FEAT: Optimize executemany() performance

[gargsaumya]

ADO Work Item Reference

AB#37935

---

Summary

This pull request introduces significant updates to the executemany method in the mssql_python library to enable column-wise parameter binding for batched SQL execution. It also includes enhancements to the underlying C++ bindings to support this functionality. The most important changes involve implementing column-wise binding logic, adding helper functions for parameter buffer allocation, and exposing a new SQLExecuteMany method for executing batched statements.

Enhancements to Python executemany method:

• mssql_python/cursor.py: Refactored the executemany method to use column-wise parameter binding for improved performance and scalability. Added _transpose_rowwise_to_columnwise helper function to convert row-wise parameters into column-wise format.

Updates to C++ bindings:

• mssql_python/pybind/ddbc_bindings.cpp: Added BindParameterArray function to handle column-wise parameter binding, supporting multiple data types such as SQL_C_LONG, SQL_C_DOUBLE, and SQL_C_WCHAR.
• mssql_python/pybind/ddbc_bindings.cpp: Introduced SQLExecuteMany_wrap function to enable batched execution of SQL statements with multiple parameter sets.
• mssql_python/pybind/ddbc_bindings.cpp: Exposed the new SQLExecuteMany method to Python via PYBIND11_MODULE.

Helper functions for buffer allocation:

• mssql_python/pybind/ddbc_bindings.cpp: Added AllocateParamBufferArray template function to allocate buffers for column-wise parameter binding.

[Copilot]

Pull Request Overview

This pull request introduces column-wise parameter binding optimization for the executemany() method in the mssql_python library to improve performance for batched SQL execution. The changes enable binding multiple parameter sets at once rather than executing each parameter set individually.

• Refactored Python executemany() method to use column-wise parameter binding instead of iterative execution
• Added C++ binding functions to support array parameter binding for various SQL data types
• Introduced helper functions for efficient parameter buffer allocation and memory management

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 7 comments.

File	Description
mssql_python/cursor.py	Completely rewrote executemany() to transpose row-wise parameters to column-wise format and use new batched execution method
mssql_python/pybind/ddbc_bindings.cpp	Added BindParameterArray(), SQLExecuteMany_wrap(), and AllocateParamBufferArray() functions to support column-wise parameter binding