Python-oracledb 3.0 supports dataframes, AQ in thin mode, SPARSE vectors, configuration providers, and much more

[cjbj]

Python-oracledb 3.0 is available on PyPI.

My blog announcement is at python-oracledb 3.0 Fetches Directly to Data Frames.

The full release notes are here.

The release highlights are:

• Fetching as data frames usable directly in PyTorch, PyArrow, Pandas, Polars, NumPy etc
• Advanced Queueing support in python-oracledb Thin mode
• Support for Oracle Database 23.7 SPARSE vector data format
• Centralized Configuration Provider support for connection management
• Cloud Native Authentication support giving automatic token retrieval
• Plugins and hooks to extend python-oracledb capabilities
• Naming and caching of connection pools
• A new connection "Use SNI" flag to improve connection performance
• A setting to align python-oracledb Thin and Thick mode connection handling
• Transaction Guard support in python-oracledb Thin mode
• Pipelining is production