SQLAlchemy release notes for 2022

This article contains the release notes for the SQLAlchemy, including the following when applicable:

  • Behavior changes

  • New features

  • Customer-facing bug fixes

Snowflake uses semantic versioning for SQLAlchemy updates.

Version 1.4.5 (December 9, 2022)

New features

  • Updated the application name for the driver connection from SnowflakeConnection to SnowflakeSQLAlchemy.

Version 1.4.4 (November 16, 2022)

Bug Fixes

  • Fixed a bug where percent signs (%) in a non-compiled statement should not be interpolated with empty sequence when executed.

Version 1.4.3 (October 21, 2022)

Bug fixes

  • Fixed an issue whereSnowflakeDialect.normalize_name and SnowflakeDialect.denormalize_name could not handle empty strings.

  • Fixed a compatibility issue to vendor function sqlalchemy.engine.url._rfc_1738_quote as it is removed from SQLAlchemy v1.4.42.

Version 1.4.2 (September 28, 2022)

Updates

  • Improved reliability by always using context managers.

Version 1.4.1 (August 23, 2022)

Updates

  • None.

Bug Fixes

  • Fixed an issue where DATE was incorrectly removed from SnowflakeDialect.ischema_names.

  • Fixed breaking changes introduced in release 1.4.0 that:

    • Changed the behavior of processing numeric, datetime, and timestamp values returned from service.

    • Changed the sequence order of primary/foreign keys in list returned by inspect.get_foreign_keys and inspect.get_pk_constraint.

Version 1.4.0 (July 21, 2022)

New Features

  • Added support for regexp_match and regexp_replace in sqlalchemy.sql.expression.ColumnOperators.

  • Added support for Identity Column.

  • Added support for handling literal values for the sql types: Date, DateTime, Time, Float, and Numeric; also added support for converting the values into corresponding Python objects.

  • Added support for get_sequence_names in SnowflakeDialect.

Bug Fixes

  • Fixed a bug where insert with autoincrement failed due to incompatible column type affinity.

  • Fixed a bug when creating a column with sequence, default value was set incorrectly.

  • Fixed a bug that identifier having percents in a compiled statement was not interpolated.

  • Fixed a bug when visiting sequence value from another schema, the sequence name is not formatted with the schema name.

  • Fixed a bug where the sequence order of columns were not maintained when retrieving primary keys and foreign keys for a table.