#290 Improve generator tests and assertion rules

[sanelli]

Summary

• Require full HasNextSyntaxNode syntax walks after every HaveGeneratedSourceCode() (or assert stubs where generation still emits a NoMap return).
• Replace HaveOnlyWarnings with HaveDiagnostics + HaveDiagnostic(MappaDiagnosticDescriptors.…) and remove the assertion helpers.
• Tighten the cursor workflow Tests rules and bump MappaVersion to 10.2.0-alpha.29.

Closes #290.

Test plan

• dotnet test Mappa.Generator.Tests/Mappa.Generator.Tests.csproj (1877 passed)
• .\Scripts\RunTestsAndReportCoverage.ps1 (2652 passed; line 98.4% / branch 90.6% / method 100% ≥ baseline)

[github-actions]

Code coverage

Compared against the latest published coverage history from the code coverage gist.

Metric	Baseline	PR	Δ	Trend
Line	98.4%	98.4%	0	=
Branch	90.6%	90.6%	0	=
Method	100%	100%	0	=

No coverage metric dropped by more than 1 percentage point(s) versus baseline.

Benchmarks

AutoMapper is the BenchmarkDotNet ratio baseline. Absolute values are mean time in nanoseconds and allocated bytes. Ratio columns are competitor / Mappa as a percentage (n/a when either side is missing or Mappa is zero).

Benchmark	Automapper Mean (ns)	Automapper Allocated (B)	Mapster Mean (ns)	Mapster Allocated (B)	Mapperly Mean (ns)	Mapperly Allocated (B)	Mappa Mean (ns)	Mappa Allocated (B)	Automapper/Mappa Time	Automapper/Mappa Alloc	Mapperly/Mappa Time	Mapperly/Mappa Alloc	Mapster/Mappa Time	Mapster/Mappa Alloc
ArrayToListBenchmark	72.69	32	77987.9	194456	7578.28	38456	7673.31	38456	0.9%	0.1%	98.8%	100%	1016.4%	505.7%
DictionaryBenchmark	142820	185221.12	89140	214364.16	11490	46018.56	13730	58316.8	1040.2%	317.6%	83.7%	78.9%	649.2%	367.6%
FastListToArrayBenchmark	n/a	n/a	82525	194426.88	7563	38420.48	7433	38420.48	n/a	n/a	101.7%	100%	1110.3%	506.1%
IQueryableProjectionBenchmark	614900	31784.96	591100	31016.96	599800	33105.92	596600	32983.04	103.1%	96.4%	100.5%	100.4%	99.1%	94%
ListToArrayBenchmark	n/a	n/a	80664	194426.88	7503	38420.48	7223	38420.48	n/a	n/a	103.9%	100%	1116.8%	506.1%
ListToHashSetBenchmark	361.5	5765.12	2285.6	5765.12	2515.4	5806.08	2619.8	5765.12	13.8%	100%	96%	100.7%	87.2%	100%
NestedDtoBenchmark	128253	225320.96	94943	227471.36	7755	38942.72	7731	38942.72	1658.9%	578.6%	100.3%	100%	1228.1%	584.1%

Benchmark trends (Mappa)

Lower is better. Compared against the latest published Mappa-only rows from MAPPA-BENCHMARK-HISTORY.md.

Benchmark	Measure	Baseline	PR	Δ	Trend
ArrayToListBenchmark	TIME_NS	7458.92	7673.31	+214.39	▼
ArrayToListBenchmark	ALLOC_B	38456	38456	0	=
DictionaryBenchmark	TIME_NS	13290	13730	+440	▼
DictionaryBenchmark	ALLOC_B	58316.8	58316.8	0	=
ListToArrayBenchmark	TIME_NS	7267	7223	-44	▲
ListToArrayBenchmark	ALLOC_B	38420.48	38420.48	0	=
FastListToArrayBenchmark	TIME_NS	7179	7433	+254	▼
FastListToArrayBenchmark	ALLOC_B	38420.48	38420.48	0	=
IQueryableProjectionBenchmark	TIME_NS	598600	596600	-2000	▲
IQueryableProjectionBenchmark	ALLOC_B	32983.04	32983.04	0	=
NestedDtoBenchmark	TIME_NS	6996	7731	+735	▼
NestedDtoBenchmark	ALLOC_B	38942.72	38942.72	0	=
ListToHashSetBenchmark	TIME_NS	2509.3	2619.8	+110.5	▼
ListToHashSetBenchmark	ALLOC_B	5765.12	5765.12	0	=