- MySQL: Add an optional `charset_collation` to the connector, as PyMySQL >=1.1.0 always runs a `SET NAMES` on connection, which breaks on servers using a non-default collation
Fix
- MySQL: Allow dict parameters to be used with PyMySQL 1.1.1 - MySQL: Use a regular PyMySQL Cursor rather than a DictCursor when pandas 2.x is used
6.1.2
Fixed
- Google Big Query: the query generated by `get_model` now correctly quotes the dataset name, which allows to build a DB models for datasets starting with a number
6.1.1
Fixed
- Elasticsearch: force `widget="json"` on `body` so the form is properly filled when updating a data source
6.1.0
Changed
- Added support for Python 3.12
Fixed
- Restored the `HubspotPrivateApp` connector, which was deleted by error in v6.0.0
6.0.0
Changed
- **Breaking**: Support for Python 3.10 has been dropped. - **Breaking**: The following connectors have been removed: * Wootric * Trello * Toucan Toco * Net Explorer * Linkedin Ads * Microstrategy * Hubspot * Google My Business * Google Adwords * Facebook Insights * Facebook Ads * Anaplan * Adobe Analytics
Fixed
- Google Big Query: do not exclude partitioning columns when listing table structure
5.3.0
Changed
- Mongo: maximal connection pool size is now configurable via the `max_pool_size` parameter. It defaults to 1