[SPARK-55361][DOCS][PYTHON] suggest the use of to_arrow_schema to avoid specifying schema twice

[casgie]

suggest the use of to_arrow_schema to avoid specifying schema twice

What changes were proposed in this pull request?

python/docs/source/tutorial/sql/python_data_source.rst gives an example for using PyArrow RecordBatch.

In the example, the Schema is specified twice, once for Spark

def schema(self):
   return "key int, value string" 

and then again for PyArrow

def read(self, partition):
   ...
   schema = pa.schema([("key", pa.int32()), ("value", pa.string())]) 

I am proposing to change the documentation to only specify a Spark schema and to use 
pyspark.sql.pandas.types.to_arrow_schema() to convert the Spark schema to arrow.

Why are the changes needed?

Using Python Data Source in production, having to specify the schema twice is a hassle and also a source for errors.

Does this PR introduce any user-facing change?

Yes, changes documentation.

How was this patch tested?

N/A

Was this patch authored or co-authored using generative AI tooling?

No

[github-actions]

JIRA Issue Information

=== Documentation SPARK-55361 ===
Summary: add to_arrow_schema python_data_source.rst to avoid double-specifying schema
Assignee: None
Status: Open
Affected: ["4.0.0","4.0.1","4.1.1"]

---

This comment was automatically generated by GitHub Actions