- Added support for joining tables in anonymizing queries.
- Added support for Windows.
- Added support for
date_truncandextract(... from ...).
- Added support for simple pre-anonymization
WHEREclauses. - Added
sumandavgaggregrators. - Added
*_noisevariants of aggregrators. - Added
count_histogramaggregrator. - Allow more metadata discovery queries.
- Various tweaks and fixes.
- Reject unsupported column types during AID labeling.
- Allow casts between
int4andint8. - Allow more metadata discovery queries.
- Allow more statement types.
- Fixed handling of
IN (subquery)expressions. - Added analyst guide section for allowed type conversions.
- Small fixes to docs.
- Added support for parameterized queries.
- Queries selecting or grouping by non-generalized AID columns are now rejected.
- Fixed some docs links.
- Fixed setup script version.
- Vastly expanded docs.
- Fixed a noise generation bug.
- Salt is randomly generated per-database during extension setup.
- Simplified and improved utility methods.
- The low count filter is not seeded with bucket data anymore.
- Extension now skips non-associated databases when loaded globally.
- Fixed handling of post-anonymization conditions.
- Minor updates to docs and config terminology.
- Tweaked README file for better rendering on PGXN.