Improve SSE chunk buffering performance

[Kludex]

Summary

• Accumulate incomplete SSE lines as string fragments and join them when a line separator arrives.
• Track the pending UTF-8 byte count incrementally while preserving max_event_size behavior.
• Add coverage for events split across many chunks.

Motivation

Reduce repeated copying and rescanning when SSE event lines are delivered in highly fragmented streams. This keeps parsing performance predictable without changing the public API or event semantics.

Validation

• pytest tests/httpx2/test_sse.py -q
• pytest tests/test_benchmark.py::test_bench_sse_many_chunks_without_line_separator -q
• Ruff formatting and lint checks
• mypy

[github-actions]

Docs preview: https://3c394336-httpx2-docs.pydantic.workers.dev

[codspeed-hq]

Merging this PR will improve performance by ×35

⚠️ Different runtime environments detected

Some benchmarks with significant performance changes were compared across different runtime environments,
which may affect the accuracy of the results.

Open the report in CodSpeed to investigate

⚡ 1 improved benchmark
✅ 15 untouched benchmarks
⏩ 7 skipped benchmarks¹

Performance Changes

	Mode	Benchmark	BASE	HEAD	Efficiency
⚡	Simulation	test_bench_sse_many_chunks_without_line_separator	3,071.8 ms	86.9 ms	×35

Tip

Curious why this is faster? Comment @codspeedbot explain why this is faster on this PR, or directly use the CodSpeed MCP with your agent.

---

_{Comparing agent/improve-sse-chunk-buffering (614b433) with main (ad141d8)}

Footnotes

1. 7 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[cubic-dev-ai]

No issues found across 2 files

_{Re-trigger cubic}