• Upgraded for compatibility with SQLAlchemy 2.0. #46
• Now documented to work with MS SQL Server. Thanks, Ryan Cheley. #37

• New --postgres-schema option for specifying a different PostgreSQL schema, thanks Jeremy Dormitzer. #29
• Correctly handles table names that need escaping, such as user in PostgreSQL. #25
• Empty tables are now created in the SQLite database with a matching schema, rather than being skipped. #28
• Connection strings starting with postgres:// are treated as if they started with postgresql://. #34

• Added support for compound primary keys. #26

• Changelog badge and project_urls for the PyPI listing.

Bumped to sqlite-utils 2.9.1. This fixes two things:

• Decimal columns are now supported. They become FLOAT columns in SQLite.
• Fixed a bug where columns that were a combination of integers and nulls were incorrectly stored as TEXT, not INTEGER.

I made a mistake with the release of 1.1 to PyPI - this fixes it.

Fixed compatibility with sqlite-utils 2.x - #20

Fix for #18 - no longer throws error on empty tables

• Improvements to README

See the README for full usage instructions.

• Instead of using --connection the connection string is now a required positional argument, #14
• --sql must now be accompanied by --output specifying the table the query results should be written to
• --redact tablename columnname option can be used to redact values, #2
• Foreign keys are now created with indexes, use --no-index-fks to disable this, #12
• --table can now be used multiple times, #6
• README and --help now include example connection strings
• README also details how this can be used with Heroku Postgres