Improve performance

[samdark]

Q	A
Is bugfix?	❌
New feature?	❌
Breaks BC?	❌
Fixed issues	-

Benchmark results against master after removing optimizations that did not show a reliable gain.

The suite uses 15 iterations now. Very small subjects use higher rev counts, and the table includes relative standard deviation. Deltas inside the measured deviation should be treated as noise rather than a reliable gain or regression.

Benchmark	Master (μs, rstdev)	Branch (μs, rstdev)	Branch vs master
TypicalUsageBench::benchArrayDefinitionObjectGraph	6.596 (±1.16%)	6.524 (±1.64%)	-1.1%
TypicalUsageBench::benchArrayDefinitionMethodsAndProperties	11.661 (±0.97%)	11.620 (±1.36%)	-0.4%
TypicalUsageBench::benchStaticFactoryDefinition	2.920 (±1.32%)	2.879 (±1.46%)	-1.4%
TypicalUsageBench::benchServiceFactoryDefinition	3.816 (±1.31%)	3.766 (±1.38%)	-1.3%
TypicalUsageBench::benchReferenceResolution	0.289 (±1.53%)	0.297 (±1.20%)	+2.8%
TypicalUsageBench::benchMixedDefinitionResolution	504.110 (±0.74%)	504.206 (±0.87%)	+0.0%
DefinitionStorageBench::benchConstruct	0.232 (±1.32%)	0.235 (±1.61%)	+1.3%
DefinitionStorageBench::benchSequentialExplicitLookupsColdStorage	49.121 (±1.15%)	47.063 (±0.95%)	-4.2%
DefinitionStorageBench::benchSequentialExplicitLookupsWarmStorage	49.415 (±1.58%)	47.694 (±1.25%)	-3.5%
DefinitionStorageBench::benchResolvableObjectGraphColdStorage	1.927 (±1.15%)	1.926 (±0.78%)	-0.1%
DefinitionStorageBench::benchResolvableObjectGraphWarmStorage	0.299 (±1.22%)	0.302 (±1.34%)	+1.0%
DefinitionStorageBench::benchUnresolvableObjectGraphColdStorage	1.827 (±1.12%)	1.803 (±1.70%)	-1.3%
DefinitionStorageBench::benchUnresolvableObjectGraphWarmStorage	1.634 (±1.26%)	1.630 (±1.33%)	-0.2%
DefinitionStorageBench::benchResolvableObjectGraphWithDelegateColdStorage	3.226 (±0.94%)	3.277 (±1.26%)	+1.6%
DefinitionStorageBench::benchResolvableObjectGraphWithDelegateWarmStorage	0.297 (±1.22%)	0.299 (±1.38%)	+0.7%
DefinitionStorageBench::benchStrictModeClassMisses	60.744 (±0.89%)	60.526 (±1.29%)	-0.4%
NormalizerBench::benchWarmClassDefinitions	57.296 (±1.34%)	56.750 (±1.14%)	-1.0%
NormalizerBench::benchColdClassDefinitions	56.080 (±1.10%)	56.637 (±1.41%)	+1.0%
NormalizerBench::benchWarmReferenceDefinitions	57.574 (±1.39%)	57.056 (±1.25%)	-0.9%
NormalizerBench::benchColdReferenceDefinitions	57.555 (±1.43%)	57.409 (±1.09%)	-0.3%
NormalizerBench::benchArrayDefinitions	139.804 (±1.32%)	141.808 (±1.54%)	+1.4%
NormalizerBench::benchCallableDefinitions	42.009 (±1.31%)	42.499 (±0.75%)	+1.2%
NormalizerBench::benchWarmObjectDefinitions	49.256 (±1.45%)	49.038 (±0.92%)	-0.4%
NormalizerBench::benchColdObjectDefinitions	48.820 (±1.73%)	49.935 (±1.39%)	+2.3%
NormalizerBench::benchArrayDefinitionsWithInferredClass	153.715 (±1.24%)	153.567 (±1.42%)	-0.1%
NormalizerBench::benchMixedApplicationDefinitionsCold	94.108 (±1.21%)	94.532 (±1.29%)	+0.5%
NormalizerBench::benchMixedApplicationDefinitionsWarm	93.596 (±1.08%)	94.111 (±1.11%)	+0.6%

Retained changes:

• Optimize array definition config parsing and align it with validator rules.
• Add benchmark coverage for typical definition usage, definition storage checks, and normalizer paths.
• Reduce benchmark noise by using 15 iterations, higher rev counts for tiny subjects, and reporting relative standard deviation.

Not retained as optimizations:

• ArrayDefinition method dependency cache.
• DefinitionStorage::has() explicit-definition and strict-mode fast paths.
• CallableDefinition dependency cache and direct callable invocation.
• Normalizer caches.

Note: Yii DI services are shared, so definition resolution normally happens once per service per container lifetime. The benchmark data does not show a meaningful typical request-time performance gain from this PR.

[Copilot]

Pull request overview

This PR focuses on performance improvements in the definitions normalization/validation path by introducing caching and fast-path checks, and adds PhpBench benchmarks to measure typical usage scenarios.

Changes:

• Added multiple caches and fast-path branches in Normalizer, ArrayDefinition, CallableDefinition, and DefinitionStorage to reduce repeated parsing/reflection work.
• Expanded unit tests to cover the new caching/fast-path behaviors (normalization, validator callable handling, build stack clearing).
• Introduced a PhpBench benchmark suite and configuration, plus a dev dependency to run it.

Reviewed changes

Copilot reviewed 14 out of 14 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
src/Helpers/Normalizer.php	Adds static caches and fast-path logic for classes/references/callables/values during normalization.
src/ArrayDefinition.php	Caches parsed config keys and reuses prepared “class-only” definitions; avoids some getter calls.
src/CallableDefinition.php	Caches extracted callable dependencies and invokes via prepared closure.
src/DefinitionStorage.php	Adds fast-path for explicit definitions and strict-mode misses; reduces build stack resets.
src/Helpers/DefinitionValidator.php	Adds coverage-ignore markers for PHP 8.4+ asymmetric visibility modifiers branch.
tests/Unit/Helpers/NormalizerTest.php	Adds tests for caching behavior and autoload avoidance for plain references.
tests/Unit/Helpers/DefinitionValidatorTest.php	Adds a test ensuring callable array definitions validate successfully.
tests/Unit/DefinitionStorageTest.php	Adds a test ensuring explicit lookups clear previous build stack state.
tests/Unit/ArrayDefinitionTest.php	Adds coverage for prepared-data arguments and ignored config keys.
tests/Benchmark/TypicalUsageBench.php	Adds a benchmark for common resolution patterns (array defs, callable defs, references).
tests/Benchmark/NormalizerBench.php	Adds a benchmark for normalization hot paths across multiple input types.
tests/Benchmark/DefinitionStorageBench.php	Adds a benchmark for storage construction and lookup patterns (strict/non-strict).
phpbench.json	Adds PhpBench configuration (bootstrap, paths, CSV output).
composer.json	Adds phpbench/phpbench to require-dev and suggests it for running benchmarks.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 13 out of 13 changed files in this pull request and generated 1 comment.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 14 out of 14 changed files in this pull request and generated 2 comments.

[Copilot]

Pull request overview

Copilot reviewed 12 out of 12 changed files in this pull request and generated 1 comment.

[Tigrov]

There were more optimizations, were they ineffective?

[samdark]

Yes. I've updated the benchmark to match real usage, and the majority of optimizations revealed to be either not necessary or making things worse.