1
# Copyright (C) 2005-2010 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
# TODO: Perhaps there should be an API to find out if bzr running under the
20
# test suite -- some plugins might want to avoid making intrusive changes if
21
# this is the case. However, we want behaviour under to test to diverge as
22
# little as possible, so this should be used rarely if it's added at all.
23
# (Suggestion from j-a-meinel, 2005-11-24)
25
# NOTE: Some classes in here use camelCaseNaming() rather than
26
# underscore_naming(). That's for consistency with unittest; it's not the
27
# general style of bzrlib. Please continue that consistency when adding e.g.
28
# new assertFoo() methods.
33
from cStringIO import StringIO
56
# nb: check this before importing anything else from within it
57
_testtools_version = getattr(testtools, '__version__', ())
58
if _testtools_version < (0, 9, 2):
59
raise ImportError("need at least testtools 0.9.2: %s is %r"
60
% (testtools.__file__, _testtools_version))
61
from testtools import content
77
transport as _mod_transport,
81
import bzrlib.commands
82
import bzrlib.timestamp
84
import bzrlib.inventory
85
import bzrlib.iterablefile
90
# lsprof not available
92
from bzrlib.merge import merge_inner
95
from bzrlib.smart import client, request, server
97
from bzrlib import symbol_versioning
98
from bzrlib.symbol_versioning import (
106
from bzrlib.transport import (
110
from bzrlib.trace import mutter, note
111
from bzrlib.tests import (
116
from bzrlib.ui import NullProgressView
117
from bzrlib.ui.text import TextUIFactory
118
import bzrlib.version_info_formats.format_custom
119
from bzrlib.workingtree import WorkingTree, WorkingTreeFormat2
121
# Mark this python module as being part of the implementation
122
# of unittest: this gives us better tracebacks where the last
123
# shown frame is the test code, not our assertXYZ.
126
default_transport = test_server.LocalURLServer
129
_unitialized_attr = object()
130
"""A sentinel needed to act as a default value in a method signature."""
133
# Subunit result codes, defined here to prevent a hard dependency on subunit.
137
# These are intentionally brought into this namespace. That way plugins, etc
138
# can just "from bzrlib.tests import TestCase, TestLoader, etc"
139
TestSuite = TestUtil.TestSuite
140
TestLoader = TestUtil.TestLoader
142
class ExtendedTestResult(testtools.TextTestResult):
143
"""Accepts, reports and accumulates the results of running tests.
145
Compared to the unittest version this class adds support for
146
profiling, benchmarking, stopping as soon as a test fails, and
147
skipping tests. There are further-specialized subclasses for
148
different types of display.
150
When a test finishes, in whatever way, it calls one of the addSuccess,
151
addFailure or addError classes. These in turn may redirect to a more
152
specific case for the special test results supported by our extended
155
Note that just one of these objects is fed the results from many tests.
160
def __init__(self, stream, descriptions, verbosity,
164
"""Construct new TestResult.
166
:param bench_history: Optionally, a writable file object to accumulate
169
testtools.TextTestResult.__init__(self, stream)
170
if bench_history is not None:
171
from bzrlib.version import _get_bzr_source_tree
172
src_tree = _get_bzr_source_tree()
175
revision_id = src_tree.get_parent_ids()[0]
177
# XXX: if this is a brand new tree, do the same as if there
181
# XXX: If there's no branch, what should we do?
183
bench_history.write("--date %s %s\n" % (time.time(), revision_id))
184
self._bench_history = bench_history
185
self.ui = ui.ui_factory
188
self.failure_count = 0
189
self.known_failure_count = 0
191
self.not_applicable_count = 0
192
self.unsupported = {}
194
self._overall_start_time = time.time()
195
self._strict = strict
196
self._first_thread_leaker_id = None
197
self._tests_leaking_threads_count = 0
199
def stopTestRun(self):
202
stopTime = time.time()
203
timeTaken = stopTime - self.startTime
204
# GZ 2010-07-19: Seems testtools has no printErrors method, and though
205
# the parent class method is similar have to duplicate
206
self._show_list('ERROR', self.errors)
207
self._show_list('FAIL', self.failures)
208
self.stream.write(self.sep2)
209
self.stream.write("%s %d test%s in %.3fs\n\n" % (actionTaken,
210
run, run != 1 and "s" or "", timeTaken))
211
if not self.wasSuccessful():
212
self.stream.write("FAILED (")
213
failed, errored = map(len, (self.failures, self.errors))
215
self.stream.write("failures=%d" % failed)
217
if failed: self.stream.write(", ")
218
self.stream.write("errors=%d" % errored)
219
if self.known_failure_count:
220
if failed or errored: self.stream.write(", ")
221
self.stream.write("known_failure_count=%d" %
222
self.known_failure_count)
223
self.stream.write(")\n")
225
if self.known_failure_count:
226
self.stream.write("OK (known_failures=%d)\n" %
227
self.known_failure_count)
229
self.stream.write("OK\n")
230
if self.skip_count > 0:
231
skipped = self.skip_count
232
self.stream.write('%d test%s skipped\n' %
233
(skipped, skipped != 1 and "s" or ""))
235
for feature, count in sorted(self.unsupported.items()):
236
self.stream.write("Missing feature '%s' skipped %d tests.\n" %
239
ok = self.wasStrictlySuccessful()
241
ok = self.wasSuccessful()
242
if self._first_thread_leaker_id:
244
'%s is leaking threads among %d leaking tests.\n' % (
245
self._first_thread_leaker_id,
246
self._tests_leaking_threads_count))
247
# We don't report the main thread as an active one.
249
'%d non-main threads were left active in the end.\n'
250
% (len(self._active_threads) - 1))
252
def getDescription(self, test):
255
def _extractBenchmarkTime(self, testCase, details=None):
256
"""Add a benchmark time for the current test case."""
257
if details and 'benchtime' in details:
258
return float(''.join(details['benchtime'].iter_bytes()))
259
return getattr(testCase, "_benchtime", None)
261
def _elapsedTestTimeString(self):
262
"""Return a time string for the overall time the current test has taken."""
263
return self._formatTime(self._delta_to_float(
264
self._now() - self._start_datetime))
266
def _testTimeString(self, testCase):
267
benchmark_time = self._extractBenchmarkTime(testCase)
268
if benchmark_time is not None:
269
return self._formatTime(benchmark_time) + "*"
271
return self._elapsedTestTimeString()
273
def _formatTime(self, seconds):
274
"""Format seconds as milliseconds with leading spaces."""
275
# some benchmarks can take thousands of seconds to run, so we need 8
277
return "%8dms" % (1000 * seconds)
279
def _shortened_test_description(self, test):
281
what = re.sub(r'^bzrlib\.tests\.', '', what)
284
def startTest(self, test):
285
super(ExtendedTestResult, self).startTest(test)
289
self.report_test_start(test)
290
test.number = self.count
291
self._recordTestStartTime()
292
# Only check for thread leaks if the test case supports cleanups
293
addCleanup = getattr(test, "addCleanup", None)
294
if addCleanup is not None:
295
addCleanup(self._check_leaked_threads, test)
297
def startTests(self):
298
self.report_tests_starting()
299
self._active_threads = threading.enumerate()
301
def _check_leaked_threads(self, test):
302
"""See if any threads have leaked since last call
304
A sample of live threads is stored in the _active_threads attribute,
305
when this method runs it compares the current live threads and any not
306
in the previous sample are treated as having leaked.
308
now_active_threads = set(threading.enumerate())
309
threads_leaked = now_active_threads.difference(self._active_threads)
311
self._report_thread_leak(test, threads_leaked, now_active_threads)
312
self._tests_leaking_threads_count += 1
313
if self._first_thread_leaker_id is None:
314
self._first_thread_leaker_id = test.id()
315
self._active_threads = now_active_threads
317
def _recordTestStartTime(self):
318
"""Record that a test has started."""
319
self._start_datetime = self._now()
321
def addError(self, test, err):
322
"""Tell result that test finished with an error.
324
Called from the TestCase run() method when the test
325
fails with an unexpected error.
328
super(ExtendedTestResult, self).addError(test, err)
329
self.error_count += 1
330
self.report_error(test, err)
334
def addFailure(self, test, err):
335
"""Tell result that test failed.
337
Called from the TestCase run() method when the test
338
fails because e.g. an assert() method failed.
341
super(ExtendedTestResult, self).addFailure(test, err)
342
self.failure_count += 1
343
self.report_failure(test, err)
347
def addSuccess(self, test, details=None):
348
"""Tell result that test completed successfully.
350
Called from the TestCase run()
352
if self._bench_history is not None:
353
benchmark_time = self._extractBenchmarkTime(test, details)
354
if benchmark_time is not None:
355
self._bench_history.write("%s %s\n" % (
356
self._formatTime(benchmark_time),
358
self.report_success(test)
359
super(ExtendedTestResult, self).addSuccess(test)
360
test._log_contents = ''
362
def addExpectedFailure(self, test, err):
363
self.known_failure_count += 1
364
self.report_known_failure(test, err)
366
def addNotSupported(self, test, feature):
367
"""The test will not be run because of a missing feature.
369
# this can be called in two different ways: it may be that the
370
# test started running, and then raised (through requireFeature)
371
# UnavailableFeature. Alternatively this method can be called
372
# while probing for features before running the test code proper; in
373
# that case we will see startTest and stopTest, but the test will
374
# never actually run.
375
self.unsupported.setdefault(str(feature), 0)
376
self.unsupported[str(feature)] += 1
377
self.report_unsupported(test, feature)
379
def addSkip(self, test, reason):
380
"""A test has not run for 'reason'."""
382
self.report_skip(test, reason)
384
def addNotApplicable(self, test, reason):
385
self.not_applicable_count += 1
386
self.report_not_applicable(test, reason)
388
def _post_mortem(self):
389
"""Start a PDB post mortem session."""
390
if os.environ.get('BZR_TEST_PDB', None):
391
import pdb;pdb.post_mortem()
393
def progress(self, offset, whence):
394
"""The test is adjusting the count of tests to run."""
395
if whence == SUBUNIT_SEEK_SET:
396
self.num_tests = offset
397
elif whence == SUBUNIT_SEEK_CUR:
398
self.num_tests += offset
400
raise errors.BzrError("Unknown whence %r" % whence)
402
def report_tests_starting(self):
403
"""Display information before the test run begins"""
404
if getattr(sys, 'frozen', None) is None:
405
bzr_path = osutils.realpath(sys.argv[0])
407
bzr_path = sys.executable
409
'bzr selftest: %s\n' % (bzr_path,))
412
bzrlib.__path__[0],))
414
' bzr-%s python-%s %s\n' % (
415
bzrlib.version_string,
416
bzrlib._format_version_tuple(sys.version_info),
417
platform.platform(aliased=1),
419
self.stream.write('\n')
421
def report_test_start(self, test):
422
"""Display information on the test just about to be run"""
424
def _report_thread_leak(self, test, leaked_threads, active_threads):
425
"""Display information on a test that leaked one or more threads"""
426
# GZ 2010-09-09: A leak summary reported separately from the general
427
# thread debugging would be nice. Tests under subunit
428
# need something not using stream, perhaps adding a
429
# testtools details object would be fitting.
430
if 'threads' in selftest_debug_flags:
431
self.stream.write('%s is leaking, active is now %d\n' %
432
(test.id(), len(active_threads)))
434
def startTestRun(self):
435
self.startTime = time.time()
437
def report_success(self, test):
440
def wasStrictlySuccessful(self):
441
if self.unsupported or self.known_failure_count:
443
return self.wasSuccessful()
446
class TextTestResult(ExtendedTestResult):
447
"""Displays progress and results of tests in text form"""
449
def __init__(self, stream, descriptions, verbosity,
454
ExtendedTestResult.__init__(self, stream, descriptions, verbosity,
455
bench_history, strict)
456
# We no longer pass them around, but just rely on the UIFactory stack
459
warnings.warn("Passing pb to TextTestResult is deprecated")
460
self.pb = self.ui.nested_progress_bar()
461
self.pb.show_pct = False
462
self.pb.show_spinner = False
463
self.pb.show_eta = False,
464
self.pb.show_count = False
465
self.pb.show_bar = False
466
self.pb.update_latency = 0
467
self.pb.show_transport_activity = False
469
def stopTestRun(self):
470
# called when the tests that are going to run have run
473
super(TextTestResult, self).stopTestRun()
475
def report_tests_starting(self):
476
super(TextTestResult, self).report_tests_starting()
477
self.pb.update('[test 0/%d] Starting' % (self.num_tests))
479
def _progress_prefix_text(self):
480
# the longer this text, the less space we have to show the test
482
a = '[%d' % self.count # total that have been run
483
# tests skipped as known not to be relevant are not important enough
485
## if self.skip_count:
486
## a += ', %d skip' % self.skip_count
487
## if self.known_failure_count:
488
## a += '+%dX' % self.known_failure_count
490
a +='/%d' % self.num_tests
492
runtime = time.time() - self._overall_start_time
494
a += '%dm%ds' % (runtime / 60, runtime % 60)
497
total_fail_count = self.error_count + self.failure_count
499
a += ', %d failed' % total_fail_count
500
# if self.unsupported:
501
# a += ', %d missing' % len(self.unsupported)
505
def report_test_start(self, test):
507
self._progress_prefix_text()
509
+ self._shortened_test_description(test))
511
def _test_description(self, test):
512
return self._shortened_test_description(test)
514
def report_error(self, test, err):
515
self.stream.write('ERROR: %s\n %s\n' % (
516
self._test_description(test),
520
def report_failure(self, test, err):
521
self.stream.write('FAIL: %s\n %s\n' % (
522
self._test_description(test),
526
def report_known_failure(self, test, err):
529
def report_skip(self, test, reason):
532
def report_not_applicable(self, test, reason):
535
def report_unsupported(self, test, feature):
536
"""test cannot be run because feature is missing."""
539
class VerboseTestResult(ExtendedTestResult):
540
"""Produce long output, with one line per test run plus times"""
542
def _ellipsize_to_right(self, a_string, final_width):
543
"""Truncate and pad a string, keeping the right hand side"""
544
if len(a_string) > final_width:
545
result = '...' + a_string[3-final_width:]
548
return result.ljust(final_width)
550
def report_tests_starting(self):
551
self.stream.write('running %d tests...\n' % self.num_tests)
552
super(VerboseTestResult, self).report_tests_starting()
554
def report_test_start(self, test):
555
name = self._shortened_test_description(test)
556
width = osutils.terminal_width()
557
if width is not None:
558
# width needs space for 6 char status, plus 1 for slash, plus an
559
# 11-char time string, plus a trailing blank
560
# when NUMBERED_DIRS: plus 5 chars on test number, plus 1 char on
562
self.stream.write(self._ellipsize_to_right(name, width-18))
564
self.stream.write(name)
567
def _error_summary(self, err):
569
return '%s%s' % (indent, err[1])
571
def report_error(self, test, err):
572
self.stream.write('ERROR %s\n%s\n'
573
% (self._testTimeString(test),
574
self._error_summary(err)))
576
def report_failure(self, test, err):
577
self.stream.write(' FAIL %s\n%s\n'
578
% (self._testTimeString(test),
579
self._error_summary(err)))
581
def report_known_failure(self, test, err):
582
self.stream.write('XFAIL %s\n%s\n'
583
% (self._testTimeString(test),
584
self._error_summary(err)))
586
def report_success(self, test):
587
self.stream.write(' OK %s\n' % self._testTimeString(test))
588
for bench_called, stats in getattr(test, '_benchcalls', []):
589
self.stream.write('LSProf output for %s(%s, %s)\n' % bench_called)
590
stats.pprint(file=self.stream)
591
# flush the stream so that we get smooth output. This verbose mode is
592
# used to show the output in PQM.
595
def report_skip(self, test, reason):
596
self.stream.write(' SKIP %s\n%s\n'
597
% (self._testTimeString(test), reason))
599
def report_not_applicable(self, test, reason):
600
self.stream.write(' N/A %s\n %s\n'
601
% (self._testTimeString(test), reason))
603
def report_unsupported(self, test, feature):
604
"""test cannot be run because feature is missing."""
605
self.stream.write("NODEP %s\n The feature '%s' is not available.\n"
606
%(self._testTimeString(test), feature))
609
class TextTestRunner(object):
610
stop_on_failure = False
618
result_decorators=None,
620
"""Create a TextTestRunner.
622
:param result_decorators: An optional list of decorators to apply
623
to the result object being used by the runner. Decorators are
624
applied left to right - the first element in the list is the
627
# stream may know claim to know to write unicode strings, but in older
628
# pythons this goes sufficiently wrong that it is a bad idea. (
629
# specifically a built in file with encoding 'UTF-8' will still try
630
# to encode using ascii.
631
new_encoding = osutils.get_terminal_encoding()
632
codec = codecs.lookup(new_encoding)
633
if type(codec) is tuple:
637
encode = codec.encode
638
stream = osutils.UnicodeOrBytesToBytesWriter(encode, stream)
639
stream.encoding = new_encoding
641
self.descriptions = descriptions
642
self.verbosity = verbosity
643
self._bench_history = bench_history
644
self._strict = strict
645
self._result_decorators = result_decorators or []
648
"Run the given test case or test suite."
649
if self.verbosity == 1:
650
result_class = TextTestResult
651
elif self.verbosity >= 2:
652
result_class = VerboseTestResult
653
original_result = result_class(self.stream,
656
bench_history=self._bench_history,
659
# Signal to result objects that look at stop early policy to stop,
660
original_result.stop_early = self.stop_on_failure
661
result = original_result
662
for decorator in self._result_decorators:
663
result = decorator(result)
664
result.stop_early = self.stop_on_failure
665
result.startTestRun()
670
# higher level code uses our extended protocol to determine
671
# what exit code to give.
672
return original_result
675
def iter_suite_tests(suite):
676
"""Return all tests in a suite, recursing through nested suites"""
677
if isinstance(suite, unittest.TestCase):
679
elif isinstance(suite, unittest.TestSuite):
681
for r in iter_suite_tests(item):
684
raise Exception('unknown type %r for object %r'
685
% (type(suite), suite))
688
TestSkipped = testtools.testcase.TestSkipped
691
class TestNotApplicable(TestSkipped):
692
"""A test is not applicable to the situation where it was run.
694
This is only normally raised by parameterized tests, if they find that
695
the instance they're constructed upon does not support one aspect
700
# traceback._some_str fails to format exceptions that have the default
701
# __str__ which does an implicit ascii conversion. However, repr() on those
702
# objects works, for all that its not quite what the doctor may have ordered.
703
def _clever_some_str(value):
708
return repr(value).replace('\\n', '\n')
710
return '<unprintable %s object>' % type(value).__name__
712
traceback._some_str = _clever_some_str
715
# deprecated - use self.knownFailure(), or self.expectFailure.
716
KnownFailure = testtools.testcase._ExpectedFailure
719
class UnavailableFeature(Exception):
720
"""A feature required for this test was not available.
722
This can be considered a specialised form of SkippedTest.
724
The feature should be used to construct the exception.
728
class StringIOWrapper(object):
729
"""A wrapper around cStringIO which just adds an encoding attribute.
731
Internally we can check sys.stdout to see what the output encoding
732
should be. However, cStringIO has no encoding attribute that we can
733
set. So we wrap it instead.
738
def __init__(self, s=None):
740
self.__dict__['_cstring'] = StringIO(s)
742
self.__dict__['_cstring'] = StringIO()
744
def __getattr__(self, name, getattr=getattr):
745
return getattr(self.__dict__['_cstring'], name)
747
def __setattr__(self, name, val):
748
if name == 'encoding':
749
self.__dict__['encoding'] = val
751
return setattr(self._cstring, name, val)
754
class TestUIFactory(TextUIFactory):
755
"""A UI Factory for testing.
757
Hide the progress bar but emit note()s.
759
Allows get_password to be tested without real tty attached.
761
See also CannedInputUIFactory which lets you provide programmatic input in
764
# TODO: Capture progress events at the model level and allow them to be
765
# observed by tests that care.
767
# XXX: Should probably unify more with CannedInputUIFactory or a
768
# particular configuration of TextUIFactory, or otherwise have a clearer
769
# idea of how they're supposed to be different.
770
# See https://bugs.launchpad.net/bzr/+bug/408213
772
def __init__(self, stdout=None, stderr=None, stdin=None):
773
if stdin is not None:
774
# We use a StringIOWrapper to be able to test various
775
# encodings, but the user is still responsible to
776
# encode the string and to set the encoding attribute
777
# of StringIOWrapper.
778
stdin = StringIOWrapper(stdin)
779
super(TestUIFactory, self).__init__(stdin, stdout, stderr)
781
def get_non_echoed_password(self):
782
"""Get password from stdin without trying to handle the echo mode"""
783
password = self.stdin.readline()
786
if password[-1] == '\n':
787
password = password[:-1]
790
def make_progress_view(self):
791
return NullProgressView()
794
class TestCase(testtools.TestCase):
795
"""Base class for bzr unit tests.
797
Tests that need access to disk resources should subclass
798
TestCaseInTempDir not TestCase.
800
Error and debug log messages are redirected from their usual
801
location into a temporary file, the contents of which can be
802
retrieved by _get_log(). We use a real OS file, not an in-memory object,
803
so that it can also capture file IO. When the test completes this file
804
is read into memory and removed from disk.
806
There are also convenience functions to invoke bzr's command-line
807
routine, and to build and check bzr trees.
809
In addition to the usual method of overriding tearDown(), this class also
810
allows subclasses to register cleanup functions via addCleanup, which are
811
run in order as the object is torn down. It's less likely this will be
812
accidentally overlooked.
816
# record lsprof data when performing benchmark calls.
817
_gather_lsprof_in_benchmarks = False
819
def __init__(self, methodName='testMethod'):
820
super(TestCase, self).__init__(methodName)
821
self._directory_isolation = True
822
self.exception_handlers.insert(0,
823
(UnavailableFeature, self._do_unsupported_or_skip))
824
self.exception_handlers.insert(0,
825
(TestNotApplicable, self._do_not_applicable))
828
super(TestCase, self).setUp()
829
for feature in getattr(self, '_test_needs_features', []):
830
self.requireFeature(feature)
831
self._log_contents = None
832
self.addDetail("log", content.Content(content.ContentType("text",
833
"plain", {"charset": "utf8"}),
834
lambda:[self._get_log(keep_log_file=True)]))
835
self._cleanEnvironment()
838
self._benchcalls = []
839
self._benchtime = None
841
self._track_transports()
843
self._clear_debug_flags()
848
pdb.Pdb().set_trace(sys._getframe().f_back)
850
def discardDetail(self, name):
851
"""Extend the addDetail, getDetails api so we can remove a detail.
853
eg. bzr always adds the 'log' detail at startup, but we don't want to
854
include it for skipped, xfail, etc tests.
856
It is safe to call this for a detail that doesn't exist, in case this
857
gets called multiple times.
859
# We cheat. details is stored in __details which means we shouldn't
860
# touch it. but getDetails() returns the dict directly, so we can
862
details = self.getDetails()
866
def _clear_debug_flags(self):
867
"""Prevent externally set debug flags affecting tests.
869
Tests that want to use debug flags can just set them in the
870
debug_flags set during setup/teardown.
872
# Start with a copy of the current debug flags we can safely modify.
873
self.overrideAttr(debug, 'debug_flags', set(debug.debug_flags))
874
if 'allow_debug' not in selftest_debug_flags:
875
debug.debug_flags.clear()
876
if 'disable_lock_checks' not in selftest_debug_flags:
877
debug.debug_flags.add('strict_locks')
879
def _clear_hooks(self):
880
# prevent hooks affecting tests
881
self._preserved_hooks = {}
882
for key, factory in hooks.known_hooks.items():
883
parent, name = hooks.known_hooks_key_to_parent_and_attribute(key)
884
current_hooks = hooks.known_hooks_key_to_object(key)
885
self._preserved_hooks[parent] = (name, current_hooks)
886
self.addCleanup(self._restoreHooks)
887
for key, factory in hooks.known_hooks.items():
888
parent, name = hooks.known_hooks_key_to_parent_and_attribute(key)
889
setattr(parent, name, factory())
890
# this hook should always be installed
891
request._install_hook()
893
def disable_directory_isolation(self):
894
"""Turn off directory isolation checks."""
895
self._directory_isolation = False
897
def enable_directory_isolation(self):
898
"""Enable directory isolation checks."""
899
self._directory_isolation = True
901
def _silenceUI(self):
902
"""Turn off UI for duration of test"""
903
# by default the UI is off; tests can turn it on if they want it.
904
self.overrideAttr(ui, 'ui_factory', ui.SilentUIFactory())
906
def _check_locks(self):
907
"""Check that all lock take/release actions have been paired."""
908
# We always check for mismatched locks. If a mismatch is found, we
909
# fail unless -Edisable_lock_checks is supplied to selftest, in which
910
# case we just print a warning.
912
acquired_locks = [lock for action, lock in self._lock_actions
913
if action == 'acquired']
914
released_locks = [lock for action, lock in self._lock_actions
915
if action == 'released']
916
broken_locks = [lock for action, lock in self._lock_actions
917
if action == 'broken']
918
# trivially, given the tests for lock acquistion and release, if we
919
# have as many in each list, it should be ok. Some lock tests also
920
# break some locks on purpose and should be taken into account by
921
# considering that breaking a lock is just a dirty way of releasing it.
922
if len(acquired_locks) != (len(released_locks) + len(broken_locks)):
923
message = ('Different number of acquired and '
924
'released or broken locks. (%s, %s + %s)' %
925
(acquired_locks, released_locks, broken_locks))
926
if not self._lock_check_thorough:
927
# Rather than fail, just warn
928
print "Broken test %s: %s" % (self, message)
932
def _track_locks(self):
933
"""Track lock activity during tests."""
934
self._lock_actions = []
935
if 'disable_lock_checks' in selftest_debug_flags:
936
self._lock_check_thorough = False
938
self._lock_check_thorough = True
940
self.addCleanup(self._check_locks)
941
_mod_lock.Lock.hooks.install_named_hook('lock_acquired',
942
self._lock_acquired, None)
943
_mod_lock.Lock.hooks.install_named_hook('lock_released',
944
self._lock_released, None)
945
_mod_lock.Lock.hooks.install_named_hook('lock_broken',
946
self._lock_broken, None)
948
def _lock_acquired(self, result):
949
self._lock_actions.append(('acquired', result))
951
def _lock_released(self, result):
952
self._lock_actions.append(('released', result))
954
def _lock_broken(self, result):
955
self._lock_actions.append(('broken', result))
957
def permit_dir(self, name):
958
"""Permit a directory to be used by this test. See permit_url."""
959
name_transport = _mod_transport.get_transport(name)
960
self.permit_url(name)
961
self.permit_url(name_transport.base)
963
def permit_url(self, url):
964
"""Declare that url is an ok url to use in this test.
966
Do this for memory transports, temporary test directory etc.
968
Do not do this for the current working directory, /tmp, or any other
969
preexisting non isolated url.
971
if not url.endswith('/'):
973
self._bzr_selftest_roots.append(url)
975
def permit_source_tree_branch_repo(self):
976
"""Permit the source tree bzr is running from to be opened.
978
Some code such as bzrlib.version attempts to read from the bzr branch
979
that bzr is executing from (if any). This method permits that directory
980
to be used in the test suite.
982
path = self.get_source_path()
983
self.record_directory_isolation()
986
workingtree.WorkingTree.open(path)
987
except (errors.NotBranchError, errors.NoWorkingTree):
988
raise TestSkipped('Needs a working tree of bzr sources')
990
self.enable_directory_isolation()
992
def _preopen_isolate_transport(self, transport):
993
"""Check that all transport openings are done in the test work area."""
994
while isinstance(transport, pathfilter.PathFilteringTransport):
995
# Unwrap pathfiltered transports
996
transport = transport.server.backing_transport.clone(
997
transport._filter('.'))
999
# ReadonlySmartTCPServer_for_testing decorates the backing transport
1000
# urls it is given by prepending readonly+. This is appropriate as the
1001
# client shouldn't know that the server is readonly (or not readonly).
1002
# We could register all servers twice, with readonly+ prepending, but
1003
# that makes for a long list; this is about the same but easier to
1005
if url.startswith('readonly+'):
1006
url = url[len('readonly+'):]
1007
self._preopen_isolate_url(url)
1009
def _preopen_isolate_url(self, url):
1010
if not self._directory_isolation:
1012
if self._directory_isolation == 'record':
1013
self._bzr_selftest_roots.append(url)
1015
# This prevents all transports, including e.g. sftp ones backed on disk
1016
# from working unless they are explicitly granted permission. We then
1017
# depend on the code that sets up test transports to check that they are
1018
# appropriately isolated and enable their use by calling
1019
# self.permit_transport()
1020
if not osutils.is_inside_any(self._bzr_selftest_roots, url):
1021
raise errors.BzrError("Attempt to escape test isolation: %r %r"
1022
% (url, self._bzr_selftest_roots))
1024
def record_directory_isolation(self):
1025
"""Gather accessed directories to permit later access.
1027
This is used for tests that access the branch bzr is running from.
1029
self._directory_isolation = "record"
1031
def start_server(self, transport_server, backing_server=None):
1032
"""Start transport_server for this test.
1034
This starts the server, registers a cleanup for it and permits the
1035
server's urls to be used.
1037
if backing_server is None:
1038
transport_server.start_server()
1040
transport_server.start_server(backing_server)
1041
self.addCleanup(transport_server.stop_server)
1042
# Obtain a real transport because if the server supplies a password, it
1043
# will be hidden from the base on the client side.
1044
t = _mod_transport.get_transport(transport_server.get_url())
1045
# Some transport servers effectively chroot the backing transport;
1046
# others like SFTPServer don't - users of the transport can walk up the
1047
# transport to read the entire backing transport. This wouldn't matter
1048
# except that the workdir tests are given - and that they expect the
1049
# server's url to point at - is one directory under the safety net. So
1050
# Branch operations into the transport will attempt to walk up one
1051
# directory. Chrooting all servers would avoid this but also mean that
1052
# we wouldn't be testing directly against non-root urls. Alternatively
1053
# getting the test framework to start the server with a backing server
1054
# at the actual safety net directory would work too, but this then
1055
# means that the self.get_url/self.get_transport methods would need
1056
# to transform all their results. On balance its cleaner to handle it
1057
# here, and permit a higher url when we have one of these transports.
1058
if t.base.endswith('/work/'):
1059
# we have safety net/test root/work
1060
t = t.clone('../..')
1061
elif isinstance(transport_server,
1062
test_server.SmartTCPServer_for_testing):
1063
# The smart server adds a path similar to work, which is traversed
1064
# up from by the client. But the server is chrooted - the actual
1065
# backing transport is not escaped from, and VFS requests to the
1066
# root will error (because they try to escape the chroot).
1068
while t2.base != t.base:
1071
self.permit_url(t.base)
1073
def _track_transports(self):
1074
"""Install checks for transport usage."""
1075
# TestCase has no safe place it can write to.
1076
self._bzr_selftest_roots = []
1077
# Currently the easiest way to be sure that nothing is going on is to
1078
# hook into bzr dir opening. This leaves a small window of error for
1079
# transport tests, but they are well known, and we can improve on this
1081
bzrdir.BzrDir.hooks.install_named_hook("pre_open",
1082
self._preopen_isolate_transport, "Check bzr directories are safe.")
1084
def _ndiff_strings(self, a, b):
1085
"""Return ndiff between two strings containing lines.
1087
A trailing newline is added if missing to make the strings
1089
if b and b[-1] != '\n':
1091
if a and a[-1] != '\n':
1093
difflines = difflib.ndiff(a.splitlines(True),
1095
linejunk=lambda x: False,
1096
charjunk=lambda x: False)
1097
return ''.join(difflines)
1099
def assertEqual(self, a, b, message=''):
1103
except UnicodeError, e:
1104
# If we can't compare without getting a UnicodeError, then
1105
# obviously they are different
1106
mutter('UnicodeError: %s', e)
1109
raise AssertionError("%snot equal:\na = %s\nb = %s\n"
1111
pprint.pformat(a), pprint.pformat(b)))
1113
assertEquals = assertEqual
1115
def assertEqualDiff(self, a, b, message=None):
1116
"""Assert two texts are equal, if not raise an exception.
1118
This is intended for use with multi-line strings where it can
1119
be hard to find the differences by eye.
1121
# TODO: perhaps override assertEquals to call this for strings?
1125
message = "texts not equal:\n"
1127
message = 'first string is missing a final newline.\n'
1129
message = 'second string is missing a final newline.\n'
1130
raise AssertionError(message +
1131
self._ndiff_strings(a, b))
1133
def assertEqualMode(self, mode, mode_test):
1134
self.assertEqual(mode, mode_test,
1135
'mode mismatch %o != %o' % (mode, mode_test))
1137
def assertEqualStat(self, expected, actual):
1138
"""assert that expected and actual are the same stat result.
1140
:param expected: A stat result.
1141
:param actual: A stat result.
1142
:raises AssertionError: If the expected and actual stat values differ
1143
other than by atime.
1145
self.assertEqual(expected.st_size, actual.st_size,
1146
'st_size did not match')
1147
self.assertEqual(expected.st_mtime, actual.st_mtime,
1148
'st_mtime did not match')
1149
self.assertEqual(expected.st_ctime, actual.st_ctime,
1150
'st_ctime did not match')
1151
if sys.platform != 'win32':
1152
# On Win32 both 'dev' and 'ino' cannot be trusted. In python2.4 it
1153
# is 'dev' that varies, in python 2.5 (6?) it is st_ino that is
1154
# odd. Regardless we shouldn't actually try to assert anything
1155
# about their values
1156
self.assertEqual(expected.st_dev, actual.st_dev,
1157
'st_dev did not match')
1158
self.assertEqual(expected.st_ino, actual.st_ino,
1159
'st_ino did not match')
1160
self.assertEqual(expected.st_mode, actual.st_mode,
1161
'st_mode did not match')
1163
def assertLength(self, length, obj_with_len):
1164
"""Assert that obj_with_len is of length length."""
1165
if len(obj_with_len) != length:
1166
self.fail("Incorrect length: wanted %d, got %d for %r" % (
1167
length, len(obj_with_len), obj_with_len))
1169
def assertLogsError(self, exception_class, func, *args, **kwargs):
1170
"""Assert that func(*args, **kwargs) quietly logs a specific exception.
1172
from bzrlib import trace
1174
orig_log_exception_quietly = trace.log_exception_quietly
1177
orig_log_exception_quietly()
1178
captured.append(sys.exc_info())
1179
trace.log_exception_quietly = capture
1180
func(*args, **kwargs)
1182
trace.log_exception_quietly = orig_log_exception_quietly
1183
self.assertLength(1, captured)
1184
err = captured[0][1]
1185
self.assertIsInstance(err, exception_class)
1188
def assertPositive(self, val):
1189
"""Assert that val is greater than 0."""
1190
self.assertTrue(val > 0, 'expected a positive value, but got %s' % val)
1192
def assertNegative(self, val):
1193
"""Assert that val is less than 0."""
1194
self.assertTrue(val < 0, 'expected a negative value, but got %s' % val)
1196
def assertStartsWith(self, s, prefix):
1197
if not s.startswith(prefix):
1198
raise AssertionError('string %r does not start with %r' % (s, prefix))
1200
def assertEndsWith(self, s, suffix):
1201
"""Asserts that s ends with suffix."""
1202
if not s.endswith(suffix):
1203
raise AssertionError('string %r does not end with %r' % (s, suffix))
1205
def assertContainsRe(self, haystack, needle_re, flags=0):
1206
"""Assert that a contains something matching a regular expression."""
1207
if not re.search(needle_re, haystack, flags):
1208
if '\n' in haystack or len(haystack) > 60:
1209
# a long string, format it in a more readable way
1210
raise AssertionError(
1211
'pattern "%s" not found in\n"""\\\n%s"""\n'
1212
% (needle_re, haystack))
1214
raise AssertionError('pattern "%s" not found in "%s"'
1215
% (needle_re, haystack))
1217
def assertNotContainsRe(self, haystack, needle_re, flags=0):
1218
"""Assert that a does not match a regular expression"""
1219
if re.search(needle_re, haystack, flags):
1220
raise AssertionError('pattern "%s" found in "%s"'
1221
% (needle_re, haystack))
1223
def assertContainsString(self, haystack, needle):
1224
if haystack.find(needle) == -1:
1225
self.fail("string %r not found in '''%s'''" % (needle, haystack))
1227
def assertSubset(self, sublist, superlist):
1228
"""Assert that every entry in sublist is present in superlist."""
1229
missing = set(sublist) - set(superlist)
1230
if len(missing) > 0:
1231
raise AssertionError("value(s) %r not present in container %r" %
1232
(missing, superlist))
1234
def assertListRaises(self, excClass, func, *args, **kwargs):
1235
"""Fail unless excClass is raised when the iterator from func is used.
1237
Many functions can return generators this makes sure
1238
to wrap them in a list() call to make sure the whole generator
1239
is run, and that the proper exception is raised.
1242
list(func(*args, **kwargs))
1246
if getattr(excClass,'__name__', None) is not None:
1247
excName = excClass.__name__
1249
excName = str(excClass)
1250
raise self.failureException, "%s not raised" % excName
1252
def assertRaises(self, excClass, callableObj, *args, **kwargs):
1253
"""Assert that a callable raises a particular exception.
1255
:param excClass: As for the except statement, this may be either an
1256
exception class, or a tuple of classes.
1257
:param callableObj: A callable, will be passed ``*args`` and
1260
Returns the exception so that you can examine it.
1263
callableObj(*args, **kwargs)
1267
if getattr(excClass,'__name__', None) is not None:
1268
excName = excClass.__name__
1271
excName = str(excClass)
1272
raise self.failureException, "%s not raised" % excName
1274
def assertIs(self, left, right, message=None):
1275
if not (left is right):
1276
if message is not None:
1277
raise AssertionError(message)
1279
raise AssertionError("%r is not %r." % (left, right))
1281
def assertIsNot(self, left, right, message=None):
1283
if message is not None:
1284
raise AssertionError(message)
1286
raise AssertionError("%r is %r." % (left, right))
1288
def assertTransportMode(self, transport, path, mode):
1289
"""Fail if a path does not have mode "mode".
1291
If modes are not supported on this transport, the assertion is ignored.
1293
if not transport._can_roundtrip_unix_modebits():
1295
path_stat = transport.stat(path)
1296
actual_mode = stat.S_IMODE(path_stat.st_mode)
1297
self.assertEqual(mode, actual_mode,
1298
'mode of %r incorrect (%s != %s)'
1299
% (path, oct(mode), oct(actual_mode)))
1301
def assertIsSameRealPath(self, path1, path2):
1302
"""Fail if path1 and path2 points to different files"""
1303
self.assertEqual(osutils.realpath(path1),
1304
osutils.realpath(path2),
1305
"apparent paths:\na = %s\nb = %s\n," % (path1, path2))
1307
def assertIsInstance(self, obj, kls, msg=None):
1308
"""Fail if obj is not an instance of kls
1310
:param msg: Supplementary message to show if the assertion fails.
1312
if not isinstance(obj, kls):
1313
m = "%r is an instance of %s rather than %s" % (
1314
obj, obj.__class__, kls)
1319
def assertFileEqual(self, content, path):
1320
"""Fail if path does not contain 'content'."""
1321
self.failUnlessExists(path)
1322
f = file(path, 'rb')
1327
self.assertEqualDiff(content, s)
1329
def assertDocstring(self, expected_docstring, obj):
1330
"""Fail if obj does not have expected_docstring"""
1332
# With -OO the docstring should be None instead
1333
self.assertIs(obj.__doc__, None)
1335
self.assertEqual(expected_docstring, obj.__doc__)
1337
def failUnlessExists(self, path):
1338
"""Fail unless path or paths, which may be abs or relative, exist."""
1339
if not isinstance(path, basestring):
1341
self.failUnlessExists(p)
1343
self.failUnless(osutils.lexists(path),path+" does not exist")
1345
def failIfExists(self, path):
1346
"""Fail if path or paths, which may be abs or relative, exist."""
1347
if not isinstance(path, basestring):
1349
self.failIfExists(p)
1351
self.failIf(osutils.lexists(path),path+" exists")
1353
def _capture_deprecation_warnings(self, a_callable, *args, **kwargs):
1354
"""A helper for callDeprecated and applyDeprecated.
1356
:param a_callable: A callable to call.
1357
:param args: The positional arguments for the callable
1358
:param kwargs: The keyword arguments for the callable
1359
:return: A tuple (warnings, result). result is the result of calling
1360
a_callable(``*args``, ``**kwargs``).
1363
def capture_warnings(msg, cls=None, stacklevel=None):
1364
# we've hooked into a deprecation specific callpath,
1365
# only deprecations should getting sent via it.
1366
self.assertEqual(cls, DeprecationWarning)
1367
local_warnings.append(msg)
1368
original_warning_method = symbol_versioning.warn
1369
symbol_versioning.set_warning_method(capture_warnings)
1371
result = a_callable(*args, **kwargs)
1373
symbol_versioning.set_warning_method(original_warning_method)
1374
return (local_warnings, result)
1376
def applyDeprecated(self, deprecation_format, a_callable, *args, **kwargs):
1377
"""Call a deprecated callable without warning the user.
1379
Note that this only captures warnings raised by symbol_versioning.warn,
1380
not other callers that go direct to the warning module.
1382
To test that a deprecated method raises an error, do something like
1385
self.assertRaises(errors.ReservedId,
1386
self.applyDeprecated,
1387
deprecated_in((1, 5, 0)),
1391
:param deprecation_format: The deprecation format that the callable
1392
should have been deprecated with. This is the same type as the
1393
parameter to deprecated_method/deprecated_function. If the
1394
callable is not deprecated with this format, an assertion error
1396
:param a_callable: A callable to call. This may be a bound method or
1397
a regular function. It will be called with ``*args`` and
1399
:param args: The positional arguments for the callable
1400
:param kwargs: The keyword arguments for the callable
1401
:return: The result of a_callable(``*args``, ``**kwargs``)
1403
call_warnings, result = self._capture_deprecation_warnings(a_callable,
1405
expected_first_warning = symbol_versioning.deprecation_string(
1406
a_callable, deprecation_format)
1407
if len(call_warnings) == 0:
1408
self.fail("No deprecation warning generated by call to %s" %
1410
self.assertEqual(expected_first_warning, call_warnings[0])
1413
def callCatchWarnings(self, fn, *args, **kw):
1414
"""Call a callable that raises python warnings.
1416
The caller's responsible for examining the returned warnings.
1418
If the callable raises an exception, the exception is not
1419
caught and propagates up to the caller. In that case, the list
1420
of warnings is not available.
1422
:returns: ([warning_object, ...], fn_result)
1424
# XXX: This is not perfect, because it completely overrides the
1425
# warnings filters, and some code may depend on suppressing particular
1426
# warnings. It's the easiest way to insulate ourselves from -Werror,
1427
# though. -- Andrew, 20071062
1429
def _catcher(message, category, filename, lineno, file=None, line=None):
1430
# despite the name, 'message' is normally(?) a Warning subclass
1432
wlist.append(message)
1433
saved_showwarning = warnings.showwarning
1434
saved_filters = warnings.filters
1436
warnings.showwarning = _catcher
1437
warnings.filters = []
1438
result = fn(*args, **kw)
1440
warnings.showwarning = saved_showwarning
1441
warnings.filters = saved_filters
1442
return wlist, result
1444
def callDeprecated(self, expected, callable, *args, **kwargs):
1445
"""Assert that a callable is deprecated in a particular way.
1447
This is a very precise test for unusual requirements. The
1448
applyDeprecated helper function is probably more suited for most tests
1449
as it allows you to simply specify the deprecation format being used
1450
and will ensure that that is issued for the function being called.
1452
Note that this only captures warnings raised by symbol_versioning.warn,
1453
not other callers that go direct to the warning module. To catch
1454
general warnings, use callCatchWarnings.
1456
:param expected: a list of the deprecation warnings expected, in order
1457
:param callable: The callable to call
1458
:param args: The positional arguments for the callable
1459
:param kwargs: The keyword arguments for the callable
1461
call_warnings, result = self._capture_deprecation_warnings(callable,
1463
self.assertEqual(expected, call_warnings)
1466
def _startLogFile(self):
1467
"""Send bzr and test log messages to a temporary file.
1469
The file is removed as the test is torn down.
1471
self._log_file = StringIO()
1472
self._log_memento = bzrlib.trace.push_log_file(self._log_file)
1473
self.addCleanup(self._finishLogFile)
1475
def _finishLogFile(self):
1476
"""Finished with the log file.
1478
Close the file and delete it, unless setKeepLogfile was called.
1480
if bzrlib.trace._trace_file:
1481
# flush the log file, to get all content
1482
bzrlib.trace._trace_file.flush()
1483
bzrlib.trace.pop_log_file(self._log_memento)
1484
# Cache the log result and delete the file on disk
1485
self._get_log(False)
1487
def thisFailsStrictLockCheck(self):
1488
"""It is known that this test would fail with -Dstrict_locks.
1490
By default, all tests are run with strict lock checking unless
1491
-Edisable_lock_checks is supplied. However there are some tests which
1492
we know fail strict locks at this point that have not been fixed.
1493
They should call this function to disable the strict checking.
1495
This should be used sparingly, it is much better to fix the locking
1496
issues rather than papering over the problem by calling this function.
1498
debug.debug_flags.discard('strict_locks')
1500
def overrideAttr(self, obj, attr_name, new=_unitialized_attr):
1501
"""Overrides an object attribute restoring it after the test.
1503
:param obj: The object that will be mutated.
1505
:param attr_name: The attribute name we want to preserve/override in
1508
:param new: The optional value we want to set the attribute to.
1510
:returns: The actual attr value.
1512
value = getattr(obj, attr_name)
1513
# The actual value is captured by the call below
1514
self.addCleanup(setattr, obj, attr_name, value)
1515
if new is not _unitialized_attr:
1516
setattr(obj, attr_name, new)
1519
def _cleanEnvironment(self):
1521
'BZR_HOME': None, # Don't inherit BZR_HOME to all the tests.
1522
'HOME': os.getcwd(),
1523
# bzr now uses the Win32 API and doesn't rely on APPDATA, but the
1524
# tests do check our impls match APPDATA
1525
'BZR_EDITOR': None, # test_msgeditor manipulates this variable
1529
'BZREMAIL': None, # may still be present in the environment
1530
'EMAIL': 'jrandom@example.com', # set EMAIL as bzr does not guess
1531
'BZR_PROGRESS_BAR': None,
1533
'BZR_PLUGIN_PATH': None,
1534
'BZR_DISABLE_PLUGINS': None,
1535
'BZR_PLUGINS_AT': None,
1536
'BZR_CONCURRENCY': None,
1537
# Make sure that any text ui tests are consistent regardless of
1538
# the environment the test case is run in; you may want tests that
1539
# test other combinations. 'dumb' is a reasonable guess for tests
1540
# going to a pipe or a StringIO.
1544
'BZR_COLUMNS': '80',
1546
'SSH_AUTH_SOCK': None,
1550
'https_proxy': None,
1551
'HTTPS_PROXY': None,
1556
# Nobody cares about ftp_proxy, FTP_PROXY AFAIK. So far at
1557
# least. If you do (care), please update this comment
1561
'BZR_REMOTE_PATH': None,
1562
# Generally speaking, we don't want apport reporting on crashes in
1563
# the test envirnoment unless we're specifically testing apport,
1564
# so that it doesn't leak into the real system environment. We
1565
# use an env var so it propagates to subprocesses.
1566
'APPORT_DISABLE': '1',
1569
self.addCleanup(self._restoreEnvironment)
1570
for name, value in new_env.iteritems():
1571
self._captureVar(name, value)
1573
def _captureVar(self, name, newvalue):
1574
"""Set an environment variable, and reset it when finished."""
1575
self._old_env[name] = osutils.set_or_unset_env(name, newvalue)
1577
def _restoreEnvironment(self):
1578
for name, value in self._old_env.iteritems():
1579
osutils.set_or_unset_env(name, value)
1581
def _restoreHooks(self):
1582
for klass, (name, hooks) in self._preserved_hooks.items():
1583
setattr(klass, name, hooks)
1585
def knownFailure(self, reason):
1586
"""This test has failed for some known reason."""
1587
raise KnownFailure(reason)
1589
def _suppress_log(self):
1590
"""Remove the log info from details."""
1591
self.discardDetail('log')
1593
def _do_skip(self, result, reason):
1594
self._suppress_log()
1595
addSkip = getattr(result, 'addSkip', None)
1596
if not callable(addSkip):
1597
result.addSuccess(result)
1599
addSkip(self, reason)
1602
def _do_known_failure(self, result, e):
1603
self._suppress_log()
1604
err = sys.exc_info()
1605
addExpectedFailure = getattr(result, 'addExpectedFailure', None)
1606
if addExpectedFailure is not None:
1607
addExpectedFailure(self, err)
1609
result.addSuccess(self)
1612
def _do_not_applicable(self, result, e):
1614
reason = 'No reason given'
1617
self._suppress_log ()
1618
addNotApplicable = getattr(result, 'addNotApplicable', None)
1619
if addNotApplicable is not None:
1620
result.addNotApplicable(self, reason)
1622
self._do_skip(result, reason)
1625
def _report_skip(self, result, err):
1626
"""Override the default _report_skip.
1628
We want to strip the 'log' detail. If we waint until _do_skip, it has
1629
already been formatted into the 'reason' string, and we can't pull it
1632
self._suppress_log()
1633
super(TestCase, self)._report_skip(self, result, err)
1636
def _report_expected_failure(self, result, err):
1639
See _report_skip for motivation.
1641
self._suppress_log()
1642
super(TestCase, self)._report_expected_failure(self, result, err)
1645
def _do_unsupported_or_skip(self, result, e):
1647
self._suppress_log()
1648
addNotSupported = getattr(result, 'addNotSupported', None)
1649
if addNotSupported is not None:
1650
result.addNotSupported(self, reason)
1652
self._do_skip(result, reason)
1654
def time(self, callable, *args, **kwargs):
1655
"""Run callable and accrue the time it takes to the benchmark time.
1657
If lsprofiling is enabled (i.e. by --lsprof-time to bzr selftest) then
1658
this will cause lsprofile statistics to be gathered and stored in
1661
if self._benchtime is None:
1662
self.addDetail('benchtime', content.Content(content.ContentType(
1663
"text", "plain"), lambda:[str(self._benchtime)]))
1667
if not self._gather_lsprof_in_benchmarks:
1668
return callable(*args, **kwargs)
1670
# record this benchmark
1671
ret, stats = bzrlib.lsprof.profile(callable, *args, **kwargs)
1673
self._benchcalls.append(((callable, args, kwargs), stats))
1676
self._benchtime += time.time() - start
1678
def log(self, *args):
1681
def _get_log(self, keep_log_file=False):
1682
"""Internal helper to get the log from bzrlib.trace for this test.
1684
Please use self.getDetails, or self.get_log to access this in test case
1687
:param keep_log_file: When True, if the log is still a file on disk
1688
leave it as a file on disk. When False, if the log is still a file
1689
on disk, the log file is deleted and the log preserved as
1691
:return: A string containing the log.
1693
if self._log_contents is not None:
1695
self._log_contents.decode('utf8')
1696
except UnicodeDecodeError:
1697
unicodestr = self._log_contents.decode('utf8', 'replace')
1698
self._log_contents = unicodestr.encode('utf8')
1699
return self._log_contents
1700
if self._log_file is not None:
1701
log_contents = self._log_file.getvalue()
1703
log_contents.decode('utf8')
1704
except UnicodeDecodeError:
1705
unicodestr = log_contents.decode('utf8', 'replace')
1706
log_contents = unicodestr.encode('utf8')
1707
if not keep_log_file:
1708
self._log_file = None
1709
# Permit multiple calls to get_log until we clean it up in
1711
self._log_contents = log_contents
1714
return "No log file content."
1717
"""Get a unicode string containing the log from bzrlib.trace.
1719
Undecodable characters are replaced.
1721
return u"".join(self.getDetails()['log'].iter_text())
1723
def requireFeature(self, feature):
1724
"""This test requires a specific feature is available.
1726
:raises UnavailableFeature: When feature is not available.
1728
if not feature.available():
1729
raise UnavailableFeature(feature)
1731
def _run_bzr_autosplit(self, args, retcode, encoding, stdin,
1733
"""Run bazaar command line, splitting up a string command line."""
1734
if isinstance(args, basestring):
1735
# shlex don't understand unicode strings,
1736
# so args should be plain string (bialix 20070906)
1737
args = list(shlex.split(str(args)))
1738
return self._run_bzr_core(args, retcode=retcode,
1739
encoding=encoding, stdin=stdin, working_dir=working_dir,
1742
def _run_bzr_core(self, args, retcode, encoding, stdin,
1744
# Clear chk_map page cache, because the contents are likely to mask
1746
chk_map.clear_cache()
1747
if encoding is None:
1748
encoding = osutils.get_user_encoding()
1749
stdout = StringIOWrapper()
1750
stderr = StringIOWrapper()
1751
stdout.encoding = encoding
1752
stderr.encoding = encoding
1754
self.log('run bzr: %r', args)
1755
# FIXME: don't call into logging here
1756
handler = logging.StreamHandler(stderr)
1757
handler.setLevel(logging.INFO)
1758
logger = logging.getLogger('')
1759
logger.addHandler(handler)
1760
old_ui_factory = ui.ui_factory
1761
ui.ui_factory = TestUIFactory(stdin=stdin, stdout=stdout, stderr=stderr)
1764
if working_dir is not None:
1765
cwd = osutils.getcwd()
1766
os.chdir(working_dir)
1770
result = self.apply_redirected(ui.ui_factory.stdin,
1772
bzrlib.commands.run_bzr_catch_user_errors,
1774
except KeyboardInterrupt:
1775
# Reraise KeyboardInterrupt with contents of redirected stdout
1776
# and stderr as arguments, for tests which are interested in
1777
# stdout and stderr and are expecting the exception.
1778
out = stdout.getvalue()
1779
err = stderr.getvalue()
1781
self.log('output:\n%r', out)
1783
self.log('errors:\n%r', err)
1784
raise KeyboardInterrupt(out, err)
1786
logger.removeHandler(handler)
1787
ui.ui_factory = old_ui_factory
1791
out = stdout.getvalue()
1792
err = stderr.getvalue()
1794
self.log('output:\n%r', out)
1796
self.log('errors:\n%r', err)
1797
if retcode is not None:
1798
self.assertEquals(retcode, result,
1799
message='Unexpected return code')
1800
return result, out, err
1802
def run_bzr(self, args, retcode=0, encoding=None, stdin=None,
1803
working_dir=None, error_regexes=[], output_encoding=None):
1804
"""Invoke bzr, as if it were run from the command line.
1806
The argument list should not include the bzr program name - the
1807
first argument is normally the bzr command. Arguments may be
1808
passed in three ways:
1810
1- A list of strings, eg ["commit", "a"]. This is recommended
1811
when the command contains whitespace or metacharacters, or
1812
is built up at run time.
1814
2- A single string, eg "add a". This is the most convenient
1815
for hardcoded commands.
1817
This runs bzr through the interface that catches and reports
1818
errors, and with logging set to something approximating the
1819
default, so that error reporting can be checked.
1821
This should be the main method for tests that want to exercise the
1822
overall behavior of the bzr application (rather than a unit test
1823
or a functional test of the library.)
1825
This sends the stdout/stderr results into the test's log,
1826
where it may be useful for debugging. See also run_captured.
1828
:keyword stdin: A string to be used as stdin for the command.
1829
:keyword retcode: The status code the command should return;
1831
:keyword working_dir: The directory to run the command in
1832
:keyword error_regexes: A list of expected error messages. If
1833
specified they must be seen in the error output of the command.
1835
retcode, out, err = self._run_bzr_autosplit(
1840
working_dir=working_dir,
1842
self.assertIsInstance(error_regexes, (list, tuple))
1843
for regex in error_regexes:
1844
self.assertContainsRe(err, regex)
1847
def run_bzr_error(self, error_regexes, *args, **kwargs):
1848
"""Run bzr, and check that stderr contains the supplied regexes
1850
:param error_regexes: Sequence of regular expressions which
1851
must each be found in the error output. The relative ordering
1853
:param args: command-line arguments for bzr
1854
:param kwargs: Keyword arguments which are interpreted by run_bzr
1855
This function changes the default value of retcode to be 3,
1856
since in most cases this is run when you expect bzr to fail.
1858
:return: (out, err) The actual output of running the command (in case
1859
you want to do more inspection)
1863
# Make sure that commit is failing because there is nothing to do
1864
self.run_bzr_error(['no changes to commit'],
1865
['commit', '-m', 'my commit comment'])
1866
# Make sure --strict is handling an unknown file, rather than
1867
# giving us the 'nothing to do' error
1868
self.build_tree(['unknown'])
1869
self.run_bzr_error(['Commit refused because there are unknown files'],
1870
['commit', --strict', '-m', 'my commit comment'])
1872
kwargs.setdefault('retcode', 3)
1873
kwargs['error_regexes'] = error_regexes
1874
out, err = self.run_bzr(*args, **kwargs)
1877
def run_bzr_subprocess(self, *args, **kwargs):
1878
"""Run bzr in a subprocess for testing.
1880
This starts a new Python interpreter and runs bzr in there.
1881
This should only be used for tests that have a justifiable need for
1882
this isolation: e.g. they are testing startup time, or signal
1883
handling, or early startup code, etc. Subprocess code can't be
1884
profiled or debugged so easily.
1886
:keyword retcode: The status code that is expected. Defaults to 0. If
1887
None is supplied, the status code is not checked.
1888
:keyword env_changes: A dictionary which lists changes to environment
1889
variables. A value of None will unset the env variable.
1890
The values must be strings. The change will only occur in the
1891
child, so you don't need to fix the environment after running.
1892
:keyword universal_newlines: Convert CRLF => LF
1893
:keyword allow_plugins: By default the subprocess is run with
1894
--no-plugins to ensure test reproducibility. Also, it is possible
1895
for system-wide plugins to create unexpected output on stderr,
1896
which can cause unnecessary test failures.
1898
env_changes = kwargs.get('env_changes', {})
1899
working_dir = kwargs.get('working_dir', None)
1900
allow_plugins = kwargs.get('allow_plugins', False)
1902
if isinstance(args[0], list):
1904
elif isinstance(args[0], basestring):
1905
args = list(shlex.split(args[0]))
1907
raise ValueError("passing varargs to run_bzr_subprocess")
1908
process = self.start_bzr_subprocess(args, env_changes=env_changes,
1909
working_dir=working_dir,
1910
allow_plugins=allow_plugins)
1911
# We distinguish between retcode=None and retcode not passed.
1912
supplied_retcode = kwargs.get('retcode', 0)
1913
return self.finish_bzr_subprocess(process, retcode=supplied_retcode,
1914
universal_newlines=kwargs.get('universal_newlines', False),
1917
def start_bzr_subprocess(self, process_args, env_changes=None,
1918
skip_if_plan_to_signal=False,
1920
allow_plugins=False):
1921
"""Start bzr in a subprocess for testing.
1923
This starts a new Python interpreter and runs bzr in there.
1924
This should only be used for tests that have a justifiable need for
1925
this isolation: e.g. they are testing startup time, or signal
1926
handling, or early startup code, etc. Subprocess code can't be
1927
profiled or debugged so easily.
1929
:param process_args: a list of arguments to pass to the bzr executable,
1930
for example ``['--version']``.
1931
:param env_changes: A dictionary which lists changes to environment
1932
variables. A value of None will unset the env variable.
1933
The values must be strings. The change will only occur in the
1934
child, so you don't need to fix the environment after running.
1935
:param skip_if_plan_to_signal: raise TestSkipped when true and system
1936
doesn't support signalling subprocesses.
1937
:param allow_plugins: If False (default) pass --no-plugins to bzr.
1939
:returns: Popen object for the started process.
1941
if skip_if_plan_to_signal:
1942
if os.name != "posix":
1943
raise TestSkipped("Sending signals not supported")
1945
if env_changes is None:
1949
def cleanup_environment():
1950
for env_var, value in env_changes.iteritems():
1951
old_env[env_var] = osutils.set_or_unset_env(env_var, value)
1953
def restore_environment():
1954
for env_var, value in old_env.iteritems():
1955
osutils.set_or_unset_env(env_var, value)
1957
bzr_path = self.get_bzr_path()
1960
if working_dir is not None:
1961
cwd = osutils.getcwd()
1962
os.chdir(working_dir)
1965
# win32 subprocess doesn't support preexec_fn
1966
# so we will avoid using it on all platforms, just to
1967
# make sure the code path is used, and we don't break on win32
1968
cleanup_environment()
1969
command = [sys.executable]
1970
# frozen executables don't need the path to bzr
1971
if getattr(sys, "frozen", None) is None:
1972
command.append(bzr_path)
1973
if not allow_plugins:
1974
command.append('--no-plugins')
1975
command.extend(process_args)
1976
process = self._popen(command, stdin=subprocess.PIPE,
1977
stdout=subprocess.PIPE,
1978
stderr=subprocess.PIPE)
1980
restore_environment()
1986
def _popen(self, *args, **kwargs):
1987
"""Place a call to Popen.
1989
Allows tests to override this method to intercept the calls made to
1990
Popen for introspection.
1992
return subprocess.Popen(*args, **kwargs)
1994
def get_source_path(self):
1995
"""Return the path of the directory containing bzrlib."""
1996
return os.path.dirname(os.path.dirname(bzrlib.__file__))
1998
def get_bzr_path(self):
1999
"""Return the path of the 'bzr' executable for this test suite."""
2000
bzr_path = os.path.join(self.get_source_path(), "bzr")
2001
if not os.path.isfile(bzr_path):
2002
# We are probably installed. Assume sys.argv is the right file
2003
bzr_path = sys.argv[0]
2006
def finish_bzr_subprocess(self, process, retcode=0, send_signal=None,
2007
universal_newlines=False, process_args=None):
2008
"""Finish the execution of process.
2010
:param process: the Popen object returned from start_bzr_subprocess.
2011
:param retcode: The status code that is expected. Defaults to 0. If
2012
None is supplied, the status code is not checked.
2013
:param send_signal: an optional signal to send to the process.
2014
:param universal_newlines: Convert CRLF => LF
2015
:returns: (stdout, stderr)
2017
if send_signal is not None:
2018
os.kill(process.pid, send_signal)
2019
out, err = process.communicate()
2021
if universal_newlines:
2022
out = out.replace('\r\n', '\n')
2023
err = err.replace('\r\n', '\n')
2025
if retcode is not None and retcode != process.returncode:
2026
if process_args is None:
2027
process_args = "(unknown args)"
2028
mutter('Output of bzr %s:\n%s', process_args, out)
2029
mutter('Error for bzr %s:\n%s', process_args, err)
2030
self.fail('Command bzr %s failed with retcode %s != %s'
2031
% (process_args, retcode, process.returncode))
2034
def check_inventory_shape(self, inv, shape):
2035
"""Compare an inventory to a list of expected names.
2037
Fail if they are not precisely equal.
2040
shape = list(shape) # copy
2041
for path, ie in inv.entries():
2042
name = path.replace('\\', '/')
2043
if ie.kind == 'directory':
2050
self.fail("expected paths not found in inventory: %r" % shape)
2052
self.fail("unexpected paths found in inventory: %r" % extras)
2054
def apply_redirected(self, stdin=None, stdout=None, stderr=None,
2055
a_callable=None, *args, **kwargs):
2056
"""Call callable with redirected std io pipes.
2058
Returns the return code."""
2059
if not callable(a_callable):
2060
raise ValueError("a_callable must be callable.")
2062
stdin = StringIO("")
2064
if getattr(self, "_log_file", None) is not None:
2065
stdout = self._log_file
2069
if getattr(self, "_log_file", None is not None):
2070
stderr = self._log_file
2073
real_stdin = sys.stdin
2074
real_stdout = sys.stdout
2075
real_stderr = sys.stderr
2080
return a_callable(*args, **kwargs)
2082
sys.stdout = real_stdout
2083
sys.stderr = real_stderr
2084
sys.stdin = real_stdin
2086
def reduceLockdirTimeout(self):
2087
"""Reduce the default lock timeout for the duration of the test, so that
2088
if LockContention occurs during a test, it does so quickly.
2090
Tests that expect to provoke LockContention errors should call this.
2092
self.overrideAttr(bzrlib.lockdir, '_DEFAULT_TIMEOUT_SECONDS', 0)
2094
def make_utf8_encoded_stringio(self, encoding_type=None):
2095
"""Return a StringIOWrapper instance, that will encode Unicode
2098
if encoding_type is None:
2099
encoding_type = 'strict'
2101
output_encoding = 'utf-8'
2102
sio = codecs.getwriter(output_encoding)(sio, errors=encoding_type)
2103
sio.encoding = output_encoding
2106
def disable_verb(self, verb):
2107
"""Disable a smart server verb for one test."""
2108
from bzrlib.smart import request
2109
request_handlers = request.request_handlers
2110
orig_method = request_handlers.get(verb)
2111
request_handlers.remove(verb)
2112
self.addCleanup(request_handlers.register, verb, orig_method)
2115
class CapturedCall(object):
2116
"""A helper for capturing smart server calls for easy debug analysis."""
2118
def __init__(self, params, prefix_length):
2119
"""Capture the call with params and skip prefix_length stack frames."""
2122
# The last 5 frames are the __init__, the hook frame, and 3 smart
2123
# client frames. Beyond this we could get more clever, but this is good
2125
stack = traceback.extract_stack()[prefix_length:-5]
2126
self.stack = ''.join(traceback.format_list(stack))
2129
return self.call.method
2132
return self.call.method
2138
class TestCaseWithMemoryTransport(TestCase):
2139
"""Common test class for tests that do not need disk resources.
2141
Tests that need disk resources should derive from TestCaseWithTransport.
2143
TestCaseWithMemoryTransport sets the TEST_ROOT variable for all bzr tests.
2145
For TestCaseWithMemoryTransport the test_home_dir is set to the name of
2146
a directory which does not exist. This serves to help ensure test isolation
2147
is preserved. test_dir is set to the TEST_ROOT, as is cwd, because they
2148
must exist. However, TestCaseWithMemoryTransport does not offer local
2149
file defaults for the transport in tests, nor does it obey the command line
2150
override, so tests that accidentally write to the common directory should
2153
:cvar TEST_ROOT: Directory containing all temporary directories, plus
2154
a .bzr directory that stops us ascending higher into the filesystem.
2160
def __init__(self, methodName='runTest'):
2161
# allow test parameterization after test construction and before test
2162
# execution. Variables that the parameterizer sets need to be
2163
# ones that are not set by setUp, or setUp will trash them.
2164
super(TestCaseWithMemoryTransport, self).__init__(methodName)
2165
self.vfs_transport_factory = default_transport
2166
self.transport_server = None
2167
self.transport_readonly_server = None
2168
self.__vfs_server = None
2170
def get_transport(self, relpath=None):
2171
"""Return a writeable transport.
2173
This transport is for the test scratch space relative to
2176
:param relpath: a path relative to the base url.
2178
t = _mod_transport.get_transport(self.get_url(relpath))
2179
self.assertFalse(t.is_readonly())
2182
def get_readonly_transport(self, relpath=None):
2183
"""Return a readonly transport for the test scratch space
2185
This can be used to test that operations which should only need
2186
readonly access in fact do not try to write.
2188
:param relpath: a path relative to the base url.
2190
t = _mod_transport.get_transport(self.get_readonly_url(relpath))
2191
self.assertTrue(t.is_readonly())
2194
def create_transport_readonly_server(self):
2195
"""Create a transport server from class defined at init.
2197
This is mostly a hook for daughter classes.
2199
return self.transport_readonly_server()
2201
def get_readonly_server(self):
2202
"""Get the server instance for the readonly transport
2204
This is useful for some tests with specific servers to do diagnostics.
2206
if self.__readonly_server is None:
2207
if self.transport_readonly_server is None:
2208
# readonly decorator requested
2209
self.__readonly_server = test_server.ReadonlyServer()
2211
# explicit readonly transport.
2212
self.__readonly_server = self.create_transport_readonly_server()
2213
self.start_server(self.__readonly_server,
2214
self.get_vfs_only_server())
2215
return self.__readonly_server
2217
def get_readonly_url(self, relpath=None):
2218
"""Get a URL for the readonly transport.
2220
This will either be backed by '.' or a decorator to the transport
2221
used by self.get_url()
2222
relpath provides for clients to get a path relative to the base url.
2223
These should only be downwards relative, not upwards.
2225
base = self.get_readonly_server().get_url()
2226
return self._adjust_url(base, relpath)
2228
def get_vfs_only_server(self):
2229
"""Get the vfs only read/write server instance.
2231
This is useful for some tests with specific servers that need
2234
For TestCaseWithMemoryTransport this is always a MemoryServer, and there
2235
is no means to override it.
2237
if self.__vfs_server is None:
2238
self.__vfs_server = memory.MemoryServer()
2239
self.start_server(self.__vfs_server)
2240
return self.__vfs_server
2242
def get_server(self):
2243
"""Get the read/write server instance.
2245
This is useful for some tests with specific servers that need
2248
This is built from the self.transport_server factory. If that is None,
2249
then the self.get_vfs_server is returned.
2251
if self.__server is None:
2252
if (self.transport_server is None or self.transport_server is
2253
self.vfs_transport_factory):
2254
self.__server = self.get_vfs_only_server()
2256
# bring up a decorated means of access to the vfs only server.
2257
self.__server = self.transport_server()
2258
self.start_server(self.__server, self.get_vfs_only_server())
2259
return self.__server
2261
def _adjust_url(self, base, relpath):
2262
"""Get a URL (or maybe a path) for the readwrite transport.
2264
This will either be backed by '.' or to an equivalent non-file based
2266
relpath provides for clients to get a path relative to the base url.
2267
These should only be downwards relative, not upwards.
2269
if relpath is not None and relpath != '.':
2270
if not base.endswith('/'):
2272
# XXX: Really base should be a url; we did after all call
2273
# get_url()! But sometimes it's just a path (from
2274
# LocalAbspathServer), and it'd be wrong to append urlescaped data
2275
# to a non-escaped local path.
2276
if base.startswith('./') or base.startswith('/'):
2279
base += urlutils.escape(relpath)
2282
def get_url(self, relpath=None):
2283
"""Get a URL (or maybe a path) for the readwrite transport.
2285
This will either be backed by '.' or to an equivalent non-file based
2287
relpath provides for clients to get a path relative to the base url.
2288
These should only be downwards relative, not upwards.
2290
base = self.get_server().get_url()
2291
return self._adjust_url(base, relpath)
2293
def get_vfs_only_url(self, relpath=None):
2294
"""Get a URL (or maybe a path for the plain old vfs transport.
2296
This will never be a smart protocol. It always has all the
2297
capabilities of the local filesystem, but it might actually be a
2298
MemoryTransport or some other similar virtual filesystem.
2300
This is the backing transport (if any) of the server returned by
2301
get_url and get_readonly_url.
2303
:param relpath: provides for clients to get a path relative to the base
2304
url. These should only be downwards relative, not upwards.
2307
base = self.get_vfs_only_server().get_url()
2308
return self._adjust_url(base, relpath)
2310
def _create_safety_net(self):
2311
"""Make a fake bzr directory.
2313
This prevents any tests propagating up onto the TEST_ROOT directory's
2316
root = TestCaseWithMemoryTransport.TEST_ROOT
2317
bzrdir.BzrDir.create_standalone_workingtree(root)
2319
def _check_safety_net(self):
2320
"""Check that the safety .bzr directory have not been touched.
2322
_make_test_root have created a .bzr directory to prevent tests from
2323
propagating. This method ensures than a test did not leaked.
2325
root = TestCaseWithMemoryTransport.TEST_ROOT
2326
self.permit_url(_mod_transport.get_transport(root).base)
2327
wt = workingtree.WorkingTree.open(root)
2328
last_rev = wt.last_revision()
2329
if last_rev != 'null:':
2330
# The current test have modified the /bzr directory, we need to
2331
# recreate a new one or all the followng tests will fail.
2332
# If you need to inspect its content uncomment the following line
2333
# import pdb; pdb.set_trace()
2334
_rmtree_temp_dir(root + '/.bzr', test_id=self.id())
2335
self._create_safety_net()
2336
raise AssertionError('%s/.bzr should not be modified' % root)
2338
def _make_test_root(self):
2339
if TestCaseWithMemoryTransport.TEST_ROOT is None:
2340
# Watch out for tricky test dir (on OSX /tmp -> /private/tmp)
2341
root = osutils.realpath(osutils.mkdtemp(prefix='testbzr-',
2343
TestCaseWithMemoryTransport.TEST_ROOT = root
2345
self._create_safety_net()
2347
# The same directory is used by all tests, and we're not
2348
# specifically told when all tests are finished. This will do.
2349
atexit.register(_rmtree_temp_dir, root)
2351
self.permit_dir(TestCaseWithMemoryTransport.TEST_ROOT)
2352
self.addCleanup(self._check_safety_net)
2354
def makeAndChdirToTestDir(self):
2355
"""Create a temporary directories for this one test.
2357
This must set self.test_home_dir and self.test_dir and chdir to
2360
For TestCaseWithMemoryTransport we chdir to the TEST_ROOT for this test.
2362
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2363
self.test_dir = TestCaseWithMemoryTransport.TEST_ROOT
2364
self.test_home_dir = self.test_dir + "/MemoryTransportMissingHomeDir"
2365
self.permit_dir(self.test_dir)
2367
def make_branch(self, relpath, format=None):
2368
"""Create a branch on the transport at relpath."""
2369
repo = self.make_repository(relpath, format=format)
2370
return repo.bzrdir.create_branch()
2372
def make_bzrdir(self, relpath, format=None):
2374
# might be a relative or absolute path
2375
maybe_a_url = self.get_url(relpath)
2376
segments = maybe_a_url.rsplit('/', 1)
2377
t = _mod_transport.get_transport(maybe_a_url)
2378
if len(segments) > 1 and segments[-1] not in ('', '.'):
2382
if isinstance(format, basestring):
2383
format = bzrdir.format_registry.make_bzrdir(format)
2384
return format.initialize_on_transport(t)
2385
except errors.UninitializableFormat:
2386
raise TestSkipped("Format %s is not initializable." % format)
2388
def make_repository(self, relpath, shared=False, format=None):
2389
"""Create a repository on our default transport at relpath.
2391
Note that relpath must be a relative path, not a full url.
2393
# FIXME: If you create a remoterepository this returns the underlying
2394
# real format, which is incorrect. Actually we should make sure that
2395
# RemoteBzrDir returns a RemoteRepository.
2396
# maybe mbp 20070410
2397
made_control = self.make_bzrdir(relpath, format=format)
2398
return made_control.create_repository(shared=shared)
2400
def make_smart_server(self, path, backing_server=None):
2401
if backing_server is None:
2402
backing_server = self.get_server()
2403
smart_server = test_server.SmartTCPServer_for_testing()
2404
self.start_server(smart_server, backing_server)
2405
remote_transport = _mod_transport.get_transport(smart_server.get_url()
2407
return remote_transport
2409
def make_branch_and_memory_tree(self, relpath, format=None):
2410
"""Create a branch on the default transport and a MemoryTree for it."""
2411
b = self.make_branch(relpath, format=format)
2412
return memorytree.MemoryTree.create_on_branch(b)
2414
def make_branch_builder(self, relpath, format=None):
2415
branch = self.make_branch(relpath, format=format)
2416
return branchbuilder.BranchBuilder(branch=branch)
2418
def overrideEnvironmentForTesting(self):
2419
test_home_dir = self.test_home_dir
2420
if isinstance(test_home_dir, unicode):
2421
test_home_dir = test_home_dir.encode(sys.getfilesystemencoding())
2422
os.environ['HOME'] = test_home_dir
2423
os.environ['BZR_HOME'] = test_home_dir
2426
super(TestCaseWithMemoryTransport, self).setUp()
2427
# Ensure that ConnectedTransport doesn't leak sockets
2428
def get_transport_with_cleanup(*args, **kwargs):
2429
t = orig_get_transport(*args, **kwargs)
2430
if isinstance(t, _mod_transport.ConnectedTransport):
2431
self.addCleanup(t.disconnect)
2434
orig_get_transport = self.overrideAttr(_mod_transport, 'get_transport',
2435
get_transport_with_cleanup)
2436
self._make_test_root()
2437
self.addCleanup(os.chdir, os.getcwdu())
2438
self.makeAndChdirToTestDir()
2439
self.overrideEnvironmentForTesting()
2440
self.__readonly_server = None
2441
self.__server = None
2442
self.reduceLockdirTimeout()
2444
def setup_smart_server_with_call_log(self):
2445
"""Sets up a smart server as the transport server with a call log."""
2446
self.transport_server = test_server.SmartTCPServer_for_testing
2447
self.hpss_calls = []
2449
# Skip the current stack down to the caller of
2450
# setup_smart_server_with_call_log
2451
prefix_length = len(traceback.extract_stack()) - 2
2452
def capture_hpss_call(params):
2453
self.hpss_calls.append(
2454
CapturedCall(params, prefix_length))
2455
client._SmartClient.hooks.install_named_hook(
2456
'call', capture_hpss_call, None)
2458
def reset_smart_call_log(self):
2459
self.hpss_calls = []
2462
class TestCaseInTempDir(TestCaseWithMemoryTransport):
2463
"""Derived class that runs a test within a temporary directory.
2465
This is useful for tests that need to create a branch, etc.
2467
The directory is created in a slightly complex way: for each
2468
Python invocation, a new temporary top-level directory is created.
2469
All test cases create their own directory within that. If the
2470
tests complete successfully, the directory is removed.
2472
:ivar test_base_dir: The path of the top-level directory for this
2473
test, which contains a home directory and a work directory.
2475
:ivar test_home_dir: An initially empty directory under test_base_dir
2476
which is used as $HOME for this test.
2478
:ivar test_dir: A directory under test_base_dir used as the current
2479
directory when the test proper is run.
2482
OVERRIDE_PYTHON = 'python'
2484
def check_file_contents(self, filename, expect):
2485
self.log("check contents of file %s" % filename)
2491
if contents != expect:
2492
self.log("expected: %r" % expect)
2493
self.log("actually: %r" % contents)
2494
self.fail("contents of %s not as expected" % filename)
2496
def _getTestDirPrefix(self):
2497
# create a directory within the top level test directory
2498
if sys.platform in ('win32', 'cygwin'):
2499
name_prefix = re.sub('[<>*=+",:;_/\\-]', '_', self.id())
2500
# windows is likely to have path-length limits so use a short name
2501
name_prefix = name_prefix[-30:]
2503
name_prefix = re.sub('[/]', '_', self.id())
2506
def makeAndChdirToTestDir(self):
2507
"""See TestCaseWithMemoryTransport.makeAndChdirToTestDir().
2509
For TestCaseInTempDir we create a temporary directory based on the test
2510
name and then create two subdirs - test and home under it.
2512
name_prefix = osutils.pathjoin(TestCaseWithMemoryTransport.TEST_ROOT,
2513
self._getTestDirPrefix())
2515
for i in range(100):
2516
if os.path.exists(name):
2517
name = name_prefix + '_' + str(i)
2519
# now create test and home directories within this dir
2520
self.test_base_dir = name
2521
self.addCleanup(self.deleteTestDir)
2522
os.mkdir(self.test_base_dir)
2524
self.permit_dir(self.test_base_dir)
2525
# 'sprouting' and 'init' of a branch both walk up the tree to find
2526
# stacking policy to honour; create a bzr dir with an unshared
2527
# repository (but not a branch - our code would be trying to escape
2528
# then!) to stop them, and permit it to be read.
2529
# control = bzrdir.BzrDir.create(self.test_base_dir)
2530
# control.create_repository()
2531
self.test_home_dir = self.test_base_dir + '/home'
2532
os.mkdir(self.test_home_dir)
2533
self.test_dir = self.test_base_dir + '/work'
2534
os.mkdir(self.test_dir)
2535
os.chdir(self.test_dir)
2536
# put name of test inside
2537
f = file(self.test_base_dir + '/name', 'w')
2543
def deleteTestDir(self):
2544
os.chdir(TestCaseWithMemoryTransport.TEST_ROOT)
2545
_rmtree_temp_dir(self.test_base_dir, test_id=self.id())
2547
def build_tree(self, shape, line_endings='binary', transport=None):
2548
"""Build a test tree according to a pattern.
2550
shape is a sequence of file specifications. If the final
2551
character is '/', a directory is created.
2553
This assumes that all the elements in the tree being built are new.
2555
This doesn't add anything to a branch.
2557
:type shape: list or tuple.
2558
:param line_endings: Either 'binary' or 'native'
2559
in binary mode, exact contents are written in native mode, the
2560
line endings match the default platform endings.
2561
:param transport: A transport to write to, for building trees on VFS's.
2562
If the transport is readonly or None, "." is opened automatically.
2565
if type(shape) not in (list, tuple):
2566
raise AssertionError("Parameter 'shape' should be "
2567
"a list or a tuple. Got %r instead" % (shape,))
2568
# It's OK to just create them using forward slashes on windows.
2569
if transport is None or transport.is_readonly():
2570
transport = _mod_transport.get_transport(".")
2572
self.assertIsInstance(name, basestring)
2574
transport.mkdir(urlutils.escape(name[:-1]))
2576
if line_endings == 'binary':
2578
elif line_endings == 'native':
2581
raise errors.BzrError(
2582
'Invalid line ending request %r' % line_endings)
2583
content = "contents of %s%s" % (name.encode('utf-8'), end)
2584
transport.put_bytes_non_atomic(urlutils.escape(name), content)
2586
build_tree_contents = staticmethod(treeshape.build_tree_contents)
2588
def assertInWorkingTree(self, path, root_path='.', tree=None):
2589
"""Assert whether path or paths are in the WorkingTree"""
2591
tree = workingtree.WorkingTree.open(root_path)
2592
if not isinstance(path, basestring):
2594
self.assertInWorkingTree(p, tree=tree)
2596
self.assertIsNot(tree.path2id(path), None,
2597
path+' not in working tree.')
2599
def assertNotInWorkingTree(self, path, root_path='.', tree=None):
2600
"""Assert whether path or paths are not in the WorkingTree"""
2602
tree = workingtree.WorkingTree.open(root_path)
2603
if not isinstance(path, basestring):
2605
self.assertNotInWorkingTree(p,tree=tree)
2607
self.assertIs(tree.path2id(path), None, path+' in working tree.')
2610
class TestCaseWithTransport(TestCaseInTempDir):
2611
"""A test case that provides get_url and get_readonly_url facilities.
2613
These back onto two transport servers, one for readonly access and one for
2616
If no explicit class is provided for readonly access, a
2617
ReadonlyTransportDecorator is used instead which allows the use of non disk
2618
based read write transports.
2620
If an explicit class is provided for readonly access, that server and the
2621
readwrite one must both define get_url() as resolving to os.getcwd().
2624
def get_vfs_only_server(self):
2625
"""See TestCaseWithMemoryTransport.
2627
This is useful for some tests with specific servers that need
2630
if self.__vfs_server is None:
2631
self.__vfs_server = self.vfs_transport_factory()
2632
self.start_server(self.__vfs_server)
2633
return self.__vfs_server
2635
def make_branch_and_tree(self, relpath, format=None):
2636
"""Create a branch on the transport and a tree locally.
2638
If the transport is not a LocalTransport, the Tree can't be created on
2639
the transport. In that case if the vfs_transport_factory is
2640
LocalURLServer the working tree is created in the local
2641
directory backing the transport, and the returned tree's branch and
2642
repository will also be accessed locally. Otherwise a lightweight
2643
checkout is created and returned.
2645
We do this because we can't physically create a tree in the local
2646
path, with a branch reference to the transport_factory url, and
2647
a branch + repository in the vfs_transport, unless the vfs_transport
2648
namespace is distinct from the local disk - the two branch objects
2649
would collide. While we could construct a tree with its branch object
2650
pointing at the transport_factory transport in memory, reopening it
2651
would behaving unexpectedly, and has in the past caused testing bugs
2652
when we tried to do it that way.
2654
:param format: The BzrDirFormat.
2655
:returns: the WorkingTree.
2657
# TODO: always use the local disk path for the working tree,
2658
# this obviously requires a format that supports branch references
2659
# so check for that by checking bzrdir.BzrDirFormat.get_default_format()
2661
b = self.make_branch(relpath, format=format)
2663
return b.bzrdir.create_workingtree()
2664
except errors.NotLocalUrl:
2665
# We can only make working trees locally at the moment. If the
2666
# transport can't support them, then we keep the non-disk-backed
2667
# branch and create a local checkout.
2668
if self.vfs_transport_factory is test_server.LocalURLServer:
2669
# the branch is colocated on disk, we cannot create a checkout.
2670
# hopefully callers will expect this.
2671
local_controldir= bzrdir.BzrDir.open(self.get_vfs_only_url(relpath))
2672
wt = local_controldir.create_workingtree()
2673
if wt.branch._format != b._format:
2675
# Make sure that assigning to wt._branch fixes wt.branch,
2676
# in case the implementation details of workingtree objects
2678
self.assertIs(b, wt.branch)
2681
return b.create_checkout(relpath, lightweight=True)
2683
def assertIsDirectory(self, relpath, transport):
2684
"""Assert that relpath within transport is a directory.
2686
This may not be possible on all transports; in that case it propagates
2687
a TransportNotPossible.
2690
mode = transport.stat(relpath).st_mode
2691
except errors.NoSuchFile:
2692
self.fail("path %s is not a directory; no such file"
2694
if not stat.S_ISDIR(mode):
2695
self.fail("path %s is not a directory; has mode %#o"
2698
def assertTreesEqual(self, left, right):
2699
"""Check that left and right have the same content and properties."""
2700
# we use a tree delta to check for equality of the content, and we
2701
# manually check for equality of other things such as the parents list.
2702
self.assertEqual(left.get_parent_ids(), right.get_parent_ids())
2703
differences = left.changes_from(right)
2704
self.assertFalse(differences.has_changed(),
2705
"Trees %r and %r are different: %r" % (left, right, differences))
2708
super(TestCaseWithTransport, self).setUp()
2709
self.__vfs_server = None
2711
def disable_missing_extensions_warning(self):
2712
"""Some tests expect a precise stderr content.
2714
There is no point in forcing them to duplicate the extension related
2717
config.GlobalConfig().set_user_option('ignore_missing_extensions', True)
2720
class ChrootedTestCase(TestCaseWithTransport):
2721
"""A support class that provides readonly urls outside the local namespace.
2723
This is done by checking if self.transport_server is a MemoryServer. if it
2724
is then we are chrooted already, if it is not then an HttpServer is used
2727
TODO RBC 20060127: make this an option to TestCaseWithTransport so it can
2728
be used without needed to redo it when a different
2729
subclass is in use ?
2733
from bzrlib.tests import http_server
2734
super(ChrootedTestCase, self).setUp()
2735
if not self.vfs_transport_factory == memory.MemoryServer:
2736
self.transport_readonly_server = http_server.HttpServer
2739
def condition_id_re(pattern):
2740
"""Create a condition filter which performs a re check on a test's id.
2742
:param pattern: A regular expression string.
2743
:return: A callable that returns True if the re matches.
2745
filter_re = re.compile(pattern, 0)
2746
def condition(test):
2748
return filter_re.search(test_id)
2752
def condition_isinstance(klass_or_klass_list):
2753
"""Create a condition filter which returns isinstance(param, klass).
2755
:return: A callable which when called with one parameter obj return the
2756
result of isinstance(obj, klass_or_klass_list).
2759
return isinstance(obj, klass_or_klass_list)
2763
def condition_id_in_list(id_list):
2764
"""Create a condition filter which verify that test's id in a list.
2766
:param id_list: A TestIdList object.
2767
:return: A callable that returns True if the test's id appears in the list.
2769
def condition(test):
2770
return id_list.includes(test.id())
2774
def condition_id_startswith(starts):
2775
"""Create a condition filter verifying that test's id starts with a string.
2777
:param starts: A list of string.
2778
:return: A callable that returns True if the test's id starts with one of
2781
def condition(test):
2782
for start in starts:
2783
if test.id().startswith(start):
2789
def exclude_tests_by_condition(suite, condition):
2790
"""Create a test suite which excludes some tests from suite.
2792
:param suite: The suite to get tests from.
2793
:param condition: A callable whose result evaluates True when called with a
2794
test case which should be excluded from the result.
2795
:return: A suite which contains the tests found in suite that fail
2799
for test in iter_suite_tests(suite):
2800
if not condition(test):
2802
return TestUtil.TestSuite(result)
2805
def filter_suite_by_condition(suite, condition):
2806
"""Create a test suite by filtering another one.
2808
:param suite: The source suite.
2809
:param condition: A callable whose result evaluates True when called with a
2810
test case which should be included in the result.
2811
:return: A suite which contains the tests found in suite that pass
2815
for test in iter_suite_tests(suite):
2818
return TestUtil.TestSuite(result)
2821
def filter_suite_by_re(suite, pattern):
2822
"""Create a test suite by filtering another one.
2824
:param suite: the source suite
2825
:param pattern: pattern that names must match
2826
:returns: the newly created suite
2828
condition = condition_id_re(pattern)
2829
result_suite = filter_suite_by_condition(suite, condition)
2833
def filter_suite_by_id_list(suite, test_id_list):
2834
"""Create a test suite by filtering another one.
2836
:param suite: The source suite.
2837
:param test_id_list: A list of the test ids to keep as strings.
2838
:returns: the newly created suite
2840
condition = condition_id_in_list(test_id_list)
2841
result_suite = filter_suite_by_condition(suite, condition)
2845
def filter_suite_by_id_startswith(suite, start):
2846
"""Create a test suite by filtering another one.
2848
:param suite: The source suite.
2849
:param start: A list of string the test id must start with one of.
2850
:returns: the newly created suite
2852
condition = condition_id_startswith(start)
2853
result_suite = filter_suite_by_condition(suite, condition)
2857
def exclude_tests_by_re(suite, pattern):
2858
"""Create a test suite which excludes some tests from suite.
2860
:param suite: The suite to get tests from.
2861
:param pattern: A regular expression string. Test ids that match this
2862
pattern will be excluded from the result.
2863
:return: A TestSuite that contains all the tests from suite without the
2864
tests that matched pattern. The order of tests is the same as it was in
2867
return exclude_tests_by_condition(suite, condition_id_re(pattern))
2870
def preserve_input(something):
2871
"""A helper for performing test suite transformation chains.
2873
:param something: Anything you want to preserve.
2879
def randomize_suite(suite):
2880
"""Return a new TestSuite with suite's tests in random order.
2882
The tests in the input suite are flattened into a single suite in order to
2883
accomplish this. Any nested TestSuites are removed to provide global
2886
tests = list(iter_suite_tests(suite))
2887
random.shuffle(tests)
2888
return TestUtil.TestSuite(tests)
2891
def split_suite_by_condition(suite, condition):
2892
"""Split a test suite into two by a condition.
2894
:param suite: The suite to split.
2895
:param condition: The condition to match on. Tests that match this
2896
condition are returned in the first test suite, ones that do not match
2897
are in the second suite.
2898
:return: A tuple of two test suites, where the first contains tests from
2899
suite matching the condition, and the second contains the remainder
2900
from suite. The order within each output suite is the same as it was in
2905
for test in iter_suite_tests(suite):
2907
matched.append(test)
2909
did_not_match.append(test)
2910
return TestUtil.TestSuite(matched), TestUtil.TestSuite(did_not_match)
2913
def split_suite_by_re(suite, pattern):
2914
"""Split a test suite into two by a regular expression.
2916
:param suite: The suite to split.
2917
:param pattern: A regular expression string. Test ids that match this
2918
pattern will be in the first test suite returned, and the others in the
2919
second test suite returned.
2920
:return: A tuple of two test suites, where the first contains tests from
2921
suite matching pattern, and the second contains the remainder from
2922
suite. The order within each output suite is the same as it was in
2925
return split_suite_by_condition(suite, condition_id_re(pattern))
2928
def run_suite(suite, name='test', verbose=False, pattern=".*",
2929
stop_on_failure=False,
2930
transport=None, lsprof_timed=None, bench_history=None,
2931
matching_tests_first=None,
2934
exclude_pattern=None,
2937
suite_decorators=None,
2939
result_decorators=None,
2941
"""Run a test suite for bzr selftest.
2943
:param runner_class: The class of runner to use. Must support the
2944
constructor arguments passed by run_suite which are more than standard
2946
:return: A boolean indicating success.
2948
TestCase._gather_lsprof_in_benchmarks = lsprof_timed
2953
if runner_class is None:
2954
runner_class = TextTestRunner
2957
runner = runner_class(stream=stream,
2959
verbosity=verbosity,
2960
bench_history=bench_history,
2962
result_decorators=result_decorators,
2964
runner.stop_on_failure=stop_on_failure
2965
# built in decorator factories:
2967
random_order(random_seed, runner),
2968
exclude_tests(exclude_pattern),
2970
if matching_tests_first:
2971
decorators.append(tests_first(pattern))
2973
decorators.append(filter_tests(pattern))
2974
if suite_decorators:
2975
decorators.extend(suite_decorators)
2976
# tell the result object how many tests will be running: (except if
2977
# --parallel=fork is being used. Robert said he will provide a better
2978
# progress design later -- vila 20090817)
2979
if fork_decorator not in decorators:
2980
decorators.append(CountingDecorator)
2981
for decorator in decorators:
2982
suite = decorator(suite)
2984
# Done after test suite decoration to allow randomisation etc
2985
# to take effect, though that is of marginal benefit.
2987
stream.write("Listing tests only ...\n")
2988
for t in iter_suite_tests(suite):
2989
stream.write("%s\n" % (t.id()))
2991
result = runner.run(suite)
2993
return result.wasStrictlySuccessful()
2995
return result.wasSuccessful()
2998
# A registry where get() returns a suite decorator.
2999
parallel_registry = registry.Registry()
3002
def fork_decorator(suite):
3003
if getattr(os, "fork", None) is None:
3004
raise errors.BzrCommandError("platform does not support fork,"
3005
" try --parallel=subprocess instead.")
3006
concurrency = osutils.local_concurrency()
3007
if concurrency == 1:
3009
from testtools import ConcurrentTestSuite
3010
return ConcurrentTestSuite(suite, fork_for_tests)
3011
parallel_registry.register('fork', fork_decorator)
3014
def subprocess_decorator(suite):
3015
concurrency = osutils.local_concurrency()
3016
if concurrency == 1:
3018
from testtools import ConcurrentTestSuite
3019
return ConcurrentTestSuite(suite, reinvoke_for_tests)
3020
parallel_registry.register('subprocess', subprocess_decorator)
3023
def exclude_tests(exclude_pattern):
3024
"""Return a test suite decorator that excludes tests."""
3025
if exclude_pattern is None:
3026
return identity_decorator
3027
def decorator(suite):
3028
return ExcludeDecorator(suite, exclude_pattern)
3032
def filter_tests(pattern):
3034
return identity_decorator
3035
def decorator(suite):
3036
return FilterTestsDecorator(suite, pattern)
3040
def random_order(random_seed, runner):
3041
"""Return a test suite decorator factory for randomising tests order.
3043
:param random_seed: now, a string which casts to a long, or a long.
3044
:param runner: A test runner with a stream attribute to report on.
3046
if random_seed is None:
3047
return identity_decorator
3048
def decorator(suite):
3049
return RandomDecorator(suite, random_seed, runner.stream)
3053
def tests_first(pattern):
3055
return identity_decorator
3056
def decorator(suite):
3057
return TestFirstDecorator(suite, pattern)
3061
def identity_decorator(suite):
3066
class TestDecorator(TestUtil.TestSuite):
3067
"""A decorator for TestCase/TestSuite objects.
3069
Usually, subclasses should override __iter__(used when flattening test
3070
suites), which we do to filter, reorder, parallelise and so on, run() and
3074
def __init__(self, suite):
3075
TestUtil.TestSuite.__init__(self)
3078
def countTestCases(self):
3081
cases += test.countTestCases()
3088
def run(self, result):
3089
# Use iteration on self, not self._tests, to allow subclasses to hook
3092
if result.shouldStop:
3098
class CountingDecorator(TestDecorator):
3099
"""A decorator which calls result.progress(self.countTestCases)."""
3101
def run(self, result):
3102
progress_method = getattr(result, 'progress', None)
3103
if callable(progress_method):
3104
progress_method(self.countTestCases(), SUBUNIT_SEEK_SET)
3105
return super(CountingDecorator, self).run(result)
3108
class ExcludeDecorator(TestDecorator):
3109
"""A decorator which excludes test matching an exclude pattern."""
3111
def __init__(self, suite, exclude_pattern):
3112
TestDecorator.__init__(self, suite)
3113
self.exclude_pattern = exclude_pattern
3114
self.excluded = False
3118
return iter(self._tests)
3119
self.excluded = True
3120
suite = exclude_tests_by_re(self, self.exclude_pattern)
3122
self.addTests(suite)
3123
return iter(self._tests)
3126
class FilterTestsDecorator(TestDecorator):
3127
"""A decorator which filters tests to those matching a pattern."""
3129
def __init__(self, suite, pattern):
3130
TestDecorator.__init__(self, suite)
3131
self.pattern = pattern
3132
self.filtered = False
3136
return iter(self._tests)
3137
self.filtered = True
3138
suite = filter_suite_by_re(self, self.pattern)
3140
self.addTests(suite)
3141
return iter(self._tests)
3144
class RandomDecorator(TestDecorator):
3145
"""A decorator which randomises the order of its tests."""
3147
def __init__(self, suite, random_seed, stream):
3148
TestDecorator.__init__(self, suite)
3149
self.random_seed = random_seed
3150
self.randomised = False
3151
self.stream = stream
3155
return iter(self._tests)
3156
self.randomised = True
3157
self.stream.write("Randomizing test order using seed %s\n\n" %
3158
(self.actual_seed()))
3159
# Initialise the random number generator.
3160
random.seed(self.actual_seed())
3161
suite = randomize_suite(self)
3163
self.addTests(suite)
3164
return iter(self._tests)
3166
def actual_seed(self):
3167
if self.random_seed == "now":
3168
# We convert the seed to a long to make it reuseable across
3169
# invocations (because the user can reenter it).
3170
self.random_seed = long(time.time())
3172
# Convert the seed to a long if we can
3174
self.random_seed = long(self.random_seed)
3177
return self.random_seed
3180
class TestFirstDecorator(TestDecorator):
3181
"""A decorator which moves named tests to the front."""
3183
def __init__(self, suite, pattern):
3184
TestDecorator.__init__(self, suite)
3185
self.pattern = pattern
3186
self.filtered = False
3190
return iter(self._tests)
3191
self.filtered = True
3192
suites = split_suite_by_re(self, self.pattern)
3194
self.addTests(suites)
3195
return iter(self._tests)
3198
def partition_tests(suite, count):
3199
"""Partition suite into count lists of tests."""
3200
# This just assigns tests in a round-robin fashion. On one hand this
3201
# splits up blocks of related tests that might run faster if they shared
3202
# resources, but on the other it avoids assigning blocks of slow tests to
3203
# just one partition. So the slowest partition shouldn't be much slower
3205
partitions = [list() for i in range(count)]
3206
tests = iter_suite_tests(suite)
3207
for partition, test in itertools.izip(itertools.cycle(partitions), tests):
3208
partition.append(test)
3212
def workaround_zealous_crypto_random():
3213
"""Crypto.Random want to help us being secure, but we don't care here.
3215
This workaround some test failure related to the sftp server. Once paramiko
3216
stop using the controversial API in Crypto.Random, we may get rid of it.
3219
from Crypto.Random import atfork
3225
def fork_for_tests(suite):
3226
"""Take suite and start up one runner per CPU by forking()
3228
:return: An iterable of TestCase-like objects which can each have
3229
run(result) called on them to feed tests to result.
3231
concurrency = osutils.local_concurrency()
3233
from subunit import TestProtocolClient, ProtocolTestCase
3234
from subunit.test_results import AutoTimingTestResultDecorator
3235
class TestInOtherProcess(ProtocolTestCase):
3236
# Should be in subunit, I think. RBC.
3237
def __init__(self, stream, pid):
3238
ProtocolTestCase.__init__(self, stream)
3241
def run(self, result):
3243
ProtocolTestCase.run(self, result)
3245
os.waitpid(self.pid, 0)
3247
test_blocks = partition_tests(suite, concurrency)
3248
for process_tests in test_blocks:
3249
process_suite = TestUtil.TestSuite()
3250
process_suite.addTests(process_tests)
3251
c2pread, c2pwrite = os.pipe()
3254
workaround_zealous_crypto_random()
3257
# Leave stderr and stdout open so we can see test noise
3258
# Close stdin so that the child goes away if it decides to
3259
# read from stdin (otherwise its a roulette to see what
3260
# child actually gets keystrokes for pdb etc).
3263
stream = os.fdopen(c2pwrite, 'wb', 1)
3264
subunit_result = AutoTimingTestResultDecorator(
3265
TestProtocolClient(stream))
3266
process_suite.run(subunit_result)
3271
stream = os.fdopen(c2pread, 'rb', 1)
3272
test = TestInOtherProcess(stream, pid)
3277
def reinvoke_for_tests(suite):
3278
"""Take suite and start up one runner per CPU using subprocess().
3280
:return: An iterable of TestCase-like objects which can each have
3281
run(result) called on them to feed tests to result.
3283
concurrency = osutils.local_concurrency()
3285
from subunit import ProtocolTestCase
3286
class TestInSubprocess(ProtocolTestCase):
3287
def __init__(self, process, name):
3288
ProtocolTestCase.__init__(self, process.stdout)
3289
self.process = process
3290
self.process.stdin.close()
3293
def run(self, result):
3295
ProtocolTestCase.run(self, result)
3298
os.unlink(self.name)
3299
# print "pid %d finished" % finished_process
3300
test_blocks = partition_tests(suite, concurrency)
3301
for process_tests in test_blocks:
3302
# ugly; currently reimplement rather than reuses TestCase methods.
3303
bzr_path = os.path.dirname(os.path.dirname(bzrlib.__file__))+'/bzr'
3304
if not os.path.isfile(bzr_path):
3305
# We are probably installed. Assume sys.argv is the right file
3306
bzr_path = sys.argv[0]
3307
bzr_path = [bzr_path]
3308
if sys.platform == "win32":
3309
# if we're on windows, we can't execute the bzr script directly
3310
bzr_path = [sys.executable] + bzr_path
3311
fd, test_list_file_name = tempfile.mkstemp()
3312
test_list_file = os.fdopen(fd, 'wb', 1)
3313
for test in process_tests:
3314
test_list_file.write(test.id() + '\n')
3315
test_list_file.close()
3317
argv = bzr_path + ['selftest', '--load-list', test_list_file_name,
3319
if '--no-plugins' in sys.argv:
3320
argv.append('--no-plugins')
3321
# stderr=subprocess.STDOUT would be ideal, but until we prevent
3322
# noise on stderr it can interrupt the subunit protocol.
3323
process = subprocess.Popen(argv, stdin=subprocess.PIPE,
3324
stdout=subprocess.PIPE,
3325
stderr=subprocess.PIPE,
3327
test = TestInSubprocess(process, test_list_file_name)
3330
os.unlink(test_list_file_name)
3335
class ForwardingResult(unittest.TestResult):
3337
def __init__(self, target):
3338
unittest.TestResult.__init__(self)
3339
self.result = target
3341
def startTest(self, test):
3342
self.result.startTest(test)
3344
def stopTest(self, test):
3345
self.result.stopTest(test)
3347
def startTestRun(self):
3348
self.result.startTestRun()
3350
def stopTestRun(self):
3351
self.result.stopTestRun()
3353
def addSkip(self, test, reason):
3354
self.result.addSkip(test, reason)
3356
def addSuccess(self, test):
3357
self.result.addSuccess(test)
3359
def addError(self, test, err):
3360
self.result.addError(test, err)
3362
def addFailure(self, test, err):
3363
self.result.addFailure(test, err)
3364
ForwardingResult = testtools.ExtendedToOriginalDecorator
3367
class ProfileResult(ForwardingResult):
3368
"""Generate profiling data for all activity between start and success.
3370
The profile data is appended to the test's _benchcalls attribute and can
3371
be accessed by the forwarded-to TestResult.
3373
While it might be cleaner do accumulate this in stopTest, addSuccess is
3374
where our existing output support for lsprof is, and this class aims to
3375
fit in with that: while it could be moved it's not necessary to accomplish
3376
test profiling, nor would it be dramatically cleaner.
3379
def startTest(self, test):
3380
self.profiler = bzrlib.lsprof.BzrProfiler()
3381
# Prevent deadlocks in tests that use lsprof: those tests will
3383
bzrlib.lsprof.BzrProfiler.profiler_block = 0
3384
self.profiler.start()
3385
ForwardingResult.startTest(self, test)
3387
def addSuccess(self, test):
3388
stats = self.profiler.stop()
3390
calls = test._benchcalls
3391
except AttributeError:
3392
test._benchcalls = []
3393
calls = test._benchcalls
3394
calls.append(((test.id(), "", ""), stats))
3395
ForwardingResult.addSuccess(self, test)
3397
def stopTest(self, test):
3398
ForwardingResult.stopTest(self, test)
3399
self.profiler = None
3402
# Controlled by "bzr selftest -E=..." option
3403
# Currently supported:
3404
# -Eallow_debug Will no longer clear debug.debug_flags() so it
3405
# preserves any flags supplied at the command line.
3406
# -Edisable_lock_checks Turns errors in mismatched locks into simple prints
3407
# rather than failing tests. And no longer raise
3408
# LockContention when fctnl locks are not being used
3409
# with proper exclusion rules.
3410
# -Ethreads Will display thread ident at creation/join time to
3411
# help track thread leaks
3412
selftest_debug_flags = set()
3415
def selftest(verbose=False, pattern=".*", stop_on_failure=True,
3417
test_suite_factory=None,
3420
matching_tests_first=None,
3423
exclude_pattern=None,
3429
suite_decorators=None,
3433
"""Run the whole test suite under the enhanced runner"""
3434
# XXX: Very ugly way to do this...
3435
# Disable warning about old formats because we don't want it to disturb
3436
# any blackbox tests.
3437
from bzrlib import repository
3438
repository._deprecation_warning_done = True
3440
global default_transport
3441
if transport is None:
3442
transport = default_transport
3443
old_transport = default_transport
3444
default_transport = transport
3445
global selftest_debug_flags
3446
old_debug_flags = selftest_debug_flags
3447
if debug_flags is not None:
3448
selftest_debug_flags = set(debug_flags)
3450
if load_list is None:
3453
keep_only = load_test_id_list(load_list)
3455
starting_with = [test_prefix_alias_registry.resolve_alias(start)
3456
for start in starting_with]
3457
if test_suite_factory is None:
3458
# Reduce loading time by loading modules based on the starting_with
3460
suite = test_suite(keep_only, starting_with)
3462
suite = test_suite_factory()
3464
# But always filter as requested.
3465
suite = filter_suite_by_id_startswith(suite, starting_with)
3466
result_decorators = []
3468
result_decorators.append(ProfileResult)
3469
return run_suite(suite, 'testbzr', verbose=verbose, pattern=pattern,
3470
stop_on_failure=stop_on_failure,
3471
transport=transport,
3472
lsprof_timed=lsprof_timed,
3473
bench_history=bench_history,
3474
matching_tests_first=matching_tests_first,
3475
list_only=list_only,
3476
random_seed=random_seed,
3477
exclude_pattern=exclude_pattern,
3479
runner_class=runner_class,
3480
suite_decorators=suite_decorators,
3482
result_decorators=result_decorators,
3485
default_transport = old_transport
3486
selftest_debug_flags = old_debug_flags
3489
def load_test_id_list(file_name):
3490
"""Load a test id list from a text file.
3492
The format is one test id by line. No special care is taken to impose
3493
strict rules, these test ids are used to filter the test suite so a test id
3494
that do not match an existing test will do no harm. This allows user to add
3495
comments, leave blank lines, etc.
3499
ftest = open(file_name, 'rt')
3501
if e.errno != errno.ENOENT:
3504
raise errors.NoSuchFile(file_name)
3506
for test_name in ftest.readlines():
3507
test_list.append(test_name.strip())
3512
def suite_matches_id_list(test_suite, id_list):
3513
"""Warns about tests not appearing or appearing more than once.
3515
:param test_suite: A TestSuite object.
3516
:param test_id_list: The list of test ids that should be found in
3519
:return: (absents, duplicates) absents is a list containing the test found
3520
in id_list but not in test_suite, duplicates is a list containing the
3521
test found multiple times in test_suite.
3523
When using a prefined test id list, it may occurs that some tests do not
3524
exist anymore or that some tests use the same id. This function warns the
3525
tester about potential problems in his workflow (test lists are volatile)
3526
or in the test suite itself (using the same id for several tests does not
3527
help to localize defects).
3529
# Build a dict counting id occurrences
3531
for test in iter_suite_tests(test_suite):
3533
tests[id] = tests.get(id, 0) + 1
3538
occurs = tests.get(id, 0)
3540
not_found.append(id)
3542
duplicates.append(id)
3544
return not_found, duplicates
3547
class TestIdList(object):
3548
"""Test id list to filter a test suite.
3550
Relying on the assumption that test ids are built as:
3551
<module>[.<class>.<method>][(<param>+)], <module> being in python dotted
3552
notation, this class offers methods to :
3553
- avoid building a test suite for modules not refered to in the test list,
3554
- keep only the tests listed from the module test suite.
3557
def __init__(self, test_id_list):
3558
# When a test suite needs to be filtered against us we compare test ids
3559
# for equality, so a simple dict offers a quick and simple solution.
3560
self.tests = dict().fromkeys(test_id_list, True)
3562
# While unittest.TestCase have ids like:
3563
# <module>.<class>.<method>[(<param+)],
3564
# doctest.DocTestCase can have ids like:
3567
# <module>.<function>
3568
# <module>.<class>.<method>
3570
# Since we can't predict a test class from its name only, we settle on
3571
# a simple constraint: a test id always begins with its module name.
3574
for test_id in test_id_list:
3575
parts = test_id.split('.')
3576
mod_name = parts.pop(0)
3577
modules[mod_name] = True
3579
mod_name += '.' + part
3580
modules[mod_name] = True
3581
self.modules = modules
3583
def refers_to(self, module_name):
3584
"""Is there tests for the module or one of its sub modules."""
3585
return self.modules.has_key(module_name)
3587
def includes(self, test_id):
3588
return self.tests.has_key(test_id)
3591
class TestPrefixAliasRegistry(registry.Registry):
3592
"""A registry for test prefix aliases.
3594
This helps implement shorcuts for the --starting-with selftest
3595
option. Overriding existing prefixes is not allowed but not fatal (a
3596
warning will be emitted).
3599
def register(self, key, obj, help=None, info=None,
3600
override_existing=False):
3601
"""See Registry.register.
3603
Trying to override an existing alias causes a warning to be emitted,
3604
not a fatal execption.
3607
super(TestPrefixAliasRegistry, self).register(
3608
key, obj, help=help, info=info, override_existing=False)
3610
actual = self.get(key)
3611
note('Test prefix alias %s is already used for %s, ignoring %s'
3612
% (key, actual, obj))
3614
def resolve_alias(self, id_start):
3615
"""Replace the alias by the prefix in the given string.
3617
Using an unknown prefix is an error to help catching typos.
3619
parts = id_start.split('.')
3621
parts[0] = self.get(parts[0])
3623
raise errors.BzrCommandError(
3624
'%s is not a known test prefix alias' % parts[0])
3625
return '.'.join(parts)
3628
test_prefix_alias_registry = TestPrefixAliasRegistry()
3629
"""Registry of test prefix aliases."""
3632
# This alias allows to detect typos ('bzrlin.') by making all valid test ids
3633
# appear prefixed ('bzrlib.' is "replaced" by 'bzrlib.').
3634
test_prefix_alias_registry.register('bzrlib', 'bzrlib')
3636
# Obvious highest levels prefixes, feel free to add your own via a plugin
3637
test_prefix_alias_registry.register('bd', 'bzrlib.doc')
3638
test_prefix_alias_registry.register('bu', 'bzrlib.utils')
3639
test_prefix_alias_registry.register('bt', 'bzrlib.tests')
3640
test_prefix_alias_registry.register('bb', 'bzrlib.tests.blackbox')
3641
test_prefix_alias_registry.register('bp', 'bzrlib.plugins')
3644
def _test_suite_testmod_names():
3645
"""Return the standard list of test module names to test."""
3648
'bzrlib.tests.blackbox',
3649
'bzrlib.tests.commands',
3650
'bzrlib.tests.doc_generate',
3651
'bzrlib.tests.per_branch',
3652
'bzrlib.tests.per_bzrdir',
3653
'bzrlib.tests.per_controldir',
3654
'bzrlib.tests.per_controldir_colo',
3655
'bzrlib.tests.per_foreign_vcs',
3656
'bzrlib.tests.per_interrepository',
3657
'bzrlib.tests.per_intertree',
3658
'bzrlib.tests.per_inventory',
3659
'bzrlib.tests.per_interbranch',
3660
'bzrlib.tests.per_lock',
3661
'bzrlib.tests.per_merger',
3662
'bzrlib.tests.per_transport',
3663
'bzrlib.tests.per_tree',
3664
'bzrlib.tests.per_pack_repository',
3665
'bzrlib.tests.per_repository',
3666
'bzrlib.tests.per_repository_chk',
3667
'bzrlib.tests.per_repository_reference',
3668
'bzrlib.tests.per_uifactory',
3669
'bzrlib.tests.per_versionedfile',
3670
'bzrlib.tests.per_workingtree',
3671
'bzrlib.tests.test__annotator',
3672
'bzrlib.tests.test__bencode',
3673
'bzrlib.tests.test__btree_serializer',
3674
'bzrlib.tests.test__chk_map',
3675
'bzrlib.tests.test__dirstate_helpers',
3676
'bzrlib.tests.test__groupcompress',
3677
'bzrlib.tests.test__known_graph',
3678
'bzrlib.tests.test__rio',
3679
'bzrlib.tests.test__simple_set',
3680
'bzrlib.tests.test__static_tuple',
3681
'bzrlib.tests.test__walkdirs_win32',
3682
'bzrlib.tests.test_ancestry',
3683
'bzrlib.tests.test_annotate',
3684
'bzrlib.tests.test_api',
3685
'bzrlib.tests.test_atomicfile',
3686
'bzrlib.tests.test_bad_files',
3687
'bzrlib.tests.test_bisect_multi',
3688
'bzrlib.tests.test_branch',
3689
'bzrlib.tests.test_branchbuilder',
3690
'bzrlib.tests.test_btree_index',
3691
'bzrlib.tests.test_bugtracker',
3692
'bzrlib.tests.test_bundle',
3693
'bzrlib.tests.test_bzrdir',
3694
'bzrlib.tests.test__chunks_to_lines',
3695
'bzrlib.tests.test_cache_utf8',
3696
'bzrlib.tests.test_chk_map',
3697
'bzrlib.tests.test_chk_serializer',
3698
'bzrlib.tests.test_chunk_writer',
3699
'bzrlib.tests.test_clean_tree',
3700
'bzrlib.tests.test_cleanup',
3701
'bzrlib.tests.test_cmdline',
3702
'bzrlib.tests.test_commands',
3703
'bzrlib.tests.test_commit',
3704
'bzrlib.tests.test_commit_merge',
3705
'bzrlib.tests.test_config',
3706
'bzrlib.tests.test_conflicts',
3707
'bzrlib.tests.test_counted_lock',
3708
'bzrlib.tests.test_crash',
3709
'bzrlib.tests.test_decorators',
3710
'bzrlib.tests.test_delta',
3711
'bzrlib.tests.test_debug',
3712
'bzrlib.tests.test_deprecated_graph',
3713
'bzrlib.tests.test_diff',
3714
'bzrlib.tests.test_directory_service',
3715
'bzrlib.tests.test_dirstate',
3716
'bzrlib.tests.test_email_message',
3717
'bzrlib.tests.test_eol_filters',
3718
'bzrlib.tests.test_errors',
3719
'bzrlib.tests.test_export',
3720
'bzrlib.tests.test_extract',
3721
'bzrlib.tests.test_fetch',
3722
'bzrlib.tests.test_fixtures',
3723
'bzrlib.tests.test_fifo_cache',
3724
'bzrlib.tests.test_filters',
3725
'bzrlib.tests.test_ftp_transport',
3726
'bzrlib.tests.test_foreign',
3727
'bzrlib.tests.test_generate_docs',
3728
'bzrlib.tests.test_generate_ids',
3729
'bzrlib.tests.test_globbing',
3730
'bzrlib.tests.test_gpg',
3731
'bzrlib.tests.test_graph',
3732
'bzrlib.tests.test_groupcompress',
3733
'bzrlib.tests.test_hashcache',
3734
'bzrlib.tests.test_help',
3735
'bzrlib.tests.test_hooks',
3736
'bzrlib.tests.test_http',
3737
'bzrlib.tests.test_http_response',
3738
'bzrlib.tests.test_https_ca_bundle',
3739
'bzrlib.tests.test_identitymap',
3740
'bzrlib.tests.test_ignores',
3741
'bzrlib.tests.test_index',
3742
'bzrlib.tests.test_import_tariff',
3743
'bzrlib.tests.test_info',
3744
'bzrlib.tests.test_inv',
3745
'bzrlib.tests.test_inventory_delta',
3746
'bzrlib.tests.test_knit',
3747
'bzrlib.tests.test_lazy_import',
3748
'bzrlib.tests.test_lazy_regex',
3749
'bzrlib.tests.test_library_state',
3750
'bzrlib.tests.test_lock',
3751
'bzrlib.tests.test_lockable_files',
3752
'bzrlib.tests.test_lockdir',
3753
'bzrlib.tests.test_log',
3754
'bzrlib.tests.test_lru_cache',
3755
'bzrlib.tests.test_lsprof',
3756
'bzrlib.tests.test_mail_client',
3757
'bzrlib.tests.test_matchers',
3758
'bzrlib.tests.test_memorytree',
3759
'bzrlib.tests.test_merge',
3760
'bzrlib.tests.test_merge3',
3761
'bzrlib.tests.test_merge_core',
3762
'bzrlib.tests.test_merge_directive',
3763
'bzrlib.tests.test_missing',
3764
'bzrlib.tests.test_msgeditor',
3765
'bzrlib.tests.test_multiparent',
3766
'bzrlib.tests.test_mutabletree',
3767
'bzrlib.tests.test_nonascii',
3768
'bzrlib.tests.test_options',
3769
'bzrlib.tests.test_osutils',
3770
'bzrlib.tests.test_osutils_encodings',
3771
'bzrlib.tests.test_pack',
3772
'bzrlib.tests.test_patch',
3773
'bzrlib.tests.test_patches',
3774
'bzrlib.tests.test_permissions',
3775
'bzrlib.tests.test_plugins',
3776
'bzrlib.tests.test_progress',
3777
'bzrlib.tests.test_read_bundle',
3778
'bzrlib.tests.test_reconcile',
3779
'bzrlib.tests.test_reconfigure',
3780
'bzrlib.tests.test_registry',
3781
'bzrlib.tests.test_remote',
3782
'bzrlib.tests.test_rename_map',
3783
'bzrlib.tests.test_repository',
3784
'bzrlib.tests.test_revert',
3785
'bzrlib.tests.test_revision',
3786
'bzrlib.tests.test_revisionspec',
3787
'bzrlib.tests.test_revisiontree',
3788
'bzrlib.tests.test_rio',
3789
'bzrlib.tests.test_rules',
3790
'bzrlib.tests.test_sampler',
3791
'bzrlib.tests.test_script',
3792
'bzrlib.tests.test_selftest',
3793
'bzrlib.tests.test_serializer',
3794
'bzrlib.tests.test_setup',
3795
'bzrlib.tests.test_sftp_transport',
3796
'bzrlib.tests.test_shelf',
3797
'bzrlib.tests.test_shelf_ui',
3798
'bzrlib.tests.test_smart',
3799
'bzrlib.tests.test_smart_add',
3800
'bzrlib.tests.test_smart_request',
3801
'bzrlib.tests.test_smart_transport',
3802
'bzrlib.tests.test_smtp_connection',
3803
'bzrlib.tests.test_source',
3804
'bzrlib.tests.test_ssh_transport',
3805
'bzrlib.tests.test_status',
3806
'bzrlib.tests.test_store',
3807
'bzrlib.tests.test_strace',
3808
'bzrlib.tests.test_subsume',
3809
'bzrlib.tests.test_switch',
3810
'bzrlib.tests.test_symbol_versioning',
3811
'bzrlib.tests.test_tag',
3812
'bzrlib.tests.test_test_server',
3813
'bzrlib.tests.test_testament',
3814
'bzrlib.tests.test_textfile',
3815
'bzrlib.tests.test_textmerge',
3816
'bzrlib.tests.test_timestamp',
3817
'bzrlib.tests.test_trace',
3818
'bzrlib.tests.test_transactions',
3819
'bzrlib.tests.test_transform',
3820
'bzrlib.tests.test_transport',
3821
'bzrlib.tests.test_transport_log',
3822
'bzrlib.tests.test_tree',
3823
'bzrlib.tests.test_treebuilder',
3824
'bzrlib.tests.test_treeshape',
3825
'bzrlib.tests.test_tsort',
3826
'bzrlib.tests.test_tuned_gzip',
3827
'bzrlib.tests.test_ui',
3828
'bzrlib.tests.test_uncommit',
3829
'bzrlib.tests.test_upgrade',
3830
'bzrlib.tests.test_upgrade_stacked',
3831
'bzrlib.tests.test_urlutils',
3832
'bzrlib.tests.test_version',
3833
'bzrlib.tests.test_version_info',
3834
'bzrlib.tests.test_versionedfile',
3835
'bzrlib.tests.test_weave',
3836
'bzrlib.tests.test_whitebox',
3837
'bzrlib.tests.test_win32utils',
3838
'bzrlib.tests.test_workingtree',
3839
'bzrlib.tests.test_workingtree_4',
3840
'bzrlib.tests.test_wsgi',
3841
'bzrlib.tests.test_xml',
3845
def _test_suite_modules_to_doctest():
3846
"""Return the list of modules to doctest."""
3848
# GZ 2009-03-31: No docstrings with -OO so there's nothing to doctest
3852
'bzrlib.branchbuilder',
3853
'bzrlib.decorators',
3856
'bzrlib.iterablefile',
3860
'bzrlib.symbol_versioning',
3862
'bzrlib.tests.fixtures',
3864
'bzrlib.version_info_formats.format_custom',
3868
def test_suite(keep_only=None, starting_with=None):
3869
"""Build and return TestSuite for the whole of bzrlib.
3871
:param keep_only: A list of test ids limiting the suite returned.
3873
:param starting_with: An id limiting the suite returned to the tests
3876
This function can be replaced if you need to change the default test
3877
suite on a global basis, but it is not encouraged.
3880
loader = TestUtil.TestLoader()
3882
if keep_only is not None:
3883
id_filter = TestIdList(keep_only)
3885
# We take precedence over keep_only because *at loading time* using
3886
# both options means we will load less tests for the same final result.
3887
def interesting_module(name):
3888
for start in starting_with:
3890
# Either the module name starts with the specified string
3891
name.startswith(start)
3892
# or it may contain tests starting with the specified string
3893
or start.startswith(name)
3897
loader = TestUtil.FilteredByModuleTestLoader(interesting_module)
3899
elif keep_only is not None:
3900
loader = TestUtil.FilteredByModuleTestLoader(id_filter.refers_to)
3901
def interesting_module(name):
3902
return id_filter.refers_to(name)
3905
loader = TestUtil.TestLoader()
3906
def interesting_module(name):
3907
# No filtering, all modules are interesting
3910
suite = loader.suiteClass()
3912
# modules building their suite with loadTestsFromModuleNames
3913
suite.addTest(loader.loadTestsFromModuleNames(_test_suite_testmod_names()))
3915
for mod in _test_suite_modules_to_doctest():
3916
if not interesting_module(mod):
3917
# No tests to keep here, move along
3920
# note that this really does mean "report only" -- doctest
3921
# still runs the rest of the examples
3922
doc_suite = doctest.DocTestSuite(mod,
3923
optionflags=doctest.REPORT_ONLY_FIRST_FAILURE)
3924
except ValueError, e:
3925
print '**failed to get doctest for: %s\n%s' % (mod, e)
3927
if len(doc_suite._tests) == 0:
3928
raise errors.BzrError("no doctests found in %s" % (mod,))
3929
suite.addTest(doc_suite)
3931
default_encoding = sys.getdefaultencoding()
3932
for name, plugin in bzrlib.plugin.plugins().items():
3933
if not interesting_module(plugin.module.__name__):
3935
plugin_suite = plugin.test_suite()
3936
# We used to catch ImportError here and turn it into just a warning,
3937
# but really if you don't have --no-plugins this should be a failure.
3938
# mbp 20080213 - see http://bugs.launchpad.net/bugs/189771
3939
if plugin_suite is None:
3940
plugin_suite = plugin.load_plugin_tests(loader)
3941
if plugin_suite is not None:
3942
suite.addTest(plugin_suite)
3943
if default_encoding != sys.getdefaultencoding():
3944
bzrlib.trace.warning(
3945
'Plugin "%s" tried to reset default encoding to: %s', name,
3946
sys.getdefaultencoding())
3948
sys.setdefaultencoding(default_encoding)
3950
if keep_only is not None:
3951
# Now that the referred modules have loaded their tests, keep only the
3953
suite = filter_suite_by_id_list(suite, id_filter)
3954
# Do some sanity checks on the id_list filtering
3955
not_found, duplicates = suite_matches_id_list(suite, keep_only)
3957
# The tester has used both keep_only and starting_with, so he is
3958
# already aware that some tests are excluded from the list, there
3959
# is no need to tell him which.
3962
# Some tests mentioned in the list are not in the test suite. The
3963
# list may be out of date, report to the tester.
3964
for id in not_found:
3965
bzrlib.trace.warning('"%s" not found in the test suite', id)
3966
for id in duplicates:
3967
bzrlib.trace.warning('"%s" is used as an id by several tests', id)
3972
def multiply_scenarios(scenarios_left, scenarios_right):
3973
"""Multiply two sets of scenarios.
3975
:returns: the cartesian product of the two sets of scenarios, that is
3976
a scenario for every possible combination of a left scenario and a
3980
('%s,%s' % (left_name, right_name),
3981
dict(left_dict.items() + right_dict.items()))
3982
for left_name, left_dict in scenarios_left
3983
for right_name, right_dict in scenarios_right]
3986
def multiply_tests(tests, scenarios, result):
3987
"""Multiply tests_list by scenarios into result.
3989
This is the core workhorse for test parameterisation.
3991
Typically the load_tests() method for a per-implementation test suite will
3992
call multiply_tests and return the result.
3994
:param tests: The tests to parameterise.
3995
:param scenarios: The scenarios to apply: pairs of (scenario_name,
3996
scenario_param_dict).
3997
:param result: A TestSuite to add created tests to.
3999
This returns the passed in result TestSuite with the cross product of all
4000
the tests repeated once for each scenario. Each test is adapted by adding
4001
the scenario name at the end of its id(), and updating the test object's
4002
__dict__ with the scenario_param_dict.
4004
>>> import bzrlib.tests.test_sampler
4005
>>> r = multiply_tests(
4006
... bzrlib.tests.test_sampler.DemoTest('test_nothing'),
4007
... [('one', dict(param=1)),
4008
... ('two', dict(param=2))],
4009
... TestUtil.TestSuite())
4010
>>> tests = list(iter_suite_tests(r))
4014
'bzrlib.tests.test_sampler.DemoTest.test_nothing(one)'
4020
for test in iter_suite_tests(tests):
4021
apply_scenarios(test, scenarios, result)
4025
def apply_scenarios(test, scenarios, result):
4026
"""Apply the scenarios in scenarios to test and add to result.
4028
:param test: The test to apply scenarios to.
4029
:param scenarios: An iterable of scenarios to apply to test.
4031
:seealso: apply_scenario
4033
for scenario in scenarios:
4034
result.addTest(apply_scenario(test, scenario))
4038
def apply_scenario(test, scenario):
4039
"""Copy test and apply scenario to it.
4041
:param test: A test to adapt.
4042
:param scenario: A tuple describing the scenarion.
4043
The first element of the tuple is the new test id.
4044
The second element is a dict containing attributes to set on the
4046
:return: The adapted test.
4048
new_id = "%s(%s)" % (test.id(), scenario[0])
4049
new_test = clone_test(test, new_id)
4050
for name, value in scenario[1].items():
4051
setattr(new_test, name, value)
4055
def clone_test(test, new_id):
4056
"""Clone a test giving it a new id.
4058
:param test: The test to clone.
4059
:param new_id: The id to assign to it.
4060
:return: The new test.
4062
new_test = copy.copy(test)
4063
new_test.id = lambda: new_id
4064
# XXX: Workaround <https://bugs.launchpad.net/testtools/+bug/637725>, which
4065
# causes cloned tests to share the 'details' dict. This makes it hard to
4066
# read the test output for parameterized tests, because tracebacks will be
4067
# associated with irrelevant tests.
4069
details = new_test._TestCase__details
4070
except AttributeError:
4071
# must be a different version of testtools than expected. Do nothing.
4074
# Reset the '__details' dict.
4075
new_test._TestCase__details = {}
4079
def permute_tests_for_extension(standard_tests, loader, py_module_name,
4081
"""Helper for permutating tests against an extension module.
4083
This is meant to be used inside a modules 'load_tests()' function. It will
4084
create 2 scenarios, and cause all tests in the 'standard_tests' to be run
4085
against both implementations. Setting 'test.module' to the appropriate
4086
module. See bzrlib.tests.test__chk_map.load_tests as an example.
4088
:param standard_tests: A test suite to permute
4089
:param loader: A TestLoader
4090
:param py_module_name: The python path to a python module that can always
4091
be loaded, and will be considered the 'python' implementation. (eg
4092
'bzrlib._chk_map_py')
4093
:param ext_module_name: The python path to an extension module. If the
4094
module cannot be loaded, a single test will be added, which notes that
4095
the module is not available. If it can be loaded, all standard_tests
4096
will be run against that module.
4097
:return: (suite, feature) suite is a test-suite that has all the permuted
4098
tests. feature is the Feature object that can be used to determine if
4099
the module is available.
4102
py_module = __import__(py_module_name, {}, {}, ['NO_SUCH_ATTRIB'])
4104
('python', {'module': py_module}),
4106
suite = loader.suiteClass()
4107
feature = ModuleAvailableFeature(ext_module_name)
4108
if feature.available():
4109
scenarios.append(('C', {'module': feature.module}))
4111
# the compiled module isn't available, so we add a failing test
4112
class FailWithoutFeature(TestCase):
4113
def test_fail(self):
4114
self.requireFeature(feature)
4115
suite.addTest(loader.loadTestsFromTestCase(FailWithoutFeature))
4116
result = multiply_tests(standard_tests, scenarios, suite)
4117
return result, feature
4120
def _rmtree_temp_dir(dirname, test_id=None):
4121
# If LANG=C we probably have created some bogus paths
4122
# which rmtree(unicode) will fail to delete
4123
# so make sure we are using rmtree(str) to delete everything
4124
# except on win32, where rmtree(str) will fail
4125
# since it doesn't have the property of byte-stream paths
4126
# (they are either ascii or mbcs)
4127
if sys.platform == 'win32':
4128
# make sure we are using the unicode win32 api
4129
dirname = unicode(dirname)
4131
dirname = dirname.encode(sys.getfilesystemencoding())
4133
osutils.rmtree(dirname)
4135
# We don't want to fail here because some useful display will be lost
4136
# otherwise. Polluting the tmp dir is bad, but not giving all the
4137
# possible info to the test runner is even worse.
4139
ui.ui_factory.clear_term()
4140
sys.stderr.write('\nWhile running: %s\n' % (test_id,))
4141
# Ugly, but the last thing we want here is fail, so bear with it.
4142
printable_e = str(e).decode(osutils.get_user_encoding(), 'replace'
4143
).encode('ascii', 'replace')
4144
sys.stderr.write('Unable to remove testing dir %s\n%s'
4145
% (os.path.basename(dirname), printable_e))
4148
class Feature(object):
4149
"""An operating system Feature."""
4152
self._available = None
4154
def available(self):
4155
"""Is the feature available?
4157
:return: True if the feature is available.
4159
if self._available is None:
4160
self._available = self._probe()
4161
return self._available
4164
"""Implement this method in concrete features.
4166
:return: True if the feature is available.
4168
raise NotImplementedError
4171
if getattr(self, 'feature_name', None):
4172
return self.feature_name()
4173
return self.__class__.__name__
4176
class _SymlinkFeature(Feature):
4179
return osutils.has_symlinks()
4181
def feature_name(self):
4184
SymlinkFeature = _SymlinkFeature()
4187
class _HardlinkFeature(Feature):
4190
return osutils.has_hardlinks()
4192
def feature_name(self):
4195
HardlinkFeature = _HardlinkFeature()
4198
class _OsFifoFeature(Feature):
4201
return getattr(os, 'mkfifo', None)
4203
def feature_name(self):
4204
return 'filesystem fifos'
4206
OsFifoFeature = _OsFifoFeature()
4209
class _UnicodeFilenameFeature(Feature):
4210
"""Does the filesystem support Unicode filenames?"""
4214
# Check for character combinations unlikely to be covered by any
4215
# single non-unicode encoding. We use the characters
4216
# - greek small letter alpha (U+03B1) and
4217
# - braille pattern dots-123456 (U+283F).
4218
os.stat(u'\u03b1\u283f')
4219
except UnicodeEncodeError:
4221
except (IOError, OSError):
4222
# The filesystem allows the Unicode filename but the file doesn't
4226
# The filesystem allows the Unicode filename and the file exists,
4230
UnicodeFilenameFeature = _UnicodeFilenameFeature()
4233
class _CompatabilityThunkFeature(Feature):
4234
"""This feature is just a thunk to another feature.
4236
It issues a deprecation warning if it is accessed, to let you know that you
4237
should really use a different feature.
4240
def __init__(self, dep_version, module, name,
4241
replacement_name, replacement_module=None):
4242
super(_CompatabilityThunkFeature, self).__init__()
4243
self._module = module
4244
if replacement_module is None:
4245
replacement_module = module
4246
self._replacement_module = replacement_module
4248
self._replacement_name = replacement_name
4249
self._dep_version = dep_version
4250
self._feature = None
4253
if self._feature is None:
4254
depr_msg = self._dep_version % ('%s.%s'
4255
% (self._module, self._name))
4256
use_msg = ' Use %s.%s instead.' % (self._replacement_module,
4257
self._replacement_name)
4258
symbol_versioning.warn(depr_msg + use_msg, DeprecationWarning)
4259
# Import the new feature and use it as a replacement for the
4261
mod = __import__(self._replacement_module, {}, {},
4262
[self._replacement_name])
4263
self._feature = getattr(mod, self._replacement_name)
4267
return self._feature._probe()
4270
class ModuleAvailableFeature(Feature):
4271
"""This is a feature than describes a module we want to be available.
4273
Declare the name of the module in __init__(), and then after probing, the
4274
module will be available as 'self.module'.
4276
:ivar module: The module if it is available, else None.
4279
def __init__(self, module_name):
4280
super(ModuleAvailableFeature, self).__init__()
4281
self.module_name = module_name
4285
self._module = __import__(self.module_name, {}, {}, [''])
4292
if self.available(): # Make sure the probe has been done
4296
def feature_name(self):
4297
return self.module_name
4300
# This is kept here for compatibility, it is recommended to use
4301
# 'bzrlib.tests.feature.paramiko' instead
4302
ParamikoFeature = _CompatabilityThunkFeature(
4303
deprecated_in((2,1,0)),
4304
'bzrlib.tests.features', 'ParamikoFeature', 'paramiko')
4307
def probe_unicode_in_user_encoding():
4308
"""Try to encode several unicode strings to use in unicode-aware tests.
4309
Return first successfull match.
4311
:return: (unicode value, encoded plain string value) or (None, None)
4313
possible_vals = [u'm\xb5', u'\xe1', u'\u0410']
4314
for uni_val in possible_vals:
4316
str_val = uni_val.encode(osutils.get_user_encoding())
4317
except UnicodeEncodeError:
4318
# Try a different character
4321
return uni_val, str_val
4325
def probe_bad_non_ascii(encoding):
4326
"""Try to find [bad] character with code [128..255]
4327
that cannot be decoded to unicode in some encoding.
4328
Return None if all non-ascii characters is valid
4331
for i in xrange(128, 256):
4334
char.decode(encoding)
4335
except UnicodeDecodeError:
4340
class _HTTPSServerFeature(Feature):
4341
"""Some tests want an https Server, check if one is available.
4343
Right now, the only way this is available is under python2.6 which provides
4354
def feature_name(self):
4355
return 'HTTPSServer'
4358
HTTPSServerFeature = _HTTPSServerFeature()
4361
class _UnicodeFilename(Feature):
4362
"""Does the filesystem support Unicode filenames?"""
4367
except UnicodeEncodeError:
4369
except (IOError, OSError):
4370
# The filesystem allows the Unicode filename but the file doesn't
4374
# The filesystem allows the Unicode filename and the file exists,
4378
UnicodeFilename = _UnicodeFilename()
4381
class _ByteStringNamedFilesystem(Feature):
4382
"""Is the filesystem based on bytes?"""
4385
if os.name == "posix":
4389
ByteStringNamedFilesystem = _ByteStringNamedFilesystem()
4392
class _UTF8Filesystem(Feature):
4393
"""Is the filesystem UTF-8?"""
4396
if osutils._fs_enc.upper() in ('UTF-8', 'UTF8'):
4400
UTF8Filesystem = _UTF8Filesystem()
4403
class _BreakinFeature(Feature):
4404
"""Does this platform support the breakin feature?"""
4407
from bzrlib import breakin
4408
if breakin.determine_signal() is None:
4410
if sys.platform == 'win32':
4411
# Windows doesn't have os.kill, and we catch the SIGBREAK signal.
4412
# We trigger SIGBREAK via a Console api so we need ctypes to
4413
# access the function
4420
def feature_name(self):
4421
return "SIGQUIT or SIGBREAK w/ctypes on win32"
4424
BreakinFeature = _BreakinFeature()
4427
class _CaseInsCasePresFilenameFeature(Feature):
4428
"""Is the file-system case insensitive, but case-preserving?"""
4431
fileno, name = tempfile.mkstemp(prefix='MixedCase')
4433
# first check truly case-preserving for created files, then check
4434
# case insensitive when opening existing files.
4435
name = osutils.normpath(name)
4436
base, rel = osutils.split(name)
4437
found_rel = osutils.canonical_relpath(base, name)
4438
return (found_rel == rel
4439
and os.path.isfile(name.upper())
4440
and os.path.isfile(name.lower()))
4445
def feature_name(self):
4446
return "case-insensitive case-preserving filesystem"
4448
CaseInsCasePresFilenameFeature = _CaseInsCasePresFilenameFeature()
4451
class _CaseInsensitiveFilesystemFeature(Feature):
4452
"""Check if underlying filesystem is case-insensitive but *not* case
4455
# Note that on Windows, Cygwin, MacOS etc, the file-systems are far
4456
# more likely to be case preserving, so this case is rare.
4459
if CaseInsCasePresFilenameFeature.available():
4462
if TestCaseWithMemoryTransport.TEST_ROOT is None:
4463
root = osutils.mkdtemp(prefix='testbzr-', suffix='.tmp')
4464
TestCaseWithMemoryTransport.TEST_ROOT = root
4466
root = TestCaseWithMemoryTransport.TEST_ROOT
4467
tdir = osutils.mkdtemp(prefix='case-sensitive-probe-', suffix='',
4469
name_a = osutils.pathjoin(tdir, 'a')
4470
name_A = osutils.pathjoin(tdir, 'A')
4472
result = osutils.isdir(name_A)
4473
_rmtree_temp_dir(tdir)
4476
def feature_name(self):
4477
return 'case-insensitive filesystem'
4479
CaseInsensitiveFilesystemFeature = _CaseInsensitiveFilesystemFeature()
4482
class _CaseSensitiveFilesystemFeature(Feature):
4485
if CaseInsCasePresFilenameFeature.available():
4487
elif CaseInsensitiveFilesystemFeature.available():
4492
def feature_name(self):
4493
return 'case-sensitive filesystem'
4495
# new coding style is for feature instances to be lowercase
4496
case_sensitive_filesystem_feature = _CaseSensitiveFilesystemFeature()
4499
# Kept for compatibility, use bzrlib.tests.features.subunit instead
4500
SubUnitFeature = _CompatabilityThunkFeature(
4501
deprecated_in((2,1,0)),
4502
'bzrlib.tests.features', 'SubUnitFeature', 'subunit')
4503
# Only define SubUnitBzrRunner if subunit is available.
4505
from subunit import TestProtocolClient
4506
from subunit.test_results import AutoTimingTestResultDecorator
4507
class SubUnitBzrProtocolClient(TestProtocolClient):
4509
def addSuccess(self, test, details=None):
4510
# The subunit client always includes the details in the subunit
4511
# stream, but we don't want to include it in ours.
4512
if details is not None and 'log' in details:
4514
return super(SubUnitBzrProtocolClient, self).addSuccess(
4517
class SubUnitBzrRunner(TextTestRunner):
4518
def run(self, test):
4519
result = AutoTimingTestResultDecorator(
4520
SubUnitBzrProtocolClient(self.stream))
4526
class _PosixPermissionsFeature(Feature):
4530
# create temporary file and check if specified perms are maintained.
4533
write_perms = stat.S_IRUSR | stat.S_IWUSR | stat.S_IXUSR
4534
f = tempfile.mkstemp(prefix='bzr_perms_chk_')
4537
os.chmod(name, write_perms)
4539
read_perms = os.stat(name).st_mode & 0777
4541
return (write_perms == read_perms)
4543
return (os.name == 'posix') and has_perms()
4545
def feature_name(self):
4546
return 'POSIX permissions support'
4548
posix_permissions_feature = _PosixPermissionsFeature()
added
removed
bzrlib/tests/__init__.py
