summaryrefslogtreecommitdiff
path: root/Test/runTests.py
blob: 8caa65d1e4310936df1fd2479a58746c70deb902 (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
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
import os
import re
import sys
import csv
import shutil
import argparse
import operator
import platform
from math import floor, ceil
from enum import Enum
from time import time, strftime
from collections import defaultdict
from multiprocessing import Pool, Manager
from subprocess import Popen, call, PIPE, TimeoutExpired

# C:/Python34/python.exe runTests.py --compiler "c:/MSR/dafny/Binaries/Dafny.exe" --flags "/useBaseNameForFileName /compile:1 /nologo" --difftool "C:\Program Files (x86)\Meld\Meld.exe" -j4 --flags "/dprelude preludes\AlmostAllTriggers.bpl" dafny0\SeqFromArray.dfy

# c:/Python34/python.exe runTests.py --compare ../TestStable/results/SequenceAxioms/2015-06-06-00-54-52--PrettyPrinted.report.csv ../TestStable/results/SequenceAxioms/*.csv

VERBOSITY = None
KILLED = False
ANSI = False

try:
    import colorama
    no_native_ansi = os.name == 'nt' and os.environ.get("TERM") in [None, "cygwin"]
    tty = all(hasattr(stream, 'isatty') and stream.isatty() for stream in (sys.stdout, sys.stderr))
    colorama.init(strip=no_native_ansi, convert=no_native_ansi and tty)
    ANSI = True
except ImportError:
    try:
        import tendo.ansiterm
        ANSI = True
    except ImportError:
        pass

class Defaults:
    EXCLUDED_FILES = ["^flycheck_"]
    EXCLUDED_FOLDERS = ["Inputs", "Output", "sandbox", "desktop"]
    DAFNY_BIN = os.path.realpath(os.path.join(os.path.dirname(__file__), "../Binaries/Dafny.exe"))
    COMPILER = [DAFNY_BIN]
    FLAGS = ["/useBaseNameForFileName", "/compile:1", "/nologo", "/timeLimit:300"]
    EXTENSIONS = [".dfy", ".transcript"]

class Colors:
    RED = '\033[91m'
    GREEN = '\033[92m'
    YELLOW = '\033[93m'
    BRIGHT = '\033[1m'
    DIM = '\033[2m'
    RESET = '\033[0m'

class Debug(Enum):
    ERROR   = (-1, Colors.RED)
    WARNING = (-1, Colors.YELLOW)
    REPORT  = (0, Colors.RESET, True)
    INFO    = (1, Colors.RESET, True)
    DEBUG   = (2, Colors.RESET)
    TRACE   = (3, Colors.RESET)

    def __init__(self, index, color, elide=False):
        self.index = index
        self.color = color
        self.elide = elide

def wrap_color(string, color, silent=False):
    if silent:
        return " " * len(string)
    elif ANSI:
        return color + string + Colors.RESET
    else:
        return string

def debug(level, *args, **kwargs):
    kwargs["file"] = sys.stderr
    kwargs["flush"] = True

    headers = kwargs.pop("headers", [])
    if isinstance(headers, Enum):
        headers = [headers]

    silentheaders = kwargs.pop("silentheaders", False)

    if level and level.index <= VERBOSITY:
        if level:
            headers = [level] + headers

        headers = tuple(wrap_color("{: <8}".format("[" + h.name + "]"), h.color, silent = silentheaders)
                        for h in headers if not h.elide)
        print(*(headers + args), **kwargs)

class TestStatus(Enum):
    PENDING = (0, Colors.RESET)
    PASSED  = (1, Colors.GREEN)
    FAILED  = (2, Colors.RED)
    UNKNOWN = (3, Colors.RED)
    TIMEOUT = (4, Colors.RED)

    def __init__(self, index, color):
        self.index = index
        self.color = color
        self.elide = False

class Test:
    COLUMNS = ["name", "status", "start", "end", "duration", "returncodes", "suite_time", "njobs", "proc_info", "source_path", "temp_directory", "cmds", "expected", "output"]

    def __init__(self, name, source_path, cmds, timeout, compiler_id = 0):
        self.name = name
        self.source_path = Test.uncygdrive(source_path)
        self.expect_path = Test.source_to_expect_path(self.source_path)
        self.source_directory, self.fname = os.path.split(self.source_path)
        self.temp_directory = os.path.join(self.source_directory, "Output")
        self.temp_output_path = os.path.join(self.temp_directory, self.fname + ".tmp")

        self.output = None
        self.expected = Test.read_normalize(self.expect_path)

        self.cmds = cmds
        self.timeout = timeout
        self.compiler_id = compiler_id
        self.cmds = [cmd.replace("%s", self.source_path) for cmd in self.cmds]
        self.cmds = [cmd.replace("%S", self.source_directory) for cmd in self.cmds]
        self.cmds = [cmd.replace("%t", self.temp_output_path) for cmd in self.cmds]
        self.cmds = [cmd.replace("%T", self.temp_directory) for cmd in self.cmds]

        self.status = TestStatus.PENDING
        self.proc_info = platform.processor()

        self.time, self.suite_time = None, None
        self.njobs, self.returncodes = None, []
        self.start, self.end, self.duration = None, None, None

    @staticmethod
    def source_to_expect_path(source):
        return source + ".expect"

    @staticmethod
    def uncygdrive(path):
        return re.sub("^/cygdrive/([a-zA-Z])/", r"\1:/", path)

    @staticmethod
    def read_normalize(path):
        try:
            with open(path, mode="rb") as reader:
                return reader.read().replace(b'\r\n', b'\n').replace(b'\r', b'\n')
        except FileNotFoundError:
            debug(Debug.WARNING, "{} not found".format(path))
            return ""

    @staticmethod
    def build_report(tests, name):
        now = strftime("%Y-%m-%d-%H-%M-%S")
        if name:
            directory, fname = os.path.split(name)
            name = os.path.join(directory, now + "--" + fname)
        else:
            name = now

        with open(name + ".csv", mode='w', newline='') as writer:
            csv_writer = csv.DictWriter(writer, Test.COLUMNS, dialect='excel')
            csv_writer.writeheader()
            for test in tests:
                test.serialize(csv_writer)

    @staticmethod
    def load_report(path):
        results = []
        with open(path) as csvfile:
            for row in csv.DictReader(csvfile):  #, fieldnames=Test.COLUMNS):
                results.append(Test.deserialize(row))
        return results

    @staticmethod
    def mean_duration(results, margin):
        durations = sorted(result.duration for result in results
                           if result.status in (TestStatus.PASSED, TestStatus.FAILED))
        if len(durations) >= 15:
            lq = durations[floor(0.25 * len(durations))]
            hq = durations[ceil(0.85 * len(durations))]
            iqr = hq - lq
            filtered = [d for d in durations if (lq - margin * iqr) <= d <= (hq + margin * iqr)]
            if filtered:
                avg = sum(durations) / len(durations)
                trimmed_avg = sum(filtered) / len(filtered)
                outliers_count = len(durations) - len(filtered)
                msg = "mean completion time: {:.2f}s".format(avg)
                if outliers_count > 0:
                    msg += "; ignoring {} outliers: {:.2f}s".format(outliers_count, trimmed_avg)
                return " ({})".format(msg)
        return ""

    @staticmethod
    def summarize(results):
        debug(Debug.INFO, "\nTesting complete ({} test(s))".format(len(results)))

        if results:
            grouped = defaultdict(list)
            for test in results:
                grouped[test.status].append(test)

            for status, tests in sorted(grouped.items(), key=lambda x: x[0].index):
                if tests:
                    debug(Debug.REPORT, "{} of {}".format(len(tests), len(results)), headers=status)
                    if status != TestStatus.PASSED:
                        for test in tests:
                            debug(Debug.REPORT, "* " + test.name, headers=status, silentheaders=True)

            debug(Debug.REPORT)

            failing = [t for t in results if t.status != TestStatus.PASSED]
            if failing:
                with open("failing.lst", mode='w') as writer:
                    for t in failing:
                        writer.write("{}\n".format(t.name))
                debug(Debug.REPORT, "Some tests failed: use [runTests.py failing.lst] to rerun the failing tests")

            debug(Debug.REPORT, "Testing took {:.2f}s on {} thread(s){}".format(
                results[0].suite_time, results[0].njobs, Test.mean_duration(results, 1.5)))


    def run(self):
        debug(Debug.DEBUG, "Starting {}".format(self.name))
        os.makedirs(self.temp_directory, exist_ok=True)
        # os.chdir(self.source_directory)

        stdout, stderr = b'', b''
        self.start = time()

        try:
            for cmd in self.cmds:
                debug(Debug.DEBUG, "> {}".format(cmd))
                try:
                    proc = Popen(cmd, stdin=PIPE, stdout=PIPE, stderr=PIPE, shell=True)
                    _stdout, _stderr = proc.communicate(timeout=self.timeout)
                    stdout, stderr = stdout + _stdout, stderr + _stderr
                    self.returncodes.append(proc.returncode)
                except FileNotFoundError:
                    debug(Debug.ERROR, "Program '{}' not found".format(cmd))
                    self.status = TestStatus.UNKNOWN
                    return
                except TimeoutExpired:
                    self.status = TestStatus.TIMEOUT
                    self.end = self.start + self.timeout
                    self.duration = self.timeout
                    return

            self.end = time()
            self.duration = self.end - self.start

            stdout, stderr = stdout.strip(), stderr.strip()
            if stdout != b"":
                debug(Debug.TRACE, "Writing the output of {} to {}".format(self.name, self.temp_output_path))
                with open(self.temp_output_path, mode='ab') as writer:
                    writer.write(stdout)
            if stderr != b"":
                debug(Debug.INFO, stderr.decode("utf-8"))

            self.update_status()
        except TimeoutExpired:
            self.status = TestStatus.TIMEOUT
        except KeyboardInterrupt:
            raise

    def update_status(self):
        self.output = Test.read_normalize(self.temp_output_path)
        self.status = TestStatus.PASSED if self.expected == self.output else TestStatus.FAILED

    def report(self, tid, running, alltests):
        running = [alltests[rid].fname for rid in running]
        running = "; oldest: {}".format(running[0]) if running else ""

        fstring = "[{:5.2f}s] {} ({}{})"
        progress = "{}/{}".format(tid, len(alltests))
        message = fstring.format(self.duration, wrap_color(self.name, Colors.BRIGHT),
                                 wrap_color(progress, Colors.BRIGHT), running)

        debug(Debug.INFO, message, headers=self.status)

    @staticmethod
    def write_bytes(base_directory, relative_path, extension, contents):
        with open(os.path.join(base_directory, relative_path + extension), mode='wb') as writer:
            writer.write(contents)

    def serialize(self, csv_writer):
        csv_writer.writerow({col: getattr(self, col) for col in Test.COLUMNS})

    @classmethod
    def deserialize(cls, row):
        test = cls.__new__(cls)
        for col, val in row.items():
            setattr(test, col, val)
        test.duration = float(test.duration)
        test.status = next(x for x in TestStatus if str(x) == test.status)
        return test

def setup_parser():
    parser = argparse.ArgumentParser(description='Run the Dafny test suite.')

    parser.add_argument('path', type=str, action='store', nargs='+',
                        help='Input files or folders. Folders are searched for test files. Lists of files can also be specified by passing a .lst file (for an example of such a file, look at failing.lst after running failing tests.')

    parser.add_argument('--compiler', type=str, action='append', default=None,
                        help='Dafny executable. Default: {}'.format(Defaults.DAFNY_BIN))

    parser.add_argument('--base-flags', type=str, action='append', default=None,
                        help='Arguments to pass to dafny. Multiple --flags are concatenated. Default: {}'.format(Defaults.FLAGS))

    parser.add_argument('--flags', '-f', type=str, action='append', default=[],
                        help='Additional arguments to pass to dafny. Useful to override some of the defaults found in --base-flags.')

    parser.add_argument('--njobs', '-j', action='store', type=int, default=None,
                        help='Number of test workers.')

    parser.add_argument('--exclude', action='append', type=str, default=[],
                        help='Excluded directories. {} are automatically added.'.format(Defaults.EXCLUDED_FOLDERS))

    parser.add_argument('--verbosity', action='store', type=int, default=1,
                        help='Set verbosity level. 0: Minimal; 1: Some info; 2: More info; 3: Trace.')

    parser.add_argument('-v', action='store_const', default=1, dest="verbosity", const=2,
                        help='Short for --verbosity 2.')

    parser.add_argument('-vv', action='store_const', default=1, dest="verbosity", const=3,
                        help='Short for --verbosity 3.')

    parser.add_argument('--report', '-r', action='store', type=str, default=None,
                        help='Give an explicit name to the report file. Defaults to the current date and time.')

    parser.add_argument('--timeout', action='store', type=float, default=15*60.0,
                        help='Prover timeout')

    parser.add_argument('--compare', action='store_true',
                        help="Compare two previously generated reports.")

    parser.add_argument('--time-all', action='store_true',
                        help="When comparing, include all timings.")

    parser.add_argument('--diff', '-d', action='store_true',
                        help="Don't run tests; show differences between outputs and .expect files, optionally overwritting .expect files.")

    parser.add_argument('--accept', '-a', action='store_true',
                        help="Don't run tests; copy outputs to .expect files.")

    parser.add_argument('--open', '-o', action='store_true',
                        help="Don't run tests; open one file.")

    parser.add_argument('--difftool', action='store', type=str, default="diff",
                        help='Diff program. Default: diff.')

    return parser

def run_one_internal(test, test_id, args, running):
    global KILLED
    global VERBOSITY
    VERBOSITY = args.verbosity

    if not KILLED:
        try:
            running.append(test_id)
            test.run()
        except KeyboardInterrupt:
            # There's no reliable way to handle this cleanly on Windows: if one
            # of the worker dies, it gets respawned. The reliable solution is to
            # ignore further work once you receive a kill signal
            KILLED = True
        except Exception as e:
            debug(Debug.ERROR, "[{}] {}".format(test.name, e))
            test.status = TestStatus.UNKNOWN
        finally:
            running.remove(test_id)

    return test

def run_one(args):
    return run_one_internal(*args)

def get_server_path(compiler):
    REGEXP = r"\bDafny.exe\b.*"
    if re.search(REGEXP, compiler):
        return re.sub(REGEXP, "DafnyServer.exe", compiler)
    else:
        return None

def substitute_binaries(cmd, compiler):
    cmd = cmd.replace("%dafny", compiler)
    cmd = cmd.replace("%server", get_server_path(compiler))
    return cmd

def read_one_test(fname, compiler_cmds, timeout):
    for cid, compiler_cmd in enumerate(compiler_cmds):
        source_path = os.path.realpath(fname)
        with open(source_path, mode='r') as reader:
            cmds = []
            for line in reader:
                line = line.strip()
                match = re.match("^[/# ]*RUN: *(?!%diff)([^ ].*)$", line)
                if match:
                    debug(Debug.TRACE, "Found RUN spec: {}".format(line))
                    cmds.append(substitute_binaries(match.groups()[0], compiler_cmd))
                else:
                    break
        if cmds:
            yield Test(fname, source_path, cmds, timeout, cid)
        else:
            debug(Debug.WARNING, "Test file {} has no RUN specification".format(fname))


def find_one(fname, compiler_cmds, timeout):
    _, name = os.path.split(fname)
    _, ext = os.path.splitext(name)
    if ext in Defaults.EXTENSIONS and not any(re.search(pattern, name, re.IGNORECASE) for pattern in Defaults.EXCLUDED_FILES):
        if os.path.exists(fname):
            debug(Debug.TRACE, "Found test file: {}".format(fname))
            yield from read_one_test(fname, compiler_cmds, timeout)
        else:
            debug(Debug.ERROR, "Test file {} not found".format(fname))
    else:
        debug(Debug.TRACE, "Ignoring {}".format(fname))


def expand_lsts(paths):
    for path in paths:
        _, ext = os.path.splitext(path)
        if ext == ".lst": #lst files are only read if explicitly listed on the CLI
            debug(Debug.INFO, "Loading tests from {}".format(path))
            with open(path) as reader:
                for line in reader:
                    _path = line.strip()
                    yield _path
        else:
            yield path

def find_tests(paths, compiler_cmds, excluded, timeout):
    for path in expand_lsts(paths):
        if os.path.isdir(path):
            debug(Debug.TRACE, "Searching for tests in {}".format(path))
            for base, dirnames, fnames in os.walk(path):
                dirnames[:] = [d for d in dirnames if d not in excluded]
                for fname in fnames:
                    yield from find_one(os.path.join(base, fname), compiler_cmds, timeout)
        else:
            yield from find_one(path, compiler_cmds, timeout)

def run_tests(args):
    if args.compiler is None:
        args.compiler = Defaults.COMPILER
    if args.base_flags is None:
        args.base_flags = Defaults.FLAGS

    for compiler in args.compiler:
        server = get_server_path(compiler)
        if not os.path.exists(compiler):
            debug(Debug.ERROR, "Compiler not found: {}".format(compiler))
            return
        if not os.path.exists(server):
            debug(Debug.WARNING, "Server not found")

    tests = list(find_tests(args.path, [compiler + ' ' + " ".join(args.base_flags + args.flags)
                                        for compiler in args.compiler],
                            args.exclude + Defaults.EXCLUDED_FOLDERS, args.timeout))
    tests.sort(key=operator.attrgetter("name"))

    args.njobs = max(1, min(args.njobs or os.cpu_count() or 1, len(tests)))
    debug(Debug.INFO, "\nRunning {} test(s) on {} testing thread(s), timeout is {:.2f}s, started at {}".format(len(tests), args.njobs, args.timeout, strftime("%H:%M:%S")))

    try:
        pool = Pool(args.njobs)

        results = []
        start = time()
        with Manager() as manager:
            running = manager.list()
            payloads = [(t, tid, args, running) for (tid, t) in enumerate(tests)]
            for tid, test in enumerate(pool.imap_unordered(run_one, payloads, 1)):
                test.report(tid + 1, running, tests)
                results.append(test)
            pool.close()
            pool.join()
        suite_time = time() - start

        for t in results:
            t.njobs = args.njobs
            t.suite_time = suite_time

        Test.summarize(results)
        Test.build_report(results, args.report)
    except KeyboardInterrupt:
        try:
            pool.terminate()
            pool.join()
        except (FileNotFoundError, EOFError, ConnectionAbortedError):
            pass
        debug(Debug.ERROR, "Testing interrupted")


def diff(paths, force_accept, difftool):
    for path in expand_lsts(paths):
        if not os.path.exists(path):
            debug(Debug.ERROR, "Not found: {}".format(path))
        else:
            test = Test(None, path, [], None)
            accept = force_accept

            if not accept:
                call([difftool, test.expect_path, test.temp_output_path])
                accept = input("Accept this change? (y/N) ") == "y"

            if accept:
                debug(Debug.INFO, path, "accepted.")
                shutil.copy(test.temp_output_path, test.expect_path)
            else:
                debug(Debug.INFO, path, "not accepted.")

def compare_results(globs, time_all):
    from glob import glob
    paths = [path for g in globs for path in glob(g)]
    reports = {path: Test.load_report(path) for path in paths}
    resultsets = {path: {test.name: (test.status, test.duration) for test in report}
                  for path, report in reports.items()}

    all_tests = set(name for resultset in resultsets.values() for name in resultset.keys())

    reference = resultsets[paths[0]]
    for path, resultset in resultsets.items():
        resultset["$$TOTAL$$"] = None, sum(v[1] for v in resultset.values() if v[1] and v[0] != TestStatus.TIMEOUT)

    with open("compare.csv", mode='w', newline='') as writer:
        csv_writer = csv.writer(writer, dialect='excel')
        csv_writer.writerow(["Name"] + [os.path.split(path)[1].lstrip("0123456789-") for path in paths])

        for name in sorted(all_tests) + ["$$TOTAL$$"]:
            ref_status, ref_duration = reference[name]

            row = []
            row.append(name)
            row.append(ref_duration)
            for path in paths[1:]:
                res = resultsets[path].get(name)
                test_status, test_duration = res if res else (TestStatus.UNKNOWN, None)
                if res is not None and (test_status == ref_status or time_all):
                    result = "{:.2%}".format((test_duration - ref_duration) / ref_duration)
                else:
                    result = test_status.name + "?!"
                row.append(result)

            csv_writer.writerow(row)

def main():
    global VERBOSITY
    parser = setup_parser()
    args = parser.parse_args()
    VERBOSITY = args.verbosity

    if os.name != 'nt' and os.environ.get("TERM") == "cygwin":
        debug(Debug.WARNING, "If you run into issues, try using Windows' Python instead of Cygwin's")

    if args.diff or args.accept:
        diff(args.path, args.accept, args.difftool)
    elif args.open:
        os.startfile(args.path[0])
    elif args.compare:
        compare_results(args.path, args.time_all)
    else:
        run_tests(args)

if __name__ == '__main__':
    main()