dict in scenarios_right]
4308
def multiply_tests(tests, scenarios, result):
4309
"""Multiply tests_list by scenarios into result.
4311
This is the core workhorse for test parameterisation.
4313
Typically the load_tests() method for a per-implementation test suite will
4314
call multiply_tests and return the result.
4316
:param tests: The tests to parameterise.
4317
:param scenarios: The scenarios to apply: pairs of (scenario_name,
4318
scenario_param_dict).
4319
:param result: A TestSuite to add created tests to.
4321
This returns the passed in result TestSuite with the cross product of all
4322
the tests repeated once for each scenario. Each test is adapted by adding
4323
the scenario name at the end of its id(), and updating the test object's
4324
__dict__ with the scenario_param_dict.
4326
>>> import breezy.tests.test_sampler
4327
>>> r = multiply_tests(
4328
... breezy.tests.test_sampler.DemoTest('test_nothing'),
4329
... [('one', dict(param=1)),
4330
... ('two', dict(param=2))],
4331
... TestUtil.TestSuite())
4332
>>> tests = list(iter_suite_tests(r))
4336
'breezy.tests.test_sampler.DemoTest.test_nothing(one)'
4342
for test in iter_suite_tests(tests):
4343
apply_scenarios(test, scenarios, result)
4347
def apply_scenarios(test, scenarios, result):
4348
"""Apply the scenarios in scenarios to test and add to result.
4350
:param test: The test to apply scenarios to.
4351
:param scenarios: An iterable of scenarios to apply to test.
4353
:seealso: apply_scenario
4355
for scenario in scenarios:
4356
result.addTest(apply_scenario(test, scenario))
4360
def apply_scenario(test, scenario):
4361
"""Copy test and apply scenario to it.
4363
:param test: A test to adapt.
4364
:param scenario: A tuple describing the scenario.
4365
The first element of the tuple is the new test id.
4366
The second element is a dict containing attributes to set on the
4368
:return: The adapted test.
4370
new_id = "%s(%s)" % (test.id(), scenario[0])
4371
new_test = clone_test(test, new_id)
4372
for name, value in scenario[1].items():
4373
setattr(new_test, name, value)
4377
def clone_test(test, new_id):
4378
"""Clone a test giving it a new id.
4380
:param test: The test to clone.
4381
:param new_id: The id to assign to it.
4382
:return: The new test.
4384
new_test = copy.copy(test)
4385
new_test.id = lambda: new_id
4386
# XXX: Workaround <https://bugs.launchpad.net/testtools/+bug/637725>, which
4387
# causes cloned tests to share the 'details' dict. This makes it hard to
4388
# read the test output for parameterized tests, because tracebacks will be
4389
# associated with irrelevant tests.
4391
details = new_test._TestCase__details
4392
except AttributeError:
4393
# must be a different version of testtools than expected. Do nothing.
4396
# Reset the '__details' dict.
4397
new_test._TestCase__details = {}
4402
def permute_tests_for_extension(standard_tests, loader, py_module_name,
4404
"""Helper for permutating tests against an extension module.
4406
This is meant to be used inside a modules 'load_tests()' function. It will
4407
create 2 scenarios, and cause all tests in the 'standard_tests' to be run
4408
against both implementations. Setting 'test.module' to the appropriate
4409
module. See breezy.tests.test__chk_map.load_tests as an example.
4411
:param standard_tests: A test suite to permute
4412
:param loader: A TestLoader
4413
:param py_module_name: The python path to a python module that can always
4414
be loaded, and will be considered the 'python' implementation. (eg
4415
'breezy._chk_map_py')
4416
:param ext_module_name: The python path to an extension module. If the
4417
module cannot be loaded, a single test will be added, which notes that
4418
the module is not available. If it can be loaded, all standard_tests
4419
will be run against that module.
4420
:return: (suite, feature) suite is a test-suite that has all the permuted
4421
tests. feature is the Feature object that can be used to determine if
4422
the module is available.
4425
from .features import ModuleAvailableFeature
4426
py_module = pyutils.get_named_object(py_module_name)
4428
('python', {'module': py_module}),
4430
suite = loader.suiteClass()
4431
feature = ModuleAvailableFeature(ext_module_name)
4432
if feature.available():
4433
scenarios.append(('C', {'module': feature.module}))
4435
class FailWithoutFeature(TestCase):
4437
return ext_module_name + '.' + super(FailWithoutFeature, self).id()
4438
def test_fail(self):
4439
self.requireFeature(feature)
4440
# the compiled module isn't available, so we add a failing test
4441
suite.addTest(loader.loadTestsFromTestCase(FailWithoutFeature))
4442
result = multiply_tests(standard_tests, scenarios, suite)
4443
return result, feature
4446
def _rmtree_temp_dir(dirname, test_id=None):
4447
# If LANG=C we probably have created some bogus paths
4448
# which rmtree(unicode) will fail to delete
4449
# so make sure we are using rmtree(str) to delete everything
4450
# except on win32, where rmtree(str) will fail
4451
# since it doesn't have the property of byte-stream paths
4452
# (they are either ascii or mbcs)
4453
if sys.platform == 'win32' and isinstance(dirname, bytes):
4454
# make sure we are using the unicode win32 api
4455
dirname = dirname.decode('mbcs')
4457
dirname = dirname.encode(sys.getfilesystemencoding())
4459
osutils.rmtree(dirname)
4460
except OSError as e:
4461
# We don't want to fail here because some useful display will be lost
4462
# otherwise. Polluting the tmp dir is bad, but not giving all the
4463
# possible info to the test runner is even worse.
4464
if test_id is not None:
4465
ui.ui_factory.clear_term()
4466
sys.stderr.write('\nWhile running: %s\n' % (test_id,))
4467
# Ugly, but the last thing we want here is fail, so bear with it.
4468
printable_e = str(e).decode(osutils.get_user_encoding(), 'replace'
4469
).encode('ascii', 'replace')
4470
sys.stderr.write('Unable to remove testing dir %s\n%s'
4471
% (os.path.basename(dirname), printable_e))
4474
def probe_unicode_in_user_encoding():
4475
"""Try to encode several unicode strings to use in unicode-aware tests.
4476
Return first successfull match.
4478
:return: (unicode value, encoded plain string value) or (None, None)
4480
possible_vals = [u'm\xb5', u'\xe1', u'\u0410']
4481
for uni_val in possible_vals:
4483
str_val = uni_val.encode(osutils.get_user_encoding())
4484
except UnicodeEncodeError:
4485
# Try a different character
4488
return uni_val, str_val
4492
def probe_bad_non_ascii(encoding):
4493
"""Try to find [bad] character with code [128..255]
4494
that cannot be decoded to unicode in some encoding.
4495
Return None if all non-ascii characters is valid
4498
for i in range(128, 256):
4501
char.decode(encoding)
4502
except UnicodeDecodeError:
4507
# Only define SubUnitBzrRunner if subunit is available.
4509
from subunit import TestProtocolClient
4510
from subunit.test_results import AutoTimingTestResultDecorator
4512
class SubUnitBzrProtocolClientv1(TestProtocolClient):
4514
def stopTest(self, test):
4515
super(SubUnitBzrProtocolClientv1, self).stopTest(test)
4516
_clear__type_equality_funcs(test)
4518
def addSuccess(self, test, details=None):
4519
# The subunit client always includes the details in the subunit
4520
# stream, but we don't want to include it in ours.
4521
if details is not None and 'log' in details:
4523
return super(SubUnitBzrProtocolClientv1, self).addSuccess(
4526
class SubUnitBzrRunnerv1(TextTestRunner):
4528
def run(self, test):
4529
result = AutoTimingTestResultDecorator(
4530
SubUnitBzrProtocolClientv1(self.stream))
4538
from subunit.run import SubunitTestRunner
4540
class SubUnitBzrRunnerv2(TextTestRunner, SubunitTestRunner):
4542
def __init__(self, stream=sys.stderr, descriptions=0, verbosity=1,
4543
bench_history=None, strict=False, result_decorators=None):
4544
TextTestRunner.__init__(
4545
self, stream=stream,
4546
descriptions=descriptions, verbosity=verbosity,
4547
bench_history=bench_history, strict=strict,
4548
result_decorators=result_decorators)
4549
SubunitTestRunner.__init__(self, verbosity=verbosity,
4552
run = SubunitTestRunner.run
added
removed
breezy/tests/__init__.py
