Add async SQLAlchemy dialect using native asyncio cursors

[laughingman7743]

Summary

PyAthena now has native asyncio cursor implementations (pyathena/aio/) added in PRs #666, #667, #668, but the SQLAlchemy dialects (pyathena/sqlalchemy/) remain synchronous only. Adding async SQLAlchemy dialects would allow PyAthena to be used with async ORMs and frameworks that rely on SQLAlchemy's async engine (e.g., FastAPI + SQLAlchemy async session).

Background

SQLAlchemy 2.0+ supports async dialects via AdaptedConnection and AsyncDialect / create_async_engine. The existing synchronous dialects are:

Dialect	Connection string	Cursor
AthenaRestDialect	awsathena+rest://	Cursor
AthenaPandasDialect	awsathena+pandas://	PandasCursor
AthenaArrowDialect	awsathena+arrow://	ArrowCursor
AthenaPolarsDialect	awsathena+polars://	PolarsCursor
AthenaS3FSDialect	awsathena+s3fs://	S3FSCursor

Each of these could have an async counterpart using the corresponding Aio*Cursor.

Scope

• Implement async SQLAlchemy dialect(s) using AioConnection and native asyncio cursors
• Connection string convention (e.g., awsathena+aiorest:// or similar)
• Register as SQLAlchemy entry points in pyproject.toml
• Add documentation to docs/sqlalchemy.md

Related

• Comprehensive cursor benchmark with memory and performance metrics #644 — Native asyncio implementation tracking issue
• Add native asyncio cursor support (AioCursor) #662 — Native asyncio implementation plan
• PRs Add native asyncio cursor support (Phase 1) #666, Add native asyncio specialized cursors and DRY refactors (Phase 2) #667, Add native asyncio S3FS and Spark cursors with boilerplate deduplication (Phase 3) #668 — Native asyncio cursor implementations
• PR Add documentation for native asyncio cursors #671 — Documentation for aio cursors