setting up GDUnit
Some checks failed
Create tag and build when new code gets to main / Export (push) Failing after 3m40s

This commit is contained in:
2026-01-25 18:19:26 +01:00
parent 39d6ab1c5f
commit c28d97de2d
471 changed files with 29716 additions and 16 deletions

View File

@@ -0,0 +1,269 @@
## The execution context
## It contains all the necessary information about the executed stage, such as memory observers, reports, orphan monitor
class_name GdUnitExecutionContext
enum GC_ORPHANS_CHECK {
NONE,
SUITE_HOOK_AFTER,
TEST_HOOK_AFTER,
TEST_CASE
}
var _parent_context: GdUnitExecutionContext
var _sub_context: Array[GdUnitExecutionContext] = []
var _orphan_monitor: GdUnitOrphanNodesMonitor
var _memory_observer: GdUnitMemoryObserver
var _report_collector: GdUnitTestReportCollector
var _timer: LocalTime
var _test_case_name: StringName
var _test_case_parameter_set: Array
var _name: String
var _test_execution_iteration: int = 0
var _flaky_test_check := GdUnitSettings.is_test_flaky_check_enabled()
var _flaky_test_retries := GdUnitSettings.get_flaky_max_retries()
var _orphans := -1
var error_monitor: GodotGdErrorMonitor = null:
get:
if _parent_context != null:
return _parent_context.error_monitor
if error_monitor == null:
error_monitor = GodotGdErrorMonitor.new()
return error_monitor
var test_suite: GdUnitTestSuite = null:
get:
if _parent_context != null:
return _parent_context.test_suite
return test_suite
var test_case: _TestCase = null:
get:
if test_case == null and _parent_context != null:
return _parent_context.test_case
return test_case
func _init(name: StringName, parent_context: GdUnitExecutionContext = null) -> void:
_name = name
_parent_context = parent_context
_timer = LocalTime.now()
_orphan_monitor = GdUnitOrphanNodesMonitor.new(name)
_orphan_monitor.start()
_memory_observer = GdUnitMemoryObserver.new()
_report_collector = GdUnitTestReportCollector.new()
if parent_context != null:
parent_context._sub_context.append(self)
func dispose() -> void:
_timer = null
_orphan_monitor = null
_report_collector = null
_memory_observer = null
_parent_context = null
test_suite = null
test_case = null
dispose_sub_contexts()
func dispose_sub_contexts() -> void:
for context in _sub_context:
context.dispose()
_sub_context.clear()
static func of(pe: GdUnitExecutionContext) -> GdUnitExecutionContext:
var context := GdUnitExecutionContext.new(pe._test_case_name, pe)
context._test_case_name = pe._test_case_name
context._test_execution_iteration = pe._test_execution_iteration
return context
static func of_test_suite(p_test_suite: GdUnitTestSuite) -> GdUnitExecutionContext:
assert(p_test_suite, "test_suite is null")
var context := GdUnitExecutionContext.new(p_test_suite.get_name())
context.test_suite = p_test_suite
return context
static func of_test_case(pe: GdUnitExecutionContext, p_test_case: _TestCase) -> GdUnitExecutionContext:
assert(p_test_case, "test_case is null")
var context := GdUnitExecutionContext.new(p_test_case.get_name(), pe)
context.test_case = p_test_case
return context
static func of_parameterized_test(pe: GdUnitExecutionContext, test_case_name: String, test_case_parameter_set: Array) -> GdUnitExecutionContext:
var context := GdUnitExecutionContext.new(test_case_name, pe)
context._test_case_name = test_case_name
context._test_case_parameter_set = test_case_parameter_set
return context
func get_test_suite_path() -> String:
return test_suite.get_script().resource_path
func get_test_suite_name() -> StringName:
return test_suite.get_name()
func get_test_case_name() -> StringName:
if _test_case_name.is_empty():
return test_case._test_case.display_name
return _test_case_name
func error_monitor_start() -> void:
error_monitor.start()
func error_monitor_stop() -> void:
await error_monitor.scan()
for error_report in error_monitor.to_reports():
if error_report.is_error():
_report_collector.push_back(error_report)
func orphan_monitor_start() -> void:
_orphan_monitor.start()
func orphan_monitor_stop() -> void:
_orphan_monitor.stop()
func add_report(report: GdUnitReport) -> GdUnitReport:
_report_collector.push_back(report)
return report
func reports() -> Array[GdUnitReport]:
return _report_collector.reports()
func collect_reports(recursive: bool) -> Array[GdUnitReport]:
if not recursive:
return reports()
# we combine the reports of test_before(), test_after() and test() to be reported by `fire_test_ended`
# we strictly need to copy the reports before adding sub context reports to avoid manipulation of the current context
var current_reports := reports().duplicate()
for sub_context in _sub_context:
current_reports.append_array(sub_context.collect_reports(true))
return current_reports
func calculate_statistics(reports_: Array[GdUnitReport]) -> Dictionary:
var failed_count := GdUnitTestReportCollector.count_failures(reports_)
var error_count := GdUnitTestReportCollector.count_errors(reports_)
var warn_count := GdUnitTestReportCollector.count_warnings(reports_)
var skip_count := GdUnitTestReportCollector.count_skipped(reports_)
var is_failed := !is_success()
var orphan_count := _count_orphans()
var elapsed_time := _timer.elapsed_since_ms()
var retries := 1 if _parent_context == null else _sub_context.size()
# Mark as flaky if it is successful, but errors were counted
var is_flaky := retries > 1 and not is_failed
# In the case of a flakiness test, we do not report an error counter, as an unreliable test is considered successful
# after a certain number of repetitions.
if is_flaky:
failed_count = 0
return {
GdUnitEvent.RETRY_COUNT: retries,
GdUnitEvent.ELAPSED_TIME: elapsed_time,
GdUnitEvent.FAILED: is_failed,
GdUnitEvent.ERRORS: error_count > 0,
GdUnitEvent.WARNINGS: warn_count > 0,
GdUnitEvent.FLAKY: is_flaky,
GdUnitEvent.SKIPPED: skip_count > 0,
GdUnitEvent.FAILED_COUNT: failed_count,
GdUnitEvent.ERROR_COUNT: error_count,
GdUnitEvent.SKIPPED_COUNT: skip_count,
GdUnitEvent.ORPHAN_NODES: orphan_count,
}
func is_success() -> bool:
if _sub_context.is_empty():
return not _report_collector.has_failures()
# we on test suite level?
if _parent_context == null:
return not _report_collector.has_failures()
return _sub_context[-1].is_success() and not _report_collector.has_failures()
func is_skipped() -> bool:
return (
_sub_context.any(func(c :GdUnitExecutionContext) -> bool:
return c.is_skipped())
or test_case.is_skipped() if test_case != null else false
)
func is_interupted() -> bool:
return false if test_case == null else test_case.is_interupted()
func _count_orphans() -> int:
if _orphans != -1:
return _orphans
var orphans := 0
for c in _sub_context:
if _orphan_monitor.orphan_nodes() != c._orphan_monitor.orphan_nodes():
orphans += c._count_orphans()
_orphans = _orphan_monitor.orphan_nodes()
if _orphan_monitor.orphan_nodes() != orphans:
_orphans -= orphans
return _orphans
func sum(accum: int, number: int) -> int:
return accum + number
func retry_execution() -> bool:
var retry := _test_execution_iteration < 1 if not _flaky_test_check else _test_execution_iteration < _flaky_test_retries
if retry:
_test_execution_iteration += 1
return retry
func register_auto_free(obj: Variant) -> Variant:
return _memory_observer.register_auto_free(obj)
## Runs the gdunit garbage collector to free registered object and handle orphan node reporting
func gc(gc_orphan_check: GC_ORPHANS_CHECK = GC_ORPHANS_CHECK.NONE) -> void:
# unreference last used assert form the test to prevent memory leaks
GdUnitThreadManager.get_current_context().clear_assert()
await _memory_observer.gc()
orphan_monitor_stop()
var orphans := _count_orphans()
match(gc_orphan_check):
GC_ORPHANS_CHECK.SUITE_HOOK_AFTER:
if orphans > 0:
reports().push_front(GdUnitReport.new() \
.create(GdUnitReport.WARN, 1, GdAssertMessages.orphan_detected_on_suite_setup(orphans)))
GC_ORPHANS_CHECK.TEST_HOOK_AFTER:
if orphans > 0:
reports().push_front(GdUnitReport.new()\
.create(GdUnitReport.WARN, 1, GdAssertMessages.orphan_detected_on_test_setup(orphans)))
GC_ORPHANS_CHECK.TEST_CASE:
if orphans > 0:
reports().push_front(GdUnitReport.new()\
.create(GdUnitReport.WARN, test_case.line_number(), GdAssertMessages.orphan_detected_on_test(orphans)))