1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
|
#!/usr/bin/env python
# Copyright (c) 2016 Ericsson AB and others.
#
# All rights reserved. This program and the accompanying materials
# are made available under the terms of the Apache License, Version 2.0
# which accompanies this distribution, and is available at
# http://www.apache.org/licenses/LICENSE-2.0
""" The entry of running tests:
1) Parses functest/ci/testcases.yaml to check which testcase(s) to be run
2) Execute the common operations on every testcase (run, push results to db...)
3) Return the right status code
"""
import argparse
import importlib
import logging
import logging.config
import os
import re
import sys
import textwrap
import pkg_resources
import enum
import prettytable
import yaml
import functest.ci.tier_builder as tb
import functest.core.testcase as testcase
LOGGER = logging.getLogger('functest.ci.run_tests')
ENV_FILE = "/home/opnfv/functest/conf/env_file"
class Result(enum.Enum):
"""The overall result in enumerated type"""
# pylint: disable=too-few-public-methods
EX_OK = os.EX_OK
EX_ERROR = -1
class BlockingTestFailed(Exception):
"""Exception when the blocking test fails"""
pass
class TestNotEnabled(Exception):
"""Exception when the test is not enabled"""
pass
class RunTestsParser(object):
"""Parser to run tests"""
# pylint: disable=too-few-public-methods
def __init__(self):
self.parser = argparse.ArgumentParser()
self.parser.add_argument("-t", "--test", dest="test", action='store',
help="Test case or tier (group of tests) "
"to be executed. It will run all the test "
"if not specified.")
self.parser.add_argument("-n", "--noclean", help="Do not clean "
"OpenStack resources after running each "
"test (default=false).",
action="store_true")
self.parser.add_argument("-r", "--report", help="Push results to "
"database (default=false).",
action="store_true")
def parse_args(self, argv=None):
"""Parse arguments.
It can call sys.exit if arguments are incorrect.
Returns:
the arguments from cmdline
"""
return vars(self.parser.parse_args(argv))
class Runner(object):
"""Runner class"""
def __init__(self):
self.executed_test_cases = {}
self.overall_result = Result.EX_OK
self.clean_flag = True
self.report_flag = False
self.tiers = tb.TierBuilder(
os.environ.get('INSTALLER_TYPE', None),
os.environ.get('DEPLOY_SCENARIO', None),
pkg_resources.resource_filename('functest', 'ci/testcases.yaml'))
@staticmethod
def source_envfile(rc_file=ENV_FILE):
"""Source the env file passed as arg"""
if not os.path.isfile(rc_file):
LOGGER.debug("No env file %s found", rc_file)
return
with open(rc_file, "r") as rcfd:
LOGGER.info("Sourcing env file %s", rc_file)
for line in rcfd:
var = (line.rstrip('"\n').replace('export ', '').split(
"=") if re.search(r'(.*)=(.*)', line) else None)
# The two next lines should be modified as soon as rc_file
# conforms with common rules. Be aware that it could induce
# issues if value starts with '
if var:
key = re.sub(r'^["\' ]*|[ \'"]*$', '', var[0])
value = re.sub(r'^["\' ]*|[ \'"]*$', '', "".join(var[1:]))
os.environ[key] = value
@staticmethod
def get_dict_by_test(testname):
# pylint: disable=bad-continuation,missing-docstring
with open(pkg_resources.resource_filename(
'functest', 'ci/testcases.yaml')) as tyaml:
testcases_yaml = yaml.safe_load(tyaml)
for dic_tier in testcases_yaml.get("tiers"):
for dic_testcase in dic_tier['testcases']:
if dic_testcase['case_name'] == testname:
return dic_testcase
LOGGER.error('Project %s is not defined in testcases.yaml', testname)
return None
@staticmethod
def get_run_dict(testname):
"""Obtain the 'run' block of the testcase from testcases.yaml"""
try:
dic_testcase = Runner.get_dict_by_test(testname)
if not dic_testcase:
LOGGER.error("Cannot get %s's config options", testname)
elif 'run' in dic_testcase:
return dic_testcase['run']
return None
except Exception: # pylint: disable=broad-except
LOGGER.exception("Cannot get %s's config options", testname)
return None
def run_test(self, test):
"""Run one test case"""
if not test.is_enabled():
raise TestNotEnabled(
"The test case {} is not enabled".format(test.get_name()))
LOGGER.info("Running test case '%s'...", test.get_name())
result = testcase.TestCase.EX_RUN_ERROR
run_dict = self.get_run_dict(test.get_name())
if run_dict:
try:
module = importlib.import_module(run_dict['module'])
cls = getattr(module, run_dict['class'])
test_dict = Runner.get_dict_by_test(test.get_name())
test_case = cls(**test_dict)
self.executed_test_cases[test.get_name()] = test_case
try:
kwargs = run_dict['args']
test_case.run(**kwargs)
except KeyError:
test_case.run()
if self.report_flag:
test_case.push_to_db()
if test.get_project() == "functest":
result = test_case.is_successful()
else:
result = testcase.TestCase.EX_OK
LOGGER.info("Test result:\n\n%s\n", test_case)
if self.clean_flag:
test_case.clean()
except ImportError:
LOGGER.exception("Cannot import module %s", run_dict['module'])
except AttributeError:
LOGGER.exception("Cannot get class %s", run_dict['class'])
else:
raise Exception("Cannot import the class for the test case.")
return result
def run_tier(self, tier):
"""Run one tier"""
tier_name = tier.get_name()
tests = tier.get_tests()
if not tests:
LOGGER.info("There are no supported test cases in this tier "
"for the given scenario")
self.overall_result = Result.EX_ERROR
else:
LOGGER.info("Running tier '%s'", tier_name)
for test in tests:
self.run_test(test)
test_case = self.executed_test_cases[test.get_name()]
if test_case.is_successful() != testcase.TestCase.EX_OK:
LOGGER.error("The test case '%s' failed.", test.get_name())
if test.get_project() == "functest":
self.overall_result = Result.EX_ERROR
if test.is_blocking():
raise BlockingTestFailed(
"The test case {} failed and is blocking".format(
test.get_name()))
return self.overall_result
def run_all(self):
"""Run all available testcases"""
tiers_to_run = []
msg = prettytable.PrettyTable(
header_style='upper', padding_width=5,
field_names=['tiers', 'order', 'CI Loop', 'description',
'testcases'])
for tier in self.tiers.get_tiers():
ci_loop = os.environ.get('CI_LOOP', 'daily')
if (tier.get_tests() and
re.search(ci_loop, tier.get_ci_loop()) is not None):
tiers_to_run.append(tier)
msg.add_row([tier.get_name(), tier.get_order(),
tier.get_ci_loop(),
textwrap.fill(tier.description, width=40),
textwrap.fill(' '.join([str(x.get_name(
)) for x in tier.get_tests()]), width=40)])
LOGGER.info("TESTS TO BE EXECUTED:\n\n%s\n", msg)
for tier in tiers_to_run:
self.run_tier(tier)
def main(self, **kwargs):
"""Entry point of class Runner"""
if 'noclean' in kwargs:
self.clean_flag = not kwargs['noclean']
if 'report' in kwargs:
self.report_flag = kwargs['report']
try:
if 'test' in kwargs:
LOGGER.debug("Sourcing the credential file...")
self.source_envfile()
LOGGER.debug("Test args: %s", kwargs['test'])
if self.tiers.get_tier(kwargs['test']):
self.run_tier(self.tiers.get_tier(kwargs['test']))
elif self.tiers.get_test(kwargs['test']):
result = self.run_test(
self.tiers.get_test(kwargs['test']))
if result != testcase.TestCase.EX_OK:
LOGGER.error("The test case '%s' failed.",
kwargs['test'])
self.overall_result = Result.EX_ERROR
elif kwargs['test'] == "all":
self.run_all()
else:
LOGGER.error("Unknown test case or tier '%s', or not "
"supported by the given scenario '%s'.",
kwargs['test'],
os.environ.get('DEPLOY_SCENARIO', ""))
LOGGER.debug("Available tiers are:\n\n%s",
self.tiers)
return Result.EX_ERROR
else:
self.run_all()
except BlockingTestFailed:
pass
except Exception: # pylint: disable=broad-except
LOGGER.exception("Failures when running testcase(s)")
self.overall_result = Result.EX_ERROR
if not self.tiers.get_test(kwargs['test']):
self.summary(self.tiers.get_tier(kwargs['test']))
LOGGER.info("Execution exit value: %s", self.overall_result)
return self.overall_result
def summary(self, tier=None):
"""To generate functest report showing the overall results"""
msg = prettytable.PrettyTable(
header_style='upper', padding_width=5,
field_names=['env var', 'value'])
for env_var in ['INSTALLER_TYPE', 'DEPLOY_SCENARIO', 'BUILD_TAG',
'CI_LOOP']:
msg.add_row([env_var, os.environ.get(env_var, "")])
LOGGER.info("Deployment description:\n\n%s\n", msg)
msg = prettytable.PrettyTable(
header_style='upper', padding_width=5,
field_names=['test case', 'project', 'tier',
'duration', 'result'])
tiers = [tier] if tier else self.tiers.get_tiers()
for each_tier in tiers:
for test in each_tier.get_tests():
try:
test_case = self.executed_test_cases[test.get_name()]
except KeyError:
msg.add_row([test.get_name(), test.get_project(),
each_tier.get_name(), "00:00", "SKIP"])
else:
result = 'PASS' if(test_case.is_successful(
) == test_case.EX_OK) else 'FAIL'
msg.add_row(
[test_case.case_name, test_case.project_name,
self.tiers.get_tier_name(test_case.case_name),
test_case.get_duration(), result])
for test in each_tier.get_skipped_test():
msg.add_row([test.get_name(), test.get_project(),
each_tier.get_name(), "00:00", "SKIP"])
LOGGER.info("FUNCTEST REPORT:\n\n%s\n", msg)
def main():
"""Entry point"""
logging.config.fileConfig(pkg_resources.resource_filename(
'functest', 'ci/logging.ini'))
logging.captureWarnings(True)
parser = RunTestsParser()
args = parser.parse_args(sys.argv[1:])
runner = Runner()
return runner.main(**args).value
|