a65e4418e5
PR-URL: https://github.com/nodejs/node/pull/54947 Reviewed-By: Luigi Pinca <luigipinca@gmail.com> Reviewed-By: James M Snell <jasnell@gmail.com> |
||
---|---|---|
.. | ||
wpt | ||
arraystream.js | ||
assertSnapshot.js | ||
benchmark.js | ||
child_process.js | ||
countdown.js | ||
cpu-prof.js | ||
crypto.js | ||
debugger.js | ||
dns.js | ||
fixtures.js | ||
fixtures.mjs | ||
gc.js | ||
globals.js | ||
heap.js | ||
hijackstdio.js | ||
http2.js | ||
index.js | ||
index.mjs | ||
inspector-helper.js | ||
internet.js | ||
measure-memory.js | ||
package.json | ||
process-exit-code-cases.js | ||
prof.js | ||
README.md | ||
report.js | ||
require-as.js | ||
sea.js | ||
shared-lib-util.js | ||
snapshot.js | ||
tick.js | ||
tls.js | ||
tmpdir.js | ||
v8.js | ||
wasi.js | ||
wpt.js |
Node.js Core Test Common Modules
This directory contains modules used to test the Node.js implementation.
Table of contents
- ArrayStream module
- Benchmark module
- Child process module
- Common module API
- Countdown module
- CPU Profiler module
- Debugger module
- DNS module
- Environment variables
- Fixtures module
- Heap dump checker module
- hijackstdio module
- HTTP2 module
- Internet module
- ongc module
- process-exit-code-test-cases module
- Report module
- tick module
- tmpdir module
- UDP pair helper
- WPT module
Benchmark module
The benchmark
module is used by tests to run benchmarks.
runBenchmark(name, env)
name
<string> Name of benchmark suite to be run.env
<Object> Environment variables to be applied during the run.
Child Process module
The child_process
module is used by tests that launch child processes.
spawnSyncAndExit(command[, args][, spawnOptions], expectations)
Spawns a child process synchronously using child_process.spawnSync()
and
check if it runs in the way expected. If it does not, print the stdout and
stderr output from the child process and additional information about it to
the stderr of the current process before throwing and error. This helps
gathering more information about test failures coming from child processes.
command
,args
,spawnOptions
Seechild_process.spawnSync()
expectations
<Object>status
<number> Expectedchild.status
signal
<string> |null
Expectedchild.signal
stderr
<string> | <RegExp> | <Function> Optional. If it's a string, check that the output to the stderr of the child process is exactly the same as the string. If it's a regular expression, check that the stderr matches it. If it's a function, invoke it with the stderr output as a string and check that it returns true. The function can just throw errors (e.g. assertion errors) to provide more information if the check fails.stdout
<string> | <RegExp> | <Function> Optional. Similar tostderr
but for the stdout.trim
<boolean> Optional. Whether this method should trim out the whitespace characters when checkingstderr
andstdout
outputs. Defaults tofalse
.
- return <Object>
child
<ChildProcess> The child process returned bychild_process.spawnSync()
.stderr
<string> The output from the child process to stderr.stdout
<string> The output from the child process to stdout.
spawnSyncAndExitWithoutError(command[, args][, spawnOptions])
Similar to expectSyncExit()
with the status
expected to be 0 and
signal
expected to be null
.
spawnSyncAndAssert(command[, args][, spawnOptions], expectations)
Similar to spawnSyncAndExitWithoutError()
, but with an additional
expectations
parameter.
Common module API
The common
module is used by tests for consistency across repeated
tasks.
allowGlobals(...allowlist)
Takes allowlist
and concats that with predefined knownGlobals
.
canCreateSymLink()
- return <boolean>
Checks whether the current running process can create symlinks. On Windows, this
returns false
if the process running doesn't have privileges to create
symlinks
(SeCreateSymbolicLinkPrivilege).
On non-Windows platforms, this always returns true
.
createZeroFilledFile(filename)
Creates a 10 MiB file of all null characters.
enoughTestMem
Indicates if there is more than 1gb of total memory.
expectsError(validator[, exact])
validator
<Object> | <RegExp> | <Function> | <Error> The validator behaves identical toassert.throws(fn, validator)
.exact
<number> default = 1- return <Function> A callback function that expects an error.
A function suitable as callback to validate callback based errors. The error is
validated using assert.throws(() => { throw error; }, validator)
. If the
returned function has not been called exactly exact
number of times when the
test is complete, then the test will fail.
expectWarning(name[, expected[, code]])
Tests whether name
, expected
, and code
are part of a raised warning.
The code is required in case the name is set to 'DeprecationWarning'
.
Examples:
const { expectWarning } = require('../common');
expectWarning('Warning', 'Foobar is really bad');
expectWarning('DeprecationWarning', 'Foobar is deprecated', 'DEP0XXX');
expectWarning('DeprecationWarning', [
'Foobar is deprecated', 'DEP0XXX',
]);
expectWarning('DeprecationWarning', [
['Foobar is deprecated', 'DEP0XXX'],
['Baz is also deprecated', 'DEP0XX2'],
]);
expectWarning('DeprecationWarning', {
DEP0XXX: 'Foobar is deprecated',
DEP0XX2: 'Baz is also deprecated',
});
expectWarning({
DeprecationWarning: {
DEP0XXX: 'Foobar is deprecated',
DEP0XX1: 'Baz is also deprecated',
},
Warning: [
['Multiple array entries are fine', 'SpecialWarningCode'],
['No code is also fine'],
],
SingleEntry: ['This will also work', 'WarningCode'],
SingleString: 'Single string entries without code will also work',
});
getArrayBufferViews(buf)
buf
<Buffer>- return <ArrayBufferView>[]
Returns an instance of all possible ArrayBufferView
s of the provided Buffer.
getBufferSources(buf)
buf
<Buffer>- return <BufferSource>[]
Returns an instance of all possible BufferSource
s of the provided Buffer,
consisting of all ArrayBufferView
and an ArrayBuffer
.
getTTYfd()
Attempts to get a valid TTY file descriptor. Returns -1
if it fails.
The TTY file descriptor is assumed to be capable of being writable.
hasCrypto
Indicates whether OpenSSL is available.
hasFipsCrypto
Indicates that Node.js has been linked with a FIPS compatible OpenSSL library,
and that FIPS as been enabled using --enable-fips
.
To only detect if the OpenSSL library is FIPS compatible, regardless if it has
been enabled or not, then process.config.variables.openssl_is_fips
can be
used to determine that situation.
hasIntl
Indicates if internationalization is supported.
hasIPv6
Indicates whether IPv6
is supported on this platform.
hasMultiLocalhost
Indicates if there are multiple localhosts available.
inFreeBSDJail
Checks whether free BSD Jail is true or false.
isAIX
Platform check for Advanced Interactive eXecutive (AIX).
isAlive(pid)
Attempts to 'kill' pid
isDumbTerminal
isFreeBSD
Platform check for Free BSD.
isIBMi
Platform check for IBMi.
isLinux
Platform check for Linux.
isLinuxPPCBE
Platform check for Linux on PowerPC.
isMacOS
Platform check for macOS.
isSunOS
Platform check for SunOS.
isWindows
Platform check for Windows.
localhostIPv4
IP of localhost
.
localIPv6Hosts
Array of IPV6 representations for localhost
.
mustCall([fn][, exact])
fn
<Function> default = () => {}exact
<number> default = 1- return <Function>
Returns a function that calls fn
. If the returned function has not been called
exactly exact
number of times when the test is complete, then the test will
fail.
If fn
is not provided, an empty function will be used.
mustCallAtLeast([fn][, minimum])
fn
<Function> default = () => {}minimum
<number> default = 1- return <Function>
Returns a function that calls fn
. If the returned function has not been called
at least minimum
number of times when the test is complete, then the test will
fail.
If fn
is not provided, an empty function will be used.
mustNotCall([msg])
msg
<string> default = 'function should not have been called'- return <Function>
Returns a function that triggers an AssertionError
if it is invoked. msg
is
used as the error message for the AssertionError
.
mustNotMutateObjectDeep([target])
If target
is an Object, returns a proxy object that triggers
an AssertionError
on mutation attempt, including mutation of deeply nested
Objects. Otherwise, it returns target
directly.
Use of this function is encouraged for relevant regression tests.
import { open } from 'node:fs/promises';
import { mustNotMutateObjectDeep } from '../common/index.mjs';
const _mutableOptions = { length: 4, position: 8 };
const options = mustNotMutateObjectDeep(_mutableOptions);
// In filehandle.read or filehandle.write, attempt to mutate options will throw
// In the test code, options can still be mutated via _mutableOptions
const fh = await open('/path/to/file', 'r+');
const { buffer } = await fh.read(options);
_mutableOptions.position = 4;
await fh.write(buffer, options);
// Inline usage
const stats = await fh.stat(mustNotMutateObjectDeep({ bigint: true }));
console.log(stats.size);
Caveats: built-in objects that make use of their internal slots (for example,
Map
s and Set
s) might not work with this function. It returns Functions
directly, not preventing their mutation.
mustSucceed([fn])
fn
<Function> default = () => {}- return <Function>
Returns a function that accepts arguments (err, ...args)
. If err
is not
undefined
or null
, it triggers an AssertionError
. Otherwise, it calls
fn(...args)
.
nodeProcessAborted(exitCode, signal)
Returns true
if the exit code exitCode
and/or signal name signal
represent
the exit code and/or signal name of a node process that aborted, false
otherwise.
opensslCli
Indicates whether 'opensslCli' is supported.
platformTimeout(ms)
Returns a timeout value based on detected conditions. For example, a debug build may need extra time so the returned value will be larger than on a release build.
PIPE
Path to the test socket.
PORT
A port number for tests to use if one is needed.
printSkipMessage(msg)
msg
<string>
Logs '1..0 # Skipped: ' + msg
pwdCommand
- <array> First two argument for the
spawn
/exec
functions.
Platform normalized pwd
command options. Usage example:
const common = require('../common');
const { spawn } = require('node:child_process');
spawn(...common.pwdCommand, { stdio: ['pipe'] });
requireNoPackageJSONAbove([dir])
dir
<string> default = __dirname
Throws an AssertionError
if a package.json
file exists in any ancestor
directory above dir
. Such files may interfere with proper test functionality.
runWithInvalidFD(func)
func
<Function>
Runs func
with an invalid file descriptor that is an unsigned integer and
can be used to trigger EBADF
as the first argument. If no such file
descriptor could be generated, a skip message will be printed and the func
will not be run.
skip(msg)
msg
<string>
Logs '1..0 # Skipped: ' + msg
and exits with exit code 0
.
skipIfDumbTerminal()
Skip the rest of the tests if the current terminal is a dumb terminal
skipIfEslintMissing()
Skip the rest of the tests in the current file when ESLint
is not available
at tools/eslint/node_modules/eslint
skipIfInspectorDisabled()
Skip the rest of the tests in the current file when the Inspector was disabled at compile time.
skipIf32Bits()
Skip the rest of the tests in the current file when the Node.js executable was compiled with a pointer size smaller than 64 bits.
skipIfWorker()
Skip the rest of the tests in the current file when not running on a main thread.
ArrayStream module
The ArrayStream
module provides a simple Stream
that pushes elements from
a given array.
const ArrayStream = require('../common/arraystream');
const stream = new ArrayStream();
stream.run(['a', 'b', 'c']);
It can be used within tests as a simple mock stream.
Countdown module
The Countdown
module provides a simple countdown mechanism for tests that
require a particular action to be taken after a given number of completed
tasks (for instance, shutting down an HTTP server after a specific number of
requests). The Countdown will fail the test if the remainder did not reach 0.
const Countdown = require('../common/countdown');
function doSomething() {
console.log('.');
}
const countdown = new Countdown(2, doSomething);
countdown.dec();
countdown.dec();
new Countdown(limit, callback)
limit
{number}callback
{function}
Creates a new Countdown
instance.
Countdown.prototype.dec()
Decrements the Countdown
counter.
Countdown.prototype.remaining
Specifies the remaining number of times Countdown.prototype.dec()
must be
called before the callback is invoked.
CPU Profiler module
The cpu-prof
module provides utilities related to CPU profiling tests.
env
- Default: { ...process.env, NODE_DEBUG_NATIVE: 'INSPECTOR_PROFILER' }
Environment variables used in profiled processes.
getCpuProfiles(dir)
dir
{string} The directory containing the CPU profile files.- return <string>
Returns an array of all .cpuprofile
files found in dir
.
getFrames(file, suffix)
file
{string} Path to a.cpuprofile
file.suffix
{string} Suffix of the URL of call frames to retrieve.- returns { frames: <Object>, nodes: <Object> }
Returns an object containing an array of the relevant call frames and an array of all the profile nodes.
kCpuProfInterval
Sampling interval in microseconds.
verifyFrames(output, file, suffix)
output
{string}file
{string}suffix
{string}
Throws an AssertionError
if there are no call frames with the expected
suffix
in the profiling data contained in file
.
Debugger module
Provides common functionality for tests for node inspect
.
startCLI(args[[, flags], spawnOpts])
Returns a null-prototype object with properties that are functions and getters
used to interact with the node inspect
CLI. These functions are:
flushOutput()
waitFor()
waitForPrompt()
waitForInitialBreak()
breakInfo
ctrlC()
output
rawOutput
parseSourceLines()
writeLine()
command()
stepCommand()
quit()
DNS
module
The DNS
module provides utilities related to the dns
built-in module.
errorLookupMock(code, syscall)
code
<string> Defaults todns.mockedErrorCode
.syscall
<string> Defaults todns.mockedSysCall
.- return <Function>
A mock for the lookup
option of net.connect()
that would result in an error
with the code
and the syscall
specified. Returns a function that has the
same signature as dns.lookup()
.
mockedErrorCode
The default code
of errors generated by errorLookupMock
.
mockedSysCall
The default syscall
of errors generated by errorLookupMock
.
readDomainFromPacket(buffer, offset)
Reads the domain string from a packet and returns an object containing the number of bytes read and the domain.
parseDNSPacket(buffer)
Parses a DNS packet. Returns an object with the values of the various flags of the packet depending on the type of packet.
writeIPv6(ip)
Reads an IPv6 String and returns a Buffer containing the parts.
writeDomainName(domain)
Reads a Domain String and returns a Buffer containing the domain.
writeDNSPacket(parsed)
Takes in a parsed Object and writes its fields to a DNS packet as a Buffer object.
Environment variables
The behavior of the Node.js test suite can be altered using the following environment variables.
NODE_COMMON_PORT
If set, NODE_COMMON_PORT
's value overrides the common.PORT
default value of
12346.
NODE_REGENERATE_SNAPSHOTS
If set, test snapshots for a the current test are regenerated.
for example NODE_REGENERATE_SNAPSHOTS=1 out/Release/node test/parallel/test-runner-output.mjs
will update all the test runner output snapshots.
NODE_SKIP_FLAG_CHECK
If set, command line arguments passed to individual tests are not validated.
NODE_SKIP_CRYPTO
If set, crypto tests are skipped.
NODE_TEST_KNOWN_GLOBALS
A comma-separated list of variables names that are appended to the global
variable allowlist. Alternatively, if NODE_TEST_KNOWN_GLOBALS
is set to '0'
,
global leak detection is disabled.
Fixtures module
The common/fixtures
module provides convenience methods for working with
files in the test/fixtures
directory.
fixtures.fixturesDir
The absolute path to the test/fixtures/
directory.
fixtures.path(...args)
...args
<string>
Returns the result of path.join(fixtures.fixturesDir, ...args)
.
fixtures.fileURL(...args)
...args
<string>
Returns the result of url.pathToFileURL(fixtures.path(...args))
.
fixtures.readSync(args[, enc])
Returns the result of
fs.readFileSync(path.join(fixtures.fixturesDir, ...args), 'enc')
.
fixtures.readKey(arg[, enc])
arg
<string>
Returns the result of
fs.readFileSync(path.join(fixtures.fixturesDir, 'keys', arg), 'enc')
.
Heap dump checker module
This provides utilities for checking the validity of heap dumps.
This requires the usage of --expose-internals
.
heap.recordState()
Create a heap dump and an embedder graph copy for inspection.
The returned object has a validateSnapshotNodes
function similar to the
one listed below. (heap.validateSnapshotNodes(...)
is a shortcut for
heap.recordState().validateSnapshotNodes(...)
.)
heap.validateSnapshotNodes(name, expected, options)
name
<string> Look for this string as the name of heap dump nodes.expected
<Array> A list of objects, possibly with anchildren
property that points to expected other adjacent nodes.options
<Array>loose
<boolean> Do not expect an exact listing of occurrences of nodes with namename
inexpected
.
Create a heap dump and an embedder graph copy and validate occurrences.
validateSnapshotNodes('TLSWRAP', [
{
children: [
{ name: 'enc_out' },
{ name: 'enc_in' },
{ name: 'TLSWrap' },
],
},
]);
hijackstdio module
The hijackstdio
module provides utility functions for temporarily redirecting
stdout
and stderr
output.
const { hijackStdout, restoreStdout } = require('../common/hijackstdio');
hijackStdout((data) => {
/* Do something with data */
restoreStdout();
});
console.log('this is sent to the hijacked listener');
hijackStderr(listener)
listener
<Function>: a listener with a single parameter calleddata
.
Eavesdrop to process.stderr.write()
calls. Once process.stderr.write()
is
called, listener
will also be called and the data
of write
function will
be passed to listener
. What's more, process.stderr.writeTimes
is a count of
the number of calls.
hijackStdout(listener)
listener
<Function>: a listener with a single parameter calleddata
.
Eavesdrop to process.stdout.write()
calls. Once process.stdout.write()
is
called, listener
will also be called and the data
of write
function will
be passed to listener
. What's more, process.stdout.writeTimes
is a count of
the number of calls.
restoreStderr()
Restore the original process.stderr.write()
. Used to restore stderr
to its
original state after calling hijackstdio.hijackStdErr()
.
restoreStdout()
Restore the original process.stdout.write()
. Used to restore stdout
to its
original state after calling hijackstdio.hijackStdOut()
.
HTTP/2 module
The http2.js module provides a handful of utilities for creating mock HTTP/2 frames for testing of HTTP/2 endpoints
const http2 = require('../common/http2');
Class: Frame
The http2.Frame
is a base class that creates a Buffer
containing a
serialized HTTP/2 frame header.
// length is a 24-bit unsigned integer
// type is an 8-bit unsigned integer identifying the frame type
// flags is an 8-bit unsigned integer containing the flag bits
// id is the 32-bit stream identifier, if any.
const frame = new http2.Frame(length, type, flags, id);
// Write the frame data to a socket
socket.write(frame.data);
The serialized Buffer
may be retrieved using the frame.data
property.
Class: HeadersFrame
The http2.HeadersFrame
is a subclass of http2.Frame
that serializes a
HEADERS
frame.
// id is the 32-bit stream identifier
// payload is a Buffer containing the HEADERS payload (see either
// http2.kFakeRequestHeaders or http2.kFakeResponseHeaders).
// padlen is an 8-bit integer giving the number of padding bytes to include
// final is a boolean indicating whether the End-of-stream flag should be set,
// defaults to false.
const frame = new http2.HeadersFrame(id, payload, padlen, final);
socket.write(frame.data);
Class: SettingsFrame
The http2.SettingsFrame
is a subclass of http2.Frame
that serializes an
empty SETTINGS
frame.
// ack is a boolean indicating whether or not to set the ACK flag.
const frame = new http2.SettingsFrame(ack);
socket.write(frame.data);
http2.kFakeRequestHeaders
Set to a Buffer
instance that contains a minimal set of serialized HTTP/2
request headers to be used as the payload of a http2.HeadersFrame
.
const frame = new http2.HeadersFrame(1, http2.kFakeRequestHeaders, 0, true);
socket.write(frame.data);
http2.kFakeResponseHeaders
Set to a Buffer
instance that contains a minimal set of serialized HTTP/2
response headers to be used as the payload a http2.HeadersFrame
.
const frame = new http2.HeadersFrame(1, http2.kFakeResponseHeaders, 0, true);
socket.write(frame.data);
http2.kClientMagic
Set to a Buffer
containing the preamble bytes an HTTP/2 client must send
upon initial establishment of a connection.
socket.write(http2.kClientMagic);
Internet module
The common/internet
module provides utilities for working with
internet-related tests.
internet.addresses
- <Object>
INET_HOST
<string> A generic host that has registered common DNS records, supports both IPv4 and IPv6, and provides basic HTTP/HTTPS servicesINET4_HOST
<string> A host that provides IPv4 servicesINET6_HOST
<string> A host that provides IPv6 servicesINET4_IP
<string> An accessible IPv4 IP, defaults to the Google Public DNS IPv4 addressINET6_IP
<string> An accessible IPv6 IP, defaults to the Google Public DNS IPv6 addressINVALID_HOST
<string> An invalid host that cannot be resolvedMX_HOST
<string> A host with MX records registeredSRV_HOST
<string> A host with SRV records registeredPTR_HOST
<string> A host with PTR records registeredNAPTR_HOST
<string> A host with NAPTR records registeredSOA_HOST
<string> A host with SOA records registeredCNAME_HOST
<string> A host with CNAME records registeredNS_HOST
<string> A host with NS records registeredTXT_HOST
<string> A host with TXT records registeredDNS4_SERVER
<string> An accessible IPv4 DNS serverDNS6_SERVER
<string> An accessible IPv6 DNS server
A set of addresses for internet-related tests. All properties are configurable
via NODE_TEST_*
environment variables. For example, to configure
internet.addresses.INET_HOST
, set the environment
variable NODE_TEST_INET_HOST
to a specified host.
ongc module
The ongc
module allows a garbage collection listener to be installed. The
module exports a single onGC()
function.
require('../common');
const { onGC } = require('../common/gc');
onGC({}, { ongc() { console.log('collected'); } });
onGC(target, listener)
target
<Object>listener
<Object>ongc
<Function>
Installs a GC listener for the collection of target
.
This uses async_hooks
for GC tracking. This means that it enables
async_hooks
tracking, which may affect the test functionality. It also
means that between a global.gc()
call and the listener being invoked
a full setImmediate()
invocation passes.
listener
is an object to make it easier to use a closure; the target object
should not be in scope when listener.ongc()
is created.
process-exit-code-test-cases module
The process-exit-code-test-cases
module provides a set of shared test cases
for testing the exit codes of the process
object. The test cases are shared
between test/parallel/test-process-exit-code.js
and
test/parallel/test-worker-exit-code.js
.
getTestCases(isWorker)
Returns an array of test cases for testing the exit codes of the process
. Each
test case is an object with a func
property that is a function that runs the
test case, a result
property that is the expected exit code, and sometimes an
error
property that is a regular expression that the error message should
match when the test case is run in a worker thread.
The isWorker
parameter is used to adjust the test cases for worker threads.
The default value is false
.
Report module
The report
module provides helper functions for testing diagnostic reporting
functionality.
findReports(pid, dir)
pid
<number> Process ID to retrieve diagnostic report files for.dir
<string> Directory to search for diagnostic report files.- return <Array>
Returns an array of diagnostic report file names found in dir
. The files
should have been generated by a process whose PID matches pid
.
validate(filepath)
filepath
<string> Diagnostic report filepath to validate.
Validates the schema of a diagnostic report file whose path is specified in
filepath
. If the report fails validation, an exception is thrown.
validateContent(report)
report
<Object> | <string> JSON contents of a diagnostic report file, the parsed Object thereof, or the result ofprocess.report.getReport()
.
Validates the schema of a diagnostic report whose content is specified in
report
. If the report fails validation, an exception is thrown.
SEA Module
The sea
module provides helper functions for testing Single Executable
Application functionality.
skipIfSingleExecutableIsNotSupported()
Skip the rest of the tests if single executable applications are not supported in the current configuration.
generateSEA(targetExecutable, sourceExecutable, seaBlob, verifyWorkflow)
Copy sourceExecutable
to targetExecutable
, use postject to inject seaBlob
into targetExecutable
and sign it if necessary.
If verifyWorkflow
is false (default) and any of the steps fails,
it skips the tests. Otherwise, an error is thrown.
tick module
The tick
module provides a helper function that can be used to call a callback
after a given number of event loop "ticks".
tick(x, cb)
x
<number> Number of event loop "ticks".cb
<Function> A callback function.
tmpdir module
The tmpdir
module supports the use of a temporary directory for testing.
path
The realpath of the testing temporary directory.
fileURL([...paths])
Resolves a sequence of paths into absolute url in the temporary directory.
When called without arguments, returns absolute url of the testing
temporary directory with explicit trailing /
.
refresh(useSpawn)
useSpawn
<boolean> default = false
Deletes and recreates the testing temporary directory. When useSpawn
is true
this action is performed using child_process.spawnSync
.
The first time refresh()
runs, it adds a listener to process 'exit'
that
cleans the temporary directory. Thus, every file under tmpdir.path
needs to
be closed before the test completes. A good way to do this is to add a
listener to process 'beforeExit'
. If a file needs to be left open until
Node.js completes, use a child process and call refresh()
only in the
parent.
It is usually only necessary to call refresh()
once in a test file.
Avoid calling it more than once in an asynchronous context as one call
might refresh the temporary directory of a different context, causing
the test to fail somewhat mysteriously.
resolve([...paths])
Resolves a sequence of paths into absolute path in the temporary directory.
hasEnoughSpace(size)
size
<number> Required size, in bytes.
Returns true
if the available blocks of the file system underlying path
are likely sufficient to hold a single file of size
bytes. This is useful for
skipping tests that require hundreds of megabytes or even gigabytes of temporary
files, but it is inaccurate and susceptible to race conditions.
WPT module
harness
A legacy port of Web Platform Tests harness.
See the source code for definitions. Please avoid using it in new code - the current usage of this port in tests is being migrated to the original WPT harness, see the WPT tests README.
Class: WPTRunner
A driver class for running WPT with the WPT harness in a worker thread.
See the WPT tests README for details.