This commit updates the way the test runner processes 'only'
tests when process-based test isolation is disabled. The
--test-only flag is no longer necessary in this scenario. The
test runner will automatically detect 'only' tests and apply the
appropriate filtering.
PR-URL: https://github.com/nodejs/node/pull/54832
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
This commit updates the way filtering is applied to tests and
suites. After this change, filters are applied just before the
test/suite is started. The results are the same, but this allows
us to eventually move away from the --test-only flag except
when process level isolation is used.
PR-URL: https://github.com/nodejs/node/pull/54832
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
This commit introduces a new --experimental-test-isolation flag
that, when set to 'none', causes the test runner to execute all
tests in the same process. By default, this is the main test
runner process, but if watch mode is enabled, it spawns a separate
process that runs all of the tests.
The default value of the new flag is 'process', which uses the
existing behavior of running each test file in its own child
process.
It is worth noting that when the isolation mode is 'none', globals
and all other top level logic (such as top level before() and after()
hooks) is shared among all files.
Co-authored-by: Moshe Atlow <moshe@atlow.co.il>
PR-URL: https://github.com/nodejs/node/pull/53927
Fixes: https://github.com/nodejs/node/issues/51548
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
This commit updates the way the test runner computes inherited
hooks. Instead of computing them when the Test/Suite is
constructed, they are now computed just prior to running the
Test/Suite. The reason is because when multiple test files are
run in the same process, it is possible for the inherited hooks
to change as more files are loaded.
PR-URL: https://github.com/nodejs/node/pull/53927
Fixes: https://github.com/nodejs/node/issues/51548
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Now that parseCommandLine() returns run() compatible arguments,
it makes sense to return setupTestReporters() as the setup()
argument to run(). This also removes another problematic use of
parseCommandLine() in setupTestReporters().
PR-URL: https://github.com/nodejs/node/pull/54353
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
This commit initializes the root harness object before the root
test and passes the harness as an option to the root test
constructor. This commit also attaches the global configuration
to the harness. This will allow the parseCommandLine() call in
test.js to be removed, as those values are now available via
the root test.
PR-URL: https://github.com/nodejs/node/pull/54353
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
The test runner bootstrap process awaits a Promise and then sets
a boolean flag. This commit consolidates the Promise and boolean
into a single value. This commit also ensures that the globalRoot
test is always assigned in createTestTree() in order to better
consolidate the CLI/run() and non-CLI configuration.
PR-URL: https://github.com/nodejs/node/pull/54013
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
The current test is large and can time out. It should be split
into multiple smaller tests as done in #54014. However, that
approach appears to change GC behavior such that the database
files are not cleaned up quickly enough on Windows. Forcing
any unfinalized SQL statements to be GC'ed appears to fix the
problem. Mark the original test as flaky until the necessary
code changes are made.
PR-URL: https://github.com/nodejs/node/pull/54031
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Jake Yuesong Li <jake.yuesong@gmail.com>
Reviewed-By: Stefan Stojanovic <stefan.stojanovic@janeasystems.com>
Reviewed-By: Richard Lau <rlau@redhat.com>
This commit refactors the internals of snapshot tests to get the
name of the test file from the test context instead of passing
it to the SnapshotManager constructor. This is prep work for
supporting running test files in the test runner process.
PR-URL: https://github.com/nodejs/node/pull/53853
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
This commit adds a filePath getter to the TestContext and
SuiteContext classes. This allows a context to be mapped back to
the original test file that created it, even if it was imported
from another file. This is useful for mapping features like test
snapshots to the correct test file. This is also prep work for
supporting running test files in the test runner process.
PR-URL: https://github.com/nodejs/node/pull/53853
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
This commit consolidates all option parsing for the test runner
in the parseCommandLine() internal helper function. The exception
is a couple of temporary flags used for feature gating which
will eventually become no-ops. This consolidation is prep work
for supporting running test files in the test runner process.
PR-URL: https://github.com/nodejs/node/pull/53849
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
This commit updates the logic that checks for the end of the
test run. Prior to this change, it was possible for root.run() to
be called multiple times because of the way pending subtests
were tracked. The extra calls to root.run() were harmless, but
could trigger an EventEmitter leak warning due to 'abort'
listeners being created.
PR-URL: https://github.com/nodejs/node/pull/52326
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
The default highWatermark of 16 on the TestsStream class can
have a substantial impact on reporting performance. This commit
sets the TestsStream highWatermark to a very large value and
lets the destination streams (which are more likely to have
meaningful highWatermarks) handle backpressure.
PR-URL: https://github.com/nodejs/node/pull/52287
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
This commit updates the test runner to suppress any output for
filtered tests. Filtered tests no longer generate reporter events,
and the unfiltered tests are renumbered in the output as though
the filtered tests were not present. Skipped tests that are not
filtered are still included in the output.
This change is particularly useful when filtering a large number
of tests, as the previously displayed skip output could be
distracting.
Fixes: https://github.com/nodejs/node/issues/51383
PR-URL: https://github.com/nodejs/node/pull/52221
Reviewed-By: Yagiz Nizipli <yagiz.nizipli@sentry.io>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
This commit simplifies the logic for tracking test start time.
The start time is now set only when a test/suite begins running.
If the test/suite never runs, a fallback is provided in postRun().
PR-URL: https://github.com/nodejs/node/pull/52182
Reviewed-By: Yagiz Nizipli <yagiz.nizipli@sentry.io>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
This commit removes a check for the todo flag when determining
if a suite should run. In general, the todo flag should have
no impact on whether or not a test/suite runs. Instead, it should
only impact how the result of the test/suite is handled.
PR-URL: https://github.com/nodejs/node/pull/52117
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
This commit updates the test harness to prevent top level
tests from executing immediately. This allows certain config
data, such as filtering options, to be discovered before running
the tests.
PR-URL: https://github.com/nodejs/node/pull/52092
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
This commit removes a redundant call to `reportStarted()`. It is
redundant because a few lines down, `subtest.finalize()` is
called. `finalize()` will find the first test that is ready to
report its data, and then call `report()`, which also calls
`reportStarted()`. This will trigger the `'test:start'` as high
up the test tree as necessary.
PR-URL: https://github.com/nodejs/node/pull/52089
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Marco Ippolito <marcoippolito54@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Raz Luvaton <rluvaton@gmail.com>
This commit updates the built in reporters to check for the
documented case of a test's location being undefined.
As a drive by fix, the C++ code for computing the test location
now returns undefined if the script location is empty. This lets
tests run inside of eval().
PR-URL: https://github.com/nodejs/node/pull/52036
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
This test was flaky on Windows when trying to clean up the
tmp directory, probably because it relied on child processes
timing out and being killed.
This commit updates the test to check for debug output
from the test runner. This should be adequate because the
original change was effectively:
let concurrency = getOptionValue('--test-concurrency') || true;
The test runner now logs the value of the concurrency variable.
Fixes: https://github.com/nodejs/node/issues/50101
PR-URL: https://github.com/nodejs/node/pull/50108
Reviewed-By: Antoine du Hamel <duhamelantoine1995@gmail.com>
Reviewed-By: Geoffrey Booth <webadmin@geoffreybooth.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
This commit adds each test's line and column number to the reporter
output. This will aid in debugging test suite failures when error
stacks are not helpful, test suites are large, or tests have the
same name. This data is also exposed on the spec reporter.
This commit also replaces the filename that was previously being
reported, with the filename where the test actually exists. These
are normally correct, but could be wrong if tests were run from
a file other than the user's entrypoint.
PR-URL: https://github.com/nodejs/node/pull/48975
Fixes: https://github.com/nodejs/node/issues/48457
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Chemi Atlow <chemi@atlow.co.il>
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
The test test/parallel/test-experimental-shared-value-conveyor.js was
added to test the --harmony-struct feature of V8. However, when used
with pointer compression, the process crashes. This commit skips
the test for pointer compression builds. This change uses a child
process because starting a Node pointer compression build with
--harmony-struct immediately crashes the process. Once this crash
is addresses, this commit can be reverted.
PR-URL: https://github.com/nodejs/node/pull/48738
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Michael Dawson <midawson@redhat.com>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Debadree Chatterjee <debadree333@gmail.com>
Reviewed-By: Luigi Pinca <luigipinca@gmail.com>
Reviewed-By: Richard Lau <rlau@redhat.com>
This commit moves code coverage collection from the test
harness exit handler to the postRun() function of the root
test.
This is necessary preparatory work for supporting
code coverage with --test. The reason is that --test is
implemented on top of run(), and that function calls the root
test's postRun() function, which outputs the test summary. This
happens before the harness exit handler.
PR-URL: https://github.com/nodejs/node/pull/47651
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Matteo Collina <matteo.collina@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
This commit addresses a previously untested branch of the code.
It is possible when using the test runner that an error occurs
outside of a test. In this case, the test runner would simply
rethrow the error. This commit updates the logic to handle the
error in the same fashion as other uncaughtExceptions.
PR-URL: https://github.com/nodejs/node/pull/46962
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
This commit updates the test harness to re-throw uncaught errors
if bootstrapping has not completed. This updates the existing
logic which tried to detect a specific error code.
PR-URL: https://github.com/nodejs/node/pull/46962
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
This commit replaces the 'coverage' reference inside of the Test
class with a more generic harness reference which includes
coverage. This will let the root test more easily track process
level state such as code coverage, uncaughtException handlers,
and the state of bootstrapping.
PR-URL: https://github.com/nodejs/node/pull/46962
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: Yagiz Nizipli <yagiz@nizipli.com>
The test runner is bootstrapped synchronously, with the exception
of importing custom reporters. To better handle asynchronously
throw errors, this commit introduces an internal error type that
can be checked for from the test runner's uncaughtException
handler.
After https://github.com/nodejs/node/pull/46707 and this change
land, the other throw statement in the uncaughtException handler
can be removed. This will allow the test runner to handle errors
thrown from outside of tests (which currently prevents the test
runner from reporting results).
PR-URL: https://github.com/nodejs/node/pull/46720
Reviewed-By: Moshe Atlow <moshe@atlow.co.il>
Reviewed-By: James M Snell <jasnell@gmail.com>
Reviewed-By: Benjamin Gruenbaum <benjamingr@gmail.com>
Colin Ihrig