test: add benchmarks, Unicode edge cases, and stress tests
Performance benchmarks across all major APIs:
- BenchmarkRender (depth 1/3/5/7 tree, full page)
- BenchmarkImprint (small/large pipeline)
- BenchmarkCompareVariants (2/3 variants)
- BenchmarkLayout (C, HCF, HLCRF, nested, many children)
- BenchmarkEach (10/100/1000 items)
- BenchmarkResponsive, BenchmarkStripTags
- Codegen: GenerateClass, TagToClassName, GenerateBundle, GenerateRegistration
Edge case tests:
- Unicode: emoji, RTL (Arabic/Hebrew), zero-width chars, mixed scripts
- Deep nesting: 10/20 levels, mixed slot types
- Large Each iterations: 1000/5000 items, nested Each
- Layout variant validation: invalid chars, lowercase, duplicates, empty
- Nil context handling for Render, Imprint, CompareVariants
- Switch no-match, Entitled nil context, empty tag El
Co-Authored-By: Virgil <virgil@lethean.io>