/usr/lib/python3/dist-packages/pytest_benchmark/plugin.py is in python3-pytest-benchmark 3.0.0-1.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 | from __future__ import division
from __future__ import print_function
import argparse
import gc
import json
import operator
import os
import platform
import sys
import time
import traceback
from collections import defaultdict
from datetime import datetime
from math import ceil
from math import isinf
import py
import pytest
from . import __version__
from .compat import INT
from .compat import XRANGE
from .timers import compute_timer_precision
from .timers import default_timer
from .utils import NameWrapper
from .utils import SecondsDecimal
from .utils import cached_property
from .utils import first_or_value
from .utils import format_dict
from .utils import format_time
from .utils import get_commit_info
from .utils import get_current_time
from .utils import get_tag
from .utils import load_timer
from .utils import parse_compare_fail
from .utils import parse_rounds
from .utils import parse_save
from .utils import parse_seconds
from .utils import parse_sort
from .utils import parse_timer
from .utils import parse_warmup
from .utils import report_progress
from .utils import time_unit
try:
import statistics
except (ImportError, SyntaxError):
statistics = False
statistics_error = traceback.format_exc()
else:
from .stats import Stats
NUMBER_FMT = "{0:,.4f}" if sys.version_info[:2] > (2, 6) else "{0:.4f}"
ALIGNED_NUMBER_FMT = "{0:>{1},.4f}{2:<{3}}" if sys.version_info[:2] > (2, 6) else "{0:>{1}.4f}{2:<{3}}"
HISTOGRAM_CURRENT = "now"
class PerformanceRegression(pytest.UsageError):
pass
class FixtureAlreadyUsed(Exception):
pass
def pytest_report_header(config):
bs = config._benchmarksession
return ("benchmark: %(version)s (defaults:"
" timer=%(timer)s"
" disable_gc=%(disable_gc)s"
" min_rounds=%(min_rounds)s"
" min_time=%(min_time)s"
" max_time=%(max_time)s"
" calibration_precision=%(calibration_precision)s"
" warmup=%(warmup)s"
" warmup_iterations=%(warmup_iterations)s"
")") % dict(
bs.options,
version=__version__,
timer=bs.options.get("timer"),
)
def pytest_addoption(parser):
group = parser.getgroup("benchmark")
group.addoption(
"--benchmark-min-time",
metavar="SECONDS", type=parse_seconds, default="0.000005",
help="Minimum time per round in seconds. Default: %(default)r"
)
group.addoption(
"--benchmark-max-time",
metavar="SECONDS", type=parse_seconds, default="1.0",
help="Maximum run time per test - it will be repeated until this total time is reached. It may be "
"exceeded if test function is very slow or --benchmark-min-rounds is large (it takes precedence). "
"Default: %(default)r"
)
group.addoption(
"--benchmark-min-rounds",
metavar="NUM", type=parse_rounds, default=5,
help="Minimum rounds, even if total time would exceed `--max-time`. Default: %(default)r"
)
group.addoption(
"--benchmark-sort",
metavar="COL", type=parse_sort, default="min",
help="Column to sort on. Can be one of: 'min', 'max', 'mean' or 'stddev'. Default: %(default)r"
)
group.addoption(
"--benchmark-group-by",
metavar="LABEL", default="group",
help="How to group tests. Can be one of: 'group', 'name', 'fullname', 'func', 'fullfunc' or 'param'."
" Default: %(default)r"
)
group.addoption(
"--benchmark-timer",
metavar="FUNC", type=parse_timer, default=str(NameWrapper(default_timer)),
help="Timer to use when measuring time. Default: %(default)r"
)
group.addoption(
"--benchmark-calibration-precision",
metavar="NUM", type=int, default=10,
help="Precision to use when calibrating number of iterations. Precision of 10 will make the timer look 10 times"
" more accurate, at a cost of less precise measure of deviations. Default: %(default)r"
)
group.addoption(
"--benchmark-warmup",
metavar="KIND", nargs="?", default=parse_warmup("auto"), type=parse_warmup,
help="Activates warmup. Will run the test function up to number of times in the calibration phase. "
"See `--benchmark-warmup-iterations`. Note: Even the warmup phase obeys --benchmark-max-time. "
"Available KIND: 'auto', 'off', 'on'. Default: 'auto' (automatically activate on PyPy)."
)
group.addoption(
"--benchmark-warmup-iterations",
metavar="NUM", type=int, default=100000,
help="Max number of iterations to run in the warmup phase. Default: %(default)r"
)
group.addoption(
"--benchmark-verbose",
action="store_true", default=False,
help="Dump diagnostic and progress information."
)
group.addoption(
"--benchmark-disable-gc",
action="store_true", default=False,
help="Disable GC during benchmarks."
)
group.addoption(
"--benchmark-skip",
action="store_true", default=False,
help="Skip running any tests that contain benchmarks."
)
group.addoption(
"--benchmark-disable",
action="store_true", default=False,
help="Disable benchmarks. Benchmarked functions are only ran once and no stats are reported. Use this is you "
"want to run the test but don't do any benchmarking."
)
group.addoption(
"--benchmark-only",
action="store_true", default=False,
help="Only run benchmarks."
)
group.addoption(
"--benchmark-save",
metavar="NAME", type=parse_save,
help="Save the current run into 'STORAGE-PATH/counter_NAME.json'."
)
tag = get_tag()
group.addoption(
"--benchmark-autosave",
action='store_const', const=tag,
help="Autosave the current run into 'STORAGE-PATH/counter_%s.json" % tag,
)
group.addoption(
"--benchmark-save-data",
action="store_true",
help="Use this to make --benchmark-save and --benchmark-autosave include all the timing data,"
" not just the stats.",
)
group.addoption(
"--benchmark-compare",
metavar="NUM", nargs="?", default=[], const=True,
help="Compare the current run against run NUM or the latest saved run if unspecified."
)
group.addoption(
"--benchmark-compare-fail",
metavar="EXPR", nargs="+", type=parse_compare_fail,
help="Fail test if performance regresses according to given EXPR"
" (eg: min:5%% or mean:0.001 for number of seconds). Can be used multiple times."
)
group.addoption(
"--benchmark-storage",
metavar="STORAGE-PATH", default="./.benchmarks/%s-%s-%s-%s" % (
platform.system(),
platform.python_implementation(),
".".join(platform.python_version_tuple()[:2]),
platform.architecture()[0]
),
help="Specify a different path to store the runs (when --benchmark-save or --benchmark-autosave are used). "
"Default: %(default)r",
)
prefix = "benchmark_%s" % get_current_time()
group.addoption(
"--benchmark-histogram",
action='append', metavar="FILENAME-PREFIX", nargs="?", default=[], const=prefix,
help="Plot graphs of min/max/avg/stddev over time in FILENAME-PREFIX-test_name.svg. If FILENAME-PREFIX contains"
" slashes ('/') then directories will be created. Default: %r" % prefix
)
group.addoption(
"--benchmark-json",
metavar="PATH", type=argparse.FileType('wb'),
help="Dump a JSON report into PATH. "
"Note that this will include the complete data (all the timings, not just the stats)."
)
def pytest_addhooks(pluginmanager):
from . import hookspec
method = getattr(pluginmanager, "add_hookspecs", None)
if method is None:
method = pluginmanager.addhooks
method(hookspec)
class BenchmarkStats(object):
def __init__(self, fixture, iterations, options):
self.name = fixture.name
self.fullname = fixture.fullname
self.group = fixture.group
self.param = fixture.param
self.iterations = iterations
self.stats = Stats()
self.options = options
self.fixture = fixture
def __bool__(self):
return bool(self.stats)
def __nonzero__(self):
return bool(self.stats)
def get(self, key, default=None):
try:
return getattr(self.stats, key)
except AttributeError:
return getattr(self, key, default)
def __getitem__(self, key):
try:
return getattr(self.stats, key)
except AttributeError:
return getattr(self, key)
@property
def has_error(self):
return self.fixture.has_error
def json(self, include_data=True):
if include_data:
return dict(self.stats.as_dict, data=self.stats.data)
else:
return self.stats.as_dict
def update(self, duration):
self.stats.update(duration / self.iterations)
class BenchmarkFixture(object):
_precisions = {}
@classmethod
def _get_precision(cls, timer):
if timer in cls._precisions:
return cls._precisions[timer]
else:
return cls._precisions.setdefault(timer, compute_timer_precision(timer))
def __init__(self, node, disable_gc, timer, min_rounds, min_time, max_time, warmup, warmup_iterations,
calibration_precision, add_stats, logger, warner, disable, group=None):
self.name = node.name
self.fullname = node._nodeid
self.disable = disable
self.param = node.callspec.id if hasattr(node, 'callspec') else None
self.group = group
self.has_error = False
self._disable_gc = disable_gc
self._timer = timer.target
self._min_rounds = min_rounds
self._max_time = float(max_time)
self._min_time = float(min_time)
self._add_stats = add_stats
self._calibration_precision = calibration_precision
self._warmup = warmup and warmup_iterations
self._logger = logger
self._warner = warner
self._cleanup_callbacks = []
self._mode = None
def _make_runner(self, function_to_benchmark, args, kwargs):
def runner(loops_range, timer=self._timer):
gc_enabled = gc.isenabled()
if self._disable_gc:
gc.disable()
tracer = sys.gettrace()
sys.settrace(None)
try:
if loops_range:
start = timer()
for _ in loops_range:
function_to_benchmark(*args, **kwargs)
end = timer()
return end - start
else:
start = timer()
result = function_to_benchmark(*args, **kwargs)
end = timer()
return end - start, result
finally:
sys.settrace(tracer)
if gc_enabled:
gc.enable()
return runner
def _make_stats(self, iterations):
stats = BenchmarkStats(self, iterations=iterations, options={
"disable_gc": self._disable_gc,
"timer": self._timer,
"min_rounds": self._min_rounds,
"max_time": self._max_time,
"min_time": self._min_time,
"warmup": self._warmup,
})
self._add_stats(stats)
return stats
def __call__(self, function_to_benchmark, *args, **kwargs):
if self._mode:
self.has_error = True
raise FixtureAlreadyUsed(
"Fixture can only be used once. Previously it was used in %s mode." % self._mode)
try:
self._mode = 'benchmark(...)'
return self._raw(function_to_benchmark, *args, **kwargs)
except Exception:
self.has_error = True
raise
def pedantic(self, target, args=(), kwargs=None, setup=None, rounds=1, warmup_rounds=0, iterations=1):
if self._mode:
self.has_error = True
raise FixtureAlreadyUsed(
"Fixture can only be used once. Previously it was used in %s mode." % self._mode)
try:
self._mode = 'benchmark.pedantic(...)'
return self._raw_pedantic(target, args=args, kwargs=kwargs, setup=setup, rounds=rounds,
warmup_rounds=warmup_rounds, iterations=iterations)
except Exception:
self.has_error = True
raise
def _raw(self, function_to_benchmark, *args, **kwargs):
if not self.disable:
runner = self._make_runner(function_to_benchmark, args, kwargs)
duration, iterations, loops_range = self._calibrate_timer(runner)
# Choose how many time we must repeat the test
rounds = int(ceil(self._max_time / duration))
rounds = max(rounds, self._min_rounds)
rounds = min(rounds, sys.maxsize)
stats = self._make_stats(iterations)
self._logger.debug(" Running %s rounds x %s iterations ..." % (rounds, iterations), yellow=True, bold=True)
run_start = time.time()
if self._warmup:
warmup_rounds = min(rounds, max(1, int(self._warmup / iterations)))
self._logger.debug(" Warmup %s rounds x %s iterations ..." % (warmup_rounds, iterations))
for _ in XRANGE(warmup_rounds):
runner(loops_range)
for _ in XRANGE(rounds):
stats.update(runner(loops_range))
self._logger.debug(" Ran for %ss." % format_time(time.time() - run_start), yellow=True, bold=True)
return function_to_benchmark(*args, **kwargs)
def _raw_pedantic(self, target, args=(), kwargs=None, setup=None, rounds=1, warmup_rounds=0, iterations=1):
if kwargs is None:
kwargs = {}
has_args = bool(args or kwargs)
if not isinstance(iterations, INT) or iterations < 1:
raise ValueError("Must have positive int for `iterations`.")
if not isinstance(rounds, INT) or rounds < 1:
raise ValueError("Must have positive int for `rounds`.")
if not isinstance(warmup_rounds, INT) or warmup_rounds < 0:
raise ValueError("Must have positive int for `warmup_rounds`.")
if iterations > 1 and setup:
raise ValueError("Can't use more than 1 `iterations` with a `setup` function.")
def make_arguments(args=args, kwargs=kwargs):
if setup:
maybe_args = setup()
if maybe_args:
if has_args:
raise TypeError("Can't use `args` or `kwargs` if `setup` returns the arguments.")
args, kwargs = maybe_args
return args, kwargs
if self.disable:
args, kwargs = make_arguments()
return target(*args, **kwargs)
stats = self._make_stats(iterations)
loops_range = XRANGE(iterations) if iterations > 1 else None
for _ in XRANGE(warmup_rounds):
args, kwargs = make_arguments()
runner = self._make_runner(target, args, kwargs)
runner(loops_range)
for _ in XRANGE(rounds):
args, kwargs = make_arguments()
runner = self._make_runner(target, args, kwargs)
if loops_range:
duration = runner(loops_range)
else:
duration, result = runner(loops_range)
stats.update(duration)
if loops_range:
args, kwargs = make_arguments()
result = target(*args, **kwargs)
return result
def weave(self, target, **kwargs):
try:
import aspectlib
except ImportError as exc:
raise ImportError(exc.args, "Please install aspectlib or pytest-benchmark[aspect]")
def aspect(function):
def wrapper(*args, **kwargs):
return self(function, *args, **kwargs)
return wrapper
self._cleanup_callbacks.append(aspectlib.weave(target, aspect, **kwargs).rollback)
patch = weave
def _cleanup(self):
while self._cleanup_callbacks:
callback = self._cleanup_callbacks.pop()
callback()
if not self._mode:
self._logger.warn("BENCHMARK-U1", "Benchmark fixture was not used at all in this test!",
warner=self._warner, suspend=True)
def _calibrate_timer(self, runner):
timer_precision = self._get_precision(self._timer)
min_time = max(self._min_time, timer_precision * self._calibration_precision)
min_time_estimate = min_time * 5 / self._calibration_precision
self._logger.debug("")
self._logger.debug(" Timer precision: %ss" % format_time(timer_precision), yellow=True, bold=True)
self._logger.debug(" Calibrating to target round %ss; will estimate when reaching %ss." % (
format_time(min_time), format_time(min_time_estimate)), yellow=True, bold=True)
loops = 1
while True:
loops_range = XRANGE(loops)
duration = runner(loops_range)
if self._warmup:
warmup_start = time.time()
warmup_iterations = 0
warmup_rounds = 0
while time.time() - warmup_start < self._max_time and warmup_iterations < self._warmup:
duration = min(duration, runner(loops_range))
warmup_rounds += 1
warmup_iterations += loops
self._logger.debug(" Warmup: %ss (%s x %s iterations)." % (
format_time(time.time() - warmup_start),
warmup_rounds, loops
))
self._logger.debug(" Measured %s iterations: %ss." % (loops, format_time(duration)), yellow=True)
if duration >= min_time:
break
if duration >= min_time_estimate:
# coarse estimation of the number of loops
loops = int(ceil(min_time * loops / duration))
self._logger.debug(" Estimating %s iterations." % loops, green=True)
if loops == 1:
# If we got a single loop then bail early - nothing to calibrate if the the
# test function is 100 times slower than the timer resolution.
loops_range = XRANGE(loops)
break
else:
loops *= 10
return duration, loops, loops_range
class Logger(object):
def __init__(self, verbose, config):
self.verbose = verbose
self.term = py.io.TerminalWriter(file=sys.stderr)
self.capman = config.pluginmanager.getplugin("capturemanager")
self.pytest_warn = config.warn
try:
self.pytest_warn_has_fslocation = 'fslocation' in config.warn.func_code.co_varnames
except AttributeError:
self.pytest_warn_has_fslocation = False
def warn(self, code, text, warner=None, suspend=False, fslocation=None):
if self.verbose:
if suspend and self.capman:
self.capman.suspendcapture(in_=True)
self.term.line("")
self.term.sep("-", red=True, bold=True)
self.term.write(" WARNING: ", red=True, bold=True)
self.term.line(text, red=True)
self.term.sep("-", red=True, bold=True)
if suspend and self.capman:
self.capman.resumecapture()
if warner is None:
warner = self.pytest_warn
if fslocation and self.pytest_warn_has_fslocation:
warner(code=code, message=text, fslocation=fslocation)
else:
warner(code=code, message=text)
def error(self, text):
self.term.line("")
self.term.sep("-", red=True, bold=True)
self.term.line(text, red=True, bold=True)
self.term.sep("-", red=True, bold=True)
def info(self, text, **kwargs):
if not kwargs or kwargs == {'bold': True}:
kwargs['purple'] = True
self.term.line(text, **kwargs)
def debug(self, text, **kwargs):
if self.verbose:
if self.capman:
self.capman.suspendcapture(in_=True)
self.info(text, **kwargs)
if self.capman:
self.capman.resumecapture()
class BenchmarkSession(object):
compare_mapping = None
def __init__(self, config):
self.verbose = config.getoption("benchmark_verbose")
self.logger = Logger(self.verbose, config)
self.config = config
self.options = dict(
min_time=SecondsDecimal(config.getoption("benchmark_min_time")),
min_rounds=config.getoption("benchmark_min_rounds"),
max_time=SecondsDecimal(config.getoption("benchmark_max_time")),
timer=load_timer(config.getoption("benchmark_timer")),
calibration_precision=config.getoption("benchmark_calibration_precision"),
disable_gc=config.getoption("benchmark_disable_gc"),
warmup=config.getoption("benchmark_warmup"),
warmup_iterations=config.getoption("benchmark_warmup_iterations"),
)
self.skip = config.getoption("benchmark_skip")
self.disable = config.getoption("benchmark_disable")
if config.getoption("dist", "no") != "no" and not self.skip:
self.logger.warn(
"BENCHMARK-U2",
"Benchmarks are automatically disabled because xdist plugin is active."
"Benchmarks cannot be performed reliably in a parallelized environment.",
fslocation="::"
)
self.disable = True
if hasattr(config, "slaveinput"):
self.disable = True
if not statistics:
self.logger.warn(
"BENCHMARK-U3",
"Benchmarks are automatically disabled because we could not import `statistics`\n\n%s" %
statistics_error,
fslocation="::"
)
self.disable = True
self.only = config.getoption("benchmark_only")
self.sort = config.getoption("benchmark_sort")
if self.skip and self.only:
raise pytest.UsageError("Can't have both --benchmark-only and --benchmark-skip options.")
if self.disable and self.only:
raise pytest.UsageError(
"Can't have both --benchmark-only and --benchmark-disable options. Note that --benchmark-disable is "
"automatically activated if xdist is on or you're missing the statistics dependency.")
self._benchmarks = []
self.group_by = config.getoption("benchmark_group_by")
self.save = config.getoption("benchmark_save")
self.autosave = config.getoption("benchmark_autosave")
self.save_data = config.getoption("benchmark_save_data")
self.json = config.getoption("benchmark_json")
self.compare = config.getoption("benchmark_compare")
self.compare_fail = config.getoption("benchmark_compare_fail")
self.performance_regressions = []
self.storage = py.path.local(config.getoption("benchmark_storage"))
self.storage.ensure(dir=1)
self.histogram = first_or_value(config.getoption("benchmark_histogram"), False)
@property
def benchmarks(self):
return [bench for bench in self._benchmarks if bench]
@property
def storage_fslocation(self):
return self.storage.relto(os.getcwd())
@cached_property
def compare_file(self):
if self.compare:
files = self.storage.listdir("[0-9][0-9][0-9][0-9]_*.json", sort=True)
if files:
if self.compare is True:
files.sort()
return files[-1]
else:
files = [f for f in files if str(f.basename).startswith(self.compare)]
if len(files) == 1:
return files[0]
if not files:
self.logger.warn("BENCHMARK-C1", "Can't compare. No benchmark files matched %r" % self.compare,
fslocation=self.storage_fslocation)
elif len(files) > 1:
self.logger.warn(
"BENCHMARK-C2", "Can't compare. Too many benchmark files matched %r:\n - %s" % (
self.compare, '\n - '.join(map(str, files))
),
fslocation=self.storage_fslocation)
else:
msg = "Can't compare. No benchmark files in %r. " \
"Expected files matching [0-9][0-9][0-9][0-9]_*.json." % str(self.storage)
if self.compare is True:
msg += " Can't load the previous benchmark."
code = "BENCHMARK-C3"
else:
msg += " Can't match anything to %r." % self.compare
code = "BENCHMARK-C4"
self.logger.warn(code, msg, fslocation=self.storage_fslocation)
return
@property
def next_num(self):
files = self.storage.listdir("[0-9][0-9][0-9][0-9]_*.json")
files.sort(reverse=True)
if not files:
return "0001"
for f in files:
try:
return "%04i" % (int(str(f.basename).split('_')[0]) + 1)
except ValueError:
raise
def handle_saving(self):
if self.json:
output_json = self.config.hook.pytest_benchmark_generate_json(
config=self.config,
benchmarks=self.benchmarks,
include_data=True
)
self.config.hook.pytest_benchmark_update_json(
config=self.config,
benchmarks=self.benchmarks,
output_json=output_json
)
with self.json as fh:
fh.write(json.dumps(output_json, ensure_ascii=True, indent=4).encode())
self.logger.info("Wrote benchmark data in %s" % self.json, purple=True)
save = self.save or self.autosave
if save:
output_json = self.config.hook.pytest_benchmark_generate_json(
config=self.config,
benchmarks=self.benchmarks,
include_data=self.save_data
)
self.config.hook.pytest_benchmark_update_json(
config=self.config,
benchmarks=self.benchmarks,
output_json=output_json
)
output_file = self.storage.join("%s_%s.json" % (self.next_num, save))
assert not output_file.exists()
with output_file.open('wb') as fh:
fh.write(json.dumps(output_json, ensure_ascii=True, indent=4).encode())
self.logger.info("Saved benchmark data in %s" % output_file)
def handle_loading(self):
if self.compare_file:
self.compare_name = self.compare_file.basename.split('_')[0]
with self.compare_file.open('rU') as fh:
try:
compared_benchmark = json.load(fh)
except Exception as exc:
self.logger.warn("BENCHMARK-C5", "Failed to load %s: %s" % (self.compare_file, exc),
fslocation=self.storage_fslocation)
return
machine_info = self.config.hook.pytest_benchmark_generate_machine_info(config=self.config)
self.config.hook.pytest_benchmark_update_machine_info(config=self.config, machine_info=machine_info)
self.config.hook.pytest_benchmark_compare_machine_info(config=self.config, benchmarksession=self,
machine_info=machine_info,
compared_benchmark=compared_benchmark)
self.compare_mapping = dict((bench['fullname'], bench) for bench in compared_benchmark['benchmarks'])
self.logger.info("Comparing against benchmark %s:" % self.compare_file.basename, bold=True)
self.logger.info("| commit info: %s" % format_dict(compared_benchmark['commit_info']))
self.logger.info("| saved at: %s" % compared_benchmark['datetime'])
self.logger.info("| saved using pytest-benchmark %s:" % compared_benchmark['version'])
def display(self, tr):
if not self.benchmarks:
return
tr.ensure_newline()
self.handle_saving()
self.handle_loading()
if self.benchmarks:
self.display_results_table(tr)
self.check_regressions()
self.handle_histogram()
def check_regressions(self):
if self.compare_fail and not self.compare_file:
raise pytest.UsageError("--benchmark-compare-fail requires valid --benchmark-compare.")
if self.performance_regressions:
self.logger.error("Performance has regressed:\n%s" % "\n".join(
"\t%s - %s" % line for line in self.performance_regressions
))
raise PerformanceRegression("Performance has regressed.")
def handle_histogram(self):
if self.histogram:
from .histogram import make_plot
history = {}
for bench_file in self.storage.listdir("[0-9][0-9][0-9][0-9]_*.json"):
with bench_file.open('rU') as fh:
fullname = bench_file.purebasename
if '_' in fullname:
id_, name = fullname.split('_', 1)
else:
id_, name = fullname, ''
data = history[id_] = json.load(fh)
data['name'] = name
data['mapping'] = dict((bench['fullname'], bench) for bench in data['benchmarks'])
for bench in self.benchmarks:
name = bench.fullname
for c in "\/:*?<>|":
name = name.replace(c, '_').replace('__', '_')
output_file = py.path.local("%s-%s.svg" % (self.histogram, name)).ensure()
table = list(self.generate_histogram_table(bench, history, sorted(history)))
plot = make_plot(
bench_name=bench.fullname,
table=table,
compare=self.compare_file,
annotations=history,
sort=self.sort,
current=HISTOGRAM_CURRENT,
)
plot.render_to_file(str(output_file))
self.logger.info("Generated histogram %s" % output_file, bold=True)
@staticmethod
def generate_histogram_table(current, history, sequence):
for name in sequence:
trial = history[name]
for bench in trial["benchmarks"]:
if bench["fullname"] == current.fullname:
found = True
else:
found = False
if found:
yield "%s" % name, bench["stats"]
break
yield HISTOGRAM_CURRENT, current.json()
def apply_compare(self, benchmarks, compare_name, compare_mapping):
result = []
for bench in benchmarks:
if bench.fullname in compare_mapping:
stats = compare_mapping[bench.fullname]["stats"]
result.extend([
dict(bench.json(include_data=False),
name="{0} ({1})".format(bench.name, "NOW"),
iterations=bench.iterations),
dict(stats, name="{0} ({1})".format(bench.name, compare_name)),
])
if self.compare_fail:
for check in self.compare_fail:
fail = check.fails(bench, stats)
if fail:
self.performance_regressions.append((bench.fullname, fail))
else:
result.append(bench)
return result
def display_results_table(self, tr):
tr.write_line("")
tr.rewrite("Computing stats ...", black=True, bold=True)
groups = self.config.hook.pytest_benchmark_group_stats(
config=self.config,
benchmarks=self.benchmarks,
group_by=self.group_by
)
for line, (group, benchmarks) in report_progress(groups, tr, "Computing stats ... group {pos}/{total}"):
if self.compare_file:
benchmarks = self.apply_compare(benchmarks, self.compare_name, self.compare_mapping)
benchmarks = sorted(benchmarks, key=operator.itemgetter(self.sort))
worst = {}
best = {}
solo = len(benchmarks) == 1
for line, prop in report_progress(("min", "max", "mean", "median", "iqr", "stddev"), tr, "{line}: {value}", line=line):
worst[prop] = max(bench[prop] for _, bench in report_progress(
benchmarks, tr, "{line} ({pos}/{total})", line=line))
best[prop] = min(bench[prop] for _, bench in report_progress(
benchmarks, tr, "{line} ({pos}/{total})", line=line))
for line, prop in report_progress(("outliers", "rounds", "iterations"), tr, "{line}: {value}", line=line):
worst[prop] = max(benchmark[prop] for _, benchmark in report_progress(
benchmarks, tr, "{line} ({pos}/{total})", line=line))
unit, adjustment = time_unit(best.get(self.sort, benchmarks[0][self.sort]))
labels = {
"name": "Name (time in %ss)" % unit,
"min": "Min",
"max": "Max",
"mean": "Mean",
"stddev": "StdDev",
"rounds": "Rounds",
"iterations": "Iterations",
"iqr": "IQR",
"median": "Median",
"outliers": "Outliers(*)",
}
widths = {
"name": 3 + max(len(labels["name"]), max(len(benchmark["name"]) for benchmark in benchmarks)),
"rounds": 2 + max(len(labels["rounds"]), len(str(worst["rounds"]))),
"iterations": 2 + max(len(labels["iterations"]), len(str(worst["iterations"]))),
"outliers": 2 + max(len(labels["outliers"]), len(str(worst["outliers"]))),
}
for prop in "min", "max", "mean", "stddev", "median", "iqr":
widths[prop] = 2 + max(len(labels[prop]), max(
len(NUMBER_FMT.format(bench[prop] * adjustment))
for bench in benchmarks
))
rpadding = 0 if solo else 10
labels_line = labels["name"].ljust(widths["name"]) + "".join(
labels[prop].rjust(widths[prop]) + (
" " * rpadding
if prop not in ["outliers", "rounds", "iterations"]
else ""
)
for prop in ("min", "max", "mean", "stddev", "median", "iqr", "outliers", "rounds", "iterations")
)
tr.rewrite("")
tr.write_line(
(" benchmark%(name)s: %(count)s tests " % dict(
count=len(benchmarks),
name="" if group is None else " %r" % group,
)).center(len(labels_line), "-"),
yellow=True,
)
tr.write_line(labels_line)
tr.write_line("-" * len(labels_line), yellow=True)
for bench in benchmarks:
has_error = bench.get("has_error")
tr.write(bench["name"].ljust(widths["name"]), red=has_error, invert=has_error)
for prop in "min", "max", "mean", "stddev", "median", "iqr":
tr.write(
ALIGNED_NUMBER_FMT.format(
bench[prop] * adjustment,
widths[prop],
self.compute_baseline_scale(best[prop], bench[prop], rpadding),
rpadding
),
green=not solo and bench[prop] == best.get(prop),
red=not solo and bench[prop] == worst.get(prop),
bold=True,
)
for prop in "outliers", "rounds", "iterations":
tr.write("{0:>{1}}".format(bench[prop], widths[prop]))
tr.write("\n")
tr.write_line("-" * len(labels_line), yellow=True)
tr.write_line("")
tr.write_line("(*) Outliers: 1 Standard Deviation from Mean; "
"1.5 IQR (InterQuartile Range) from 1st Quartile and 3rd Quartile.", bold=True, black=True)
def compute_baseline_scale(self, baseline, value, width):
if not width:
return ""
if value == baseline:
return " (1.0)".ljust(width)
scale = abs(value / baseline) if baseline else float("inf")
if scale > 1000:
if isinf(scale):
return " (inf)".ljust(width)
else:
return " (>1000.0)".ljust(width)
else:
return " ({0:.2f})".format(scale).ljust(width)
def pytest_benchmark_compare_machine_info(config, benchmarksession, machine_info, compared_benchmark):
if compared_benchmark["machine_info"] != machine_info:
benchmarksession.logger.warn(
"BENCHMARK-C6",
"Benchmark machine_info is different. Current: %s VS saved: %s." % (
format_dict(machine_info),
format_dict(compared_benchmark["machine_info"]),
),
fslocation=benchmarksession.storage_fslocation
)
if hasattr(pytest, 'hookimpl'):
_hookwrapper = pytest.hookimpl(hookwrapper=True)
else:
_hookwrapper = pytest.mark.hookwrapper
@_hookwrapper
def pytest_runtest_call(item):
bs = item.config._benchmarksession
fixure = hasattr(item, "funcargs") and item.funcargs.get("benchmark")
if isinstance(fixure, BenchmarkFixture):
if bs.skip:
pytest.skip("Skipping benchmark (--benchmark-skip active).")
else:
yield
else:
if bs.only:
pytest.skip("Skipping non-benchmark (--benchmark-only active).")
else:
yield
def pytest_benchmark_group_stats(config, benchmarks, group_by):
groups = defaultdict(list)
for bench in benchmarks:
if group_by == "group":
groups[bench.group].append(bench)
elif group_by == "name":
groups[bench.name].append(bench)
elif group_by == "func":
groups[bench.name.split("[")[0]].append(bench)
elif group_by == "fullfunc":
groups[bench.fullname.split("[")[0]].append(bench)
elif group_by == "fullname":
groups[bench.fullname].append(bench)
elif group_by == "param":
groups[bench.param].append(bench)
else:
raise NotImplementedError("Unsupported grouping %r." % group_by)
for grouped_benchmarks in groups.values():
grouped_benchmarks.sort(key=operator.attrgetter("fullname" if "full" in group_by else "name"))
return sorted(groups.items(), key=lambda pair: pair[0] or "")
def pytest_terminal_summary(terminalreporter):
try:
terminalreporter.config._benchmarksession.display(terminalreporter)
except PerformanceRegression:
raise
except Exception:
terminalreporter.config._benchmarksession.logger.error("\n%s" % traceback.format_exc())
raise
def pytest_benchmark_generate_machine_info():
return {
"node": platform.node(),
"processor": platform.processor(),
"machine": platform.machine(),
"python_compiler": platform.python_compiler(),
"python_implementation": platform.python_implementation(),
"python_version": platform.python_version(),
"release": platform.release(),
"system": platform.system()
}
def pytest_benchmark_generate_commit_info(config):
return get_commit_info()
def pytest_benchmark_generate_json(config, benchmarks, include_data):
machine_info = config.hook.pytest_benchmark_generate_machine_info(config=config)
config.hook.pytest_benchmark_update_machine_info(config=config, machine_info=machine_info)
commit_info = config.hook.pytest_benchmark_generate_commit_info(config=config)
config.hook.pytest_benchmark_update_commit_info(config=config, commit_info=commit_info)
benchmarks_json = []
output_json = {
"machine_info": machine_info,
"commit_info": commit_info,
"benchmarks": benchmarks_json,
"datetime": datetime.utcnow().isoformat(),
"version": __version__,
}
for bench in benchmarks:
if not bench.has_error:
benchmarks_json.append({
"group": bench.group,
"name": bench.name,
"fullname": bench.fullname,
"stats": dict(bench.json(include_data=include_data), iterations=bench.iterations),
"options": dict(
(k, v.__name__ if callable(v) else v) for k, v in bench.options.items()
)
})
return output_json
@pytest.fixture(scope="function")
def benchmark(request):
bs = request.config._benchmarksession
if bs.skip:
pytest.skip("Benchmarks are skipped (--benchmark-skip was used).")
else:
node = request.node
marker = node.get_marker("benchmark")
options = marker.kwargs if marker else {}
if "timer" in options:
options["timer"] = NameWrapper(options["timer"])
fixture = BenchmarkFixture(
node,
add_stats=bs._benchmarks.append,
logger=bs.logger,
warner=request.node.warn,
disable=bs.disable,
**dict(bs.options, **options)
)
request.addfinalizer(fixture._cleanup)
return fixture
@pytest.fixture(scope="function")
def benchmark_weave(benchmark):
return benchmark.weave
def pytest_runtest_setup(item):
marker = item.get_marker("benchmark")
if marker:
if marker.args:
raise ValueError("benchmark mark can't have positional arguments.")
for name in marker.kwargs:
if name not in (
"max_time", "min_rounds", "min_time", "timer", "group", "disable_gc", "warmup",
"warmup_iterations", "calibration_precision"):
raise ValueError("benchmark mark can't have %r keyword argument." % name)
@pytest.mark.trylast # force the other plugins to initialise, fixes issue with capture not being properly initialised
def pytest_configure(config):
config.addinivalue_line("markers", "benchmark: mark a test with custom benchmark settings.")
config._benchmarksession = BenchmarkSession(config)
config.pluginmanager.register(config._benchmarksession, "pytest-benchmark")
|