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
from __future__ import absolute_import
21
# NOTE: Some classes in here use camelCaseNaming() rather than
22
# underscore_naming(). That's for consistency with unittest; it's not the
23
# general style of breezy. Please continue that consistency when adding e.g.
24
# new assertFoo() methods.
59
# nb: check this before importing anything else from within it
60
_testtools_version = getattr(testtools, '__version__', ())
61
if _testtools_version < (0, 9, 5):
62
raise ImportError("need at least testtools 0.9.5: %s is %r"
63
% (testtools.__file__, _testtools_version))
64
from testtools import content
70
commands as _mod_commands,
79
plugin as _mod_plugin,
86
transport as _mod_transport,
89
from breezy.bzr import (
95
# lsprof not available
97
from ..sixish import (
103
from ..bzr.smart import client, request
104
from ..transport import (
108
from ..tests import (
116
# Mark this python module as being part of the implementation
117
# of unittest: this gives us better tracebacks where the last
118
# shown frame is the test code, not our assertXYZ.
121
default_transport = test_server.LocalURLServer
124
_unitialized_attr = object()
125
"""A sentinel needed to act as a default value in a method signature."""
128
# Subunit result codes, defined here to prevent a hard dependency on subunit.
132
# These are intentionally brought into this namespace. That way plugins, etc
133
# can just "from breezy.tests import TestCase, TestLoader, etc"
134
TestSuite = TestUtil.TestSuite
135
TestLoader = TestUtil.TestLoader
137
# Tests should run in a clean and clearly defined environment. The goal is to
138
# keep them isolated from the running environment as mush as possible. The test
139
# framework ensures the variables defined below are set (or deleted if the
140
# value is None) before a test is run and reset to their original value after
141
# the test is run. Generally if some code depends on an environment variable,
142
# the tests should start without this variable in the environment. There are a
143
# few exceptions but you shouldn't violate this rule lightly.
148
'XDG_CONFIG_HOME': None,
149
# brz now uses the Win32 API and doesn't rely on APPDATA, but the
150
# tests do check our impls match APPDATA
151
'BRZ_EDITOR': None, # test_msgeditor manipulates this variable
155
'BZREMAIL': None, # may still be present in the environment
156
'EMAIL': 'jrandom@example.com', # set EMAIL as brz does not guess
157
'BRZ_PROGRESS_BAR': None,
158
# This should trap leaks to ~/.brz.log. This occurs when tests use TestCase
159
# as a base class instead of TestCaseInTempDir. Tests inheriting from
160
# TestCase should not use disk resources, BRZ_LOG is one.
161
'BRZ_LOG': '/you-should-use-TestCaseInTempDir-if-you-need-a-log-file',
162
'BRZ_PLUGIN_PATH': '-site',
163
'BRZ_DISABLE_PLUGINS': None,
164
'BRZ_PLUGINS_AT': None,
165
'BRZ_CONCURRENCY': None,
166
# Make sure that any text ui tests are consistent regardless of
167
# the environment the test case is run in; you may want tests that
168
# test other combinations. 'dumb' is a reasonable guess for tests
169
# going to a pipe or a BytesIO.
175
'SSH_AUTH_SOCK': None,
185
'BZR_REMOTE_PATH': None,
186
# Generally speaking, we don't want apport reporting on crashes in
187
# the test envirnoment unless we're specifically testing apport,
188
# so that it doesn't leak into the real system environment. We
189
# use an env var so it propagates to subprocesses.
190
'APPORT_DISABLE': '1',
194
def override_os_environ(test, env=None):
195
"""Modify os.environ keeping a copy.
197
:param test: A test instance
199
:param env: A dict containing variable definitions to be installed
202
env = isolated_environ
203
test._original_os_environ = dict(**os.environ)
205
osutils.set_or_unset_env(var, env[var])
206
if var not in test._original_os_environ:
207
# The var is new, add it with a value of None, so
208
# restore_os_environ will delete it
209
test._original_os_environ[var] = None
212
def restore_os_environ(test):
213
"""Restore os.environ to its original state.
215
:param test: A test instance previously passed to override_os_environ.
217
for var, value in test._original_os_environ.items():
218
# Restore the original value (or delete it if the value has been set to
219
# None in override_os_environ).
220
osutils.set_or_unset_env(var, value)
223
def _clear__type_equality_funcs(test):
224
"""Cleanup bound methods stored on TestCase instances
226
Clear the dict breaking a few (mostly) harmless cycles in the affected
227
unittests released with Python 2.6 and initial Python 2.7 versions.
229
For a few revisions between Python 2.7.1 and Python 2.7.2 that annoyingly
230
shipped in Oneiric, an object with no clear method was used, hence the
231
extra complications, see bug 809048 for details.
233
type_equality_funcs = getattr(test, "_type_equality_funcs", None)
234
if type_equality_funcs is not None:
235
tef_clear = getattr(type_equality_funcs, "clear", None)
236
if tef_clear is None:
237
tef_instance_dict = getattr(type_equality_funcs, "__dict__", None)
238
if tef_instance_dict is not None:
239
tef_clear = tef_instance_dict.clear
240
if tef_clear is not None:
244
class ExtendedTestResult(testtools.TextTestResult):
245
"""Accepts, reports and accumulates the results of running tests.
247
Compared to the unittest version this class adds support for
248
profiling, benchmarking, stopping as soon as a test fails, and
249
skipping tests. There are further-specialized subclasses for
250
different types of display.
252
When a test finishes, in whatever way, it calls one of the addSuccess,
253
addFailure or addError methods. These in turn may redirect to a more
254
specific case for the special test results supported by our extended
257
Note that just one of these objects is fed the results from many tests.
262
def __init__(self, stream, descriptions, verbosity,
266
"""Construct new TestResult.
268
:param bench_history: Optionally, a writable file object to accumulate
271
testtools.TextTestResult.__init__(self, stream)
272
if bench_history is not None:
273
from breezy.version import _get_bzr_source_tree
274
src_tree = _get_bzr_source_tree()
277
revision_id = src_tree.get_parent_ids()[0]
279
# XXX: if this is a brand new tree, do the same as if there
283
# XXX: If there's no branch, what should we do?
285
bench_history.write("--date %s %s\n" % (time.time(), revision_id))
286
self._bench_history = bench_history
287
self.ui = ui.ui_factory
290
self.failure_count = 0
291
self.known_failure_count = 0
293
self.not_applicable_count = 0
294
self.unsupported = {}
296
self._overall_start_time = time.time()
297
self._strict = strict
298
self._first_thread_leaker_id = None
299
self._tests_leaking_threads_count = 0
300
self._traceback_from_test = None
302
def stopTestRun(self):
305
stopTime = time.time()
306
timeTaken = stopTime - self.startTime
307
# GZ 2010-07-19: Seems testtools has no printErrors method, and though
308
# the parent class method is similar have to duplicate
309
self._show_list('ERROR', self.errors)
310
self._show_list('FAIL', self.failures)
311
self.stream.write(self.sep2)
312
self.stream.write("%s %d test%s in %.3fs\n\n" % (
313
actionTaken, run, run != 1 and "s" or "", timeTaken))
314
if not self.wasSuccessful():
315
self.stream.write("FAILED (")
316
failed, errored = map(len, (self.failures, self.errors))
318
self.stream.write("failures=%d" % failed)
321
self.stream.write(", ")
322
self.stream.write("errors=%d" % errored)
323
if self.known_failure_count:
324
if failed or errored:
325
self.stream.write(", ")
326
self.stream.write("known_failure_count=%d" %
327
self.known_failure_count)
328
self.stream.write(")\n")
330
if self.known_failure_count:
331
self.stream.write("OK (known_failures=%d)\n" %
332
self.known_failure_count)
334
self.stream.write("OK\n")
335
if self.skip_count > 0:
336
skipped = self.skip_count
337
self.stream.write('%d test%s skipped\n' %
338
(skipped, skipped != 1 and "s" or ""))
340
for feature, count in sorted(self.unsupported.items()):
341
self.stream.write("Missing feature '%s' skipped %d tests.\n" %
344
ok = self.wasStrictlySuccessful()
346
ok = self.wasSuccessful()
347
if self._first_thread_leaker_id:
349
'%s is leaking threads among %d leaking tests.\n' % (
350
self._first_thread_leaker_id,
351
self._tests_leaking_threads_count))
352
# We don't report the main thread as an active one.
354
'%d non-main threads were left active in the end.\n'
355
% (len(self._active_threads) - 1))
357
def getDescription(self, test):
360
def _extractBenchmarkTime(self, testCase, details=None):
361
"""Add a benchmark time for the current test case."""
362
if details and 'benchtime' in details:
363
return float(''.join(details['benchtime'].iter_bytes()))
364
return getattr(testCase, "_benchtime", None)
366
def _delta_to_float(self, a_timedelta, precision):
367
# This calls ceiling to ensure that the most pessimistic view of time
368
# taken is shown (rather than leaving it to the Python %f operator
369
# to decide whether to round/floor/ceiling. This was added when we
370
# had pyp3 test failures that suggest a floor was happening.
371
shift = 10 ** precision
373
(a_timedelta.days * 86400.0 + a_timedelta.seconds +
374
a_timedelta.microseconds / 1000000.0) * shift) / shift
376
def _elapsedTestTimeString(self):
377
"""Return time string for overall time the current test has taken."""
378
return self._formatTime(self._delta_to_float(
379
self._now() - self._start_datetime, 3))
381
def _testTimeString(self, testCase):
382
benchmark_time = self._extractBenchmarkTime(testCase)
383
if benchmark_time is not None:
384
return self._formatTime(benchmark_time) + "*"
386
return self._elapsedTestTimeString()
388
def _formatTime(self, seconds):
389
"""Format seconds as milliseconds with leading spaces."""
390
# some benchmarks can take thousands of seconds to run, so we need 8
392
return "%8dms" % (1000 * seconds)
394
def _shortened_test_description(self, test):
396
what = re.sub(r'^breezy\.tests\.', '', what)
399
# GZ 2010-10-04: Cloned tests may end up harmlessly calling this method
400
# multiple times in a row, because the handler is added for
401
# each test but the container list is shared between cases.
402
# See lp:498869 lp:625574 and lp:637725 for background.
403
def _record_traceback_from_test(self, exc_info):
404
"""Store the traceback from passed exc_info tuple till"""
405
self._traceback_from_test = exc_info[2]
407
def startTest(self, test):
408
super(ExtendedTestResult, self).startTest(test)
412
self.report_test_start(test)
413
test.number = self.count
414
self._recordTestStartTime()
415
# Make testtools cases give us the real traceback on failure
416
addOnException = getattr(test, "addOnException", None)
417
if addOnException is not None:
418
addOnException(self._record_traceback_from_test)
419
# Only check for thread leaks on breezy derived test cases
420
if isinstance(test, TestCase):
421
test.addCleanup(self._check_leaked_threads, test)
423
def stopTest(self, test):
424
super(ExtendedTestResult, self).stopTest(test)
425
# Manually break cycles, means touching various private things but hey
426
getDetails = getattr(test, "getDetails", None)
427
if getDetails is not None:
429
_clear__type_equality_funcs(test)
430
self._traceback_from_test = None
432
def startTests(self):
433
self.report_tests_starting()
434
self._active_threads = threading.enumerate()
436
def _check_leaked_threads(self, test):
437
"""See if any threads have leaked since last call
439
A sample of live threads is stored in the _active_threads attribute,
440
when this method runs it compares the current live threads and any not
441
in the previous sample are treated as having leaked.
443
now_active_threads = set(threading.enumerate())
444
threads_leaked = now_active_threads.difference(self._active_threads)
446
self._report_thread_leak(test, threads_leaked, now_active_threads)
447
self._tests_leaking_threads_count += 1
448
if self._first_thread_leaker_id is None:
449
self._first_thread_leaker_id = test.id()
450
self._active_threads = now_active_threads
452
def _recordTestStartTime(self):
453
"""Record that a test has started."""
454
self._start_datetime = self._now()
456
def addError(self, test, err):
457
"""Tell result that test finished with an error.
459
Called from the TestCase run() method when the test
460
fails with an unexpected error.
462
self._post_mortem(self._traceback_from_test or err[2])
463
super(ExtendedTestResult, self).addError(test, err)
464
self.error_count += 1
465
self.report_error(test, err)
469
def addFailure(self, test, err):
470
"""Tell result that test failed.
472
Called from the TestCase run() method when the test
473
fails because e.g. an assert() method failed.
475
self._post_mortem(self._traceback_from_test or err[2])
476
super(ExtendedTestResult, self).addFailure(test, err)
477
self.failure_count += 1
478
self.report_failure(test, err)
482
def addSuccess(self, test, details=None):
483
"""Tell result that test completed successfully.
485
Called from the TestCase run()
487
if self._bench_history is not None:
488
benchmark_time = self._extractBenchmarkTime(test, details)
489
if benchmark_time is not None:
490
self._bench_history.write("%s %s\n" % (
491
self._formatTime(benchmark_time),
493
self.report_success(test)
494
super(ExtendedTestResult, self).addSuccess(test)
495
test._log_contents = ''
497
def addExpectedFailure(self, test, err):
498
self.known_failure_count += 1
499
self.report_known_failure(test, err)
501
def addUnexpectedSuccess(self, test, details=None):
502
"""Tell result the test unexpectedly passed, counting as a failure
504
When the minimum version of testtools required becomes 0.9.8 this
505
can be updated to use the new handling there.
507
super(ExtendedTestResult, self).addFailure(test, details=details)
508
self.failure_count += 1
509
self.report_unexpected_success(test,
510
"".join(details["reason"].iter_text()))
514
def addNotSupported(self, test, feature):
515
"""The test will not be run because of a missing feature.
517
# this can be called in two different ways: it may be that the
518
# test started running, and then raised (through requireFeature)
519
# UnavailableFeature. Alternatively this method can be called
520
# while probing for features before running the test code proper; in
521
# that case we will see startTest and stopTest, but the test will
522
# never actually run.
523
self.unsupported.setdefault(str(feature), 0)
524
self.unsupported[str(feature)] += 1
525
self.report_unsupported(test, feature)
527
def addSkip(self, test, reason):
528
"""A test has not run for 'reason'."""
530
self.report_skip(test, reason)
532
def addNotApplicable(self, test, reason):
533
self.not_applicable_count += 1
534
self.report_not_applicable(test, reason)
536
def _count_stored_tests(self):
537
"""Count of tests instances kept alive due to not succeeding"""
538
return self.error_count + self.failure_count + self.known_failure_count
540
def _post_mortem(self, tb=None):
541
"""Start a PDB post mortem session."""
542
if os.environ.get('BRZ_TEST_PDB', None):
546
def progress(self, offset, whence):
547
"""The test is adjusting the count of tests to run."""
548
if whence == SUBUNIT_SEEK_SET:
549
self.num_tests = offset
550
elif whence == SUBUNIT_SEEK_CUR:
551
self.num_tests += offset
553
raise errors.BzrError("Unknown whence %r" % whence)
555
def report_tests_starting(self):
556
"""Display information before the test run begins"""
557
if getattr(sys, 'frozen', None) is None:
558
bzr_path = osutils.realpath(sys.argv[0])
560
bzr_path = sys.executable
562
'brz selftest: %s\n' % (bzr_path,))
565
breezy.__path__[0],))
567
' bzr-%s python-%s %s\n' % (
568
breezy.version_string,
569
breezy._format_version_tuple(sys.version_info),
570
platform.platform(aliased=1),
572
self.stream.write('\n')
574
def report_test_start(self, test):
575
"""Display information on the test just about to be run"""
577
def _report_thread_leak(self, test, leaked_threads, active_threads):
578
"""Display information on a test that leaked one or more threads"""
579
# GZ 2010-09-09: A leak summary reported separately from the general
580
# thread debugging would be nice. Tests under subunit
581
# need something not using stream, perhaps adding a
582
# testtools details object would be fitting.
583
if 'threads' in selftest_debug_flags:
584
self.stream.write('%s is leaking, active is now %d\n' %
585
(test.id(), len(active_threads)))
587
def startTestRun(self):
588
self.startTime = time.time()
590
def report_success(self, test):
593
def wasStrictlySuccessful(self):
594
if self.unsupported or self.known_failure_count:
596
return self.wasSuccessful()
599
class TextTestResult(ExtendedTestResult):
600
"""Displays progress and results of tests in text form"""
602
def __init__(self, stream, descriptions, verbosity,
606
ExtendedTestResult.__init__(self, stream, descriptions, verbosity,
607
bench_history, strict)
608
self.pb = self.ui.nested_progress_bar()
609
self.pb.show_pct = False
610
self.pb.show_spinner = False
611
self.pb.show_eta = False,
612
self.pb.show_count = False
613
self.pb.show_bar = False
614
self.pb.update_latency = 0
615
self.pb.show_transport_activity = False
617
def stopTestRun(self):
618
# called when the tests that are going to run have run
621
super(TextTestResult, self).stopTestRun()
623
def report_tests_starting(self):
624
super(TextTestResult, self).report_tests_starting()
625
self.pb.update('[test 0/%d] Starting' % (self.num_tests))
627
def _progress_prefix_text(self):
628
# the longer this text, the less space we have to show the test
630
a = '[%d' % self.count # total that have been run
631
# tests skipped as known not to be relevant are not important enough
633
# if self.skip_count:
634
# a += ', %d skip' % self.skip_count
635
# if self.known_failure_count:
636
# a += '+%dX' % self.known_failure_count
638
a += '/%d' % self.num_tests
640
runtime = time.time() - self._overall_start_time
642
a += '%dm%ds' % (runtime / 60, runtime % 60)
645
total_fail_count = self.error_count + self.failure_count
647
a += ', %d failed' % total_fail_count
648
# if self.unsupported:
649
# a += ', %d missing' % len(self.unsupported)
653
def report_test_start(self, test):
655
self._progress_prefix_text() +
657
self._shortened_test_description(test))
659
def _test_description(self, test):
660
return self._shortened_test_description(test)
662
def report_error(self, test, err):
663
self.stream.write('ERROR: %s\n %s\n' % (
664
self._test_description(test),
668
def report_failure(self, test, err):
669
self.stream.write('FAIL: %s\n %s\n' % (
670
self._test_description(test),
674
def report_known_failure(self, test, err):
677
def report_unexpected_success(self, test, reason):
678
self.stream.write('FAIL: %s\n %s: %s\n' % (
679
self._test_description(test),
680
"Unexpected success. Should have failed",
684
def report_skip(self, test, reason):
687
def report_not_applicable(self, test, reason):
690
def report_unsupported(self, test, feature):
691
"""test cannot be run because feature is missing."""
694
class VerboseTestResult(ExtendedTestResult):
695
"""Produce long output, with one line per test run plus times"""
697
def _ellipsize_to_right(self, a_string, final_width):
698
"""Truncate and pad a string, keeping the right hand side"""
699
if len(a_string) > final_width:
700
result = '...' + a_string[3 - final_width:]
703
return result.ljust(final_width)
705
def report_tests_starting(self):
706
self.stream.write('running %d tests...\n' % self.num_tests)
707
super(VerboseTestResult, self).report_tests_starting()
709
def report_test_start(self, test):
710
name = self._shortened_test_description(test)
711
width = osutils.terminal_width()
712
if width is not None:
713
# width needs space for 6 char status, plus 1 for slash, plus an
714
# 11-char time string, plus a trailing blank
715
# when NUMBERED_DIRS: plus 5 chars on test number, plus 1 char on
717
self.stream.write(self._ellipsize_to_right(name, width - 18))
719
self.stream.write(name)
722
def _error_summary(self, err):
724
return '%s%s' % (indent, err[1])
726
def report_error(self, test, err):
727
self.stream.write('ERROR %s\n%s\n'
728
% (self._testTimeString(test),
729
self._error_summary(err)))
731
def report_failure(self, test, err):
732
self.stream.write(' FAIL %s\n%s\n'
733
% (self._testTimeString(test),
734
self._error_summary(err)))
736
def report_known_failure(self, test, err):
737
self.stream.write('XFAIL %s\n%s\n'
738
% (self._testTimeString(test),
739
self._error_summary(err)))
741
def report_unexpected_success(self, test, reason):
742
self.stream.write(' FAIL %s\n%s: %s\n'
743
% (self._testTimeString(test),
744
"Unexpected success. Should have failed",
747
def report_success(self, test):
748
self.stream.write(' OK %s\n' % self._testTimeString(test))
749
for bench_called, stats in getattr(test, '_benchcalls', []):
750
self.stream.write('LSProf output for %s(%s, %s)\n' % bench_called)
751
stats.pprint(file=self.stream)
752
# flush the stream so that we get smooth output. This verbose mode is
753
# used to show the output in PQM.
756
def report_skip(self, test, reason):
757
self.stream.write(' SKIP %s\n%s\n'
758
% (self._testTimeString(test), reason))
760
def report_not_applicable(self, test, reason):
761
self.stream.write(' N/A %s\n %s\n'
762
% (self._testTimeString(test), reason))
764
def report_unsupported(self, test, feature):
765
"""test cannot be run because feature is missing."""
766
self.stream.write("NODEP %s\n The feature '%s' is not available.\n"
767
% (self._testTimeString(test), feature))
770
class TextTestRunner(object):
771
stop_on_failure = False
779
result_decorators=None,
781
"""Create a TextTestRunner.
783
:param result_decorators: An optional list of decorators to apply
784
to the result object being used by the runner. Decorators are
785
applied left to right - the first element in the list is the
788
# stream may know claim to know to write unicode strings, but in older
789
# pythons this goes sufficiently wrong that it is a bad idea. (
790
# specifically a built in file with encoding 'UTF-8' will still try
791
# to encode using ascii.
792
new_encoding = osutils.get_terminal_encoding()
793
codec = codecs.lookup(new_encoding)
794
encode = codec.encode
795
# GZ 2010-09-08: Really we don't want to be writing arbitrary bytes,
796
# so should swap to the plain codecs.StreamWriter
797
stream = osutils.UnicodeOrBytesToBytesWriter(encode, stream,
799
stream.encoding = new_encoding
801
self.descriptions = descriptions
802
self.verbosity = verbosity
803
self._bench_history = bench_history
804
self._strict = strict
805
self._result_decorators = result_decorators or []
808
"Run the given test case or test suite."
809
if self.verbosity == 1:
810
result_class = TextTestResult
811
elif self.verbosity >= 2:
812
result_class = VerboseTestResult
813
original_result = result_class(self.stream,
816
bench_history=self._bench_history,
819
# Signal to result objects that look at stop early policy to stop,
820
original_result.stop_early = self.stop_on_failure
821
result = original_result
822
for decorator in self._result_decorators:
823
result = decorator(result)
824
result.stop_early = self.stop_on_failure
825
result.startTestRun()
830
# higher level code uses our extended protocol to determine
831
# what exit code to give.
832
return original_result
835
def iter_suite_tests(suite):
836
"""Return all tests in a suite, recursing through nested suites"""
837
if isinstance(suite, unittest.TestCase):
839
elif isinstance(suite, unittest.TestSuite):
841
for r in iter_suite_tests(item):
844
raise Exception('unknown type %r for object %r'
845
% (type(suite), suite))
848
TestSkipped = testtools.testcase.TestSkipped
851
class TestNotApplicable(TestSkipped):
852
"""A test is not applicable to the situation where it was run.
854
This is only normally raised by parameterized tests, if they find that
855
the instance they're constructed upon does not support one aspect
860
# traceback._some_str fails to format exceptions that have the default
861
# __str__ which does an implicit ascii conversion. However, repr() on those
862
# objects works, for all that its not quite what the doctor may have ordered.
863
def _clever_some_str(value):
866
except BaseException:
868
return repr(value).replace('\\n', '\n')
869
except BaseException:
870
return '<unprintable %s object>' % type(value).__name__
873
traceback._some_str = _clever_some_str
876
# deprecated - use self.knownFailure(), or self.expectFailure.
877
KnownFailure = testtools.testcase._ExpectedFailure
880
class UnavailableFeature(Exception):
881
"""A feature required for this test was not available.
883
This can be considered a specialised form of SkippedTest.
885
The feature should be used to construct the exception.
889
class StringIOWrapper(ui_testing.BytesIOWithEncoding):
891
@symbol_versioning.deprecated_method(
892
symbol_versioning.deprecated_in((3, 0)))
893
def __init__(self, s=None):
894
super(StringIOWrapper, self).__init__(s)
897
TestUIFactory = ui_testing.TestUIFactory
900
def isolated_doctest_setUp(test):
901
override_os_environ(test)
902
osutils.set_or_unset_env('BRZ_HOME', '/nonexistent')
903
test._orig_ui_factory = ui.ui_factory
904
ui.ui_factory = ui.SilentUIFactory()
907
def isolated_doctest_tearDown(test):
908
restore_os_environ(test)
909
ui.ui_factory = test._orig_ui_factory
912
def IsolatedDocTestSuite(*args, **kwargs):
913
"""Overrides doctest.DocTestSuite to handle isolation.
915
The method is really a factory and users are expected to use it as such.
918
kwargs['setUp'] = isolated_doctest_setUp
919
kwargs['tearDown'] = isolated_doctest_tearDown
920
return doctest.DocTestSuite(*args, **kwargs)
923
class TestCase(testtools.TestCase):
924
"""Base class for brz unit tests.
926
Tests that need access to disk resources should subclass
927
TestCaseInTempDir not TestCase.
929
Error and debug log messages are redirected from their usual
930
location into a temporary file, the contents of which can be
931
retrieved by _get_log(). We use a real OS file, not an in-memory object,
932
so that it can also capture file IO. When the test completes this file
933
is read into memory and removed from disk.
935
There are also convenience functions to invoke bzr's command-line
936
routine, and to build and check brz trees.
938
In addition to the usual method of overriding tearDown(), this class also
939
allows subclasses to register cleanup functions via addCleanup, which are
940
run in order as the object is torn down. It's less likely this will be
941
accidentally overlooked.
945
# record lsprof data when performing benchmark calls.
946
_gather_lsprof_in_benchmarks = False
948
def __init__(self, methodName='testMethod'):
949
super(TestCase, self).__init__(methodName)
950
self._directory_isolation = True
951
self.exception_handlers.insert(
952
0, (UnavailableFeature, self._do_unsupported_or_skip))
953
self.exception_handlers.insert(
954
0, (TestNotApplicable, self._do_not_applicable))
957
super(TestCase, self).setUp()
959
# At this point we're still accessing the config files in $BRZ_HOME (as
960
# set by the user running selftest).
961
timeout = config.GlobalStack().get('selftest.timeout')
963
timeout_fixture = fixtures.TimeoutFixture(timeout)
964
timeout_fixture.setUp()
965
self.addCleanup(timeout_fixture.cleanUp)
967
for feature in getattr(self, '_test_needs_features', []):
968
self.requireFeature(feature)
969
self._cleanEnvironment()
971
self.overrideAttr(breezy.get_global_state(), 'cmdline_overrides',
972
config.CommandLineStore())
976
self._benchcalls = []
977
self._benchtime = None
979
self._track_transports()
981
self._clear_debug_flags()
982
# Isolate global verbosity level, to make sure it's reproducible
983
# between tests. We should get rid of this altogether: bug 656694. --
985
self.overrideAttr(breezy.trace, '_verbosity_level', 0)
986
self._log_files = set()
987
# Each key in the ``_counters`` dict holds a value for a different
988
# counter. When the test ends, addDetail() should be used to output the
989
# counter values. This happens in install_counter_hook().
991
if 'config_stats' in selftest_debug_flags:
992
self._install_config_stats_hooks()
993
# Do not use i18n for tests (unless the test reverses this)
999
# The sys preserved stdin/stdout should allow blackbox tests debugging
1000
pdb.Pdb(stdin=sys.__stdin__, stdout=sys.__stdout__
1001
).set_trace(sys._getframe().f_back)
1003
def discardDetail(self, name):
1004
"""Extend the addDetail, getDetails api so we can remove a detail.
1006
eg. brz always adds the 'log' detail at startup, but we don't want to
1007
include it for skipped, xfail, etc tests.
1009
It is safe to call this for a detail that doesn't exist, in case this
1010
gets called multiple times.
1012
# We cheat. details is stored in __details which means we shouldn't
1013
# touch it. but getDetails() returns the dict directly, so we can
1015
details = self.getDetails()
1019
def install_counter_hook(self, hooks, name, counter_name=None):
1020
"""Install a counting hook.
1022
Any hook can be counted as long as it doesn't need to return a value.
1024
:param hooks: Where the hook should be installed.
1026
:param name: The hook name that will be counted.
1028
:param counter_name: The counter identifier in ``_counters``, defaults
1031
_counters = self._counters # Avoid closing over self
1032
if counter_name is None:
1034
if counter_name in _counters:
1035
raise AssertionError('%s is already used as a counter name'
1037
_counters[counter_name] = 0
1038
self.addDetail(counter_name, content.Content(
1040
lambda: [b'%d' % (_counters[counter_name],)]))
1042
def increment_counter(*args, **kwargs):
1043
_counters[counter_name] += 1
1044
label = 'count %s calls' % (counter_name,)
1045
hooks.install_named_hook(name, increment_counter, label)
1046
self.addCleanup(hooks.uninstall_named_hook, name, label)
1048
def _install_config_stats_hooks(self):
1049
"""Install config hooks to count hook calls.
1052
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1053
self.install_counter_hook(config.ConfigHooks, hook_name,
1054
'config.%s' % (hook_name,))
1056
# The OldConfigHooks are private and need special handling to protect
1057
# against recursive tests (tests that run other tests), so we just do
1058
# manually what registering them into _builtin_known_hooks will provide
1060
self.overrideAttr(config, 'OldConfigHooks', config._OldConfigHooks())
1061
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1062
self.install_counter_hook(config.OldConfigHooks, hook_name,
1063
'old_config.%s' % (hook_name,))
1065
def _clear_debug_flags(self):
1066
"""Prevent externally set debug flags affecting tests.
1068
Tests that want to use debug flags can just set them in the
1069
debug_flags set during setup/teardown.
1071
# Start with a copy of the current debug flags we can safely modify.
1072
self.overrideAttr(debug, 'debug_flags', set(debug.debug_flags))
1073
if 'allow_debug' not in selftest_debug_flags:
1074
debug.debug_flags.clear()
1075
if 'disable_lock_checks' not in selftest_debug_flags:
1076
debug.debug_flags.add('strict_locks')
1078
def _clear_hooks(self):
1079
# prevent hooks affecting tests
1080
known_hooks = hooks.known_hooks
1081
self._preserved_hooks = {}
1082
for key, (parent, name) in known_hooks.iter_parent_objects():
1083
current_hooks = getattr(parent, name)
1084
self._preserved_hooks[parent] = (name, current_hooks)
1085
self._preserved_lazy_hooks = hooks._lazy_hooks
1086
hooks._lazy_hooks = {}
1087
self.addCleanup(self._restoreHooks)
1088
for key, (parent, name) in known_hooks.iter_parent_objects():
1089
factory = known_hooks.get(key)
1090
setattr(parent, name, factory())
1091
# this hook should always be installed
1092
request._install_hook()
1094
def disable_directory_isolation(self):
1095
"""Turn off directory isolation checks."""
1096
self._directory_isolation = False
1098
def enable_directory_isolation(self):
1099
"""Enable directory isolation checks."""
1100
self._directory_isolation = True
1102
def _silenceUI(self):
1103
"""Turn off UI for duration of test"""
1104
# by default the UI is off; tests can turn it on if they want it.
1105
self.overrideAttr(ui, 'ui_factory', ui.SilentUIFactory())
1107
def _check_locks(self):
1108
"""Check that all lock take/release actions have been paired."""
1109
# We always check for mismatched locks. If a mismatch is found, we
1110
# fail unless -Edisable_lock_checks is supplied to selftest, in which
1111
# case we just print a warning.
1113
acquired_locks = [lock for action, lock in self._lock_actions
1114
if action == 'acquired']
1115
released_locks = [lock for action, lock in self._lock_actions
1116
if action == 'released']
1117
broken_locks = [lock for action, lock in self._lock_actions
1118
if action == 'broken']
1119
# trivially, given the tests for lock acquistion and release, if we
1120
# have as many in each list, it should be ok. Some lock tests also
1121
# break some locks on purpose and should be taken into account by
1122
# considering that breaking a lock is just a dirty way of releasing it.
1123
if len(acquired_locks) != (len(released_locks) + len(broken_locks)):
1125
'Different number of acquired and '
1126
'released or broken locks.\n'
1130
(acquired_locks, released_locks, broken_locks))
1131
if not self._lock_check_thorough:
1132
# Rather than fail, just warn
1133
print("Broken test %s: %s" % (self, message))
1137
def _track_locks(self):
1138
"""Track lock activity during tests."""
1139
self._lock_actions = []
1140
if 'disable_lock_checks' in selftest_debug_flags:
1141
self._lock_check_thorough = False
1143
self._lock_check_thorough = True
1145
self.addCleanup(self._check_locks)
1146
_mod_lock.Lock.hooks.install_named_hook('lock_acquired',
1147
self._lock_acquired, None)
1148
_mod_lock.Lock.hooks.install_named_hook('lock_released',
1149
self._lock_released, None)
1150
_mod_lock.Lock.hooks.install_named_hook('lock_broken',
1151
self._lock_broken, None)
1153
def _lock_acquired(self, result):
1154
self._lock_actions.append(('acquired', result))
1156
def _lock_released(self, result):
1157
self._lock_actions.append(('released', result))
1159
def _lock_broken(self, result):
1160
self._lock_actions.append(('broken', result))
1162
def permit_dir(self, name):
1163
"""Permit a directory to be used by this test. See permit_url."""
1164
name_transport = _mod_transport.get_transport_from_path(name)
1165
self.permit_url(name)
1166
self.permit_url(name_transport.base)
1168
def permit_url(self, url):
1169
"""Declare that url is an ok url to use in this test.
1171
Do this for memory transports, temporary test directory etc.
1173
Do not do this for the current working directory, /tmp, or any other
1174
preexisting non isolated url.
1176
if not url.endswith('/'):
1178
self._bzr_selftest_roots.append(url)
1180
def permit_source_tree_branch_repo(self):
1181
"""Permit the source tree brz is running from to be opened.
1183
Some code such as breezy.version attempts to read from the brz branch
1184
that brz is executing from (if any). This method permits that directory
1185
to be used in the test suite.
1187
path = self.get_source_path()
1188
self.record_directory_isolation()
1191
workingtree.WorkingTree.open(path)
1192
except (errors.NotBranchError, errors.NoWorkingTree):
1193
raise TestSkipped('Needs a working tree of brz sources')
1195
self.enable_directory_isolation()
1197
def _preopen_isolate_transport(self, transport):
1198
"""Check that all transport openings are done in the test work area."""
1199
while isinstance(transport, pathfilter.PathFilteringTransport):
1200
# Unwrap pathfiltered transports
1201
transport = transport.server.backing_transport.clone(
1202
transport._filter('.'))
1203
url = transport.base
1204
# ReadonlySmartTCPServer_for_testing decorates the backing transport
1205
# urls it is given by prepending readonly+. This is appropriate as the
1206
# client shouldn't know that the server is readonly (or not readonly).
1207
# We could register all servers twice, with readonly+ prepending, but
1208
# that makes for a long list; this is about the same but easier to
1210
if url.startswith('readonly+'):
1211
url = url[len('readonly+'):]
1212
self._preopen_isolate_url(url)
1214
def _preopen_isolate_url(self, url):
1215
if not self._directory_isolation:
1217
if self._directory_isolation == 'record':
1218
self._bzr_selftest_roots.append(url)
1220
# This prevents all transports, including e.g. sftp ones backed on disk
1221
# from working unless they are explicitly granted permission. We then
1222
# depend on the code that sets up test transports to check that they
1223
# are appropriately isolated and enable their use by calling
1224
# self.permit_transport()
1225
if not osutils.is_inside_any(self._bzr_selftest_roots, url):
1226
raise errors.BzrError("Attempt to escape test isolation: %r %r"
1227
% (url, self._bzr_selftest_roots))
1229
def record_directory_isolation(self):
1230
"""Gather accessed directories to permit later access.
1232
This is used for tests that access the branch brz is running from.
1234
self._directory_isolation = "record"
1236
def start_server(self, transport_server, backing_server=None):
1237
"""Start transport_server for this test.
1239
This starts the server, registers a cleanup for it and permits the
1240
server's urls to be used.
1242
if backing_server is None:
1243
transport_server.start_server()
1245
transport_server.start_server(backing_server)
1246
self.addCleanup(transport_server.stop_server)
1247
# Obtain a real transport because if the server supplies a password, it
1248
# will be hidden from the base on the client side.
1249
t = _mod_transport.get_transport_from_url(transport_server.get_url())
1250
# Some transport servers effectively chroot the backing transport;
1251
# others like SFTPServer don't - users of the transport can walk up the
1252
# transport to read the entire backing transport. This wouldn't matter
1253
# except that the workdir tests are given - and that they expect the
1254
# server's url to point at - is one directory under the safety net. So
1255
# Branch operations into the transport will attempt to walk up one
1256
# directory. Chrooting all servers would avoid this but also mean that
1257
# we wouldn't be testing directly against non-root urls. Alternatively
1258
# getting the test framework to start the server with a backing server
1259
# at the actual safety net directory would work too, but this then
1260
# means that the self.get_url/self.get_transport methods would need
1261
# to transform all their results. On balance its cleaner to handle it
1262
# here, and permit a higher url when we have one of these transports.
1263
if t.base.endswith('/work/'):
1264
# we have safety net/test root/work
1265
t = t.clone('../..')
1266
elif isinstance(transport_server,
1267
test_server.SmartTCPServer_for_testing):
1268
# The smart server adds a path similar to work, which is traversed
1269
# up from by the client. But the server is chrooted - the actual
1270
# backing transport is not escaped from, and VFS requests to the
1271
# root will error (because they try to escape the chroot).
1273
while t2.base != t.base:
1276
self.permit_url(t.base)
1278
def _track_transports(self):
1279
"""Install checks for transport usage."""
1280
# TestCase has no safe place it can write to.
1281
self._bzr_selftest_roots = []
1282
# Currently the easiest way to be sure that nothing is going on is to
1283
# hook into brz dir opening. This leaves a small window of error for
1284
# transport tests, but they are well known, and we can improve on this
1286
controldir.ControlDir.hooks.install_named_hook(
1287
"pre_open", self._preopen_isolate_transport,
1288
"Check brz directories are safe.")
1290
def _ndiff_strings(self, a, b):
1291
"""Return ndiff between two strings containing lines.
1293
A trailing newline is added if missing to make the strings
1295
if b and not b.endswith('\n'):
1297
if a and not a.endswith('\n'):
1299
difflines = difflib.ndiff(a.splitlines(True),
1301
linejunk=lambda x: False,
1302
charjunk=lambda x: False)
1303
return ''.join(difflines)
1305
def assertEqual(self, a, b, message=''):
1309
except UnicodeError as e:
1310
# If we can't compare without getting a UnicodeError, then
1311
# obviously they are different
1312
trace.mutter('UnicodeError: %s', e)
1315
raise AssertionError("%snot equal:\na = %s\nb = %s\n"
1317
pprint.pformat(a), pprint.pformat(b)))
1319
# FIXME: This is deprecated in unittest2 but plugins may still use it so we
1320
# need a deprecation period for them -- vila 2016-02-01
1321
assertEquals = assertEqual
1323
def assertEqualDiff(self, a, b, message=None):
1324
"""Assert two texts are equal, if not raise an exception.
1326
This is intended for use with multi-line strings where it can
1327
be hard to find the differences by eye.
1329
# TODO: perhaps override assertEqual to call this for strings?
1333
message = "texts not equal:\n"
1334
if a + ('\n' if isinstance(a, text_type) else b'\n') == b:
1335
message = 'first string is missing a final newline.\n'
1336
if a == b + ('\n' if isinstance(b, text_type) else b'\n'):
1337
message = 'second string is missing a final newline.\n'
1338
raise AssertionError(message
1339
+ self._ndiff_strings(a, b))
1341
def assertEqualMode(self, mode, mode_test):
1342
self.assertEqual(mode, mode_test,
1343
'mode mismatch %o != %o' % (mode, mode_test))
1345
def assertEqualStat(self, expected, actual):
1346
"""assert that expected and actual are the same stat result.
1348
:param expected: A stat result.
1349
:param actual: A stat result.
1350
:raises AssertionError: If the expected and actual stat values differ
1351
other than by atime.
1353
self.assertEqual(expected.st_size, actual.st_size,
1354
'st_size did not match')
1355
self.assertEqual(expected.st_mtime, actual.st_mtime,
1356
'st_mtime did not match')
1357
self.assertEqual(expected.st_ctime, actual.st_ctime,
1358
'st_ctime did not match')
1359
if sys.platform == 'win32':
1360
# On Win32 both 'dev' and 'ino' cannot be trusted. In python2.4 it
1361
# is 'dev' that varies, in python 2.5 (6?) it is st_ino that is
1362
# odd. We just force it to always be 0 to avoid any problems.
1363
self.assertEqual(0, expected.st_dev)
1364
self.assertEqual(0, actual.st_dev)
1365
self.assertEqual(0, expected.st_ino)
1366
self.assertEqual(0, actual.st_ino)
1368
self.assertEqual(expected.st_dev, actual.st_dev,
1369
'st_dev did not match')
1370
self.assertEqual(expected.st_ino, actual.st_ino,
1371
'st_ino did not match')
1372
self.assertEqual(expected.st_mode, actual.st_mode,
1373
'st_mode did not match')
1375
def assertLength(self, length, obj_with_len):
1376
"""Assert that obj_with_len is of length length."""
1377
if len(obj_with_len) != length:
1378
self.fail("Incorrect length: wanted %d, got %d for %r" % (
1379
length, len(obj_with_len), obj_with_len))
1381
def assertLogsError(self, exception_class, func, *args, **kwargs):
1382
"""Assert that `func(*args, **kwargs)` quietly logs a specific error.
1385
orig_log_exception_quietly = trace.log_exception_quietly
1388
orig_log_exception_quietly()
1389
captured.append(sys.exc_info()[1])
1390
trace.log_exception_quietly = capture
1391
func(*args, **kwargs)
1393
trace.log_exception_quietly = orig_log_exception_quietly
1394
self.assertLength(1, captured)
1396
self.assertIsInstance(err, exception_class)
1399
def assertPositive(self, val):
1400
"""Assert that val is greater than 0."""
1401
self.assertTrue(val > 0, 'expected a positive value, but got %s' % val)
1403
def assertNegative(self, val):
1404
"""Assert that val is less than 0."""
1405
self.assertTrue(val < 0, 'expected a negative value, but got %s' % val)
1407
def assertStartsWith(self, s, prefix):
1408
if not s.startswith(prefix):
1409
raise AssertionError(
1410
'string %r does not start with %r' % (s, prefix))
1412
def assertEndsWith(self, s, suffix):
1413
"""Asserts that s ends with suffix."""
1414
if not s.endswith(suffix):
1415
raise AssertionError(
1416
'string %r does not end with %r' % (s, suffix))
1418
def assertContainsRe(self, haystack, needle_re, flags=0):
1419
"""Assert that a contains something matching a regular expression."""
1420
if not re.search(needle_re, haystack, flags):
1421
if (('\n' if isinstance(haystack, str) else b'\n') in haystack or
1422
len(haystack) > 60):
1423
# a long string, format it in a more readable way
1424
raise AssertionError(
1425
'pattern "%s" not found in\n"""\\\n%s"""\n'
1426
% (needle_re, haystack))
1428
raise AssertionError('pattern "%s" not found in "%s"'
1429
% (needle_re, haystack))
1431
def assertNotContainsRe(self, haystack, needle_re, flags=0):
1432
"""Assert that a does not match a regular expression"""
1433
if re.search(needle_re, haystack, flags):
1434
raise AssertionError('pattern "%s" found in "%s"'
1435
% (needle_re, haystack))
1437
def assertContainsString(self, haystack, needle):
1438
if haystack.find(needle) == -1:
1439
self.fail("string %r not found in '''%s'''" % (needle, haystack))
1441
def assertNotContainsString(self, haystack, needle):
1442
if haystack.find(needle) != -1:
1443
self.fail("string %r found in '''%s'''" % (needle, haystack))
1445
def assertSubset(self, sublist, superlist):
1446
"""Assert that every entry in sublist is present in superlist."""
1447
missing = set(sublist) - set(superlist)
1448
if len(missing) > 0:
1449
raise AssertionError("value(s) %r not present in container %r" %
1450
(missing, superlist))
1452
def assertListRaises(self, excClass, func, *args, **kwargs):
1453
"""Fail unless excClass is raised when the iterator from func is used.
1455
Many functions can return generators this makes sure
1456
to wrap them in a list() call to make sure the whole generator
1457
is run, and that the proper exception is raised.
1460
list(func(*args, **kwargs))
1461
except excClass as e:
1464
if getattr(excClass, '__name__', None) is not None:
1465
excName = excClass.__name__
1467
excName = str(excClass)
1468
raise self.failureException("%s not raised" % excName)
1470
def assertRaises(self, excClass, callableObj, *args, **kwargs):
1471
"""Assert that a callable raises a particular exception.
1473
:param excClass: As for the except statement, this may be either an
1474
exception class, or a tuple of classes.
1475
:param callableObj: A callable, will be passed ``*args`` and
1478
Returns the exception so that you can examine it.
1481
callableObj(*args, **kwargs)
1482
except excClass as e:
1485
if getattr(excClass, '__name__', None) is not None:
1486
excName = excClass.__name__
1489
excName = str(excClass)
1490
raise self.failureException("%s not raised" % excName)
1492
def assertIs(self, left, right, message=None):
1493
if not (left is right):
1494
if message is not None:
1495
raise AssertionError(message)
1497
raise AssertionError("%r is not %r." % (left, right))
1499
def assertIsNot(self, left, right, message=None):
1501
if message is not None:
1502
raise AssertionError(message)
1504
raise AssertionError("%r is %r." % (left, right))
1506
def assertTransportMode(self, transport, path, mode):
1507
"""Fail if a path does not have mode "mode".
1509
If modes are not supported on this transport, the assertion is ignored.
1511
if not transport._can_roundtrip_unix_modebits():
1513
path_stat = transport.stat(path)
1514
actual_mode = stat.S_IMODE(path_stat.st_mode)
1515
self.assertEqual(mode, actual_mode,
1516
'mode of %r incorrect (%s != %s)'
1517
% (path, oct(mode), oct(actual_mode)))
1519
def assertIsSameRealPath(self, path1, path2):
1520
"""Fail if path1 and path2 points to different files"""
1521
self.assertEqual(osutils.realpath(path1),
1522
osutils.realpath(path2),
1523
"apparent paths:\na = %s\nb = %s\n," % (path1, path2))
1525
def assertIsInstance(self, obj, kls, msg=None):
1526
"""Fail if obj is not an instance of kls
1528
:param msg: Supplementary message to show if the assertion fails.
1530
if not isinstance(obj, kls):
1531
m = "%r is an instance of %s rather than %s" % (
1532
obj, obj.__class__, kls)
1537
def assertFileEqual(self, content, path):
1538
"""Fail if path does not contain 'content'."""
1539
self.assertPathExists(path)
1541
mode = 'r' + ('b' if isinstance(content, bytes) else '')
1542
with open(path, mode) as f:
1544
self.assertEqualDiff(content, s)
1546
def assertDocstring(self, expected_docstring, obj):
1547
"""Fail if obj does not have expected_docstring"""
1549
# With -OO the docstring should be None instead
1550
self.assertIs(obj.__doc__, None)
1552
self.assertEqual(expected_docstring, obj.__doc__)
1554
def assertPathExists(self, path):
1555
"""Fail unless path or paths, which may be abs or relative, exist."""
1556
# TODO(jelmer): Clean this up for pad.lv/1696545
1557
if not isinstance(path, (bytes, str, text_type)):
1559
self.assertPathExists(p)
1561
self.assertTrue(osutils.lexists(path),
1562
path + " does not exist")
1564
def assertPathDoesNotExist(self, path):
1565
"""Fail if path or paths, which may be abs or relative, exist."""
1566
if not isinstance(path, (str, text_type)):
1568
self.assertPathDoesNotExist(p)
1570
self.assertFalse(osutils.lexists(path),
1573
def _capture_deprecation_warnings(self, a_callable, *args, **kwargs):
1574
"""A helper for callDeprecated and applyDeprecated.
1576
:param a_callable: A callable to call.
1577
:param args: The positional arguments for the callable
1578
:param kwargs: The keyword arguments for the callable
1579
:return: A tuple (warnings, result). result is the result of calling
1580
a_callable(``*args``, ``**kwargs``).
1584
def capture_warnings(msg, cls=None, stacklevel=None):
1585
# we've hooked into a deprecation specific callpath,
1586
# only deprecations should getting sent via it.
1587
self.assertEqual(cls, DeprecationWarning)
1588
local_warnings.append(msg)
1589
original_warning_method = symbol_versioning.warn
1590
symbol_versioning.set_warning_method(capture_warnings)
1592
result = a_callable(*args, **kwargs)
1594
symbol_versioning.set_warning_method(original_warning_method)
1595
return (local_warnings, result)
1597
def applyDeprecated(self, deprecation_format, a_callable, *args, **kwargs):
1598
"""Call a deprecated callable without warning the user.
1600
Note that this only captures warnings raised by symbol_versioning.warn,
1601
not other callers that go direct to the warning module.
1603
To test that a deprecated method raises an error, do something like
1604
this (remember that both assertRaises and applyDeprecated delays *args
1605
and **kwargs passing)::
1607
self.assertRaises(errors.ReservedId,
1608
self.applyDeprecated,
1609
deprecated_in((1, 5, 0)),
1613
:param deprecation_format: The deprecation format that the callable
1614
should have been deprecated with. This is the same type as the
1615
parameter to deprecated_method/deprecated_function. If the
1616
callable is not deprecated with this format, an assertion error
1618
:param a_callable: A callable to call. This may be a bound method or
1619
a regular function. It will be called with ``*args`` and
1621
:param args: The positional arguments for the callable
1622
:param kwargs: The keyword arguments for the callable
1623
:return: The result of a_callable(``*args``, ``**kwargs``)
1625
call_warnings, result = self._capture_deprecation_warnings(
1626
a_callable, *args, **kwargs)
1627
expected_first_warning = symbol_versioning.deprecation_string(
1628
a_callable, deprecation_format)
1629
if len(call_warnings) == 0:
1630
self.fail("No deprecation warning generated by call to %s" %
1632
self.assertEqual(expected_first_warning, call_warnings[0])
1635
def callCatchWarnings(self, fn, *args, **kw):
1636
"""Call a callable that raises python warnings.
1638
The caller's responsible for examining the returned warnings.
1640
If the callable raises an exception, the exception is not
1641
caught and propagates up to the caller. In that case, the list
1642
of warnings is not available.
1644
:returns: ([warning_object, ...], fn_result)
1646
# XXX: This is not perfect, because it completely overrides the
1647
# warnings filters, and some code may depend on suppressing particular
1648
# warnings. It's the easiest way to insulate ourselves from -Werror,
1649
# though. -- Andrew, 20071062
1652
def _catcher(message, category, filename, lineno, file=None, line=None):
1653
# despite the name, 'message' is normally(?) a Warning subclass
1655
wlist.append(message)
1656
saved_showwarning = warnings.showwarning
1657
saved_filters = warnings.filters
1659
warnings.showwarning = _catcher
1660
warnings.filters = []
1661
result = fn(*args, **kw)
1663
warnings.showwarning = saved_showwarning
1664
warnings.filters = saved_filters
1665
return wlist, result
1667
def callDeprecated(self, expected, callable, *args, **kwargs):
1668
"""Assert that a callable is deprecated in a particular way.
1670
This is a very precise test for unusual requirements. The
1671
applyDeprecated helper function is probably more suited for most tests
1672
as it allows you to simply specify the deprecation format being used
1673
and will ensure that that is issued for the function being called.
1675
Note that this only captures warnings raised by symbol_versioning.warn,
1676
not other callers that go direct to the warning module. To catch
1677
general warnings, use callCatchWarnings.
1679
:param expected: a list of the deprecation warnings expected, in order
1680
:param callable: The callable to call
1681
:param args: The positional arguments for the callable
1682
:param kwargs: The keyword arguments for the callable
1684
call_warnings, result = self._capture_deprecation_warnings(
1685
callable, *args, **kwargs)
1686
self.assertEqual(expected, call_warnings)
1689
def _startLogFile(self):
1690
"""Setup a in-memory target for bzr and testcase log messages"""
1691
pseudo_log_file = BytesIO()
1693
def _get_log_contents_for_weird_testtools_api():
1694
return [pseudo_log_file.getvalue().decode(
1695
"utf-8", "replace").encode("utf-8")]
1697
"log", content.Content(
1698
content.ContentType("text", "plain", {"charset": "utf8"}),
1699
_get_log_contents_for_weird_testtools_api))
1700
self._log_file = pseudo_log_file
1701
self._log_memento = trace.push_log_file(self._log_file)
1702
self.addCleanup(self._finishLogFile)
1704
def _finishLogFile(self):
1705
"""Flush and dereference the in-memory log for this testcase"""
1706
if trace._trace_file:
1707
# flush the log file, to get all content
1708
trace._trace_file.flush()
1709
trace.pop_log_file(self._log_memento)
1710
# The logging module now tracks references for cleanup so discard ours
1711
del self._log_memento
1713
def thisFailsStrictLockCheck(self):
1714
"""It is known that this test would fail with -Dstrict_locks.
1716
By default, all tests are run with strict lock checking unless
1717
-Edisable_lock_checks is supplied. However there are some tests which
1718
we know fail strict locks at this point that have not been fixed.
1719
They should call this function to disable the strict checking.
1721
This should be used sparingly, it is much better to fix the locking
1722
issues rather than papering over the problem by calling this function.
1724
debug.debug_flags.discard('strict_locks')
1726
def overrideAttr(self, obj, attr_name, new=_unitialized_attr):
1727
"""Overrides an object attribute restoring it after the test.
1729
:note: This should be used with discretion; you should think about
1730
whether it's better to make the code testable without monkey-patching.
1732
:param obj: The object that will be mutated.
1734
:param attr_name: The attribute name we want to preserve/override in
1737
:param new: The optional value we want to set the attribute to.
1739
:returns: The actual attr value.
1741
# The actual value is captured by the call below
1742
value = getattr(obj, attr_name, _unitialized_attr)
1743
if value is _unitialized_attr:
1744
# When the test completes, the attribute should not exist, but if
1745
# we aren't setting a value, we don't need to do anything.
1746
if new is not _unitialized_attr:
1747
self.addCleanup(delattr, obj, attr_name)
1749
self.addCleanup(setattr, obj, attr_name, value)
1750
if new is not _unitialized_attr:
1751
setattr(obj, attr_name, new)
1754
def overrideEnv(self, name, new):
1755
"""Set an environment variable, and reset it after the test.
1757
:param name: The environment variable name.
1759
:param new: The value to set the variable to. If None, the
1760
variable is deleted from the environment.
1762
:returns: The actual variable value.
1764
value = osutils.set_or_unset_env(name, new)
1765
self.addCleanup(osutils.set_or_unset_env, name, value)
1768
def recordCalls(self, obj, attr_name):
1769
"""Monkeypatch in a wrapper that will record calls.
1771
The monkeypatch is automatically removed when the test concludes.
1773
:param obj: The namespace holding the reference to be replaced;
1774
typically a module, class, or object.
1775
:param attr_name: A string for the name of the attribute to patch.
1776
:returns: A list that will be extended with one item every time the
1777
function is called, with a tuple of (args, kwargs).
1781
def decorator(*args, **kwargs):
1782
calls.append((args, kwargs))
1783
return orig(*args, **kwargs)
1784
orig = self.overrideAttr(obj, attr_name, decorator)
1787
def _cleanEnvironment(self):
1788
for name, value in isolated_environ.items():
1789
self.overrideEnv(name, value)
1791
def _restoreHooks(self):
1792
for klass, (name, hooks) in self._preserved_hooks.items():
1793
setattr(klass, name, hooks)
1794
self._preserved_hooks.clear()
1795
breezy.hooks._lazy_hooks = self._preserved_lazy_hooks
1796
self._preserved_lazy_hooks.clear()
1798
def knownFailure(self, reason):
1799
"""Declare that this test fails for a known reason
1801
Tests that are known to fail should generally be using expectedFailure
1802
with an appropriate reverse assertion if a change could cause the test
1803
to start passing. Conversely if the test has no immediate prospect of
1804
succeeding then using skip is more suitable.
1806
When this method is called while an exception is being handled, that
1807
traceback will be used, otherwise a new exception will be thrown to
1808
provide one but won't be reported.
1810
self._add_reason(reason)
1812
exc_info = sys.exc_info()
1813
if exc_info != (None, None, None):
1814
self._report_traceback(exc_info)
1817
raise self.failureException(reason)
1818
except self.failureException:
1819
exc_info = sys.exc_info()
1820
# GZ 02-08-2011: Maybe cleanup this err.exc_info attribute too?
1821
raise testtools.testcase._ExpectedFailure(exc_info)
1825
def _suppress_log(self):
1826
"""Remove the log info from details."""
1827
self.discardDetail('log')
1829
def _do_skip(self, result, reason):
1830
self._suppress_log()
1831
addSkip = getattr(result, 'addSkip', None)
1832
if not callable(addSkip):
1833
result.addSuccess(result)
1835
addSkip(self, str(reason))
1838
def _do_known_failure(self, result, e):
1839
self._suppress_log()
1840
err = sys.exc_info()
1841
addExpectedFailure = getattr(result, 'addExpectedFailure', None)
1842
if addExpectedFailure is not None:
1843
addExpectedFailure(self, err)
1845
result.addSuccess(self)
1848
def _do_not_applicable(self, result, e):
1850
reason = 'No reason given'
1853
self._suppress_log()
1854
addNotApplicable = getattr(result, 'addNotApplicable', None)
1855
if addNotApplicable is not None:
1856
result.addNotApplicable(self, reason)
1858
self._do_skip(result, reason)
1861
def _report_skip(self, result, err):
1862
"""Override the default _report_skip.
1864
We want to strip the 'log' detail. If we waint until _do_skip, it has
1865
already been formatted into the 'reason' string, and we can't pull it
1868
self._suppress_log()
1869
super(TestCase, self)._report_skip(self, result, err)
1872
def _report_expected_failure(self, result, err):
1875
See _report_skip for motivation.
1877
self._suppress_log()
1878
super(TestCase, self)._report_expected_failure(self, result, err)
1881
def _do_unsupported_or_skip(self, result, e):
1883
self._suppress_log()
1884
addNotSupported = getattr(result, 'addNotSupported', None)
1885
if addNotSupported is not None:
1886
result.addNotSupported(self, reason)
1888
self._do_skip(result, reason)
1890
def time(self, callable, *args, **kwargs):
1891
"""Run callable and accrue the time it takes to the benchmark time.
1893
If lsprofiling is enabled (i.e. by --lsprof-time to brz selftest) then
1894
this will cause lsprofile statistics to be gathered and stored in
1897
if self._benchtime is None:
1898
self.addDetail('benchtime', content.Content(
1900
lambda: [str(self._benchtime).encode('utf-8')]))
1904
if not self._gather_lsprof_in_benchmarks:
1905
return callable(*args, **kwargs)
1907
# record this benchmark
1908
ret, stats = breezy.lsprof.profile(callable, *args, **kwargs)
1910
self._benchcalls.append(((callable, args, kwargs), stats))
1913
self._benchtime += time.time() - start
1915
def log(self, *args):
1919
"""Get a unicode string containing the log from breezy.trace.
1921
Undecodable characters are replaced.
1923
return u"".join(self.getDetails()['log'].iter_text())
1925
def requireFeature(self, feature):
1926
"""This test requires a specific feature is available.
1928
:raises UnavailableFeature: When feature is not available.
1930
if not feature.available():
1931
raise UnavailableFeature(feature)
1933
def _run_bzr_core(self, args, encoding, stdin, stdout, stderr,
1935
# Clear chk_map page cache, because the contents are likely to mask
1937
chk_map.clear_cache()
1939
self.log('run brz: %r', args)
1942
self._last_cmd_stdout = stdout
1943
self._last_cmd_stderr = stderr
1945
self._last_cmd_stdout = codecs.getwriter(encoding)(stdout)
1946
self._last_cmd_stderr = codecs.getwriter(encoding)(stderr)
1948
old_ui_factory = ui.ui_factory
1949
ui.ui_factory = ui_testing.TestUIFactory(
1951
stdout=self._last_cmd_stdout,
1952
stderr=self._last_cmd_stderr)
1955
if working_dir is not None:
1956
cwd = osutils.getcwd()
1957
os.chdir(working_dir)
1961
result = self.apply_redirected(
1962
ui.ui_factory.stdin,
1964
_mod_commands.run_bzr_catch_user_errors,
1967
ui.ui_factory = old_ui_factory
1973
def run_bzr_raw(self, args, retcode=0, stdin=None, encoding=None,
1974
working_dir=None, error_regexes=[]):
1975
"""Invoke brz, as if it were run from the command line.
1977
The argument list should not include the brz program name - the
1978
first argument is normally the brz command. Arguments may be
1979
passed in three ways:
1981
1- A list of strings, eg ["commit", "a"]. This is recommended
1982
when the command contains whitespace or metacharacters, or
1983
is built up at run time.
1985
2- A single string, eg "add a". This is the most convenient
1986
for hardcoded commands.
1988
This runs brz through the interface that catches and reports
1989
errors, and with logging set to something approximating the
1990
default, so that error reporting can be checked.
1992
This should be the main method for tests that want to exercise the
1993
overall behavior of the brz application (rather than a unit test
1994
or a functional test of the library.)
1996
This sends the stdout/stderr results into the test's log,
1997
where it may be useful for debugging. See also run_captured.
1999
:keyword stdin: A string to be used as stdin for the command.
2000
:keyword retcode: The status code the command should return;
2002
:keyword working_dir: The directory to run the command in
2003
:keyword error_regexes: A list of expected error messages. If
2004
specified they must be seen in the error output of the command.
2006
if isinstance(args, string_types):
2007
args = shlex.split(args)
2009
if encoding is None:
2010
encoding = osutils.get_user_encoding()
2012
if sys.version_info[0] == 2:
2013
wrapped_stdout = stdout = ui_testing.BytesIOWithEncoding()
2014
wrapped_stderr = stderr = ui_testing.BytesIOWithEncoding()
2015
stdout.encoding = stderr.encoding = encoding
2017
# FIXME: don't call into logging here
2018
handler = trace.EncodedStreamHandler(
2019
stderr, errors="replace")
2023
wrapped_stdout = TextIOWrapper(stdout, encoding)
2024
wrapped_stderr = TextIOWrapper(stderr, encoding)
2025
handler = logging.StreamHandler(wrapped_stderr)
2026
handler.setLevel(logging.INFO)
2028
logger = logging.getLogger('')
2029
logger.addHandler(handler)
2031
result = self._run_bzr_core(
2032
args, encoding=encoding, stdin=stdin, stdout=wrapped_stdout,
2033
stderr=wrapped_stderr, working_dir=working_dir,
2036
logger.removeHandler(handler)
2039
wrapped_stdout.flush()
2040
wrapped_stderr.flush()
2042
out = stdout.getvalue()
2043
err = stderr.getvalue()
2045
self.log('output:\n%r', out)
2047
self.log('errors:\n%r', err)
2048
if retcode is not None:
2049
self.assertEqual(retcode, result,
2050
message='Unexpected return code')
2051
self.assertIsInstance(error_regexes, (list, tuple))
2052
for regex in error_regexes:
2053
self.assertContainsRe(err, regex)
2056
def run_bzr(self, args, retcode=0, stdin=None, encoding=None,
2057
working_dir=None, error_regexes=[]):
2058
"""Invoke brz, as if it were run from the command line.
2060
The argument list should not include the brz program name - the
2061
first argument is normally the brz command. Arguments may be
2062
passed in three ways:
2064
1- A list of strings, eg ["commit", "a"]. This is recommended
2065
when the command contains whitespace or metacharacters, or
2066
is built up at run time.
2068
2- A single string, eg "add a". This is the most convenient
2069
for hardcoded commands.
2071
This runs brz through the interface that catches and reports
2072
errors, and with logging set to something approximating the
2073
default, so that error reporting can be checked.
2075
This should be the main method for tests that want to exercise the
2076
overall behavior of the brz application (rather than a unit test
2077
or a functional test of the library.)
2079
This sends the stdout/stderr results into the test's log,
2080
where it may be useful for debugging. See also run_captured.
2082
:keyword stdin: A string to be used as stdin for the command.
2083
:keyword retcode: The status code the command should return;
2085
:keyword working_dir: The directory to run the command in
2086
:keyword error_regexes: A list of expected error messages. If
2087
specified they must be seen in the error output of the command.
2089
if isinstance(args, string_types):
2090
args = shlex.split(args)
2092
if encoding is None:
2093
encoding = osutils.get_user_encoding()
2095
if sys.version_info[0] == 2:
2096
stdout = ui_testing.BytesIOWithEncoding()
2097
stderr = ui_testing.BytesIOWithEncoding()
2098
stdout.encoding = stderr.encoding = encoding
2099
# FIXME: don't call into logging here
2100
handler = trace.EncodedStreamHandler(
2101
stderr, errors="replace")
2103
stdout = ui_testing.StringIOWithEncoding()
2104
stderr = ui_testing.StringIOWithEncoding()
2105
stdout.encoding = stderr.encoding = encoding
2106
handler = logging.StreamHandler(stream=stderr)
2107
handler.setLevel(logging.INFO)
2109
logger = logging.getLogger('')
2110
logger.addHandler(handler)
2113
result = self._run_bzr_core(
2114
args, encoding=encoding, stdin=stdin, stdout=stdout,
2115
stderr=stderr, working_dir=working_dir)
2117
logger.removeHandler(handler)
2119
out = stdout.getvalue()
2120
err = stderr.getvalue()
2122
self.log('output:\n%r', out)
2124
self.log('errors:\n%r', err)
2125
if retcode is not None:
2126
self.assertEqual(retcode, result,
2127
message='Unexpected return code')
2128
self.assertIsInstance(error_regexes, (list, tuple))
2129
for regex in error_regexes:
2130
self.assertContainsRe(err, regex)
2133
def run_bzr_error(self, error_regexes, *args, **kwargs):
2134
"""Run brz, and check that stderr contains the supplied regexes
2136
:param error_regexes: Sequence of regular expressions which
2137
must each be found in the error output. The relative ordering
2139
:param args: command-line arguments for brz
2140
:param kwargs: Keyword arguments which are interpreted by run_brz
2141
This function changes the default value of retcode to be 3,
2142
since in most cases this is run when you expect brz to fail.
2144
:return: (out, err) The actual output of running the command (in case
2145
you want to do more inspection)
2149
# Make sure that commit is failing because there is nothing to do
2150
self.run_bzr_error(['no changes to commit'],
2151
['commit', '-m', 'my commit comment'])
2152
# Make sure --strict is handling an unknown file, rather than
2153
# giving us the 'nothing to do' error
2154
self.build_tree(['unknown'])
2156
['Commit refused because there are unknown files'],
2157
['commit', --strict', '-m', 'my commit comment'])
2159
kwargs.setdefault('retcode', 3)
2160
kwargs['error_regexes'] = error_regexes
2161
out, err = self.run_bzr(*args, **kwargs)
2164
def run_bzr_subprocess(self, *args, **kwargs):
2165
"""Run brz in a subprocess for testing.
2167
This starts a new Python interpreter and runs brz in there.
2168
This should only be used for tests that have a justifiable need for
2169
this isolation: e.g. they are testing startup time, or signal
2170
handling, or early startup code, etc. Subprocess code can't be
2171
profiled or debugged so easily.
2173
:keyword retcode: The status code that is expected. Defaults to 0. If
2174
None is supplied, the status code is not checked.
2175
:keyword env_changes: A dictionary which lists changes to environment
2176
variables. A value of None will unset the env variable.
2177
The values must be strings. The change will only occur in the
2178
child, so you don't need to fix the environment after running.
2179
:keyword universal_newlines: Convert CRLF => LF
2180
:keyword allow_plugins: By default the subprocess is run with
2181
--no-plugins to ensure test reproducibility. Also, it is possible
2182
for system-wide plugins to create unexpected output on stderr,
2183
which can cause unnecessary test failures.
2185
env_changes = kwargs.get('env_changes', {})
2186
working_dir = kwargs.get('working_dir', None)
2187
allow_plugins = kwargs.get('allow_plugins', False)
2189
if isinstance(args[0], list):
2191
elif isinstance(args[0], (str, text_type)):
2192
args = list(shlex.split(args[0]))
2194
raise ValueError("passing varargs to run_bzr_subprocess")
2195
process = self.start_bzr_subprocess(args, env_changes=env_changes,
2196
working_dir=working_dir,
2197
allow_plugins=allow_plugins)
2198
# We distinguish between retcode=None and retcode not passed.
2199
supplied_retcode = kwargs.get('retcode', 0)
2200
return self.finish_bzr_subprocess(process, retcode=supplied_retcode,
2201
universal_newlines=kwargs.get(
2202
'universal_newlines', False),
2205
def start_bzr_subprocess(self, process_args, env_changes=None,
2206
skip_if_plan_to_signal=False,
2208
allow_plugins=False, stderr=subprocess.PIPE):
2209
"""Start brz in a subprocess for testing.
2211
This starts a new Python interpreter and runs brz in there.
2212
This should only be used for tests that have a justifiable need for
2213
this isolation: e.g. they are testing startup time, or signal
2214
handling, or early startup code, etc. Subprocess code can't be
2215
profiled or debugged so easily.
2217
:param process_args: a list of arguments to pass to the brz executable,
2218
for example ``['--version']``.
2219
:param env_changes: A dictionary which lists changes to environment
2220
variables. A value of None will unset the env variable.
2221
The values must be strings. The change will only occur in the
2222
child, so you don't need to fix the environment after running.
2223
:param skip_if_plan_to_signal: raise TestSkipped when true and system
2224
doesn't support signalling subprocesses.
2225
:param allow_plugins: If False (default) pass --no-plugins to brz.
2226
:param stderr: file to use for the subprocess's stderr. Valid values
2227
are those valid for the stderr argument of `subprocess.Popen`.
2228
Default value is ``subprocess.PIPE``.
2230
:returns: Popen object for the started process.
2232
if skip_if_plan_to_signal:
2233
if os.name != "posix":
2234
raise TestSkipped("Sending signals not supported")
2236
if env_changes is None:
2238
# Because $HOME is set to a tempdir for the context of a test, modules
2239
# installed in the user dir will not be found unless $PYTHONUSERBASE
2240
# gets set to the computed directory of this parent process.
2241
if site.USER_BASE is not None:
2242
env_changes["PYTHONUSERBASE"] = site.USER_BASE
2245
def cleanup_environment():
2246
for env_var, value in env_changes.items():
2247
old_env[env_var] = osutils.set_or_unset_env(env_var, value)
2249
def restore_environment():
2250
for env_var, value in old_env.items():
2251
osutils.set_or_unset_env(env_var, value)
2253
bzr_path = self.get_brz_path()
2256
if working_dir is not None:
2257
cwd = osutils.getcwd()
2258
os.chdir(working_dir)
2261
# win32 subprocess doesn't support preexec_fn
2262
# so we will avoid using it on all platforms, just to
2263
# make sure the code path is used, and we don't break on win32
2264
cleanup_environment()
2265
# Include the subprocess's log file in the test details, in case
2266
# the test fails due to an error in the subprocess.
2267
self._add_subprocess_log(trace._get_brz_log_filename())
2268
command = [sys.executable]
2269
# frozen executables don't need the path to bzr
2270
if getattr(sys, "frozen", None) is None:
2271
command.append(bzr_path)
2272
if not allow_plugins:
2273
command.append('--no-plugins')
2274
command.extend(process_args)
2275
process = self._popen(command, stdin=subprocess.PIPE,
2276
stdout=subprocess.PIPE,
2277
stderr=stderr, bufsize=0)
2279
restore_environment()
2285
def _add_subprocess_log(self, log_file_path):
2286
if len(self._log_files) == 0:
2287
# Register an addCleanup func. We do this on the first call to
2288
# _add_subprocess_log rather than in TestCase.setUp so that this
2289
# addCleanup is registered after any cleanups for tempdirs that
2290
# subclasses might create, which will probably remove the log file
2292
self.addCleanup(self._subprocess_log_cleanup)
2293
# self._log_files is a set, so if a log file is reused we won't grab it
2295
self._log_files.add(log_file_path)
2297
def _subprocess_log_cleanup(self):
2298
for count, log_file_path in enumerate(self._log_files):
2299
# We use buffer_now=True to avoid holding the file open beyond
2300
# the life of this function, which might interfere with e.g.
2301
# cleaning tempdirs on Windows.
2302
# XXX: Testtools 0.9.5 doesn't have the content_from_file helper
2303
# detail_content = content.content_from_file(
2304
# log_file_path, buffer_now=True)
2305
with open(log_file_path, 'rb') as log_file:
2306
log_file_bytes = log_file.read()
2307
detail_content = content.Content(
2308
content.ContentType("text", "plain", {"charset": "utf8"}),
2309
lambda: [log_file_bytes])
2310
self.addDetail("start_bzr_subprocess-log-%d" % (count,),
2313
def _popen(self, *args, **kwargs):
2314
"""Place a call to Popen.
2316
Allows tests to override this method to intercept the calls made to
2317
Popen for introspection.
2319
return subprocess.Popen(*args, **kwargs)
2321
def get_source_path(self):
2322
"""Return the path of the directory containing breezy."""
2323
return os.path.dirname(os.path.dirname(breezy.__file__))
2325
def get_brz_path(self):
2326
"""Return the path of the 'brz' executable for this test suite."""
2327
brz_path = os.path.join(self.get_source_path(), "brz")
2328
if not os.path.isfile(brz_path):
2329
# We are probably installed. Assume sys.argv is the right file
2330
brz_path = sys.argv[0]
2333
def finish_bzr_subprocess(self, process, retcode=0, send_signal=None,
2334
universal_newlines=False, process_args=None):
2335
"""Finish the execution of process.
2337
:param process: the Popen object returned from start_bzr_subprocess.
2338
:param retcode: The status code that is expected. Defaults to 0. If
2339
None is supplied, the status code is not checked.
2340
:param send_signal: an optional signal to send to the process.
2341
:param universal_newlines: Convert CRLF => LF
2342
:returns: (stdout, stderr)
2344
if send_signal is not None:
2345
os.kill(process.pid, send_signal)
2346
out, err = process.communicate()
2348
if universal_newlines:
2349
out = out.replace(b'\r\n', b'\n')
2350
err = err.replace(b'\r\n', b'\n')
2352
if retcode is not None and retcode != process.returncode:
2353
if process_args is None:
2354
process_args = "(unknown args)"
2355
trace.mutter('Output of brz %r:\n%s', process_args, out)
2356
trace.mutter('Error for brz %r:\n%s', process_args, err)
2357
self.fail('Command brz %r failed with retcode %d != %d'
2358
% (process_args, retcode, process.returncode))
2361
def check_tree_shape(self, tree, shape):
2362
"""Compare a tree to a list of expected names.
2364
Fail if they are not precisely equal.
2367
shape = list(shape) # copy
2368
for path, ie in tree.iter_entries_by_dir():
2369
name = path.replace('\\', '/')
2370
if ie.kind == 'directory':
2373
pass # ignore root entry
2379
self.fail("expected paths not found in inventory: %r" % shape)
2381
self.fail("unexpected paths found in inventory: %r" % extras)
2383
def apply_redirected(self, stdin=None, stdout=None, stderr=None,
2384
a_callable=None, *args, **kwargs):
2385
"""Call callable with redirected std io pipes.
2387
Returns the return code."""
2388
if not callable(a_callable):
2389
raise ValueError("a_callable must be callable.")
2391
stdin = BytesIO(b"")
2393
if getattr(self, "_log_file", None) is not None:
2394
stdout = self._log_file
2396
if sys.version_info[0] == 2:
2401
if getattr(self, "_log_file", None is not None):
2402
stderr = self._log_file
2404
if sys.version_info[0] == 2:
2408
real_stdin = sys.stdin
2409
real_stdout = sys.stdout
2410
real_stderr = sys.stderr
2415
return a_callable(*args, **kwargs)
2417
sys.stdout = real_stdout
2418
sys.stderr = real_stderr
2419
sys.stdin = real_stdin
2421
def reduceLockdirTimeout(self):
2422
"""Reduce the default lock timeout for the duration of the test, so that
2423
if LockContention occurs during a test, it does so quickly.
2425
Tests that expect to provoke LockContention errors should call this.
2427
self.overrideAttr(lockdir, '_DEFAULT_TIMEOUT_SECONDS', 0)
2429
def make_utf8_encoded_stringio(self, encoding_type=None):
2430
"""Return a wrapped BytesIO, that will encode text input to UTF-8."""
2431
if encoding_type is None:
2432
encoding_type = 'strict'
2434
output_encoding = 'utf-8'
2435
sio = codecs.getwriter(output_encoding)(bio, errors=encoding_type)
2436
sio.encoding = output_encoding
2439
def disable_verb(self, verb):
2440
"""Disable a smart server verb for one test."""
2441
from breezy.bzr.smart import request
2442
request_handlers = request.request_handlers
2443
orig_method = request_handlers.get(verb)
2444
orig_info = request_handlers.get_info(verb)
2445
request_handlers.remove(verb)
2446
self.addCleanup(request_handlers.register, verb, orig_method,
2453
class CapturedCall(object):
2454
"""A helper for capturing smart server calls for easy debug analysis."""
2456
def __init__(self, params, prefix_length):
2457
"""Capture the call with params and skip prefix_length stack frames."""
2460
# The last 5 frames are the __init__, the hook frame, and 3 smart
2461
# client frames. Beyond this we could get more clever, but this is good
2463
stack = traceback.extract_stack()[prefix_length:-5]
2464
self.stack = ''.join(traceback.format_list(stack))
2467
return self.call.method
2470
return self.call.method
2476
class TestCaseWithMemoryTransport(TestCase):
2477
"""Common test class for tests that do not need disk resources.
2479
Tests that need disk resources should derive from TestCaseInTempDir
2480
orTestCaseWithTransport.
2482
TestCaseWithMemoryTransport sets the TEST_ROOT variable for all brz tests.
2484
For TestCaseWithMemoryTransport the ``test_home_dir`` is set to the name of
2485
a directory which does not exist. This serves to help ensure test isolation
2486
is preserved. ``test_dir`` is set to the TEST_ROOT, as is cwd, because they
2487
must exist. However, TestCaseWithMemoryTransport does not offer local file
2488
defaults for the transport in tests, nor does it obey the command line
2489
override, so tests that accidentally write to the common directory should
2492
:cvar TEST_ROOT: Directory containing all temporary directories, plus a
2493
``.bzr`` directory that stops us ascending higher into the filesystem.
2499
def __init__(self, methodName='runTest'):
2500
# allow test parameterization after test construction and before test
2501
# execution. Variables that the parameterizer sets need to be
2502
# ones that are not set by setUp, or setUp will trash them.
2503
super(TestCaseWithMemoryTransport, self).__init__(methodName)
2504
self.vfs_transport_factory = default_transport
2505
self.transport_server = None
2506
self.transport_readonly_server = None
2507
self.__vfs_server = None
2510
super(TestCaseWithMemoryTransport, self).setUp()
2512
def _add_disconnect_cleanup(transport):
2513
"""Schedule disconnection of given transport at test cleanup
2515
This needs to happen for all connected transports or leaks occur.
2517
Note reconnections may mean we call disconnect multiple times per
2518
transport which is suboptimal but seems harmless.
2520
self.addCleanup(transport.disconnect)
2522
_mod_transport.Transport.hooks.install_named_hook(
2523
'post_connect', _add_disconnect_cleanup, None)
2525
self._make_test_root()
2526
self.addCleanup(os.chdir, osutils.getcwd())
2527
self.makeAndChdirToTestDir()
2528
self.overrideEnvironmentForTesting()
2529
self.__readonly_server = None
2530
self.__server = None
2531
self.reduceLockdirTimeout()
2532
# Each test may use its own config files even if the local config files
2533
# don't actually exist. They'll rightly fail if they try to create them
2535
self.overrideAttr(config, '_shared_stores', {})
2537
def get_transport(self, relpath=None):
2538
"""Return a writeable transport.
2540
This transport is for the test scratch space relative to
2543
:param relpath: a path relative to the base url.
2545
t = _mod_transport.get_transport_from_url(self.get_url(relpath))
2546
self.assertFalse(t.is_readonly())
2549
def get_readonly_transport(self, relpath=None):
2550
"""Return a readonly transport for the test scratch space
2552
This can be used to test that operations which should only need
2553
readonly access in fact do not try to write.
2555
:param relpath: a path relative to the base url.
2557
t = _mod_transport.get_transport_from_url(
2558
self.get_readonly_url(relpath))
2559
self.assertTrue(t.is_readonly())
2562
def create_transport_readonly_server(self):
2563
"""Create a transport server from class defined at init.
2565
This is mostly a hook for daughter classes.
2567
return self.transport_readonly_server()
2569
def get_readonly_server(self):
2570
"""Get the server instance for the readonly transport
2572
This is useful for some tests with specific servers to do diagnostics.
2574
if self.__readonly_server is None:
2575
if self.transport_readonly_server is None:
2576
# readonly decorator requested
2577
self.__readonly_server = test_server.ReadonlyServer()
2579
# explicit readonly transport.
2580
self.__readonly_server = (
2581
self.create_transport_readonly_server())
2582
self.start_server(self.__readonly_server,
2583
self.get_vfs_only_server())
2584
return self.__readonly_server
2586
def get_readonly_url(self, relpath=None):
2587
"""Get a URL for the readonly transport.
2589
This will either be backed by '.' or a decorator to the transport
2590
used by self.get_url()
2591
relpath provides for clients to get a path relative to the base url.
2592
These should only be downwards relative, not upwards.
2594
base = self.get_readonly_server().get_url()
2595
return self._adjust_url(base, relpath)
2597
def get_vfs_only_server(self):
2598
"""Get the vfs only read/write server instance.
2600
This is useful for some tests with specific servers that need
2603
For TestCaseWithMemoryTransport this is always a MemoryServer, and there
2604
is no means to override it.
2606
if self.__vfs_server is None:
2607
self.__vfs_server = memory.MemoryServer()
2608
self.start_server(self.__vfs_server)
2609
return self.__vfs_server
2611
def get_server(self):
2612
"""Get the read/write server instance.
2614
This is useful for some tests with specific servers that need
2617
This is built from the self.transport_server factory. If that is None,
2618
then the self.get_vfs_server is returned.
2620
if self.__server is None:
2621
if (self.transport_server is None or self.transport_server is
2622
self.vfs_transport_factory):
2623
self.__server = self.get_vfs_only_server()
2625
# bring up a decorated means of access to the vfs only server.
2626
self.__server = self.transport_server()
2627
self.start_server(self.__server, self.get_vfs_only_server())
2628
return self.__server
2630
def _adjust_url(self, base, relpath):
2631
"""Get a URL (or maybe a path) for the readwrite transport.
2633
This will either be backed by '.' or to an equivalent non-file based
2635
relpath provides for clients to get a path relative to the base url.
2636
These should only be downwards relative, not upwards.
2638
if relpath is not None and relpath != '.':
2639
if not base.endswith('/'):
2641
# XXX: Really base should be a url; we did after all call
2642
# get_url()! But sometimes it's just a path (from
2643
# LocalAbspathServer), and it'd be wrong to append urlescaped data
2644
# to a non-escaped local path.
2645
if base.startswith('./') or base.startswith('/'):
2648
base += urlutils.escape(relpath)
2651
def get_url(self, relpath=None):
2652
"""Get a URL (or maybe a path) for the readwrite transport.
2654
This will either be backed by '.' or to an equivalent non-file based
2656
relpath provides for clients to get a path relative to the base url.
2657
These should only be downwards relative, not upwards.
2659
base = self.get_server().get_url()
2660
return self._adjust_url(base, relpath)
2662
def get_vfs_only_url(self, relpath=None):
2663
"""Get a URL (or maybe a path for the plain old vfs transport.
2665
This will never be a smart protocol. It always has all the
2666
capabilities of the local filesystem, but it might actually be a
2667
MemoryTransport or some other similar virtual filesystem.
2669
This is the backing transport (if any) of the server returned by
2670
get_url and get_readonly_url.
2672
:param relpath: provides for clients to get a path relative to the base
2673
url. These should only be downwards relative, not upwards.
2676
base = self.get_vfs_only_server().get_url()
2677
return self._adjust_url(base, relpath)
2679
def _create_safety_net(self):
2680
"""Make a fake bzr directory.
2682
This prevents any tests propagating up onto the TEST_ROOT directory's
2685
root = TestCaseWithMemoryTransport.TEST_ROOT
2687
# Make sure we get a readable and accessible home for .brz.log
2688
# and/or config files, and not fallback to weird defaults (see
2689
# http://pad.lv/825027).
2690
self.assertIs(None, os.environ.get('BRZ_HOME', None))
2691
os.environ['BRZ_HOME'] = root
2692
wt = controldir.ControlDir.create_standalone_workingtree(root)
2693
del os.environ['BRZ_HOME']
2694
except Exception as e:
2695
self.fail("Fail to initialize the safety net: %r\n" % (e,))
2696
# Hack for speed: remember the raw bytes of the dirstate file so that
2697
# we don't need to re-open the wt to check it hasn't changed.
2698
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE = (
2699
wt.control_transport.get_bytes('dirstate'))
2701
def _check_safety_net(self):
2702
"""Check that the safety .bzr directory have not been touched.
2704
_make_test_root have created a .bzr directory to prevent tests from
2705
propagating. This method ensures than a test did not leaked.
2707
root = TestCaseWithMemoryTransport.TEST_ROOT
2708
t = _mod_transport.get_transport_from_path(root)
2709
self.permit_url(t.base)
2710
if (t.get_bytes('.bzr/checkout/dirstate') !=
2711
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE):
2712
# The current test have modified the /bzr directory, we need to
2713
# recreate a new one or all the followng tests will fail.
2714
# If you need to inspect its content uncomment the following line
2715
# import pdb; pdb.set_trace()
2716
_rmtree_temp_dir(root + '/.bzr', test_id=self.id())
2717
self._create_safety_net()
2718
raise AssertionError('%s/.bzr should not be modified' % root)
2720
def _make_test_root(self):
2721
if TestCaseWithMemoryTransport.TEST_ROOT is None:
2722
# Watch out for tricky test dir (on OSX /tmp -> /private/tmp)
2723
root = osutils.realpath(osutils.mkdtemp(prefix='testbzr-',
2725
TestCaseWithMemoryTransport.TEST_ROOT = root
2727
self._create_safety_net()
2729
# The same directory is used by all tests, and we're not
2730
# specifically told when all tests are finished. This will do.
2731
atexit.register(_rmtree_temp_dir, root)
2733
self.permit_dir(TestCaseWithMemoryTransport.TEST_ROOT)
2734
self.addCleanup(self._check_safety_net)
2736
def makeAndChdirToTestDir(self):
2737
"""Create a temporary directories for this one test.
2739
This must set self.test_home_dir and self.test_dir and chdir to
2742
For TestCaseWithMemoryTransport we chdir to the TEST_ROOT for this
2745
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2746
self.test_dir = TestCaseWithMemoryTransport.TEST_ROOT
2747
self.test_home_dir = self.test_dir + "/MemoryTransportMissingHomeDir"
2748
self.permit_dir(self.test_dir)
2750
def make_branch(self, relpath, format=None, name=None):
2751
"""Create a branch on the transport at relpath."""
2752
repo = self.make_repository(relpath, format=format)
2753
return repo.controldir.create_branch(append_revisions_only=False, name=name)
2755
def get_default_format(self):
2758
def resolve_format(self, format):
2759
"""Resolve an object to a ControlDir format object.
2761
The initial format object can either already be
2762
a ControlDirFormat, None (for the default format),
2763
or a string with the name of the control dir format.
2765
:param format: Object to resolve
2766
:return A ControlDirFormat instance
2769
format = self.get_default_format()
2770
if isinstance(format, str):
2771
format = controldir.format_registry.make_controldir(format)
2774
def make_controldir(self, relpath, format=None):
2776
# might be a relative or absolute path
2777
maybe_a_url = self.get_url(relpath)
2778
segments = maybe_a_url.rsplit('/', 1)
2779
t = _mod_transport.get_transport(maybe_a_url)
2780
if len(segments) > 1 and segments[-1] not in ('', '.'):
2782
format = self.resolve_format(format)
2783
return format.initialize_on_transport(t)
2784
except errors.UninitializableFormat:
2785
raise TestSkipped("Format %s is not initializable." % format)
2787
def make_repository(self, relpath, shared=None, format=None):
2788
"""Create a repository on our default transport at relpath.
2790
Note that relpath must be a relative path, not a full url.
2792
# FIXME: If you create a remoterepository this returns the underlying
2793
# real format, which is incorrect. Actually we should make sure that
2794
# RemoteBzrDir returns a RemoteRepository.
2795
# maybe mbp 20070410
2796
made_control = self.make_controldir(relpath, format=format)
2797
return made_control.create_repository(shared=shared)
2799
def make_smart_server(self, path, backing_server=None):
2800
if backing_server is None:
2801
backing_server = self.get_server()
2802
smart_server = test_server.SmartTCPServer_for_testing()
2803
self.start_server(smart_server, backing_server)
2804
remote_transport = _mod_transport.get_transport_from_url(smart_server.get_url()
2806
return remote_transport
2808
def make_branch_and_memory_tree(self, relpath, format=None):
2809
"""Create a branch on the default transport and a MemoryTree for it."""
2810
b = self.make_branch(relpath, format=format)
2811
return b.create_memorytree()
2813
def make_branch_builder(self, relpath, format=None):
2814
branch = self.make_branch(relpath, format=format)
2815
return branchbuilder.BranchBuilder(branch=branch)
2817
def overrideEnvironmentForTesting(self):
2818
test_home_dir = self.test_home_dir
2819
if not PY3 and isinstance(test_home_dir, text_type):
2820
test_home_dir = test_home_dir.encode(sys.getfilesystemencoding())
2821
self.overrideEnv('HOME', test_home_dir)
2822
self.overrideEnv('BRZ_HOME', test_home_dir)
2823
self.overrideEnv('GNUPGHOME', os.path.join(test_home_dir, '.gnupg'))
2825
def setup_smart_server_with_call_log(self):
2826
"""Sets up a smart server as the transport server with a call log."""
2827
self.transport_server = test_server.SmartTCPServer_for_testing
2828
self.hpss_connections = []
2829
self.hpss_calls = []
2831
# Skip the current stack down to the caller of
2832
# setup_smart_server_with_call_log
2833
prefix_length = len(traceback.extract_stack()) - 2
2835
def capture_hpss_call(params):
2836
self.hpss_calls.append(
2837
CapturedCall(params, prefix_length))
2839
def capture_connect(transport):
2840
self.hpss_connections.append(transport)
2841
client._SmartClient.hooks.install_named_hook(
2842
'call', capture_hpss_call, None)
2843
_mod_transport.Transport.hooks.install_named_hook(
2844
'post_connect', capture_connect, None)
2846
def reset_smart_call_log(self):
2847
self.hpss_calls = []
2848
self.hpss_connections = []
2851
class TestCaseInTempDir(TestCaseWithMemoryTransport):
2852
"""Derived class that runs a test within a temporary directory.
2854
This is useful for tests that need to create a branch, etc.
2856
The directory is created in a slightly complex way: for each
2857
Python invocation, a new temporary top-level directory is created.
2858
All test cases create their own directory within that. If the
2859
tests complete successfully, the directory is removed.
2861
:ivar test_base_dir: The path of the top-level directory for this
2862
test, which contains a home directory and a work directory.
2864
:ivar test_home_dir: An initially empty directory under test_base_dir
2865
which is used as $HOME for this test.
2867
:ivar test_dir: A directory under test_base_dir used as the current
2868
directory when the test proper is run.
2871
OVERRIDE_PYTHON = 'python'
2874
super(TestCaseInTempDir, self).setUp()
2875
# Remove the protection set in isolated_environ, we have a proper
2876
# access to disk resources now.
2877
self.overrideEnv('BRZ_LOG', None)
2879
def check_file_contents(self, filename, expect):
2880
self.log("check contents of file %s" % filename)
2881
with open(filename, 'rb') as f:
2883
if contents != expect:
2884
self.log("expected: %r" % expect)
2885
self.log("actually: %r" % contents)
2886
self.fail("contents of %s not as expected" % filename)
2888
def _getTestDirPrefix(self):
2889
# create a directory within the top level test directory
2890
if sys.platform in ('win32', 'cygwin'):
2891
name_prefix = re.sub('[<>*=+",:;_/\\-]', '_', self.id())
2892
# windows is likely to have path-length limits so use a short name
2893
name_prefix = name_prefix[-30:]
2895
name_prefix = re.sub('[/]', '_', self.id())
2898
def makeAndChdirToTestDir(self):
2899
"""See TestCaseWithMemoryTransport.makeAndChdirToTestDir().
2901
For TestCaseInTempDir we create a temporary directory based on the test
2902
name and then create two subdirs - test and home under it.
2904
name_prefix = osutils.pathjoin(TestCaseWithMemoryTransport.TEST_ROOT,
2905
self._getTestDirPrefix())
2907
for i in range(100):
2908
if os.path.exists(name):
2909
name = name_prefix + '_' + str(i)
2911
# now create test and home directories within this dir
2912
self.test_base_dir = name
2913
self.addCleanup(self.deleteTestDir)
2914
os.mkdir(self.test_base_dir)
2916
self.permit_dir(self.test_base_dir)
2917
# 'sprouting' and 'init' of a branch both walk up the tree to find
2918
# stacking policy to honour; create a bzr dir with an unshared
2919
# repository (but not a branch - our code would be trying to escape
2920
# then!) to stop them, and permit it to be read.
2921
# control = controldir.ControlDir.create(self.test_base_dir)
2922
# control.create_repository()
2923
self.test_home_dir = self.test_base_dir + '/home'
2924
os.mkdir(self.test_home_dir)
2925
self.test_dir = self.test_base_dir + '/work'
2926
os.mkdir(self.test_dir)
2927
os.chdir(self.test_dir)
2928
# put name of test inside
2929
with open(self.test_base_dir + '/name', 'w') as f:
2932
def deleteTestDir(self):
2933
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2934
_rmtree_temp_dir(self.test_base_dir, test_id=self.id())
2936
def build_tree(self, shape, line_endings='binary', transport=None):
2937
"""Build a test tree according to a pattern.
2939
shape is a sequence of file specifications. If the final
2940
character is '/', a directory is created.
2942
This assumes that all the elements in the tree being built are new.
2944
This doesn't add anything to a branch.
2946
:type shape: list or tuple.
2947
:param line_endings: Either 'binary' or 'native'
2948
in binary mode, exact contents are written in native mode, the
2949
line endings match the default platform endings.
2950
:param transport: A transport to write to, for building trees on VFS's.
2951
If the transport is readonly or None, "." is opened automatically.
2954
if type(shape) not in (list, tuple):
2955
raise AssertionError("Parameter 'shape' should be "
2956
"a list or a tuple. Got %r instead" % (shape,))
2957
# It's OK to just create them using forward slashes on windows.
2958
if transport is None or transport.is_readonly():
2959
transport = _mod_transport.get_transport_from_path(".")
2961
self.assertIsInstance(name, (str, text_type))
2963
transport.mkdir(urlutils.escape(name[:-1]))
2965
if line_endings == 'binary':
2967
elif line_endings == 'native':
2968
end = os.linesep.encode('ascii')
2970
raise errors.BzrError(
2971
'Invalid line ending request %r' % line_endings)
2972
content = b"contents of %s%s" % (name.encode('utf-8'), end)
2973
transport.put_bytes_non_atomic(urlutils.escape(name), content)
2975
build_tree_contents = staticmethod(treeshape.build_tree_contents)
2977
def assertInWorkingTree(self, path, root_path='.', tree=None):
2978
"""Assert whether path or paths are in the WorkingTree"""
2980
tree = workingtree.WorkingTree.open(root_path)
2981
if not isinstance(path, (str, text_type)):
2983
self.assertInWorkingTree(p, tree=tree)
2985
self.assertTrue(tree.is_versioned(path),
2986
path + ' not in working tree.')
2988
def assertNotInWorkingTree(self, path, root_path='.', tree=None):
2989
"""Assert whether path or paths are not in the WorkingTree"""
2991
tree = workingtree.WorkingTree.open(root_path)
2992
if not isinstance(path, (str, text_type)):
2994
self.assertNotInWorkingTree(p, tree=tree)
2996
self.assertFalse(tree.is_versioned(
2997
path), path + ' in working tree.')
3000
class TestCaseWithTransport(TestCaseInTempDir):
3001
"""A test case that provides get_url and get_readonly_url facilities.
3003
These back onto two transport servers, one for readonly access and one for
3006
If no explicit class is provided for readonly access, a
3007
ReadonlyTransportDecorator is used instead which allows the use of non disk
3008
based read write transports.
3010
If an explicit class is provided for readonly access, that server and the
3011
readwrite one must both define get_url() as resolving to os.getcwd().
3015
super(TestCaseWithTransport, self).setUp()
3016
self.__vfs_server = None
3018
def get_vfs_only_server(self):
3019
"""See TestCaseWithMemoryTransport.
3021
This is useful for some tests with specific servers that need
3024
if self.__vfs_server is None:
3025
self.__vfs_server = self.vfs_transport_factory()
3026
self.start_server(self.__vfs_server)
3027
return self.__vfs_server
3029
def make_branch_and_tree(self, relpath, format=None):
3030
"""Create a branch on the transport and a tree locally.
3032
If the transport is not a LocalTransport, the Tree can't be created on
3033
the transport. In that case if the vfs_transport_factory is
3034
LocalURLServer the working tree is created in the local
3035
directory backing the transport, and the returned tree's branch and
3036
repository will also be accessed locally. Otherwise a lightweight
3037
checkout is created and returned.
3039
We do this because we can't physically create a tree in the local
3040
path, with a branch reference to the transport_factory url, and
3041
a branch + repository in the vfs_transport, unless the vfs_transport
3042
namespace is distinct from the local disk - the two branch objects
3043
would collide. While we could construct a tree with its branch object
3044
pointing at the transport_factory transport in memory, reopening it
3045
would behaving unexpectedly, and has in the past caused testing bugs
3046
when we tried to do it that way.
3048
:param format: The BzrDirFormat.
3049
:returns: the WorkingTree.
3051
# TODO: always use the local disk path for the working tree,
3052
# this obviously requires a format that supports branch references
3053
# so check for that by checking bzrdir.BzrDirFormat.get_default_format()
3055
format = self.resolve_format(format=format)
3056
if not format.supports_workingtrees:
3057
b = self.make_branch(relpath + '.branch', format=format)
3058
return b.create_checkout(relpath, lightweight=True)
3059
b = self.make_branch(relpath, format=format)
3061
return b.controldir.create_workingtree()
3062
except errors.NotLocalUrl:
3063
# We can only make working trees locally at the moment. If the
3064
# transport can't support them, then we keep the non-disk-backed
3065
# branch and create a local checkout.
3066
if self.vfs_transport_factory is test_server.LocalURLServer:
3067
# the branch is colocated on disk, we cannot create a checkout.
3068
# hopefully callers will expect this.
3069
local_controldir = controldir.ControlDir.open(
3070
self.get_vfs_only_url(relpath))
3071
wt = local_controldir.create_workingtree()
3072
if wt.branch._format != b._format:
3074
# Make sure that assigning to wt._branch fixes wt.branch,
3075
# in case the implementation details of workingtree objects
3077
self.assertIs(b, wt.branch)
3080
return b.create_checkout(relpath, lightweight=True)
3082
def assertIsDirectory(self, relpath, transport):
3083
"""Assert that relpath within transport is a directory.
3085
This may not be possible on all transports; in that case it propagates
3086
a TransportNotPossible.
3089
mode = transport.stat(relpath).st_mode
3090
except errors.NoSuchFile:
3091
self.fail("path %s is not a directory; no such file"
3093
if not stat.S_ISDIR(mode):
3094
self.fail("path %s is not a directory; has mode %#o"
3097
def assertTreesEqual(self, left, right):
3098
"""Check that left and right have the same content and properties."""
3099
# we use a tree delta to check for equality of the content, and we
3100
# manually check for equality of other things such as the parents list.
3101
self.assertEqual(left.get_parent_ids(), right.get_parent_ids())
3102
differences = left.changes_from(right)
3103
self.assertFalse(differences.has_changed(),
3104
"Trees %r and %r are different: %r" % (left, right, differences))
3106
def disable_missing_extensions_warning(self):
3107
"""Some tests expect a precise stderr content.
3109
There is no point in forcing them to duplicate the extension related
3112
config.GlobalConfig().set_user_option(
3113
'suppress_warnings', 'missing_extensions')
3116
class ChrootedTestCase(TestCaseWithTransport):
3117
"""A support class that provides readonly urls outside the local namespace.
3119
This is done by checking if self.transport_server is a MemoryServer. if it
3120
is then we are chrooted already, if it is not then an HttpServer is used
3123
TODO RBC 20060127: make this an option to TestCaseWithTransport so it can
3124
be used without needed to redo it when a different
3125
subclass is in use ?
3129
from breezy.tests import http_server
3130
super(ChrootedTestCase, self).setUp()
3131
if not self.vfs_transport_factory == memory.MemoryServer:
3132
self.transport_readonly_server = http_server.HttpServer
3135
def condition_id_re(pattern):
3136
"""Create a condition filter which performs a re check on a test's id.
3138
:param pattern: A regular expression string.
3139
:return: A callable that returns True if the re matches.
3141
filter_re = re.compile(pattern, 0)
3143
def condition(test):
3145
return filter_re.search(test_id)
3149
def condition_isinstance(klass_or_klass_list):
3150
"""Create a condition filter which returns isinstance(param, klass).
3152
:return: A callable which when called with one parameter obj return the
3153
result of isinstance(obj, klass_or_klass_list).
3156
return isinstance(obj, klass_or_klass_list)
3160
def condition_id_in_list(id_list):
3161
"""Create a condition filter which verify that test's id in a list.
3163
:param id_list: A TestIdList object.
3164
:return: A callable that returns True if the test's id appears in the list.
3166
def condition(test):
3167
return id_list.includes(test.id())
3171
def condition_id_startswith(starts):
3172
"""Create a condition filter verifying that test's id starts with a string.
3174
:param starts: A list of string.
3175
:return: A callable that returns True if the test's id starts with one of
3178
def condition(test):
3179
for start in starts:
3180
if test.id().startswith(start):
3186
def exclude_tests_by_condition(suite, condition):
3187
"""Create a test suite which excludes some tests from suite.
3189
:param suite: The suite to get tests from.
3190
:param condition: A callable whose result evaluates True when called with a
3191
test case which should be excluded from the result.
3192
:return: A suite which contains the tests found in suite that fail
3196
for test in iter_suite_tests(suite):
3197
if not condition(test):
3199
return TestUtil.TestSuite(result)
3202
def filter_suite_by_condition(suite, condition):
3203
"""Create a test suite by filtering another one.
3205
:param suite: The source suite.
3206
:param condition: A callable whose result evaluates True when called with a
3207
test case which should be included in the result.
3208
:return: A suite which contains the tests found in suite that pass
3212
for test in iter_suite_tests(suite):
3215
return TestUtil.TestSuite(result)
3218
def filter_suite_by_re(suite, pattern):
3219
"""Create a test suite by filtering another one.
3221
:param suite: the source suite
3222
:param pattern: pattern that names must match
3223
:returns: the newly created suite
3225
condition = condition_id_re(pattern)
3226
result_suite = filter_suite_by_condition(suite, condition)
3230
def filter_suite_by_id_list(suite, test_id_list):
3231
"""Create a test suite by filtering another one.
3233
:param suite: The source suite.
3234
:param test_id_list: A list of the test ids to keep as strings.
3235
:returns: the newly created suite
3237
condition = condition_id_in_list(test_id_list)
3238
result_suite = filter_suite_by_condition(suite, condition)
3242
def filter_suite_by_id_startswith(suite, start):
3243
"""Create a test suite by filtering another one.
3245
:param suite: The source suite.
3246
:param start: A list of string the test id must start with one of.
3247
:returns: the newly created suite
3249
condition = condition_id_startswith(start)
3250
result_suite = filter_suite_by_condition(suite, condition)
3254
def exclude_tests_by_re(suite, pattern):
3255
"""Create a test suite which excludes some tests from suite.
3257
:param suite: The suite to get tests from.
3258
:param pattern: A regular expression string. Test ids that match this
3259
pattern will be excluded from the result.
3260
:return: A TestSuite that contains all the tests from suite without the
3261
tests that matched pattern. The order of tests is the same as it was in
3264
return exclude_tests_by_condition(suite, condition_id_re(pattern))
3267
def preserve_input(something):
3268
"""A helper for performing test suite transformation chains.
3270
:param something: Anything you want to preserve.
3276
def randomize_suite(suite):
3277
"""Return a new TestSuite with suite's tests in random order.
3279
The tests in the input suite are flattened into a single suite in order to
3280
accomplish this. Any nested TestSuites are removed to provide global
3283
tests = list(iter_suite_tests(suite))
3284
random.shuffle(tests)
3285
return TestUtil.TestSuite(tests)
3288
def split_suite_by_condition(suite, condition):
3289
"""Split a test suite into two by a condition.
3291
:param suite: The suite to split.
3292
:param condition: The condition to match on. Tests that match this
3293
condition are returned in the first test suite, ones that do not match
3294
are in the second suite.
3295
:return: A tuple of two test suites, where the first contains tests from
3296
suite matching the condition, and the second contains the remainder
3297
from suite. The order within each output suite is the same as it was in
3302
for test in iter_suite_tests(suite):
3304
matched.append(test)
3306
did_not_match.append(test)
3307
return TestUtil.TestSuite(matched), TestUtil.TestSuite(did_not_match)
3310
def split_suite_by_re(suite, pattern):
3311
"""Split a test suite into two by a regular expression.
3313
:param suite: The suite to split.
3314
:param pattern: A regular expression string. Test ids that match this
3315
pattern will be in the first test suite returned, and the others in the
3316
second test suite returned.
3317
:return: A tuple of two test suites, where the first contains tests from
3318
suite matching pattern, and the second contains the remainder from
3319
suite. The order within each output suite is the same as it was in
3322
return split_suite_by_condition(suite, condition_id_re(pattern))
3325
def run_suite(suite, name='test', verbose=False, pattern=".*",
3326
stop_on_failure=False,
3327
transport=None, lsprof_timed=None, bench_history=None,
3328
matching_tests_first=None,
3331
exclude_pattern=None,
3334
suite_decorators=None,
3336
result_decorators=None,
3338
"""Run a test suite for brz selftest.
3340
:param runner_class: The class of runner to use. Must support the
3341
constructor arguments passed by run_suite which are more than standard
3343
:return: A boolean indicating success.
3345
TestCase._gather_lsprof_in_benchmarks = lsprof_timed
3350
if runner_class is None:
3351
runner_class = TextTestRunner
3354
runner = runner_class(stream=stream,
3356
verbosity=verbosity,
3357
bench_history=bench_history,
3359
result_decorators=result_decorators,
3361
runner.stop_on_failure = stop_on_failure
3362
if isinstance(suite, unittest.TestSuite):
3363
# Empty out _tests list of passed suite and populate new TestSuite
3364
suite._tests[:], suite = [], TestSuite(suite)
3365
# built in decorator factories:
3367
random_order(random_seed, runner),
3368
exclude_tests(exclude_pattern),
3370
if matching_tests_first:
3371
decorators.append(tests_first(pattern))
3373
decorators.append(filter_tests(pattern))
3374
if suite_decorators:
3375
decorators.extend(suite_decorators)
3376
# tell the result object how many tests will be running: (except if
3377
# --parallel=fork is being used. Robert said he will provide a better
3378
# progress design later -- vila 20090817)
3379
if fork_decorator not in decorators:
3380
decorators.append(CountingDecorator)
3381
for decorator in decorators:
3382
suite = decorator(suite)
3384
# Done after test suite decoration to allow randomisation etc
3385
# to take effect, though that is of marginal benefit.
3387
stream.write("Listing tests only ...\n")
3388
if getattr(runner, 'list', None) is not None:
3391
for t in iter_suite_tests(suite):
3392
stream.write("%s\n" % (t.id()))
3394
result = runner.run(suite)
3395
if strict and getattr(result, 'wasStrictlySuccessful', False):
3396
return result.wasStrictlySuccessful()
3398
return result.wasSuccessful()
3401
# A registry where get() returns a suite decorator.
3402
parallel_registry = registry.Registry()
3405
def fork_decorator(suite):
3406
if getattr(os, "fork", None) is None:
3407
raise errors.BzrCommandError("platform does not support fork,"
3408
" try --parallel=subprocess instead.")
3409
concurrency = osutils.local_concurrency()
3410
if concurrency == 1:
3412
from testtools import ConcurrentTestSuite
3413
return ConcurrentTestSuite(suite, fork_for_tests)
3416
parallel_registry.register('fork', fork_decorator)
3419
def subprocess_decorator(suite):
3420
concurrency = osutils.local_concurrency()
3421
if concurrency == 1:
3423
from testtools import ConcurrentTestSuite
3424
return ConcurrentTestSuite(suite, reinvoke_for_tests)
3427
parallel_registry.register('subprocess', subprocess_decorator)
3430
def exclude_tests(exclude_pattern):
3431
"""Return a test suite decorator that excludes tests."""
3432
if exclude_pattern is None:
3433
return identity_decorator
3435
def decorator(suite):
3436
return ExcludeDecorator(suite, exclude_pattern)
3440
def filter_tests(pattern):
3442
return identity_decorator
3444
def decorator(suite):
3445
return FilterTestsDecorator(suite, pattern)
3449
def random_order(random_seed, runner):
3450
"""Return a test suite decorator factory for randomising tests order.
3452
:param random_seed: now, a string which casts to an integer, or an integer.
3453
:param runner: A test runner with a stream attribute to report on.
3455
if random_seed is None:
3456
return identity_decorator
3458
def decorator(suite):
3459
return RandomDecorator(suite, random_seed, runner.stream)
3463
def tests_first(pattern):
3465
return identity_decorator
3467
def decorator(suite):
3468
return TestFirstDecorator(suite, pattern)
3472
def identity_decorator(suite):
3477
class TestDecorator(TestUtil.TestSuite):
3478
"""A decorator for TestCase/TestSuite objects.
3480
Contains rather than flattening suite passed on construction
3483
def __init__(self, suite=None):
3484
super(TestDecorator, self).__init__()
3485
if suite is not None:
3488
# Don't need subclass run method with suite emptying
3489
run = unittest.TestSuite.run
3492
class CountingDecorator(TestDecorator):
3493
"""A decorator which calls result.progress(self.countTestCases)."""
3495
def run(self, result):
3496
progress_method = getattr(result, 'progress', None)
3497
if callable(progress_method):
3498
progress_method(self.countTestCases(), SUBUNIT_SEEK_SET)
3499
return super(CountingDecorator, self).run(result)
3502
class ExcludeDecorator(TestDecorator):
3503
"""A decorator which excludes test matching an exclude pattern."""
3505
def __init__(self, suite, exclude_pattern):
3506
super(ExcludeDecorator, self).__init__(
3507
exclude_tests_by_re(suite, exclude_pattern))
3510
class FilterTestsDecorator(TestDecorator):
3511
"""A decorator which filters tests to those matching a pattern."""
3513
def __init__(self, suite, pattern):
3514
super(FilterTestsDecorator, self).__init__(
3515
filter_suite_by_re(suite, pattern))
3518
class RandomDecorator(TestDecorator):
3519
"""A decorator which randomises the order of its tests."""
3521
def __init__(self, suite, random_seed, stream):
3522
random_seed = self.actual_seed(random_seed)
3523
stream.write("Randomizing test order using seed %s\n\n" %
3525
# Initialise the random number generator.
3526
random.seed(random_seed)
3527
super(RandomDecorator, self).__init__(randomize_suite(suite))
3530
def actual_seed(seed):
3532
# We convert the seed to an integer to make it reuseable across
3533
# invocations (because the user can reenter it).
3534
return int(time.time())
3536
# Convert the seed to an integer if we can
3539
except (TypeError, ValueError):
3544
class TestFirstDecorator(TestDecorator):
3545
"""A decorator which moves named tests to the front."""
3547
def __init__(self, suite, pattern):
3548
super(TestFirstDecorator, self).__init__()
3549
self.addTests(split_suite_by_re(suite, pattern))
3552
def partition_tests(suite, count):
3553
"""Partition suite into count lists of tests."""
3554
# This just assigns tests in a round-robin fashion. On one hand this
3555
# splits up blocks of related tests that might run faster if they shared
3556
# resources, but on the other it avoids assigning blocks of slow tests to
3557
# just one partition. So the slowest partition shouldn't be much slower
3559
partitions = [list() for i in range(count)]
3560
tests = iter_suite_tests(suite)
3561
for partition, test in zip(itertools.cycle(partitions), tests):
3562
partition.append(test)
3566
def workaround_zealous_crypto_random():
3567
"""Crypto.Random want to help us being secure, but we don't care here.
3569
This workaround some test failure related to the sftp server. Once paramiko
3570
stop using the controversial API in Crypto.Random, we may get rid of it.
3573
from Crypto.Random import atfork
3579
def fork_for_tests(suite):
3580
"""Take suite and start up one runner per CPU by forking()
3582
:return: An iterable of TestCase-like objects which can each have
3583
run(result) called on them to feed tests to result.
3585
concurrency = osutils.local_concurrency()
3587
from subunit import ProtocolTestCase
3588
from subunit.test_results import AutoTimingTestResultDecorator
3590
class TestInOtherProcess(ProtocolTestCase):
3591
# Should be in subunit, I think. RBC.
3592
def __init__(self, stream, pid):
3593
ProtocolTestCase.__init__(self, stream)
3596
def run(self, result):
3598
ProtocolTestCase.run(self, result)
3600
pid, status = os.waitpid(self.pid, 0)
3601
# GZ 2011-10-18: If status is nonzero, should report to the result
3602
# that something went wrong.
3604
test_blocks = partition_tests(suite, concurrency)
3605
# Clear the tests from the original suite so it doesn't keep them alive
3606
suite._tests[:] = []
3607
for process_tests in test_blocks:
3608
process_suite = TestUtil.TestSuite(process_tests)
3609
# Also clear each split list so new suite has only reference
3610
process_tests[:] = []
3611
c2pread, c2pwrite = os.pipe()
3615
stream = os.fdopen(c2pwrite, 'wb', 1)
3616
workaround_zealous_crypto_random()
3622
coverage.process_startup()
3624
# Leave stderr and stdout open so we can see test noise
3625
# Close stdin so that the child goes away if it decides to
3626
# read from stdin (otherwise its a roulette to see what
3627
# child actually gets keystrokes for pdb etc).
3629
subunit_result = AutoTimingTestResultDecorator(
3630
SubUnitBzrProtocolClientv1(stream))
3631
process_suite.run(subunit_result)
3633
# Try and report traceback on stream, but exit with error even
3634
# if stream couldn't be created or something else goes wrong.
3635
# The traceback is formatted to a string and written in one go
3636
# to avoid interleaving lines from multiple failing children.
3637
tb = traceback.format_exc()
3638
if isinstance(tb, text_type):
3639
tb = tb.encode('utf-8')
3648
stream = os.fdopen(c2pread, 'rb', 1)
3649
test = TestInOtherProcess(stream, pid)
3654
def reinvoke_for_tests(suite):
3655
"""Take suite and start up one runner per CPU using subprocess().
3657
:return: An iterable of TestCase-like objects which can each have
3658
run(result) called on them to feed tests to result.
3660
concurrency = osutils.local_concurrency()
3662
from subunit import ProtocolTestCase
3664
class TestInSubprocess(ProtocolTestCase):
3665
def __init__(self, process, name):
3666
ProtocolTestCase.__init__(self, process.stdout)
3667
self.process = process
3668
self.process.stdin.close()
3671
def run(self, result):
3673
ProtocolTestCase.run(self, result)
3676
os.unlink(self.name)
3677
# print "pid %d finished" % finished_process
3678
test_blocks = partition_tests(suite, concurrency)
3679
for process_tests in test_blocks:
3680
# ugly; currently reimplement rather than reuses TestCase methods.
3681
bzr_path = os.path.dirname(os.path.dirname(breezy.__file__)) + '/bzr'
3682
if not os.path.isfile(bzr_path):
3683
# We are probably installed. Assume sys.argv is the right file
3684
bzr_path = sys.argv[0]
3685
bzr_path = [bzr_path]
3686
if sys.platform == "win32":
3687
# if we're on windows, we can't execute the bzr script directly
3688
bzr_path = [sys.executable] + bzr_path
3689
fd, test_list_file_name = tempfile.mkstemp()
3690
test_list_file = os.fdopen(fd, 'wb', 1)
3691
for test in process_tests:
3692
test_list_file.write(test.id() + '\n')
3693
test_list_file.close()
3695
argv = bzr_path + ['selftest', '--load-list', test_list_file_name,
3697
if '--no-plugins' in sys.argv:
3698
argv.append('--no-plugins')
3699
# stderr=subprocess.STDOUT would be ideal, but until we prevent
3700
# noise on stderr it can interrupt the subunit protocol.
3701
process = subprocess.Popen(argv, stdin=subprocess.PIPE,
3702
stdout=subprocess.PIPE,
3703
stderr=subprocess.PIPE,
3705
test = TestInSubprocess(process, test_list_file_name)
3708
os.unlink(test_list_file_name)
3713
class ProfileResult(testtools.ExtendedToOriginalDecorator):
3714
"""Generate profiling data for all activity between start and success.
3716
The profile data is appended to the test's _benchcalls attribute and can
3717
be accessed by the forwarded-to TestResult.
3719
While it might be cleaner do accumulate this in stopTest, addSuccess is
3720
where our existing output support for lsprof is, and this class aims to
3721
fit in with that: while it could be moved it's not necessary to accomplish
3722
test profiling, nor would it be dramatically cleaner.
3725
def startTest(self, test):
3726
self.profiler = breezy.lsprof.BzrProfiler()
3727
# Prevent deadlocks in tests that use lsprof: those tests will
3729
breezy.lsprof.BzrProfiler.profiler_block = 0
3730
self.profiler.start()
3731
testtools.ExtendedToOriginalDecorator.startTest(self, test)
3733
def addSuccess(self, test):
3734
stats = self.profiler.stop()
3736
calls = test._benchcalls
3737
except AttributeError:
3738
test._benchcalls = []
3739
calls = test._benchcalls
3740
calls.append(((test.id(), "", ""), stats))
3741
testtools.ExtendedToOriginalDecorator.addSuccess(self, test)
3743
def stopTest(self, test):
3744
testtools.ExtendedToOriginalDecorator.stopTest(self, test)
3745
self.profiler = None
3748
# Controlled by "brz selftest -E=..." option
3749
# Currently supported:
3750
# -Eallow_debug Will no longer clear debug.debug_flags() so it
3751
# preserves any flags supplied at the command line.
3752
# -Edisable_lock_checks Turns errors in mismatched locks into simple prints
3753
# rather than failing tests. And no longer raise
3754
# LockContention when fctnl locks are not being used
3755
# with proper exclusion rules.
3756
# -Ethreads Will display thread ident at creation/join time to
3757
# help track thread leaks
3758
# -Euncollected_cases Display the identity of any test cases that weren't
3759
# deallocated after being completed.
3760
# -Econfig_stats Will collect statistics using addDetail
3761
selftest_debug_flags = set()
3764
def selftest(verbose=False, pattern=".*", stop_on_failure=True,
3766
test_suite_factory=None,
3769
matching_tests_first=None,
3772
exclude_pattern=None,
3778
suite_decorators=None,
3782
"""Run the whole test suite under the enhanced runner"""
3783
# XXX: Very ugly way to do this...
3784
# Disable warning about old formats because we don't want it to disturb
3785
# any blackbox tests.
3786
from breezy import repository
3787
repository._deprecation_warning_done = True
3789
global default_transport
3790
if transport is None:
3791
transport = default_transport
3792
old_transport = default_transport
3793
default_transport = transport
3794
global selftest_debug_flags
3795
old_debug_flags = selftest_debug_flags
3796
if debug_flags is not None:
3797
selftest_debug_flags = set(debug_flags)
3799
if load_list is None:
3802
keep_only = load_test_id_list(load_list)
3804
starting_with = [test_prefix_alias_registry.resolve_alias(start)
3805
for start in starting_with]
3806
# Always consider 'unittest' an interesting name so that failed
3807
# suites wrapped as test cases appear in the output.
3808
starting_with.append('unittest')
3809
if test_suite_factory is None:
3810
# Reduce loading time by loading modules based on the starting_with
3812
suite = test_suite(keep_only, starting_with)
3814
suite = test_suite_factory()
3816
# But always filter as requested.
3817
suite = filter_suite_by_id_startswith(suite, starting_with)
3818
result_decorators = []
3820
result_decorators.append(ProfileResult)
3821
return run_suite(suite, 'testbzr', verbose=verbose, pattern=pattern,
3822
stop_on_failure=stop_on_failure,
3823
transport=transport,
3824
lsprof_timed=lsprof_timed,
3825
bench_history=bench_history,
3826
matching_tests_first=matching_tests_first,
3827
list_only=list_only,
3828
random_seed=random_seed,
3829
exclude_pattern=exclude_pattern,
3831
runner_class=runner_class,
3832
suite_decorators=suite_decorators,
3834
result_decorators=result_decorators,
3837
default_transport = old_transport
3838
selftest_debug_flags = old_debug_flags
3841
def load_test_id_list(file_name):
3842
"""Load a test id list from a text file.
3844
The format is one test id by line. No special care is taken to impose
3845
strict rules, these test ids are used to filter the test suite so a test id
3846
that do not match an existing test will do no harm. This allows user to add
3847
comments, leave blank lines, etc.
3851
ftest = open(file_name, 'rt')
3852
except IOError as e:
3853
if e.errno != errno.ENOENT:
3856
raise errors.NoSuchFile(file_name)
3858
for test_name in ftest.readlines():
3859
test_list.append(test_name.strip())
3864
def suite_matches_id_list(test_suite, id_list):
3865
"""Warns about tests not appearing or appearing more than once.
3867
:param test_suite: A TestSuite object.
3868
:param test_id_list: The list of test ids that should be found in
3871
:return: (absents, duplicates) absents is a list containing the test found
3872
in id_list but not in test_suite, duplicates is a list containing the
3873
tests found multiple times in test_suite.
3875
When using a prefined test id list, it may occurs that some tests do not
3876
exist anymore or that some tests use the same id. This function warns the
3877
tester about potential problems in his workflow (test lists are volatile)
3878
or in the test suite itself (using the same id for several tests does not
3879
help to localize defects).
3881
# Build a dict counting id occurrences
3883
for test in iter_suite_tests(test_suite):
3885
tests[id] = tests.get(id, 0) + 1
3890
occurs = tests.get(id, 0)
3892
not_found.append(id)
3894
duplicates.append(id)
3896
return not_found, duplicates
3899
class TestIdList(object):
3900
"""Test id list to filter a test suite.
3902
Relying on the assumption that test ids are built as:
3903
<module>[.<class>.<method>][(<param>+)], <module> being in python dotted
3904
notation, this class offers methods to :
3905
- avoid building a test suite for modules not refered to in the test list,
3906
- keep only the tests listed from the module test suite.
3909
def __init__(self, test_id_list):
3910
# When a test suite needs to be filtered against us we compare test ids
3911
# for equality, so a simple dict offers a quick and simple solution.
3912
self.tests = dict().fromkeys(test_id_list, True)
3914
# While unittest.TestCase have ids like:
3915
# <module>.<class>.<method>[(<param+)],
3916
# doctest.DocTestCase can have ids like:
3919
# <module>.<function>
3920
# <module>.<class>.<method>
3922
# Since we can't predict a test class from its name only, we settle on
3923
# a simple constraint: a test id always begins with its module name.
3926
for test_id in test_id_list:
3927
parts = test_id.split('.')
3928
mod_name = parts.pop(0)
3929
modules[mod_name] = True
3931
mod_name += '.' + part
3932
modules[mod_name] = True
3933
self.modules = modules
3935
def refers_to(self, module_name):
3936
"""Is there tests for the module or one of its sub modules."""
3937
return module_name in self.modules
3939
def includes(self, test_id):
3940
return test_id in self.tests
3943
class TestPrefixAliasRegistry(registry.Registry):
3944
"""A registry for test prefix aliases.
3946
This helps implement shorcuts for the --starting-with selftest
3947
option. Overriding existing prefixes is not allowed but not fatal (a
3948
warning will be emitted).
3951
def register(self, key, obj, help=None, info=None,
3952
override_existing=False):
3953
"""See Registry.register.
3955
Trying to override an existing alias causes a warning to be emitted,
3956
not a fatal execption.
3959
super(TestPrefixAliasRegistry, self).register(
3960
key, obj, help=help, info=info, override_existing=False)
3962
actual = self.get(key)
3964
'Test prefix alias %s is already used for %s, ignoring %s'
3965
% (key, actual, obj))
3967
def resolve_alias(self, id_start):
3968
"""Replace the alias by the prefix in the given string.
3970
Using an unknown prefix is an error to help catching typos.
3972
parts = id_start.split('.')
3974
parts[0] = self.get(parts[0])
3976
raise errors.BzrCommandError(
3977
'%s is not a known test prefix alias' % parts[0])
3978
return '.'.join(parts)
3981
test_prefix_alias_registry = TestPrefixAliasRegistry()
3982
"""Registry of test prefix aliases."""
3985
# This alias allows to detect typos ('bzrlin.') by making all valid test ids
3986
# appear prefixed ('breezy.' is "replaced" by 'breezy.').
3987
test_prefix_alias_registry.register('breezy', 'breezy')
3989
# Obvious highest levels prefixes, feel free to add your own via a plugin
3990
test_prefix_alias_registry.register('bd', 'breezy.doc')
3991
test_prefix_alias_registry.register('bu', 'breezy.utils')
3992
test_prefix_alias_registry.register('bt', 'breezy.tests')
3993
test_prefix_alias_registry.register('bb', 'breezy.tests.blackbox')
3994
test_prefix_alias_registry.register('bp', 'breezy.plugins')
3997
def _test_suite_testmod_names():
3998
"""Return the standard list of test module names to test."""
4000
'breezy.git.tests.test_blackbox',
4001
'breezy.git.tests.test_builder',
4002
'breezy.git.tests.test_branch',
4003
'breezy.git.tests.test_cache',
4004
'breezy.git.tests.test_dir',
4005
'breezy.git.tests.test_fetch',
4006
'breezy.git.tests.test_git_remote_helper',
4007
'breezy.git.tests.test_mapping',
4008
'breezy.git.tests.test_memorytree',
4009
'breezy.git.tests.test_object_store',
4010
'breezy.git.tests.test_pristine_tar',
4011
'breezy.git.tests.test_push',
4012
'breezy.git.tests.test_remote',
4013
'breezy.git.tests.test_repository',
4014
'breezy.git.tests.test_refs',
4015
'breezy.git.tests.test_revspec',
4016
'breezy.git.tests.test_roundtrip',
4017
'breezy.git.tests.test_server',
4018
'breezy.git.tests.test_transportgit',
4019
'breezy.git.tests.test_unpeel_map',
4020
'breezy.git.tests.test_urls',
4021
'breezy.git.tests.test_workingtree',
4022
'breezy.tests.blackbox',
4023
'breezy.tests.commands',
4024
'breezy.tests.per_branch',
4025
'breezy.tests.per_bzrdir',
4026
'breezy.tests.per_controldir',
4027
'breezy.tests.per_controldir_colo',
4028
'breezy.tests.per_foreign_vcs',
4029
'breezy.tests.per_interrepository',
4030
'breezy.tests.per_intertree',
4031
'breezy.tests.per_inventory',
4032
'breezy.tests.per_interbranch',
4033
'breezy.tests.per_lock',
4034
'breezy.tests.per_merger',
4035
'breezy.tests.per_transport',
4036
'breezy.tests.per_tree',
4037
'breezy.tests.per_pack_repository',
4038
'breezy.tests.per_repository',
4039
'breezy.tests.per_repository_chk',
4040
'breezy.tests.per_repository_reference',
4041
'breezy.tests.per_repository_vf',
4042
'breezy.tests.per_uifactory',
4043
'breezy.tests.per_versionedfile',
4044
'breezy.tests.per_workingtree',
4045
'breezy.tests.test__annotator',
4046
'breezy.tests.test__bencode',
4047
'breezy.tests.test__btree_serializer',
4048
'breezy.tests.test__chk_map',
4049
'breezy.tests.test__dirstate_helpers',
4050
'breezy.tests.test__groupcompress',
4051
'breezy.tests.test__known_graph',
4052
'breezy.tests.test__rio',
4053
'breezy.tests.test__simple_set',
4054
'breezy.tests.test__static_tuple',
4055
'breezy.tests.test__walkdirs_win32',
4056
'breezy.tests.test_ancestry',
4057
'breezy.tests.test_annotate',
4058
'breezy.tests.test_atomicfile',
4059
'breezy.tests.test_bad_files',
4060
'breezy.tests.test_bisect',
4061
'breezy.tests.test_bisect_multi',
4062
'breezy.tests.test_branch',
4063
'breezy.tests.test_branchbuilder',
4064
'breezy.tests.test_btree_index',
4065
'breezy.tests.test_bugtracker',
4066
'breezy.tests.test_bundle',
4067
'breezy.tests.test_bzrdir',
4068
'breezy.tests.test__chunks_to_lines',
4069
'breezy.tests.test_cache_utf8',
4070
'breezy.tests.test_chk_map',
4071
'breezy.tests.test_chk_serializer',
4072
'breezy.tests.test_chunk_writer',
4073
'breezy.tests.test_clean_tree',
4074
'breezy.tests.test_cleanup',
4075
'breezy.tests.test_cmdline',
4076
'breezy.tests.test_commands',
4077
'breezy.tests.test_commit',
4078
'breezy.tests.test_commit_merge',
4079
'breezy.tests.test_config',
4080
'breezy.tests.test_conflicts',
4081
'breezy.tests.test_controldir',
4082
'breezy.tests.test_counted_lock',
4083
'breezy.tests.test_crash',
4084
'breezy.tests.test_decorators',
4085
'breezy.tests.test_delta',
4086
'breezy.tests.test_debug',
4087
'breezy.tests.test_diff',
4088
'breezy.tests.test_directory_service',
4089
'breezy.tests.test_dirstate',
4090
'breezy.tests.test_email_message',
4091
'breezy.tests.test_eol_filters',
4092
'breezy.tests.test_errors',
4093
'breezy.tests.test_estimate_compressed_size',
4094
'breezy.tests.test_export',
4095
'breezy.tests.test_export_pot',
4096
'breezy.tests.test_extract',
4097
'breezy.tests.test_features',
4098
'breezy.tests.test_fetch',
4099
'breezy.tests.test_fetch_ghosts',
4100
'breezy.tests.test_fixtures',
4101
'breezy.tests.test_fifo_cache',
4102
'breezy.tests.test_filters',
4103
'breezy.tests.test_filter_tree',
4104
'breezy.tests.test_foreign',
4105
'breezy.tests.test_generate_docs',
4106
'breezy.tests.test_generate_ids',
4107
'breezy.tests.test_globbing',
4108
'breezy.tests.test_gpg',
4109
'breezy.tests.test_graph',
4110
'breezy.tests.test_grep',
4111
'breezy.tests.test_groupcompress',
4112
'breezy.tests.test_hashcache',
4113
'breezy.tests.test_help',
4114
'breezy.tests.test_hooks',
4115
'breezy.tests.test_http',
4116
'breezy.tests.test_http_response',
4117
'breezy.tests.test_https_ca_bundle',
4118
'breezy.tests.test_https_urllib',
4119
'breezy.tests.test_i18n',
4120
'breezy.tests.test_identitymap',
4121
'breezy.tests.test_ignores',
4122
'breezy.tests.test_index',
4123
'breezy.tests.test_import_tariff',
4124
'breezy.tests.test_info',
4125
'breezy.tests.test_inv',
4126
'breezy.tests.test_inventory_delta',
4127
'breezy.tests.test_knit',
4128
'breezy.tests.test_lazy_import',
4129
'breezy.tests.test_lazy_regex',
4130
'breezy.tests.test_library_state',
4131
'breezy.tests.test_location',
4132
'breezy.tests.test_lock',
4133
'breezy.tests.test_lockable_files',
4134
'breezy.tests.test_lockdir',
4135
'breezy.tests.test_log',
4136
'breezy.tests.test_lru_cache',
4137
'breezy.tests.test_lsprof',
4138
'breezy.tests.test_mail_client',
4139
'breezy.tests.test_matchers',
4140
'breezy.tests.test_memorytree',
4141
'breezy.tests.test_merge',
4142
'breezy.tests.test_merge3',
4143
'breezy.tests.test_merge_core',
4144
'breezy.tests.test_merge_directive',
4145
'breezy.tests.test_mergetools',
4146
'breezy.tests.test_missing',
4147
'breezy.tests.test_msgeditor',
4148
'breezy.tests.test_multiparent',
4149
'breezy.tests.test_mutabletree',
4150
'breezy.tests.test_nonascii',
4151
'breezy.tests.test_options',
4152
'breezy.tests.test_osutils',
4153
'breezy.tests.test_osutils_encodings',
4154
'breezy.tests.test_pack',
4155
'breezy.tests.test_patch',
4156
'breezy.tests.test_patches',
4157
'breezy.tests.test_permissions',
4158
'breezy.tests.test_plugins',
4159
'breezy.tests.test_progress',
4160
'breezy.tests.test_pyutils',
4161
'breezy.tests.test_read_bundle',
4162
'breezy.tests.test_reconcile',
4163
'breezy.tests.test_reconfigure',
4164
'breezy.tests.test_registry',
4165
'breezy.tests.test_remote',
4166
'breezy.tests.test_rename_map',
4167
'breezy.tests.test_repository',
4168
'breezy.tests.test_revert',
4169
'breezy.tests.test_revision',
4170
'breezy.tests.test_revisionspec',
4171
'breezy.tests.test_revisiontree',
4172
'breezy.tests.test_rio',
4173
'breezy.tests.test_rules',
4174
'breezy.tests.test_url_policy_open',
4175
'breezy.tests.test_sampler',
4176
'breezy.tests.test_scenarios',
4177
'breezy.tests.test_script',
4178
'breezy.tests.test_selftest',
4179
'breezy.tests.test_serializer',
4180
'breezy.tests.test_setup',
4181
'breezy.tests.test_sftp_transport',
4182
'breezy.tests.test_shelf',
4183
'breezy.tests.test_shelf_ui',
4184
'breezy.tests.test_smart',
4185
'breezy.tests.test_smart_add',
4186
'breezy.tests.test_smart_request',
4187
'breezy.tests.test_smart_signals',
4188
'breezy.tests.test_smart_transport',
4189
'breezy.tests.test_smtp_connection',
4190
'breezy.tests.test_source',
4191
'breezy.tests.test_ssh_transport',
4192
'breezy.tests.test_status',
4193
'breezy.tests.test_strace',
4194
'breezy.tests.test_subsume',
4195
'breezy.tests.test_switch',
4196
'breezy.tests.test_symbol_versioning',
4197
'breezy.tests.test_tag',
4198
'breezy.tests.test_test_server',
4199
'breezy.tests.test_testament',
4200
'breezy.tests.test_textfile',
4201
'breezy.tests.test_textmerge',
4202
'breezy.tests.test_cethread',
4203
'breezy.tests.test_timestamp',
4204
'breezy.tests.test_trace',
4205
'breezy.tests.test_transactions',
4206
'breezy.tests.test_transform',
4207
'breezy.tests.test_transport',
4208
'breezy.tests.test_transport_log',
4209
'breezy.tests.test_tree',
4210
'breezy.tests.test_treebuilder',
4211
'breezy.tests.test_treeshape',
4212
'breezy.tests.test_tsort',
4213
'breezy.tests.test_tuned_gzip',
4214
'breezy.tests.test_ui',
4215
'breezy.tests.test_uncommit',
4216
'breezy.tests.test_upgrade',
4217
'breezy.tests.test_upgrade_stacked',
4218
'breezy.tests.test_upstream_import',
4219
'breezy.tests.test_urlutils',
4220
'breezy.tests.test_utextwrap',
4221
'breezy.tests.test_version',
4222
'breezy.tests.test_version_info',
4223
'breezy.tests.test_versionedfile',
4224
'breezy.tests.test_vf_search',
4225
'breezy.tests.test_views',
4226
'breezy.tests.test_weave',
4227
'breezy.tests.test_whitebox',
4228
'breezy.tests.test_win32utils',
4229
'breezy.tests.test_workingtree',
4230
'breezy.tests.test_workingtree_4',
4231
'breezy.tests.test_wsgi',
4232
'breezy.tests.test_xml',
4236
def _test_suite_modules_to_doctest():
4237
"""Return the list of modules to doctest."""
4239
# GZ 2009-03-31: No docstrings with -OO so there's nothing to doctest
4243
'breezy.branchbuilder',
4244
'breezy.bzr.inventory',
4245
'breezy.decorators',
4246
'breezy.iterablefile',
4251
'breezy.symbol_versioning',
4253
'breezy.tests.fixtures',
4255
'breezy.transport.http',
4256
'breezy.version_info_formats.format_custom',
4260
def test_suite(keep_only=None, starting_with=None):
4261
"""Build and return TestSuite for the whole of breezy.
4263
:param keep_only: A list of test ids limiting the suite returned.
4265
:param starting_with: An id limiting the suite returned to the tests
4268
This function can be replaced if you need to change the default test
4269
suite on a global basis, but it is not encouraged.
4272
loader = TestUtil.TestLoader()
4274
if keep_only is not None:
4275
id_filter = TestIdList(keep_only)
4277
# We take precedence over keep_only because *at loading time* using
4278
# both options means we will load less tests for the same final result.
4279
def interesting_module(name):
4280
for start in starting_with:
4281
# Either the module name starts with the specified string
4282
# or it may contain tests starting with the specified string
4283
if name.startswith(start) or start.startswith(name):
4286
loader = TestUtil.FilteredByModuleTestLoader(interesting_module)
4288
elif keep_only is not None:
4289
loader = TestUtil.FilteredByModuleTestLoader(id_filter.refers_to)
4291
def interesting_module(name):
4292
return id_filter.refers_to(name)
4295
loader = TestUtil.TestLoader()
4297
def interesting_module(name):
4298
# No filtering, all modules are interesting
4301
suite = loader.suiteClass()
4303
# modules building their suite with loadTestsFromModuleNames
4304
suite.addTest(loader.loadTestsFromModuleNames(_test_suite_testmod_names()))
4307
suite.addTest(loader.loadTestsFromModuleNames(['breezy.doc']))
4309
# It's pretty much impossible to write readable doctests that work on
4310
# both Python 2 and Python 3 because of their overreliance on
4311
# consistent repr() return values.
4312
# For now, just run doctests on Python 2 so we now they haven't broken.
4313
for mod in _test_suite_modules_to_doctest():
4314
if not interesting_module(mod):
4315
# No tests to keep here, move along
4318
# note that this really does mean "report only" -- doctest
4319
# still runs the rest of the examples
4320
doc_suite = IsolatedDocTestSuite(
4321
mod, optionflags=doctest.REPORT_ONLY_FIRST_FAILURE)
4322
except ValueError as e:
4323
print('**failed to get doctest for: %s\n%s' % (mod, e))
4325
if len(doc_suite._tests) == 0:
4326
raise errors.BzrError("no doctests found in %s" % (mod,))
4327
suite.addTest(doc_suite)
4329
default_encoding = sys.getdefaultencoding()
4330
for name, plugin in _mod_plugin.plugins().items():
4331
if not interesting_module(plugin.module.__name__):
4333
plugin_suite = plugin.test_suite()
4334
# We used to catch ImportError here and turn it into just a warning,
4335
# but really if you don't have --no-plugins this should be a failure.
4336
# mbp 20080213 - see http://bugs.launchpad.net/bugs/189771
4337
if plugin_suite is None:
4338
plugin_suite = plugin.load_plugin_tests(loader)
4339
if plugin_suite is not None:
4340
suite.addTest(plugin_suite)
4341
if default_encoding != sys.getdefaultencoding():
4343
'Plugin "%s" tried to reset default encoding to: %s', name,
4344
sys.getdefaultencoding())
4346
sys.setdefaultencoding(default_encoding)
4348
if keep_only is not None:
4349
# Now that the referred modules have loaded their tests, keep only the
4351
suite = filter_suite_by_id_list(suite, id_filter)
4352
# Do some sanity checks on the id_list filtering
4353
not_found, duplicates = suite_matches_id_list(suite, keep_only)
4355
# The tester has used both keep_only and starting_with, so he is
4356
# already aware that some tests are excluded from the list, there
4357
# is no need to tell him which.
4360
# Some tests mentioned in the list are not in the test suite. The
4361
# list may be out of date, report to the tester.
4362
for id in not_found:
4363
trace.warning('"%s" not found in the test suite', id)
4364
for id in duplicates:
4365
trace.warning('"%s" is used as an id by several tests', id)
4370
def multiply_scenarios(*scenarios):
4371
"""Multiply two or more iterables of scenarios.
4373
It is safe to pass scenario generators or iterators.
4375
:returns: A list of compound scenarios: the cross-product of all
4376
scenarios, with the names concatenated and the parameters
4379
return functools.reduce(_multiply_two_scenarios, map(list, scenarios))
4382
def _multiply_two_scenarios(scenarios_left, scenarios_right):
4383
"""Multiply two sets of scenarios.
4385
:returns: the cartesian product of the two sets of scenarios, that is
4386
a scenario for every possible combination of a left scenario and a
4390
('%s,%s' % (left_name, right_name),
4391
dict(left_dict, **right_dict))
4392
for left_name, left_dict in scenarios_left
4393
for right_name, right_dict in scenarios_right]
4396
def multiply_tests(tests, scenarios, result):
4397
"""Multiply tests_list by scenarios into result.
4399
This is the core workhorse for test parameterisation.
4401
Typically the load_tests() method for a per-implementation test suite will
4402
call multiply_tests and return the result.
4404
:param tests: The tests to parameterise.
4405
:param scenarios: The scenarios to apply: pairs of (scenario_name,
4406
scenario_param_dict).
4407
:param result: A TestSuite to add created tests to.
4409
This returns the passed in result TestSuite with the cross product of all
4410
the tests repeated once for each scenario. Each test is adapted by adding
4411
the scenario name at the end of its id(), and updating the test object's
4412
__dict__ with the scenario_param_dict.
4414
>>> import breezy.tests.test_sampler
4415
>>> r = multiply_tests(
4416
... breezy.tests.test_sampler.DemoTest('test_nothing'),
4417
... [('one', dict(param=1)),
4418
... ('two', dict(param=2))],
4419
... TestUtil.TestSuite())
4420
>>> tests = list(iter_suite_tests(r))
4424
'breezy.tests.test_sampler.DemoTest.test_nothing(one)'
4430
for test in iter_suite_tests(tests):
4431
apply_scenarios(test, scenarios, result)
4435
def apply_scenarios(test, scenarios, result):
4436
"""Apply the scenarios in scenarios to test and add to result.
4438
:param test: The test to apply scenarios to.
4439
:param scenarios: An iterable of scenarios to apply to test.
4441
:seealso: apply_scenario
4443
for scenario in scenarios:
4444
result.addTest(apply_scenario(test, scenario))
4448
def apply_scenario(test, scenario):
4449
"""Copy test and apply scenario to it.
4451
:param test: A test to adapt.
4452
:param scenario: A tuple describing the scenario.
4453
The first element of the tuple is the new test id.
4454
The second element is a dict containing attributes to set on the
4456
:return: The adapted test.
4458
new_id = "%s(%s)" % (test.id(), scenario[0])
4459
new_test = clone_test(test, new_id)
4460
for name, value in scenario[1].items():
4461
setattr(new_test, name, value)
4465
def clone_test(test, new_id):
4466
"""Clone a test giving it a new id.
4468
:param test: The test to clone.
4469
:param new_id: The id to assign to it.
4470
:return: The new test.
4472
new_test = copy.copy(test)
4473
new_test.id = lambda: new_id
4474
# XXX: Workaround <https://bugs.launchpad.net/testtools/+bug/637725>, which
4475
# causes cloned tests to share the 'details' dict. This makes it hard to
4476
# read the test output for parameterized tests, because tracebacks will be
4477
# associated with irrelevant tests.
4479
details = new_test._TestCase__details
4480
except AttributeError:
4481
# must be a different version of testtools than expected. Do nothing.
4484
# Reset the '__details' dict.
4485
new_test._TestCase__details = {}
4489
def permute_tests_for_extension(standard_tests, loader, py_module_name,
4491
"""Helper for permutating tests against an extension module.
4493
This is meant to be used inside a modules 'load_tests()' function. It will
4494
create 2 scenarios, and cause all tests in the 'standard_tests' to be run
4495
against both implementations. Setting 'test.module' to the appropriate
4496
module. See breezy.tests.test__chk_map.load_tests as an example.
4498
:param standard_tests: A test suite to permute
4499
:param loader: A TestLoader
4500
:param py_module_name: The python path to a python module that can always
4501
be loaded, and will be considered the 'python' implementation. (eg
4502
'breezy._chk_map_py')
4503
:param ext_module_name: The python path to an extension module. If the
4504
module cannot be loaded, a single test will be added, which notes that
4505
the module is not available. If it can be loaded, all standard_tests
4506
will be run against that module.
4507
:return: (suite, feature) suite is a test-suite that has all the permuted
4508
tests. feature is the Feature object that can be used to determine if
4509
the module is available.
4512
from .features import ModuleAvailableFeature
4513
py_module = pyutils.get_named_object(py_module_name)
4515
('python', {'module': py_module}),
4517
suite = loader.suiteClass()
4518
feature = ModuleAvailableFeature(ext_module_name)
4519
if feature.available():
4520
scenarios.append(('C', {'module': feature.module}))
4522
# the compiled module isn't available, so we add a failing test
4523
class FailWithoutFeature(TestCase):
4524
def test_fail(self):
4525
self.requireFeature(feature)
4526
suite.addTest(loader.loadTestsFromTestCase(FailWithoutFeature))
4527
result = multiply_tests(standard_tests, scenarios, suite)
4528
return result, feature
4531
def _rmtree_temp_dir(dirname, test_id=None):
4532
# If LANG=C we probably have created some bogus paths
4533
# which rmtree(unicode) will fail to delete
4534
# so make sure we are using rmtree(str) to delete everything
4535
# except on win32, where rmtree(str) will fail
4536
# since it doesn't have the property of byte-stream paths
4537
# (they are either ascii or mbcs)
4538
if sys.platform == 'win32' and isinstance(dirname, bytes):
4539
# make sure we are using the unicode win32 api
4540
dirname = dirname.decode('mbcs')
4542
dirname = dirname.encode(sys.getfilesystemencoding())
4544
osutils.rmtree(dirname)
4545
except OSError as e:
4546
# We don't want to fail here because some useful display will be lost
4547
# otherwise. Polluting the tmp dir is bad, but not giving all the
4548
# possible info to the test runner is even worse.
4549
if test_id is not None:
4550
ui.ui_factory.clear_term()
4551
sys.stderr.write('\nWhile running: %s\n' % (test_id,))
4552
# Ugly, but the last thing we want here is fail, so bear with it.
4553
printable_e = str(e).decode(osutils.get_user_encoding(), 'replace'
4554
).encode('ascii', 'replace')
4555
sys.stderr.write('Unable to remove testing dir %s\n%s'
4556
% (os.path.basename(dirname), printable_e))
4559
def probe_unicode_in_user_encoding():
4560
"""Try to encode several unicode strings to use in unicode-aware tests.
4561
Return first successfull match.
4563
:return: (unicode value, encoded plain string value) or (None, None)
4565
possible_vals = [u'm\xb5', u'\xe1', u'\u0410']
4566
for uni_val in possible_vals:
4568
str_val = uni_val.encode(osutils.get_user_encoding())
4569
except UnicodeEncodeError:
4570
# Try a different character
4573
return uni_val, str_val
4577
def probe_bad_non_ascii(encoding):
4578
"""Try to find [bad] character with code [128..255]
4579
that cannot be decoded to unicode in some encoding.
4580
Return None if all non-ascii characters is valid
4583
for i in range(128, 256):
4586
char.decode(encoding)
4587
except UnicodeDecodeError:
4592
# Only define SubUnitBzrRunner if subunit is available.
4594
from subunit import TestProtocolClient
4595
from subunit.test_results import AutoTimingTestResultDecorator
4597
class SubUnitBzrProtocolClientv1(TestProtocolClient):
4599
def stopTest(self, test):
4600
super(SubUnitBzrProtocolClientv1, self).stopTest(test)
4601
_clear__type_equality_funcs(test)
4603
def addSuccess(self, test, details=None):
4604
# The subunit client always includes the details in the subunit
4605
# stream, but we don't want to include it in ours.
4606
if details is not None and 'log' in details:
4608
return super(SubUnitBzrProtocolClientv1, self).addSuccess(
4611
class SubUnitBzrRunnerv1(TextTestRunner):
4613
def run(self, test):
4614
result = AutoTimingTestResultDecorator(
4615
SubUnitBzrProtocolClientv1(self.stream))
4623
from subunit.run import SubunitTestRunner
4625
class SubUnitBzrRunnerv2(TextTestRunner, SubunitTestRunner):
4627
def __init__(self, stream=sys.stderr, descriptions=0, verbosity=1,
4628
bench_history=None, strict=False, result_decorators=None):
4629
TextTestRunner.__init__(
4630
self, stream=stream,
4631
descriptions=descriptions, verbosity=verbosity,
4632
bench_history=bench_history, strict=strict,
4633
result_decorators=result_decorators)
4634
SubunitTestRunner.__init__(self, verbosity=verbosity,
4637
run = SubunitTestRunner.run
added
removed
breezy/tests/__init__.py
