feat(examples): add FastAPI asynchronous cleaning microservice recipe #514

Context

Many production systems consume data-cleaning libraries behind microservices or ETL worker endpoints. Adding a lightweight, standalone FastAPI integration recipe in examples/integrations/ helps backend engineers integrate FreshData into web APIs and production serving stacks.

Task

Create a standalone runnable FastAPI recipe in examples/integrations/fastapi_clean_service.py that implements:

1. A POST /clean endpoint accepting CSV file uploads (UploadFile) or JSON tabular records.
2. Optional query parameters / request body options (strategy, impute, drop_duplicates).
3. Clean execution using freshdata.clean() inside a thread pool (asyncio.to_thread or run_in_threadpool) so CPU-bound dataframe processing does not block the async event loop.
4. Returns the cleaned data along with the structured audit actions from CleanReport.to_dict().
5. Includes a runnable if __name__ == "__main__": uvicorn.run(...) block so contributors can test immediately.

Acceptance Criteria

• Create examples/integrations/fastapi_clean_service.py with standalone, clean, and well-typed code.
• Document required dependencies (pip install fastapi uvicorn "freshdata-cleaner[dev]") in the module docstring.
• Register the new recipe in examples/README.md.
• Script runs and passes lint checks (ruff check examples/integrations/fastapi_clean_service.py).

Pointers

• Existing integration recipes: examples/integrations/
• Recipe index: examples/README.md
• Core API reference: src/freshdata/api.py

New to the codebase? Comment /assign to claim this issue! You can also check our First Contribution Guide.