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.
54
# nb: check this before importing anything else from within it
55
_testtools_version = getattr(testtools, '__version__', ())
56
if _testtools_version < (0, 9, 5):
57
raise ImportError("need at least testtools 0.9.5: %s is %r"
58
% (testtools.__file__, _testtools_version))
59
from testtools import content
65
commands as _mod_commands,
75
plugin as _mod_plugin,
82
transport as _mod_transport,
85
from breezy.bzr import (
91
# lsprof not available
93
from ..sixish import (
99
from ..bzr.smart import client, request
100
from ..transport import (
104
from ..tests import (
111
from ..tests.features import _CompatabilityThunkFeature
113
# Mark this python module as being part of the implementation
114
# of unittest: this gives us better tracebacks where the last
115
# shown frame is the test code, not our assertXYZ.
118
default_transport = test_server.LocalURLServer
121
_unitialized_attr = object()
122
"""A sentinel needed to act as a default value in a method signature."""
125
# Subunit result codes, defined here to prevent a hard dependency on subunit.
129
# These are intentionally brought into this namespace. That way plugins, etc
130
# can just "from breezy.tests import TestCase, TestLoader, etc"
131
TestSuite = TestUtil.TestSuite
132
TestLoader = TestUtil.TestLoader
134
# Tests should run in a clean and clearly defined environment. The goal is to
135
# keep them isolated from the running environment as mush as possible. The test
136
# framework ensures the variables defined below are set (or deleted if the
137
# value is None) before a test is run and reset to their original value after
138
# the test is run. Generally if some code depends on an environment variable,
139
# the tests should start without this variable in the environment. There are a
140
# few exceptions but you shouldn't violate this rule lightly.
145
'XDG_CONFIG_HOME': None,
146
# brz now uses the Win32 API and doesn't rely on APPDATA, but the
147
# tests do check our impls match APPDATA
148
'BRZ_EDITOR': None, # test_msgeditor manipulates this variable
152
'BZREMAIL': None, # may still be present in the environment
153
'EMAIL': 'jrandom@example.com', # set EMAIL as brz does not guess
154
'BRZ_PROGRESS_BAR': None,
155
# This should trap leaks to ~/.brz.log. This occurs when tests use TestCase
156
# as a base class instead of TestCaseInTempDir. Tests inheriting from
157
# TestCase should not use disk resources, BRZ_LOG is one.
158
'BRZ_LOG': '/you-should-use-TestCaseInTempDir-if-you-need-a-log-file',
159
'BRZ_PLUGIN_PATH': '-site',
160
'BRZ_DISABLE_PLUGINS': None,
161
'BRZ_PLUGINS_AT': None,
162
'BRZ_CONCURRENCY': None,
163
# Make sure that any text ui tests are consistent regardless of
164
# the environment the test case is run in; you may want tests that
165
# test other combinations. 'dumb' is a reasonable guess for tests
166
# going to a pipe or a BytesIO.
172
'SSH_AUTH_SOCK': None,
182
# Nobody cares about ftp_proxy, FTP_PROXY AFAIK. So far at
183
# least. If you do (care), please update this comment
187
'BZR_REMOTE_PATH': None,
188
# Generally speaking, we don't want apport reporting on crashes in
189
# the test envirnoment unless we're specifically testing apport,
190
# so that it doesn't leak into the real system environment. We
191
# use an env var so it propagates to subprocesses.
192
'APPORT_DISABLE': '1',
196
def override_os_environ(test, env=None):
197
"""Modify os.environ keeping a copy.
199
:param test: A test instance
201
:param env: A dict containing variable definitions to be installed
204
env = isolated_environ
205
test._original_os_environ = dict(**os.environ)
207
osutils.set_or_unset_env(var, env[var])
208
if var not in test._original_os_environ:
209
# The var is new, add it with a value of None, so
210
# restore_os_environ will delete it
211
test._original_os_environ[var] = None
214
def restore_os_environ(test):
215
"""Restore os.environ to its original state.
217
:param test: A test instance previously passed to override_os_environ.
219
for var, value in test._original_os_environ.items():
220
# Restore the original value (or delete it if the value has been set to
221
# None in override_os_environ).
222
osutils.set_or_unset_env(var, value)
225
def _clear__type_equality_funcs(test):
226
"""Cleanup bound methods stored on TestCase instances
228
Clear the dict breaking a few (mostly) harmless cycles in the affected
229
unittests released with Python 2.6 and initial Python 2.7 versions.
231
For a few revisions between Python 2.7.1 and Python 2.7.2 that annoyingly
232
shipped in Oneiric, an object with no clear method was used, hence the
233
extra complications, see bug 809048 for details.
235
type_equality_funcs = getattr(test, "_type_equality_funcs", None)
236
if type_equality_funcs is not None:
237
tef_clear = getattr(type_equality_funcs, "clear", None)
238
if tef_clear is None:
239
tef_instance_dict = getattr(type_equality_funcs, "__dict__", None)
240
if tef_instance_dict is not None:
241
tef_clear = tef_instance_dict.clear
242
if tef_clear is not None:
246
class ExtendedTestResult(testtools.TextTestResult):
247
"""Accepts, reports and accumulates the results of running tests.
249
Compared to the unittest version this class adds support for
250
profiling, benchmarking, stopping as soon as a test fails, and
251
skipping tests. There are further-specialized subclasses for
252
different types of display.
254
When a test finishes, in whatever way, it calls one of the addSuccess,
255
addFailure or addError methods. These in turn may redirect to a more
256
specific case for the special test results supported by our extended
259
Note that just one of these objects is fed the results from many tests.
264
def __init__(self, stream, descriptions, verbosity,
268
"""Construct new TestResult.
270
:param bench_history: Optionally, a writable file object to accumulate
273
testtools.TextTestResult.__init__(self, stream)
274
if bench_history is not None:
275
from breezy.version import _get_bzr_source_tree
276
src_tree = _get_bzr_source_tree()
279
revision_id = src_tree.get_parent_ids()[0]
281
# XXX: if this is a brand new tree, do the same as if there
285
# XXX: If there's no branch, what should we do?
287
bench_history.write("--date %s %s\n" % (time.time(), revision_id))
288
self._bench_history = bench_history
289
self.ui = ui.ui_factory
292
self.failure_count = 0
293
self.known_failure_count = 0
295
self.not_applicable_count = 0
296
self.unsupported = {}
298
self._overall_start_time = time.time()
299
self._strict = strict
300
self._first_thread_leaker_id = None
301
self._tests_leaking_threads_count = 0
302
self._traceback_from_test = None
304
def stopTestRun(self):
307
stopTime = time.time()
308
timeTaken = stopTime - self.startTime
309
# GZ 2010-07-19: Seems testtools has no printErrors method, and though
310
# the parent class method is similar have to duplicate
311
self._show_list('ERROR', self.errors)
312
self._show_list('FAIL', self.failures)
313
self.stream.write(self.sep2)
314
self.stream.write("%s %d test%s in %.3fs\n\n" % (actionTaken,
315
run, run != 1 and "s" or "", timeTaken))
316
if not self.wasSuccessful():
317
self.stream.write("FAILED (")
318
failed, errored = map(len, (self.failures, self.errors))
320
self.stream.write("failures=%d" % failed)
322
if failed: self.stream.write(", ")
323
self.stream.write("errors=%d" % errored)
324
if self.known_failure_count:
325
if failed or errored: 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
372
return math.ceil((a_timedelta.days * 86400.0 + a_timedelta.seconds +
373
a_timedelta.microseconds / 1000000.0) * shift) / shift
375
def _elapsedTestTimeString(self):
376
"""Return a time string for the overall time the current test has taken."""
377
return self._formatTime(self._delta_to_float(
378
self._now() - self._start_datetime, 3))
380
def _testTimeString(self, testCase):
381
benchmark_time = self._extractBenchmarkTime(testCase)
382
if benchmark_time is not None:
383
return self._formatTime(benchmark_time) + "*"
385
return self._elapsedTestTimeString()
387
def _formatTime(self, seconds):
388
"""Format seconds as milliseconds with leading spaces."""
389
# some benchmarks can take thousands of seconds to run, so we need 8
391
return "%8dms" % (1000 * seconds)
393
def _shortened_test_description(self, test):
395
what = re.sub(r'^breezy\.tests\.', '', what)
398
# GZ 2010-10-04: Cloned tests may end up harmlessly calling this method
399
# multiple times in a row, because the handler is added for
400
# each test but the container list is shared between cases.
401
# See lp:498869 lp:625574 and lp:637725 for background.
402
def _record_traceback_from_test(self, exc_info):
403
"""Store the traceback from passed exc_info tuple till"""
404
self._traceback_from_test = exc_info[2]
406
def startTest(self, test):
407
super(ExtendedTestResult, self).startTest(test)
411
self.report_test_start(test)
412
test.number = self.count
413
self._recordTestStartTime()
414
# Make testtools cases give us the real traceback on failure
415
addOnException = getattr(test, "addOnException", None)
416
if addOnException is not None:
417
addOnException(self._record_traceback_from_test)
418
# Only check for thread leaks on breezy derived test cases
419
if isinstance(test, TestCase):
420
test.addCleanup(self._check_leaked_threads, test)
422
def stopTest(self, test):
423
super(ExtendedTestResult, self).stopTest(test)
424
# Manually break cycles, means touching various private things but hey
425
getDetails = getattr(test, "getDetails", None)
426
if getDetails is not None:
428
_clear__type_equality_funcs(test)
429
self._traceback_from_test = None
431
def startTests(self):
432
self.report_tests_starting()
433
self._active_threads = threading.enumerate()
435
def _check_leaked_threads(self, test):
436
"""See if any threads have leaked since last call
438
A sample of live threads is stored in the _active_threads attribute,
439
when this method runs it compares the current live threads and any not
440
in the previous sample are treated as having leaked.
442
now_active_threads = set(threading.enumerate())
443
threads_leaked = now_active_threads.difference(self._active_threads)
445
self._report_thread_leak(test, threads_leaked, now_active_threads)
446
self._tests_leaking_threads_count += 1
447
if self._first_thread_leaker_id is None:
448
self._first_thread_leaker_id = test.id()
449
self._active_threads = now_active_threads
451
def _recordTestStartTime(self):
452
"""Record that a test has started."""
453
self._start_datetime = self._now()
455
def addError(self, test, err):
456
"""Tell result that test finished with an error.
458
Called from the TestCase run() method when the test
459
fails with an unexpected error.
461
self._post_mortem(self._traceback_from_test)
462
super(ExtendedTestResult, self).addError(test, err)
463
self.error_count += 1
464
self.report_error(test, err)
468
def addFailure(self, test, err):
469
"""Tell result that test failed.
471
Called from the TestCase run() method when the test
472
fails because e.g. an assert() method failed.
474
self._post_mortem(self._traceback_from_test)
475
super(ExtendedTestResult, self).addFailure(test, err)
476
self.failure_count += 1
477
self.report_failure(test, err)
481
def addSuccess(self, test, details=None):
482
"""Tell result that test completed successfully.
484
Called from the TestCase run()
486
if self._bench_history is not None:
487
benchmark_time = self._extractBenchmarkTime(test, details)
488
if benchmark_time is not None:
489
self._bench_history.write("%s %s\n" % (
490
self._formatTime(benchmark_time),
492
self.report_success(test)
493
super(ExtendedTestResult, self).addSuccess(test)
494
test._log_contents = ''
496
def addExpectedFailure(self, test, err):
497
self.known_failure_count += 1
498
self.report_known_failure(test, err)
500
def addUnexpectedSuccess(self, test, details=None):
501
"""Tell result the test unexpectedly passed, counting as a failure
503
When the minimum version of testtools required becomes 0.9.8 this
504
can be updated to use the new handling there.
506
super(ExtendedTestResult, self).addFailure(test, details=details)
507
self.failure_count += 1
508
self.report_unexpected_success(test,
509
"".join(details["reason"].iter_text()))
513
def addNotSupported(self, test, feature):
514
"""The test will not be run because of a missing feature.
516
# this can be called in two different ways: it may be that the
517
# test started running, and then raised (through requireFeature)
518
# UnavailableFeature. Alternatively this method can be called
519
# while probing for features before running the test code proper; in
520
# that case we will see startTest and stopTest, but the test will
521
# never actually run.
522
self.unsupported.setdefault(str(feature), 0)
523
self.unsupported[str(feature)] += 1
524
self.report_unsupported(test, feature)
526
def addSkip(self, test, reason):
527
"""A test has not run for 'reason'."""
529
self.report_skip(test, reason)
531
def addNotApplicable(self, test, reason):
532
self.not_applicable_count += 1
533
self.report_not_applicable(test, reason)
535
def _count_stored_tests(self):
536
"""Count of tests instances kept alive due to not succeeding"""
537
return self.error_count + self.failure_count + self.known_failure_count
539
def _post_mortem(self, tb=None):
540
"""Start a PDB post mortem session."""
541
if os.environ.get('BRZ_TEST_PDB', None):
545
def progress(self, offset, whence):
546
"""The test is adjusting the count of tests to run."""
547
if whence == SUBUNIT_SEEK_SET:
548
self.num_tests = offset
549
elif whence == SUBUNIT_SEEK_CUR:
550
self.num_tests += offset
552
raise errors.BzrError("Unknown whence %r" % whence)
554
def report_tests_starting(self):
555
"""Display information before the test run begins"""
556
if getattr(sys, 'frozen', None) is None:
557
bzr_path = osutils.realpath(sys.argv[0])
559
bzr_path = sys.executable
561
'brz selftest: %s\n' % (bzr_path,))
564
breezy.__path__[0],))
566
' bzr-%s python-%s %s\n' % (
567
breezy.version_string,
568
breezy._format_version_tuple(sys.version_info),
569
platform.platform(aliased=1),
571
self.stream.write('\n')
573
def report_test_start(self, test):
574
"""Display information on the test just about to be run"""
576
def _report_thread_leak(self, test, leaked_threads, active_threads):
577
"""Display information on a test that leaked one or more threads"""
578
# GZ 2010-09-09: A leak summary reported separately from the general
579
# thread debugging would be nice. Tests under subunit
580
# need something not using stream, perhaps adding a
581
# testtools details object would be fitting.
582
if 'threads' in selftest_debug_flags:
583
self.stream.write('%s is leaking, active is now %d\n' %
584
(test.id(), len(active_threads)))
586
def startTestRun(self):
587
self.startTime = time.time()
589
def report_success(self, test):
592
def wasStrictlySuccessful(self):
593
if self.unsupported or self.known_failure_count:
595
return self.wasSuccessful()
598
class TextTestResult(ExtendedTestResult):
599
"""Displays progress and results of tests in text form"""
601
def __init__(self, stream, descriptions, verbosity,
605
ExtendedTestResult.__init__(self, stream, descriptions, verbosity,
606
bench_history, strict)
607
self.pb = self.ui.nested_progress_bar()
608
self.pb.show_pct = False
609
self.pb.show_spinner = False
610
self.pb.show_eta = False,
611
self.pb.show_count = False
612
self.pb.show_bar = False
613
self.pb.update_latency = 0
614
self.pb.show_transport_activity = False
616
def stopTestRun(self):
617
# called when the tests that are going to run have run
620
super(TextTestResult, self).stopTestRun()
622
def report_tests_starting(self):
623
super(TextTestResult, self).report_tests_starting()
624
self.pb.update('[test 0/%d] Starting' % (self.num_tests))
626
def _progress_prefix_text(self):
627
# the longer this text, the less space we have to show the test
629
a = '[%d' % self.count # total that have been run
630
# tests skipped as known not to be relevant are not important enough
632
## if self.skip_count:
633
## a += ', %d skip' % self.skip_count
634
## if self.known_failure_count:
635
## a += '+%dX' % self.known_failure_count
637
a +='/%d' % self.num_tests
639
runtime = time.time() - self._overall_start_time
641
a += '%dm%ds' % (runtime / 60, runtime % 60)
644
total_fail_count = self.error_count + self.failure_count
646
a += ', %d failed' % total_fail_count
647
# if self.unsupported:
648
# a += ', %d missing' % len(self.unsupported)
652
def report_test_start(self, test):
654
self._progress_prefix_text()
656
+ self._shortened_test_description(test))
658
def _test_description(self, test):
659
return self._shortened_test_description(test)
661
def report_error(self, test, err):
662
self.stream.write('ERROR: %s\n %s\n' % (
663
self._test_description(test),
667
def report_failure(self, test, err):
668
self.stream.write('FAIL: %s\n %s\n' % (
669
self._test_description(test),
673
def report_known_failure(self, test, err):
676
def report_unexpected_success(self, test, reason):
677
self.stream.write('FAIL: %s\n %s: %s\n' % (
678
self._test_description(test),
679
"Unexpected success. Should have failed",
683
def report_skip(self, test, reason):
686
def report_not_applicable(self, test, reason):
689
def report_unsupported(self, test, feature):
690
"""test cannot be run because feature is missing."""
693
class VerboseTestResult(ExtendedTestResult):
694
"""Produce long output, with one line per test run plus times"""
696
def _ellipsize_to_right(self, a_string, final_width):
697
"""Truncate and pad a string, keeping the right hand side"""
698
if len(a_string) > final_width:
699
result = '...' + a_string[3-final_width:]
702
return result.ljust(final_width)
704
def report_tests_starting(self):
705
self.stream.write('running %d tests...\n' % self.num_tests)
706
super(VerboseTestResult, self).report_tests_starting()
708
def report_test_start(self, test):
709
name = self._shortened_test_description(test)
710
width = osutils.terminal_width()
711
if width is not None:
712
# width needs space for 6 char status, plus 1 for slash, plus an
713
# 11-char time string, plus a trailing blank
714
# when NUMBERED_DIRS: plus 5 chars on test number, plus 1 char on
716
self.stream.write(self._ellipsize_to_right(name, width-18))
718
self.stream.write(name)
721
def _error_summary(self, err):
723
return '%s%s' % (indent, err[1])
725
def report_error(self, test, err):
726
self.stream.write('ERROR %s\n%s\n'
727
% (self._testTimeString(test),
728
self._error_summary(err)))
730
def report_failure(self, test, err):
731
self.stream.write(' FAIL %s\n%s\n'
732
% (self._testTimeString(test),
733
self._error_summary(err)))
735
def report_known_failure(self, test, err):
736
self.stream.write('XFAIL %s\n%s\n'
737
% (self._testTimeString(test),
738
self._error_summary(err)))
740
def report_unexpected_success(self, test, reason):
741
self.stream.write(' FAIL %s\n%s: %s\n'
742
% (self._testTimeString(test),
743
"Unexpected success. Should have failed",
746
def report_success(self, test):
747
self.stream.write(' OK %s\n' % self._testTimeString(test))
748
for bench_called, stats in getattr(test, '_benchcalls', []):
749
self.stream.write('LSProf output for %s(%s, %s)\n' % bench_called)
750
stats.pprint(file=self.stream)
751
# flush the stream so that we get smooth output. This verbose mode is
752
# used to show the output in PQM.
755
def report_skip(self, test, reason):
756
self.stream.write(' SKIP %s\n%s\n'
757
% (self._testTimeString(test), reason))
759
def report_not_applicable(self, test, reason):
760
self.stream.write(' N/A %s\n %s\n'
761
% (self._testTimeString(test), reason))
763
def report_unsupported(self, test, feature):
764
"""test cannot be run because feature is missing."""
765
self.stream.write("NODEP %s\n The feature '%s' is not available.\n"
766
%(self._testTimeString(test), feature))
769
class TextTestRunner(object):
770
stop_on_failure = False
778
result_decorators=None,
780
"""Create a TextTestRunner.
782
:param result_decorators: An optional list of decorators to apply
783
to the result object being used by the runner. Decorators are
784
applied left to right - the first element in the list is the
787
# stream may know claim to know to write unicode strings, but in older
788
# pythons this goes sufficiently wrong that it is a bad idea. (
789
# specifically a built in file with encoding 'UTF-8' will still try
790
# to encode using ascii.
791
new_encoding = osutils.get_terminal_encoding()
792
codec = codecs.lookup(new_encoding)
793
encode = codec.encode
794
# GZ 2010-09-08: Really we don't want to be writing arbitrary bytes,
795
# so should swap to the plain codecs.StreamWriter
796
stream = osutils.UnicodeOrBytesToBytesWriter(encode, stream,
798
stream.encoding = new_encoding
800
self.descriptions = descriptions
801
self.verbosity = verbosity
802
self._bench_history = bench_history
803
self._strict = strict
804
self._result_decorators = result_decorators or []
807
"Run the given test case or test suite."
808
if self.verbosity == 1:
809
result_class = TextTestResult
810
elif self.verbosity >= 2:
811
result_class = VerboseTestResult
812
original_result = result_class(self.stream,
815
bench_history=self._bench_history,
818
# Signal to result objects that look at stop early policy to stop,
819
original_result.stop_early = self.stop_on_failure
820
result = original_result
821
for decorator in self._result_decorators:
822
result = decorator(result)
823
result.stop_early = self.stop_on_failure
824
result.startTestRun()
829
# higher level code uses our extended protocol to determine
830
# what exit code to give.
831
return original_result
834
def iter_suite_tests(suite):
835
"""Return all tests in a suite, recursing through nested suites"""
836
if isinstance(suite, unittest.TestCase):
838
elif isinstance(suite, unittest.TestSuite):
840
for r in iter_suite_tests(item):
843
raise Exception('unknown type %r for object %r'
844
% (type(suite), suite))
847
TestSkipped = testtools.testcase.TestSkipped
850
class TestNotApplicable(TestSkipped):
851
"""A test is not applicable to the situation where it was run.
853
This is only normally raised by parameterized tests, if they find that
854
the instance they're constructed upon does not support one aspect
859
# traceback._some_str fails to format exceptions that have the default
860
# __str__ which does an implicit ascii conversion. However, repr() on those
861
# objects works, for all that its not quite what the doctor may have ordered.
862
def _clever_some_str(value):
867
return repr(value).replace('\\n', '\n')
869
return '<unprintable %s object>' % type(value).__name__
871
traceback._some_str = _clever_some_str
874
# deprecated - use self.knownFailure(), or self.expectFailure.
875
KnownFailure = testtools.testcase._ExpectedFailure
878
class UnavailableFeature(Exception):
879
"""A feature required for this test was not available.
881
This can be considered a specialised form of SkippedTest.
883
The feature should be used to construct the exception.
887
class StringIOWrapper(ui_testing.BytesIOWithEncoding):
889
@symbol_versioning.deprecated_method(
890
symbol_versioning.deprecated_in((3, 0)))
891
def __init__(self, s=None):
892
super(StringIOWrapper, self).__init__(s)
895
TestUIFactory = ui_testing.TestUIFactory
898
def isolated_doctest_setUp(test):
899
override_os_environ(test)
900
osutils.set_or_unset_env('BRZ_HOME', '/nonexistent')
901
test._orig_ui_factory = ui.ui_factory
902
ui.ui_factory = ui.SilentUIFactory()
905
def isolated_doctest_tearDown(test):
906
restore_os_environ(test)
907
ui.ui_factory = test._orig_ui_factory
910
def IsolatedDocTestSuite(*args, **kwargs):
911
"""Overrides doctest.DocTestSuite to handle isolation.
913
The method is really a factory and users are expected to use it as such.
916
kwargs['setUp'] = isolated_doctest_setUp
917
kwargs['tearDown'] = isolated_doctest_tearDown
918
return doctest.DocTestSuite(*args, **kwargs)
921
class TestCase(testtools.TestCase):
922
"""Base class for brz unit tests.
924
Tests that need access to disk resources should subclass
925
TestCaseInTempDir not TestCase.
927
Error and debug log messages are redirected from their usual
928
location into a temporary file, the contents of which can be
929
retrieved by _get_log(). We use a real OS file, not an in-memory object,
930
so that it can also capture file IO. When the test completes this file
931
is read into memory and removed from disk.
933
There are also convenience functions to invoke bzr's command-line
934
routine, and to build and check brz trees.
936
In addition to the usual method of overriding tearDown(), this class also
937
allows subclasses to register cleanup functions via addCleanup, which are
938
run in order as the object is torn down. It's less likely this will be
939
accidentally overlooked.
943
# record lsprof data when performing benchmark calls.
944
_gather_lsprof_in_benchmarks = False
946
def __init__(self, methodName='testMethod'):
947
super(TestCase, self).__init__(methodName)
948
self._directory_isolation = True
949
self.exception_handlers.insert(0,
950
(UnavailableFeature, self._do_unsupported_or_skip))
951
self.exception_handlers.insert(0,
952
(TestNotApplicable, self._do_not_applicable))
955
super(TestCase, self).setUp()
957
# At this point we're still accessing the config files in $BRZ_HOME (as
958
# set by the user running selftest).
959
timeout = config.GlobalStack().get('selftest.timeout')
961
timeout_fixture = fixtures.TimeoutFixture(timeout)
962
timeout_fixture.setUp()
963
self.addCleanup(timeout_fixture.cleanUp)
965
for feature in getattr(self, '_test_needs_features', []):
966
self.requireFeature(feature)
967
self._cleanEnvironment()
969
self.overrideAttr(breezy.get_global_state(), 'cmdline_overrides',
970
config.CommandLineStore())
974
self._benchcalls = []
975
self._benchtime = None
977
self._track_transports()
979
self._clear_debug_flags()
980
# Isolate global verbosity level, to make sure it's reproducible
981
# between tests. We should get rid of this altogether: bug 656694. --
983
self.overrideAttr(breezy.trace, '_verbosity_level', 0)
984
self._log_files = set()
985
# Each key in the ``_counters`` dict holds a value for a different
986
# counter. When the test ends, addDetail() should be used to output the
987
# counter values. This happens in install_counter_hook().
989
if 'config_stats' in selftest_debug_flags:
990
self._install_config_stats_hooks()
991
# Do not use i18n for tests (unless the test reverses this)
997
# The sys preserved stdin/stdout should allow blackbox tests debugging
998
pdb.Pdb(stdin=sys.__stdin__, stdout=sys.__stdout__
999
).set_trace(sys._getframe().f_back)
1001
def discardDetail(self, name):
1002
"""Extend the addDetail, getDetails api so we can remove a detail.
1004
eg. brz always adds the 'log' detail at startup, but we don't want to
1005
include it for skipped, xfail, etc tests.
1007
It is safe to call this for a detail that doesn't exist, in case this
1008
gets called multiple times.
1010
# We cheat. details is stored in __details which means we shouldn't
1011
# touch it. but getDetails() returns the dict directly, so we can
1013
details = self.getDetails()
1017
def install_counter_hook(self, hooks, name, counter_name=None):
1018
"""Install a counting hook.
1020
Any hook can be counted as long as it doesn't need to return a value.
1022
:param hooks: Where the hook should be installed.
1024
:param name: The hook name that will be counted.
1026
:param counter_name: The counter identifier in ``_counters``, defaults
1029
_counters = self._counters # Avoid closing over self
1030
if counter_name is None:
1032
if counter_name in _counters:
1033
raise AssertionError('%s is already used as a counter name'
1035
_counters[counter_name] = 0
1036
self.addDetail(counter_name, content.Content(content.UTF8_TEXT,
1037
lambda: [b'%d' % (_counters[counter_name],)]))
1038
def increment_counter(*args, **kwargs):
1039
_counters[counter_name] += 1
1040
label = 'count %s calls' % (counter_name,)
1041
hooks.install_named_hook(name, increment_counter, label)
1042
self.addCleanup(hooks.uninstall_named_hook, name, label)
1044
def _install_config_stats_hooks(self):
1045
"""Install config hooks to count hook calls.
1048
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1049
self.install_counter_hook(config.ConfigHooks, hook_name,
1050
'config.%s' % (hook_name,))
1052
# The OldConfigHooks are private and need special handling to protect
1053
# against recursive tests (tests that run other tests), so we just do
1054
# manually what registering them into _builtin_known_hooks will provide
1056
self.overrideAttr(config, 'OldConfigHooks', config._OldConfigHooks())
1057
for hook_name in ('get', 'set', 'remove', 'load', 'save'):
1058
self.install_counter_hook(config.OldConfigHooks, hook_name,
1059
'old_config.%s' % (hook_name,))
1061
def _clear_debug_flags(self):
1062
"""Prevent externally set debug flags affecting tests.
1064
Tests that want to use debug flags can just set them in the
1065
debug_flags set during setup/teardown.
1067
# Start with a copy of the current debug flags we can safely modify.
1068
self.overrideAttr(debug, 'debug_flags', set(debug.debug_flags))
1069
if 'allow_debug' not in selftest_debug_flags:
1070
debug.debug_flags.clear()
1071
if 'disable_lock_checks' not in selftest_debug_flags:
1072
debug.debug_flags.add('strict_locks')
1074
def _clear_hooks(self):
1075
# prevent hooks affecting tests
1076
known_hooks = hooks.known_hooks
1077
self._preserved_hooks = {}
1078
for key, (parent, name) in known_hooks.iter_parent_objects():
1079
current_hooks = getattr(parent, name)
1080
self._preserved_hooks[parent] = (name, current_hooks)
1081
self._preserved_lazy_hooks = hooks._lazy_hooks
1082
hooks._lazy_hooks = {}
1083
self.addCleanup(self._restoreHooks)
1084
for key, (parent, name) in known_hooks.iter_parent_objects():
1085
factory = known_hooks.get(key)
1086
setattr(parent, name, factory())
1087
# this hook should always be installed
1088
request._install_hook()
1090
def disable_directory_isolation(self):
1091
"""Turn off directory isolation checks."""
1092
self._directory_isolation = False
1094
def enable_directory_isolation(self):
1095
"""Enable directory isolation checks."""
1096
self._directory_isolation = True
1098
def _silenceUI(self):
1099
"""Turn off UI for duration of test"""
1100
# by default the UI is off; tests can turn it on if they want it.
1101
self.overrideAttr(ui, 'ui_factory', ui.SilentUIFactory())
1103
def _check_locks(self):
1104
"""Check that all lock take/release actions have been paired."""
1105
# We always check for mismatched locks. If a mismatch is found, we
1106
# fail unless -Edisable_lock_checks is supplied to selftest, in which
1107
# case we just print a warning.
1109
acquired_locks = [lock for action, lock in self._lock_actions
1110
if action == 'acquired']
1111
released_locks = [lock for action, lock in self._lock_actions
1112
if action == 'released']
1113
broken_locks = [lock for action, lock in self._lock_actions
1114
if action == 'broken']
1115
# trivially, given the tests for lock acquistion and release, if we
1116
# have as many in each list, it should be ok. Some lock tests also
1117
# break some locks on purpose and should be taken into account by
1118
# considering that breaking a lock is just a dirty way of releasing it.
1119
if len(acquired_locks) != (len(released_locks) + len(broken_locks)):
1121
'Different number of acquired and '
1122
'released or broken locks.\n'
1126
(acquired_locks, released_locks, broken_locks))
1127
if not self._lock_check_thorough:
1128
# Rather than fail, just warn
1129
print("Broken test %s: %s" % (self, message))
1133
def _track_locks(self):
1134
"""Track lock activity during tests."""
1135
self._lock_actions = []
1136
if 'disable_lock_checks' in selftest_debug_flags:
1137
self._lock_check_thorough = False
1139
self._lock_check_thorough = True
1141
self.addCleanup(self._check_locks)
1142
_mod_lock.Lock.hooks.install_named_hook('lock_acquired',
1143
self._lock_acquired, None)
1144
_mod_lock.Lock.hooks.install_named_hook('lock_released',
1145
self._lock_released, None)
1146
_mod_lock.Lock.hooks.install_named_hook('lock_broken',
1147
self._lock_broken, None)
1149
def _lock_acquired(self, result):
1150
self._lock_actions.append(('acquired', result))
1152
def _lock_released(self, result):
1153
self._lock_actions.append(('released', result))
1155
def _lock_broken(self, result):
1156
self._lock_actions.append(('broken', result))
1158
def permit_dir(self, name):
1159
"""Permit a directory to be used by this test. See permit_url."""
1160
name_transport = _mod_transport.get_transport_from_path(name)
1161
self.permit_url(name)
1162
self.permit_url(name_transport.base)
1164
def permit_url(self, url):
1165
"""Declare that url is an ok url to use in this test.
1167
Do this for memory transports, temporary test directory etc.
1169
Do not do this for the current working directory, /tmp, or any other
1170
preexisting non isolated url.
1172
if not url.endswith('/'):
1174
self._bzr_selftest_roots.append(url)
1176
def permit_source_tree_branch_repo(self):
1177
"""Permit the source tree brz is running from to be opened.
1179
Some code such as breezy.version attempts to read from the brz branch
1180
that brz is executing from (if any). This method permits that directory
1181
to be used in the test suite.
1183
path = self.get_source_path()
1184
self.record_directory_isolation()
1187
workingtree.WorkingTree.open(path)
1188
except (errors.NotBranchError, errors.NoWorkingTree):
1189
raise TestSkipped('Needs a working tree of brz sources')
1191
self.enable_directory_isolation()
1193
def _preopen_isolate_transport(self, transport):
1194
"""Check that all transport openings are done in the test work area."""
1195
while isinstance(transport, pathfilter.PathFilteringTransport):
1196
# Unwrap pathfiltered transports
1197
transport = transport.server.backing_transport.clone(
1198
transport._filter('.'))
1199
url = transport.base
1200
# ReadonlySmartTCPServer_for_testing decorates the backing transport
1201
# urls it is given by prepending readonly+. This is appropriate as the
1202
# client shouldn't know that the server is readonly (or not readonly).
1203
# We could register all servers twice, with readonly+ prepending, but
1204
# that makes for a long list; this is about the same but easier to
1206
if url.startswith('readonly+'):
1207
url = url[len('readonly+'):]
1208
self._preopen_isolate_url(url)
1210
def _preopen_isolate_url(self, url):
1211
if not self._directory_isolation:
1213
if self._directory_isolation == 'record':
1214
self._bzr_selftest_roots.append(url)
1216
# This prevents all transports, including e.g. sftp ones backed on disk
1217
# from working unless they are explicitly granted permission. We then
1218
# depend on the code that sets up test transports to check that they are
1219
# appropriately isolated and enable their use by calling
1220
# self.permit_transport()
1221
if not osutils.is_inside_any(self._bzr_selftest_roots, url):
1222
raise errors.BzrError("Attempt to escape test isolation: %r %r"
1223
% (url, self._bzr_selftest_roots))
1225
def record_directory_isolation(self):
1226
"""Gather accessed directories to permit later access.
1228
This is used for tests that access the branch brz is running from.
1230
self._directory_isolation = "record"
1232
def start_server(self, transport_server, backing_server=None):
1233
"""Start transport_server for this test.
1235
This starts the server, registers a cleanup for it and permits the
1236
server's urls to be used.
1238
if backing_server is None:
1239
transport_server.start_server()
1241
transport_server.start_server(backing_server)
1242
self.addCleanup(transport_server.stop_server)
1243
# Obtain a real transport because if the server supplies a password, it
1244
# will be hidden from the base on the client side.
1245
t = _mod_transport.get_transport_from_url(transport_server.get_url())
1246
# Some transport servers effectively chroot the backing transport;
1247
# others like SFTPServer don't - users of the transport can walk up the
1248
# transport to read the entire backing transport. This wouldn't matter
1249
# except that the workdir tests are given - and that they expect the
1250
# server's url to point at - is one directory under the safety net. So
1251
# Branch operations into the transport will attempt to walk up one
1252
# directory. Chrooting all servers would avoid this but also mean that
1253
# we wouldn't be testing directly against non-root urls. Alternatively
1254
# getting the test framework to start the server with a backing server
1255
# at the actual safety net directory would work too, but this then
1256
# means that the self.get_url/self.get_transport methods would need
1257
# to transform all their results. On balance its cleaner to handle it
1258
# here, and permit a higher url when we have one of these transports.
1259
if t.base.endswith('/work/'):
1260
# we have safety net/test root/work
1261
t = t.clone('../..')
1262
elif isinstance(transport_server,
1263
test_server.SmartTCPServer_for_testing):
1264
# The smart server adds a path similar to work, which is traversed
1265
# up from by the client. But the server is chrooted - the actual
1266
# backing transport is not escaped from, and VFS requests to the
1267
# root will error (because they try to escape the chroot).
1269
while t2.base != t.base:
1272
self.permit_url(t.base)
1274
def _track_transports(self):
1275
"""Install checks for transport usage."""
1276
# TestCase has no safe place it can write to.
1277
self._bzr_selftest_roots = []
1278
# Currently the easiest way to be sure that nothing is going on is to
1279
# hook into brz dir opening. This leaves a small window of error for
1280
# transport tests, but they are well known, and we can improve on this
1282
controldir.ControlDir.hooks.install_named_hook("pre_open",
1283
self._preopen_isolate_transport, "Check brz directories are safe.")
1285
def _ndiff_strings(self, a, b):
1286
"""Return ndiff between two strings containing lines.
1288
A trailing newline is added if missing to make the strings
1290
if b and b[-1] != '\n':
1292
if a and a[-1] != '\n':
1294
difflines = difflib.ndiff(a.splitlines(True),
1296
linejunk=lambda x: False,
1297
charjunk=lambda x: False)
1298
return ''.join(difflines)
1300
def assertEqual(self, a, b, message=''):
1304
except UnicodeError as e:
1305
# If we can't compare without getting a UnicodeError, then
1306
# obviously they are different
1307
trace.mutter('UnicodeError: %s', e)
1310
raise AssertionError("%snot equal:\na = %s\nb = %s\n"
1312
pprint.pformat(a), pprint.pformat(b)))
1314
# FIXME: This is deprecated in unittest2 but plugins may still use it so we
1315
# need a deprecation period for them -- vila 2016-02-01
1316
assertEquals = assertEqual
1318
def assertEqualDiff(self, a, b, message=None):
1319
"""Assert two texts are equal, if not raise an exception.
1321
This is intended for use with multi-line strings where it can
1322
be hard to find the differences by eye.
1324
# TODO: perhaps override assertEqual to call this for strings?
1328
message = "texts not equal:\n"
1330
message = 'first string is missing a final newline.\n'
1332
message = 'second string is missing a final newline.\n'
1333
raise AssertionError(message +
1334
self._ndiff_strings(a, b))
1336
def assertEqualMode(self, mode, mode_test):
1337
self.assertEqual(mode, mode_test,
1338
'mode mismatch %o != %o' % (mode, mode_test))
1340
def assertEqualStat(self, expected, actual):
1341
"""assert that expected and actual are the same stat result.
1343
:param expected: A stat result.
1344
:param actual: A stat result.
1345
:raises AssertionError: If the expected and actual stat values differ
1346
other than by atime.
1348
self.assertEqual(expected.st_size, actual.st_size,
1349
'st_size did not match')
1350
self.assertEqual(expected.st_mtime, actual.st_mtime,
1351
'st_mtime did not match')
1352
self.assertEqual(expected.st_ctime, actual.st_ctime,
1353
'st_ctime did not match')
1354
if sys.platform == 'win32':
1355
# On Win32 both 'dev' and 'ino' cannot be trusted. In python2.4 it
1356
# is 'dev' that varies, in python 2.5 (6?) it is st_ino that is
1357
# odd. We just force it to always be 0 to avoid any problems.
1358
self.assertEqual(0, expected.st_dev)
1359
self.assertEqual(0, actual.st_dev)
1360
self.assertEqual(0, expected.st_ino)
1361
self.assertEqual(0, actual.st_ino)
1363
self.assertEqual(expected.st_dev, actual.st_dev,
1364
'st_dev did not match')
1365
self.assertEqual(expected.st_ino, actual.st_ino,
1366
'st_ino did not match')
1367
self.assertEqual(expected.st_mode, actual.st_mode,
1368
'st_mode did not match')
1370
def assertLength(self, length, obj_with_len):
1371
"""Assert that obj_with_len is of length length."""
1372
if len(obj_with_len) != length:
1373
self.fail("Incorrect length: wanted %d, got %d for %r" % (
1374
length, len(obj_with_len), obj_with_len))
1376
def assertLogsError(self, exception_class, func, *args, **kwargs):
1377
"""Assert that `func(*args, **kwargs)` quietly logs a specific error.
1380
orig_log_exception_quietly = trace.log_exception_quietly
1383
orig_log_exception_quietly()
1384
captured.append(sys.exc_info()[1])
1385
trace.log_exception_quietly = capture
1386
func(*args, **kwargs)
1388
trace.log_exception_quietly = orig_log_exception_quietly
1389
self.assertLength(1, captured)
1391
self.assertIsInstance(err, exception_class)
1394
def assertPositive(self, val):
1395
"""Assert that val is greater than 0."""
1396
self.assertTrue(val > 0, 'expected a positive value, but got %s' % val)
1398
def assertNegative(self, val):
1399
"""Assert that val is less than 0."""
1400
self.assertTrue(val < 0, 'expected a negative value, but got %s' % val)
1402
def assertStartsWith(self, s, prefix):
1403
if not s.startswith(prefix):
1404
raise AssertionError('string %r does not start with %r' % (s, prefix))
1406
def assertEndsWith(self, s, suffix):
1407
"""Asserts that s ends with suffix."""
1408
if not s.endswith(suffix):
1409
raise AssertionError('string %r does not end with %r' % (s, suffix))
1411
def assertContainsRe(self, haystack, needle_re, flags=0):
1412
"""Assert that a contains something matching a regular expression."""
1413
if not re.search(needle_re, haystack, flags):
1414
if '\n' in haystack or len(haystack) > 60:
1415
# a long string, format it in a more readable way
1416
raise AssertionError(
1417
'pattern "%s" not found in\n"""\\\n%s"""\n'
1418
% (needle_re, haystack))
1420
raise AssertionError('pattern "%s" not found in "%s"'
1421
% (needle_re, haystack))
1423
def assertNotContainsRe(self, haystack, needle_re, flags=0):
1424
"""Assert that a does not match a regular expression"""
1425
if re.search(needle_re, haystack, flags):
1426
raise AssertionError('pattern "%s" found in "%s"'
1427
% (needle_re, haystack))
1429
def assertContainsString(self, haystack, needle):
1430
if haystack.find(needle) == -1:
1431
self.fail("string %r not found in '''%s'''" % (needle, haystack))
1433
def assertNotContainsString(self, haystack, needle):
1434
if haystack.find(needle) != -1:
1435
self.fail("string %r found in '''%s'''" % (needle, haystack))
1437
def assertSubset(self, sublist, superlist):
1438
"""Assert that every entry in sublist is present in superlist."""
1439
missing = set(sublist) - set(superlist)
1440
if len(missing) > 0:
1441
raise AssertionError("value(s) %r not present in container %r" %
1442
(missing, superlist))
1444
def assertListRaises(self, excClass, func, *args, **kwargs):
1445
"""Fail unless excClass is raised when the iterator from func is used.
1447
Many functions can return generators this makes sure
1448
to wrap them in a list() call to make sure the whole generator
1449
is run, and that the proper exception is raised.
1452
list(func(*args, **kwargs))
1453
except excClass as e:
1456
if getattr(excClass, '__name__', None) is not None:
1457
excName = excClass.__name__
1459
excName = str(excClass)
1460
raise self.failureException("%s not raised" % excName)
1462
def assertRaises(self, excClass, callableObj, *args, **kwargs):
1463
"""Assert that a callable raises a particular exception.
1465
:param excClass: As for the except statement, this may be either an
1466
exception class, or a tuple of classes.
1467
:param callableObj: A callable, will be passed ``*args`` and
1470
Returns the exception so that you can examine it.
1473
callableObj(*args, **kwargs)
1474
except excClass as e:
1477
if getattr(excClass, '__name__', None) is not None:
1478
excName = excClass.__name__
1481
excName = str(excClass)
1482
raise self.failureException("%s not raised" % excName)
1484
def assertIs(self, left, right, message=None):
1485
if not (left is right):
1486
if message is not None:
1487
raise AssertionError(message)
1489
raise AssertionError("%r is not %r." % (left, right))
1491
def assertIsNot(self, left, right, message=None):
1493
if message is not None:
1494
raise AssertionError(message)
1496
raise AssertionError("%r is %r." % (left, right))
1498
def assertTransportMode(self, transport, path, mode):
1499
"""Fail if a path does not have mode "mode".
1501
If modes are not supported on this transport, the assertion is ignored.
1503
if not transport._can_roundtrip_unix_modebits():
1505
path_stat = transport.stat(path)
1506
actual_mode = stat.S_IMODE(path_stat.st_mode)
1507
self.assertEqual(mode, actual_mode,
1508
'mode of %r incorrect (%s != %s)'
1509
% (path, oct(mode), oct(actual_mode)))
1511
def assertIsSameRealPath(self, path1, path2):
1512
"""Fail if path1 and path2 points to different files"""
1513
self.assertEqual(osutils.realpath(path1),
1514
osutils.realpath(path2),
1515
"apparent paths:\na = %s\nb = %s\n," % (path1, path2))
1517
def assertIsInstance(self, obj, kls, msg=None):
1518
"""Fail if obj is not an instance of kls
1520
:param msg: Supplementary message to show if the assertion fails.
1522
if not isinstance(obj, kls):
1523
m = "%r is an instance of %s rather than %s" % (
1524
obj, obj.__class__, kls)
1529
def assertFileEqual(self, content, path):
1530
"""Fail if path does not contain 'content'."""
1531
self.assertPathExists(path)
1532
with open(path, 'rb') as f:
1534
self.assertEqualDiff(content, s)
1536
def assertDocstring(self, expected_docstring, obj):
1537
"""Fail if obj does not have expected_docstring"""
1539
# With -OO the docstring should be None instead
1540
self.assertIs(obj.__doc__, None)
1542
self.assertEqual(expected_docstring, obj.__doc__)
1544
def assertPathExists(self, path):
1545
"""Fail unless path or paths, which may be abs or relative, exist."""
1546
if not isinstance(path, (str, text_type)):
1548
self.assertPathExists(p)
1550
self.assertTrue(osutils.lexists(path),
1551
path + " does not exist")
1553
def assertPathDoesNotExist(self, path):
1554
"""Fail if path or paths, which may be abs or relative, exist."""
1555
if not isinstance(path, (str, text_type)):
1557
self.assertPathDoesNotExist(p)
1559
self.assertFalse(osutils.lexists(path),
1562
def _capture_deprecation_warnings(self, a_callable, *args, **kwargs):
1563
"""A helper for callDeprecated and applyDeprecated.
1565
:param a_callable: A callable to call.
1566
:param args: The positional arguments for the callable
1567
:param kwargs: The keyword arguments for the callable
1568
:return: A tuple (warnings, result). result is the result of calling
1569
a_callable(``*args``, ``**kwargs``).
1572
def capture_warnings(msg, cls=None, stacklevel=None):
1573
# we've hooked into a deprecation specific callpath,
1574
# only deprecations should getting sent via it.
1575
self.assertEqual(cls, DeprecationWarning)
1576
local_warnings.append(msg)
1577
original_warning_method = symbol_versioning.warn
1578
symbol_versioning.set_warning_method(capture_warnings)
1580
result = a_callable(*args, **kwargs)
1582
symbol_versioning.set_warning_method(original_warning_method)
1583
return (local_warnings, result)
1585
def applyDeprecated(self, deprecation_format, a_callable, *args, **kwargs):
1586
"""Call a deprecated callable without warning the user.
1588
Note that this only captures warnings raised by symbol_versioning.warn,
1589
not other callers that go direct to the warning module.
1591
To test that a deprecated method raises an error, do something like
1592
this (remember that both assertRaises and applyDeprecated delays *args
1593
and **kwargs passing)::
1595
self.assertRaises(errors.ReservedId,
1596
self.applyDeprecated,
1597
deprecated_in((1, 5, 0)),
1601
:param deprecation_format: The deprecation format that the callable
1602
should have been deprecated with. This is the same type as the
1603
parameter to deprecated_method/deprecated_function. If the
1604
callable is not deprecated with this format, an assertion error
1606
:param a_callable: A callable to call. This may be a bound method or
1607
a regular function. It will be called with ``*args`` and
1609
:param args: The positional arguments for the callable
1610
:param kwargs: The keyword arguments for the callable
1611
:return: The result of a_callable(``*args``, ``**kwargs``)
1613
call_warnings, result = self._capture_deprecation_warnings(a_callable,
1615
expected_first_warning = symbol_versioning.deprecation_string(
1616
a_callable, deprecation_format)
1617
if len(call_warnings) == 0:
1618
self.fail("No deprecation warning generated by call to %s" %
1620
self.assertEqual(expected_first_warning, call_warnings[0])
1623
def callCatchWarnings(self, fn, *args, **kw):
1624
"""Call a callable that raises python warnings.
1626
The caller's responsible for examining the returned warnings.
1628
If the callable raises an exception, the exception is not
1629
caught and propagates up to the caller. In that case, the list
1630
of warnings is not available.
1632
:returns: ([warning_object, ...], fn_result)
1634
# XXX: This is not perfect, because it completely overrides the
1635
# warnings filters, and some code may depend on suppressing particular
1636
# warnings. It's the easiest way to insulate ourselves from -Werror,
1637
# though. -- Andrew, 20071062
1639
def _catcher(message, category, filename, lineno, file=None, line=None):
1640
# despite the name, 'message' is normally(?) a Warning subclass
1642
wlist.append(message)
1643
saved_showwarning = warnings.showwarning
1644
saved_filters = warnings.filters
1646
warnings.showwarning = _catcher
1647
warnings.filters = []
1648
result = fn(*args, **kw)
1650
warnings.showwarning = saved_showwarning
1651
warnings.filters = saved_filters
1652
return wlist, result
1654
def callDeprecated(self, expected, callable, *args, **kwargs):
1655
"""Assert that a callable is deprecated in a particular way.
1657
This is a very precise test for unusual requirements. The
1658
applyDeprecated helper function is probably more suited for most tests
1659
as it allows you to simply specify the deprecation format being used
1660
and will ensure that that is issued for the function being called.
1662
Note that this only captures warnings raised by symbol_versioning.warn,
1663
not other callers that go direct to the warning module. To catch
1664
general warnings, use callCatchWarnings.
1666
:param expected: a list of the deprecation warnings expected, in order
1667
:param callable: The callable to call
1668
:param args: The positional arguments for the callable
1669
:param kwargs: The keyword arguments for the callable
1671
call_warnings, result = self._capture_deprecation_warnings(callable,
1673
self.assertEqual(expected, call_warnings)
1676
def _startLogFile(self):
1677
"""Setup a in-memory target for bzr and testcase log messages"""
1678
pseudo_log_file = BytesIO()
1679
def _get_log_contents_for_weird_testtools_api():
1680
return [pseudo_log_file.getvalue().decode(
1681
"utf-8", "replace").encode("utf-8")]
1682
self.addDetail("log", content.Content(content.ContentType("text",
1683
"plain", {"charset": "utf8"}),
1684
_get_log_contents_for_weird_testtools_api))
1685
self._log_file = pseudo_log_file
1686
self._log_memento = trace.push_log_file(self._log_file)
1687
self.addCleanup(self._finishLogFile)
1689
def _finishLogFile(self):
1690
"""Flush and dereference the in-memory log for this testcase"""
1691
if trace._trace_file:
1692
# flush the log file, to get all content
1693
trace._trace_file.flush()
1694
trace.pop_log_file(self._log_memento)
1695
# The logging module now tracks references for cleanup so discard ours
1696
del self._log_memento
1698
def thisFailsStrictLockCheck(self):
1699
"""It is known that this test would fail with -Dstrict_locks.
1701
By default, all tests are run with strict lock checking unless
1702
-Edisable_lock_checks is supplied. However there are some tests which
1703
we know fail strict locks at this point that have not been fixed.
1704
They should call this function to disable the strict checking.
1706
This should be used sparingly, it is much better to fix the locking
1707
issues rather than papering over the problem by calling this function.
1709
debug.debug_flags.discard('strict_locks')
1711
def overrideAttr(self, obj, attr_name, new=_unitialized_attr):
1712
"""Overrides an object attribute restoring it after the test.
1714
:note: This should be used with discretion; you should think about
1715
whether it's better to make the code testable without monkey-patching.
1717
:param obj: The object that will be mutated.
1719
:param attr_name: The attribute name we want to preserve/override in
1722
:param new: The optional value we want to set the attribute to.
1724
:returns: The actual attr value.
1726
# The actual value is captured by the call below
1727
value = getattr(obj, attr_name, _unitialized_attr)
1728
if value is _unitialized_attr:
1729
# When the test completes, the attribute should not exist, but if
1730
# we aren't setting a value, we don't need to do anything.
1731
if new is not _unitialized_attr:
1732
self.addCleanup(delattr, obj, attr_name)
1734
self.addCleanup(setattr, obj, attr_name, value)
1735
if new is not _unitialized_attr:
1736
setattr(obj, attr_name, new)
1739
def overrideEnv(self, name, new):
1740
"""Set an environment variable, and reset it after the test.
1742
:param name: The environment variable name.
1744
:param new: The value to set the variable to. If None, the
1745
variable is deleted from the environment.
1747
:returns: The actual variable value.
1749
value = osutils.set_or_unset_env(name, new)
1750
self.addCleanup(osutils.set_or_unset_env, name, value)
1753
def recordCalls(self, obj, attr_name):
1754
"""Monkeypatch in a wrapper that will record calls.
1756
The monkeypatch is automatically removed when the test concludes.
1758
:param obj: The namespace holding the reference to be replaced;
1759
typically a module, class, or object.
1760
:param attr_name: A string for the name of the attribute to
1762
:returns: A list that will be extended with one item every time the
1763
function is called, with a tuple of (args, kwargs).
1767
def decorator(*args, **kwargs):
1768
calls.append((args, kwargs))
1769
return orig(*args, **kwargs)
1770
orig = self.overrideAttr(obj, attr_name, decorator)
1773
def _cleanEnvironment(self):
1774
for name, value in isolated_environ.items():
1775
self.overrideEnv(name, value)
1777
def _restoreHooks(self):
1778
for klass, (name, hooks) in self._preserved_hooks.items():
1779
setattr(klass, name, hooks)
1780
self._preserved_hooks.clear()
1781
breezy.hooks._lazy_hooks = self._preserved_lazy_hooks
1782
self._preserved_lazy_hooks.clear()
1784
def knownFailure(self, reason):
1785
"""Declare that this test fails for a known reason
1787
Tests that are known to fail should generally be using expectedFailure
1788
with an appropriate reverse assertion if a change could cause the test
1789
to start passing. Conversely if the test has no immediate prospect of
1790
succeeding then using skip is more suitable.
1792
When this method is called while an exception is being handled, that
1793
traceback will be used, otherwise a new exception will be thrown to
1794
provide one but won't be reported.
1796
self._add_reason(reason)
1798
exc_info = sys.exc_info()
1799
if exc_info != (None, None, None):
1800
self._report_traceback(exc_info)
1803
raise self.failureException(reason)
1804
except self.failureException:
1805
exc_info = sys.exc_info()
1806
# GZ 02-08-2011: Maybe cleanup this err.exc_info attribute too?
1807
raise testtools.testcase._ExpectedFailure(exc_info)
1811
def _suppress_log(self):
1812
"""Remove the log info from details."""
1813
self.discardDetail('log')
1815
def _do_skip(self, result, reason):
1816
self._suppress_log()
1817
addSkip = getattr(result, 'addSkip', None)
1818
if not callable(addSkip):
1819
result.addSuccess(result)
1821
addSkip(self, str(reason))
1824
def _do_known_failure(self, result, e):
1825
self._suppress_log()
1826
err = sys.exc_info()
1827
addExpectedFailure = getattr(result, 'addExpectedFailure', None)
1828
if addExpectedFailure is not None:
1829
addExpectedFailure(self, err)
1831
result.addSuccess(self)
1834
def _do_not_applicable(self, result, e):
1836
reason = 'No reason given'
1839
self._suppress_log ()
1840
addNotApplicable = getattr(result, 'addNotApplicable', None)
1841
if addNotApplicable is not None:
1842
result.addNotApplicable(self, reason)
1844
self._do_skip(result, reason)
1847
def _report_skip(self, result, err):
1848
"""Override the default _report_skip.
1850
We want to strip the 'log' detail. If we waint until _do_skip, it has
1851
already been formatted into the 'reason' string, and we can't pull it
1854
self._suppress_log()
1855
super(TestCase, self)._report_skip(self, result, err)
1858
def _report_expected_failure(self, result, err):
1861
See _report_skip for motivation.
1863
self._suppress_log()
1864
super(TestCase, self)._report_expected_failure(self, result, err)
1867
def _do_unsupported_or_skip(self, result, e):
1869
self._suppress_log()
1870
addNotSupported = getattr(result, 'addNotSupported', None)
1871
if addNotSupported is not None:
1872
result.addNotSupported(self, reason)
1874
self._do_skip(result, reason)
1876
def time(self, callable, *args, **kwargs):
1877
"""Run callable and accrue the time it takes to the benchmark time.
1879
If lsprofiling is enabled (i.e. by --lsprof-time to brz selftest) then
1880
this will cause lsprofile statistics to be gathered and stored in
1883
if self._benchtime is None:
1884
self.addDetail('benchtime', content.Content(content.UTF8_TEXT,
1885
lambda:[str(self._benchtime).encode('utf-8')]))
1889
if not self._gather_lsprof_in_benchmarks:
1890
return callable(*args, **kwargs)
1892
# record this benchmark
1893
ret, stats = breezy.lsprof.profile(callable, *args, **kwargs)
1895
self._benchcalls.append(((callable, args, kwargs), stats))
1898
self._benchtime += time.time() - start
1900
def log(self, *args):
1904
"""Get a unicode string containing the log from breezy.trace.
1906
Undecodable characters are replaced.
1908
return u"".join(self.getDetails()['log'].iter_text())
1910
def requireFeature(self, feature):
1911
"""This test requires a specific feature is available.
1913
:raises UnavailableFeature: When feature is not available.
1915
if not feature.available():
1916
raise UnavailableFeature(feature)
1918
def _run_bzr_autosplit(self, args, retcode, encoding, stdin,
1920
"""Run bazaar command line, splitting up a string command line."""
1921
if isinstance(args, string_types):
1922
args = shlex.split(args)
1923
return self._run_bzr_core(args, retcode=retcode,
1924
encoding=encoding, stdin=stdin, working_dir=working_dir,
1927
def _run_bzr_core(self, args, retcode, encoding, stdin,
1929
# Clear chk_map page cache, because the contents are likely to mask
1931
chk_map.clear_cache()
1932
if encoding is None:
1933
encoding = osutils.get_user_encoding()
1935
self.log('run brz: %r', args)
1937
stdout = ui_testing.BytesIOWithEncoding()
1938
stderr = ui_testing.BytesIOWithEncoding()
1939
stdout.encoding = stderr.encoding = encoding
1941
# FIXME: don't call into logging here
1942
handler = trace.EncodedStreamHandler(
1943
stderr, errors="replace", level=logging.INFO)
1944
logger = logging.getLogger('')
1945
logger.addHandler(handler)
1947
self._last_cmd_stdout = codecs.getwriter(encoding)(stdout)
1948
self._last_cmd_stderr = codecs.getwriter(encoding)(stderr)
1950
old_ui_factory = ui.ui_factory
1951
ui.ui_factory = ui_testing.TestUIFactory(
1953
stdout=self._last_cmd_stdout,
1954
stderr=self._last_cmd_stderr)
1957
if working_dir is not None:
1958
cwd = osutils.getcwd()
1959
os.chdir(working_dir)
1962
result = self.apply_redirected(
1963
ui.ui_factory.stdin,
1965
_mod_commands.run_bzr_catch_user_errors,
1968
logger.removeHandler(handler)
1969
ui.ui_factory = old_ui_factory
1973
out = stdout.getvalue()
1974
err = stderr.getvalue()
1976
self.log('output:\n%r', out)
1978
self.log('errors:\n%r', err)
1979
if retcode is not None:
1980
self.assertEqual(retcode, result,
1981
message='Unexpected return code')
1982
return result, out, err
1984
def run_bzr(self, args, retcode=0, stdin=None, encoding=None,
1985
working_dir=None, error_regexes=[]):
1986
"""Invoke brz, as if it were run from the command line.
1988
The argument list should not include the brz program name - the
1989
first argument is normally the brz command. Arguments may be
1990
passed in three ways:
1992
1- A list of strings, eg ["commit", "a"]. This is recommended
1993
when the command contains whitespace or metacharacters, or
1994
is built up at run time.
1996
2- A single string, eg "add a". This is the most convenient
1997
for hardcoded commands.
1999
This runs brz through the interface that catches and reports
2000
errors, and with logging set to something approximating the
2001
default, so that error reporting can be checked.
2003
This should be the main method for tests that want to exercise the
2004
overall behavior of the brz application (rather than a unit test
2005
or a functional test of the library.)
2007
This sends the stdout/stderr results into the test's log,
2008
where it may be useful for debugging. See also run_captured.
2010
:keyword stdin: A string to be used as stdin for the command.
2011
:keyword retcode: The status code the command should return;
2013
:keyword working_dir: The directory to run the command in
2014
:keyword error_regexes: A list of expected error messages. If
2015
specified they must be seen in the error output of the command.
2017
retcode, out, err = self._run_bzr_autosplit(
2022
working_dir=working_dir,
2024
self.assertIsInstance(error_regexes, (list, tuple))
2025
for regex in error_regexes:
2026
self.assertContainsRe(err, regex)
2029
def run_bzr_error(self, error_regexes, *args, **kwargs):
2030
"""Run brz, and check that stderr contains the supplied regexes
2032
:param error_regexes: Sequence of regular expressions which
2033
must each be found in the error output. The relative ordering
2035
:param args: command-line arguments for brz
2036
:param kwargs: Keyword arguments which are interpreted by run_brz
2037
This function changes the default value of retcode to be 3,
2038
since in most cases this is run when you expect brz to fail.
2040
:return: (out, err) The actual output of running the command (in case
2041
you want to do more inspection)
2045
# Make sure that commit is failing because there is nothing to do
2046
self.run_bzr_error(['no changes to commit'],
2047
['commit', '-m', 'my commit comment'])
2048
# Make sure --strict is handling an unknown file, rather than
2049
# giving us the 'nothing to do' error
2050
self.build_tree(['unknown'])
2051
self.run_bzr_error(['Commit refused because there are unknown files'],
2052
['commit', --strict', '-m', 'my commit comment'])
2054
kwargs.setdefault('retcode', 3)
2055
kwargs['error_regexes'] = error_regexes
2056
out, err = self.run_bzr(*args, **kwargs)
2059
def run_bzr_subprocess(self, *args, **kwargs):
2060
"""Run brz in a subprocess for testing.
2062
This starts a new Python interpreter and runs brz in there.
2063
This should only be used for tests that have a justifiable need for
2064
this isolation: e.g. they are testing startup time, or signal
2065
handling, or early startup code, etc. Subprocess code can't be
2066
profiled or debugged so easily.
2068
:keyword retcode: The status code that is expected. Defaults to 0. If
2069
None is supplied, the status code is not checked.
2070
:keyword env_changes: A dictionary which lists changes to environment
2071
variables. A value of None will unset the env variable.
2072
The values must be strings. The change will only occur in the
2073
child, so you don't need to fix the environment after running.
2074
:keyword universal_newlines: Convert CRLF => LF
2075
:keyword allow_plugins: By default the subprocess is run with
2076
--no-plugins to ensure test reproducibility. Also, it is possible
2077
for system-wide plugins to create unexpected output on stderr,
2078
which can cause unnecessary test failures.
2080
env_changes = kwargs.get('env_changes', {})
2081
working_dir = kwargs.get('working_dir', None)
2082
allow_plugins = kwargs.get('allow_plugins', False)
2084
if isinstance(args[0], list):
2086
elif isinstance(args[0], (str, text_type)):
2087
args = list(shlex.split(args[0]))
2089
raise ValueError("passing varargs to run_bzr_subprocess")
2090
process = self.start_bzr_subprocess(args, env_changes=env_changes,
2091
working_dir=working_dir,
2092
allow_plugins=allow_plugins)
2093
# We distinguish between retcode=None and retcode not passed.
2094
supplied_retcode = kwargs.get('retcode', 0)
2095
return self.finish_bzr_subprocess(process, retcode=supplied_retcode,
2096
universal_newlines=kwargs.get('universal_newlines', False),
2099
def start_bzr_subprocess(self, process_args, env_changes=None,
2100
skip_if_plan_to_signal=False,
2102
allow_plugins=False, stderr=subprocess.PIPE):
2103
"""Start brz in a subprocess for testing.
2105
This starts a new Python interpreter and runs brz in there.
2106
This should only be used for tests that have a justifiable need for
2107
this isolation: e.g. they are testing startup time, or signal
2108
handling, or early startup code, etc. Subprocess code can't be
2109
profiled or debugged so easily.
2111
:param process_args: a list of arguments to pass to the brz executable,
2112
for example ``['--version']``.
2113
:param env_changes: A dictionary which lists changes to environment
2114
variables. A value of None will unset the env variable.
2115
The values must be strings. The change will only occur in the
2116
child, so you don't need to fix the environment after running.
2117
:param skip_if_plan_to_signal: raise TestSkipped when true and system
2118
doesn't support signalling subprocesses.
2119
:param allow_plugins: If False (default) pass --no-plugins to brz.
2120
:param stderr: file to use for the subprocess's stderr. Valid values
2121
are those valid for the stderr argument of `subprocess.Popen`.
2122
Default value is ``subprocess.PIPE``.
2124
:returns: Popen object for the started process.
2126
if skip_if_plan_to_signal:
2127
if os.name != "posix":
2128
raise TestSkipped("Sending signals not supported")
2130
if env_changes is None:
2132
# Because $HOME is set to a tempdir for the context of a test, modules
2133
# installed in the user dir will not be found unless $PYTHONUSERBASE
2134
# gets set to the computed directory of this parent process.
2135
if site.USER_BASE is not None:
2136
env_changes["PYTHONUSERBASE"] = site.USER_BASE
2139
def cleanup_environment():
2140
for env_var, value in env_changes.items():
2141
old_env[env_var] = osutils.set_or_unset_env(env_var, value)
2143
def restore_environment():
2144
for env_var, value in old_env.items():
2145
osutils.set_or_unset_env(env_var, value)
2147
bzr_path = self.get_brz_path()
2150
if working_dir is not None:
2151
cwd = osutils.getcwd()
2152
os.chdir(working_dir)
2155
# win32 subprocess doesn't support preexec_fn
2156
# so we will avoid using it on all platforms, just to
2157
# make sure the code path is used, and we don't break on win32
2158
cleanup_environment()
2159
# Include the subprocess's log file in the test details, in case
2160
# the test fails due to an error in the subprocess.
2161
self._add_subprocess_log(trace._get_brz_log_filename())
2162
command = [sys.executable]
2163
# frozen executables don't need the path to bzr
2164
if getattr(sys, "frozen", None) is None:
2165
command.append(bzr_path)
2166
if not allow_plugins:
2167
command.append('--no-plugins')
2168
command.extend(process_args)
2169
process = self._popen(command, stdin=subprocess.PIPE,
2170
stdout=subprocess.PIPE,
2173
restore_environment()
2179
def _add_subprocess_log(self, log_file_path):
2180
if len(self._log_files) == 0:
2181
# Register an addCleanup func. We do this on the first call to
2182
# _add_subprocess_log rather than in TestCase.setUp so that this
2183
# addCleanup is registered after any cleanups for tempdirs that
2184
# subclasses might create, which will probably remove the log file
2186
self.addCleanup(self._subprocess_log_cleanup)
2187
# self._log_files is a set, so if a log file is reused we won't grab it
2189
self._log_files.add(log_file_path)
2191
def _subprocess_log_cleanup(self):
2192
for count, log_file_path in enumerate(self._log_files):
2193
# We use buffer_now=True to avoid holding the file open beyond
2194
# the life of this function, which might interfere with e.g.
2195
# cleaning tempdirs on Windows.
2196
# XXX: Testtools 0.9.5 doesn't have the content_from_file helper
2197
#detail_content = content.content_from_file(
2198
# log_file_path, buffer_now=True)
2199
with open(log_file_path, 'rb') as log_file:
2200
log_file_bytes = log_file.read()
2201
detail_content = content.Content(content.ContentType("text",
2202
"plain", {"charset": "utf8"}), lambda: [log_file_bytes])
2203
self.addDetail("start_bzr_subprocess-log-%d" % (count,),
2206
def _popen(self, *args, **kwargs):
2207
"""Place a call to Popen.
2209
Allows tests to override this method to intercept the calls made to
2210
Popen for introspection.
2212
return subprocess.Popen(*args, **kwargs)
2214
def get_source_path(self):
2215
"""Return the path of the directory containing breezy."""
2216
return os.path.dirname(os.path.dirname(breezy.__file__))
2218
def get_brz_path(self):
2219
"""Return the path of the 'brz' executable for this test suite."""
2220
brz_path = os.path.join(self.get_source_path(), "brz")
2221
if not os.path.isfile(brz_path):
2222
# We are probably installed. Assume sys.argv is the right file
2223
brz_path = sys.argv[0]
2226
def finish_bzr_subprocess(self, process, retcode=0, send_signal=None,
2227
universal_newlines=False, process_args=None):
2228
"""Finish the execution of process.
2230
:param process: the Popen object returned from start_bzr_subprocess.
2231
:param retcode: The status code that is expected. Defaults to 0. If
2232
None is supplied, the status code is not checked.
2233
:param send_signal: an optional signal to send to the process.
2234
:param universal_newlines: Convert CRLF => LF
2235
:returns: (stdout, stderr)
2237
if send_signal is not None:
2238
os.kill(process.pid, send_signal)
2239
out, err = process.communicate()
2241
if universal_newlines:
2242
out = out.replace('\r\n', '\n')
2243
err = err.replace('\r\n', '\n')
2245
if retcode is not None and retcode != process.returncode:
2246
if process_args is None:
2247
process_args = "(unknown args)"
2248
trace.mutter('Output of brz %r:\n%s', process_args, out)
2249
trace.mutter('Error for brz %r:\n%s', process_args, err)
2250
self.fail('Command brz %r failed with retcode %d != %d'
2251
% (process_args, retcode, process.returncode))
2254
def check_tree_shape(self, tree, shape):
2255
"""Compare a tree to a list of expected names.
2257
Fail if they are not precisely equal.
2260
shape = list(shape) # copy
2261
for path, ie in tree.iter_entries_by_dir():
2262
name = path.replace('\\', '/')
2263
if ie.kind == 'directory':
2266
pass # ignore root entry
2272
self.fail("expected paths not found in inventory: %r" % shape)
2274
self.fail("unexpected paths found in inventory: %r" % extras)
2276
def apply_redirected(self, stdin=None, stdout=None, stderr=None,
2277
a_callable=None, *args, **kwargs):
2278
"""Call callable with redirected std io pipes.
2280
Returns the return code."""
2281
if not callable(a_callable):
2282
raise ValueError("a_callable must be callable.")
2284
stdin = BytesIO(b"")
2286
if getattr(self, "_log_file", None) is not None:
2287
stdout = self._log_file
2291
if getattr(self, "_log_file", None is not None):
2292
stderr = self._log_file
2295
real_stdin = sys.stdin
2296
real_stdout = sys.stdout
2297
real_stderr = sys.stderr
2302
return a_callable(*args, **kwargs)
2304
sys.stdout = real_stdout
2305
sys.stderr = real_stderr
2306
sys.stdin = real_stdin
2308
def reduceLockdirTimeout(self):
2309
"""Reduce the default lock timeout for the duration of the test, so that
2310
if LockContention occurs during a test, it does so quickly.
2312
Tests that expect to provoke LockContention errors should call this.
2314
self.overrideAttr(lockdir, '_DEFAULT_TIMEOUT_SECONDS', 0)
2316
def make_utf8_encoded_stringio(self, encoding_type=None):
2317
"""Return a wrapped BytesIO, that will encode text input to UTF-8."""
2318
if encoding_type is None:
2319
encoding_type = 'strict'
2321
output_encoding = 'utf-8'
2322
sio = codecs.getwriter(output_encoding)(bio, errors=encoding_type)
2323
sio.encoding = output_encoding
2326
def disable_verb(self, verb):
2327
"""Disable a smart server verb for one test."""
2328
from breezy.bzr.smart import request
2329
request_handlers = request.request_handlers
2330
orig_method = request_handlers.get(verb)
2331
orig_info = request_handlers.get_info(verb)
2332
request_handlers.remove(verb)
2333
self.addCleanup(request_handlers.register, verb, orig_method,
2340
class CapturedCall(object):
2341
"""A helper for capturing smart server calls for easy debug analysis."""
2343
def __init__(self, params, prefix_length):
2344
"""Capture the call with params and skip prefix_length stack frames."""
2347
# The last 5 frames are the __init__, the hook frame, and 3 smart
2348
# client frames. Beyond this we could get more clever, but this is good
2350
stack = traceback.extract_stack()[prefix_length:-5]
2351
self.stack = ''.join(traceback.format_list(stack))
2354
return self.call.method
2357
return self.call.method
2363
class TestCaseWithMemoryTransport(TestCase):
2364
"""Common test class for tests that do not need disk resources.
2366
Tests that need disk resources should derive from TestCaseInTempDir
2367
orTestCaseWithTransport.
2369
TestCaseWithMemoryTransport sets the TEST_ROOT variable for all brz tests.
2371
For TestCaseWithMemoryTransport the ``test_home_dir`` is set to the name of
2372
a directory which does not exist. This serves to help ensure test isolation
2373
is preserved. ``test_dir`` is set to the TEST_ROOT, as is cwd, because they
2374
must exist. However, TestCaseWithMemoryTransport does not offer local file
2375
defaults for the transport in tests, nor does it obey the command line
2376
override, so tests that accidentally write to the common directory should
2379
:cvar TEST_ROOT: Directory containing all temporary directories, plus a
2380
``.bzr`` directory that stops us ascending higher into the filesystem.
2386
def __init__(self, methodName='runTest'):
2387
# allow test parameterization after test construction and before test
2388
# execution. Variables that the parameterizer sets need to be
2389
# ones that are not set by setUp, or setUp will trash them.
2390
super(TestCaseWithMemoryTransport, self).__init__(methodName)
2391
self.vfs_transport_factory = default_transport
2392
self.transport_server = None
2393
self.transport_readonly_server = None
2394
self.__vfs_server = None
2397
super(TestCaseWithMemoryTransport, self).setUp()
2399
def _add_disconnect_cleanup(transport):
2400
"""Schedule disconnection of given transport at test cleanup
2402
This needs to happen for all connected transports or leaks occur.
2404
Note reconnections may mean we call disconnect multiple times per
2405
transport which is suboptimal but seems harmless.
2407
self.addCleanup(transport.disconnect)
2409
_mod_transport.Transport.hooks.install_named_hook('post_connect',
2410
_add_disconnect_cleanup, None)
2412
self._make_test_root()
2413
self.addCleanup(os.chdir, osutils.getcwd())
2414
self.makeAndChdirToTestDir()
2415
self.overrideEnvironmentForTesting()
2416
self.__readonly_server = None
2417
self.__server = None
2418
self.reduceLockdirTimeout()
2419
# Each test may use its own config files even if the local config files
2420
# don't actually exist. They'll rightly fail if they try to create them
2422
self.overrideAttr(config, '_shared_stores', {})
2424
def get_transport(self, relpath=None):
2425
"""Return a writeable transport.
2427
This transport is for the test scratch space relative to
2430
:param relpath: a path relative to the base url.
2432
t = _mod_transport.get_transport_from_url(self.get_url(relpath))
2433
self.assertFalse(t.is_readonly())
2436
def get_readonly_transport(self, relpath=None):
2437
"""Return a readonly transport for the test scratch space
2439
This can be used to test that operations which should only need
2440
readonly access in fact do not try to write.
2442
:param relpath: a path relative to the base url.
2444
t = _mod_transport.get_transport_from_url(
2445
self.get_readonly_url(relpath))
2446
self.assertTrue(t.is_readonly())
2449
def create_transport_readonly_server(self):
2450
"""Create a transport server from class defined at init.
2452
This is mostly a hook for daughter classes.
2454
return self.transport_readonly_server()
2456
def get_readonly_server(self):
2457
"""Get the server instance for the readonly transport
2459
This is useful for some tests with specific servers to do diagnostics.
2461
if self.__readonly_server is None:
2462
if self.transport_readonly_server is None:
2463
# readonly decorator requested
2464
self.__readonly_server = test_server.ReadonlyServer()
2466
# explicit readonly transport.
2467
self.__readonly_server = self.create_transport_readonly_server()
2468
self.start_server(self.__readonly_server,
2469
self.get_vfs_only_server())
2470
return self.__readonly_server
2472
def get_readonly_url(self, relpath=None):
2473
"""Get a URL for the readonly transport.
2475
This will either be backed by '.' or a decorator to the transport
2476
used by self.get_url()
2477
relpath provides for clients to get a path relative to the base url.
2478
These should only be downwards relative, not upwards.
2480
base = self.get_readonly_server().get_url()
2481
return self._adjust_url(base, relpath)
2483
def get_vfs_only_server(self):
2484
"""Get the vfs only read/write server instance.
2486
This is useful for some tests with specific servers that need
2489
For TestCaseWithMemoryTransport this is always a MemoryServer, and there
2490
is no means to override it.
2492
if self.__vfs_server is None:
2493
self.__vfs_server = memory.MemoryServer()
2494
self.start_server(self.__vfs_server)
2495
return self.__vfs_server
2497
def get_server(self):
2498
"""Get the read/write server instance.
2500
This is useful for some tests with specific servers that need
2503
This is built from the self.transport_server factory. If that is None,
2504
then the self.get_vfs_server is returned.
2506
if self.__server is None:
2507
if (self.transport_server is None or self.transport_server is
2508
self.vfs_transport_factory):
2509
self.__server = self.get_vfs_only_server()
2511
# bring up a decorated means of access to the vfs only server.
2512
self.__server = self.transport_server()
2513
self.start_server(self.__server, self.get_vfs_only_server())
2514
return self.__server
2516
def _adjust_url(self, base, relpath):
2517
"""Get a URL (or maybe a path) for the readwrite transport.
2519
This will either be backed by '.' or to an equivalent non-file based
2521
relpath provides for clients to get a path relative to the base url.
2522
These should only be downwards relative, not upwards.
2524
if relpath is not None and relpath != '.':
2525
if not base.endswith('/'):
2527
# XXX: Really base should be a url; we did after all call
2528
# get_url()! But sometimes it's just a path (from
2529
# LocalAbspathServer), and it'd be wrong to append urlescaped data
2530
# to a non-escaped local path.
2531
if base.startswith('./') or base.startswith('/'):
2534
base += urlutils.escape(relpath)
2537
def get_url(self, relpath=None):
2538
"""Get a URL (or maybe a path) for the readwrite transport.
2540
This will either be backed by '.' or to an equivalent non-file based
2542
relpath provides for clients to get a path relative to the base url.
2543
These should only be downwards relative, not upwards.
2545
base = self.get_server().get_url()
2546
return self._adjust_url(base, relpath)
2548
def get_vfs_only_url(self, relpath=None):
2549
"""Get a URL (or maybe a path for the plain old vfs transport.
2551
This will never be a smart protocol. It always has all the
2552
capabilities of the local filesystem, but it might actually be a
2553
MemoryTransport or some other similar virtual filesystem.
2555
This is the backing transport (if any) of the server returned by
2556
get_url and get_readonly_url.
2558
:param relpath: provides for clients to get a path relative to the base
2559
url. These should only be downwards relative, not upwards.
2562
base = self.get_vfs_only_server().get_url()
2563
return self._adjust_url(base, relpath)
2565
def _create_safety_net(self):
2566
"""Make a fake bzr directory.
2568
This prevents any tests propagating up onto the TEST_ROOT directory's
2571
root = TestCaseWithMemoryTransport.TEST_ROOT
2573
# Make sure we get a readable and accessible home for .brz.log
2574
# and/or config files, and not fallback to weird defaults (see
2575
# http://pad.lv/825027).
2576
self.assertIs(None, os.environ.get('BRZ_HOME', None))
2577
os.environ['BRZ_HOME'] = root
2578
wt = controldir.ControlDir.create_standalone_workingtree(root)
2579
del os.environ['BRZ_HOME']
2580
except Exception as e:
2581
self.fail("Fail to initialize the safety net: %r\n" % (e,))
2582
# Hack for speed: remember the raw bytes of the dirstate file so that
2583
# we don't need to re-open the wt to check it hasn't changed.
2584
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE = (
2585
wt.control_transport.get_bytes('dirstate'))
2587
def _check_safety_net(self):
2588
"""Check that the safety .bzr directory have not been touched.
2590
_make_test_root have created a .bzr directory to prevent tests from
2591
propagating. This method ensures than a test did not leaked.
2593
root = TestCaseWithMemoryTransport.TEST_ROOT
2594
t = _mod_transport.get_transport_from_path(root)
2595
self.permit_url(t.base)
2596
if (t.get_bytes('.bzr/checkout/dirstate') !=
2597
TestCaseWithMemoryTransport._SAFETY_NET_PRISTINE_DIRSTATE):
2598
# The current test have modified the /bzr directory, we need to
2599
# recreate a new one or all the followng tests will fail.
2600
# If you need to inspect its content uncomment the following line
2601
# import pdb; pdb.set_trace()
2602
_rmtree_temp_dir(root + '/.bzr', test_id=self.id())
2603
self._create_safety_net()
2604
raise AssertionError('%s/.bzr should not be modified' % root)
2606
def _make_test_root(self):
2607
if TestCaseWithMemoryTransport.TEST_ROOT is None:
2608
# Watch out for tricky test dir (on OSX /tmp -> /private/tmp)
2609
root = osutils.realpath(osutils.mkdtemp(prefix='testbzr-',
2611
TestCaseWithMemoryTransport.TEST_ROOT = root
2613
self._create_safety_net()
2615
# The same directory is used by all tests, and we're not
2616
# specifically told when all tests are finished. This will do.
2617
atexit.register(_rmtree_temp_dir, root)
2619
self.permit_dir(TestCaseWithMemoryTransport.TEST_ROOT)
2620
self.addCleanup(self._check_safety_net)
2622
def makeAndChdirToTestDir(self):
2623
"""Create a temporary directories for this one test.
2625
This must set self.test_home_dir and self.test_dir and chdir to
2628
For TestCaseWithMemoryTransport we chdir to the TEST_ROOT for this test.
2630
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2631
self.test_dir = TestCaseWithMemoryTransport.TEST_ROOT
2632
self.test_home_dir = self.test_dir + "/MemoryTransportMissingHomeDir"
2633
self.permit_dir(self.test_dir)
2635
def make_branch(self, relpath, format=None, name=None):
2636
"""Create a branch on the transport at relpath."""
2637
repo = self.make_repository(relpath, format=format)
2638
return repo.controldir.create_branch(append_revisions_only=False, name=name)
2640
def get_default_format(self):
2643
def resolve_format(self, format):
2644
"""Resolve an object to a ControlDir format object.
2646
The initial format object can either already be
2647
a ControlDirFormat, None (for the default format),
2648
or a string with the name of the control dir format.
2650
:param format: Object to resolve
2651
:return A ControlDirFormat instance
2654
format = self.get_default_format()
2655
if isinstance(format, str):
2656
format = controldir.format_registry.make_controldir(format)
2659
def make_controldir(self, relpath, format=None):
2661
# might be a relative or absolute path
2662
maybe_a_url = self.get_url(relpath)
2663
segments = maybe_a_url.rsplit('/', 1)
2664
t = _mod_transport.get_transport(maybe_a_url)
2665
if len(segments) > 1 and segments[-1] not in ('', '.'):
2667
format = self.resolve_format(format)
2668
return format.initialize_on_transport(t)
2669
except errors.UninitializableFormat:
2670
raise TestSkipped("Format %s is not initializable." % format)
2672
def make_repository(self, relpath, shared=None, format=None):
2673
"""Create a repository on our default transport at relpath.
2675
Note that relpath must be a relative path, not a full url.
2677
# FIXME: If you create a remoterepository this returns the underlying
2678
# real format, which is incorrect. Actually we should make sure that
2679
# RemoteBzrDir returns a RemoteRepository.
2680
# maybe mbp 20070410
2681
made_control = self.make_controldir(relpath, format=format)
2682
return made_control.create_repository(shared=shared)
2684
def make_smart_server(self, path, backing_server=None):
2685
if backing_server is None:
2686
backing_server = self.get_server()
2687
smart_server = test_server.SmartTCPServer_for_testing()
2688
self.start_server(smart_server, backing_server)
2689
remote_transport = _mod_transport.get_transport_from_url(smart_server.get_url()
2691
return remote_transport
2693
def make_branch_and_memory_tree(self, relpath, format=None):
2694
"""Create a branch on the default transport and a MemoryTree for it."""
2695
b = self.make_branch(relpath, format=format)
2696
return memorytree.MemoryTree.create_on_branch(b)
2698
def make_branch_builder(self, relpath, format=None):
2699
branch = self.make_branch(relpath, format=format)
2700
return branchbuilder.BranchBuilder(branch=branch)
2702
def overrideEnvironmentForTesting(self):
2703
test_home_dir = self.test_home_dir
2704
if not PY3 and isinstance(test_home_dir, text_type):
2705
test_home_dir = test_home_dir.encode(sys.getfilesystemencoding())
2706
self.overrideEnv('HOME', test_home_dir)
2707
self.overrideEnv('BRZ_HOME', test_home_dir)
2708
self.overrideEnv('GNUPGHOME', os.path.join(test_home_dir, '.gnupg'))
2710
def setup_smart_server_with_call_log(self):
2711
"""Sets up a smart server as the transport server with a call log."""
2712
self.transport_server = test_server.SmartTCPServer_for_testing
2713
self.hpss_connections = []
2714
self.hpss_calls = []
2716
# Skip the current stack down to the caller of
2717
# setup_smart_server_with_call_log
2718
prefix_length = len(traceback.extract_stack()) - 2
2719
def capture_hpss_call(params):
2720
self.hpss_calls.append(
2721
CapturedCall(params, prefix_length))
2722
def capture_connect(transport):
2723
self.hpss_connections.append(transport)
2724
client._SmartClient.hooks.install_named_hook(
2725
'call', capture_hpss_call, None)
2726
_mod_transport.Transport.hooks.install_named_hook(
2727
'post_connect', capture_connect, None)
2729
def reset_smart_call_log(self):
2730
self.hpss_calls = []
2731
self.hpss_connections = []
2734
class TestCaseInTempDir(TestCaseWithMemoryTransport):
2735
"""Derived class that runs a test within a temporary directory.
2737
This is useful for tests that need to create a branch, etc.
2739
The directory is created in a slightly complex way: for each
2740
Python invocation, a new temporary top-level directory is created.
2741
All test cases create their own directory within that. If the
2742
tests complete successfully, the directory is removed.
2744
:ivar test_base_dir: The path of the top-level directory for this
2745
test, which contains a home directory and a work directory.
2747
:ivar test_home_dir: An initially empty directory under test_base_dir
2748
which is used as $HOME for this test.
2750
:ivar test_dir: A directory under test_base_dir used as the current
2751
directory when the test proper is run.
2754
OVERRIDE_PYTHON = 'python'
2757
super(TestCaseInTempDir, self).setUp()
2758
# Remove the protection set in isolated_environ, we have a proper
2759
# access to disk resources now.
2760
self.overrideEnv('BRZ_LOG', None)
2762
def check_file_contents(self, filename, expect):
2763
self.log("check contents of file %s" % filename)
2764
with open(filename) as f:
2766
if contents != expect:
2767
self.log("expected: %r" % expect)
2768
self.log("actually: %r" % contents)
2769
self.fail("contents of %s not as expected" % filename)
2771
def _getTestDirPrefix(self):
2772
# create a directory within the top level test directory
2773
if sys.platform in ('win32', 'cygwin'):
2774
name_prefix = re.sub('[<>*=+",:;_/\\-]', '_', self.id())
2775
# windows is likely to have path-length limits so use a short name
2776
name_prefix = name_prefix[-30:]
2778
name_prefix = re.sub('[/]', '_', self.id())
2781
def makeAndChdirToTestDir(self):
2782
"""See TestCaseWithMemoryTransport.makeAndChdirToTestDir().
2784
For TestCaseInTempDir we create a temporary directory based on the test
2785
name and then create two subdirs - test and home under it.
2787
name_prefix = osutils.pathjoin(TestCaseWithMemoryTransport.TEST_ROOT,
2788
self._getTestDirPrefix())
2790
for i in range(100):
2791
if os.path.exists(name):
2792
name = name_prefix + '_' + str(i)
2794
# now create test and home directories within this dir
2795
self.test_base_dir = name
2796
self.addCleanup(self.deleteTestDir)
2797
os.mkdir(self.test_base_dir)
2799
self.permit_dir(self.test_base_dir)
2800
# 'sprouting' and 'init' of a branch both walk up the tree to find
2801
# stacking policy to honour; create a bzr dir with an unshared
2802
# repository (but not a branch - our code would be trying to escape
2803
# then!) to stop them, and permit it to be read.
2804
# control = controldir.ControlDir.create(self.test_base_dir)
2805
# control.create_repository()
2806
self.test_home_dir = self.test_base_dir + '/home'
2807
os.mkdir(self.test_home_dir)
2808
self.test_dir = self.test_base_dir + '/work'
2809
os.mkdir(self.test_dir)
2810
os.chdir(self.test_dir)
2811
# put name of test inside
2812
with open(self.test_base_dir + '/name', 'w') as f:
2815
def deleteTestDir(self):
2816
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2817
_rmtree_temp_dir(self.test_base_dir, test_id=self.id())
2819
def build_tree(self, shape, line_endings='binary', transport=None):
2820
"""Build a test tree according to a pattern.
2822
shape is a sequence of file specifications. If the final
2823
character is '/', a directory is created.
2825
This assumes that all the elements in the tree being built are new.
2827
This doesn't add anything to a branch.
2829
:type shape: list or tuple.
2830
:param line_endings: Either 'binary' or 'native'
2831
in binary mode, exact contents are written in native mode, the
2832
line endings match the default platform endings.
2833
:param transport: A transport to write to, for building trees on VFS's.
2834
If the transport is readonly or None, "." is opened automatically.
2837
if type(shape) not in (list, tuple):
2838
raise AssertionError("Parameter 'shape' should be "
2839
"a list or a tuple. Got %r instead" % (shape,))
2840
# It's OK to just create them using forward slashes on windows.
2841
if transport is None or transport.is_readonly():
2842
transport = _mod_transport.get_transport_from_path(".")
2844
self.assertIsInstance(name, (str, text_type))
2846
transport.mkdir(urlutils.escape(name[:-1]))
2848
if line_endings == 'binary':
2850
elif line_endings == 'native':
2851
end = os.linesep.encode('ascii')
2853
raise errors.BzrError(
2854
'Invalid line ending request %r' % line_endings)
2855
content = b"contents of %s%s" % (name.encode('utf-8'), end)
2856
transport.put_bytes_non_atomic(urlutils.escape(name), content)
2858
build_tree_contents = staticmethod(treeshape.build_tree_contents)
2860
def assertInWorkingTree(self, path, root_path='.', tree=None):
2861
"""Assert whether path or paths are in the WorkingTree"""
2863
tree = workingtree.WorkingTree.open(root_path)
2864
if not isinstance(path, (str, text_type)):
2866
self.assertInWorkingTree(p, tree=tree)
2868
self.assertTrue(tree.is_versioned(path),
2869
path+' not in working tree.')
2871
def assertNotInWorkingTree(self, path, root_path='.', tree=None):
2872
"""Assert whether path or paths are not in the WorkingTree"""
2874
tree = workingtree.WorkingTree.open(root_path)
2875
if not isinstance(path, (str, text_type)):
2877
self.assertNotInWorkingTree(p, tree=tree)
2879
self.assertFalse(tree.is_versioned(path), path+' in working tree.')
2882
class TestCaseWithTransport(TestCaseInTempDir):
2883
"""A test case that provides get_url and get_readonly_url facilities.
2885
These back onto two transport servers, one for readonly access and one for
2888
If no explicit class is provided for readonly access, a
2889
ReadonlyTransportDecorator is used instead which allows the use of non disk
2890
based read write transports.
2892
If an explicit class is provided for readonly access, that server and the
2893
readwrite one must both define get_url() as resolving to os.getcwd().
2897
super(TestCaseWithTransport, self).setUp()
2898
self.__vfs_server = None
2900
def get_vfs_only_server(self):
2901
"""See TestCaseWithMemoryTransport.
2903
This is useful for some tests with specific servers that need
2906
if self.__vfs_server is None:
2907
self.__vfs_server = self.vfs_transport_factory()
2908
self.start_server(self.__vfs_server)
2909
return self.__vfs_server
2911
def make_branch_and_tree(self, relpath, format=None):
2912
"""Create a branch on the transport and a tree locally.
2914
If the transport is not a LocalTransport, the Tree can't be created on
2915
the transport. In that case if the vfs_transport_factory is
2916
LocalURLServer the working tree is created in the local
2917
directory backing the transport, and the returned tree's branch and
2918
repository will also be accessed locally. Otherwise a lightweight
2919
checkout is created and returned.
2921
We do this because we can't physically create a tree in the local
2922
path, with a branch reference to the transport_factory url, and
2923
a branch + repository in the vfs_transport, unless the vfs_transport
2924
namespace is distinct from the local disk - the two branch objects
2925
would collide. While we could construct a tree with its branch object
2926
pointing at the transport_factory transport in memory, reopening it
2927
would behaving unexpectedly, and has in the past caused testing bugs
2928
when we tried to do it that way.
2930
:param format: The BzrDirFormat.
2931
:returns: the WorkingTree.
2933
# TODO: always use the local disk path for the working tree,
2934
# this obviously requires a format that supports branch references
2935
# so check for that by checking bzrdir.BzrDirFormat.get_default_format()
2937
format = self.resolve_format(format=format)
2938
if not format.supports_workingtrees:
2939
b = self.make_branch(relpath+'.branch', format=format)
2940
return b.create_checkout(relpath, lightweight=True)
2941
b = self.make_branch(relpath, format=format)
2943
return b.controldir.create_workingtree()
2944
except errors.NotLocalUrl:
2945
# We can only make working trees locally at the moment. If the
2946
# transport can't support them, then we keep the non-disk-backed
2947
# branch and create a local checkout.
2948
if self.vfs_transport_factory is test_server.LocalURLServer:
2949
# the branch is colocated on disk, we cannot create a checkout.
2950
# hopefully callers will expect this.
2951
local_controldir = controldir.ControlDir.open(
2952
self.get_vfs_only_url(relpath))
2953
wt = local_controldir.create_workingtree()
2954
if wt.branch._format != b._format:
2956
# Make sure that assigning to wt._branch fixes wt.branch,
2957
# in case the implementation details of workingtree objects
2959
self.assertIs(b, wt.branch)
2962
return b.create_checkout(relpath, lightweight=True)
2964
def assertIsDirectory(self, relpath, transport):
2965
"""Assert that relpath within transport is a directory.
2967
This may not be possible on all transports; in that case it propagates
2968
a TransportNotPossible.
2971
mode = transport.stat(relpath).st_mode
2972
except errors.NoSuchFile:
2973
self.fail("path %s is not a directory; no such file"
2975
if not stat.S_ISDIR(mode):
2976
self.fail("path %s is not a directory; has mode %#o"
2979
def assertTreesEqual(self, left, right):
2980
"""Check that left and right have the same content and properties."""
2981
# we use a tree delta to check for equality of the content, and we
2982
# manually check for equality of other things such as the parents list.
2983
self.assertEqual(left.get_parent_ids(), right.get_parent_ids())
2984
differences = left.changes_from(right)
2985
self.assertFalse(differences.has_changed(),
2986
"Trees %r and %r are different: %r" % (left, right, differences))
2988
def disable_missing_extensions_warning(self):
2989
"""Some tests expect a precise stderr content.
2991
There is no point in forcing them to duplicate the extension related
2994
config.GlobalConfig().set_user_option(
2995
'suppress_warnings', 'missing_extensions')
2998
class ChrootedTestCase(TestCaseWithTransport):
2999
"""A support class that provides readonly urls outside the local namespace.
3001
This is done by checking if self.transport_server is a MemoryServer. if it
3002
is then we are chrooted already, if it is not then an HttpServer is used
3005
TODO RBC 20060127: make this an option to TestCaseWithTransport so it can
3006
be used without needed to redo it when a different
3007
subclass is in use ?
3011
from breezy.tests import http_server
3012
super(ChrootedTestCase, self).setUp()
3013
if not self.vfs_transport_factory == memory.MemoryServer:
3014
self.transport_readonly_server = http_server.HttpServer
3017
def condition_id_re(pattern):
3018
"""Create a condition filter which performs a re check on a test's id.
3020
:param pattern: A regular expression string.
3021
:return: A callable that returns True if the re matches.
3023
filter_re = re.compile(pattern, 0)
3024
def condition(test):
3026
return filter_re.search(test_id)
3030
def condition_isinstance(klass_or_klass_list):
3031
"""Create a condition filter which returns isinstance(param, klass).
3033
:return: A callable which when called with one parameter obj return the
3034
result of isinstance(obj, klass_or_klass_list).
3037
return isinstance(obj, klass_or_klass_list)
3041
def condition_id_in_list(id_list):
3042
"""Create a condition filter which verify that test's id in a list.
3044
:param id_list: A TestIdList object.
3045
:return: A callable that returns True if the test's id appears in the list.
3047
def condition(test):
3048
return id_list.includes(test.id())
3052
def condition_id_startswith(starts):
3053
"""Create a condition filter verifying that test's id starts with a string.
3055
:param starts: A list of string.
3056
:return: A callable that returns True if the test's id starts with one of
3059
def condition(test):
3060
for start in starts:
3061
if test.id().startswith(start):
3067
def exclude_tests_by_condition(suite, condition):
3068
"""Create a test suite which excludes some tests from suite.
3070
:param suite: The suite to get tests from.
3071
:param condition: A callable whose result evaluates True when called with a
3072
test case which should be excluded from the result.
3073
:return: A suite which contains the tests found in suite that fail
3077
for test in iter_suite_tests(suite):
3078
if not condition(test):
3080
return TestUtil.TestSuite(result)
3083
def filter_suite_by_condition(suite, condition):
3084
"""Create a test suite by filtering another one.
3086
:param suite: The source suite.
3087
:param condition: A callable whose result evaluates True when called with a
3088
test case which should be included in the result.
3089
:return: A suite which contains the tests found in suite that pass
3093
for test in iter_suite_tests(suite):
3096
return TestUtil.TestSuite(result)
3099
def filter_suite_by_re(suite, pattern):
3100
"""Create a test suite by filtering another one.
3102
:param suite: the source suite
3103
:param pattern: pattern that names must match
3104
:returns: the newly created suite
3106
condition = condition_id_re(pattern)
3107
result_suite = filter_suite_by_condition(suite, condition)
3111
def filter_suite_by_id_list(suite, test_id_list):
3112
"""Create a test suite by filtering another one.
3114
:param suite: The source suite.
3115
:param test_id_list: A list of the test ids to keep as strings.
3116
:returns: the newly created suite
3118
condition = condition_id_in_list(test_id_list)
3119
result_suite = filter_suite_by_condition(suite, condition)
3123
def filter_suite_by_id_startswith(suite, start):
3124
"""Create a test suite by filtering another one.
3126
:param suite: The source suite.
3127
:param start: A list of string the test id must start with one of.
3128
:returns: the newly created suite
3130
condition = condition_id_startswith(start)
3131
result_suite = filter_suite_by_condition(suite, condition)
3135
def exclude_tests_by_re(suite, pattern):
3136
"""Create a test suite which excludes some tests from suite.
3138
:param suite: The suite to get tests from.
3139
:param pattern: A regular expression string. Test ids that match this
3140
pattern will be excluded from the result.
3141
:return: A TestSuite that contains all the tests from suite without the
3142
tests that matched pattern. The order of tests is the same as it was in
3145
return exclude_tests_by_condition(suite, condition_id_re(pattern))
3148
def preserve_input(something):
3149
"""A helper for performing test suite transformation chains.
3151
:param something: Anything you want to preserve.
3157
def randomize_suite(suite):
3158
"""Return a new TestSuite with suite's tests in random order.
3160
The tests in the input suite are flattened into a single suite in order to
3161
accomplish this. Any nested TestSuites are removed to provide global
3164
tests = list(iter_suite_tests(suite))
3165
random.shuffle(tests)
3166
return TestUtil.TestSuite(tests)
3169
def split_suite_by_condition(suite, condition):
3170
"""Split a test suite into two by a condition.
3172
:param suite: The suite to split.
3173
:param condition: The condition to match on. Tests that match this
3174
condition are returned in the first test suite, ones that do not match
3175
are in the second suite.
3176
:return: A tuple of two test suites, where the first contains tests from
3177
suite matching the condition, and the second contains the remainder
3178
from suite. The order within each output suite is the same as it was in
3183
for test in iter_suite_tests(suite):
3185
matched.append(test)
3187
did_not_match.append(test)
3188
return TestUtil.TestSuite(matched), TestUtil.TestSuite(did_not_match)
3191
def split_suite_by_re(suite, pattern):
3192
"""Split a test suite into two by a regular expression.
3194
:param suite: The suite to split.
3195
:param pattern: A regular expression string. Test ids that match this
3196
pattern will be in the first test suite returned, and the others in the
3197
second test suite returned.
3198
:return: A tuple of two test suites, where the first contains tests from
3199
suite matching pattern, and the second contains the remainder from
3200
suite. The order within each output suite is the same as it was in
3203
return split_suite_by_condition(suite, condition_id_re(pattern))
3206
def run_suite(suite, name='test', verbose=False, pattern=".*",
3207
stop_on_failure=False,
3208
transport=None, lsprof_timed=None, bench_history=None,
3209
matching_tests_first=None,
3212
exclude_pattern=None,
3215
suite_decorators=None,
3217
result_decorators=None,
3219
"""Run a test suite for brz selftest.
3221
:param runner_class: The class of runner to use. Must support the
3222
constructor arguments passed by run_suite which are more than standard
3224
:return: A boolean indicating success.
3226
TestCase._gather_lsprof_in_benchmarks = lsprof_timed
3231
if runner_class is None:
3232
runner_class = TextTestRunner
3235
runner = runner_class(stream=stream,
3237
verbosity=verbosity,
3238
bench_history=bench_history,
3240
result_decorators=result_decorators,
3242
runner.stop_on_failure=stop_on_failure
3243
if isinstance(suite, unittest.TestSuite):
3244
# Empty out _tests list of passed suite and populate new TestSuite
3245
suite._tests[:], suite = [], TestSuite(suite)
3246
# built in decorator factories:
3248
random_order(random_seed, runner),
3249
exclude_tests(exclude_pattern),
3251
if matching_tests_first:
3252
decorators.append(tests_first(pattern))
3254
decorators.append(filter_tests(pattern))
3255
if suite_decorators:
3256
decorators.extend(suite_decorators)
3257
# tell the result object how many tests will be running: (except if
3258
# --parallel=fork is being used. Robert said he will provide a better
3259
# progress design later -- vila 20090817)
3260
if fork_decorator not in decorators:
3261
decorators.append(CountingDecorator)
3262
for decorator in decorators:
3263
suite = decorator(suite)
3265
# Done after test suite decoration to allow randomisation etc
3266
# to take effect, though that is of marginal benefit.
3268
stream.write("Listing tests only ...\n")
3269
if getattr(runner, 'list', None) is not None:
3272
for t in iter_suite_tests(suite):
3273
stream.write("%s\n" % (t.id()))
3275
result = runner.run(suite)
3277
return result.wasStrictlySuccessful()
3279
return result.wasSuccessful()
3282
# A registry where get() returns a suite decorator.
3283
parallel_registry = registry.Registry()
3286
def fork_decorator(suite):
3287
if getattr(os, "fork", None) is None:
3288
raise errors.BzrCommandError("platform does not support fork,"
3289
" try --parallel=subprocess instead.")
3290
concurrency = osutils.local_concurrency()
3291
if concurrency == 1:
3293
from testtools import ConcurrentTestSuite
3294
return ConcurrentTestSuite(suite, fork_for_tests)
3295
parallel_registry.register('fork', fork_decorator)
3298
def subprocess_decorator(suite):
3299
concurrency = osutils.local_concurrency()
3300
if concurrency == 1:
3302
from testtools import ConcurrentTestSuite
3303
return ConcurrentTestSuite(suite, reinvoke_for_tests)
3304
parallel_registry.register('subprocess', subprocess_decorator)
3307
def exclude_tests(exclude_pattern):
3308
"""Return a test suite decorator that excludes tests."""
3309
if exclude_pattern is None:
3310
return identity_decorator
3311
def decorator(suite):
3312
return ExcludeDecorator(suite, exclude_pattern)
3316
def filter_tests(pattern):
3318
return identity_decorator
3319
def decorator(suite):
3320
return FilterTestsDecorator(suite, pattern)
3324
def random_order(random_seed, runner):
3325
"""Return a test suite decorator factory for randomising tests order.
3327
:param random_seed: now, a string which casts to an integer, or an integer.
3328
:param runner: A test runner with a stream attribute to report on.
3330
if random_seed is None:
3331
return identity_decorator
3332
def decorator(suite):
3333
return RandomDecorator(suite, random_seed, runner.stream)
3337
def tests_first(pattern):
3339
return identity_decorator
3340
def decorator(suite):
3341
return TestFirstDecorator(suite, pattern)
3345
def identity_decorator(suite):
3350
class TestDecorator(TestUtil.TestSuite):
3351
"""A decorator for TestCase/TestSuite objects.
3353
Contains rather than flattening suite passed on construction
3356
def __init__(self, suite=None):
3357
super(TestDecorator, self).__init__()
3358
if suite is not None:
3361
# Don't need subclass run method with suite emptying
3362
run = unittest.TestSuite.run
3365
class CountingDecorator(TestDecorator):
3366
"""A decorator which calls result.progress(self.countTestCases)."""
3368
def run(self, result):
3369
progress_method = getattr(result, 'progress', None)
3370
if callable(progress_method):
3371
progress_method(self.countTestCases(), SUBUNIT_SEEK_SET)
3372
return super(CountingDecorator, self).run(result)
3375
class ExcludeDecorator(TestDecorator):
3376
"""A decorator which excludes test matching an exclude pattern."""
3378
def __init__(self, suite, exclude_pattern):
3379
super(ExcludeDecorator, self).__init__(
3380
exclude_tests_by_re(suite, exclude_pattern))
3383
class FilterTestsDecorator(TestDecorator):
3384
"""A decorator which filters tests to those matching a pattern."""
3386
def __init__(self, suite, pattern):
3387
super(FilterTestsDecorator, self).__init__(
3388
filter_suite_by_re(suite, pattern))
3391
class RandomDecorator(TestDecorator):
3392
"""A decorator which randomises the order of its tests."""
3394
def __init__(self, suite, random_seed, stream):
3395
random_seed = self.actual_seed(random_seed)
3396
stream.write("Randomizing test order using seed %s\n\n" %
3398
# Initialise the random number generator.
3399
random.seed(random_seed)
3400
super(RandomDecorator, self).__init__(randomize_suite(suite))
3403
def actual_seed(seed):
3405
# We convert the seed to an integer to make it reuseable across
3406
# invocations (because the user can reenter it).
3407
return int(time.time())
3409
# Convert the seed to an integer if we can
3412
except (TypeError, ValueError):
3417
class TestFirstDecorator(TestDecorator):
3418
"""A decorator which moves named tests to the front."""
3420
def __init__(self, suite, pattern):
3421
super(TestFirstDecorator, self).__init__()
3422
self.addTests(split_suite_by_re(suite, pattern))
3425
def partition_tests(suite, count):
3426
"""Partition suite into count lists of tests."""
3427
# This just assigns tests in a round-robin fashion. On one hand this
3428
# splits up blocks of related tests that might run faster if they shared
3429
# resources, but on the other it avoids assigning blocks of slow tests to
3430
# just one partition. So the slowest partition shouldn't be much slower
3432
partitions = [list() for i in range(count)]
3433
tests = iter_suite_tests(suite)
3434
for partition, test in zip(itertools.cycle(partitions), tests):
3435
partition.append(test)
3439
def workaround_zealous_crypto_random():
3440
"""Crypto.Random want to help us being secure, but we don't care here.
3442
This workaround some test failure related to the sftp server. Once paramiko
3443
stop using the controversial API in Crypto.Random, we may get rid of it.
3446
from Crypto.Random import atfork
3452
def fork_for_tests(suite):
3453
"""Take suite and start up one runner per CPU by forking()
3455
:return: An iterable of TestCase-like objects which can each have
3456
run(result) called on them to feed tests to result.
3458
concurrency = osutils.local_concurrency()
3460
from subunit import ProtocolTestCase
3461
from subunit.test_results import AutoTimingTestResultDecorator
3462
class TestInOtherProcess(ProtocolTestCase):
3463
# Should be in subunit, I think. RBC.
3464
def __init__(self, stream, pid):
3465
ProtocolTestCase.__init__(self, stream)
3468
def run(self, result):
3470
ProtocolTestCase.run(self, result)
3472
pid, status = os.waitpid(self.pid, 0)
3473
# GZ 2011-10-18: If status is nonzero, should report to the result
3474
# that something went wrong.
3476
test_blocks = partition_tests(suite, concurrency)
3477
# Clear the tests from the original suite so it doesn't keep them alive
3478
suite._tests[:] = []
3479
for process_tests in test_blocks:
3480
process_suite = TestUtil.TestSuite(process_tests)
3481
# Also clear each split list so new suite has only reference
3482
process_tests[:] = []
3483
c2pread, c2pwrite = os.pipe()
3487
stream = os.fdopen(c2pwrite, 'wb', 1)
3488
workaround_zealous_crypto_random()
3490
# Leave stderr and stdout open so we can see test noise
3491
# Close stdin so that the child goes away if it decides to
3492
# read from stdin (otherwise its a roulette to see what
3493
# child actually gets keystrokes for pdb etc).
3495
subunit_result = AutoTimingTestResultDecorator(
3496
SubUnitBzrProtocolClientv1(stream))
3497
process_suite.run(subunit_result)
3499
# Try and report traceback on stream, but exit with error even
3500
# if stream couldn't be created or something else goes wrong.
3501
# The traceback is formatted to a string and written in one go
3502
# to avoid interleaving lines from multiple failing children.
3504
stream.write(traceback.format_exc())
3510
stream = os.fdopen(c2pread, 'rb', 1)
3511
test = TestInOtherProcess(stream, pid)
3516
def reinvoke_for_tests(suite):
3517
"""Take suite and start up one runner per CPU using subprocess().
3519
:return: An iterable of TestCase-like objects which can each have
3520
run(result) called on them to feed tests to result.
3522
concurrency = osutils.local_concurrency()
3524
from subunit import ProtocolTestCase
3525
class TestInSubprocess(ProtocolTestCase):
3526
def __init__(self, process, name):
3527
ProtocolTestCase.__init__(self, process.stdout)
3528
self.process = process
3529
self.process.stdin.close()
3532
def run(self, result):
3534
ProtocolTestCase.run(self, result)
3537
os.unlink(self.name)
3538
# print "pid %d finished" % finished_process
3539
test_blocks = partition_tests(suite, concurrency)
3540
for process_tests in test_blocks:
3541
# ugly; currently reimplement rather than reuses TestCase methods.
3542
bzr_path = os.path.dirname(os.path.dirname(breezy.__file__))+'/bzr'
3543
if not os.path.isfile(bzr_path):
3544
# We are probably installed. Assume sys.argv is the right file
3545
bzr_path = sys.argv[0]
3546
bzr_path = [bzr_path]
3547
if sys.platform == "win32":
3548
# if we're on windows, we can't execute the bzr script directly
3549
bzr_path = [sys.executable] + bzr_path
3550
fd, test_list_file_name = tempfile.mkstemp()
3551
test_list_file = os.fdopen(fd, 'wb', 1)
3552
for test in process_tests:
3553
test_list_file.write(test.id() + '\n')
3554
test_list_file.close()
3556
argv = bzr_path + ['selftest', '--load-list', test_list_file_name,
3558
if '--no-plugins' in sys.argv:
3559
argv.append('--no-plugins')
3560
# stderr=subprocess.STDOUT would be ideal, but until we prevent
3561
# noise on stderr it can interrupt the subunit protocol.
3562
process = subprocess.Popen(argv, stdin=subprocess.PIPE,
3563
stdout=subprocess.PIPE,
3564
stderr=subprocess.PIPE,
3566
test = TestInSubprocess(process, test_list_file_name)
3569
os.unlink(test_list_file_name)
3574
class ProfileResult(testtools.ExtendedToOriginalDecorator):
3575
"""Generate profiling data for all activity between start and success.
3577
The profile data is appended to the test's _benchcalls attribute and can
3578
be accessed by the forwarded-to TestResult.
3580
While it might be cleaner do accumulate this in stopTest, addSuccess is
3581
where our existing output support for lsprof is, and this class aims to
3582
fit in with that: while it could be moved it's not necessary to accomplish
3583
test profiling, nor would it be dramatically cleaner.
3586
def startTest(self, test):
3587
self.profiler = breezy.lsprof.BzrProfiler()
3588
# Prevent deadlocks in tests that use lsprof: those tests will
3590
breezy.lsprof.BzrProfiler.profiler_block = 0
3591
self.profiler.start()
3592
testtools.ExtendedToOriginalDecorator.startTest(self, test)
3594
def addSuccess(self, test):
3595
stats = self.profiler.stop()
3597
calls = test._benchcalls
3598
except AttributeError:
3599
test._benchcalls = []
3600
calls = test._benchcalls
3601
calls.append(((test.id(), "", ""), stats))
3602
testtools.ExtendedToOriginalDecorator.addSuccess(self, test)
3604
def stopTest(self, test):
3605
testtools.ExtendedToOriginalDecorator.stopTest(self, test)
3606
self.profiler = None
3609
# Controlled by "brz selftest -E=..." option
3610
# Currently supported:
3611
# -Eallow_debug Will no longer clear debug.debug_flags() so it
3612
# preserves any flags supplied at the command line.
3613
# -Edisable_lock_checks Turns errors in mismatched locks into simple prints
3614
# rather than failing tests. And no longer raise
3615
# LockContention when fctnl locks are not being used
3616
# with proper exclusion rules.
3617
# -Ethreads Will display thread ident at creation/join time to
3618
# help track thread leaks
3619
# -Euncollected_cases Display the identity of any test cases that weren't
3620
# deallocated after being completed.
3621
# -Econfig_stats Will collect statistics using addDetail
3622
selftest_debug_flags = set()
3625
def selftest(verbose=False, pattern=".*", stop_on_failure=True,
3627
test_suite_factory=None,
3630
matching_tests_first=None,
3633
exclude_pattern=None,
3639
suite_decorators=None,
3643
"""Run the whole test suite under the enhanced runner"""
3644
# XXX: Very ugly way to do this...
3645
# Disable warning about old formats because we don't want it to disturb
3646
# any blackbox tests.
3647
from breezy import repository
3648
repository._deprecation_warning_done = True
3650
global default_transport
3651
if transport is None:
3652
transport = default_transport
3653
old_transport = default_transport
3654
default_transport = transport
3655
global selftest_debug_flags
3656
old_debug_flags = selftest_debug_flags
3657
if debug_flags is not None:
3658
selftest_debug_flags = set(debug_flags)
3660
if load_list is None:
3663
keep_only = load_test_id_list(load_list)
3665
starting_with = [test_prefix_alias_registry.resolve_alias(start)
3666
for start in starting_with]
3667
# Always consider 'unittest' an interesting name so that failed
3668
# suites wrapped as test cases appear in the output.
3669
starting_with.append('unittest')
3670
if test_suite_factory is None:
3671
# Reduce loading time by loading modules based on the starting_with
3673
suite = test_suite(keep_only, starting_with)
3675
suite = test_suite_factory()
3677
# But always filter as requested.
3678
suite = filter_suite_by_id_startswith(suite, starting_with)
3679
result_decorators = []
3681
result_decorators.append(ProfileResult)
3682
return run_suite(suite, 'testbzr', verbose=verbose, pattern=pattern,
3683
stop_on_failure=stop_on_failure,
3684
transport=transport,
3685
lsprof_timed=lsprof_timed,
3686
bench_history=bench_history,
3687
matching_tests_first=matching_tests_first,
3688
list_only=list_only,
3689
random_seed=random_seed,
3690
exclude_pattern=exclude_pattern,
3692
runner_class=runner_class,
3693
suite_decorators=suite_decorators,
3695
result_decorators=result_decorators,
3698
default_transport = old_transport
3699
selftest_debug_flags = old_debug_flags
3702
def load_test_id_list(file_name):
3703
"""Load a test id list from a text file.
3705
The format is one test id by line. No special care is taken to impose
3706
strict rules, these test ids are used to filter the test suite so a test id
3707
that do not match an existing test will do no harm. This allows user to add
3708
comments, leave blank lines, etc.
3712
ftest = open(file_name, 'rt')
3713
except IOError as e:
3714
if e.errno != errno.ENOENT:
3717
raise errors.NoSuchFile(file_name)
3719
for test_name in ftest.readlines():
3720
test_list.append(test_name.strip())
3725
def suite_matches_id_list(test_suite, id_list):
3726
"""Warns about tests not appearing or appearing more than once.
3728
:param test_suite: A TestSuite object.
3729
:param test_id_list: The list of test ids that should be found in
3732
:return: (absents, duplicates) absents is a list containing the test found
3733
in id_list but not in test_suite, duplicates is a list containing the
3734
tests found multiple times in test_suite.
3736
When using a prefined test id list, it may occurs that some tests do not
3737
exist anymore or that some tests use the same id. This function warns the
3738
tester about potential problems in his workflow (test lists are volatile)
3739
or in the test suite itself (using the same id for several tests does not
3740
help to localize defects).
3742
# Build a dict counting id occurrences
3744
for test in iter_suite_tests(test_suite):
3746
tests[id] = tests.get(id, 0) + 1
3751
occurs = tests.get(id, 0)
3753
not_found.append(id)
3755
duplicates.append(id)
3757
return not_found, duplicates
3760
class TestIdList(object):
3761
"""Test id list to filter a test suite.
3763
Relying on the assumption that test ids are built as:
3764
<module>[.<class>.<method>][(<param>+)], <module> being in python dotted
3765
notation, this class offers methods to :
3766
- avoid building a test suite for modules not refered to in the test list,
3767
- keep only the tests listed from the module test suite.
3770
def __init__(self, test_id_list):
3771
# When a test suite needs to be filtered against us we compare test ids
3772
# for equality, so a simple dict offers a quick and simple solution.
3773
self.tests = dict().fromkeys(test_id_list, True)
3775
# While unittest.TestCase have ids like:
3776
# <module>.<class>.<method>[(<param+)],
3777
# doctest.DocTestCase can have ids like:
3780
# <module>.<function>
3781
# <module>.<class>.<method>
3783
# Since we can't predict a test class from its name only, we settle on
3784
# a simple constraint: a test id always begins with its module name.
3787
for test_id in test_id_list:
3788
parts = test_id.split('.')
3789
mod_name = parts.pop(0)
3790
modules[mod_name] = True
3792
mod_name += '.' + part
3793
modules[mod_name] = True
3794
self.modules = modules
3796
def refers_to(self, module_name):
3797
"""Is there tests for the module or one of its sub modules."""
3798
return module_name in self.modules
3800
def includes(self, test_id):
3801
return test_id in self.tests
3804
class TestPrefixAliasRegistry(registry.Registry):
3805
"""A registry for test prefix aliases.
3807
This helps implement shorcuts for the --starting-with selftest
3808
option. Overriding existing prefixes is not allowed but not fatal (a
3809
warning will be emitted).
3812
def register(self, key, obj, help=None, info=None,
3813
override_existing=False):
3814
"""See Registry.register.
3816
Trying to override an existing alias causes a warning to be emitted,
3817
not a fatal execption.
3820
super(TestPrefixAliasRegistry, self).register(
3821
key, obj, help=help, info=info, override_existing=False)
3823
actual = self.get(key)
3825
'Test prefix alias %s is already used for %s, ignoring %s'
3826
% (key, actual, obj))
3828
def resolve_alias(self, id_start):
3829
"""Replace the alias by the prefix in the given string.
3831
Using an unknown prefix is an error to help catching typos.
3833
parts = id_start.split('.')
3835
parts[0] = self.get(parts[0])
3837
raise errors.BzrCommandError(
3838
'%s is not a known test prefix alias' % parts[0])
3839
return '.'.join(parts)
3842
test_prefix_alias_registry = TestPrefixAliasRegistry()
3843
"""Registry of test prefix aliases."""
3846
# This alias allows to detect typos ('bzrlin.') by making all valid test ids
3847
# appear prefixed ('breezy.' is "replaced" by 'breezy.').
3848
test_prefix_alias_registry.register('breezy', 'breezy')
3850
# Obvious highest levels prefixes, feel free to add your own via a plugin
3851
test_prefix_alias_registry.register('bd', 'breezy.doc')
3852
test_prefix_alias_registry.register('bu', 'breezy.utils')
3853
test_prefix_alias_registry.register('bt', 'breezy.tests')
3854
test_prefix_alias_registry.register('bb', 'breezy.tests.blackbox')
3855
test_prefix_alias_registry.register('bp', 'breezy.plugins')
3858
def _test_suite_testmod_names():
3859
"""Return the standard list of test module names to test."""
3862
'breezy.tests.blackbox',
3863
'breezy.tests.commands',
3864
'breezy.tests.per_branch',
3865
'breezy.tests.per_bzrdir',
3866
'breezy.tests.per_controldir',
3867
'breezy.tests.per_controldir_colo',
3868
'breezy.tests.per_foreign_vcs',
3869
'breezy.tests.per_interrepository',
3870
'breezy.tests.per_intertree',
3871
'breezy.tests.per_inventory',
3872
'breezy.tests.per_interbranch',
3873
'breezy.tests.per_lock',
3874
'breezy.tests.per_merger',
3875
'breezy.tests.per_transport',
3876
'breezy.tests.per_tree',
3877
'breezy.tests.per_pack_repository',
3878
'breezy.tests.per_repository',
3879
'breezy.tests.per_repository_chk',
3880
'breezy.tests.per_repository_reference',
3881
'breezy.tests.per_repository_vf',
3882
'breezy.tests.per_uifactory',
3883
'breezy.tests.per_versionedfile',
3884
'breezy.tests.per_workingtree',
3885
'breezy.tests.test__annotator',
3886
'breezy.tests.test__bencode',
3887
'breezy.tests.test__btree_serializer',
3888
'breezy.tests.test__chk_map',
3889
'breezy.tests.test__dirstate_helpers',
3890
'breezy.tests.test__groupcompress',
3891
'breezy.tests.test__known_graph',
3892
'breezy.tests.test__rio',
3893
'breezy.tests.test__simple_set',
3894
'breezy.tests.test__static_tuple',
3895
'breezy.tests.test__walkdirs_win32',
3896
'breezy.tests.test_ancestry',
3897
'breezy.tests.test_annotate',
3898
'breezy.tests.test_atomicfile',
3899
'breezy.tests.test_bad_files',
3900
'breezy.tests.test_bisect',
3901
'breezy.tests.test_bisect_multi',
3902
'breezy.tests.test_branch',
3903
'breezy.tests.test_branchbuilder',
3904
'breezy.tests.test_btree_index',
3905
'breezy.tests.test_bugtracker',
3906
'breezy.tests.test_bundle',
3907
'breezy.tests.test_bzrdir',
3908
'breezy.tests.test__chunks_to_lines',
3909
'breezy.tests.test_cache_utf8',
3910
'breezy.tests.test_chk_map',
3911
'breezy.tests.test_chk_serializer',
3912
'breezy.tests.test_chunk_writer',
3913
'breezy.tests.test_clean_tree',
3914
'breezy.tests.test_cleanup',
3915
'breezy.tests.test_cmdline',
3916
'breezy.tests.test_commands',
3917
'breezy.tests.test_commit',
3918
'breezy.tests.test_commit_merge',
3919
'breezy.tests.test_config',
3920
'breezy.tests.test_conflicts',
3921
'breezy.tests.test_controldir',
3922
'breezy.tests.test_counted_lock',
3923
'breezy.tests.test_crash',
3924
'breezy.tests.test_decorators',
3925
'breezy.tests.test_delta',
3926
'breezy.tests.test_debug',
3927
'breezy.tests.test_diff',
3928
'breezy.tests.test_directory_service',
3929
'breezy.tests.test_dirstate',
3930
'breezy.tests.test_email_message',
3931
'breezy.tests.test_eol_filters',
3932
'breezy.tests.test_errors',
3933
'breezy.tests.test_estimate_compressed_size',
3934
'breezy.tests.test_export',
3935
'breezy.tests.test_export_pot',
3936
'breezy.tests.test_extract',
3937
'breezy.tests.test_features',
3938
'breezy.tests.test_fetch',
3939
'breezy.tests.test_fetch_ghosts',
3940
'breezy.tests.test_fixtures',
3941
'breezy.tests.test_fifo_cache',
3942
'breezy.tests.test_filters',
3943
'breezy.tests.test_filter_tree',
3944
'breezy.tests.test_ftp_transport',
3945
'breezy.tests.test_foreign',
3946
'breezy.tests.test_generate_docs',
3947
'breezy.tests.test_generate_ids',
3948
'breezy.tests.test_globbing',
3949
'breezy.tests.test_gpg',
3950
'breezy.tests.test_graph',
3951
'breezy.tests.test_groupcompress',
3952
'breezy.tests.test_hashcache',
3953
'breezy.tests.test_help',
3954
'breezy.tests.test_hooks',
3955
'breezy.tests.test_http',
3956
'breezy.tests.test_http_response',
3957
'breezy.tests.test_https_ca_bundle',
3958
'breezy.tests.test_https_urllib',
3959
'breezy.tests.test_i18n',
3960
'breezy.tests.test_identitymap',
3961
'breezy.tests.test_ignores',
3962
'breezy.tests.test_index',
3963
'breezy.tests.test_import_tariff',
3964
'breezy.tests.test_info',
3965
'breezy.tests.test_inv',
3966
'breezy.tests.test_inventory_delta',
3967
'breezy.tests.test_knit',
3968
'breezy.tests.test_lazy_import',
3969
'breezy.tests.test_lazy_regex',
3970
'breezy.tests.test_library_state',
3971
'breezy.tests.test_lock',
3972
'breezy.tests.test_lockable_files',
3973
'breezy.tests.test_lockdir',
3974
'breezy.tests.test_log',
3975
'breezy.tests.test_lru_cache',
3976
'breezy.tests.test_lsprof',
3977
'breezy.tests.test_mail_client',
3978
'breezy.tests.test_matchers',
3979
'breezy.tests.test_memorytree',
3980
'breezy.tests.test_merge',
3981
'breezy.tests.test_merge3',
3982
'breezy.tests.test_merge_core',
3983
'breezy.tests.test_merge_directive',
3984
'breezy.tests.test_mergetools',
3985
'breezy.tests.test_missing',
3986
'breezy.tests.test_msgeditor',
3987
'breezy.tests.test_multiparent',
3988
'breezy.tests.test_mutabletree',
3989
'breezy.tests.test_nonascii',
3990
'breezy.tests.test_options',
3991
'breezy.tests.test_osutils',
3992
'breezy.tests.test_osutils_encodings',
3993
'breezy.tests.test_pack',
3994
'breezy.tests.test_patch',
3995
'breezy.tests.test_patches',
3996
'breezy.tests.test_permissions',
3997
'breezy.tests.test_plugins',
3998
'breezy.tests.test_progress',
3999
'breezy.tests.test_pyutils',
4000
'breezy.tests.test_read_bundle',
4001
'breezy.tests.test_reconcile',
4002
'breezy.tests.test_reconfigure',
4003
'breezy.tests.test_registry',
4004
'breezy.tests.test_remote',
4005
'breezy.tests.test_rename_map',
4006
'breezy.tests.test_repository',
4007
'breezy.tests.test_revert',
4008
'breezy.tests.test_revision',
4009
'breezy.tests.test_revisionspec',
4010
'breezy.tests.test_revisiontree',
4011
'breezy.tests.test_rio',
4012
'breezy.tests.test_rules',
4013
'breezy.tests.test_url_policy_open',
4014
'breezy.tests.test_sampler',
4015
'breezy.tests.test_scenarios',
4016
'breezy.tests.test_script',
4017
'breezy.tests.test_selftest',
4018
'breezy.tests.test_serializer',
4019
'breezy.tests.test_setup',
4020
'breezy.tests.test_sftp_transport',
4021
'breezy.tests.test_shelf',
4022
'breezy.tests.test_shelf_ui',
4023
'breezy.tests.test_smart',
4024
'breezy.tests.test_smart_add',
4025
'breezy.tests.test_smart_request',
4026
'breezy.tests.test_smart_signals',
4027
'breezy.tests.test_smart_transport',
4028
'breezy.tests.test_smtp_connection',
4029
'breezy.tests.test_source',
4030
'breezy.tests.test_ssh_transport',
4031
'breezy.tests.test_status',
4032
'breezy.tests.test_strace',
4033
'breezy.tests.test_subsume',
4034
'breezy.tests.test_switch',
4035
'breezy.tests.test_symbol_versioning',
4036
'breezy.tests.test_tag',
4037
'breezy.tests.test_test_server',
4038
'breezy.tests.test_testament',
4039
'breezy.tests.test_textfile',
4040
'breezy.tests.test_textmerge',
4041
'breezy.tests.test_cethread',
4042
'breezy.tests.test_timestamp',
4043
'breezy.tests.test_trace',
4044
'breezy.tests.test_transactions',
4045
'breezy.tests.test_transform',
4046
'breezy.tests.test_transport',
4047
'breezy.tests.test_transport_log',
4048
'breezy.tests.test_tree',
4049
'breezy.tests.test_treebuilder',
4050
'breezy.tests.test_treeshape',
4051
'breezy.tests.test_tsort',
4052
'breezy.tests.test_tuned_gzip',
4053
'breezy.tests.test_ui',
4054
'breezy.tests.test_uncommit',
4055
'breezy.tests.test_upgrade',
4056
'breezy.tests.test_upgrade_stacked',
4057
'breezy.tests.test_upstream_import',
4058
'breezy.tests.test_urlutils',
4059
'breezy.tests.test_utextwrap',
4060
'breezy.tests.test_version',
4061
'breezy.tests.test_version_info',
4062
'breezy.tests.test_versionedfile',
4063
'breezy.tests.test_vf_search',
4064
'breezy.tests.test_views',
4065
'breezy.tests.test_weave',
4066
'breezy.tests.test_whitebox',
4067
'breezy.tests.test_win32utils',
4068
'breezy.tests.test_workingtree',
4069
'breezy.tests.test_workingtree_4',
4070
'breezy.tests.test_wsgi',
4071
'breezy.tests.test_xml',
4075
def _test_suite_modules_to_doctest():
4076
"""Return the list of modules to doctest."""
4078
# GZ 2009-03-31: No docstrings with -OO so there's nothing to doctest
4082
'breezy.branchbuilder',
4083
'breezy.bzr.inventory',
4084
'breezy.decorators',
4085
'breezy.iterablefile',
4090
'breezy.symbol_versioning',
4092
'breezy.tests.fixtures',
4094
'breezy.transport.http',
4095
'breezy.version_info_formats.format_custom',
4099
def test_suite(keep_only=None, starting_with=None):
4100
"""Build and return TestSuite for the whole of breezy.
4102
:param keep_only: A list of test ids limiting the suite returned.
4104
:param starting_with: An id limiting the suite returned to the tests
4107
This function can be replaced if you need to change the default test
4108
suite on a global basis, but it is not encouraged.
4111
loader = TestUtil.TestLoader()
4113
if keep_only is not None:
4114
id_filter = TestIdList(keep_only)
4116
# We take precedence over keep_only because *at loading time* using
4117
# both options means we will load less tests for the same final result.
4118
def interesting_module(name):
4119
for start in starting_with:
4121
# Either the module name starts with the specified string
4122
name.startswith(start)
4123
# or it may contain tests starting with the specified string
4124
or start.startswith(name)
4128
loader = TestUtil.FilteredByModuleTestLoader(interesting_module)
4130
elif keep_only is not None:
4131
loader = TestUtil.FilteredByModuleTestLoader(id_filter.refers_to)
4132
def interesting_module(name):
4133
return id_filter.refers_to(name)
4136
loader = TestUtil.TestLoader()
4137
def interesting_module(name):
4138
# No filtering, all modules are interesting
4141
suite = loader.suiteClass()
4143
# modules building their suite with loadTestsFromModuleNames
4144
suite.addTest(loader.loadTestsFromModuleNames(_test_suite_testmod_names()))
4146
for mod in _test_suite_modules_to_doctest():
4147
if not interesting_module(mod):
4148
# No tests to keep here, move along
4151
# note that this really does mean "report only" -- doctest
4152
# still runs the rest of the examples
4153
doc_suite = IsolatedDocTestSuite(
4154
mod, optionflags=doctest.REPORT_ONLY_FIRST_FAILURE)
4155
except ValueError as e:
4156
print('**failed to get doctest for: %s\n%s' % (mod, e))
4158
if len(doc_suite._tests) == 0:
4159
raise errors.BzrError("no doctests found in %s" % (mod,))
4160
suite.addTest(doc_suite)
4162
default_encoding = sys.getdefaultencoding()
4163
for name, plugin in _mod_plugin.plugins().items():
4164
if not interesting_module(plugin.module.__name__):
4166
plugin_suite = plugin.test_suite()
4167
# We used to catch ImportError here and turn it into just a warning,
4168
# but really if you don't have --no-plugins this should be a failure.
4169
# mbp 20080213 - see http://bugs.launchpad.net/bugs/189771
4170
if plugin_suite is None:
4171
plugin_suite = plugin.load_plugin_tests(loader)
4172
if plugin_suite is not None:
4173
suite.addTest(plugin_suite)
4174
if default_encoding != sys.getdefaultencoding():
4176
'Plugin "%s" tried to reset default encoding to: %s', name,
4177
sys.getdefaultencoding())
4179
sys.setdefaultencoding(default_encoding)
4181
if keep_only is not None:
4182
# Now that the referred modules have loaded their tests, keep only the
4184
suite = filter_suite_by_id_list(suite, id_filter)
4185
# Do some sanity checks on the id_list filtering
4186
not_found, duplicates = suite_matches_id_list(suite, keep_only)
4188
# The tester has used both keep_only and starting_with, so he is
4189
# already aware that some tests are excluded from the list, there
4190
# is no need to tell him which.
4193
# Some tests mentioned in the list are not in the test suite. The
4194
# list may be out of date, report to the tester.
4195
for id in not_found:
4196
trace.warning('"%s" not found in the test suite', id)
4197
for id in duplicates:
4198
trace.warning('"%s" is used as an id by several tests', id)
4203
def multiply_scenarios(*scenarios):
4204
"""Multiply two or more iterables of scenarios.
4206
It is safe to pass scenario generators or iterators.
4208
:returns: A list of compound scenarios: the cross-product of all
4209
scenarios, with the names concatenated and the parameters
4212
return functools.reduce(_multiply_two_scenarios, map(list, scenarios))
4215
def _multiply_two_scenarios(scenarios_left, scenarios_right):
4216
"""Multiply two sets of scenarios.
4218
:returns: the cartesian product of the two sets of scenarios, that is
4219
a scenario for every possible combination of a left scenario and a
4223
('%s,%s' % (left_name, right_name),
4224
dict(left_dict, **right_dict))
4225
for left_name, left_dict in scenarios_left
4226
for right_name, right_dict in scenarios_right]
4229
def multiply_tests(tests, scenarios, result):
4230
"""Multiply tests_list by scenarios into result.
4232
This is the core workhorse for test parameterisation.
4234
Typically the load_tests() method for a per-implementation test suite will
4235
call multiply_tests and return the result.
4237
:param tests: The tests to parameterise.
4238
:param scenarios: The scenarios to apply: pairs of (scenario_name,
4239
scenario_param_dict).
4240
:param result: A TestSuite to add created tests to.
4242
This returns the passed in result TestSuite with the cross product of all
4243
the tests repeated once for each scenario. Each test is adapted by adding
4244
the scenario name at the end of its id(), and updating the test object's
4245
__dict__ with the scenario_param_dict.
4247
>>> import breezy.tests.test_sampler
4248
>>> r = multiply_tests(
4249
... breezy.tests.test_sampler.DemoTest('test_nothing'),
4250
... [('one', dict(param=1)),
4251
... ('two', dict(param=2))],
4252
... TestUtil.TestSuite())
4253
>>> tests = list(iter_suite_tests(r))
4257
'breezy.tests.test_sampler.DemoTest.test_nothing(one)'
4263
for test in iter_suite_tests(tests):
4264
apply_scenarios(test, scenarios, result)
4268
def apply_scenarios(test, scenarios, result):
4269
"""Apply the scenarios in scenarios to test and add to result.
4271
:param test: The test to apply scenarios to.
4272
:param scenarios: An iterable of scenarios to apply to test.
4274
:seealso: apply_scenario
4276
for scenario in scenarios:
4277
result.addTest(apply_scenario(test, scenario))
4281
def apply_scenario(test, scenario):
4282
"""Copy test and apply scenario to it.
4284
:param test: A test to adapt.
4285
:param scenario: A tuple describing the scenario.
4286
The first element of the tuple is the new test id.
4287
The second element is a dict containing attributes to set on the
4289
:return: The adapted test.
4291
new_id = "%s(%s)" % (test.id(), scenario[0])
4292
new_test = clone_test(test, new_id)
4293
for name, value in scenario[1].items():
4294
setattr(new_test, name, value)
4298
def clone_test(test, new_id):
4299
"""Clone a test giving it a new id.
4301
:param test: The test to clone.
4302
:param new_id: The id to assign to it.
4303
:return: The new test.
4305
new_test = copy.copy(test)
4306
new_test.id = lambda: new_id
4307
# XXX: Workaround <https://bugs.launchpad.net/testtools/+bug/637725>, which
4308
# causes cloned tests to share the 'details' dict. This makes it hard to
4309
# read the test output for parameterized tests, because tracebacks will be
4310
# associated with irrelevant tests.
4312
details = new_test._TestCase__details
4313
except AttributeError:
4314
# must be a different version of testtools than expected. Do nothing.
4317
# Reset the '__details' dict.
4318
new_test._TestCase__details = {}
4322
def permute_tests_for_extension(standard_tests, loader, py_module_name,
4324
"""Helper for permutating tests against an extension module.
4326
This is meant to be used inside a modules 'load_tests()' function. It will
4327
create 2 scenarios, and cause all tests in the 'standard_tests' to be run
4328
against both implementations. Setting 'test.module' to the appropriate
4329
module. See breezy.tests.test__chk_map.load_tests as an example.
4331
:param standard_tests: A test suite to permute
4332
:param loader: A TestLoader
4333
:param py_module_name: The python path to a python module that can always
4334
be loaded, and will be considered the 'python' implementation. (eg
4335
'breezy._chk_map_py')
4336
:param ext_module_name: The python path to an extension module. If the
4337
module cannot be loaded, a single test will be added, which notes that
4338
the module is not available. If it can be loaded, all standard_tests
4339
will be run against that module.
4340
:return: (suite, feature) suite is a test-suite that has all the permuted
4341
tests. feature is the Feature object that can be used to determine if
4342
the module is available.
4345
from .features import ModuleAvailableFeature
4346
py_module = pyutils.get_named_object(py_module_name)
4348
('python', {'module': py_module}),
4350
suite = loader.suiteClass()
4351
feature = ModuleAvailableFeature(ext_module_name)
4352
if feature.available():
4353
scenarios.append(('C', {'module': feature.module}))
4355
# the compiled module isn't available, so we add a failing test
4356
class FailWithoutFeature(TestCase):
4357
def test_fail(self):
4358
self.requireFeature(feature)
4359
suite.addTest(loader.loadTestsFromTestCase(FailWithoutFeature))
4360
result = multiply_tests(standard_tests, scenarios, suite)
4361
return result, feature
4364
def _rmtree_temp_dir(dirname, test_id=None):
4365
# If LANG=C we probably have created some bogus paths
4366
# which rmtree(unicode) will fail to delete
4367
# so make sure we are using rmtree(str) to delete everything
4368
# except on win32, where rmtree(str) will fail
4369
# since it doesn't have the property of byte-stream paths
4370
# (they are either ascii or mbcs)
4371
if sys.platform == 'win32' and isinstance(dirname, bytes):
4372
# make sure we are using the unicode win32 api
4373
dirname = dirname.decode('mbcs')
4375
dirname = dirname.encode(sys.getfilesystemencoding())
4377
osutils.rmtree(dirname)
4378
except OSError as e:
4379
# We don't want to fail here because some useful display will be lost
4380
# otherwise. Polluting the tmp dir is bad, but not giving all the
4381
# possible info to the test runner is even worse.
4383
ui.ui_factory.clear_term()
4384
sys.stderr.write('\nWhile running: %s\n' % (test_id,))
4385
# Ugly, but the last thing we want here is fail, so bear with it.
4386
printable_e = str(e).decode(osutils.get_user_encoding(), 'replace'
4387
).encode('ascii', 'replace')
4388
sys.stderr.write('Unable to remove testing dir %s\n%s'
4389
% (os.path.basename(dirname), printable_e))
4392
def probe_unicode_in_user_encoding():
4393
"""Try to encode several unicode strings to use in unicode-aware tests.
4394
Return first successfull match.
4396
:return: (unicode value, encoded plain string value) or (None, None)
4398
possible_vals = [u'm\xb5', u'\xe1', u'\u0410']
4399
for uni_val in possible_vals:
4401
str_val = uni_val.encode(osutils.get_user_encoding())
4402
except UnicodeEncodeError:
4403
# Try a different character
4406
return uni_val, str_val
4410
def probe_bad_non_ascii(encoding):
4411
"""Try to find [bad] character with code [128..255]
4412
that cannot be decoded to unicode in some encoding.
4413
Return None if all non-ascii characters is valid
4416
for i in range(128, 256):
4419
char.decode(encoding)
4420
except UnicodeDecodeError:
4425
# Only define SubUnitBzrRunner if subunit is available.
4427
from subunit import TestProtocolClient
4428
from subunit.test_results import AutoTimingTestResultDecorator
4430
class SubUnitBzrProtocolClientv1(TestProtocolClient):
4432
def stopTest(self, test):
4433
super(SubUnitBzrProtocolClientv1, self).stopTest(test)
4434
_clear__type_equality_funcs(test)
4436
def addSuccess(self, test, details=None):
4437
# The subunit client always includes the details in the subunit
4438
# stream, but we don't want to include it in ours.
4439
if details is not None and 'log' in details:
4441
return super(SubUnitBzrProtocolClientv1, self).addSuccess(
4444
class SubUnitBzrRunnerv1(TextTestRunner):
4446
def run(self, test):
4447
result = AutoTimingTestResultDecorator(
4448
SubUnitBzrProtocolClientv1(self.stream))
4456
from subunit.run import SubunitTestRunner
4458
class SubUnitBzrRunnerv2(TextTestRunner, SubunitTestRunner):
4460
def __init__(self, stream=sys.stderr, descriptions=0, verbosity=1,
4461
bench_history=None, strict=False, result_decorators=None):
4462
TextTestRunner.__init__(
4463
self, stream=stream,
4464
descriptions=descriptions, verbosity=verbosity,
4465
bench_history=bench_history, strict=strict,
4466
result_decorators=result_decorators)
4467
SubunitTestRunner.__init__(self, verbosity=verbosity,
4470
run = SubunitTestRunner.run
added
removed
breezy/tests/__init__.py
