summaryrefslogtreecommitdiff
path: root/buildscripts/resmokelib/testing/report.py
blob: f13cfdc9a8479366daacc0b611376d04e110eb5a (plain)
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
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
"""
Extension to the unittest.TestResult to support additional test status
and timing information for the report.json file.
"""

from __future__ import absolute_import

import copy
import threading
import time
import unittest

from .. import config as _config
from .. import logging


class TestReport(unittest.TestResult):
    """
    Records test status and timing information.
    """

    def __init__(self, job_logger, suite_options):
        """
        Initializes the TestReport with the buildlogger configuration.
        """

        unittest.TestResult.__init__(self)

        self.job_logger = job_logger
        self.suite_options = suite_options

        self._lock = threading.Lock()

        self.reset()

    @classmethod
    def combine(cls, *reports):
        """
        Merges the results from multiple TestReport instances into one.

        If the same test is present in multiple reports, then one that
        failed or errored is more preferred over one that succeeded.
        This behavior is useful for when running multiple jobs that
        dynamically add a #dbhash# test case.
        """

        # TestReports that are used when running tests need a JobLogger but combined reports don't
        # use the logger.
        combined_report = cls(logging.loggers.EXECUTOR_LOGGER,
                              _config.SuiteOptions.ALL_INHERITED.resolve())
        combining_time = time.time()

        for report in reports:
            if not isinstance(report, TestReport):
                raise TypeError("reports must be a list of TestReport instances")

            with report._lock:
                for test_info in report.test_infos:
                    # If the user triggers a KeyboardInterrupt exception while a test is running,
                    # then it is possible for 'test_info' to be modified by a job thread later on.
                    # We make a shallow copy in order to ensure 'num_interrupted' is consistent with
                    # the actual number of tests that have status equal to "timeout".
                    test_info = copy.copy(test_info)

                    # TestReport.addXX() may not have been called.
                    if test_info.status is None or test_info.return_code is None:
                        # Mark the test as having timed out if it was interrupted. It might have
                        # passed if the suite ran to completion, but we wouldn't know for sure.
                        #
                        # Until EVG-1536 is completed, we shouldn't distinguish between failures and
                        # interrupted tests in the report.json file. In Evergreen, the behavior to
                        # sort tests with the "timeout" test status after tests with the "pass" test
                        # status effectively hides interrupted tests from the test results sidebar
                        # unless sorting by the time taken.
                        test_info.status = "timeout"
                        test_info.evergreen_status = "fail"
                        test_info.return_code = -2

                    # TestReport.stopTest() may not have been called.
                    if test_info.end_time is None:
                        # Use the current time as the time that the test finished running.
                        test_info.end_time = combining_time

                    combined_report.test_infos.append(test_info)

                combined_report.num_dynamic += report.num_dynamic

        # Recompute number of success, failures, and errors.
        combined_report.num_succeeded = len(combined_report.get_successful())
        combined_report.num_failed = len(combined_report.get_failed())
        combined_report.num_errored = len(combined_report.get_errored())
        combined_report.num_interrupted = len(combined_report.get_interrupted())

        return combined_report

    def startTest(self, test, dynamic=False):
        """
        Called immediately before 'test' is run.
        """

        unittest.TestResult.startTest(self, test)

        test_info = _TestInfo(test.id(), dynamic)
        test_info.start_time = time.time()

        basename = test.basename()
        command = test.as_command()
        self.job_logger.info("Running %s...\n%s", basename, command)

        with self._lock:
            self.test_infos.append(test_info)
            if dynamic:
                self.num_dynamic += 1

        # Set up the test-specific logger.
        test_logger = self.job_logger.new_test_logger(test.short_name(), test.basename(), command,
                                                      test.logger)
        test_info.url_endpoint = test_logger.url_endpoint

        test.override_logger(test_logger)

    def stopTest(self, test):
        """
        Called immediately after 'test' has run.
        """

        unittest.TestResult.stopTest(self, test)

        with self._lock:
            test_info = self._find_test_info(test)
            test_info.end_time = time.time()

        time_taken = test_info.end_time - test_info.start_time
        self.job_logger.info("%s ran in %0.2f seconds.", test.basename(), time_taken)

        # Asynchronously closes the buildlogger test handler to avoid having too many threads open
        # on 32-bit systems.
        for handler in test.logger.handlers:
            # We ignore the cancellation token returned by close_later() since we always want the
            # logs to eventually get flushed.
            logging.flush.close_later(handler)

        # Restore the original logger for the test.
        test.reset_logger()

    def addError(self, test, err):
        """
        Called when a non-failureException was raised during the
        execution of 'test'.
        """

        unittest.TestResult.addError(self, test, err)

        with self._lock:
            self.num_errored += 1

            # We don't distinguish between test failures and Python errors in Evergreen.
            test_info = self._find_test_info(test)
            test_info.status = "error"
            test_info.evergreen_status = "fail"
            test_info.return_code = test.return_code

    def setError(self, test):
        """
        Used to change the outcome of an existing test to an error.
        """

        with self._lock:
            test_info = self._find_test_info(test)
            if test_info.end_time is None:
                raise ValueError("stopTest was not called on %s" % (test.basename()))

            # We don't distinguish between test failures and Python errors in Evergreen.
            test_info.status = "error"
            test_info.evergreen_status = "fail"
            test_info.return_code = 2

        # Recompute number of success, failures, and errors.
        self.num_succeeded = len(self.get_successful())
        self.num_failed = len(self.get_failed())
        self.num_errored = len(self.get_errored())
        self.num_interrupted = len(self.get_interrupted())

    def addFailure(self, test, err):
        """
        Called when a failureException was raised during the execution
        of 'test'.
        """

        unittest.TestResult.addFailure(self, test, err)

        with self._lock:
            self.num_failed += 1

            test_info = self._find_test_info(test)
            test_info.status = "fail"
            if test_info.dynamic:
                # Dynamic tests are used for data consistency checks, so the failures are never
                # silenced.
                test_info.evergreen_status = "fail"
            else:
                test_info.evergreen_status = self.suite_options.report_failure_status
            test_info.return_code = test.return_code

    def setFailure(self, test, return_code=1):
        """
        Used to change the outcome of an existing test to a failure.
        """

        with self._lock:
            test_info = self._find_test_info(test)
            if test_info.end_time is None:
                raise ValueError("stopTest was not called on %s" % (test.basename()))

            test_info.status = "fail"
            if test_info.dynamic:
                # Dynamic tests are used for data consistency checks, so the failures are never
                # silenced.
                test_info.evergreen_status = "fail"
            else:
                test_info.evergreen_status = self.suite_options.report_failure_status
            test_info.return_code = return_code

        # Recompute number of success, failures, and errors.
        self.num_succeeded = len(self.get_successful())
        self.num_failed = len(self.get_failed())
        self.num_errored = len(self.get_errored())
        self.num_interrupted = len(self.get_interrupted())

    def addSuccess(self, test):
        """
        Called when 'test' executed successfully.
        """

        unittest.TestResult.addSuccess(self, test)

        with self._lock:
            self.num_succeeded += 1

            test_info = self._find_test_info(test)
            test_info.status = "pass"
            test_info.evergreen_status = "pass"
            test_info.return_code = test.return_code

    def wasSuccessful(self):
        """
        Returns true if all tests executed successfully.
        """

        with self._lock:
            return self.num_failed == self.num_errored == self.num_interrupted == 0

    def get_successful(self):
        """
        Returns the status and timing information of the tests that
        executed successfully.
        """

        with self._lock:
            return [test_info for test_info in self.test_infos if test_info.status == "pass"]

    def get_failed(self):
        """
        Returns the status and timing information of the tests that
        raised a failureException during their execution.
        """

        with self._lock:
            return [test_info for test_info in self.test_infos if test_info.status == "fail"]

    def get_errored(self):
        """
        Returns the status and timing information of the tests that
        raised a non-failureException during their execution.
        """

        with self._lock:
            return [test_info for test_info in self.test_infos if test_info.status == "error"]

    def get_interrupted(self):
        """
        Returns the status and timing information of the tests that had
        their execution interrupted.
        """

        with self._lock:
            return [test_info for test_info in self.test_infos if test_info.status == "timeout"]

    def as_dict(self):
        """
        Return the test result information as a dictionary.

        Used to create the report.json file.
        """

        results = []
        with self._lock:
            for test_info in self.test_infos:
                result = {
                    "test_file": test_info.test_id,
                    "status": test_info.evergreen_status,
                    "exit_code": test_info.return_code,
                    "start": test_info.start_time,
                    "end": test_info.end_time,
                    "elapsed": test_info.end_time - test_info.start_time,
                }

                if test_info.url_endpoint is not None:
                    result["url"] = test_info.url_endpoint
                    result["url_raw"] = test_info.url_endpoint + "?raw=1"

                results.append(result)

            return {
                "results": results,
                "failures": self.num_failed + self.num_errored + self.num_interrupted,
            }

    @classmethod
    def from_dict(cls, report_dict):
        """
        Returns the test report instance copied from a dict (generated in as_dict).

        Used when combining reports instances.
        """

        report = cls(logging.loggers.EXECUTOR_LOGGER, _config.SuiteOptions.ALL_INHERITED.resolve())
        for result in report_dict["results"]:
            # By convention, dynamic tests are named "<basename>:<hook name>".
            is_dynamic = ":" in result["test_file"]
            test_info = _TestInfo(result["test_file"], is_dynamic)
            test_info.url_endpoint = result.get("url")
            test_info.status = result["status"]
            test_info.evergreen_status = test_info.status
            test_info.return_code = result["exit_code"]
            test_info.start_time = result["start"]
            test_info.end_time = result["end"]
            report.test_infos.append(test_info)

            if is_dynamic:
                report.num_dynamic += 1

        # Update cached values for number of successful and failed tests.
        report.num_failed = len(report.get_failed())
        report.num_errored = len(report.get_errored())
        report.num_interrupted = len(report.get_interrupted())
        report.num_succeeded = len(report.get_successful())

        return report

    def reset(self):
        """
        Resets the test report back to its initial state.
        """

        with self._lock:
            self.test_infos = []

            self.num_dynamic = 0
            self.num_succeeded = 0
            self.num_failed = 0
            self.num_errored = 0
            self.num_interrupted = 0

    def _find_test_info(self, test):
        """
        Returns the status and timing information associated with
        'test'.
        """

        test_id = test.id()

        # Search the list backwards to efficiently find the status and timing information of a test
        # that was recently started.
        for test_info in reversed(self.test_infos):
            if test_info.test_id == test_id:
                return test_info

        raise ValueError("Details for %s not found in the report" % (test.basename()))


class _TestInfo(object):
    """
    Holder for the test status and timing information.
    """

    def __init__(self, test_id, dynamic):
        """
        Initializes the _TestInfo instance.
        """

        self.test_id = test_id
        self.dynamic = dynamic

        self.start_time = None
        self.end_time = None
        self.status = None
        self.evergreen_status = None
        self.return_code = None
        self.url_endpoint = None