[3.5] bpo-30523, bpo-30764, bpo-30776: Sync regrtest from master (#2442)

* bpo-30523: regrtest --list-cases --match (#2401)

* regrtest --list-cases now supports --match and --match-file options.
  Example: ./python -m test --list-cases -m FileTests test_os
* --list-cases now also sets support.verbose to False to prevent
  messages to stdout when loading test modules.
* Add support._match_test() private function.
(cherry picked from commit ace56d583664f855d89d1219ece7c21c2fddcf30)

* bpo-30764: regrtest: add --fail-env-changed option (#2402)

* bpo-30764: regrtest: change exit code on failure

* Exit code 2 if failed tests ("bad")
* Exit code 3 if interrupted

* bpo-30764: regrtest: add --fail-env-changed option

If the option is set, mark a test as failed if it alters the
environment, for example if it creates a file without removing it.

(cherry picked from commit 63f54c68936d648c70ca411661e4208329edcf26)

* bpo-30776: reduce regrtest -R false positives (#2422)

* Change the regrtest --huntrleaks checker to decide if a test file
  leaks or not. Require that each run leaks at least 1 reference.
* Warmup runs are now completely ignored: ignored in the checker test
  and not used anymore to compute the sum.
* Add an unit test for a reference leak.

Example of reference differences previously considered a failure
(leak) and now considered as success (success, no leak):

    [3, 0, 0]
    [0, 1, 0]
    [8, -8, 1]
(cherry picked from commit 48b5c422ffb03affb00c184b9a99e5537be92732)

3 files changed, 135 insertions, 31 deletions

diff --git a/Lib/test/regrtest.py b/Lib/test/regrtest.py
index 299416c..339beb1 100755
--- a/Lib/test/regrtest.py
+++ b/Lib/test/regrtest.py
@@ -343,6 +343,9 @@ def _create_parser():
                             ' , don\'t execute them')
     group.add_argument('-P', '--pgo', dest='pgo', action='store_true',
                        help='enable Profile Guided Optimization training')
+    group.add_argument('--fail-env-changed', action='store_true',
+                       help='if a test file alters the environment, mark '
+                            'the test as failed')
 
     return parser
 
@@ -944,11 +947,19 @@ def main(tests=None, **kwargs):
         result = "FAILURE"
     elif interrupted:
         result = "INTERRUPTED"
+    elif environment_changed and ns.fail_env_changed:
+        result = "ENV CHANGED"
     else:
         result = "SUCCESS"
     print("Tests result: %s" % result)
 
-    sys.exit(len(bad) > 0 or interrupted)
+    if bad:
+        sys.exit(2)
+    if interrupted:
+        sys.exit(130)
+    if ns.fail_env_changed and environment_changed:
+        sys.exit(3)
+    sys.exit(0)
 
 
 # small set of tests to determine if we have a basically functioning interpreter
@@ -1510,9 +1521,21 @@ def dash_R(the_module, test, indirect_test, huntrleaks):
             alloc_deltas[i] = alloc_after - alloc_before
         alloc_before, rc_before = alloc_after, rc_after
     print(file=sys.stderr)
+
     # These checkers return False on success, True on failure
     def check_rc_deltas(deltas):
-        return any(deltas)
+        # bpo-30776: Try to ignore false positives:
+        #
+        #   [3, 0, 0]
+        #   [0, 1, 0]
+        #   [8, -8, 1]
+        #
+        # Expected leaks:
+        #
+        #   [5, 5, 6]
+        #   [10, 1, 1]
+        return all(delta >= 1 for delta in deltas)
+
     def check_alloc_deltas(deltas):
         # At least 1/3rd of 0s
         if 3 * deltas.count(0) < len(deltas):
@@ -1524,10 +1547,13 @@ def dash_R(the_module, test, indirect_test, huntrleaks):
     failed = False
     for deltas, item_name, checker in [
         (rc_deltas, 'references', check_rc_deltas),
-        (alloc_deltas, 'memory blocks', check_alloc_deltas)]:
+        (alloc_deltas, 'memory blocks', check_alloc_deltas)
+    ]:
+        # ignore warmup runs
+        deltas = deltas[nwarmup:]
         if checker(deltas):
             msg = '%s leaked %s %s, sum=%s' % (
-                test, deltas[nwarmup:], item_name, sum(deltas))
+                test, deltas, item_name, sum(deltas))
             print(msg, file=sys.stderr)
             sys.stderr.flush()
             with open(fname, "a") as refrep:
@@ -1735,10 +1761,14 @@ def _list_cases(suite):
         if isinstance(test, unittest.TestSuite):
             _list_cases(test)
         elif isinstance(test, unittest.TestCase):
-            print(test.id())
+            if support._match_test(test):
+                print(test.id())
 
 
 def list_cases(ns, selected):
+    support.verbose = False
+    support.match_tests = ns.match_tests
+
     skipped = []
     for test in selected:
         abstest = get_abs_module(ns, test)
diff --git a/Lib/test/support/__init__.py b/Lib/test/support/__init__.py
index 92599b3..85878eb 100644
--- a/Lib/test/support/__init__.py
+++ b/Lib/test/support/__init__.py
@@ -1866,6 +1866,23 @@ def _run_suite(suite):
         raise TestFailed(err)
 
 
+def _match_test(test):
+    global match_tests
+
+    if match_tests is None:
+        return True
+    test_id = test.id()
+
+    for match_test in match_tests:
+        if fnmatch.fnmatchcase(test_id, match_test):
+            return True
+
+        for name in test_id.split("."):
+            if fnmatch.fnmatchcase(name, match_test):
+                return True
+    return False
+
+
 def run_unittest(*classes):
     """Run tests from unittest.TestCase-derived classes."""
     valid_types = (unittest.TestSuite, unittest.TestCase)
@@ -1880,20 +1897,7 @@ def run_unittest(*classes):
             suite.addTest(cls)
         else:
             suite.addTest(unittest.makeSuite(cls))
-    def case_pred(test):
-        if match_tests is None:
-            return True
-        test_id = test.id()
-
-        for match_test in match_tests:
-            if fnmatch.fnmatchcase(test_id, match_test):
-                return True
-
-            for name in test_id.split("."):
-                if fnmatch.fnmatchcase(name, match_test):
-                    return True
-        return False
-    _filter_suite(suite, case_pred)
+    _filter_suite(suite, _match_test)
     _run_suite(suite)
 
 #=======================================================================
diff --git a/Lib/test/test_regrtest.py b/Lib/test/test_regrtest.py
index 3568004..0eb3f08 100644
--- a/Lib/test/test_regrtest.py
+++ b/Lib/test/test_regrtest.py
@@ -377,19 +377,19 @@ class BaseTestCase(unittest.TestCase):
         return list(match.group(1) for match in parser)
 
     def check_executed_tests(self, output, tests, skipped=(), failed=(),
-                             omitted=(), randomize=False, interrupted=False):
+                             env_changed=(), omitted=(),
+                             randomize=False, interrupted=False,
+                             fail_env_changed=False):
         if isinstance(tests, str):
             tests = [tests]
         if isinstance(skipped, str):
             skipped = [skipped]
         if isinstance(failed, str):
             failed = [failed]
+        if isinstance(env_changed, str):
+            env_changed = [env_changed]
         if isinstance(omitted, str):
             omitted = [omitted]
-        ntest = len(tests)
-        nskipped = len(skipped)
-        nfailed = len(failed)
-        nomitted = len(omitted)
 
         executed = self.parse_executed_tests(output)
         if randomize:
@@ -415,11 +415,17 @@ class BaseTestCase(unittest.TestCase):
             regex = list_regex('%s test%s failed', failed)
             self.check_line(output, regex)
 
+        if env_changed:
+            regex = list_regex('%s test%s altered the execution environment',
+                               env_changed)
+            self.check_line(output, regex)
+
         if omitted:
             regex = list_regex('%s test%s omitted', omitted)
             self.check_line(output, regex)
 
-        good = ntest - nskipped - nfailed - nomitted
+        good = (len(tests) - len(skipped) - len(failed)
+                - len(omitted) - len(env_changed))
         if good:
             regex = r'%s test%s OK\.$' % (good, plural(good))
             if not skipped and not failed and good > 1:
@@ -429,10 +435,12 @@ class BaseTestCase(unittest.TestCase):
         if interrupted:
             self.check_line(output, 'Test suite interrupted by signal SIGINT.')
 
-        if nfailed:
+        if failed:
             result = 'FAILURE'
         elif interrupted:
             result = 'INTERRUPTED'
+        elif fail_env_changed and env_changed:
+            result = 'ENV CHANGED'
         else:
             result = 'SUCCESS'
         self.check_line(output, 'Tests result: %s' % result)
@@ -604,7 +612,7 @@ class ArgsTestCase(BaseTestCase):
         test_failing = self.create_test('failing', code=code)
         tests = [test_ok, test_failing]
 
-        output = self.run_tests(*tests, exitcode=1)
+        output = self.run_tests(*tests, exitcode=2)
         self.check_executed_tests(output, tests, failed=test_failing)
 
     def test_resources(self):
@@ -703,7 +711,7 @@ class ArgsTestCase(BaseTestCase):
     def test_interrupted(self):
         code = TEST_INTERRUPTED
         test = self.create_test('sigint', code=code)
-        output = self.run_tests(test, exitcode=1)
+        output = self.run_tests(test, exitcode=130)
         self.check_executed_tests(output, test, omitted=test,
                                   interrupted=True)
 
@@ -732,7 +740,7 @@ class ArgsTestCase(BaseTestCase):
                 args = ("--slowest", "-j2", test)
             else:
                 args = ("--slowest", test)
-            output = self.run_tests(*args, exitcode=1)
+            output = self.run_tests(*args, exitcode=130)
             self.check_executed_tests(output, test,
                                       omitted=test, interrupted=True)
 
@@ -772,9 +780,43 @@ class ArgsTestCase(BaseTestCase):
                         builtins.__dict__['RUN'] = 1
         """)
         test = self.create_test('forever', code=code)
-        output = self.run_tests('--forever', test, exitcode=1)
+        output = self.run_tests('--forever', test, exitcode=2)
         self.check_executed_tests(output, [test]*3, failed=test)
 
+    def check_leak(self, code, what):
+        test = self.create_test('huntrleaks', code=code)
+
+        filename = 'reflog.txt'
+        self.addCleanup(support.unlink, filename)
+        output = self.run_tests('--huntrleaks', '3:3:', test,
+                                exitcode=2,
+                                stderr=subprocess.STDOUT)
+        self.check_executed_tests(output, [test], failed=test)
+
+        line = 'beginning 6 repetitions\n123456\n......\n'
+        self.check_line(output, re.escape(line))
+
+        line2 = '%s leaked [1, 1, 1] %s, sum=3\n' % (test, what)
+        self.assertIn(line2, output)
+
+        with open(filename) as fp:
+            reflog = fp.read()
+            self.assertIn(line2, reflog)
+
+    @unittest.skipUnless(Py_DEBUG, 'need a debug build')
+    def test_huntrleaks(self):
+        # test --huntrleaks
+        code = textwrap.dedent("""
+            import unittest
+
+            GLOBAL_LIST = []
+
+            class RefLeakTest(unittest.TestCase):
+                def test_leak(self):
+                    GLOBAL_LIST.append(object())
+        """)
+        self.check_leak(code, 'references')
+
     def test_list_tests(self):
         # test --list-tests
         tests = [self.create_test() for i in range(5)]
@@ -794,11 +836,20 @@ class ArgsTestCase(BaseTestCase):
                     pass
         """)
         testname = self.create_test(code=code)
+
+        # Test --list-cases
         all_methods = ['%s.Tests.test_method1' % testname,
                        '%s.Tests.test_method2' % testname]
         output = self.run_tests('--list-cases', testname)
         self.assertEqual(output.splitlines(), all_methods)
 
+        # Test --list-cases with --match
+        all_methods = ['%s.Tests.test_method1' % testname]
+        output = self.run_tests('--list-cases',
+                                '-m', 'test_method1',
+                                testname)
+        self.assertEqual(output.splitlines(), all_methods)
+
     def test_crashed(self):
         # Any code which causes a crash
         code = 'import faulthandler; faulthandler._sigsegv()'
@@ -806,7 +857,7 @@ class ArgsTestCase(BaseTestCase):
         ok_test = self.create_test(name="ok")
 
         tests = [crash_test, ok_test]
-        output = self.run_tests("-j2", *tests, exitcode=1)
+        output = self.run_tests("-j2", *tests, exitcode=2)
         self.check_executed_tests(output, tests, failed=crash_test,
                                   randomize=True)
 
@@ -855,6 +906,25 @@ class ArgsTestCase(BaseTestCase):
         subset = ['test_method1', 'test_method3']
         self.assertEqual(methods, subset)
 
+    def test_env_changed(self):
+        code = textwrap.dedent("""
+            import unittest
+
+            class Tests(unittest.TestCase):
+                def test_env_changed(self):
+                    open("env_changed", "w").close()
+        """)
+        testname = self.create_test(code=code)
+
+        # don't fail by default
+        output = self.run_tests(testname)
+        self.check_executed_tests(output, [testname], env_changed=testname)
+
+        # fail with --fail-env-changed
+        output = self.run_tests("--fail-env-changed", testname, exitcode=3)
+        self.check_executed_tests(output, [testname], env_changed=testname,
+                                  fail_env_changed=True)
+
 
 if __name__ == '__main__':
     unittest.main()