#!/usr/bin/env python # This Source Code Form is subject to the terms of the Mozilla Public # License, v. 2.0. If a copy of the MPL was not distributed with this # file, You can obtain one at http://mozilla.org/MPL/2.0/.
"""
The JS Shell Test Harness.
See the adjacent README.txt for more details. """
import math import os import platform import posixpath import re import shlex import sys import tempfile from contextlib import contextmanager from copy import copy from datetime import datetime from itertools import chain from os.path import abspath, dirname, isfile, realpath from subprocess import call, list2cmdline
from lib.adaptor import xdr_annotate from lib.progressbar import ProgressBar from lib.results import ResultsSink, TestOutput from lib.tempfile import TemporaryDirectory from lib.tests import (
RefTestCase,
change_env,
get_cpu_count,
get_environment_overlay,
get_jitflags,
)
if sys.platform.startswith("linux") or sys.platform.startswith("darwin"): from lib.tasks_unix import run_all_tests else: from lib.tasks_win import run_all_tests
@staticmethod def _split_files_and_dirs(location, paths): """Split up a set of paths into files and directories"""
files, dirs = set(), set() for path in paths:
fullpath = os.path.join(location, path) if path.endswith("/"):
dirs.add(path[:-1]) elif os.path.isdir(fullpath):
dirs.add(path) elif os.path.exists(fullpath):
files.add(path)
return files, dirs
def should_run(self, filename): # If any tests are requested by name, skip tests that do not match. if self.requested_paths andnot any(
req in filename for req in self.requested_paths
): returnFalse
# Skip excluded tests. if filename in self.excluded_files: returnFalse
for dir in self.excluded_dirs: if filename.startswith(dir + "/"): returnFalse
returnTrue
def parse_args(): """
Parse command line arguments.
Returns a tuple of: (options, js_shell, requested_paths, excluded_paths)
options :object: The raw OptionParser output.
js_shell :str: The absolute location of the shell to test with.
requested_paths :set<str>: Test paths specially requested on the CLI.
excluded_paths :set<str>: Test paths specifically excluded by the CLI. """ from argparse import ArgumentParser
op = ArgumentParser(
description="Run jstests JS shell tests",
epilog="Shell output format: [ pass | fail | timeout | skip ] progress | time",
)
op.add_argument( "--xul-info",
dest="xul_info_src",
help="config data for xulRuntime"" (avoids search for config/autoconf.mk)",
)
harness_og = op.add_argument_group("Harness Controls", "Control how tests are run.")
harness_og.add_argument( "-j", "--worker-count",
type=int,
default=max(1, get_cpu_count()),
help="Number of tests to run in parallel"" (default %(default)s)",
)
harness_og.add_argument( "-t", "--timeout",
type=float,
default=150.0,
help="Set maximum time a test is allows to run"" (in seconds).",
)
harness_og.add_argument( "--show-slow",
action="store_true",
help="Show tests taking longer than a minimum time"" (in seconds).",
)
harness_og.add_argument( "--slow-test-threshold",
type=float,
default=5.0,
help="Time in seconds a test can take until it is" "considered slow (default %(default)s).",
)
harness_og.add_argument( "-a", "--args",
dest="shell_args",
default="",
help="Extra args to pass to the JS shell.",
)
harness_og.add_argument( "--feature-args",
dest="feature_args",
default="",
help="Extra args to pass to the JS shell even when feature-testing.",
)
harness_og.add_argument( "--jitflags",
dest="jitflags",
default="none",
type=str,
help="IonMonkey option combinations. One of all," " debug, ion, and none (default %(default)s).",
)
harness_og.add_argument( "--tbpl",
action="store_true",
help="Runs each test in all configurations tbpl"" tests.",
)
harness_og.add_argument( "--tbpl-debug",
action="store_true",
help="Runs each test in some faster configurations"" tbpl tests.",
)
harness_og.add_argument( "-g", "--debug", action="store_true", help="Run a test in debugger."
)
harness_og.add_argument( "--debugger", default="gdb -q --args", help="Debugger command."
)
harness_og.add_argument( "-J", "--jorendb", action="store_true", help="Run under JS debugger."
)
harness_og.add_argument( "--passthrough",
action="store_true",
help="Run tests with stdin/stdout attached to"" caller.",
)
harness_og.add_argument( "--test-reflect-stringify",
dest="test_reflect_stringify",
help="instead of running tests, use them to test the " "Reflect.stringify code in specified file",
)
harness_og.add_argument( "--valgrind", action="store_true", help="Run tests in valgrind."
)
harness_og.add_argument( "--valgrind-args", default="", help="Extra args to pass to valgrind."
)
harness_og.add_argument( "--rr",
action="store_true",
help="Run tests under RR record-and-replay debugger.",
)
harness_og.add_argument( "-C", "--check-output",
action="store_true",
help="Run tests to check output for different jit-flags",
)
harness_og.add_argument( "--remote", action="store_true", help="Run tests on a remote device"
)
harness_og.add_argument( "--deviceIP",
action="store",
type=str,
dest="device_ip",
help="IP address of remote device to test",
)
harness_og.add_argument( "--devicePort",
action="store",
type=int,
dest="device_port",
default=20701,
help="port of remote device to test",
)
harness_og.add_argument( "--deviceSerial",
action="store",
type=str,
dest="device_serial",
default=None,
help="ADB device serial number of remote device to test",
)
harness_og.add_argument( "--remoteTestRoot",
dest="remote_test_root",
action="store",
type=str,
default="/data/local/tmp/test_root",
help="The remote directory to use as test root"" (e.g. %(default)s)",
)
harness_og.add_argument( "--localLib",
dest="local_lib",
action="store",
type=str,
help="The location of libraries to push -- preferably"" stripped",
)
harness_og.add_argument( "--no-xdr",
dest="use_xdr",
action="store_false",
help="Whether to disable caching of self-hosted parsed content in XDR format.",
)
input_og = op.add_argument_group("Inputs", "Change what tests are run.")
input_og.add_argument( "-f", "--file",
dest="test_file",
action="append",
help="Get tests from the given file.",
)
input_og.add_argument( "-x", "--exclude-file",
action="append",
help="Exclude tests from the given file.",
)
input_og.add_argument( "--wpt",
dest="wpt",
choices=["enabled", "disabled", "if-running-everything"],
default="if-running-everything",
help="Enable or disable shell web-platform-tests " "(default: enable if no test paths are specified).",
)
input_og.add_argument( "--include",
action="append",
dest="requested_paths",
default=[],
help="Include the given test file or directory.",
)
input_og.add_argument( "--exclude",
action="append",
dest="excluded_paths",
default=[],
help="Exclude the given test file or directory.",
)
input_og.add_argument( "-d", "--exclude-random",
dest="random",
action="store_false",
help='Exclude tests marked as "random."',
)
input_og.add_argument( "--run-skipped", action="store_true", help='Run tests marked as "skip."'
)
input_og.add_argument( "--run-only-skipped",
action="store_true",
help='Run only tests marked as "skip."',
)
input_og.add_argument( "--run-slow-tests",
action="store_true",
help='Do not skip tests marked as "slow."',
)
input_og.add_argument( "--no-extensions",
action="store_true",
help="Run only tests conforming to the ECMAScript 5"" standard.",
)
input_og.add_argument( "--repeat", type=int, default=1, help="Repeat tests the given number of times."
)
output_og = op.add_argument_group("Output", "Modify the harness and tests output.")
output_og.add_argument( "-s", "--show-cmd",
action="store_true",
help="Show exact commandline used to run each test.",
)
output_og.add_argument( "-o", "--show-output",
action="store_true",
help="Print each test's output to the file given by"" --output-file.",
)
output_og.add_argument( "-F", "--failed-only",
action="store_true",
help="If a --show-* option is given, only print"" output for failed tests.",
)
output_og.add_argument( "--no-show-failed",
action="store_true",
help="Don't print output for failed tests"" (no-op with --show-output).",
)
output_og.add_argument( "-O", "--output-file",
help="Write all output to the given file"" (default: stdout).",
)
output_og.add_argument( "--failure-file", help="Write all not-passed tests to the given file."
)
output_og.add_argument( "--no-progress",
dest="hide_progress",
action="store_true",
help="Do not show the progress bar.",
)
output_og.add_argument( "--tinderbox",
dest="format",
action="store_const",
const="automation",
help="Use automation-parseable output format.",
)
output_og.add_argument( "--format",
dest="format",
default="none",
choices=["automation", "none"],
help="Output format. Either automation or none"" (default %(default)s).",
)
output_og.add_argument( "--log-wptreport",
dest="wptreport",
action="store",
help="Path to write a Web Platform Tests report (wptreport)",
)
output_og.add_argument( "--this-chunk", type=int, default=1, help="The test chunk to run."
)
output_og.add_argument( "--total-chunks", type=int, default=1, help="The total number of test chunks."
)
special_og = op.add_argument_group( "Special", "Special modes that do not run tests."
)
special_og.add_argument( "--make-manifests",
metavar="BASE_TEST_PATH",
help="Generate reftest manifest files.",
)
op.add_argument("--js-shell", metavar="JS_SHELL", help="JS shell to run tests with")
op.add_argument( "-z", "--gc-zeal", help="GC zeal mode to use when running the shell"
)
options, args = op.parse_known_args()
# Need a shell unless in a special mode. ifnot options.make_manifests: ifnot args:
op.error("missing JS_SHELL argument")
options.js_shell = os.path.abspath(args.pop(0))
requested_paths = set(args)
# Valgrind, gdb, and rr are mutually exclusive. if sum(map(bool, (options.valgrind, options.debug, options.rr))) > 1:
op.error("--valgrind, --debug, and --rr are mutually exclusive.")
# Fill the debugger field, as needed. if options.debug: if options.debugger == "lldb":
debugger_prefix = ["lldb", "--"] else:
debugger_prefix = options.debugger.split() else:
debugger_prefix = []
if options.valgrind:
debugger_prefix = ["valgrind"] + options.valgrind_args.split() if os.uname()[0] == "Darwin":
debugger_prefix.append("--dsymutil=yes")
options.show_output = True if options.rr:
debugger_prefix = ["rr", "record"]
# If files with lists of tests to run were specified, add them to the # requested tests set. if options.test_file: for test_file in options.test_file:
requested_paths |= set(
[line.strip() for line in open(test_file).readlines()]
)
excluded_paths = set(options.excluded_paths)
# If files with lists of tests to exclude were specified, add them to the # excluded tests set. if options.exclude_file: for filename in options.exclude_file: with open(filename, "r") as fp: for line in fp: if line.startswith("#"): continue
line = line.strip() ifnot line: continue
excluded_paths.add(line)
# Handle output redirection, if requested and relevant.
options.output_fp = sys.stdout if options.output_file: ifnot options.show_cmd:
options.show_output = True try:
options.output_fp = open(options.output_file, "w") except IOError as ex: raise SystemExit("Failed to open output file: " + str(ex))
# Hide the progress bar if it will get in the way of other output.
options.hide_progress = (
options.format == "automation" ornot ProgressBar.conservative_isatty() or options.hide_progress
)
def load_wpt_tests(xul_tester, requested_paths, excluded_paths, update_manifest=True): """Return a list of `RefTestCase` objects for the jsshell testharness.js
tests filtered by the given paths and debug-ness."""
repo_root = abspath(os.path.join(here, "..", "..", ".."))
wp = os.path.join(repo_root, "testing", "web-platform")
wpt = os.path.join(wp, "tests")
def filter_jsshell_tests(it): for item_type, path, tests in it:
tests = set(item for item in tests if item.jsshell) if tests: yield item_type, path, tests
tests = [] for test in loader.tests[""]["testharness"]:
test_path = os.path.relpath(test.path, wpt)
scripts = [resolve(test_path, s) for s in test.scripts]
extra_helper_paths_for_test = extra_helper_paths + scripts
# We must create at least one test with the default options, along with # one test for each option given in a test-also annotation.
variants = [None]
flags = [] for m in test.itermeta(): # Search for prefs to enable that we recognize for pref in m.prefs:
pref_value = m.prefs[pref] ifnot pref.startswith(pref_prefix): continue
short_pref = pref.replace(pref_prefix, "") ifnot short_pref in recognized_prefs: continue
flags.append("--setpref=" + short_pref + "=" + pref_value)
if m.has_key("test-also"): # NOQA: W601
variants += m.get("test-also").split() for variant in variants:
test_case = RefTestCase(
wpt,
test_path,
extra_helper_paths=extra_helper_paths_for_test[:],
wpt=test,
) if variant:
test_case.options.append(variant) for flag in flags:
test_case.options.append(flag)
tests.append(test_case) return tests
def load_tests(options, requested_paths, excluded_paths): """
Returns a tuple: (test_count, test_gen)
test_count: [int] Number of tests that will be in test_gen
test_gen: [iterable<Test>] Tests found that should be run. """ import lib.manifest as manifest
# WPT tests are already run in the browser in their own harness.
wpt_enabled = options.wpt == "enabled"or (
options.wpt == "if-running-everything" and len(requested_paths) == 0 andnot options.make_manifests
) if wpt_enabled:
wpt_tests = load_wpt_tests(xul_tester, requested_paths, excluded_paths)
test_count += len(wpt_tests)
test_gen = chain(test_gen, wpt_tests)
if options.test_reflect_stringify isnotNone:
def trs_gen(tests): for test in tests:
test.test_reflect_stringify = options.test_reflect_stringify # Even if the test is not normally expected to pass, we still # expect reflect-stringify to be able to handle it.
test.expect = True
test.random = False
test.slow = False yield test
test_gen = trs_gen(test_gen)
if options.make_manifests:
manifest.make_manifests(options.make_manifests, test_gen)
sys.exit()
# Create a new test list. Apply each TBPL configuration to every test.
flags_list = None if options.tbpl:
flags_list = get_jitflags("all") elif options.tbpl_debug:
flags_list = get_jitflags("debug") else:
flags_list = get_jitflags(options.jitflags, none=None)
if flags_list:
def flag_gen(tests): for test in tests: for jitflags in flags_list:
tmp_test = copy(test)
tmp_test.jitflags = copy(test.jitflags)
tmp_test.jitflags.extend(jitflags) yield tmp_test
if options.test_file:
paths = set() for test_file in options.test_file:
paths |= set([line.strip() for line in open(test_file).readlines()])
test_gen = (_ for _ in test_gen if _.path in paths)
if options.no_extensions:
pattern = os.sep + "extensions" + os.sep
test_gen = (_ for _ in test_gen if pattern notin _.path)
ifnot options.random:
test_gen = (_ for _ in test_gen ifnot _.random)
if options.run_only_skipped:
options.run_skipped = True
test_gen = (_ for _ in test_gen ifnot _.enable)
ifnot options.run_slow_tests:
test_gen = (_ for _ in test_gen ifnot _.slow)
if options.repeat:
test_gen = (test for test in test_gen for i in range(options.repeat))
test_count *= options.repeat
return test_count, test_gen
def main():
options, prefix, requested_paths, excluded_paths = parse_args() if options.js_shell isnotNoneandnot (
isfile(options.js_shell) and os.access(options.js_shell, os.X_OK)
): if (
platform.system() != "Windows" or isfile(options.js_shell) ornot isfile(options.js_shell + ".exe") ornot os.access(options.js_shell + ".exe", os.X_OK)
):
print("Could not find executable shell: " + options.js_shell) return 1
if test_count == 0:
print("no tests selected") return 1
test_dir = dirname(abspath(__file__))
if options.debug: if test_count > 1:
print( "Multiple tests match command line arguments," " debugger can only run one"
) for tc in test_gen:
print(" {}".format(tc.path)) return 2
with changedir(test_dir), change_env(
test_environment
), TemporaryDirectory() as tempdir:
cmd = next(test_gen).get_command(prefix, tempdir) if options.show_cmd:
print(list2cmdline(cmd))
call(cmd) return 0
# The test_gen generator is converted into a list in # run_all_tests. Go ahead and do it here so we can apply # chunking. # # If chunking is enabled, determine which tests are part of this chunk. # This code was adapted from testing/mochitest/runtestsremote.py. if options.total_chunks > 1:
tests_per_chunk = math.ceil(test_count / float(options.total_chunks))
start = int(round((options.this_chunk - 1) * tests_per_chunk))
end = int(round(options.this_chunk * tests_per_chunk))
test_gen = list(test_gen)[start:end]
if options.remote:
results = ResultsSink("jstests", options, test_count) try: from lib.remote import init_device, init_remote_dir
device = init_device(options)
tempdir = posixpath.join(options.remote_test_root, "tmp")
jtd_tests = posixpath.join(options.remote_test_root, "tests", "tests")
init_remote_dir(device, jtd_tests)
device.push(test_dir, jtd_tests, timeout=600)
device.chmod(jtd_tests, recursive=True)
prefix[0] = options.js_shell if options.use_xdr:
test_gen = xdr_annotate(test_gen, options) for test in test_gen:
out = run_test_remote(test, device, prefix, tempdir, options)
results.push(out)
results.finish(True) except KeyboardInterrupt:
results.finish(False)
return 0 if results.all_passed() else 1
with changedir(test_dir), change_env(
test_environment
), TemporaryDirectory() as tempdir:
results = ResultsSink("jstests", options, test_count) try: for out in run_all_tests(test_gen, prefix, tempdir, results.pb, options):
results.push(out)
results.finish(True) except KeyboardInterrupt:
results.finish(False)
adb_cmd = ADBDevice._escape_command_line(cmd)
start = datetime.now() try: # Allow ADBError or ADBTimeoutError to terminate the test run, # but handle ADBProcessError in order to support the use of # non-zero exit codes in the JavaScript shell tests.
out = device.shell_output(
adb_cmd, env=env, cwd=options.remote_test_root, timeout=int(options.timeout)
)
returncode = 0 except ADBProcessError as e: # Treat ignorable intermittent adb communication errors as # skipped tests.
out = str(e.adb_process.stdout)
returncode = e.adb_process.exitcode
re_ignore = re.compile(r"error: (closed|device .* not found)") if returncode == 1 and re_ignore.search(out):
print("Skipping {} due to ignorable adb error {}".format(test.path, out))
test.skip_if_cond = "true"
returncode = test.SKIPPED_EXIT_STATUS
# We can't distinguish between stdout and stderr so we pass # the same buffer to both. return TestOutput(test, cmd, out, out, returncode, elapsed, False)
Die Informationen auf dieser Webseite wurden
nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit,
noch Qualität der bereit gestellten Informationen zugesichert.
Bemerkung:
Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.