class TestSuite:
"""Each test suite contains a number of related tests files,
located in the same directory on disk. Each test file has
extention .test and contains a listing of server commands,
followed by their output. The commands are executed, and
obtained results are compared with pre-recorded output. In case
of a comparision difference, an exception is raised. A test suite
must also contain suite.ini, which describes how to start the
server for this suite, the client program to execute individual
tests and other suite properties. The server is started once per
suite."""
def get_multirun_conf(self, suite_path):
conf_name = self.ini.get('config', None)
if conf_name is None:
return None
path = os.path.join(suite_path, conf_name)
result = None
with open(path) as cfg:
try:
result = json.load(cfg)
except ValueError:
raise RuntimeError('Ivalid multirun json')
return result
def get_multirun_params(self, test_path):
test = test_path.split('/')[-1]
if self.multi_run is None:
return
result = self.multi_run.get(test, None)
if result is not None:
return result
result = self.multi_run.get('*', None)
return result
def __init__(self, suite_path, args):
"""Initialize a test suite: check that it exists and contains
a syntactically correct configuration file. Then create
a test instance for each found test."""
self.args = args
self.tests = []
self.ini = {}
self.suite_path = suite_path
self.ini["core"] = "tarantool"
if os.access(suite_path, os.F_OK) == False:
raise RuntimeError("Suite %s doesn't exist" % repr(suite_path))
# read the suite config
config = ConfigParser.ConfigParser()
config.read(os.path.join(suite_path, "suite.ini"))
self.ini.update(dict(config.items("default")))
self.ini.update(self.args.__dict__)
self.multi_run = self.get_multirun_conf(suite_path)
if self.args.stress is None and self.ini['core'] == 'stress':
return
for i in ["script"]:
self.ini[i] = os.path.join(suite_path,
self.ini[i]) if i in self.ini else None
for i in ["disabled", "valgrind_disabled", "release_disabled"]:
self.ini[i] = dict.fromkeys(
self.ini[i].split()) if i in self.ini else dict()
for i in ["lua_libs"]:
self.ini[i] = map(
lambda x: os.path.join(suite_path, x),
dict.fromkeys(self.ini[i].split())
if i in self.ini else dict())
try:
if self.ini['core'] in ['tarantool', 'stress']:
self.server = TarantoolServer(self.ini)
else:
self.server = Server(self.ini)
self.ini["server"] = self.server
except Exception as e:
print e
raise RuntimeError("Unknown server: core = {0}".format(
self.ini["core"]))
color_stdout("Collecting tests in ", schema='ts_text')
color_stdout(repr(suite_path), schema='path')
color_stdout(": ", self.ini["description"], ".\n", schema='ts_text')
self.server.find_tests(self, suite_path)
color_stdout("Found ",
str(len(self.tests)),
" tests.\n",
schema='path')
def run_all(self):
"""For each file in the test suite, run client program
assuming each file represents an individual test."""
if not self.tests:
# noting to test, exit
return []
# fixme: remove this string if we fix all legacy tests
self.server.cls = self.tests[0].__class__
# create inspectpor daemon for cluster tests
inspector = TarantoolInspector('localhost', self.server.inspector_port)
inspector.start()
self.server.deploy(silent=False)
longsep = '=' * 80
shortsep = '-' * 75
color_stdout(longsep, "\n", schema='separator')
color_stdout("TEST".ljust(48), schema='t_name')
color_stdout("PARAMS\t\t", schema='test_var')
color_stdout("RESULT\n", schema='test_pass')
color_stdout(shortsep, "\n", schema='separator')
failed_tests = []
try:
for test in self.tests:
test.inspector = inspector
color_stdout(os.path.join(self.ini['suite'],
os.path.basename(
test.name)).ljust(48),
schema='t_name')
# for better diagnostics in case of a long-running test
conf = ''
if test.run_params:
conf = test.conf_name
color_stdout("%s" % conf.ljust(16), schema='test_var')
test_name = os.path.basename(test.name)
if (test_name in self.ini["disabled"] or not self.server.debug
and test_name in self.ini["release_disabled"]
or self.args.valgrind
and test_name in self.ini["valgrind_disabled"]
or not self.args.long
and test_name in self.ini.get("long_run", [])):
color_stdout("[ disabled ]\n", schema='t_name')
else:
test.run(self.server)
if not test.passed():
failed_tests.append(test.name)
color_stdout(shortsep, "\n", schema='separator')
self.server.stop(silent=False)
# don't delete core files or state of the data dir
# in case of exception, which is raised when the
# server crashes
inspector.stop()
self.server.cleanup()
except KeyboardInterrupt:
color_stdout("\n%s\n" % shortsep, schema='separator')
self.server.stop(silent=False)
raise
if failed_tests:
color_stdout("Failed {0} tests: {1}.\n".format(
len(failed_tests), ", ".join(failed_tests)),
schema='error')
if self.args.valgrind and check_valgrind_log(self.server.valgrind_log):
color_stdout(shortsep, "\n", schema='separator')
color_stdout(
" Error! There were warnings/errors in valgrind log file:\n",
schema='error')
print_tail_n(self.server.valgrind_log, 20)
color_stdout(shortsep, "\n", schema='separator')
return ['valgrind error in ' + self.suite_path]
return failed_tests
class TestSuite:
"""Each test suite contains a number of related tests files,
located in the same directory on disk. Each test file has
extention .test and contains a listing of server commands,
followed by their output. The commands are executed, and
obtained results are compared with pre-recorded output. In case
of a comparision difference, an exception is raised. A test suite
must also contain suite.ini, which describes how to start the
server for this suite, the client program to execute individual
tests and other suite properties. The server is started once per
suite."""
def __init__(self, suite_path, args):
"""Initialize a test suite: check that it exists and contains
a syntactically correct configuration file. Then create
a test instance for each found test."""
self.args = args
self.tests = []
self.ini = {}
self.suite_path = suite_path
self.ini["core"] = "tarantool"
if os.access(suite_path, os.F_OK) == False:
raise RuntimeError("Suite \"" + suite_path + \
"\" doesn't exist")
# read the suite config
config = ConfigParser.ConfigParser()
config.read(os.path.join(suite_path, "suite.ini"))
self.ini.update(dict(config.items("default")))
for i in ["config", "init_lua"]:
self.ini[i] = os.path.join(suite_path,
self.ini[i]) if i in self.ini else None
for i in ["disabled", "valgrind_disabled", "release_disabled"]:
self.ini[i] = dict.fromkeys(
self.ini[i].split()) if i in self.ini else dict()
for i in ["lua_libs"]:
self.ini[i] = map(
lambda x: os.path.join(suite_path, x),
dict.fromkeys(self.ini[i].split())
if i in self.ini else dict())
try:
self.server = Server(self.ini["core"])
self.ini["server"] = self.server
except Exception as e:
print e
raise RuntimeError("Unknown server: core = {0}".format(
self.ini["core"]))
print "Collecting tests in \"" + suite_path + "\": " +\
self.ini["description"] + "."
self.server.find_tests(self, suite_path)
print "Found " + str(len(self.tests)) + " tests."
def run_all(self):
"""For each file in the test suite, run client program
assuming each file represents an individual test."""
if not self.tests:
# noting to test, exit
return []
self.server.deploy(self.ini["config"],
self.server.find_exe(self.args.builddir,
silent=False),
self.args.vardir,
self.args.mem,
self.args.start_and_exit,
self.args.gdb,
self.args.valgrind,
init_lua=self.ini["init_lua"],
silent=False)
if self.ini['core'] != 'unittest':
self.ini['servers'] = {'default': self.server}
self.ini['connections'] = {'default': self.server.admin}
self.ini['vardir'] = self.args.vardir
self.ini['builddir'] = self.args.builddir
for i in self.ini['lua_libs']:
shutil.copy(i, self.args.vardir)
if self.args.start_and_exit:
print " Start and exit requested, exiting..."
exit(0)
longsep = '=' * 70
shortsep = '-' * 60
print longsep
print "TEST".ljust(48), "RESULT"
print shortsep
failed_tests = []
try:
for test in self.tests:
sys.stdout.write(test.name.ljust(48))
# for better diagnostics in case of a long-running test
sys.stdout.flush()
test_name = os.path.basename(test.name)
if (test_name in self.ini["disabled"] or not self.server.debug
and test_name in self.ini["release_disabled"]
or self.args.valgrind
and test_name in self.ini["valgrind_disabled"]):
print "[ disabled ]"
else:
test.run(self.server)
if not test.passed():
failed_tests.append(test.name)
except (KeyboardInterrupt) as e:
print '\n',
raise
finally:
print shortsep
self.server.stop(silent=False)
self.server.cleanup()
if failed_tests:
print "Failed {0} tests: {1}.".format(len(failed_tests),
", ".join(failed_tests))
if self.args.valgrind and check_valgrind_log(self.server.valgrind_log):
print " Error! There were warnings/errors in valgrind log file:"
print_tail_n(self.server.valgrind_log, 20)
return ['valgrind error in ' + self.suite_path]
return failed_tests
class TestSuite:
"""Each test suite contains a number of related tests files,
located in the same directory on disk. Each test file has
extention .test and contains a listing of server commands,
followed by their output. The commands are executed, and
obtained results are compared with pre-recorded output. In case
of a comparision difference, an exception is raised. A test suite
must also contain suite.ini, which describes how to start the
server for this suite, the client program to execute individual
tests and other suite properties. The server is started once per
suite."""
def __init__(self, suite_path, args):
"""Initialize a test suite: check that it exists and contains
a syntactically correct configuration file. Then create
a test instance for each found test."""
self.args = args
self.tests = []
self.ini = {}
self.suite_path = suite_path
self.ini["core"] = "tarantool"
if os.access(suite_path, os.F_OK) == False:
raise RuntimeError("Suite \"" + suite_path + \
"\" doesn't exist")
# read the suite config
config = ConfigParser.ConfigParser()
config.read(os.path.join(suite_path, "suite.ini"))
self.ini.update(dict(config.items("default")))
for i in ["config", "init_lua"]:
self.ini[i] = os.path.join(suite_path, self.ini[i]) if i in self.ini else None
for i in ["disabled", "valgrind_disabled", "release_disabled"]:
self.ini[i] = dict.fromkeys(self.ini[i].split()) if i in self.ini else dict()
for i in ["lua_libs"]:
self.ini[i] = map(lambda x: os.path.join(suite_path, x),
dict.fromkeys(self.ini[i].split()) if i in self.ini else
dict())
try:
self.server = Server(self.ini["core"])
self.ini["server"] = self.server
except Exception as e:
print e
raise RuntimeError("Unknown server: core = {0}".format(
self.ini["core"]))
print "Collecting tests in \"" + suite_path + "\": " +\
self.ini["description"] + "."
self.server.find_tests(self, suite_path)
print "Found " + str(len(self.tests)) + " tests."
def run_all(self):
"""For each file in the test suite, run client program
assuming each file represents an individual test."""
if not self.tests:
# noting to test, exit
return []
self.server.deploy(self.ini["config"],
self.server.find_exe(self.args.builddir, silent=False),
self.args.vardir, self.args.mem, self.args.start_and_exit,
self.args.gdb, self.args.valgrind,
init_lua=self.ini["init_lua"], silent=False)
if self.ini['core'] != 'unittest':
self.ini['servers'] = {'default' : self.server}
self.ini['connections'] = {'default' : self.server.admin}
self.ini['vardir'] = self.args.vardir
self.ini['builddir'] = self.args.builddir
for i in self.ini['lua_libs']:
shutil.copy(i, self.args.vardir)
if self.args.start_and_exit:
print " Start and exit requested, exiting..."
exit(0)
longsep = '='*70
shortsep = '-'*60
print longsep
print "TEST".ljust(48), "RESULT"
print shortsep
failed_tests = []
try:
for test in self.tests:
sys.stdout.write(test.name.ljust(48))
# for better diagnostics in case of a long-running test
sys.stdout.flush()
test_name = os.path.basename(test.name)
if (test_name in self.ini["disabled"]
or not self.server.debug and test_name in self.ini["release_disabled"]
or self.args.valgrind and test_name in self.ini["valgrind_disabled"]):
print "[ disabled ]"
else:
test.run(self.server)
if not test.passed():
failed_tests.append(test.name)
except (KeyboardInterrupt) as e:
print '\n',
raise
finally:
print shortsep
self.server.stop(silent=False)
self.server.cleanup()
if failed_tests:
print "Failed {0} tests: {1}.".format(len(failed_tests),
", ".join(failed_tests))
if self.args.valgrind and check_valgrind_log(self.server.valgrind_log):
print " Error! There were warnings/errors in valgrind log file:"
print_tail_n(self.server.valgrind_log, 20)
return ['valgrind error in ' + self.suite_path]
return failed_tests
class TestSuite:
"""Each test suite contains a number of related tests files,
located in the same directory on disk. Each test file has
extention .test and contains a listing of server commands,
followed by their output. The commands are executed, and
obtained results are compared with pre-recorded output. In case
of a comparision difference, an exception is raised. A test suite
must also contain suite.ini, which describes how to start the
server for this suite, the client program to execute individual
tests and other suite properties. The server is started once per
suite."""
def __init__(self, suite_path, args):
"""Initialize a test suite: check that it exists and contains
a syntactically correct configuration file. Then create
a test instance for each found test."""
self.args = args
self.tests = []
self.ini = {}
self.suite_path = suite_path
self.ini["core"] = "tarantool"
if os.access(suite_path, os.F_OK) == False:
raise RuntimeError("Suite %s doesn't exist" % repr(suite_path))
# read the suite config
config = ConfigParser.ConfigParser()
config.read(os.path.join(suite_path, "suite.ini"))
self.ini.update(dict(config.items("default")))
self.ini.update(self.args.__dict__)
if self.args.stress is None and self.ini['core'] == 'stress':
return
for i in ["script"]:
self.ini[i] = os.path.join(suite_path, self.ini[i]) if i in self.ini else None
for i in ["disabled", "valgrind_disabled", "release_disabled"]:
self.ini[i] = dict.fromkeys(self.ini[i].split()) if i in self.ini else dict()
for i in ["lua_libs"]:
self.ini[i] = map(lambda x: os.path.join(suite_path, x),
dict.fromkeys(self.ini[i].split()) if i in self.ini else dict())
try:
if self.ini['core'] in ['tarantool', 'stress']:
self.server = TarantoolServer(self.ini)
else:
self.server = Server(self.ini)
self.ini["server"] = self.server
except Exception as e:
print e
raise RuntimeError("Unknown server: core = {0}".format(
self.ini["core"]))
color_stdout("Collecting tests in ", schema='ts_text')
color_stdout(repr(suite_path), schema='path')
color_stdout(": ", self.ini["description"], ".\n", schema='ts_text')
self.server.find_tests(self, suite_path)
color_stdout("Found ", str(len(self.tests)), " tests.\n", schema='path')
def run_all(self):
"""For each file in the test suite, run client program
assuming each file represents an individual test."""
if not self.tests:
# noting to test, exit
return []
self.server.deploy(silent=False)
if self.args.start_and_exit:
color_stdout(" Start and exit requested, exiting...\n", schema='info')
exit(0)
longsep = '='*70
shortsep = '-'*60
color_stdout(longsep, "\n", schema='separator')
color_stdout("TEST".ljust(48), schema='t_name')
color_stdout("RESULT\n", schema='test_pass')
color_stdout(shortsep, "\n", schema='separator')
failed_tests = []
try:
for test in self.tests:
color_stdout(test.name.ljust(48), schema='t_name')
# for better diagnostics in case of a long-running test
test_name = os.path.basename(test.name)
if (test_name in self.ini["disabled"]
or not self.server.debug and test_name in self.ini["release_disabled"]
or self.args.valgrind and test_name in self.ini["valgrind_disabled"]):
color_stdout("[ disabled ]\n", schema='t_name')
else:
test.run(self.server)
if not test.passed():
failed_tests.append(test.name)
color_stdout(shortsep, "\n", schema='separator')
self.server.stop(silent=False)
except (KeyboardInterrupt) as e:
color_stdout("\n%s\n" % shortsep, schema='separator')
self.server.stop(silent=False)
raise
finally:
self.server.cleanup()
if failed_tests:
color_stdout("Failed {0} tests: {1}.\n".format(len(failed_tests),
", ".join(failed_tests)),
schema='error')
if self.args.valgrind and check_valgrind_log(self.server.valgrind_log):
color_stdout(shortsep, "\n", schema='separator')
color_stdout(" Error! There were warnings/errors in valgrind log file:\n", schema='error')
print_tail_n(self.server.valgrind_log, 20)
color_stdout(shortsep, "\n", schema='separator')
return ['valgrind error in ' + self.suite_path]
return failed_tests
class TestSuite:
"""Each test suite contains a number of related tests files,
located in the same directory on disk. Each test file has
extention .test and contains a listing of server commands,
followed by their output. The commands are executed, and
obtained results are compared with pre-recorded output. In case
of a comparision difference, an exception is raised. A test suite
must also contain suite.ini, which describes how to start the
server for this suite, the client program to execute individual
tests and other suite properties. The server is started once per
suite."""
def __init__(self, suite_path, args):
"""Initialize a test suite: check that it exists and contains
a syntactically correct configuration file. Then create
a test instance for each found test."""
self.args = args
self.tests = []
self.ini = {}
self.suite_path = suite_path
self.ini["core"] = "tarantool"
if os.access(suite_path, os.F_OK) == False:
raise RuntimeError("Suite %s doesn't exist" % repr(suite_path))
# read the suite config
config = ConfigParser.ConfigParser()
config.read(os.path.join(suite_path, "suite.ini"))
self.ini.update(dict(config.items("default")))
self.ini.update(self.args.__dict__)
if self.args.stress is None and self.ini['core'] == 'stress':
return
for i in ["script"]:
self.ini[i] = os.path.join(suite_path,
self.ini[i]) if i in self.ini else None
for i in ["disabled", "valgrind_disabled", "release_disabled"]:
self.ini[i] = dict.fromkeys(
self.ini[i].split()) if i in self.ini else dict()
for i in ["lua_libs"]:
self.ini[i] = map(
lambda x: os.path.join(suite_path, x),
dict.fromkeys(self.ini[i].split())
if i in self.ini else dict())
try:
if self.ini['core'] in ['tarantool', 'stress']:
self.server = TarantoolServer(self.ini)
else:
self.server = Server(self.ini)
self.ini["server"] = self.server
except Exception as e:
print e
raise RuntimeError("Unknown server: core = {0}".format(
self.ini["core"]))
color_stdout("Collecting tests in ", schema='ts_text')
color_stdout(repr(suite_path), schema='path')
color_stdout(": ", self.ini["description"], ".\n", schema='ts_text')
self.server.find_tests(self, suite_path)
color_stdout("Found ",
str(len(self.tests)),
" tests.\n",
schema='path')
def run_all(self):
"""For each file in the test suite, run client program
assuming each file represents an individual test."""
if not self.tests:
# noting to test, exit
return []
self.server.deploy(silent=False)
if self.args.start_and_exit:
color_stdout(" Start and exit requested, exiting...\n",
schema='info')
exit(0)
longsep = '=' * 70
shortsep = '-' * 60
color_stdout(longsep, "\n", schema='separator')
color_stdout("TEST".ljust(48), schema='t_name')
color_stdout("RESULT\n", schema='test_pass')
color_stdout(shortsep, "\n", schema='separator')
failed_tests = []
try:
for test in self.tests:
color_stdout(test.name.ljust(48), schema='t_name')
# for better diagnostics in case of a long-running test
test_name = os.path.basename(test.name)
if (test_name in self.ini["disabled"] or not self.server.debug
and test_name in self.ini["release_disabled"]
or self.args.valgrind
and test_name in self.ini["valgrind_disabled"]):
color_stdout("[ disabled ]\n", schema='t_name')
else:
test.run(self.server)
if not test.passed():
failed_tests.append(test.name)
color_stdout(shortsep, "\n", schema='separator')
self.server.stop(silent=False)
except (KeyboardInterrupt) as e:
color_stdout("\n%s\n" % shortsep, schema='separator')
self.server.stop(silent=False)
raise
finally:
self.server.cleanup()
if failed_tests:
color_stdout("Failed {0} tests: {1}.\n".format(
len(failed_tests), ", ".join(failed_tests)),
schema='error')
if self.args.valgrind and check_valgrind_log(self.server.valgrind_log):
color_stdout(shortsep, "\n", schema='separator')
color_stdout(
" Error! There were warnings/errors in valgrind log file:\n",
schema='error')
print_tail_n(self.server.valgrind_log, 20)
color_stdout(shortsep, "\n", schema='separator')
return ['valgrind error in ' + self.suite_path]
return failed_tests
