[vbreuss]

An injected property that needed a closing only an open generic registration provides was reported as AWT101, although the service is effectively registered: moving the same dependency into the constructor made it work. Expansion was seeded only from constructor and factory parameters, which pushed users to change their domain types instead of declaring the injection on the container (goal 7). Lifecycle hook parameters had the same gap.

Open generic expansion now also seeds from the injected members of every container-constructed implementation ([Inject] properties, [InjectProperty] entries and [InjectProperty(typeof(...))] entries, an open-generic target applied per closing) and from the parameters of the OnActivated/OnRelease hook a registration dispatches. Collection elements and relationship shapes are unwrapped like constructor parameters, a keyed member expands the keyed closing, and the existing constraint check (AWT126), depth and count limits (AWT129) and the rule that a dropped overridable default does not seed all apply unchanged. The [InjectProperty] entries are now read before registration collection so expansion can see them, and hook overload selection is shared with expansion so only the dispatched hook seeds.

BREAKING CHANGE: an Optional injected property whose service only an open generic registration can provide is now filled with that closing, as a constructor parameter would be, instead of being left at its default. A collection-typed injected property likewise now contains the closings of matching open registrations instead of being empty.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
100.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[github-actions]

Test Results

   27 files  ± 0     27 suites  ±0   1h 52m 46s ⏱️ + 56m 26s
1 704 tests +14  1 704 ✅ +14  0 💤 ±0  0 ❌ ±0 
8 255 runs  +70  8 255 ✅ +70  0 💤 ±0  0 ❌ ±0 

Results for commit 0e47de9. ± Comparison against base commit 2cd4104.

[github-actions]

🚀 Benchmark Results

Details

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)

AMD EPYC 7763 3.23GHz, 1 CPU, 4 logical and 2 physical cores

.NET SDK 10.0.303

[Host] : .NET 10.0.14 (10.0.14, 10.0.1426.52326), X64 RyuJIT x86-64-v3
Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15

LaunchCount=1  WarmupCount=10

Realistic	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
baseline*	251.7 ns	3.70 ns	3.46 ns	0.96	568 B	1.00
Awaiten	262.6 ns	3.30 ns	3.08 ns	1.00	568 B	1.00
MsDI	620.7 ns	15.49 ns	14.49 ns	2.36	1104 B	1.94
Autofac	6,836.7 ns	68.65 ns	57.33 ns	26.04	10784 B	18.99
Jab	180.2 ns	4.29 ns	4.02 ns	0.69	432 B	0.76
DryIoc	405.6 ns	9.44 ns	8.83 ns	1.54	944 B	1.66
SimpleInjector	831.6 ns	8.89 ns	8.32 ns	3.17	1096 B	1.93
PureDI	179.3 ns	4.83 ns	4.52 ns	0.68	632 B	1.11

Details

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)

AMD EPYC 7763 2.65GHz, 1 CPU, 4 logical and 2 physical cores

.NET SDK 10.0.303

[Host] : .NET 10.0.14 (10.0.14, 10.0.1426.52326), X64 RyuJIT x86-64-v3
Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15

LaunchCount=1  WarmupCount=10

Resolve	Size	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
baseline*	8	5.875 ns	0.1548 ns	0.1448 ns	0.82	-	NA
Awaiten	8	7.158 ns	0.0075 ns	0.0066 ns	1.00	-	NA
MsDI	8	7.791 ns	0.2418 ns	0.2262 ns	1.09	-	NA
Autofac	8	124.992 ns	0.9249 ns	0.8199 ns	17.46	656 B	NA
Jab	8	2.724 ns	0.0801 ns	0.0749 ns	0.38	-	NA
PureDI	8	5.427 ns	0.0056 ns	0.0049 ns	0.76	-	NA
DryIoc	8	9.068 ns	0.0100 ns	0.0089 ns	1.27	-	NA
SimpleInjector	8	11.067 ns	0.0071 ns	0.0055 ns	1.55	-	NA
baseline*	256	6.283 ns	0.0068 ns	0.0061 ns	0.77	-	NA
Awaiten	256	8.129 ns	0.0099 ns	0.0088 ns	1.00	-	NA
MsDI	256	7.745 ns	0.1907 ns	0.1691 ns	0.95	-	NA
Autofac	256	127.790 ns	4.4808 ns	4.1913 ns	15.72	656 B	NA
Jab	256	42.682 ns	0.0290 ns	0.0226 ns	5.25	-	NA
PureDI	256	7.334 ns	0.0057 ns	0.0050 ns	0.90	-	NA
DryIoc	256	9.088 ns	0.0883 ns	0.0826 ns	1.12	-	NA
SimpleInjector	256	14.452 ns	0.0230 ns	0.0215 ns	1.78	-	NA

Details

BenchmarkDotNet v0.15.8, Linux Ubuntu 26.04.1 LTS (Resolute Raccoon)

AMD EPYC 7763 2.63GHz, 1 CPU, 4 logical and 2 physical cores

.NET SDK 10.0.303

[Host] : .NET 10.0.14 (10.0.14, 10.0.1426.52326), X64 RyuJIT x86-64-v3
Job=InProcess  Toolchain=InProcessEmitToolchain  IterationCount=15

LaunchCount=1  WarmupCount=10

Build	Size	Mean	Error	StdDev	Ratio	Allocated	Alloc Ratio
baseline*	8	20.923 ns	0.8159 ns	0.7233 ns	0.83	208 B	1.00
Awaiten	8	25.08 ns	0.716 ns	0.598 ns	1.00	208 B	1.00
MsDI	8	1,654.52 ns	20.534 ns	18.203 ns	66.00	5688 B	27.35
Autofac	8	31,740.66 ns	126.971 ns	112.556 ns	1,266.19	32478 B	156.14
Jab	8	13.73 ns	0.702 ns	0.657 ns	0.55	96 B	0.46
PureDI	8	18.43 ns	0.307 ns	0.287 ns	0.74	128 B	0.62
DryIoc	8	733.78 ns	3.146 ns	2.627 ns	29.27	1528 B	7.35
SimpleInjector	8	13,480.87 ns	106.994 ns	94.847 ns	537.77	24760 B	119.04
baseline*	256	161.892 ns	9.0341 ns	8.4505 ns	0.71	4176 B	1.00
Awaiten	256	227.65 ns	9.938 ns	9.296 ns	1.00	4176 B	1.00
MsDI	256	16,117.61 ns	364.810 ns	341.243 ns	70.91	61016 B	14.61
Autofac	256	777,065.38 ns	6,735.426 ns	6,300.322 ns	3,418.74	736562 B	176.38
Jab	256	98.39 ns	5.950 ns	5.566 ns	0.43	2080 B	0.50
PureDI	256	114.59 ns	3.969 ns	3.713 ns	0.50	2112 B	0.51
DryIoc	256	46,593.47 ns	209.895 ns	196.336 ns	204.99	80734 B	19.33
SimpleInjector	256	408,809.52 ns	2,747.519 ns	2,435.604 ns	1,798.58	572981 B	137.21

baseline* rows show the corresponding Awaiten benchmark from the most recent successful main branch build with results, for regression comparison.

[github-actions]

This is addressed in release v0.8.0.