mirror of
https://github.com/ruby/ruby.git
synced 2026-10-29 00:46:26 +08:00
William T. NelsonandClaude Opus 5.5
b77c9f32dc
timed-out memory leak child (https://github.com/ruby/test-unit-ruby-core/pull/37) * Close stdin before terminating a timed-out memory leak child With a sample range, assert_no_memory_leak keeps the child's stdin open, and the child's ensure clause reads a "continue" line from it after each sample. When the timeout fires while the child is running the sampled code, EnvUtil.terminate sends SIGTERM, and the child's ensure clause then blocks reading stdin, which was closed only after terminate returned. SIGTERM cannot end the child, so terminate escalates to SIGABRT and then waits for the aborted child without a time limit. On ruby/ruby's YJIT macOS CI jobs the child sometimes did not exit after the RUBY_ON_BUG core dump, and the test worker hung until the runner killed it 20 minutes later. In other runs the test passed but left a crash report for the child. Close stdin first. The child's read then returns nil, and its ensure clause breaks out of the sample loop. The break discards the SignalException raised by SIGTERM, so the child exits with status 0. If SIGTERM arrives before the child reaches the sample loop, the child still dies from SIGTERM. The existing timeout test now also checks that the child exited normally or by SIGTERM. https://github.com/ruby/test-unit-ruby-core/commit/ce7a92e724 Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
1223 lines
41 KiB
Ruby
1223 lines
41 KiB
Ruby
# frozen_string_literal: true
|
|
|
|
module Test
|
|
|
|
class << self
|
|
##
|
|
# Filter object for backtraces.
|
|
|
|
attr_accessor :backtrace_filter
|
|
end
|
|
|
|
class BacktraceFilter # :nodoc:
|
|
def filter bt
|
|
return ["No backtrace"] unless bt
|
|
|
|
new_bt = []
|
|
pattern = %r[/(?:lib\/test/|core_assertions\.rb:)]
|
|
|
|
unless $DEBUG then
|
|
bt.each do |line|
|
|
break if pattern =~ line
|
|
new_bt << line
|
|
end
|
|
|
|
new_bt = bt.reject { |line| pattern =~ line } if new_bt.empty?
|
|
new_bt = bt.dup if new_bt.empty?
|
|
else
|
|
new_bt = bt.dup
|
|
end
|
|
|
|
new_bt
|
|
end
|
|
end
|
|
|
|
self.backtrace_filter = BacktraceFilter.new
|
|
|
|
Ractor.make_shareable(backtrace_filter) if defined?(Ractor.make_shareable)
|
|
|
|
def self.filter_backtrace bt # :nodoc:
|
|
backtrace_filter.filter bt
|
|
end
|
|
|
|
module Unit
|
|
module Assertions
|
|
def assert_raises(*exp, &b)
|
|
raise NoMethodError, "use assert_raise", caller
|
|
end
|
|
|
|
def _assertions= n # :nodoc:
|
|
@_assertions = n
|
|
end
|
|
|
|
def _assertions # :nodoc:
|
|
@_assertions ||= 0
|
|
end
|
|
|
|
##
|
|
# Returns a proc that will output +msg+ along with the default message.
|
|
|
|
def message msg = nil, ending = nil, &default
|
|
proc {
|
|
ending ||= (ending_pattern = /(?<!\.)\z/; ".")
|
|
ending_pattern ||= /(?<!#{Regexp.quote(ending)})\z/
|
|
msg = msg.call if Proc === msg
|
|
ary = [msg, (default.call if default)].compact.reject(&:empty?)
|
|
ary.map! {|str| str.to_s.sub(ending_pattern, ending) }
|
|
begin
|
|
ary.join("\n")
|
|
rescue Encoding::CompatibilityError
|
|
ary.map(&:b).join("\n")
|
|
end
|
|
}
|
|
end
|
|
end
|
|
|
|
module CoreAssertions
|
|
require_relative 'envutil'
|
|
require 'pp'
|
|
begin
|
|
require '-test-/sanitizers'
|
|
rescue LoadError
|
|
# in test-unit-ruby-core gem
|
|
def sanitizers
|
|
nil
|
|
end
|
|
else
|
|
def sanitizers
|
|
Test::Sanitizers
|
|
end
|
|
end
|
|
module_function :sanitizers
|
|
|
|
nil.pretty_inspect
|
|
|
|
def mu_pp(obj) #:nodoc:
|
|
obj.pretty_inspect.chomp
|
|
end
|
|
|
|
def assert_file
|
|
AssertFile
|
|
end
|
|
|
|
FailDesc = proc do |status, message = "", out = ""|
|
|
now = Time.now
|
|
proc do
|
|
EnvUtil.failure_description(status, now, message, out)
|
|
end
|
|
end
|
|
|
|
def assert_in_out_err(args, test_stdin = "", test_stdout = [], test_stderr = [], message = nil,
|
|
success: nil, failed: nil, gems: false, **opt)
|
|
args = Array(args).dup
|
|
unless gems.nil?
|
|
args.insert((Hash === args[0] ? 1 : 0), "--#{gems ? 'enable' : 'disable'}=gems")
|
|
end
|
|
stdout, stderr, status = EnvUtil.invoke_ruby(args, test_stdin, true, true, **opt)
|
|
desc = failed[status, message, stderr] if failed
|
|
desc ||= FailDesc[status, message, stderr]
|
|
if block_given?
|
|
raise "test_stdout ignored, use block only or without block" if test_stdout != []
|
|
raise "test_stderr ignored, use block only or without block" if test_stderr != []
|
|
yield(stdout.lines.map {|l| l.chomp }, stderr.lines.map {|l| l.chomp }, status)
|
|
else
|
|
all_assertions(desc) do |a|
|
|
[["stdout", test_stdout, stdout], ["stderr", test_stderr, stderr]].each do |key, exp, act|
|
|
a.for(key) do
|
|
if exp.is_a?(Regexp)
|
|
assert_match(exp, act)
|
|
elsif exp.all? {|e| String === e}
|
|
assert_equal(exp, act.lines.map {|l| l.chomp })
|
|
else
|
|
assert_pattern_list(exp, act)
|
|
end
|
|
end
|
|
end
|
|
unless success.nil?
|
|
a.for("success?") do
|
|
if success
|
|
assert_predicate(status, :success?)
|
|
else
|
|
assert_not_predicate(status, :success?)
|
|
end
|
|
end
|
|
end
|
|
end
|
|
status
|
|
end
|
|
end
|
|
|
|
if defined?(RubyVM::InstructionSequence)
|
|
def syntax_check(code, fname, line)
|
|
code = code.dup.force_encoding(Encoding::UTF_8)
|
|
RubyVM::InstructionSequence.compile(code, fname, fname, line)
|
|
:ok
|
|
ensure
|
|
raise if SyntaxError === $!
|
|
end
|
|
else
|
|
def syntax_check(code, fname, line)
|
|
code = code.b
|
|
code.sub!(/\A(?:\xef\xbb\xbf)?(\s*\#.*$)*(\n)?/n) {
|
|
"#$&#{"\n" if $1 && !$2}BEGIN{throw tag, :ok}\n"
|
|
}
|
|
code = code.force_encoding(Encoding::UTF_8)
|
|
catch {|tag| eval(code, binding, fname, line - 1)}
|
|
end
|
|
end
|
|
|
|
private def memory_growth_rate(points)
|
|
n_1 = (n = points.length) - 1
|
|
mean_x = n_1 / 2.0
|
|
mean_y = points.inject(:+).fdiv(n)
|
|
covariance = 0
|
|
points.each_with_index {|value, i| covariance += (i - mean_x) * (value - mean_y)}
|
|
variance = (n_1 * n * (2 * n - 1) / 6) - (n_1 * n * mean_x) + (n * mean_x**2)
|
|
covariance / variance
|
|
end
|
|
|
|
private def memory_growth_rates(points)
|
|
midpoint = (points.length - 1) / 2
|
|
[memory_growth_rate(points[0..midpoint]), memory_growth_rate(points[midpoint..-1])]
|
|
end
|
|
|
|
private def memory_growth_tail_rates(points)
|
|
window = [points.length / 4, 2].max
|
|
first = points[[points.length - window * 2, 0].max, window]
|
|
last = points[-window, window]
|
|
[memory_growth_rate(first), memory_growth_rate(last)]
|
|
end
|
|
|
|
private def memory_growth_settled_in_tail?(points, growth_rate_ratio)
|
|
early_rate, late_rate = memory_growth_tail_rates(points)
|
|
memory_growth_settled?(early_rate, late_rate, growth_rate_ratio)
|
|
end
|
|
|
|
private def memory_growth_settled?(early_rate, late_rate, growth_rate_ratio)
|
|
if early_rate <= 0
|
|
late_rate <= 0
|
|
else
|
|
late_rate <= early_rate * growth_rate_ratio
|
|
end
|
|
end
|
|
|
|
# growth_rate_ratio compares late and early memory growth rates. With a
|
|
# sample range, sampling stops when the late rate is at most this ratio
|
|
# of the early rate, after collecting the minimum number of samples.
|
|
def assert_no_memory_leak(args, prepare, code, message=nil, limit: 2.0, rss: false, samples: 1, growth_rate_ratio: 0.25, **opt)
|
|
# TODO: consider choosing some appropriate limit for RJIT and stop skipping this once it does not randomly fail
|
|
pend 'assert_no_memory_leak may consider RJIT memory usage as leak' if defined?(RubyVM::RJIT) && RubyVM::RJIT.enabled?
|
|
# For previous versions which implemented MJIT
|
|
pend 'assert_no_memory_leak may consider MJIT memory usage as leak' if defined?(RubyVM::MJIT) && RubyVM::MJIT.enabled?
|
|
# ASAN has the same problem - its shadow memory greatly increases memory usage
|
|
# (plus asan has better ways to detect memory leaks than this assertion)
|
|
pend 'assert_no_memory_leak may consider ASAN memory usage as leak' if sanitizers&.asan_enabled?
|
|
|
|
require_relative 'memory_status'
|
|
raise Test::Unit::PendedError, "unsupported platform" unless defined?(Memory::Status)
|
|
|
|
token_dump, token_re = new_test_token
|
|
start_re = /^#{token_re}START=(\{.*\})\n/
|
|
sample_re = /^#{token_re}SAMPLE=(\{.*\})\n/
|
|
final_re = /^#{token_re}FINAL=(\{.*\})\n/
|
|
envs = args.shift if Array === args and Hash === args.first
|
|
case samples
|
|
when Integer
|
|
unless samples > 0
|
|
raise ArgumentError, "samples must be a positive integer or a bounded range"
|
|
end
|
|
min_samples = max_samples = samples
|
|
adaptive_samples = false
|
|
when Range
|
|
min_samples = samples.begin
|
|
max_samples = samples.end
|
|
max_samples -= 1 if samples.exclude_end? && Integer === max_samples
|
|
unless Integer === min_samples and Integer === max_samples and min_samples >= 2 and max_samples >= min_samples
|
|
raise ArgumentError, "samples range must have a minimum of 2 and a bounded positive maximum"
|
|
end
|
|
adaptive_samples = true
|
|
else
|
|
raise ArgumentError, "samples must be a positive integer or a bounded range"
|
|
end
|
|
args = [
|
|
"--disable=gems",
|
|
"-r", File.expand_path("../memory_status", __FILE__),
|
|
*args,
|
|
"-v", "-",
|
|
]
|
|
if defined? Memory::NO_MEMORY_LEAK_ENVS then
|
|
envs ||= {}
|
|
newenvs = envs.merge(Memory::NO_MEMORY_LEAK_ENVS) { |_, _, _| break }
|
|
envs = newenvs if newenvs
|
|
end
|
|
args.unshift(envs) if envs
|
|
unless (0..1).cover?(growth_rate_ratio)
|
|
raise ArgumentError, "growth_rate_ratio must be between 0 and 1"
|
|
end
|
|
|
|
cmd = <<~RUBY
|
|
STDERR.sync = true
|
|
#{<<~ADAPTIVE if adaptive_samples}
|
|
memory_leak_control = STDIN.dup
|
|
STDIN.reopen(File::NULL, "r")
|
|
ADAPTIVE
|
|
END {STDERR.puts #{token_dump}"FINAL=\#{Memory::Status.new}"}
|
|
#{prepare}
|
|
STDERR.puts(#{token_dump}"START=\#{$initial_status = Memory::Status.new}")
|
|
$initial_size = $initial_status.size
|
|
#{max_samples}.times do
|
|
begin
|
|
#{code.gsub(/^/, ' ')}
|
|
ensure
|
|
GC.start
|
|
STDERR.puts(#{token_dump}"SAMPLE=\#{Memory::Status.new}")
|
|
#{<<~ADAPTIVE if adaptive_samples}
|
|
break unless memory_leak_control.gets == "continue\\n"
|
|
ADAPTIVE
|
|
end
|
|
end
|
|
__END__
|
|
RUBY
|
|
sample_statuses = []
|
|
before = after = nil
|
|
if adaptive_samples
|
|
stdout, err, status = EnvUtil.invoke_ruby(args, "", true, true, **opt) do |in_p, out_p, err_p, pid|
|
|
out_reader = Thread.new {out_p.read}
|
|
in_p.write(cmd)
|
|
in_p.flush
|
|
err = +""
|
|
timeout = EnvUtil.apply_timeout_scale(opt.fetch(:timeout, 10))
|
|
timed_out = nil
|
|
begin
|
|
Timeout.timeout(timeout) do
|
|
while line = err_p.gets
|
|
case line
|
|
when start_re
|
|
before = Memory::Status.parse($1)
|
|
when sample_re
|
|
sample_statuses << Memory::Status.parse($1)
|
|
count = sample_statuses.length
|
|
settled = false
|
|
if count >= min_samples && count < max_samples
|
|
metrics = [:size, (rss && :rss)] & sample_statuses.last.members
|
|
settled = metrics.all? do |n|
|
|
values = [before[n], *sample_statuses.map {|sample| sample[n]}]
|
|
next true unless values.all? {|value| value && value > 0}
|
|
|
|
memory_growth_settled_in_tail?(values, growth_rate_ratio)
|
|
end
|
|
end
|
|
if settled || count == max_samples
|
|
in_p.close
|
|
else
|
|
in_p.write("continue\n")
|
|
in_p.flush
|
|
end
|
|
when final_re
|
|
after = Memory::Status.parse($1)
|
|
else
|
|
err << line
|
|
end
|
|
end
|
|
in_p.close
|
|
status = Process.wait2(pid)[1]
|
|
end
|
|
rescue Timeout::Error
|
|
timed_out = Time.now
|
|
in_p.close
|
|
status = begin
|
|
EnvUtil.terminate(pid, opt.fetch(:signal, :TERM), opt[:pgroup], opt.fetch(:reprieve, 1))
|
|
rescue Errno::ESRCH, Errno::ECHILD
|
|
warn "child not found: #{pid}: #$!"
|
|
end
|
|
end
|
|
stdout = out_reader.value
|
|
if timed_out
|
|
message = "execution of assert_no_memory_leak expired timeout (#{timeout} sec)"
|
|
message = EnvUtil.failure_description(status, timed_out, message, [stdout, err].join("\n")) if status
|
|
raise opt.fetch(:timeout_error, Timeout::Error), message
|
|
end
|
|
[stdout, err, status]
|
|
end
|
|
else
|
|
stdout, err, status = EnvUtil.invoke_ruby(args, cmd, true, true, **opt)
|
|
before = err.sub!(start_re, '') && Memory::Status.parse($1)
|
|
err.gsub!(sample_re) {sample_statuses << Memory::Status.parse($1); ''}
|
|
after = err.sub!(final_re, '') && Memory::Status.parse($1)
|
|
end
|
|
assert(status.success?, FailDesc[status, message, err])
|
|
assert_operator(sample_statuses.length, :>=, min_samples, "could not collect the minimum memory samples")
|
|
assert_operator(sample_statuses.length, :<=, max_samples, "collected more than the maximum memory samples")
|
|
assert_equal(samples, sample_statuses.length, "could not collect all memory samples") unless adaptive_samples
|
|
([:size, (rss && :rss)] & after.members).each do |n|
|
|
if max_samples == 1
|
|
b = before[n]
|
|
a = after[n]
|
|
next unless a > 0 and b > 0
|
|
assert_operator(a.fdiv(b), :<, limit, message(message) {"#{n}: #{b} => #{a}"})
|
|
else
|
|
values = [before[n], *sample_statuses.map {|sample| sample[n]}]
|
|
next unless values.all? {|value| value && value > 0}
|
|
|
|
if adaptive_samples
|
|
early_rate, late_rate = memory_growth_tail_rates(values)
|
|
settled = memory_growth_settled?(early_rate, late_rate, growth_rate_ratio)
|
|
else
|
|
early_rate, late_rate = memory_growth_rates(values)
|
|
settled = memory_growth_settled?(early_rate, late_rate, growth_rate_ratio)
|
|
end
|
|
assert(settled, message(message) {
|
|
"#{n} growth did not settle: #{values.inspect} " \
|
|
"(early rate: #{early_rate}, late rate: #{late_rate})"
|
|
})
|
|
end
|
|
end
|
|
rescue LoadError
|
|
pend
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_nothing_raised( *args, &block )
|
|
#
|
|
#If any exceptions are given as arguments, the assertion will
|
|
#fail if one of those exceptions are raised. Otherwise, the test fails
|
|
#if any exceptions are raised.
|
|
#
|
|
#The final argument may be a failure message.
|
|
#
|
|
# assert_nothing_raised RuntimeError do
|
|
# raise Exception #Assertion passes, Exception is not a RuntimeError
|
|
# end
|
|
#
|
|
# assert_nothing_raised do
|
|
# raise Exception #Assertion fails
|
|
# end
|
|
def assert_nothing_raised(*args)
|
|
self._assertions += 1
|
|
if Module === args.last
|
|
msg = nil
|
|
else
|
|
msg = args.pop
|
|
end
|
|
begin
|
|
yield
|
|
rescue Test::Unit::PendedError, *(Test::Unit::AssertionFailedError if args.empty?)
|
|
raise
|
|
rescue *(args.empty? ? Exception : args) => e
|
|
msg = message(msg) {
|
|
"Exception raised:\n<#{mu_pp(e)}>\n""Backtrace:\n" <<
|
|
Test.filter_backtrace(e.backtrace).map{|frame| " #{frame}"}.join("\n")
|
|
}
|
|
raise Test::Unit::AssertionFailedError, msg.call, e.backtrace
|
|
end
|
|
end
|
|
|
|
def prepare_syntax_check(code, fname = nil, mesg = nil, verbose: nil)
|
|
fname ||= caller_locations(2, 1)[0]
|
|
mesg ||= fname.to_s
|
|
verbose, $VERBOSE = $VERBOSE, verbose
|
|
case
|
|
when Array === fname
|
|
fname, line = *fname
|
|
when defined?(fname.path) && defined?(fname.lineno)
|
|
fname, line = fname.path, fname.lineno
|
|
else
|
|
line = 1
|
|
end
|
|
yield(code, fname, line, message(mesg) {
|
|
if code.end_with?("\n")
|
|
"```\n#{code}```\n"
|
|
else
|
|
"```\n#{code}\n```\n""no-newline"
|
|
end
|
|
})
|
|
ensure
|
|
$VERBOSE = verbose
|
|
end
|
|
|
|
def assert_valid_syntax(code, *args, **opt)
|
|
prepare_syntax_check(code, *args, **opt) do |src, fname, line, mesg|
|
|
yield if defined?(yield)
|
|
assert_nothing_raised(SyntaxError, mesg) do
|
|
assert_equal(:ok, syntax_check(src, fname, line), mesg)
|
|
end
|
|
end
|
|
end
|
|
|
|
def assert_normal_exit(testsrc, message = '', child_env: nil, **opt)
|
|
assert_valid_syntax(testsrc, caller_locations(1, 1)[0])
|
|
if child_env
|
|
child_env = [child_env]
|
|
else
|
|
child_env = []
|
|
end
|
|
out, _, status = EnvUtil.invoke_ruby(child_env + %W'-W0', testsrc, true, :merge_to_stdout, **opt)
|
|
assert !status.signaled?, FailDesc[status, message, out]
|
|
end
|
|
|
|
def assert_ruby_status(args, test_stdin="", message=nil, **opt)
|
|
out, _, status = EnvUtil.invoke_ruby(args, test_stdin, true, :merge_to_stdout, **opt)
|
|
desc = FailDesc[status, message, out]
|
|
assert(!status.signaled?, desc)
|
|
message ||= "ruby exit status is not success:"
|
|
assert(status.success?, desc)
|
|
end
|
|
|
|
ABORT_SIGNALS = Signal.list.values_at(*%w"ILL ABRT BUS SEGV TERM").freeze
|
|
|
|
def separated_runner(token, out = nil)
|
|
include(*Test::Unit::TestCase.ancestors.select {|c| !c.is_a?(Class) })
|
|
|
|
out = out ? IO.new(out, 'w') : STDOUT
|
|
|
|
# avoid method redefinitions
|
|
out_write = out.method(:write)
|
|
integer_to_s = Integer.instance_method(:to_s)
|
|
array_pack = Array.instance_method(:pack)
|
|
marshal_dump = Marshal.method(:dump)
|
|
assertions_ivar_get = Test::Unit::Assertions.method(:instance_variable_get)
|
|
Test::Unit::Assertions.module_eval do
|
|
@_assertions = 0
|
|
|
|
# These need to be defined with def (not define_method with a
|
|
# Proc) so that they can be called from non-main Ractors, where
|
|
# instance variables of modules are not accessible; assertions
|
|
# made in non-main Ractors are counted separately and merged by
|
|
# assert_in_ractor instead.
|
|
undef _assertions=
|
|
def _assertions=(n)
|
|
if !defined?(Ractor) || Ractor.main?
|
|
Test::Unit::Assertions.instance_variable_set(:@_assertions, n)
|
|
end
|
|
end
|
|
|
|
undef _assertions
|
|
def _assertions
|
|
if !defined?(Ractor) || Ractor.main?
|
|
Test::Unit::Assertions.instance_variable_get(:@_assertions)
|
|
else
|
|
0
|
|
end
|
|
end
|
|
end
|
|
# assume Method#call and UnboundMethod#bind_call need to work as the original
|
|
|
|
at_exit {
|
|
assertions = assertions_ivar_get.call(:@_assertions)
|
|
out_write.call <<~OUT
|
|
<error id="#{token}" assertions=#{integer_to_s.bind(assertions).call}>
|
|
#{array_pack.bind([marshal_dump.call($!)]).call('m0')}
|
|
</error id="#{token}">
|
|
OUT
|
|
}
|
|
if defined?(Test::Unit::Runner)
|
|
Test::Unit::Runner.class_variable_set(:@@stop_auto_run, true)
|
|
elsif defined?(Test::Unit::AutoRunner)
|
|
Test::Unit::AutoRunner.need_auto_run = false
|
|
end
|
|
end
|
|
|
|
def assert_separately(args, file = nil, line = nil, src, ignore_stderr: nil, **opt)
|
|
unless file and line
|
|
loc, = caller_locations(1,1)
|
|
file ||= loc.path
|
|
line ||= loc.lineno
|
|
end
|
|
capture_stdout = true
|
|
unless /mswin|mingw/ =~ RbConfig::CONFIG['host_os']
|
|
capture_stdout = false
|
|
opt[:out] = Test::Unit::Runner.output if defined?(Test::Unit::Runner)
|
|
res_p, res_c = IO.pipe
|
|
opt[:ios] = [res_c]
|
|
end
|
|
token_dump, token_re = new_test_token
|
|
src = <<eom
|
|
# -*- coding: #{line += __LINE__; src.encoding}; -*-
|
|
BEGIN {
|
|
require "test/unit";include Test::Unit::Assertions;require #{__FILE__.dump};include Test::Unit::CoreAssertions
|
|
separated_runner #{token_dump}, #{res_c&.fileno || 'nil'}
|
|
}
|
|
#{line -= __LINE__; src}
|
|
eom
|
|
args = args.dup
|
|
# -W:no-experimental after the -w, which would otherwise turn the category
|
|
# back on. The option itself needs 2.7 or later.
|
|
warn_opts = RUBY_VERSION >= "2.7." ? ["-w", "-W:no-experimental"] : ["-w"]
|
|
args.insert((Hash === args.first ? 1 : 0), *warn_opts, "--disable=gems", *$:.map {|l| "-I#{l}"})
|
|
args << "--debug" if RUBY_ENGINE == 'jruby' # warning: tracing (e.g. set_trace_func) will not capture all events without --debug flag
|
|
# power_assert 3 requires ruby 3.1 or later
|
|
args << "-W:no-experimental" if ("2.7."..."3.1.").cover?(RUBY_VERSION)
|
|
stdout, stderr, status = EnvUtil.invoke_ruby(args, src, capture_stdout, true, **opt)
|
|
|
|
if sanitizers&.lsan_enabled?
|
|
# LSAN may output messages like the following line into stderr. We should ignore it.
|
|
# ==276855==Running thread 276851 was not suspended. False leaks are possible.
|
|
# See https://github.com/google/sanitizers/issues/1479
|
|
stderr.gsub!(/==\d+==Running thread \d+ was not suspended\. False leaks are possible\.\n/, "")
|
|
end
|
|
ensure
|
|
if res_c
|
|
res_c.close
|
|
res = res_p.read
|
|
res_p.close
|
|
else
|
|
res = stdout
|
|
end
|
|
raise if $!
|
|
abort = status.coredump? || (status.signaled? && ABORT_SIGNALS.include?(status.termsig))
|
|
marshal_error = nil
|
|
assert(!abort, FailDesc[status, nil, stderr])
|
|
res.scan(/^<error id="#{token_re}" assertions=(\d+)>\n(.*?)\n(?=<\/error id="#{token_re}">$)/m) do
|
|
self._assertions += $1.to_i
|
|
begin
|
|
res = Marshal.load($2.unpack1("m")) or next
|
|
rescue => marshal_error
|
|
ignore_stderr = nil
|
|
res = nil
|
|
else
|
|
next if SystemExit === res
|
|
if bt = res.backtrace
|
|
bt.each do |l|
|
|
l.sub!(/\A-:(\d+)/){"#{file}:#{line + $1.to_i}"}
|
|
end
|
|
bt.concat(caller)
|
|
else
|
|
res.set_backtrace(caller)
|
|
end
|
|
raise res
|
|
end
|
|
end
|
|
|
|
# really did it succeed?
|
|
unless ignore_stderr
|
|
# the body of assert_separately must not output anything to detect error
|
|
assert(stderr.empty?, FailDesc[status, "assert_separately failed with error message", stderr])
|
|
end
|
|
assert(status.success?, FailDesc[status, "assert_separately failed", stderr])
|
|
raise marshal_error if marshal_error
|
|
end
|
|
|
|
# Run Ractor-related test without influencing the main test suite
|
|
def assert_ractor(src, args: [], require: nil, require_relative: nil, file: nil, line: nil, ignore_stderr: nil, **opt)
|
|
omit unless defined?(Ractor)
|
|
|
|
# https://bugs.ruby-lang.org/issues/21262
|
|
shim_value = "class Ractor; alias value take; end" unless Ractor.method_defined?(:value)
|
|
shim_join = "class Ractor; alias join take; end" unless Ractor.method_defined?(:join)
|
|
|
|
if require
|
|
require = [require] unless require.is_a?(Array)
|
|
require = require.map {|r| "require #{r.inspect}"}.join("\n")
|
|
end
|
|
|
|
if require_relative
|
|
dir = File.dirname(caller_locations[0,1][0].absolute_path)
|
|
full_path = File.expand_path(require_relative, dir)
|
|
require = "#{require}; require #{full_path.inspect}"
|
|
end
|
|
|
|
assert_separately(args, file, line, <<~RUBY, ignore_stderr: ignore_stderr, **opt)
|
|
#{shim_value}
|
|
#{shim_join}
|
|
#{require}
|
|
previous_verbose = $VERBOSE
|
|
$VERBOSE = nil
|
|
Ractor.new {} # trigger initial warning
|
|
$VERBOSE = previous_verbose
|
|
#{src}
|
|
RUBY
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_throw( tag, failure_message = nil, &block )
|
|
#
|
|
#Fails unless the given block throws +tag+, returns the caught
|
|
#value otherwise.
|
|
#
|
|
#An optional failure message may be provided as the final argument.
|
|
#
|
|
# tag = Object.new
|
|
# assert_throw(tag, "#{tag} was not thrown!") do
|
|
# throw tag
|
|
# end
|
|
def assert_throw(tag, msg = nil)
|
|
ret = catch(tag) do
|
|
begin
|
|
yield(tag)
|
|
rescue UncaughtThrowError => e
|
|
thrown = e.tag
|
|
end
|
|
msg = message(msg) {
|
|
"Expected #{mu_pp(tag)} to have been thrown"\
|
|
"#{%Q[, not #{thrown}] if thrown}"
|
|
}
|
|
assert(false, msg)
|
|
end
|
|
assert(true)
|
|
ret
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_raise( *args, &block )
|
|
#
|
|
#Tests if the given block raises an exception. Acceptable exception
|
|
#types may be given as optional arguments. If the last argument is a
|
|
#String, it will be used as the error message.
|
|
#
|
|
# assert_raise do #Fails, no Exceptions are raised
|
|
# end
|
|
#
|
|
# assert_raise NameError do
|
|
# puts x #Raises NameError, so assertion succeeds
|
|
# end
|
|
def assert_raise(*exp, &b)
|
|
case exp.last
|
|
when String, Proc
|
|
msg = exp.pop
|
|
end
|
|
|
|
begin
|
|
yield
|
|
rescue Test::Unit::PendedError => e
|
|
return e if exp.include? Test::Unit::PendedError
|
|
raise e
|
|
rescue Exception => e
|
|
expected = exp.any? { |ex|
|
|
if ex.instance_of? Module then
|
|
e.kind_of? ex
|
|
else
|
|
e.instance_of? ex
|
|
end
|
|
}
|
|
|
|
assert expected, proc {
|
|
flunk(message(msg) {"#{mu_pp(exp)} exception expected, not #{mu_pp(e)}"})
|
|
}
|
|
|
|
return e
|
|
ensure
|
|
unless e
|
|
exp = exp.first if exp.size == 1
|
|
|
|
flunk(message(msg) {"#{mu_pp(exp)} expected but nothing was raised"})
|
|
end
|
|
end
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_raise_with_message(exception, expected, msg = nil, &block)
|
|
#
|
|
#Tests if the given block raises an exception with the expected
|
|
#message.
|
|
#
|
|
# assert_raise_with_message(RuntimeError, "foo") do
|
|
# nil #Fails, no Exceptions are raised
|
|
# end
|
|
#
|
|
# assert_raise_with_message(RuntimeError, "foo") do
|
|
# raise ArgumentError, "foo" #Fails, different Exception is raised
|
|
# end
|
|
#
|
|
# assert_raise_with_message(RuntimeError, "foo") do
|
|
# raise "bar" #Fails, RuntimeError is raised but the message differs
|
|
# end
|
|
#
|
|
# assert_raise_with_message(RuntimeError, "foo") do
|
|
# raise "foo" #Raises RuntimeError with the message, so assertion succeeds
|
|
# end
|
|
def assert_raise_with_message(exception, expected, msg = nil, &block)
|
|
case expected
|
|
when String
|
|
assert = :assert_equal
|
|
else
|
|
assert_respond_to(expected, :===)
|
|
assert = :assert_match
|
|
end
|
|
|
|
ex = assert_raise(exception, msg || proc {"Exception(#{exception}) with message matches to #{expected.inspect}"}) do
|
|
yield
|
|
end
|
|
m = ex.message
|
|
msg = message(msg, "") {"Expected Exception(#{exception}) was raised, but the message doesn't match"}
|
|
|
|
if assert == :assert_equal
|
|
assert_equal(expected, m, msg)
|
|
else
|
|
msg = message(msg) { "Expected #{mu_pp expected} to match #{mu_pp m}" }
|
|
assert expected =~ m, msg
|
|
block.binding.eval("proc{|_|$~=_}").call($~)
|
|
end
|
|
ex
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_raise_kind_of(*args, &block)
|
|
#
|
|
#Tests if the given block raises one of the given exceptions or
|
|
#sub exceptions of the given exceptions. If the last argument
|
|
#is a String, it will be used as the error message.
|
|
#
|
|
# assert_raise do #Fails, no Exceptions are raised
|
|
# end
|
|
#
|
|
# assert_raise SystemCallErr do
|
|
# Dir.chdir(__FILE__) #Raises Errno::ENOTDIR, so assertion succeeds
|
|
# end
|
|
def assert_raise_kind_of(*exp, &b)
|
|
case exp.last
|
|
when String, Proc
|
|
msg = exp.pop
|
|
end
|
|
|
|
begin
|
|
yield
|
|
rescue Test::Unit::PendedError => e
|
|
raise e unless exp.include? Test::Unit::PendedError
|
|
rescue *exp => e
|
|
pass
|
|
rescue Exception => e
|
|
flunk(message(msg) {"#{mu_pp(exp)} family exception expected, not #{mu_pp(e)}"})
|
|
ensure
|
|
unless e
|
|
exp = exp.first if exp.size == 1
|
|
|
|
flunk(message(msg) {"#{mu_pp(exp)} family expected but nothing was raised"})
|
|
end
|
|
end
|
|
e
|
|
end
|
|
|
|
TEST_DIR = File.join(__dir__, "test/unit").freeze #:nodoc:
|
|
|
|
# :call-seq:
|
|
# assert(test, [failure_message])
|
|
#
|
|
#Tests if +test+ is true.
|
|
#
|
|
#+msg+ may be a String or a Proc. If +msg+ is a String, it will be used
|
|
#as the failure message. Otherwise, the result of calling +msg+ will be
|
|
#used as the message if the assertion fails.
|
|
#
|
|
#If no +msg+ is given, a default message will be used.
|
|
#
|
|
# assert(false, "This was expected to be true")
|
|
def assert(test, *msgs)
|
|
case msg = msgs.first
|
|
when String, Proc
|
|
when nil
|
|
msgs.shift
|
|
else
|
|
bt = caller.reject { |s| s.start_with?(TEST_DIR) }
|
|
raise ArgumentError, "assertion message must be String or Proc, but #{msg.class} was given.", bt
|
|
end unless msgs.empty?
|
|
super
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_respond_to( object, method, failure_message = nil )
|
|
#
|
|
#Tests if the given Object responds to +method+.
|
|
#
|
|
#An optional failure message may be provided as the final argument.
|
|
#
|
|
# assert_respond_to("hello", :reverse) #Succeeds
|
|
# assert_respond_to("hello", :does_not_exist) #Fails
|
|
def assert_respond_to(obj, (meth, *priv), msg = nil)
|
|
unless priv.empty?
|
|
msg = message(msg) {
|
|
"Expected #{mu_pp(obj)} (#{obj.class}) to respond to ##{meth}#{" privately" if priv[0]}"
|
|
}
|
|
return assert obj.respond_to?(meth, *priv), msg
|
|
end
|
|
#get rid of overcounting
|
|
if caller_locations(1, 1)[0].path.start_with?(TEST_DIR)
|
|
return if obj.respond_to?(meth)
|
|
end
|
|
super(obj, meth, msg)
|
|
end
|
|
|
|
# :call-seq:
|
|
# assert_not_respond_to( object, method, failure_message = nil )
|
|
#
|
|
#Tests if the given Object does not respond to +method+.
|
|
#
|
|
#An optional failure message may be provided as the final argument.
|
|
#
|
|
# assert_not_respond_to("hello", :reverse) #Fails
|
|
# assert_not_respond_to("hello", :does_not_exist) #Succeeds
|
|
def assert_not_respond_to(obj, (meth, *priv), msg = nil)
|
|
unless priv.empty?
|
|
msg = message(msg) {
|
|
"Expected #{mu_pp(obj)} (#{obj.class}) to not respond to ##{meth}#{" privately" if priv[0]}"
|
|
}
|
|
return assert !obj.respond_to?(meth, *priv), msg
|
|
end
|
|
#get rid of overcounting
|
|
if caller_locations(1, 1)[0].path.start_with?(TEST_DIR)
|
|
return unless obj.respond_to?(meth)
|
|
end
|
|
refute_respond_to(obj, meth, msg)
|
|
end
|
|
|
|
# pattern_list is an array which contains regexp, string and :*.
|
|
# :* means any sequence.
|
|
#
|
|
# pattern_list is anchored.
|
|
# Use [:*, regexp/string, :*] for non-anchored match.
|
|
def assert_pattern_list(pattern_list, actual, message=nil)
|
|
rest = actual
|
|
anchored = true
|
|
pattern_list.each_with_index {|pattern, i|
|
|
if pattern == :*
|
|
anchored = false
|
|
else
|
|
if anchored
|
|
match = rest.rindex(pattern, 0)
|
|
else
|
|
match = rest.index(pattern)
|
|
end
|
|
if match
|
|
post_match = $~ ? $~.post_match : rest[match+pattern.size..-1]
|
|
else
|
|
msg = message(msg) {
|
|
expect_msg = "Expected #{mu_pp pattern}\n"
|
|
if /\n[^\n]/ =~ rest
|
|
actual_mesg = +"to match\n"
|
|
rest.scan(/.*\n+/) {
|
|
actual_mesg << ' ' << $&.inspect << "+\n"
|
|
}
|
|
actual_mesg.sub!(/\+\n\z/, '')
|
|
else
|
|
actual_mesg = "to match " + mu_pp(rest)
|
|
end
|
|
actual_mesg << "\nafter #{i} patterns with #{actual.length - rest.length} characters"
|
|
expect_msg + actual_mesg
|
|
}
|
|
assert false, msg
|
|
end
|
|
rest = post_match
|
|
anchored = true
|
|
end
|
|
}
|
|
if anchored
|
|
assert_equal("", rest)
|
|
end
|
|
end
|
|
|
|
def assert_warning(pat, msg = nil)
|
|
result = nil
|
|
stderr = EnvUtil.with_default_internal(of: pat) {
|
|
EnvUtil.verbose_warning {
|
|
result = yield
|
|
}
|
|
}
|
|
msg = message(msg) {diff pat, stderr}
|
|
assert(pat === stderr, msg)
|
|
result
|
|
end
|
|
|
|
def assert_warn(*args)
|
|
assert_warning(*args) {$VERBOSE = false; yield}
|
|
end
|
|
|
|
def assert_deprecated_warning(mesg = /deprecated/, &block)
|
|
assert_warning(mesg) do
|
|
EnvUtil.deprecation_warning(&block)
|
|
end
|
|
end
|
|
|
|
def assert_deprecated_warn(mesg = /deprecated/, &block)
|
|
assert_warn(mesg) do
|
|
EnvUtil.deprecation_warning(&block)
|
|
end
|
|
end
|
|
|
|
class << (AssertFile = Struct.new(:failure_message).new)
|
|
include Assertions
|
|
include CoreAssertions
|
|
def assert_file_predicate(predicate, *args)
|
|
if /\Anot_/ =~ predicate
|
|
predicate = $'
|
|
neg = " not"
|
|
end
|
|
result = File.__send__(predicate, *args)
|
|
result = !result if neg
|
|
mesg = "Expected file ".dup << args.shift.inspect
|
|
mesg << "#{neg} to be #{predicate}"
|
|
mesg << mu_pp(args).sub(/\A\[(.*)\]\z/m, '(\1)') unless args.empty?
|
|
mesg << " #{failure_message}" if failure_message
|
|
assert(result, mesg)
|
|
end
|
|
alias method_missing assert_file_predicate
|
|
|
|
def for(message)
|
|
clone.tap {|a| a.failure_message = message}
|
|
end
|
|
end
|
|
|
|
class AllFailures
|
|
attr_reader :failures
|
|
|
|
def initialize
|
|
@count = 0
|
|
@failures = {}
|
|
end
|
|
|
|
def for(key)
|
|
@count += 1
|
|
yield key
|
|
rescue Exception => e
|
|
@failures[key] = [@count, e]
|
|
end
|
|
|
|
def foreach(*keys)
|
|
keys.each do |key|
|
|
@count += 1
|
|
begin
|
|
yield key
|
|
rescue Exception => e
|
|
@failures[key] = [@count, e]
|
|
end
|
|
end
|
|
end
|
|
|
|
def message
|
|
i = 0
|
|
total = @count.to_s
|
|
fmt = "%#{total.size}d"
|
|
@failures.map {|k, (n, v)|
|
|
v = v.message
|
|
"\n#{i+=1}. [#{fmt%n}/#{total}] Assertion for #{k.inspect}\n#{v.b.gsub(/^/, ' | ').force_encoding(v.encoding)}"
|
|
}.join("\n")
|
|
end
|
|
|
|
def pass?
|
|
@failures.empty?
|
|
end
|
|
end
|
|
|
|
# threads should respond to shift method.
|
|
# Array can be used.
|
|
def assert_join_threads(threads, message = nil)
|
|
errs = []
|
|
values = []
|
|
while th = threads.shift
|
|
begin
|
|
values << th.value
|
|
rescue Exception
|
|
errs << [th, $!]
|
|
th = nil
|
|
end
|
|
end
|
|
values
|
|
ensure
|
|
if th&.alive?
|
|
th.raise(Timeout::Error.new)
|
|
th.join rescue errs << [th, $!]
|
|
end
|
|
if !errs.empty?
|
|
msg = "exceptions on #{errs.length} threads:\n" +
|
|
errs.map {|t, err|
|
|
"#{t.inspect}:\n" +
|
|
(err.respond_to?(:full_message) ? err.full_message(highlight: false, order: :top) : err.message)
|
|
}.join("\n---\n")
|
|
if message
|
|
msg = "#{message}\n#{msg}"
|
|
end
|
|
raise Test::Unit::AssertionFailedError, msg
|
|
end
|
|
end
|
|
|
|
def assert_all?(obj, m = nil, &blk)
|
|
failed = []
|
|
obj.each do |*a, &b|
|
|
unless blk.call(*a, &b)
|
|
failed << (a.size > 1 ? a : a[0])
|
|
end
|
|
end
|
|
assert(failed.empty?, message(m) {failed.pretty_inspect})
|
|
end
|
|
|
|
def assert_all_assertions(msg = nil)
|
|
all = AllFailures.new
|
|
yield all
|
|
ensure
|
|
assert(all.pass?, message(msg) {all.message.chomp(".")})
|
|
end
|
|
alias all_assertions assert_all_assertions
|
|
|
|
def assert_all_assertions_foreach(msg = nil, *keys, &block)
|
|
all = AllFailures.new
|
|
all.foreach(*keys, &block)
|
|
ensure
|
|
assert(all.pass?, message(msg) {all.message.chomp(".")})
|
|
end
|
|
alias all_assertions_foreach assert_all_assertions_foreach
|
|
|
|
clocks = %w[
|
|
CLOCK_THREAD_CPUTIME_ID CLOCK_PROCESS_CPUTIME_ID
|
|
CLOCK_MONOTONIC
|
|
]
|
|
# CLOCK_THREAD_CPUTIME_ID is per native thread, and an M:N thread moves
|
|
# between native threads, so it can go backwards and yield a negative
|
|
# elapsed time. Measure process CPU time there instead.
|
|
clocks.shift if RUBY_DESCRIPTION.include?("+MN")
|
|
clocks.find do |c|
|
|
if Process.const_defined?(c)
|
|
[c.to_sym, Process.const_get(c)].find do |clk|
|
|
begin
|
|
Process.clock_gettime(clk)
|
|
rescue
|
|
# Constants may be defined but not implemented, e.g., mingw.
|
|
else
|
|
unless Process.clock_getres(clk) < 1.0e-03
|
|
next # needs msec precision
|
|
end
|
|
PERFORMANCE_CLOCK = clk
|
|
end
|
|
end
|
|
end
|
|
end
|
|
|
|
# Expect +seq+ to respond to +first+ and +each+ methods, e.g.,
|
|
# Array, Range, Enumerator::ArithmeticSequence and other
|
|
# Enumerable-s, and each elements should be size factors.
|
|
#
|
|
# :yield: each elements of +seq+.
|
|
def assert_linear_performance(seq, rehearsal: nil, pre: ->(n) {n})
|
|
pend "No PERFORMANCE_CLOCK found" unless defined?(PERFORMANCE_CLOCK)
|
|
|
|
# Timeout testing generally doesn't work when RJIT compilation happens.
|
|
rjit_enabled = defined?(RubyVM::RJIT) && RubyVM::RJIT.enabled?
|
|
measure = proc do |arg, message|
|
|
st = Process.clock_gettime(PERFORMANCE_CLOCK)
|
|
yield(*arg)
|
|
t = (Process.clock_gettime(PERFORMANCE_CLOCK) - st)
|
|
assert_operator 0, :<=, t, message unless rjit_enabled
|
|
t
|
|
end
|
|
|
|
first = seq.first
|
|
*arg = pre.call(first)
|
|
raw_times = (0..(rehearsal || (2 * first))).map do
|
|
measure[arg, "rehearsal"].nonzero?
|
|
end
|
|
times = raw_times.compact
|
|
raise "all measurements are zero: #{raw_times.inspect}" if times.empty?
|
|
tmin, tmax = times.minmax
|
|
|
|
# safe_factor * tmax * rehearsal_time_variance_factor(equals to 1 when variance is small)
|
|
tbase = 10 * tmax * [(tmax / tmin) ** 2 / 4, 1].max
|
|
info = "(tmin: #{tmin}, tmax: #{tmax}, tbase: #{tbase})"
|
|
|
|
seq.each do |i|
|
|
next if i == first
|
|
t = tbase * i.fdiv(first)
|
|
*arg = pre.call(i)
|
|
message = "[#{i}]: in #{t}s #{info}"
|
|
Timeout.timeout(t, Timeout::Error, message) do
|
|
measure[arg, message]
|
|
end
|
|
end
|
|
end
|
|
|
|
def diff(exp, act)
|
|
require 'pp'
|
|
q = PP.new(+"")
|
|
q.guard_inspect_key do
|
|
q.group(2, "expected: ") do
|
|
q.pp exp
|
|
end
|
|
q.text q.newline
|
|
q.group(2, "actual: ") do
|
|
q.pp act
|
|
end
|
|
q.flush
|
|
end
|
|
q.output
|
|
end
|
|
|
|
def new_test_token
|
|
token = "\e[7;1m#{$$.to_s}:#{Time.now.strftime('%s.%L')}:#{rand(0x10000).to_s(16)}:\e[m"
|
|
return token.dump, Regexp.quote(token)
|
|
end
|
|
|
|
# Platform predicates
|
|
|
|
def self.mswin?
|
|
defined?(@mswin) ? @mswin : @mswin = RUBY_PLATFORM.include?('mswin')
|
|
end
|
|
private def mswin?
|
|
CoreAssertions.mswin?
|
|
end
|
|
|
|
def self.mingw?
|
|
defined?(@mingw) ? @mingw : @mingw = RUBY_PLATFORM.include?('mingw')
|
|
end
|
|
private def mingw?
|
|
CoreAssertions.mingw?
|
|
end
|
|
|
|
module_function def windows?
|
|
mswin? or mingw?
|
|
end
|
|
|
|
def self.version_compare(expected, actual)
|
|
expected.zip(actual).each {|e, a| z = (e <=> a); return z if z.nonzero?}
|
|
0
|
|
end
|
|
|
|
def self.version_match?(expected, actual)
|
|
if !actual
|
|
false
|
|
elsif expected.empty?
|
|
true
|
|
elsif expected.size == 1 and Range === (range = expected.first)
|
|
b, e = range.begin, range.end
|
|
return false if b and (c = version_compare(Array(b), actual)) > 0
|
|
return false if e and (c = version_compare(Array(e), actual)) < 0
|
|
return false if e and range.exclude_end? and c == 0
|
|
true
|
|
else
|
|
version_compare(expected, actual).zero?
|
|
end
|
|
end
|
|
|
|
def self.linux?(*ver)
|
|
unless defined?(@linux)
|
|
@linux = RUBY_PLATFORM.include?('linux') && `uname -r`.scan(/\d+/).map(&:to_i)
|
|
end
|
|
version_match? ver, @linux
|
|
end
|
|
private def linux?(*ver)
|
|
CoreAssertions.linux?(*ver)
|
|
end
|
|
|
|
def self.glibc?(*ver)
|
|
unless defined?(@glibc)
|
|
libc = `/usr/bin/ldd /bin/sh`[/^\s*libc.*=> *\K\S*/]
|
|
if libc and /version (\d+)\.(\d+)\.$/ =~ IO.popen([libc], &:read)[]
|
|
@glibc = [$1.to_i, $2.to_i]
|
|
else
|
|
@glibc = false
|
|
end
|
|
end
|
|
version_match? ver, @glibc
|
|
end
|
|
private def glibc?(*ver)
|
|
CoreAssertions.glibc?(*ver)
|
|
end
|
|
|
|
def self.macos?(*ver)
|
|
unless defined?(@macos)
|
|
@macos = RUBY_PLATFORM.include?('darwin') && `sw_vers -productVersion`.scan(/\d+/).map(&:to_i)
|
|
end
|
|
version_match? ver, @macos
|
|
end
|
|
private def macos?(*ver)
|
|
CoreAssertions.macos?(*ver)
|
|
end
|
|
end
|
|
end
|
|
end
|